Reply form for the ESMA MiFID II/MiFIR Discussion Paper

Transcription

1 Reply form for the ESMA MiFID II/MiFIR Discussion Paper 22 May 2014

2 Date: 22 May 2014

3 Responding to this paper The European Securities and Markets Authority (ESMA) invites responses to the specific questions listed in the ESMA MiFID II/MiFIR Discussion Paper, published on the ESMA website (here). Instructions Please note that, in order to facilitate the analysis of the large number of responses expected, you are requested to use this file to send your response to ESMA so as to allow us to process it properly. Therefore, please follow the instructions described below: i. use this form and send your responses in Word format; ii. do not remove the tags of type <ESMA_QUESTION_1> - i.e. the response to one question has to be framed by the 2 tags corresponding to the question; and iii. if you do not have a response to a question, do not delete it and leave the text TYPE YOUR TEXT HERE between the tags. Responses are most helpful: if they respond to the question stated; contain a clear rationale, including on any related costs and benefits; and describe any alternatives that ESMA should consider Given the breadth of issues covered, ESMA expects and encourages respondents to specially answer those questions relevant to their business, interest and experience. To help you navigate this document more easily, bookmarks are available in Navigation Pane for Word 2010 and in Document Map for Word Responses must reach us by 1 August All contributions should be submitted online at under the heading Your input/consultations. Publication of responses All contributions received will be published following the end of the consultation period, unless otherwise requested. Please clearly indicate by ticking the appropriate checkbox in the website submission form if you do not wish your contribution to be publicly disclosed. A standard confidentiality statement in an message will not be treated as a request for non-disclosure. Note also that a confidential response may be requested from us in accordance with ESMA s rules on access to documents. We may consult you if we receive such a request. Any decision we make is reviewable by ESMA s Board of Appeal and the European Ombudsman. Data protection Information on data protection can be found at under the heading Disclaimer. 3

4 1. Overview 2. Investor protection 2.1. Authorisation of investment firms Q1: Do you agree that the existing work/standards set out in points Error! Reference source not found. and Error! Reference source not found. Error! Reference source not found. provide a valid basis on which to develop implementing measures in respect of the authorisation of investment firms? <ESMA_QUESTION_1> No comment. Please note, we provide an introduction to LSEG in footnote 1 below 1. <ESMA_QUESTION_1> Q2: What areas of these existing standards do you consider require adjustment, and in what way should they be adjusted? <ESMA_QUESTION_2> TYPE YOUR TEXT HERE <ESMA_QUESTION_2> Q3: Do you consider that the list of information set out in point Error! Reference source not found. should be provided to Home State NCAs? If not, what other information should ES- MA consider? <ESMA_QUESTION_3> TYPE YOUR TEXT HERE <ESMA_QUESTION_3> Q4: Are there any other elements which may help to assess whether the main activities of an applicant investment firm is not in the territory where the application is made? 1 London Stock Exchange Group (LSEG) is a financial market infrastructure provider, headquartered in London, with significant operations in Europe, North America and Asia. LSEG welcomes the opportunity to respond to ESMA s Discussion Paper and Consultation Paper on MiFID II/MiFIR. LSEG operates a range of international equity, fixed income and derivatives markets, including: London Stock Exchange; Borsa Italiana; MTS Group, and Turquoise; post trade and risk management, including Cassa di Compensazione e Garanzia (CC&G), the EMIR authorised CCP, Monte Titoli, the Italian CSD and globesettle, a new CSD based in Luxembourg; and is majority owner of the multi-asset global CCP, LCH.Clearnet Group serving major international exchanges and platforms, as well as a range of OTC markets (with EMIR authorised CCPs in France and the UK). LSEG operates the EMIR authorised trade repository, UnaVista, and offers a range of real-time and reference data products, as well as access to international equity, bond and alternative asset class indices, through the leading index provider, FTSE International. In addition to the Group s markets, over 30 other organisations and exchanges around the world use the Group s MillenniumIT trading, surveillance and post trade technology. For further information contact: Steven Travers/ Shrey Kohli, Regulatory Strategy, LSEG (stravers@lseg.com/ skohli@lseg.com) or Fabrizio Plateroti/ Lucia Bordigato, Regulation and Post Trading, Borsa Italiana (fabrizio.plateroti@borsaitaliana.it/ lucia.bordigato@borsaitaliana.it) 4

5 <ESMA_QUESTION_4> TYPE YOUR TEXT HERE <ESMA_QUESTION_4> Q5: How much would one-off costs incurred during the authorisation process increase, compared to current practices, in order to meet the requirements suggested in this section? <ESMA_QUESTION_5> TYPE YOUR TEXT HERE <ESMA_QUESTION_5> Q6: Are there any particular items of information suggested above that would take significant time or cost to produce and if so, do you have alternative suggestions that would reduce the time/cost for firms yet provide the same assurance to NCAs? <ESMA_QUESTION_6> TYPE YOUR TEXT HERE <ESMA_QUESTION_6> 2.2. Freedom to provide investment services and activities / Establishment of a branch Q7: Do you agree that development of technical standards required under Articles 34 and 35 of MiFID II should be based on the existing standards and forms contained in the CESR Protocol on MiFID Notifications (CESR/07-317c)? If not, what are the specific areas in the existing CESR standards requiring review and adjustment? <ESMA_QUESTION_7> LSEG: Yes we agree. LSEG believes that the existing CESR Protocol on MiFID notifications has worked well and is a suitable basis for the development of technical standards. <ESMA_QUESTION_7> 2.3. Best execution - publication of data related to the quality of execution by trading venues for each financial instrument traded Q8: Do you agree data should be provided by all the execution venues as set out in footnote 24? If not, please state why not. <ESMA_QUESTION_8> LSEG: Yes we agree. LSEG believes that it is appropriate for all execution venues to provide this data on a periodic basis. It will enable users to get reliable information, especially for certain asset classes, where trading volumes are not concentrated in a limited number of venues. However, it should be noted that such metrics will only provide a very approximate and broad-brush indication of execution quality for non-equity products. For example, for non-equity products traded predominantly through RFQ systems, best execution is typically monitored by reference to competing prices from alternative dealers, rather than identification of best bid or best offer prices available in a central order book. <ESMA_QUESTION_8> 5

6 Q9: If you think that the different types of venues should not publish exactly the same data, please specify how the data should be adapted in each case, and the reasons for each adjustment. <ESMA_QUESTION_9> LSEG believes that it is reasonable to ask a venue to provide average price and spread statistics. However, it should be noted that cost, speed and price improvement all pose different complexities. Cost: differentiated pricing may disguise the actual cost for participants, which may differ distinctly from the average. Also, the cost at each execution venue is only a portion of the total cost of trading, and those other external costs of trading may differ from one venue to the next. Speed: a lower volume venue would typically have lower latency since it would experience much less congestion which detracts from a fair comparison. Likelihood of execution: there is no agreed universal definition, which may differ when considered across different venues, for example, a venue with a greater proportion of less-liquid securities will experience a longer average time to execution for resting orders compared to one that is concentrated more in liquid securities. <ESMA_QUESTION_9> Q10: Should the data publication obligation apply to every financial instrument traded on the execution venue? Alternatively, should there be a minimum threshold of activity and, if so, how should it be defined (for example, frequency of trades, number of trades, turnover etc.)? <ESMA_QUESTION_10> LSEG believes that a minimum threshold of activity would be a sensible solution. This should be set taking into account frequency of trades, number of trades, turnover - with all the minimum requirements to be met at the same time to qualify to provide meaningful statistics. For example, if a venue was observed to display a bid/offer spread of 5 basis points for ten minutes of the day and no best bid/offer spread for the remainder of the day, then it should not record an average spread since it would be unfair for that venue to have a better average spread than another venue which had 5 basis points spread for 60 minutes and then a spread of 10 basis points for the remaining 450 minutes of the day. <ESMA_QUESTION_10> Q11: How often should all execution data be published by trading venues? Is the minimum requirement specified in MiFID II sufficient, or should this frequency be increased? Is it reasonable or beneficial to require publication on a monthly basis and is it possible to reliably estimate the marginal cost of increased frequency? <ESMA_QUESTION_11> LSEG believes that there should be periodic reporting, which should be more frequent than on an annual basis, but this will depend on the execution venue and the type of financial instrument. However, it is difficult to estimate the marginal cost of doing so, as this will also depend on the execution venue and the type of financial instrument. <ESMA_QUESTION_11> Q12: Please provide an estimate of the cost of the necessary IT development for the production and the publication of such reporting. <ESMA_QUESTION_12> LSEG: It is not possible to provide a meaningful estimate until final details are known. However, for most regulated markets (RMs), this should not be a substantial cost, as they will already have functionality for trade reporting and subsequent publication. <ESMA_QUESTION_12> 6

7 Q13: Do you agree that trading venues should publish the data relating to the quality of execution with regard to a uniform reference period, with a minimum of specific reporting details and in a compatible format of data based on a homogeneous calculation method? If not, please state why. <ESMA_QUESTION_13> LSEG: Yes. LSEG believes that if any measures are to be introduced there should be as much harmony and consistency as possible. <ESMA_QUESTION_13> Q14: Is the volume of orders received and executed a good indicator for investment firms to compare execution venues? Would the VBBO in a single stock published at the same time also be a good indicator by facilitating the creation of a periodic European price benchmark? Are there other indicators to be considered? <ESMA_QUESTION_14> It is LSEG s view that the value and volume of orders executed on a venue may provide an indicator of the likely execution certainty to be expected if using the specific venue s order book. However, other factors such as spread and size as displayed also indicate the likely execution experience, and other attributes of varying importance to users related to the total cost and risk of trading on a specific venue. VBBO price and size published at the exact same timestamp for multiple venues could in theory be consolidated by a third party to create a multi-venue BBO measure. It is, however, prudent to note that additional sunk costs of trading, including trade fees and settlement fees, will not be reflected in this indicator and should be considered when performing TCA. <ESMA_QUESTION_14> Q15: The venue execution quality reporting obligation is intended to apply to all MiFID instruments. Is this feasible and what differences in approach will be required for different instrument types? <ESMA_QUESTION_15> LSEG believes that if execution quality measures are to be introduced, it would seem appropriate that this principle should apply to all MiFID instruments. However, in practice, different financial instruments should have different quality reporting obligations. For example, with options and futures, while spreads are a good measure for "Delta-1" product (futures, forwards), it is more complex to interpret and produce measures for Options. <ESMA_QUESTION_15> Q16: Do you consider that this requirement will generate any additional cost? If yes, could you specify in which areas and provide an estimation of these costs? <ESMA_QUESTION_16> LSEG: Yes, there will be costs involved in the specification, development, testing and implementation of the relevant programs. It is not possible to carry out a detailed assessment at this stage without further detail. <ESMA_QUESTION_16> Q17: If available liquidity and execution quality are a function of order size, is it appropriate to split trades into ranges so that they are comparable? How should they be defined (for example, as a percentage of the average trading size of the financial instrument on the execution venue; fixed ranges by volume or value; or in another manner)? <ESMA_QUESTION_17> LSEG: To the extent that execution quality varies significantly by order size, yes it would be appropriate to split into ranges. 7

8 In order to be comparable across all instruments on a platform, the trading venue should determine standard ranges across all financial instruments within a particular class of financial instruments. However, over-elaborating execution quality analysis into pre-defined size thresholds should be avoided, as this will only yield diminishing returns. <ESMA_QUESTION_17> Q18: Do you agree that a benchmark price is needed to evaluate execution quality? Would a depth-weighted benchmark that relates in size to the executed order be appropriate or, if not, could you provide alternative suggestions together with justification? <ESMA_QUESTION_18> LSEG: Yes, a benchmark price would be needed but it is unclear at this point how such metrics would be produced and what cost it would have. Whilst benchmark price offers value in certain circumstances (such as brokers measuring execution performance for clients by comparing average price achieved against market VWAP or arrival price as benchmarks), it may be misleading in other circumstances (such as individual execution price at a venue compared to a prevailing VBBO or EBBO since some participants may value immediacy of liquidity over execution price, by trading through multiple price points on a single venue rather than sweep the top of multiple venues with the risk of missing liquidity); these specific client needs may skew averages and make them misleading. <ESMA_QUESTION_18> Q19: What kind of cost should be reported (e.g. regulatory levies, taxes, mandatory clearing fees) and how should this data be presented to enable recipients to assess the total consideration of transactions? <ESMA_QUESTION_19> LSEG: Only costs that are incurred at the execution venue itself should be reported, in a clear and transparent manner. Reporting on indirect costs will place additional administrative overheads on the execution venue, with the risk that the information may not be current and adding costs for the end user. Other costs vary according to user, and their tax, clearing and regulatory arrangements. <ESMA_QUESTION_19> Q20: What would be the most appropriate way to measure the likelihood of execution in order to get useful data? Would it be a good indicator for likelihood of execution to measure the percentage of orders not executed at the end of the applicable trading period (for example the end of each trading day)? Should the modification of an order be taken into consideration? <ESMA_QUESTION_20> LSEG: An order-to-trade ratio is one measure of the likelihood of execution but by no means the only measure and given the different make-up of trading participants on any venue, an average time from order entry to execution could be misleading. For example, retail orders are likely to have a longer resting time than orders dynamically managed by an institutional broker. Therefore, any such statistics could be flawed. Order modification and the percentage of orders not executed at the end of the period could also be useful indicators. With percentage of orders, this should be weighted in order to take into consideration how far an order was from the best bid or offer when sent to the venue. <ESMA_QUESTION_20> Q21: What would be the most appropriate way to measure the speed of execution in order to get useful data? <ESMA_QUESTION_21> 8

9 LSEG: Speed of execution is only relevant for a subset of trading participants who invest in the fastest technology to optimise their own order processing, and, therefore, the majority of participants and investors would gain little or no value out of such regulatory reporting and furthermore, those latency sensitive customers would be expected to ask for latency analysis directly from the venue and test it themselves to validate. Also, an execution venue with lower order and execution volumes would typically have lower latency since it would experience much less congestion, which skews from a fair comparison <ESMA_QUESTION_21> Q22: Are there other criteria (qualitative or quantitative) that are particularly relevant (e.g. market structures providing for a guarantee of settlement of the trades vs OTC deals; robustness of the market infrastructure due to the existence of circuit breakers)? <ESMA_QUESTION_22> LSEG: No; all of these criteria are important in the context of the overall quality of a venue but are irrelevant in the context of latency. <ESMA_QUESTION_22> Q23: Is data on orders cancelled useful and if so, on what time basis should it be computed (e.g. within a single trading day)? <ESMA_QUESTION_23> LSEG: No. Deleted orders are not useful in performing TCA (especially if there is no reference to the overall number of orders submitted), as they could have been deleted for any number of reasons. <ESMA_QUESTION_23> Q24: Are there any adjustments that need to be made to the above execution quality metrics to accommodate different market microstructures? <ESMA_QUESTION_24> LSEG: The most appropriate indicators will need to be drawn up for each trading type (e.g. RFQ, executable prices, central limit order book, etc.). <ESMA_QUESTION_24> Q25: What additional measures are required to define or capture the above data and relevant additional information (e.g. depth weighted spreads, book depths, or others) How should the data be presented: on an average basis such as daily, weekly or monthly for each financial instrument (or on more than one basis)? Do you think that the metrics captured in the Annex to this chapter are relevant to European markets trading in the full range of MiFID instruments? What alternative could you propose? <ESMA_QUESTION_25> LSEG: All additional measures defined in Annex are valid metrics for European derivatives markets apart from: "Number of orders routed to another venue for execution" and any reference to BBO, since such concepts are not relevant to European markets due to the lack of fungibility across venues. <ESMA_QUESTION_25> Q26: Please provide an estimate of the costs of production and publication of all of the above data and, the IT developments required? How could these costs be minimised? <ESMA_QUESTION_26> LSEG: There will be development costs incurred. However, it is not possible to carry out a detailed assessment at this stage until the full details of the requirements are known. <ESMA_QUESTION_26> 9

10 Q27: Would increasing the frequency of venue execution quality data generate additional costs for you? Would these costs arise as a result of an increase of the frequency of the review, or because this review will require additional training for your staff in order to be able to analyse and take into account these data? Please provide an estimate of these costs. <ESMA_QUESTION_27> LSEG: Yes, increasing the frequency would generate additional costs. These costs could arise not only as a result of an increased frequency but also as a result of employing additional resources, training and IT changes. However, it is not possible to carry out a detailed assessment at this stage until the full details of the requirements are known. <ESMA_QUESTION_27> Q28: Do you agree that investment firms should take the publication of the data envisaged in this Discussion Paper into consideration, in order to determine whether they represent a material change? <ESMA_QUESTION_28> LSEG: Yes, although realistically these reports will only provide very broadly based indications of execution quality, as set out in our response to Question 8 above. Extensive refinement and further development would be necessary to form the basis of TCA, to which paragraph 29 of the Discussion Paper refers. <ESMA_QUESTION_28> 2.4. Best execution - publication of data by investment firms Q29: Do you agree that in order to allow clients to evaluate the quality of a firm s execution, any proposed standards should oblige the firm to give an appropriate picture of the venues and the different ways they execute an order? <ESMA_QUESTION_29> TYPE YOUR TEXT HERE <ESMA_QUESTION_29> Q30: Do you agree that when systematic internalisers, market makers, OTC negotiation or dealing on own account represent one of the five most important ways for the firm to execute clients orders, they should be incorporated in the reporting obligations under Article 27(6) of MiFID II? <ESMA_QUESTION_30> TYPE YOUR TEXT HERE <ESMA_QUESTION_30> Q31: Do you think that the data provided should be different in cases when the firm directly executes the orders to when the firm transmits the orders to a third-party for execution? If yes, please indicate what the differences should be, and explain why. <ESMA_QUESTION_31> TYPE YOUR TEXT HERE <ESMA_QUESTION_31> 10

11 Q32: Do you consider that information on both directed and non-directed orders is useful? Should the data be aggregated so that both types of order are shown together or separated? Should there be a similar approach to disclosure of information on market orders versus limit orders? Do you think that another categorisation of client orders could be useful? <ESMA_QUESTION_32> TYPE YOUR TEXT HERE <ESMA_QUESTION_32> Q33: Do you think that the reporting data should separate retail clients from other types of clients? Do you think that this data should be publicly disclosed or only provided to the NCA (e.g. when requested to assess whether there is unfair discrimination between retail clients and other categories)? Is there a more useful way to categorise clients for these purposes? <ESMA_QUESTION_33> TYPE YOUR TEXT HERE <ESMA_QUESTION_33> Q34: Do you agree that the investment firms should publish the data relating to their execution of orders with regard to a uniform reference period, with a minimum of specific reporting details and in a compatible format of data based on a homogeneous calculation method? If not, please state why. <ESMA_QUESTION_34> TYPE YOUR TEXT HERE <ESMA_QUESTION_34> Q35: What would be an acceptable delay for publication to provide the clients with useful data? <ESMA_QUESTION_35> TYPE YOUR TEXT HERE <ESMA_QUESTION_35> Q36: What format should the report take? Should there be any difference depending on the nature of the execution venues (MTF, OTF, Regulated Market, systematic internalisers, own account) and, if so, could you specify the precise data required for each type? <ESMA_QUESTION_36> TYPE YOUR TEXT HERE <ESMA_QUESTION_36> Q37: Do you agree that it is proportionate to require investment firms to publish on an annual basis a summary based on their internal execution quality monitoring of their top five execution venues in terms of trading volumes, subject to certain minimum standards? <ESMA_QUESTION_37> TYPE YOUR TEXT HERE <ESMA_QUESTION_37> Q38: Do you have views on how directed orders covered by client specific instructions should be captured in the information on execution quality? Is it possible to disaggregate reporting for directed orders from those for which there are no specific instructions and, if so, what the most relevant criteria would be for this exercise? <ESMA_QUESTION_38> TYPE YOUR TEXT HERE 11

12 <ESMA_QUESTION_38> Q39: Minimum standards to ensure that the summary of the firm s internal execution quality monitoring of their top five execution venues (in terms of trading volumes) is comprehensive and contains sufficient analysis or context to allow it to be understood by market participants shall include the factors set out at paragraph 29. Do you agree with this analysis or are there any other relevant factors that should be considered as minimum standards for reporting? <ESMA_QUESTION_39> TYPE YOUR TEXT HERE <ESMA_QUESTION_39> Q40: Can you recommend an alternative approach to the provision of information on execution quality obtained by investment firms, which is consistent with Article 27(6) of MiFID II and with ESMA s overall objective to ensure proportionate implementation? <ESMA_QUESTION_40> TYPE YOUR TEXT HERE <ESMA_QUESTION_40> Q41: Do you agree that ESMA should try to limit the number of definitions of classes of instruments and provide a classification that can be used for the different reports established by MiFID and MiFIR? <ESMA_QUESTION_41> TYPE YOUR TEXT HERE <ESMA_QUESTION_41> Q42: If this approach is not viable how should these classes be defined? What elements should be taken into consideration for that classification? Please explain the rationale of your classification. Is there a need to delay the publication of the reporting for particular class of financial instruments? If the schedule has to be defined, what timeframe would be the most relevant? <ESMA_QUESTION_42> TYPE YOUR TEXT HERE <ESMA_QUESTION_42> Q43: Is any additional data required (for instance, on number of trades or total value of orders routed)? <ESMA_QUESTION_43> TYPE YOUR TEXT HERE <ESMA_QUESTION_43> Q44: What information on conflicts of interest would be appropriate (inducements, capital links, payment for order flow, etc.)? <ESMA_QUESTION_44> TYPE YOUR TEXT HERE <ESMA_QUESTION_44> 12

13 3. Transparency 3.1. Pre-trade transparency - Equities Q45: What in your view would be the minimum content of information that would make an indication of interest actionable? Please provide arguments with your answer. <ESMA_QUESTION_45> Specific response to Q45 LSEG agrees with ESMA s analysis. The minimum content of information is the instrument details, direction (buy or sell), volume, limit price and time bounds. An IOI should be considered as actionable only when a market participant marks it as firm and executable on order entry. Introduction: LSEG s general comments on section 3.1 on pre-trade transparency for equities LSEG operates multiple trading venues in the equities in Europe, including the London Stock Exchange (a regulated market with over 2300 admitted shares and depositary receipts, 700 ETFs and 360 ETCs/ETNs), AIM (an MTF for SMEs looking to access public equity capital, with over 1100 admitted SMEs), Turquoise (a pan-european MTF), Borsa Italiana (a regulated market with over 280 admitted shares, 640 ETFs and 210 ETCs/ETNs) and AIM Italia MAC (an MTF for SMEs in Italy with 46 admitted companies). LSEG s markets cover shares and equity-like instruments across the liquidity spectrum. They operate through different market models to cater to the specific characteristics of the product and its liquidity including continuous order-driven, quote-driven and other models. We use our experience of operating these diverse markets in answering the questions below. In summary, LSEG s views on the topics in section 3.1 of the Discussion Paper are: 1. Large-in-scale: 1.1. For shares and DRs, LSEG does not object to the introduction of additional ADT bands to the LIS table for additional granularity in the regime. On the thresholds, we understand ESMA s position and suggestion, although we note that we have previously argued for a reduction in LIS thresholds For ETFs, Scenario A (10% of turnover being above Large-in-Scale) is appropriate. An added consideration for ETFs should be that they are usually subject to short-term periods of high volumes and trading, and may trade less than once a day (this is particularly the case for ETFs based on fixed income underlyings). This does not mean that these are less liquid or trade in small sizes, but can it artificially lower their ADT. So higher thresholds than those ESMA proposes may be required, in particular for fixed income ETFs For stubs, we support an approach which offers flexibility and incentivises investors to execute in large sizes on central limit order books. Our preference is for stubs to remain protected from the LIS waiver. 2. Reference price waiver: the additional option for the reference price should be the most relevant liquid market by turnover on an instrument basis, but this turnover should only contain pre-trade transparent trading. It should not include trading under the current RPW, NTW waiver or OTC trading reported to a venue, as this is would make the definition circular and inconsistent. 3. Double volume cap (DVC): although ESMA does not ask particular questions on this section, we note that: 13

14 3.1. ESMA should be the authority that calculates the amount of the relevant trading under the waivers to trigger application of the caps based on aggregated data from RMs and MTFs, because a harmonised CTP is not likely to be in place by 2016/17 (when the DVC calculation regime starts) Trading venues, SIs and investment firms must adopt harmonised standards for trade reporting data for the calculation mechanism to work we suggest the Market Model Typology (MMT) should be required to be used by all participants We suggest that additional ESMA guidelines will be needed by markets on how the suspension mechanism operates once the cap is breached either at venue (4%) or EU (8%) level, to ensure that the restriction should be implemented in an orderly and fair manner to minimise disruption for the market and investors. <ESMA_QUESTION_45> Q46: Do you agree with ESMA s opinion that Table 1 of Annex II of Regulation 1287/2006 is still valid for shares traded on regulated markets and MTFs? Please provide reasons for your answer. <ESMA_QUESTION_46> Yes, LSEG agrees. <ESMA_QUESTION_46> Q47: Do you agree with ESMA s view that Table 1 of Annex II of Regulation 1287/2006 is appropriate for equity-like instruments traded on regulated markets and MTFs? Are there other trading systems ESMA should take into account for these instruments? Please provide reasons for your answer. <ESMA_QUESTION_47> Yes, LSEG agrees. In addition, LSEG also suggests that ESMA clarifies that Request-For-Quote (RFQ) mechanisms may be offered by trading venues in equity-like instruments such as ETFs (similar to what has been proposed for non-equities), and that these systems are covered in the fourth row ( trading systems not covered in the first three rows ) of the transparency table. This may be analogous to what has been proposed for nonequity instruments in section 3.7 Pre-trade transparency requirements for non-equity instruments (paragraph 16, p.151). <ESMA_QUESTION_47> Q48: Do you agree with ESMA s view that ADT remains a valid measure for determining when an order is large in scale compared to normal market size? If not, what other measure would you suggest as a substitute or complement to the ADT? Please provide reasons for your answer. <ESMA_QUESTION_48> Yes, LSEG agrees. The ADT is a simple measure, well understood by the market. It continues to be a valid factor for determining the large-in-scale (LIS) threshold, so long as it has a positive and strong correlation with the observed normal/ average market size. <ESMA_QUESTION_48> Q49: Do you agree that ADT should be used as an indicator also for the MiFIR equity-like products (depositary receipts, ETFs and certificates)? Please provide reasons for your answers. <ESMA_QUESTION_49> In principle, LSEG agrees. The ADT is a valid measure to determine if an order is large in scale compared to normal market size for depositary receipts and ETFs. For ETFs in particular, LSEG notes that some instruments are generally traded in large quantities, but are not traded as frequently as shares (e.g. some may not be traded daily), which may result in a low ADT. 14

15 The liquidity of an ETF is related to the underlying basket of securities, and market makers quotes are usually in line with the underlying liquidity. The fact that trades may not be as frequent, but often execute in large size, is evidence that market makers are continuously present on the market and are willing to trade large sizes to satisfy the liquidity needs of the market. ESMA should be mindful of such specific characteristics when determining the LIS for ETFs. Please note there are no certificates on LSEG s markets. <ESMA_QUESTION_49> Q50: Do you think there is merit in creating a new ADT class of 0 to 100,ooo with an adequate new large in scale threshold and a new ADT class of 100,000 to 500,000? At what level should the thresholds be set? Please provide reasons for your answer. <ESMA_QUESTION_50> LSEG does not object to the introduction of additional ADT bands to the LIS table for additional granularity in the regime. On the thresholds, we understand ESMA s position and suggestion, although we note that we have previously argued for a reduction in LIS thresholds. <ESMA_QUESTION_50> Q51: Do you think there is merit in creating new ADT classes of 1 to 5m and 5 to 25m? At what level should the thresholds be set? Please provide reasons for your answer. <ESMA_QUESTION_51> As noted in our answer to Q50, LSEG does not object to the introduction of additional ADT bands to the LIS table for additional granularity in the regime. On the thresholds, we understand ESMA s position and suggestion, although we note that we have previously argued for a reduction in LIS thresholds. <ESMA_QUESTION_51> Q52: Do you think there is merit in creating a new ADT class for super-liquid shares with an ADT in excess of 100m and a new class of 50m to 100m? At what level should the thresholds be set? <ESMA_QUESTION_52> LSEG does not object to the introduction of additional ADT bands to the LIS table for additional granularity in the regime. On the thresholds, we understand ESMA s position and suggestion, although we note that we have previously argued for a reduction in LIS thresholds. <ESMA_QUESTION_52> Q53: What comments do you have in respect of the new large in scale transparency thresholds for shares proposed by ESMA? <ESMA_QUESTION_53> Please see our response to Q LSEG does not object to the introduction of additional ADT bands to the LIS table for additional granularity in the regime. On the thresholds, we understand ESMA s position and suggestion, although we note that we have previously argued for a reduction in LIS thresholds. <ESMA_QUESTION_53> Q54: Do you agree with the ADT ranges selected? Do you agree with the large in scale thresholds set for each ADT class? Which is your preferred option? Would you calibrate the ADT classes and related large in scale thresholds differently? Please provide reasons for your answers, including describing your own role in the market (e.g. market-maker, issuer etc). <ESMA_QUESTION_54> LSEG presents its views as an operator of Regulated Markets for listed ETFs in the UK and Italy. 15

16 For equity ETFs, LSEG s preferred approach is Scenario A where 10% of the turnover is above the LIS threshold, with 5 ADT classes. Such approach is very similar to the current approach in place for in our markets, where the LIS threshold is expressed as a multiplier of a specialist s (i.e. market maker s) quoting obligations. We note here that operationally, sometimes the same ETF is admitted to trading on different venues in the EU, but with different ISINs; this may need to be factored in to the methodology. That said, for Exchange Traded Funds/Products in general (and in particular for ETFs based on fixed income underlyings), the LIS thresholds proposed by ESMA may be too low. In our experience, such instruments are generally traded in large quantities, but are not traded as frequently as shares (some may not be traded daily). This may result in an artificially low ADT, but does not necessarily mean that they are not liquid or trade only in small sizes. We suggest that ESMA should be mindful of this when determining large in scale, and may suggest higher thresholds than those proposed in the Discussion Paper. <ESMA_QUESTION_54> Q55: Which is your preferred scenario? Would you calibrate the ADT classes differently? Please provide reasons for your answers. <ESMA_QUESTION_55> No comment. Please note there are no certificates on LSEG s markets. <ESMA_QUESTION_55> Q56: Do you agree that the same ADT classes should be used for both pre-trade and posttrade transparency? Please provide reasons for your answers. <ESMA_QUESTION_56> No comment. Please note there are no certificates on LSEG s markets. <ESMA_QUESTION_56> Q57: How would you calibrate the large in scale thresholds for each ADT class for pre- and post-trade transparency? Please provide reasons for your answers. <ESMA_QUESTION_57> No comment. Please note there are no certificates on LSEG s markets. <ESMA_QUESTION_57> Q58: Do you agree with ESMA s view that the large in scale thresholds (i.e. the minimum size of orders qualifying as large in scale and the ADT classes) should be subject to a review no earlier than two years after MiFIR and Level 2 apply in practice? <ESMA_QUESTION_58> Yes, LSEG agrees with ESMA s view, given that such thresholds are described in Level 2 implementing measures. However, this should not prevent ESMA from reviewing the thresholds and methodology earlier for equitylike products, including DRs and ETFs if, once implemented, it feels that the LIS thresholds are clearly calibrated incorrectly, for e.g. if LIS thresholds manifestly affects liquidity on transparent markets. <ESMA_QUESTION_58> Q59: How frequently do you think the calculation per financial instrument should be performed to determine within which large in scale class it falls? Which combination of frequency and period would you recommend? <ESMA_QUESTION_59> 16

17 LSEG: For shares and depositary receipts, an annual calculation is suitable using data from the previous 12 months, in line with the current provision for shares. This is consistent with the observed liquidity profile for shares and DRs, would ensure continuity and cause minimal administrative change. In our experience, ETFs are more sensitive to short-term changes in liquidity. So a more frequent calculation may be appropriate. LSEG suggests a semi-annual calculation, based on data from a previous 6-12 month period. <ESMA_QUESTION_59> Q60: Do you agree with ESMA s opinion that stubs should become transparent once they are a certain percentage below the large in scale thresholds? If yes, at what percentage would you set the transparency threshold for large in scale stubs? Please provide reasons to support your answer. <ESMA_QUESTION_60> LSEG: Practices dealing with large in scale stubs vary for different LIS orders. Some markets offer LIS trade reporting facilities for executions that are already bilaterally negotiated and agreed. These trades are fully executed in advance, so stubs are not relevant. Stubs are relevant when markets allow LIS orders to execute (as a hidden order) in the order book. For such orders, LSEG supports an approach which offers flexibility and incentivises investors to execute in large sizes on central limit order books. ESMA s suggested compromise approach in paragraph 56 (transparency below 75% of LIS) may unintentionally prevent this outcome. LSEG s experience suggests that investors prefer high fill rates for their orders. Stubs remaining protected from the LIS waiver helps deliver that these can achieve multiple fills and execute completely over time. Further, we are also mindful that 75% of the LIS threshold is substantially larger than average order and trade sizes on transparent markets. For example, average trade size in 2013 for all shares and depositary receipts in the FTSE UK index series with ADT greater than 100 million on the LSE was 14,720, i.e. 2% of the proposed LIS threshold for such shares is 650,000. If an order is exposed at 75% of LIS (i.e. 487,500), there is a substantial risk of market impact. In such situations, we feel that investors will withdraw their orders when the stub becomes transparent, or use other orders to execute large transactions (e.g. iceberg orders). Thus LSEG s preference would be for stubs to remain protected from the LIS waiver; or subject to transparency only when they reach the average market size or Standard Market Size (SMS) for that particular instrument. <ESMA_QUESTION_60> Q61: Do you agree with ESMA s view that the most relevant market in terms of liquidity should be the trading venue with the highest turnover in the relevant financial instrument? Do you agree with an annual review of the most relevant market in terms of liquidity? Please give reasons for your answer. <ESMA_QUESTION_61> Yes, LSEG agrees with ESMA s views on both questions, and suggest some additional considerations. In the context of the use of the reference price waiver from a market other than where an instrument was first admitted, LSEG agrees that the most relevant market in terms of liquidity should be based on the trading venue with the highest turnover, on an instrument-by-instrument basis. In some cases it may occur that the most relevant market is the market where the instrument was first admitted. In this case, there should only be one reference price. 17

18 This turnover should be calculated using only pre-trade transparent trading, e.g. not including any trading executed through the negotiated trade waiver, the reference price waiver or OTC trading reported to the venue. This would be a more representative measure of the liquidity accessible to all participants on that market, and an accurate and independent indicator of the price itself. It would also prevent the definition from becoming circular and inconsistent. We suggest the use of the same metric for turnover in our response to Q48 (use of ADT for calculation of LIS sizes) and Q109 of the Consultation Paper (definition of liquid market). An annual review using previous 12 months data is appropriate for shares and depositary receipts; a shorter period risks being affected by price swings or liquidity incentives which may not be lasting. This would also be consistent with the ADT calculations and cause minimal administrative change. For ETFs, a more frequent calculation may be appropriate. We suggest a semi-annual calculation, based on data from a previous 6-12 month period. This is because ETFs are more sensitive to shortterm changes in liquidity, and a semi-annual calculation would take this into account. <ESMA_QUESTION_61> Q62: Do you agree with ESMA s view on the different ways the member or participant of a trading venue can execute a negotiated trade? Please give reasons for your answer. <ESMA_QUESTION_62> Yes, LSEG agrees with ESMA. We suggest 74(ii) could be revised as follows: dealing with other members or participants, where both all are executing orders on own account, to ensure that there could be more than two parties all acting on own account to cover riskless principal trades. <ESMA_QUESTION_62> Q63: Do you agree that the proposed list of transactions are subject to conditions other than the current market price and do not contribute to the price formation process? Do you think that there are other transactions which are subject to conditions other than the current market price that should be added to the list? Please provide reasons for your answer. <ESMA_QUESTION_63> Yes, LSEG agrees with ESMA s proposed list, and notes some further considerations: In paragraph 75, ESMA implies that transactions that are not subject to the current market price are synonymous with those that do not contribute to the price formation process. This is not always the case, for example, when a market participant requests for extended settlement, the trade is subject to conditions other than the market price of the share, but is still price forming. We suggest that such extended settlement trades should be added to the list of transactions other than current market price for the use of the negotiated trade waiver. Equity swaps and forward transactions, under certain conditions, may also be added to the list. Often participants agree the number of shares in each which may not always exactly match current trading price in each (due to a previous corporate action, for example). <ESMA_QUESTION_63> Q64: Do you agree that these are the two main groups of order management facilities ESMA should focus on or are there others? <ESMA_QUESTION_64> Yes, LSEG agrees with ESMA on the two main groups of order management facilities ( stop and iceberg orders). Double volume cap mechanism: 18

19 Although ESMA does not ask any specific questions regarding its analysis in the previous subsection (double volume cap mechanism), we would like to make some general comments, as below (please also see our general comments to section 3.1 as noted in Q45): ESMA should be the authority that calculates the amount of the relevant trading under the waivers to trigger application of the caps based on aggregated data from RMs and MTFs, because a harmonised CTP is not likely to be in place by 2016/17 (when the DVC calculation regime starts). Trading venues, SIs and investment firms must adopt harmonised standards for trade reporting data for the calculation mechanism to work we suggest the Market Model Typology (MMT) should be required used by all participants. We suggest that additional ESMA guidelines will be needed by markets on how the suspension mechanism operates once the cap is breached either at venue (4%) or EU (8%) level, to ensure that the restriction should be implemented in an orderly and fair manner to minimise disruption for the market and investors. <ESMA_QUESTION_64> Q65: Do you agree with ESMA s general assessment on how to design future implementing measures for the order management facility waiver? Please provide reasons for your answer. <ESMA_QUESTION_65> Yes, LSEG agrees with ESMA. Such an approach is consistent with the current approach and will ensure that innovation is not limited. <ESMA_QUESTION_65> Q66: Are there other factors that need to be taken into consideration for equity-like instruments? Please provide reasons for your answer. <ESMA_QUESTION_66> No, LSEG is not aware of any other factors. DRs and ETFs should be treated the same way as shares. <ESMA_QUESTION_66> Q67: Do you agree that the minimum size for a stop order should be set at the minimum tradable quantity of shares in the relevant trading venue? Please provide reasons for your answer. <ESMA_QUESTION_67> Yes, LSEG agrees with ESMA s analysis and approach. Stop orders, once triggered, should be treated in the same way as other orders in the order book. <ESMA_QUESTION_67> Q68: Are there additional factors that need to be taken into consideration for equity-like instruments? <ESMA_QUESTION_68> LSEG: No additional comments. <ESMA_QUESTION_68> Q69: Which minimum overall sizes for iceberg orders are currently employed in the markets you use and how are those minimum sizes determined? <ESMA_QUESTION_69> Consistent with ESMA s analysis in paragraph 98 and 99, LSEG s experience indicates that different markets implement minimum sizes for icebergs differently: 19

20 LSE s markets across shares, DRs and ETFs prescribe a minimum iceberg peak size of 40% of the Exchange Market Size (EMS) 2. Borsa Italiana prescribes a minimum iceberg size of 40% of EMS for shares and 10% for ETFs. It follows that the minimum size of an iceberg is equal to the minimum peak size plus one share to be considered an iceberg order, as below this quantity an order will simply be a transparent limit order. <ESMA_QUESTION_69> Q70: Which minimum sizes and which methods for determining them should be prescribed via implementing measures? To what level of detail should such an implementing measure go and what should be left to the discretion of the individual market to attain an appropriate level of harmonisation? <ESMA_QUESTION_70> Consistent with ESMA s analysis in paragraph 98 and 99, LSEG suggests that implementing measures should describe principles for iceberg orders and not prescribe minimum sizes or harmonised methodologies. Trading venues should be given the flexibility to design and calibrate their iceberg orders according to the type of trading activity and participation in the instrument. This would ensure that innovation is not limited. <ESMA_QUESTION_70> Q71: Which methods for determining the individual peak sizes of iceberg orders are currently employed in European markets? <ESMA_QUESTION_71> LSEG: Please refer to our response to Q69. LSE s markets across shares, DRs and ETFs prescribe a minimum iceberg peak size of 40% of the Exchange Market Size (EMS) 3. Borsa Italiana prescribes a minimum iceberg size of 40% of EMS for shares and 10% for ETFs. <ESMA_QUESTION_71> Q72: Which methods for determining peaks should be prescribed by implementing measures, for example, should these be purely abstract criteria or a measure expressed in percentages against the overall size of the iceberg order? To what level of details should such an implementing measure go and what should be left to the discretion of the individual market to attain an appropriate level of harmonisation? <ESMA_QUESTION_72> Please refer to LSEG s response to Q70. Given that an iceberg peak is a transparent order, LSEG does not feel there is any regulatory purpose that would be achieved in defining minimum sizes. This approach is consistent with ESMA s analysis in paragraph 98 and 99, and we suggest that trading venues should be given the flexibility to design and calibrate their iceberg orders according to the type of trading activity and participation in the instrument. <ESMA_QUESTION_72> Q73: Are there additional factors that need to be taken into consideration for equity-like instruments? <ESMA_QUESTION_73> No, LSEG is not aware of any other factors. DRs and ETFs should be treated the same way as shares. 2 The Exchange Market Size (EMS) is a quantification of the minimum sizes that a registered market maker on LSE and Borsa Italiana is obliged to quote, and is linked to the ADT and average trade size on markets. In value this is numerically similar to the Standard Market Size (SMS) thresholds for Systematic Internalisers. For example, on the LSE, it ranges between 250 and 25,000 and is calculated as a percentage of the ADT (e.g. 1% or 2% for different market segments). 3 Ibid 20

21 <ESMA_QUESTION_73> 3.2. Post-trade transparency - Equities Q74: Do you agree that the content of the information currently required under existing MiFID is still valid for shares and applicable to equity-like instruments? Please provide reasons for your answer. <ESMA_QUESTION_74> Specific response to Q74 Yes, LSEG agrees with ESMA. The content under existing MiFID has become the industry standard and is appropriate. -- Introduction: LSEG s general comments to section 3.2 on post-trade transparency for equities LSEG operates multiple trading venues in equities in Europe, including the London Stock Exchange (a regulated market with over 2300 admitted shares and depositary receipts, 700 ETFs and 360 ETCs/ETNs), AIM (an MTF for SMEs looking to access public equity capital, with over 1100 admitted SMEs), Turquoise (a pan-european MTF), Borsa Italiana (a regulated market with over 280 admitted shares, 640 ETFs and 210 ETCs/ETNs) and AIM Italia MAC (an MTF for SMEs in Italy with 46 admitted companies). LSEG s markets cover shares and equity-like instruments across the liquidity spectrum. They operate through different market models to cater to the specific characteristics of asset and liquidity class including continuous order-driven, quote-driven and other models. We use our experience of operating these diverse markets in answering the questions below. In summary, LSEG s views on the topics in section 3.2 are as follows: 1. Trade reporting: 1.1. We support harmonisation across all participants, including trading venues and OTC, buy and sell sides and information vendors, and encourage the adoption of the Market Model Typology (MMT) Accurate trade reporting depends on how straightforward it is for a customer to enter an order or a trade report. Additional flags on the report should be justified, rather than complicate this further: for example, a single flag for negotiated trades (NT), trades outside current market price (SP) and deferred publication request (K) may be needed for entry of reports rather than proposing combinations of these flags with other waivers or with static instrument reference data or with data maintained on the ESMA database (e.g. liquid/ less-liquid or large-inscale). The outcome from a regulatory perspective will remain the same, and this would reduce the chances of rejects if participants use an incorrect trade type on entry Instead of focussing on reducing the maximum time for trade reporting (relevant for off-book and OTC transactions), ESMA and NCAs should focus on the effective supervision and enforcement of these limits. For portfolio transactions, 3 minutes may be needed. 2. Deferred Publication: 2.1. A modified Option B is the most appropriate approach for liquid instruments and would present enough time for intermediaries to offset any large positions taken on towards the end of the trading day For less liquid instruments, a 3 day maximum delay is merited: Access to non-bank capital, including public equity markets, is an important issue to keep in mind when assessing the proportionality for less liquid instruments. In our experience, activity in less liquid securities, including SMEs, is more reliant on registered market makers to provide liquidity, who need more time to offset their inventory when they execute in large sizes with institutional clients, compared to the normal market size. To prevent a potential increase in the cost of capital and to ensure that 21

22 issuers and investors are not adversely impacted, we suggest that instruments not having a liquid market should be allowed a maximum permissible delay in publication to the end of the 3rd trading day. A single table based on the thresholds in the current regime for ADT < 1,000,000 (i.e. column 3 in table 8 on p.84 of the Discussion Paper) would be appropriate. The second delay period could be reduced to 120 minutes (currently at 180 minutes). <ESMA_QUESTION_74> Q75: Do you think that any new field(s) should be considered? If yes, which other information should be disclosed? <ESMA_QUESTION_75> LSEG: No, the identified fields are sufficient <ESMA_QUESTION_75> Q76: Do you think that the current post-trade regime should be retained or that the identity of the systematic internaliser is relevant information which should be published? Please provide reasons for your response, distinguishing between liquid shares and illiquid shares. <ESMA_QUESTION_76> In general, LSEG supports greater post-trade transparency. It is also our view that there should be greater parity between the activity of participants trading on trading venues, and those investment firms trading as SIs. In this regard, one way of judging whether the identity of the SI should be revealed on the trade report (or not), is to consider it whether its activity is the same as that of a market making service for a set of clients (if it takes risk to provide two way liquidity), or whether its activity resembles that of a trading venue (if it facilitates trades between two different clients without taking any risk). LSEG does not agree that the identity of an SI should be revealed when an SI uses its own capital and exposes itself to risk to provide two-way liquidity and make markets (consistent with the treatment of the activity of market makers on RMs/MTFs, whose identity is not revealed on a post-trade report). This is to prevent the individual (and aggregate) risk positions taken by an SI being revealed to third parties, and protect the SI from moves against its (revealed) positions. However, LSEG agrees that the identity of an SI should be revealed if, in the MiFID II regime, an SI is allowed to execute trades without taking on risk, i.e. as riskless principal or through back-toback trades/riskless principal/agency cross trades (consistent with the treatment of trading venues, which are always identified on the post-trade report), since in this case an SI is matching trades rather than taking on risk and making markets. Our comments apply equally to all financial instruments, regardless of liquidity <ESMA_QUESTION_76> Q77: Do you agree with the proposed list of identifiers? Please provide reasons for your answer. <ESMA_QUESTION_77> LSEG agrees with ESMA, except for the proposal to have three different flags for negotiated trades. We explain why, and also note some additional considerations below. General principle: LSEG notes here that it is important that users understand the various flags/identifiers and use them appropriately and accurately when reporting trades. Accurate trade reporting depends on how straightforward it is for a market participant to enter an order or a report. Additional flags should be justified when they are an independent, otherwise they would complicate this further. NT: 22

23 For negotiated trades, LSEG believes a simpler approach would be for the participant to indicate all NTW trades with the same trade-type (NT) on the entry of trade report, and that the MiFID database and/or trading venue reference data will be used by the trading venue or APA or ESMA to identify whether an instrument is liquid or illiquid when generating the trade report for the calculation of the double volume caps, rather than the customer being required to flag as such on the entry of the report. This would simplify entry for users, avoid confusion and unnecessary rejects if an incorrect flag was used by mistake, and achieve the same regulatory outcome. Conditions other than current market price: An additional flag can be used to indicate that a trade is subject to conditions other than the current market price, which can also be used independently of the negotiated trade. For example, on the London Stock Exchange, rule on trade reports requires where a negotiated trade is subject to conditions other than the current market price of the share, a member firm should include the SP (special price) trade reporting condition on the trade report. Cancellations: We further suggest that, since cancellations will affect the net position of volume contributing the capping mechanism, more granularity of cancelled trade types may be required, e.g. CNT should be a cancelled negotiated trade, CR should be a cancelled reference price trade etc. Riskless principal: ESMA may consider an additional flag for such trades to be flagged, or clarify whether such trades can be included under the Agency Cross flag (X). Deferred publication flag: a separate delayed publication flag should be used when a participant requests for deferred publication and if the trade meets the size requirements; this should not be linked to the LIS flag (please see Q79 for further details). Harmonisation of data: LSEG notes that ESMA should link this effort to the requirement to harmonise of data standards and technical arrangements as mentioned in ESMA/2014/549 (Consultation Paper) p.212. LSEG fully supports the drive to higher standards of data quality and greater harmonisation across data sources (including OTC), and has worked closely with other market participants to shape and support the adoption of the Market Model Typology (MMT). The initiative is supported by European exchanges (including LSEG, BATS Chi-X Europe, ICE Euronext, Deutsche Borse), information vendors (including Bloomberg, Markit, Thomson Reuters) and users and the rights in the protocol have been assigned to FIX Protocol Ltd Trust for free use by the market. LSEG is already committed to providing data with this mapping in proprietary trading data feeds to ensure that data can be comprehensively mapped by users and data aggregators across information sources. This increases data quality and the ability to compare and analyse trading activity across Europe. <ESMA_QUESTION_77> Q78: Do you think that specific flags for equity-like instruments should be envisaged? Please justify your answer. <ESMA_QUESTION_78> No. LSEG encourages harmonisation and consistency across equity and equity-like instruments, and suggest the use of the same flags. <ESMA_QUESTION_78> 4 Rules of the London Stock Exchange: 23

24 Q79: Do you support the proposal to introduce a flag for trades that benefit from the large in scale deferral? Please provide reasons for your response. <ESMA_QUESTION_79> LSEG: It is not clear what ESMA s use of large-in-scale deferral refers to. We suggest that ESMA clarifies this further. In our view, there are three considerations here: a) Not all large-in-scale transactions necessarily receive delayed publication; b) Even when the size conditions for delayed publication are fulfilled for LIS transactions, participants may not request for delayed publication; and c) For some ADT classes, the thresholds for deferred publication, as proposed by ESMA on p.89 of the Discussion Paper, may fall below the large in scale threshold. Thus, LSEG s view is that deferred publication is independent from the large in scale waiver, and is elective rather than mandatory. A separate delayed publication flag can be used when a participant requests deferred publication, on the basis of a real economic impact, and if the trade meets the size requirements; and this should not be linked to the LIS flag for the above mentioned reasons. <ESMA_QUESTION_79> Q80: What is your view on requiring post-trade reports to identify the market mechanism, the trading mode and the publication mode in addition to the flags for the different types of transactions proposed in the table above? Please provide reasons for your answer. <ESMA_QUESTION_80> LSEG supports ESMA s approach to enhance transparency on the market mechanism of execution, and efforts to adopt the Market Model Typology (MMT) across all participants. Market mechanism would appear to be best assigned to trading venue instrument reference data. Trading mode and publication mode would be suitable to apply to post-trade reports to assist a trading participant in understanding the nature of the trade and any delayed publication that was applied. <ESMA_QUESTION_80> Q81: For which transactions captured by Article 20(1) would you consider specifying additional flags as foreseen by Article 20(3)(b) as useful? <ESMA_QUESTION_81> LSEG: In our view, securities financing transactions where there is legal transfer of instruments for lending or borrowing purposes are technical transactions which are not subject to the current market price of the instrument or set of instruments. Thus, these trades should be technical or non-publishing trades. <ESMA_QUESTION_81> Q82: Do you agree with the definition of normal trading hours given above? <ESMA_QUESTION_82> Yes, LSEG agrees with ESMA. <ESMA_QUESTION_82> Q83: Do you agree with the proposed shortening of the maximum permissible delay to 1 minute? Do you see any reason to have a different maximum permissible deferral of publication for any equity-like instrument? Please provide reasons for your answer <ESMA_QUESTION_83> LSEG: In our view, shortening the maximum permissible delay from 3 to 1 minute is technically achievable; however the emphasis should be on the improvement of the supervision and enforcement of the current rule, rather than simply setting a shorter permissible delay. 24

25 LSEG also notes that there may be challenges for some users to report in one minute, e.g. portfolio trades where a number of trades are executed simultaneously and manually reported- it may take more than 1 minute to assemble the relevant details for the report. The same maximum delay should be applied for equity and equity-like instruments. <ESMA_QUESTION_83> Q84: Should the deferred publication regime be subject to the condition that the transaction is between an investment firm dealing on own account and a client of the firm? Please provide reasons for your answer. <ESMA_QUESTION_84> Yes, this is an appropriate condition. In LSEG s view, the purpose of this regime is to allow intermediaries a reasonable amount of time to offset positions they acquire from their clients as a result of providing risk capital. Deferred publication would not be appropriate for other types of transactions where offsetting trades are not required as the delay will unnecessarily deny the market price-forming information. <ESMA_QUESTION_84> Q85: Which of the two options do you prefer in relation to the deferral periods for large in scale transactions (or do you prefer another option that has not been proposed)? Please provide reasons for your answer <ESMA_QUESTION_85> LSEG prefers Option B with one modification the option for a longer maximum delay (of up to the end of the 2 nd or 3 rd day) for instruments not having a liquid market (as defined in MiFIR article 2(1)(17)(b)). For liquid instruments: The additional time in Option B will recognise the importance of risk capital for intermediation, allow brokers to offset positions resulting from large orders late in the day, and prevent liquidity from drying up late in the day, particularly for less liquid and SME securities. For less liquid instruments: Access to non-bank capital, including via public equity markets, is an important issue to keep in mind when assessing the proportionality for less liquid instruments. These are often the securities of issuers in the small cap and SME sectors that are crucial to the development of jobs and growth for rebuilding the economy. In our experience, markets in less liquid securities, including SME markets, are more reliant on liquidity from registered market makers. This is particularly true in the UK, which has significantly more SME listings than other European markets. Trades in a number of these less liquid securities are also guaranteed and cleared by a CCP. Activity in these securities is more reliant on market makers, who may need more time to offset their inventory, if they execute in large sizes compared to the normal market size. This is typically how institutional investors execute blocks against market makers. For example, for trading in the week commencing 19 May 2014, if a maximum delay of noon of the next trading day (Option B) was introduced: Only 8% of delayed trades in the FTSE 100 by value were for longer periods (i.e. end of next or 2 nd or 3 rd trading day); However, for the shares in the FTSE UK Small Cap Index (most of which are less liquid under the MiFID II ESMA proposed definition), 74% of delayed trades by value were for longer periods (i.e. between noon of the next trading day and the end of the 3 rd day). These trades account for over a third of total value traded in the FTSE UK Small Cap Index; thus imposition of even shorter delays will clearly have a significant price and liquidity impact on participants and so issuers in this sector. 25

26 Therefore, we suggest there is a need for greater protection for large executions in less liquid securities in order to encourage competitive risk pricing promoting liquidity in the less liquid and SME sectors. To prevent a potential increase in the cost of capital for issuers and transaction costs for investors from the impact on market maker spreads/liquidity caused by shorter deferred publication periods, LSEG suggests that instruments not having a liquid market (as under MiFIR article 2(1)(17)(b)) should be allowed a maximum permissible delay till the 2 nd or 3 rd trading day. For less liquid instruments, a single table based on the thresholds in the current regime for ADT < 1,000,000 (i.e. column 3 in table 8 on p.84 of the Discussion Paper) would be appropriate. The second delay period could be reduced to 120 minutes (currently at 180 minutes). <ESMA_QUESTION_85> Q86: Do you see merit in adding more ADT classes and adjusting the large in scale thresholds as proposed? Please provide alternatives if you disagree with ESMA s proposal <ESMA_QUESTION_86> LSEG supports the addition of more ADT classes and harmonising these with the pre-trade transparency regime. The additional granularity will enable a more tailored approach to defining appropriate transaction sizes. We are also supportive in principle of fixed values for minimum qualifying size as this will be easier to apply, monitor and enforce for the market. <ESMA_QUESTION_86> Q87: Do you consider the thresholds proposed as appropriate for SME shares? <ESMA_QUESTION_87> LSEG: The thresholds may be appropriate, but a maximum delay of noon of the next trading day is not appropriate for the reasons we give in our response to Q85. In our experience, markets in less liquid securities, including SME markets, are more reliant on liquidity from registered market makers. This is particularly true in the UK which has significantly more SME listings than other European markets. A number of these less liquid securities are also guaranteed and cleared by a CCP. Activity in these securities is more reliant on market makers, who may need more time to offset their inventory, if they execute in large sizes compared to the normal market size. This is typically how institutional investors execute blocks against market makers. For example, for trading in the week commencing 19 May 2014, if a maximum delay of noon of the next trading day (Option B) was introduced: Only 8% of delayed trades in the FTSE 100 by value were for longer periods (i.e. end of next or 2 nd or 3 rd trading day); However, for the shares in the FTSE UK Small Cap Index (most of which are less liquid under the MiFID II ESMA proposed definition), 74% of delayed trades by value were for longer periods (i.e. between noon of the next trading day and the end of the 3 rd day). These trades account for over a third of total value traded in the FTSE UK Small Cap Index; thus imposition of even shorter delays will clearly have a significant price and liquidity impact on participants and so issuers in this sector. Therefore, we suggest there is a need for greater protection for large executions in less liquid securities in order to encourage competitive risk pricing. To prevent an increase in the cost of capital for issuers and transaction costs for investors from the impact on market maker spreads/liquidity caused by shorter deferred publication periods, we suggest that instruments not having a liquid market as under MiFIR article 2(1)(17)(b) should be allowed a maximum permissible delay of the end of the 2 nd or 3 rd trading day. A single table based on the thresholds in the current regime for ADT < 1,000,000 (i.e. column 3 in table 8 on p.84 of the Discussion 26

27 Paper) would be appropriate. The second delay period could be reduced to 120 minutes (currently at 180 minutes). For liquid instruments, Option B is appropriate. <ESMA_QUESTION_87> Q88: How frequently should the large in scale table be reviewed? Please provide reasons for your answer <ESMA_QUESTION_88> LSEG: An annual review using previous 12 months data is appropriate for shares and depositary receipts. This would also be consistent with the ADT calculations for LIS and optimise any administrative change management requirements. <ESMA_QUESTION_88> Q89: Do you have concerns regarding deferred publication occurring at the end of the trading day, during the closing auction period? <ESMA_QUESTION_89> Please also see LSEG s response to Q85. Option B is our preferred approach for liquid securities. The additional time (till noon of the next trading day) will allow for brokers to offset positions resulting from large orders late in the day (including auctions), and prevent liquidity from drying up late in the day. Systematically publishing during a closing auction could be misleading. We suggest it would be better to wait until later, although participants may choose to report in this period and, provided these are distinguished appropriately, it should cause minimal disruption. Further, LSEG believes that it is appropriate that deferred trades for publication at end of that day should be published at the end of that day s trade reporting submission session. For LSE, this is currently 17:15 BST. <ESMA_QUESTION_89> Q90: Do you agree with ESMA s preliminary view of applying the same ADT classes to the pre-trade and post-trade transparency regimes for ETFs? Please provide reasons for your answer. <ESMA_QUESTION_90> Yes, LSEG agrees with ESMA. Harmonising ADT classes for pre- and post-trade transparency regime for ETFs is a sensible approach. LSEG also agrees with ESMA s assessment in paragraph 41 that higher thresholds may be valid for ETFs. Similar to our comments for equity instruments in Q52, we suggest that the ADT calculation for ETFs may be artificially low and not capture appropriately average market sizes this is because in our experience, such instruments are generally traded in large quantities, but are not traded as frequently as shares (some may not be traded daily), which may result in a low ADT. We suggest that ESMA should be mindful of this when determining the pre- and post-trade transparency thresholds, and may suggest higher thresholds than those proposed in the discussion paper. A more frequent calculation may also be appropriate for ETFs. We suggest a semi-annual calculation, based on data from a previous 6-12 month period. This is because ETFs are more sensitive to shortterm changes in liquidity, and a semi-annual calculation would take this into account. This is consistent with the approach we suggest for the LIS threshold. <ESMA_QUESTION_90> 3.3. Systematic Internaliser Regime - Equities 27

28 Q91: Do you support maintaining the existing definition of quotes reflecting prevailing market conditions? Please provide reasons for your answer. <ESMA_QUESTION_91> LSEG: Greater parity between trading venues and SIs, and impact of the trading obligation: In our view, the Level II provisions must deliver a level playing field for market participants, regardless of which mechanism under the trading obligation for shares they use to trade i.e. multilateral trading venues (i.e. Regulated Markets and MTFs) and SIs (for trades executed bilaterally outside trading venues). This approach is consistent with ESMA s analysis in paragraph 9 of p.176 of the Consultation Paper, where it comments that Given that systematic internalisers are permitted platforms under the trading obligations and that their quoting obligations depend on whether the instrument is liquid, [ ] it is necessary to ensure a level playing field exists between trading venues and systematic internalisers to the extent possible. There is a concern that whilst MiFID II/ R prescribes a thorough transparency regime, microstructure and limits to dark trading for market participants in the trading venue space, the requirements are not of the same nature in the SI space, where trading might be viewed as non-pre trade transparent for sizes above SMS. Further, we suggest that using the MiFID I definitions to define the pre-trade transparency characteristics for SIs may not be appropriate, given that SIs are a permitted platform under the trading obligation, the purpose of which was to allow only genuine OTC and ensure that trades are transparent and contribute to the price-formation process. An outcome therefore might be an increase in execution through the SI category to avoid the limitations of trading in the dark (double volume cap, mid-point for RPW trades, price conditions on NTW trades), which only apply to trading venues. We recognise that in some circumstances, dark execution may offer genuinely better results for investors (e.g. through better prices or sizes). However, we suggest that these circumstances need to be defined further. One way of ensuring consistency might be to treat an SI as a private market making service for its clients (but not to the wider market), as the characteristics of a trade executed bilaterally against a market maker providing pre-trade transparency (up to SMS) or a negotiated trade, can be seen as very similar to the activities of an SI. Thus, similar conditions should apply to the activities of SIs and those of a registered market maker. Therefore, since one of the objectives of MIFID II is to protect price formation in pre-trade transparent markets, we suggest that the existing definition of quotes reflecting prevailing market conditions is developed further; particularly to clearly specify when it is appropriate for SIs to offer price improvement to their published quotes. Therefore, LSEG suggests that for trades up to Standard Market Size, if price improvement can be offered in certain justified cases for small sizes: This should be at the reference price as defined in MiFIR, i.e. the mid-point of the best bid or offer on the market where that instrument was first admitted to trading or the most relevant liquid market. This would ensure parity of conditions under which price improvement can be offered in trading venues and through SIs. We suggest that price improvement of fractions of a tick from the BBO do not present a significantly better outcome for investors nor does it contribute to the price discovery process, especially in small sizes; we suggest the mid-point is an appropriate price. For trades above Standard Market Size, LSEG suggests that SI quotes must be subject to the best execution policy of the firm, but no worse than the conditions for NTW for the same instruments on public markets. <ESMA_QUESTION_91> 28

29 Q92: Do you support maintaining the existing table for the calculation of the standard market size? If not, which of the above options do you believe provides the best trade-off between maintaining a sufficient level of transparency and ensuring that obligations for systematic internalisers remain reasonable and proportionate? Please provide reasons for your answer. <ESMA_QUESTION_92> LSEG disagrees with ESMA s suggested approaches, and suggests that some measure of upward recalibration may be required for the SMS. As in our comments to Q76 and Q91, we support greater parity between the activity of market participants on RMs and MTFs and those on SIs, to level the playing field for mechanisms under the trading obligation and reduce the potential for regulatory arbitrage. We suggest that the activity of SIs (taking risk and providing two-way liquidity to its clients) is similar to that of registered market markets on trading venues, both in purpose and outcomes. Consistent treatment of such activity would be a desirable outcome, in our view. In line with ESMA s analysis of AVT in Section 4.3, we note that all but 5 UK securities are on the lowest SMS band of 7,500. The remaining 5 are on the second band of 15,000. With a requirement to quote at 10% of SMS, the minimum SI quote size is between 750 and 1,500 for all UK securities. We also note that above these sizes, an SI does not need to offer any pre-trade transparency for its executions and does not fall under any capping mechanism either (unlike trading under the RPW or NTW in trading venues). In comparison, on regulated markets such as LSE and Borsa Italiana, registered market makers are usually subject to stricter obligations. For example, LSE has linked its minimum quote size for shares to 1% of individual security s ADT, with a minimum ( 2,500) and maximum threshold ( 25,000). So for UK liquid securities this currently means that the minimum market maker obligation is between approximately 24,000 and 30,000. This ensures that securities have higher minimum quoting obligations to reflect their overall liquidity and potential to accommodate price impact, even if average trade sizes are similar across the liquidity spectrum. We would welcome a recalibration of the SMS table to ensure that SI quotes are providing similar pretrade transparency as would be required of registered market markers on a regulated market. A possible approach could be that the SMS is at a certain percentage of the ADT capped at an appropriately suitable floor and ceiling. ESMA should also keep in mind that the minimum transparency requirement is only 10% of SMS. <ESMA_QUESTION_92> Q93: Do you agree with the proposal to set the standard market size for depositary receipts at the same level as for shares? Please provide reasons for your answer. <ESMA_QUESTION_93> Yes, LSEG agrees with ESMA s approach. <ESMA_QUESTION_93> Q94: What are your views regarding how financial instruments should be grouped into classes and/or how the standard market size for each class should be established for certificates and exchange traded funds? <ESMA_QUESTION_94> TYPE YOUR TEXT HERE <ESMA_QUESTION_94> 29

30 3.4. Trading obligation for shares (Article 23, MiFIR) Q95: Do you consider that the determination of what is non-systematic, ad-hoc, irregular and infrequent should be defined within the same parameters applicable for the systematic internaliser definition? In the case of the exemption to the trading obligation for shares, should the frequency concept be more restrictive taking into consideration the other factors, i.e. ad-hoc and irregular? <ESMA_QUESTION_95> LSEG agrees with ESMA s analysis in paragraph 8 on p.101 of the Discussion Paper. ESMA should provide maximum legal certainty through clear and precise definitions of what is and what is not in scope of the trading obligation (e.g. what trades are ad-hoc, irregular, non-price forming etc). This determination should be based on the parameters applicable for the systematic internaliser (i.e. frequent, substantial, systematic), but also needs additional clarity on the definition of ad-hoc and irregular in order to prevent the circumvention of the trading obligation as noted in paragraph 12. This is consistent with ESMA s general approach in the Discussion and Consultation Papers, where it proposes further details to terms and definitions in the MiFID legislation should it help to offer clarification or avoid unintended consequences. An exhaustive list may avoid legal uncertainty. We also note here that if MiFID II focuses on perfecting the definitions and controls on trading venue and SI activity, then anything that is not on venue or SI is by definition OTC. This puts the onus on firms to justify why a trade cannot be transacted on a trading venue or SI. LSEG also agrees with ESMA s clarification at the Open Hearing on Markets Issues in MiFID II on 07 July 2014 that Large-In-Scale transactions (i.e. with sizes above the relevant LIS thresholds) should be subject to the trading obligation. Internal matching system In addition, clarity is required from the European Commission or ESMA on what systems operated by investment firms qualify as internal matching systems for the purpose of authorisation as MTFs, as there is no such definition in MiFID II, although we recognise this is not strictly part of ESMA s mandate. LSEG s initial view is that an internal matching system must be a multilateral system ( any system or facility in which multiple third-party buying and selling interests in financial instruments are able to interact in the system ), as defined in MiFID Art 4(1)(19), or functionally perform the same activity (e.g. through back-to-back trading, or through an agency cross, or as riskless principal, where the intermediary takes no market risk or proprietary position). ESMA may wish to clarify which systems it believes to be multilateral, to avoid trading systems operating outside the scope of the MiFID II legislation. We note that MiFIR recital 7 says the term system encompasses all those markets that are composed of a set of rules and a trading platform as well as those that only function on the basis of a set of rules. We also note that recital 20 of MiFIR states that multiple-dealer platform [ ] should not be considered a Systematic Internaliser. There may be a case to argue that such multiple-dealer platforms are multilateral, and hence must be authorised as MTFs. <ESMA_QUESTION_95> Q96: Do you agree with the list of examples of trades that do not contribute to the price discovery process? In case of an exhaustive list would you add any other type of transaction? Would you exclude any of them? Please, provide reasons for your response. <ESMA_QUESTION_96> 30

31 Yes, LSEG agrees with ESMA s proposed list. LSEG notes that this list is similar to the list ESMA proposes in paragraph 75 on p.69 of the Discussion Paper for the negotiated trade waiver. Please also see our response to Q63 in this regard. <ESMA_QUESTION_96> Q97: Do you consider it appropriate to include benchmark and/or portfolio trades in the list of those transactions determined by factors other than the current valuation of the share? If not, please provide an explanation with your response. <ESMA_QUESTION_97> LSEG: Yes, both these involve pricing which may be based on factors other than current prices. As noted in the examples given in the proposal, benchmark trades will often include historic pricing which may not reflect current market valuations, e.g. an all day VWAP trade printing at the end of the day. In terms of portfolio risk, we note that one of the factors of valuation is the market correlation between the components of the portfolio. For example, a large portfolio of equities from the oil sector will be priced differently than if each individual component were being priced because of the combined and correlated market risk. <ESMA_QUESTION_97> 3.5. Introduction to the non-equity section and scope of non-equity financial instruments Q98: Do you agree with the proposed description of structured finance products? If not, please provide arguments and suggestions for an alternative. <ESMA_QUESTION_98> Introduction: General Comments on LSEG s experience of operating non-equity markets: In answering the questions in this and the following sections, LSEG draws on its experience of operating diverse markets across the non-equity space. A short description of our key markets is below: Fixed income: Borsa Italiana and EuroTLX Borsa Italiana and EuroTLX SIM operate multiple trading venues in the non-equities space in Europe, including MOT (a regulated market with 115 Italian Government Bonds, 554 Corporate Bonds mainly issued by banks and 478 Supranational and Foreign Government Bonds), ExtraMOT (an MTF with over 480 Corporate Bonds issued by both Italian and international companies) and EuroTLX (an MTF with over 4,100 Bonds, 3,600 of which are Corporate Bonds). In the retail/retail size segment, MOT is the largest European Market for trades and turnover, and EuroTLX is the reference market for Corporate Bonds. The tables below list the most relevant statistics for these markets in 2013: All Bonds - Activity 2013 FY 2013 Instruments as of Trades Turnover (ml) Average Ticket 31/12 MOT 1,197 5,892, ,952 55,994 ExtraMOT ,302 5,025 52,732 EuroTLX 4,104 3,082,117 89,591 29,067 Totals 4,878 9,070, ,568 46,810 Source: LSEG 31

32 Corporate Bonds - Activity 2013 FY 2013 Instruments as of Trades Turnover (ml) Average Ticket 31/12 MOT ,885 15,811 16,912 ExtraMOT ,430 1,744 74,420 EuroTLX 3,590 2,048,462 52,151 25,459 Totals 4,321 3,066,777 69,706 23,183 Source: LSEG In 2013, 9 million trades were executed in Borsa Italiana s bond markets in just under 5000 bonds (3 million in corporate bonds) with average size of 47,000 for all bonds and of 23,000 for corporate bonds only. The traded value was approximately 425 billion for all bonds and 70 billion for corporate bonds. Our markets represent significant liquidity in the fixed income space in the EU, particularly in the electronic markets. We note that according to ESMA s suggested thresholds, only 5% of securities on Borsa Italiana and EuroTLX will be deemed liquid. In our experience, most securities on these markets are suitable for electronic and transparent trading, so we are concerned the thresholds may be set too low. In line with our response to Q112 of the Discussion Paper, we suggest that for the purpose of identifying liquid instruments, the framework should consider the establishment of a clear minimum coverage ratio, which takes into consideration the number of instruments and number of trades/contracts captured, not just volume traded Borsa Italiana and EuroTLX operate markets which guarantee transparency and efficiency in the price formation mechanism and ensure regular and orderly trading. In the context of the rules relating to illiquid securities framed by the current Italian domestic regulation and requiring additional protection for retail investors, our markets are deemed liquid also by virtue of the presence of market makers; this, together with the order driven microstructure, makes our markets liquid. MTS Group The MTS Group is one of Europe s leading operators of electronic fixed income trading markets, with over 500 unique counterparties and average daily volumes exceeding 100 billion. The MTS Group operates multiple trading venues in the non-equity space (regulated markets, MTFs and Broker Dealers) across Europe, Western Asia and the USA. Key statistics from MTS s markets (bonds only) for 2013 are below: FY 2013 Instruments as of Trades Turnover (ml) Average Ticket 31/12 MTS 4, ,728 3,054,465 5,832,156 Source: LSEG LSE The London Stock Exchange operates fixed income markets as a regulated market, including for corporate, government, local authority and multilateral debt. Issuers can choose to admit securities to trading on the Main Market or the Professional Securities Market. It also operates the Order Book for Retail Bonds (ORB), an electronic regulated market for retail investors, building on the success of Borsa Italiana s MOT market. Renminbi-denominated bonds and Islamic finance bonds can also be listed or admitted to trading on the LSE. 32

33 Key statistics from LSE s markets (all admitted bonds) for are below (please note, a significant amount of activity reflects off-exchange trade reporting in UK government bonds): FY 2013 Instruments as of Trades Turnover (ml) Average Ticket 31/12 LSE fixed income 19, ,774 5,473,153 9,140,599 Source: LSEG Structured products Both London Stock Exchange and Borsa Italiana offer the listing and trading of structured products. Derivatives LSEG operates multiple trading venues for the trading of exchange traded derivatives (ETDs) e.g. LSE Derivatives, IDEM, IDEX and AGREX, and OTC derivatives (OTCDs), including MTS Swaps. -- LSEG: General Comments on Non-equity Transparency We welcome the extension of MiFID pre and post trade transparency to non-equity instruments. We support the approach of the level 1 text which recognises that such transparency requirements should be calibrated for different types of trading system. As an overriding principle, we believe that the objective of the regime should be to maximise transparency to the extent possible and we argue in particular for liquidity determination thresholds that capture a sufficiently high number of instruments, subject to appropriate granularity applied to ESMA s preferred option of class of instruments approach (COFIA). It is important that the calibration of transparency for non-equities does not lead to incentives for participants to trade away from transparent multilateral platforms to less liquid/bilateral arrangements (where less demanding calibration is required or a lower degree of enforcement is applied). These points underpin our responses to Questions 103 to 167 on non-equity transparency -- Specific response to Q98 LSEG: We agree with proposed description. Although not specifically related to the definition of structured products we would also make the following comment in relation to bonds: The definition for bonds should also include "index structured bonds", which are attributable to the category under discussion, because of the provision of the capital repayment at the nominal value (only the coupon flow is variable). We therefore suggest an amendment to Table 1 of Annex and to include "equity index linked bonds" in the other potential liquidity sub category. <ESMA_QUESTION_98> Q99: For the purposes of transparency, should structured finance products be identified in order to distinguish them from other non-equity transferable securities? If so, how should this be done? <ESMA_QUESTION_99> LSEG: We do not believe ABSs should be treated in a different way from other debt products for the purposes of transparency. See also our response to Q98. <ESMA_QUESTION_99> Q100: Do you agree with the proposed explanation for the various types of transferable securities that should be treated as derivatives for pre-trade and post trade transparency? If not, please provide arguments and suggestions for an alternative. 33

34 <ESMA_QUESTION_100> LSEG: We agree with the proposed types of transferable securities proposed by ESMA. However, in addition, we also suggest the inclusion of Exchange Traded Currencies/Commodities and Exchange Traded Notes as securitised derivatives. <ESMA_QUESTION_100> Q101: Do you agree with ESMA s proposal that for transparency purposes market operators and investment firms operating a trading venue should assume responsibility for determining to which MiFIR category the non-equity financial instruments which they intend to introduce on their trading venue belong and for providing their competent authorities and the market with this information before trading begins? <ESMA_QUESTION_101> LSEG: We understand the approach proposed by ESMA for trading venues to assume responsibility for determining the classification for bonds/structured products and recognise the need for this to ensure that the appropriate pre and post trade transparency requirements are applied. While we broadly support this approach, we would highlight the fact that, as ESMA recognises, any such determination will need to be based on information made available to the trading venue by other relevant parties, e.g. issuers, advisors, listing authorities etc. The trading venue itself would not necessarily be the only entity responsible for assessing the instruments against ESMA criteria. Any RTS in this area should therefore provide for trading venues to include, as part of the process of admitting instruments to a trading system, the need, where relevant, for third parties to provide the necessary information for the classification of that instrument to be determined. In addition, while we support this approach for new instruments admitted to a trading venue after the application of MiFIR, we believe ESMA should confirm how the classification would be determined for existing instruments. Will this be done as part of the development of any COFIA approach and the calculation of the initial liquidity determination? <ESMA_QUESTION_101> Q102: Do you agree with the definitions listed and proposed by ESMA? If not, please provide alternatives. <ESMA_QUESTION_102> LSEG: We agree with the definitions proposed by ESMA. Note, however, our response to Q98, regarding index structured bonds : The definition for bonds should also include "index structured bonds", which are attributable to the category under discussion, because of the provision of the capital repayment at the nominal value (only the coupon flow is variable). We therefore suggest an amendment to Table 1 of Annex and to include "equity index linked bonds" in the other potential liquidity sub category. <ESMA_QUESTION_102> 3.6. Liquid market definition for non-equity financial instruments Q103: Do you agree with the proposed approach? If you do not agree please provide reasons for your answers. Could you provide for an alternative approach? <ESMA_QUESTION_103> 34

35 LSEG: Yes, we agree with ESMA s preferred Option 3. However, we would welcome clarification regarding the basis of the proposed trading calculations/analysis, whether it includes OTC trading volumes and how this aligns with the approaches to data collection for the determination of LIS thresholds (page 179 DP). <ESMA_QUESTION_103> Q104: Do you agree with the proposed approach? If you do not agree please provide reasons. Could you provide an alternative approach? <ESMA_QUESTION_104> LSEG: Yes, we agree with ESMA s preferred Option 2. It is important to recognise that this option/calculation does not represent the average size of transactions it is a measure of turnover divided by the number of trading days, i.e. average daily turnover (ADT). This is consistent with the approach taken under MiFID currently (and proposed to continue under MiFID II) for equity/equity-like instruments and we support this consistent approach. <ESMA_QUESTION_104> Q105: Do you agree with the proposed approach? If you do not agree please provide reasons. Could you provide an alternative approach? <ESMA_QUESTION_105> LSEG: We support Option 1, which takes into consideration the participants being involved in at least one transaction. However, we also believe that it is appropriate, where relevant, based on the type of financial instrument and the role of participants, to consider, for the purpose of the definition of a liquid market, the presence of at least one liquidity provider with a contractual arrangement in place and which undertakes to support liquidity by means of posting firm two-way quotes (i.e. Option 2). Such activity should also cover participants who support the liquidity of the market by means of posting bid-only quotes. <ESMA_QUESTION_105> Q106: Do you agree with the proposed approach? If you do not agree please provide reasons. Could you provide an alternative approach? <ESMA_QUESTION_106> LSEG: We believe there are a number of questions that ESMA needs to consider concerning the calculation of average spreads: It is not clear what end of day relative bid-ask spread means is this an end of day snapshot or a weighted average for a specified time period at the end of the day? The current notion of "bid ask spread" is not sufficiently well specified, in the context of taking into account the concept of "volume attached". Is the intention to specify a minimum size/notional threshold or to calculate a depth weighted measure? What is the intended approach where an instrument is traded on more than one venue? In certain circumstances the end of day spread may not be the most appropriate time for measurement of the spread, therefore some flexibility should be available to use appropriate alternative times; In addition, average spreads are not always a viable way of measuring liquidity for all instruments, e.g. options; However, in resolving these issues, there is the possibility that a proposed methodology will become too complex and costly to be developed and maintained. <ESMA_QUESTION_106> Q107: Should different thresholds be applied for different (classes of) financial instruments? Please provide proposals and reasons. <ESMA_QUESTION_107> LSEG: We believe that it would be beneficial for the same methodology to be used across all relevant financial instruments, although the parameters/thresholds will vary by class of financial instrument. 35

36 It is difficult to propose specific thresholds given the outstanding questions on the proposed approach (see response to Q106). However, we believe that the approach to setting thresholds should be consistent with an objective to maximise available transparency to the extent practicable. In this context we would propose relative low thresholds, i.e. if there is a stable spread in reasonable size, then it is appropriate to categorise the market as relatively liquid even if the spread is wide. <ESMA_QUESTION_107> Q108: Do you have any proposals for appropriate spread thresholds? Please provide figures and reasons. <ESMA_QUESTION_108> LSEG: It is difficult to propose specific thresholds given the outstanding questions on the proposed approach (see response to Q106). However, we believe that the approach to setting thresholds should be consistent with an objective to maximise available transparency to the extent practicable. <ESMA_QUESTION_108> Q109: How could the data necessary for computing the average spreads be obtained? <ESMA_QUESTION_109> LSEG: For products made available to trade on trading venues, it is appropriate for those trading venues to provide this information. However, as we suggest in our response to Q106, the process required to capture and calculate the data may be complex, given the need to rebuild order books to calculate spread/depth measures. <ESMA_QUESTION_109> Q110: Do you agree with the proposed approach? If you do not agree please providereasons for your answer. Could you provide an alternative approach? <ESMA_QUESTION_110> LSEG: All of the proposed measures represent simplifications, apart, potentially, from the average spread calculation. All will have drawbacks or inaccuracies in certain circumstances (including the misleading use of the term average size of transactions see our response to Q104), and, per our response to questions , there are a number of issues regarding the average spread calculation. In addition, average spreads may not be an appropriate indicator of liquidity for some instruments, or, as ESMA recognises, this data may not be available. We believe that not all the criteria have to be met in order for the class of the financial instrument to be considered liquid and therefore we would propose Option 2. In particular, we suggest it is sufficient, for the purpose of defining liquidity of an instrument, to satisfy the following criteria: average size average frequency number of participants (see also response to Q111) <ESMA_QUESTION_110> Q111: Overall, could you think of an alternative approach on how to assess whether a market is liquid bearing in mind the various elements of the liquid market definition in MiFIR? <ESMA_QUESTION_111> LSEG: Additional measures/criteria could be: For bond markets a key parameter used extensively as an indicator of likely liquidity is issue size. We believe that a measure of issue size for bonds should be included as an additional parameter; The presence of at least one specialist/market maker (also with bid only obligations) should also be considered a sufficient condition for the qualification of an instrument as liquid. (see also our response to Q105 on this). 36

37 <ESMA_QUESTION_111> Q112: Which is your preferred scenario or which combination of thresholds would you propose for defining a liquid market for bonds or for a sub-category of bonds (sovereign, corporate, covered, convertible, etc.)? Please provide reasons for your answer. <ESMA_QUESTION_112> LSEG: We believe that the approach to setting the combination of thresholds should be consistent with an objective to maximise available transparency to the extent practicable. While we accept that that the initial analysis carried out by ESMA covers only a sample of bonds, and applies only some of the criteria, the results reflect a low number of liquid bonds. For effective pre and post trade transparency, we suggest that for the purpose of identifying liquid instruments, the framework should consider: The establishment of a clear minimum coverage ratio; That the number of instruments and number of trades/contracts captured should be taken into account, not just volume traded; That the liquidity of an individual bond should not assessed in isolation, as it is inevitably related to the liquidity of related bonds. This has implications for the way in which liquidity measures are applied (see our response to Q113) and also to the scenarios / combination of thresholds employed; That the determination of liquidity will include other criteria e.g. number of participants and we would refer to our response to Q111 where we suggest that a sufficient condition in order to qualify an instrument as liquid is the presence of at least one market maker which supports the trading. That parameters related to the number of trades, the number of days, average daily volume should be lowered if compared to those used in the scenarios considered in the DP and they should take into consideration the empirical characteristics of trades considered in the assessment. In particular: o the number of days parameter should be lowered in consideration of the dynamics that determine the frequency of trading; o the average daily volume parameter should be based on the average trade size of the bulk of trades considered in the assessment. Given the dynamics that determine the frequency of trading on bonds, the classes of securities, once classified as liquid, may continue to benefit from the transparency regime until the maturity of the bonds. <ESMA_QUESTION_112> Q113: Should the concept of liquid market be applied to financial instruments (IBIA) or to classes of financial instruments (COFIA)? Would be appropriate to apply IBIA for certain asset classes and COFIA to other asset classes? Please provide reasons for your answers <ESMA_QUESTION_113> LSEG: We believe that the concept of liquidity for non equity instruments (bonds and derivatives) should be applied on a COFIA basis. In relation to bonds as stated in our response to Q112, the liquidity of an individual bond should not be assessed in isolation, as it is inevitably related to the liquidity of related bonds. However, a deeper level of granularity would be required compared to that proposed by ESMA. In this context, we broadly support the approach reflected in the tables in Annex However, we suggest that the classes to be taken into account should incorporate additional qualitative criteria deemed relevant from a liquidity perspective (e.g. residual maturity and issuance size for bonds, maturity, underlying and strike price for derivatives). See also our response to Q98 in relation to equity index linked bonds. 37

38 <ESMA_QUESTION_113> Q114: Do you have any (alternative) proposals how to take the range of market conditions and the life-cycle of (classes of) financial instruments into account - other than the periodic reviews described in the sections periodic review of the liquidity threshold and periodic assessment of the liquidity of the instrument class, above? <ESMA_QUESTION_114> LSEG: We believe that ESMA should consider whether there is any merit in providing for a waiver from transparency arrangements for brand new products developed and admitted to trading venues, depending on the type of product, market structure and the need to set any new liquidity thresholds for that product. Any such waiver would need to take account of initial liquidity calculations and should recognise that a waiver would not be appropriate where a product fulfilled all of the liquidity thresholds for that or similar products. <ESMA_QUESTION_114> Q115: Do you have any proposals on how to form homogenous and relevant classes of financial instruments? Which specifics do you consider relevant for that purpose? Please distinguish between bonds, SFPs and (different types of) derivatives and across qualitative criteria (please refer to Annex 3.6.1). <ESMA_QUESTION_115> LSEG: We broadly support the tables proposed on Annex See also our responses to Q98 and Q100. <ESMA_QUESTION_115> Q116: Do you think that, in the context of the liquidity thresholds to be calculated under MiFID II, the classification in Annex is relevant? Which product types or sub-product types would you be inclined to create or merge? Please provide reasons for your answers <ESMA_QUESTION_116> LSEG: Yes. See responses to Q113 and Q115. <ESMA_QUESTION_116> Q117: Do you agree with the proposed approach? If not, please provide rationales and alternatives. <ESMA_QUESTION_117> LSEG: Yes, we agree with ESMA s proposal. In addition, we would support the inclusion of the presence of committed market makers in the qualitative criteria to be taken into account by NCAs. We also believe that where a decision is made to suspend transparency requirements for an instrument, it should still be possible for a trading venue to which that instrument is admitted maintain transparency requirements where it would appropriate to do so. <ESMA_QUESTION_117> Q118: Do you agree with the proposed thresholds? If not, please provide rationales and alternatives. <ESMA_QUESTION_118> LSEG: Paragraph 60 on page 130 of the DP states that temporary suspension of liquidity should only be triggered in exceptional circumstances. One of the methods proposed by ESMA is to assess current trading volumes against the latest official ADT of that instrument. However, we would argue that this approach could result in transparency being suspended for an instrument where trading volumes, while significantly lower, were still above the ADT threshold for that class which determines a liquid instrument (and for which transparency requirements would ordinarily apply). 38

39 We would suggest that current trading volumes should be assessed against that liquidity threshold, possibly against a % of that threshold, along the lines suggested by ESMA. This would be more consistent with the concept of exceptional circumstances and provide a more consistent basis to assess the need for temporary suspensions of transparency across asset classes. <ESMA_QUESTION_118> 3.7. Pre-trade transparency requirements for non-equity instruments Q119: Do you agree with the description of request-for-quote system? If not, how would you describe a request-for-quote system? Please give reasons to support your answer. <ESMA_QUESTION_119> LSEG: Yes, we agree, although we note that the RFQ functionality works slightly differently in some markets where the RFQ message sent by the requesting participant is visible to the whole market (not just market makers) and market maker quotes are not only visible but also available to any market participant (not just requesting participant). <ESMA_QUESTION_119> Q120: Do you agree with the inclusion of request-for-stream systems in the definition of request-for-quote system? Please give reasons to support your answer. <ESMA_QUESTION_120> LSEG: Yes, request-for-stream is technically different but functionally similar to request-for-quote. <ESMA_QUESTION_120> Q121: Do you think that apart from request-for-stream systems other functionalities should be included in the definition of request-for-quote system? If yes, please provide a description of this functionality and give reasons to support your answer. <ESMA_QUESTION_121> LSEG: No. <ESMA_QUESTION_121> Q122: Do you agree with the description of voice trading system? If not, how would you describe a voice trading system? <ESMA_QUESTION_122> LSEG: We broadly agree with the description provided. However, we would note that the definition is very broad, and would capture a range of hybrid systems which are largely automated but have an element of voice negotiation. We would propose that it be restricted to systems which are arranged entirely through voice negotiation, without the use of technological support. Alternatively, the definition could be restricted to voice trading systems in which relevant prices are not shown on parallel screen-based systems. <ESMA_QUESTION_122> Q123: Do you agree with the proposed table setting out different types of trading systems for non-equity instruments? <ESMA_QUESTION_123> LSEG: Yes, we agree, but as stated in our response to Q122, the definition of voice trading systems should be tightened. <ESMA_QUESTION_123> 39

40 Q124: Do you think that the information to be made public for each type of trading system provides adequate transparency for each trading system? <ESMA_QUESTION_124> LSEG: Yes, we agree and support the extension to the non equity space of the current minimum pre trade information provided for share instruments and equity like instruments. <ESMA_QUESTION_124> Q125: Besides the trading systems mentioned above, are there additional trading models that need to be considered for pre-trade transparency requirements in the non-equity market space? <ESMA_QUESTION_125> LSEG: No. <ESMA_QUESTION_125> Q126: If you think that additional trading systems should be considered, what information do you think should be made public for each additional type of trading model? <ESMA_QUESTION_126> TYPE YOUR TEXT HERE <ESMA_QUESTION_126> Q127: Based on your experience, what are the different types of voice trading systems in the market currently? What specific characteristics do these systems have? <ESMA_QUESTION_127> TYPE YOUR TEXT HERE <ESMA_QUESTION_127> Q128: How do these voice trading systems currently make information public or known to interested parties at the pre-trade stage? <ESMA_QUESTION_128> TYPE YOUR TEXT HERE <ESMA_QUESTION_128> Q129: Do you agree with ESMA s approach in relation to the content, method and timing of pre-trade information being made available to the wider public? <ESMA_QUESTION_129> TYPE YOUR TEXT HERE <ESMA_QUESTION_129> Q130: Do you agree with the above mentioned approach with regard to indicative pre-trade bid and offer prices which are close to the price of the trading interests? Please give reasons to support your answer <ESMA_QUESTION_130> LSEG: We agree that the methodology should be documented, robust and transparent. Trading venues should be entitled to specify the specific methodology an average of volume weighted bid and offer prices is one of many examples of a potentially valid methodology. <ESMA_QUESTION_130> Q131: If you do not agree with the approach described above please provide an alternative <ESMA_QUESTION_131> 40

41 TYPE YOUR TEXT HERE <ESMA_QUESTION_131> 3.8. Post-trade transparency requirements for non-equity instruments Q132: Do you agree with the proposed content of post-trade public information? If not, please provide arguments and suggestions for an alternative. <ESMA_QUESTION_132> LSEG: Yes, we agree - subject to clarification of "price notation" and "quantity notation". <ESMA_QUESTION_132> Q133: Do you think that the current post-trade regime for shares on the systematic internaliser s identity should be extended to non-equity instruments or that the systematic internaliser s identity is relevant information which should be published without exception? <ESMA_QUESTION_133> LSEG: We do not agree that the identity of an SI should be revealed when an SI uses its own capital and exposes itself to risk to provide liquidity (consistent with the treatment of the activity of market makers on RMs/MTFs, whose identity is not revealed on a post-trade report). However, if in the MiFID II regime, an SI is allowed to execute trades without taking on risk, i.e. as riskless principal or through back-to-back trades/agency cross trades, then the identity of an SI should be revealed (consistent with the treatment of a trading venue, whose identity is revealed on the post-trade report since in this case an SI is matching trades rather than making markets). <ESMA_QUESTION_133> Q134: Is there any other information that would be relevant to the market for the above mentioned asset classes? <ESMA_QUESTION_134> LSEG: No. <ESMA_QUESTION_134> Q135: Do you agree with the proposed table of identifiers for transactions executed on nonequity instruments? Please provide reasons for your answer. <ESMA_QUESTION_135> LSEG: We broadly agree with the table of identifiers, subject to the following observations: We would propose the inclusion of a separate flag for riskless principal trades; A give up/give in flag should only be relevant for trading gives ups (for example where a firm undertakes trading activity with the intent to transfer the resultant position to another firm to provide a hedge). Where a give-up is in respect of trades passed to another firm for the purpose of post trade processing, this should not be published. <ESMA_QUESTION_135> Q136: Do you support the use of flags to identify trades which have benefitted from the use of deferrals? Should separate flags be used for each type of deferral (e.g. large in scale deferral, size specific to the instrument deferral)? Please provide reasons for your answer. <ESMA_QUESTION_136> LSEG: Yes, it is appropriate for flags to be used to identify trades which are subject to deferrals. <ESMA_QUESTION_136> 41

42 Q137: Do you think a flag related to coupon payments (ex/cum) should be introduced? If yes, please describe the cases where such flags would be warranted and which information should be captured. <ESMA_QUESTION_137> LSEG: We believe that for institutional markets such information would be redundant, and is not information which is necessarily routinely captured by trading venues. <ESMA_QUESTION_137> Q138: Do you think that give-up/give-in trades (identified with a flag) should be included in post-trade reports or not made public? Please provide reasons for your answers. <ESMA_QUESTION_138> LSEG: See also our response to Q135. We do not believe that give-up / give-in trades in respect of post trade processing should be published as they are clearing related and do not contribute to the objectives of post-trade transparency, which could lead to a misleading impression of trading activity. Similarly, a number of interest rate swap systems operate on the principle of sponsored access, under which the counterparty to the client s trade is always the sponsored access dealer (sometime called the Prime dealer). Where appropriate, an additional back-to-back trade is effected between the sponsored access dealer and the executing dealer. The only difference between the original trade and the matching back-to-back trade is that the price of the original trade includes an AVC adjustment to reflect the credit risk associated with the client. We do not believe that details of the back-to-back trade should be published, for the following reasons: As for give-up / give-in trades, it does not contribute to the objectives of post-trade transparency; The only additional information contained is the effective credit quality of the client; Publication could lead to a misleading impression of trading activity. <ESMA_QUESTION_138> Q139: Do you agree that securities financing transactions should be exempted from the post-trade transparency regime? <ESMA_QUESTION_139> LSEG: See our response to Q81. In our view, securities financing transactions where there is legal transfer of instruments for lending or borrowing purposes are technical transactions which are not subject to the current market price of the instrument or set of instruments. Thus, these trades should be technical or non-publishing trades. <ESMA_QUESTION_139> Q140: Do you agree that for the initial application of the new transparency regime the information should be made public within five minutes after the relevant non-equity transaction? Please provide reasons for your answer. <ESMA_QUESTION_140> LSEG: We would suggest that the same regime envisaged for equity/equity-like instruments should also apply to non-equities. Please see our response to Q83 <ESMA_QUESTION_140> Q141: Do you agree with the proposed text or would you propose an alternative option? Please provide reasons for your answer. <ESMA_QUESTION_141> LSEG: We agree with the principle of having different deferral times for different transaction types based on the size of the trade and the liquidity of the instrument. However, we do not agree with the principle of "partial" publication. We believe this would be difficult to operate in practice and we would much rather 42

43 have a post-trade transparency regime whereby the trade publication is deferred completely for a period of time after which the all the relevant trade details are published at the same time. <ESMA_QUESTION_141> Q142: Do you agree that the intra-day deferral periods should range between 60 minutes and 120 minutes? <ESMA_QUESTION_142> LSEG: Yes, we agree. <ESMA_QUESTION_142> Q143: Do you agree that the maximum deferral period, reserved for the largest transactions, should not exceed end of day or, for transactions executed after 15.00, the opening of the following trading day? If not, could you provide alternative proposals? Please provide reasons for your answer. <ESMA_QUESTION_143> LSEG: We would suggest that for instruments for which there is a liquid market, in line with the proposed equity regime, the maximum deferral permitted is until noon of the next trading day, for those trades executed after 15:00 and where the size is equal to or above the LIS threshold. <ESMA_QUESTION_143> Q144: Do you consider there are reasons for applying different deferral periods to different asset classes, e.g. fixing specific deferral periods for sovereign bonds? Please provide arguments to support your answer. <ESMA_QUESTION_144> LSEG: No, we do not believe it would be appropriate as it would raise complexity on the application of the post trade deferral regime. The proposed framework should provide the necessary post trade transparency for liquid and illiquid non-equity instruments, including sovereign bonds. <ESMA_QUESTION_144> Q145: Do you support the proposal that the deferral for non-equity instruments which do not have a liquid market should be until the end of day + 1? Please provide reasons for your answer. <ESMA_QUESTION_145> LSEG: Yes, we agree. <ESMA_QUESTION_145> Q146: Do you think that one universal deferral period is appropriate for all non-equity instruments which do not have a liquid market or that the deferrals should be set at a more granular level, depending on asset class and even sub asset class. Please provide reasons for your answer. <ESMA_QUESTION_146> LSEG: We support a uniform deferral period. <ESMA_QUESTION_146> Q147: Do you agree with the proposal that during the deferred period for non-equity instruments which do not have a liquid market, the volume of the transaction should be omitted but all the other details of individual transactions must be published? Please provide reasons for your answer. <ESMA_QUESTION_147> LSEG: No, during deferred period no data should be published - otherwise it will be confusing and disseminating the price would make deferred publication pointless. 43

44 See our response to Q141. We do not agree with the principle of "partial" publication. We believe this would be difficult to operate in practice and we would much rather have a post-trade transparency regime whereby the trade is completely dark for a period of time after which the all the relevant trade details are published at the same time.<esma_question_147> Q148: Do you agree that publication in an aggregated form with respect to sovereign debt should be authorised for an indefinite period only in limited circumstances? Please give reasons for your answers. If you disagree, what alternative approaches would you propose? <ESMA_QUESTION_148> LSEG: Yes we believe that as relatively liquid instruments, deferral of post-trade information for sovereign bonds should only take place in limited circumstances. <ESMA_QUESTION_148> Q149: In your view, which criteria and/or conditions would it be appropriate to specify as indicating there is a need to authorise extended/indefinite deferrals for sovereign debt?? <ESMA_QUESTION_149> LSEG: Alternative measures could be a substantial reduction in trading volumes (e.g. volume reductions of over 50%) or substantial increase in price volatility. Such measurements should be taken over a suitably representative time period, e.g. three months. <ESMA_QUESTION_149> Q150: In your view, could those transactions determined by other factors than the valuation of the instrument be authorised for deferred publication to the end of day? Please provide reasons for your answer. <ESMA_QUESTION_150> LSEG: As set out in our response to Question 138, we do not believe that give-up / give-in trades that are in respect of post trade processing should be published as they do not contribute to the objectives of posttrade transparency, and could lead to a misleading impression of trading activity. Similarly, a number of interest rate swap systems operate on the principle of sponsored access, under which the counterparty to the client s trade is always the sponsored access dealer (sometime called the Prime dealer). Where appropriate, an additional back-to-back trade is effected between the sponsored access dealer and the executing dealer. The only difference between the original trade and the matching back-to-back trade is that the price of the original trade includes an AVC adjustment to reflect the credit risk associated with the client. We do not believe that details of the back-to-back trade should be published, for the following reasons: As for give-up / give-in trades, it does not contribute to the objectives of post-trade transparency The only additional information contained is the effective credit quality of the client Publication could lead to a misleading impression of trading activity. <ESMA_QUESTION_150> 3.9. The transparency regime of non-equity large in scale orders and transactions Q151: Do you agree with the proposed option? Which option would be more suitable for the calibration of the large in scale requirements within an asset class? <ESMA_QUESTION_151> 44

45 LSEG: We support option 1 which, similar to the equity regime, takes into account the different levels of liquidity. This would enhance coherence with the future development of the liquid market framework envisaged for the non equity space. However, this should be calibrated at the level of granularity established by the COFIA approach to the determination of liquid/illiquid markets (see our response to Q113) and liquidity bands assigned to the resulting product/sub-product levels. This would not require the further generation of data and would provide the necessary level of granularity to remove the need to consider different options for different asset classes (see our response to Q152). <ESMA_QUESTION_151> Q152: Do you consider there are reasons for opting for different options for different asset classes? Please provide arguments. <ESMA_QUESTION_152> LSEG: We believe that the use of different options for different asset classes could be overly complex and potentially confusing. A better approach would be to align the calculations with the approach adopted for the assessment of liquidity (see our response to Q151 and Q153). <ESMA_QUESTION_152> Q153: Do you agree that the choice between the two options should be consistent with the approach adopted for the assessment of liquidity? If not, please provide arguments. <ESMA_QUESTION_153> LSEG: Yes, see our responses to Q151 and Q152. <ESMA_QUESTION_153> Q154: Do you agree with the proposed approach? If no, which indicator would you consider more appropriate for the determination of large in scale thresholds for orders and transactions? <ESMA_QUESTION_154> LSEG: We believe that, with Option 1 as the LIS framework (per Q151), the use of ADT from the liquidity definition should be used for the liquidity determination/ bands and overall trading size. AVT could be used as a proxy for the LIS thresholds (X1, X2, Y1, Y2 etc). Again we would emphasise that while this metric is referred to in MiFIR as average size of transaction, the ADT is not a measure of this. The reference to overall trading size is more relevant and less misleading. (please refer to response to Q104). <ESMA_QUESTION_154> Q155: Do you agree that the proxy used for the determining the large in scale thresholds should be the same as the one used to assess the average size of transactions in the context of the definition of liquid markets? Please provide arguments. <ESMA_QUESTION_155> LSEG: See also our response to Q151 and Q154. <ESMA_QUESTION_155> Q156: In your view, which option would be more suitable for the determination of the large in scale thresholds? Please provide arguments. <ESMA_QUESTION_156> LSEG: We believe Option 2 would be more suitable. <ESMA_QUESTION_156> 45

46 Q157: Alternatively which method would you suggest for setting the large in scale thresholds? <ESMA_QUESTION_157> TYPE YOUR TEXT HERE.<ESMA_QUESTION_157> Q158: In your view, should large in scale thresholds for orders differ from the large in scale thresholds for transactions? If yes, which thresholds should be higher: pre-trade or posttrade? Please provide reasons to support your answer. <ESMA_QUESTION_158> LSEG: No, we believe that LIS thresholds should be the same for orders and transactions. <ESMA_QUESTION_158> Q159: Do you agree that the large in scale thresholds should be computed only on the basis of transactions carried out on trading venues following the implementation of MiFID II? Please, provide reasons for the answer. <ESMA_QUESTION_159> LSEG: We would urge ESMA to adopt a consistent approach to the calculation of liquidity and LIS thresholds. If OTC trading is included in the determination of liquidity, then ESMA should assess the impact of calculating LIS thresholds for liquid shares using trading venue volumes only. Longer term, we would support a review of the impact of OTFs and whether this could lead to the computations being limited to transactions carried out on trading venues. <ESMA_QUESTION_159> Q160: Do you think that the condition for deferred publication of large in scale transactions currently applying to shares (transaction is between an investment firm that deals on own account and a client of the investment firm) is applicable to non-equity instruments? Please provide reasons for your answer. <ESMA_QUESTION_160> LSEG: No. This condition is geared to the structure of the equities markets, which is different from that of non-equities markets. Specifically, deferred publication is equally relevant for transactions between two investment firms dealing on own account. <ESMA_QUESTION_160> Q161: Do you agree that the large in scale regime should be reviewed no earlier than two years after application of MiFIR in practice? <ESMA_QUESTION_161> LSEG: In the context of the LIS regime, we believe that 12 months would be a more appropriate period to review and remedy any issues from the implementation of MiFIR. <ESMA_QUESTION_161> Size specific to the instrument Q162: Do you agree with the above description of the applicability of the size specific to the instrument? If not please provide reasons for your answer. <ESMA_QUESTION_162> LSEG: Yes, we agree. <ESMA_QUESTION_162> 46

47 Q163: Do you agree with the proposal that the size specific to the instrument should be set as a percentage of the large in scale size? Please provide reasons for you answer. <ESMA_QUESTION_163> LSEG: No. This approach is not sufficiently flexible. The approach should be more closely geared to the objectives set out in paragraph 8 (page 163) of the Discussion Paper. We believe there should also be a distinction between retail and wholesale markets. Regarding wholesale markets, for instruments where there is a substantial risk of liquidity providers not being able to hedge their risks, the size specific to the instrument should be tailored accordingly. In addition, specific consideration should be given to the relevant threshold for publication of pre-trade transparency information other than indicative prices for request-for-quote systems. We believe that the threshold for wholesale bond contracts should be established at 500,000, which is also the threshold for retail market activity in certain EU member states. <ESMA_QUESTION_163> Q164: In your view, what methodologies would be most appropriate for measuring the undue risk in order to set the size specific threshold? <ESMA_QUESTION_164> LSEG: It would be more appropriate to have a default methodology (e.g. a proportion of the large in scale size) with the possibility for particular trading venues to apply for variations, where appropriate and justifiable for a particular instrument. In addition, specific consideration should be given to the relevant threshold for publication of pre-trade transparency information other than indicative prices for request-for-quote systems. We believe that the thresholds for bond contracts should be established at 500,000, which is also the threshold for retail market activity in certain EU Member States. <ESMA_QUESTION_164> Q165: Would you suggest any other practical ways in which ESMA could take into account whether, at such sizes, liquidity providers would be able to hedge their risks? <ESMA_QUESTION_165> LSEG: See our responses to Q163 and Q164. <ESMA_QUESTION_165> Q166: Do you agree with ESMA s description of how the size specific to the instrument waiver would interact with the large in scale waiver? Please provide reasons for your answer. <ESMA_QUESTION_166> LSEG: Yes, we agree with the proposed interpretation. <ESMA_QUESTION_166> Q167: Do you agree with ESMA s description of how the size specific to the instrument deferrals would interact with the large in scale deferrals? In particular, do you agree that the deferral periods for the size specific to the instrument and the large in scale should differ and have any specific proposals on how the deferral periods should be calibrated? Please provide reasons for your answer. <ESMA_QUESTION_167> LSEG: Yes, we agree with ESMA s description of how the size specific to an instrument deferrals would interact with the LIS deferrals. <ESMA_QUESTION_167> The Trading Obligation for Derivatives 47

48 Q168: Do you agree that there should be consistent categories of derivatives contracts throughout MiFIR/EMIR? <ESMA_QUESTION_168> LSEG: Yes, we believe that it is important to keep consistency in the categories of derivatives contracts in both regulation frameworks: MIFIR and EMIR. <ESMA_QUESTION_168> Q169: Do you agree with this approach to the treatment of third countries? <ESMA_QUESTION_169> LSEG: Yes, we agree. <ESMA_QUESTION_169> Q170: Do you agree with the proposed criteria based anti-avoidance procedure? <ESMA_QUESTION_170> LSEG: Yes, we agree. <ESMA_QUESTION_170> Q171: Do you think it would be reasonable for ESMA to consult venues with regard to which classes of derivatives contracts are traded on venue? Do you think venues would be well placed to undertake this task? <ESMA_QUESTION_171> LSEG: Yes, we agree. <ESMA_QUESTION_171> Q172: The discussion in section 3.6 on the liquid market for non-equity instruments around average frequency, average size, number and type of active market participants and average size of spreads is also relevant to this chapter and we would welcome respondent s views on any differences in how the trading obligation procedure should approach the following: <ESMA_QUESTION_172> LSEG: See also our response to Q106. Spreads data is the most complex data to obtain. Further analysis needs to be done to clearly state the computation method the ESMA will require; for example, on what frequency should spreads be computed? Should it be a "snapshot" measure or a continuous, time weighted measure? Should it look at best bid / ask or should it measure spread for a notional? This would make more sense but would be computationally even more challenging given that most equity derivatives are expressed in contracts, not notional. <ESMA_QUESTION_172> Q173: Do you have a view on how ESMA should approach data gathering about a product s life cycle, and how a dynamic calibration across that life cycle might work? How frequently should ESMA revisit its assumptions? What factors might lead the reduction of the liquidity of a contract currently traded on venue? Are you able to share with ESMA any analysis related to product lifecycles? <ESMA_QUESTION_173> LSEG: The lifecycle for derivatives must take the following aspects in account: expiry week (derivatives are usually more traded during expiry weeks) time to maturity ("front" month are usually the most liquid derivatives, whereas longer term derivatives are less traded), Money/value position (for options only, ATM options are usually more liquid than deep in/out the money options) <ESMA_QUESTION_173> 48

49 Q174: Do you have any suggestions on how ESMA should consider the anticipated effects of the trading obligation on end users and on future market behaviour? <ESMA_QUESTION_174> LSEG: For vanilla, relatively standardized derivatives we feel that the current trading obligations proposed are acceptable. ESMA needs to consider that not all derivatives are fit to trade on venues especially as they become more tailored/exotic. ESMA will need to ensure that these products are nonetheless covered to ensure there is not a biased incentive to trade more complex products to evade the trading obligation. <ESMA_QUESTION_174> Q175: Do you have any other comments on our overall approach? <ESMA_QUESTION_175> TYPE YOUR TEXT HERE <ESMA_QUESTION_175> Transparency Requirements for the Members of ESCB Q176: Do you agree that the above identifies the types of operations that can be undertaken by a member of the ESCB for the purpose of monetary, foreign exchange and financial stability policy and that are within the MiFID scope? Please give reasons to support your answer. <ESMA_QUESTION_176> TYPE YOUR TEXT HERE <ESMA_QUESTION_176> Q177: What is your view about the types of transactions for which the member of the ESCB would be able to provide prior notification that the transaction is exempt? <ESMA_QUESTION_177> LSEG: Such operations are particularly relevant for sovereign bonds. We believe that members of the ESCB should generally be able to provide prior notification for such transactions, although this is a matter primarily for members of the ESCB to provide input on. We believe that members of the ESCB should be as transparent as possible in relation to trade execution. In addition, members of the ESCB should be transparent and non-exclusive in their use of trading venues for the purposes of such execution. <ESMA_QUESTION_177> Article 22, MiFIR: Providing information for the purposes of transparency and other calculations Q178: Do you have any comments on the content of requests as outlined above? <ESMA_QUESTION_178> LSEG: Venues may run multiple trading protocols. As such the data reported by the trading venues, APAs and CTPs should include the type of trading protocol (CLOB, RfQ, Click-to-Trade, OTC registration) generating trades to better understand the source of liquidity. We urge ESMA (and NCAs) to recognise the different trading systems and consider the potential impact of these (and their participants) on the calculations it is required to manage. We believe this will also improve the specification and targeting of data requests to trading venues, APAs, CTPs etc. 49

50 We would also encourage ESMA to consider the potential synergies in collecting some of this data from trading venues with other data requirements, such as reference data, in a single file. <ESMA_QUESTION_178> Q179: Do you have proposals on how NCAs could collect specific information on the number and type of market participants in a product? <ESMA_QUESTION_179> LSEG: We believe that data on the total number of participants across particular markets, asset/product classes could be collected in the first instance by NCAs from the transaction reporting information available to them. This may involve the use of additional reference data associated with the reporting entity identifier. The use of reference data in relation to financial instruments received under Article 27 MiFIR can be used to identify these markets, asset/product groups. This data can be augmented with member/participant data from individual trading venues and APAs, including details of registered market makers. <ESMA_QUESTION_179> Q180: Do you consider the frequency of data requests proposed as appropriate? <ESMA_QUESTION_180> LSEG: We agree with ESMA s proposed approach. <ESMA_QUESTION_180> Q181: How often should data be requested in respect of newly issued instruments in order to classify them correctly based on their actual liquidity? <ESMA_QUESTION_181> LSEG: As with the current framework for shares, we believe it would be helpful for ESMA to establish, for equity-like and non-equity instruments, a standard initial trading period after which the liquidity criteria would be calculated and assessed against the thresholds (for a more accurate classification). This would need to be calibrated according to the liquidity and volatility of, and number of participants in, a product/class of instrument. <ESMA_QUESTION_181> Q182: What is your view of ESMA s initial assessment of the format of data requests and do you have any proposals for making requests cost-efficient and useful for all parties involved? <ESMA_QUESTION_182> LSEG: We agree that ESMA should develop standard template / format / delivery mechanisms and use these consistently. We suggest that it is appropriate to distinguish the format/delivery requirements for regular data (e.g. reference data) or more complex data sets (e.g. order data), which are considered in Section 8 of the DP, from more periodic and bespoke requests, such as those for annual or ad hoc liquidity data for instruments. <ESMA_QUESTION_182> Q183: Do you consider a maximum period of two weeks appropriate for responding to data requests? <ESMA_QUESTION_183> LSEG: While we recognise the need for ESMA to make ad hoc and short notice data requests, we believe that there will be a number of occasions when a longer period than 2 weeks is required. Annual liquidity data requests, for example, should allow for a minimum 4 week response period and timing should also take into account any other requests that fall in the same period, to avoid conflict/allow prioritisation. <ESMA_QUESTION_183> 50

51 Q184: Do you consider a storage time for relevant data of two years appropriate? <ESMA_QUESTION_184> LSEG: Yes, we agree. <ESMA_QUESTION_184> 51

52 4. Microstructural issues 4.1. Microstructural issues: common elements for Articles 17, 48 and 49 MiFID II Q185: Is there any element that has not been considered and/or needs to be further clarified in the ESMA Guidelines that should be addressed in the RTS relating to Articles 17, 48 and 49 of MiFID II? <ESMA_QUESTION_185> In LSEG s view, the ESMA Guidelines address all relevant aspects on systems and controls in automated trading environments. We support ESMA s approach to building its MiFID II draft technical standards using this as the base. <ESMA_QUESTION_185> Q186: Do you agree with the definition of trading systems for trading venues? <ESMA_QUESTION_186> Yes, LSEG agrees. <ESMA_QUESTION_186> Q187: Do you agree that the requirements under Articles 48 and 49 of MiFID II are only relevant for continuous auction order book systems and quote-driven trading systems and not for the other systems mentioned above? <ESMA_QUESTION_187> Whilst algorithmic trading techniques are most prevalent on continuous trading systems (including, continuous order books or quote-driven systems), LSEG suggests that appropriate consideration should be given to certain hybrid and request-for-quote systems operated by trading venues which entail algorithmic trading, with appropriate proportionality. For example, those hybrid systems that require price streaming from market makers or SI quoting engines to facilitate execution necessitate forms of algorithmic trading, hence should be considered within the scope of articles 48 and 49 for a level playing field for all trading venues. <ESMA_QUESTION_187> Q188: Which hybrid systems, if any, should be considered within the scope of Articles 48 and 49, and why? <ESMA_QUESTION_188> In general, LSEG s view is that the microstructural and organisational requirements should apply to any system where algorithmic trading is used, with appropriate proportionality, to ensure a level playing field between trading venues and avoid the scope for regulatory arbitrage. This is consistent with the approach in MiFID Article 18(5): Member States shall require that an [ ] MTF or an OTF comply with Articles 48 and 49 and have in place all the necessary effective systems, procedures and arrangements to do so. Hybrid systems which execute a reasonably significant percentage of total trading in any given financial instrument should be in scope, with the appropriate proportionality. For example, RFQ systems in some markets execute more than is seen on a trading venue. In addition, those hybrid systems that require price streaming from market makers or SI quoting engines to facilitate execution necessitate forms of algorithmic trading and should be considered within the scope of articles 48 and

53 <ESMA_QUESTION_188> Q189: Do you agree with the definition of trading system for investment firms? <ESMA_QUESTION_189> LSEG: Yes, in principle. However, it seems to be more specific to firms that use high frequency (algorithmic) trading techniques, which (as MiFID defines) is only a subset of algorithmic trading techniques. We suggest a few additional considerations. Similar to (and for consistency with) the definition of trading venues, the definition of trading systems for investment firms should include its downstream systems, which is currently not mentioned (paragraph 9(iii), p.202 of the Discussion Paper). The definition of trading system should include the use of execution algorithms, where the parent order is generated by a human or unrelated system and execution is done through an algorithm which is often taking the same signals, and automated market making systems. Finally, there is a concern that investors or fund managers that simply use DEA or AOR systems for order entry, but are not involved in the design or operation of the platforms, may be subject to the same organisational requirements as the investment firms that provide this service. We suggest that such users should be regulated as users (including due diligence), but not be subject to the organisational/ systems requirements specific to investment firms providing DEA. <ESMA_QUESTION_189> Q190: Do you agree with the definition of real time in relation to market monitoring of algorithmic trading activity by investment firms? <ESMA_QUESTION_190> Yes, LSEG agrees. <ESMA_QUESTION_190> Q191: Is the requirement that real time monitoring should take place with a delay of maximum 5 seconds appropriate for the risks inherent to algorithmic trading and from an operational perspective? Should the time frame be longer or shorter? Please state your reasons. <ESMA_QUESTION_191> Yes, LSEG agrees. 5 seconds is a reasonable maximum time for an alert to be raised. It may take longer for the underlying issue to be checked, verified and then acted upon. <ESMA_QUESTION_191> Q192: Do you agree with the definition of t+1 in relation to market monitoring of algorithmic trading activity by investment firms? <ESMA_QUESTION_192> Yes, LSEG agrees. <ESMA_QUESTION_192> Q193: Do you agree with the parameters to be considered to define situations of severe market stress and disorderly trading conditions? <ESMA_QUESTION_193> LSEG agrees with the parameters to define severe market stress. We suggest this term is expanded to include situations when investment firms trading systems are affected in the same manner to trading venues systems ( i.e. where the ability of [an investment firm] to process orders and make prices available is compromised ) for a consistent approach. Please also see LSEG s response with regards to the inclusion of downstream systems in the definition of trading system for investment firms in Q

54 LSEG also agrees with the situations considered under disorderly market conditions. We note that the same conditions could also be used to describe an efficient market reacting to unexpected news or volatility, so ESMA may need to detail this further to avoid confusing these two very different situations. <ESMA_QUESTION_193> Q194: Do you agree with the aboveapproach? <ESMA_QUESTION_194> Yes, LSEG agrees with the approach suggested by ESMA in paragraph 35 and 36 on p.206 of the Discussion Paper. Given that there is no standard trading venue, i.e. all are unique, a structured selfassessment of compliance with Article 48 is a sensible approach. <ESMA_QUESTION_194> Q195: Is there any element that should be added to/removed from the periodic selfassessment? <ESMA_QUESTION_195> In terms of the self assessment, LSEG s comments are as follows: Elements to remove: 37(i)(c) Types of strategies incentivised by the venue s fee structure We would question the relevance of this element and suggest that it is removed; it is very subjective and any consequences of fee structure, such as the attraction of latency sensitive strategies, will be borne out during the assessment of subsequent elements, such as 37(ii)(b), (d), (e) and (h). 37(ii)(l) Number of countries and regions in which the trading venue is undertaking business activity It is not clear what ESMA refers to by business activity in this context, and how is it relevant to a trading venue's self-assessment. We suggest this should be removed. Elements needing further clarification: 37(ii)(a), (d), (e) Metrics related to algorithmic trading/ strategies or HFT We note that trading venues will only be able to calculate these metrics once the algorithm/strategy flagging requirement on order entry is functional, and investment firms are designated HFT because of their use of high frequency trading techniques (HFTT, as described in Section 5.1 of the Consultation Paper). With reference to ESMA's draft technical advice in relation to the identification of HFTT at a member or participant level (also see our response to Q171 of Consultation Paper), if a member/ participant is considered to be HFT in one venue then this would extend across all markets in the EU. Without communication of HFT status, trading venues will not be able to determine whether a member/ participant should be classified as HFT except in relation to activity conducted on its own markets. Similarly, where a DEA participant is undertaking HFT activity but accesses a trading venue via member/ participant that is not accorded HFT status, trading venues may be unable to calculate true percentages accordingly. Some of these elements may also be irrelevant for certain (illiquid) instruments and market models. 37(ii)(i) Number of remote members By remote members, we assume ESMA is referring to trading members who are incorporated in a different jurisdiction to the trading venue and are not operating via a branch office in the trading venue s jurisdiction. 54

55 Elements to add: In 37(iii) Complexity in terms of ESMA may consider adding an assessment of the number of ISV (independent software vendor) solutions certified by the trading venue. <ESMA_QUESTION_195> Q196: Would the MiFID II organisational requirements for investment firms undertaking algorithmic trading fit all the types of investment firms you are aware of? Please elaborate. <ESMA_QUESTION_196> Yes, LSEG agrees. In paragraph 45 of p.209 of the Discussion Paper, ESMA states that its proposal aims to apply minimum standards to investment firms whose trading systems are indirectly linked to algorithmic trading. This needs further definition; ESMA should provide clarity on where it expects investment firms to be subject to the organisational requirements detailed under Article 17 where the firm's trading systems are not directly linked to algorithmic trading. <ESMA_QUESTION_196> Q197: Do you agree with the approach described above regarding the application of the proportionality principle by investment firms? Please elaborate. <ESMA_QUESTION_197> LSEG: Yes, the application of the proportionality principle seems sensible in this respect. However, the detail and frequency of the self-assessment process outlined by ESMA may be disproportionate to infrequent or low volume users of trading algorithms. A self-assessment on an annual rather than semi-annual basis would probably be more appropriate, given that changes in organisational structure are not very frequent. Ad-hoc changes could be notified to the NCA on an ad-hoc basis. In paragraph 45 of p.209 of the Discussion Paper, ESMA states that its proposal aims to apply minimum standards to investment firms whose trading systems are indirectly linked to algorithmic trading. This needs further definition ESMA should provide clarity on where it expects investment firms to be subject to the organisational requirements detailed under Article 17 where the firm's trading systems are not directly linked to algorithmic trading. <ESMA_QUESTION_197> Q198: Are there any additional elements that for the purpose of clarity should be added to/removed from the non-exhaustive list contained in the RTS? Please elaborate. <ESMA_QUESTION_198> LSEG: As noted in our response to Q197, the list is extensive; and possibly contains more detail than required especially for infrequent or low volume users of algorithms. Some aspects requested are very subjective, and open to numerous interpretations, while others may not be relevant to the purpose here. Some brief examples: 51(i)(c): level of automation of trading and other processes or activities of the firm Some firms consider the use of data editing and computational software to record/value execution as automation, while others will consider nothing less than STP (straight through processing) as automation. This would need to be limited to broad categories in relation to the trading system to be useful: highly sensitive, moderately or not sensitive; otherwise, we suggest it is too subjective. 51(i)(f): latency sensitivity of the firm s strategies and trading activities 55

56 All trading activities are latency sensitive a retail investor who wants to buy a stock now does not expect the order to be executed in 20 minutes time so the answers here will again be subjective to the respondent s point of view. 51(i)(m) the maturity of the firm and level of experience and competency of its personnel (i.e. whether it is a start-up or incumbent). The relevance of this element is unclear. Experience suggests that sometimes systems issues may appear in more mature firms with legacy systems than less mature firms. Further, the scale elements under 51(ii) do not appear to include an assessment of how much of a firm's trading activity is the result of algorithms versus non-algorithmic trading. We suggest questions should, wherever possible, be phrased to require quantitative answers or answers which require some form of substantiation or examples. By doing so, the questions will become more relevant and the number of questions required will be reduced, which would be to the benefit of both the respondents and regulators who will be reviewing the answers. <ESMA_QUESTION_198> 4.2. Organisational requirements for investment firms (Article 17 MiFID II) Q199: Do you agree with a restricted deployment of algorithms in a live environment? Please elaborate <ESMA_QUESTION_199> Yes, LSEG agrees. Trading venues usually have rules requiring members to test their systems or controls prior to submitting orders, quotes or trade reports to the trading system. For e.g., the London Stock Exchange offers members a separate connection to a testing environment packaged within a number of services such as the Customer Development Service, Application Certification, High Volume Testing and Participant Test Weekends, in addition to the use of specific test segments on the trading system. LSEG suggests that testing on the live production environment of the trading system should be prohibited (especially during trading hours; exceptionally it may be allowed on certain planned weekends or public holidays) as it has the potential to impact the market, particularly as testing may result in unusually priced and/or sized orders, quotes or trade reports being entered. LSEG also notes that it agrees with ESMA s view in paragraph 9 on p.214 of the Discussion Paper where it states that trading venues do not have to and should not have to provide any ex-ante sign-off or authorisation of algorithms. This is a sensible approach venues should be required to provide the necessary testing environments, but should not be policing individual algorithms. Certification on successful conformance testing can be provided. Finally, LSEG notes that in paragraph 8, ESMA requires a DEA client to use a trading venue s testing environments or require the DEA provider to conduct tests on its behalf. We suggest direct testing may be sensible for Sponsored Access, and delegation to the DEA provider for DMA. <ESMA_QUESTION_199> Q200: Do you agree with the parameters outlined for initial restriction? Please elaborate. <ESMA_QUESTION_200> Yes, if by initial restriction ESMA is referring to the initial testing of algorithms. LSEG notes that non-live testing may be done on both an initial and on-going (post deployment) basis. <ESMA_QUESTION_200> 56

57 Q201: Do you agree with the proposed testing scenarios outlined above? Would you propose any alternative or additional testing scenarios? Please elaborate. <ESMA_QUESTION_201> Yes, LSEG agrees. <ESMA_QUESTION_201> Q202: Do you agree with ESMA s approach regarding the conditions under which investment firms should make use of non-live trading venue testing environments? Please elaborate. <ESMA_QUESTION_202> Yes, LSEG agrees with ESMA, and we provide some additional comments below. Trading venues predominantly provide non-live trading environments for users to test, from a technical perspective, against the functional interfaces that they will be connecting to in order to access the live trading environments, as well as performing conformance and certification tests that validate from an application perspective that there is not expected to be any technical issues (either user specific or market wide) as a result of that user's technical connectivity. Testing other scenarios, such as trading strategies and algorithms against, for example, historical trading data, is something that trading venues could provide on a fair commercial basis, but is currently seen as being added value to what is core testing requirements. We recognise that the direction of MiFID II is for such testing to be part of minimum requirements for investment firms. <ESMA_QUESTION_202> Q203: Do you consider that ESMA should specify more in detail what should be the minimum functionality or the types of testing that should be carried out in non-live trading venue testing environments, and if so, which? <ESMA_QUESTION_203> LSEG suggests that ESMA should set a general framework for different market scenarios to be tested (e.g. high volumes, high message rates, trading gateway failures etc.), but individual markets should be responsible to ensure appropriate detailed scenarios suitable to their market model and unique functionalities or technology. <ESMA_QUESTION_203> Q204: Do you consider that the requirements around change management are appropriately laid down, especially with regard to testing? Please elaborate. <ESMA_QUESTION_204> LSEG: Change management should be done on an ad-hoc basis. Investment firms should prepare a Release Note for any production releases with the nature of change description, the impact analysis of the change and the related test check list. These should be kept and made available to the NCA if required. <ESMA_QUESTION_204> Q205: Do you agree with the proposed monitoring and review approach? Is a twice yearly review, as a minimum, appropriate? <ESMA_QUESTION_205> LSEG: Yes, this approach is appropriate. Although this is an issue which concerns investment firms, we suggest that an annual review should be enough if regression testing of algorithms is performed. LSEG also notes that paragraph 36 on p.219 of the Discussion Paper suggests that firms should ensure that the staff involved in supporting electronic trading operations should have the necessary authorisations with the relevant trading venues. In our view, whilst trading venues ask members for relevant contacts (i.e. the responsible managers/ traders) with the appropriate level of technical experiences and notify when 57

58 there are any changes, venues do not certify or authorise investment firm staff. This should be the responsibility of the NCA. <ESMA_QUESTION_205> Q206: To what extent do you agree with the usage of drop copies in the context of monitoring? Which sources of drop copies would be most important? <ESMA_QUESTION_206> Yes, LSEG agrees with the usage of drop copies in the context of monitoring. A trading member drop copy will hold additional information not contained in the trading venue drop copy. For this reason we suggest a member s drop copy is most important. <ESMA_QUESTION_206> Q207: Do you agree with the proposed approach? <ESMA_QUESTION_207> Yes, LSEG agrees. <ESMA_QUESTION_207> Q208: Is the proposed list of pre trade controls adequate? Are there any you would add to or remove from the list? <ESMA_QUESTION_208> LSEG: Yes, the list seems adequate, although investment firms would be most appropriate to comment. It may be prudent for investment firms to also have a control and/or alert on the multiple entry of orders in same instrument at the maximum order value in a short space of time; as this can be more damaging than a single large order of the same size. In addition, a trading venue would also check with the investment firm if it has the required permissions to trade the relevant financial instruments. <ESMA_QUESTION_208> Q209: To what extent do you consider it appropriate to request having all the pre-trade controls in place? In which cases would it not be appropriate? Please elaborate. <ESMA_QUESTION_209> In general, LSEG notes that it is the responsibility of the investment firm to assess the specific pre-trade risk controls and the parameters suitable to its risk profile. The principle of proportionality should apply here. A trading venue usually has in its rules a requirement that member/ participant shall, at all times, have sufficient systems, procedures and controls designed to prevent the entry of erroneous orders and quotes to the trading system <ESMA_QUESTION_209> Q210: Do you agree with the record keeping approach outlined above? <ESMA_QUESTION_210> Yes, LSEG agrees. <ESMA_QUESTION_210> Q211: In particular, what are your views regarding the storage of the parameters used to calibrate the trading algorithms and the market data messages on which the algorithm s decision is based? <ESMA_QUESTION_211> TYPE YOUR TEXT HERE <ESMA_QUESTION_211> 58

59 Q212: Do you consider that the requirements regarding the scope, capabilities, and flexibility of the monitoring system are appropriate? <ESMA_QUESTION_212> TYPE YOUR TEXT HERE <ESMA_QUESTION_212> Q213: Trade reconciliation should a more prescriptive deadline be set for reconciling trade and account information? <ESMA_QUESTION_213> TYPE YOUR TEXT HERE <ESMA_QUESTION_213> Q214: Periodic reviews would a minimum requirement of undertaking reviews on a halfyearly basis seem reasonable for investment firms engaged in algorithmic trading activity, and if not, what would be an appropriate minimum interval for undertaking such reviews? Should a more prescriptive rule be set as to when more frequent reviews need be taken? <ESMA_QUESTION_214> LSEG suggests an annual review should be appropriate, subject to a proportionality test. <ESMA_QUESTION_214> Q215: Are there any elements that have not been considered and / or need to be further clarified here? <ESMA_QUESTION_215> LSEG: No; in our view the list seems appropriate. LSEG notes here that in our experience, appropriate awareness by a trader when entering orders or a programmer when writing code is the most important criteria, as automated code can cause significant market impact over a short period of time before human intervention to prevent any unintentional consequences. We feel this is appropriately reflected in criteria 92(iv) and (v). LSEG is conscious that the sharing of source code between clients and investment firms who provide DEA, suggested by ESMA in paragraph 95 on p.231 of the Discussion Paper, is a sensitive issue and should be considered further. It may not be reasonable to expect a client to share its proprietary source code with an investment firm which may be in many certain instances a competitor as well as a supplier. The ESMA proposal makes no reference to "sub-delegation" of DEA access as per the Organisational Requirements for Investment Firms and the Proportionality Principle (refer to paragraph 51(1)(i) on p.210 of the Discussion Paper). In ESMA's proposal, it may be helpful to include additional guidance that if the DEA provider intends to allow sub-delegation of DEA, it should put in place appropriate due diligence arrangements so that it can assess how sub-delegation will be managed and mitigate any associated risks. Please also see our response to Q254. <ESMA_QUESTION_215> Q216: What is your opinion of the elements that the DEA provider should take into account when performing the due diligence assessment? In your opinion, should any elements be added or removed? If so, which? <ESMA_QUESTION_216> LSEG: In our view, the list is appropriate. We stress that DEA providers should ensure that DEA users are familiar with the rules of the trading venue to which they are connecting this requirement appears to be covered by paragraph 105(iii) on p.233 of the Discussion Paper. 59

60 <ESMA_QUESTION_216> Q217: Do you agree that for assessing the adequacy of the systems and controls of a prospective DEA user, the DEA provider should use the systems and controls requirements applied by trading venues for members as a benchmark? <ESMA_QUESTION_217> LSEG: Yes; adopting the requirements applied by trading venues for members as a benchmark appears to be a suitable minimum basis for DEA user arrangements in relation to pre-trade controls. LSEG also suggests that DEA users should be required to have dedicated user IDs to connect their algorithms to trading venues so that their activity can be monitored and participant level controls can be applied by the trading venues. Our view is that DEA users systems should also pass the DEA provider s acceptance test before connecting to a trading venue. <ESMA_QUESTION_217> Q218: Do you agree that a long term prior relationship (in other areas of service than DEA) between the investment firm and a client facilitates the due diligence process for providing DEA and, thus, additional precautions and diligence are needed when allowing a new client (to whom the investment firm has never provided any other services previously) to use DEA? If yes, to what extent does a long term relationship between the investment firm and a client facilitate the due diligence process of the DEA provider? Please elaborate. <ESMA_QUESTION_218> Whilst we cannot comment from the perspective of a DEA provider, in general LSEG suggests that a prior relationship clearly helps but this should not prevent a provider to take on new DEA clients, so long as the acceptance criteria and process is clear, documented and applies in a non-discriminatory and case-bycase manner. <ESMA_QUESTION_218> Q219: Do you agree with the above approach? Please elaborate. <ESMA_QUESTION_219> Yes, LSEG agrees. The approach is comprehensive. It may be helpful for ESMA to clarify paragraph 105(iii) such that in the event that the DEA provider allows sub-delegation of DEA access, it should have appropriate arrangements in place to ensure that that any sub-delegated user (rather than just employee of the DEA user) has adequate training on order entry procedures etc., subject to the appropriate proportionality. <ESMA_QUESTION_219> Q220: Do you agree with the above approach, specifically with regard to the granular identification of DEA user order flow as separate from the firm s other order flow? Please elaborate. <ESMA_QUESTION_220> Yes, LSEG agrees. DEA users should be required to have dedicated user IDs to connect algorithms to trading venues so that their activity can be monitored and participant level controls can be applied by the trading venues. <ESMA_QUESTION_220> Q221: Are there any criteria other than those listed above against which clearing firms should be assessing their potential clients? <ESMA_QUESTION_221> In answering the questions in this section, LSEG provides its experience of operating derivatives markets (including LSE Derivatives, IDEM, IDEX, AGREX and MTS Swaps) and EMIR authorised CCPs (LCH.Clearnet Ltd, LCH.Clearnet SA and CC&G). 60

61 There are some areas where clarification or addition would be helpful. For example: 1. Collateral provided to the clearing firm (paragraph 123(vi)) This can be expanded to ensure either that the client can post collateral eligible at CCP or that the GCM is able to substitute for the client. From the perspective of a CCP, we prefer the first option given that the potential for delays in or failure of the substitution process. 2. Indirect clients If a client has its own (indirect) clients, it should ensure that it has appropriate controls/agreements in place equivalent to those of a GCM - i.e. that the client applies the same criteria listed here and in paragraph 123 to the indirect clients. 3. Trading models GCMs should assess clients' trading models/systems, particularly for high frequency traders. 4. Business continuity GCMs should assess clients' back-up and disaster recovery provisions. 5. Organisational requirements GCMs should ensure clients enforce appropriate segregation of duties. <ESMA_QUESTION_221> Q222: Should clearing firms disclose their criteria (some or all of them) in order to help potential clients to assess their ability to become clients of clearing firms (either publicly or on request from prospective clients)? <ESMA_QUESTION_222> LSEG suggests that disclosure would aid the requirement in paragraph 125 to be transparent, but this should be at the discretion of the GCM. If GCMs disclose some of the membership criteria, then potential clients can assess their suitability to access the services of that clearing member. However, GCMs and non clearing members (i.e. direct clearing members, but with no clients) should not compete on risk, therefore certain elements of criteria that may relate to the assessment of risk (such as margin algorithms), should be disclosed at the GCM s discretion. This practice is common for CCPs as well; a CCP also discloses some of its membership criteria to clearing members, including balance sheet requirements, credit worthiness, technical capabilities and ownership and governance disclosure; however, transparency of margin algorithms are publicised at the CCP discretion. <ESMA_QUESTION_222> Q223: How often should clearing firms review their clients ongoing performance against these criteria? <ESMA_QUESTION_223> In general, LSEG suggests that reviews should be decided by the outcome of a risk assessment (i.e. credit rating, type of business cleared, nature of client, volume of business cleared) and at least annual. The criteria and the nature of the client for example will impact the frequency of the review. For example, collateral reviews will be less frequent if the client does not trade actively. Many of the criteria relate to the business as usual relationship between GCM and client and should be monitored continuously - e.g. trade processing, payment/settlement, collateral. An annual review should be made of criteria such as legal agreements, business strategy, financials, etc. Proportionality should apply, and triggers should be in place to prompt immediate reviews, e.g. change in trading behaviour, failed payment, rating change, restructuring, etc. 61

62 <ESMA_QUESTION_223> Q224: Should clearing firms have any arrangement(s) other than position limits and margins to limit their risk exposure to clients (counterparty, liquidity, operational and any other risks)? For example, should clearing firms stress-test clients positions that could pose material risk to the clearing firms, test their own ability to meet initial margin and variation margin requirements, test their own ability to liquidate their clients positions in an orderly manner and estimate the cost of the liquidation, test their own credit lines? <ESMA_QUESTION_224> Yes, and LSEG agrees with the examples quoted. GCMs should stress test their clients' positions, and test their own liquidity requirements and liquidation processes in a client default event. Further controls should include limits on concentrations in margin collateral and wrong way risk restrictions. <ESMA_QUESTION_224> Q225: How regularly should clearing firms monitor their clients compliance with such limits and margin requirements (e.g. intra-day, overnight) and any other tests, as applicable? <ESMA_QUESTION_225> LSEG: Limit monitoring should ideally be real time or at least daily, depending on the size and nature of the activity of the client. The capacity to monitor intraday margin requirements would depend upon the sophistication of the GCM's system; if GCMs are unable to monitor margins intraday, the GCM should require all clients to over-collateralise their exposures at the end of each day, or at least have standby unconditional guarantees in place. <ESMA_QUESTION_225> Q226: Should clearing firms have a real-time view on their clients positions? <ESMA_QUESTION_226> Yes, LSEG agrees. <ESMA_QUESTION_226> Q227: How should clearing firms manage their risks in relation to orders from managers on behalf of multiple clients for execution as a block and post-trade allocation to individual accounts for clearing? <ESMA_QUESTION_227> LSEG: From a CCP s perspective, clearing firms should ensure the sufficiency of margin from each client at all times, regardless of any subsequent post trade allocation to sub accounts. Any netting arrangements must be legally enforceable in all relevant jurisdictions. <ESMA_QUESTION_227> Q228: Which type(s) of automated systems would enable clearing members to monitor their risks (including clients compliance with limits)? Which criteria should apply to any such automated systems (e.g. should they enable clearing firms to screen clients orders for compliance with the relevant limits etc.)? <ESMA_QUESTION_228> LSEG: CCPs will typically require their clearing members to have and maintain systems enabling them to process trades efficiently and monitor limits and risks generated by each client. Pre-trade limit checks are appropriate particularly for clients using automated order routing systems. A GCM should retain power to kill the order routing process. System should enable real-time post trade limit monitoring within the GCM's infrastructure. A GCM should request limits to a client s trading (notional value, securities traded etc.) and be able to receive warnings when limits are near. The same requirements should apply to clients of GCMs and DEA clients of clearing firms. 62

63 Further, GCMs may wish to utilise trading venue Drop Copy feeds in order to maintain a real time view of client positions and trades, in order to monitor the trading behaviour and assess the principal risk that they have against the CCP. This should enable GCMs to manage appropriate margin requirements from the trading client without adding additional complexity to the trading firm's pre-trade control arrangements. <ESMA_QUESTION_228> 4.3. Organisational requirements for trading venues (Article 48 MiFID II) Q229: Do you agree with requiring trading venues to perform due diligence on all types of entities willing to become members/participants of a trading venue which permits algorithmic trading through its systems? <ESMA_QUESTION_229> Yes, LSEG agrees. This requirement should apply on an equal basis to all trading venues. Best practice indicates that due diligence should be performed on all applicants, not just those carrying out algorithmic trading. A trading venue usually provides for a precise set of admission requirements to be met by market participants during "on boarding" phase and then on an on-going basis. However, the performance of specific due diligence in relation to automated trading on at least a yearly basis may place an excessive burden on firms, particularly those which are members of multiple venues. It is our view that trading venues should adopt a risk based approach to determining the frequency of ongoing due diligence, recognising the scale and potential impact of trading undertaken by a member, the time elapsed since the member's last review and other pertinent factors. <ESMA_QUESTION_229> Q230: Do you agree with the list of minimum requirements that in all cases trading venues should assess prior to granting and while maintaining membership? Should the requirements for entities not authorised as credit institutions or not registered as investment firms be more stringent than for those who are qualified as such? <ESMA_QUESTION_230> LSEG: Most of the requirements are appropriate. However, the extension of member due diligence to cover organisational arrangements appears to duplicate the work undertaken by the firm's NCA. Our view is that trading venues should not seek to displace the NCA in the supervision of its members. Rather, the main focus of due diligence by a trading platform in relation to automated trading should be on arrangements for orderly behaviour including pre-trade controls, technical connectivity, resiliency and reporting obligations a member has to the trading venue. So we suggest that whilst a trading venue s rules should require that its members have appropriate organisational arrangements as per items ii [staff selection policy and training practice], iii [responsible managers]., vii [business continuity], and viii [outsourcing, as appropriate] in paragraph 5 on p.241 of the Discussion Paper, the actual requirements should be subject to NCA review or where a member is not a investment firm or credit institution, it should be subject to equivalent third country provisions. We suggest that ESMA should clarify this when drafting its technical standards. With regards to 5(v) testing of algorithms, as in our response to Q199, we note that we agree with ESMA s view in paragraph 9 on p.214 of the Discussion Paper where it states that trading venues do not have to, and should not have to, provide any ex-ante sign-off or authorisation of algorithms. This is a sensible approach venues should be required to provide the necessary testing environments, but should not be policing or certifying individual algorithms. 63

64 In our view there is no need to provide for stricter requirements for non investment firms or non credit institutions. This will usually relate only to third country firms, who (as of today) are considered as sufficiently regulated in respect of capital adequacy, and fitness and probity, and will continue to be considered as such if the third country firm is equivalent to the local Member State s rules or to EU law. <ESMA_QUESTION_230> Q231: If you agree that non-investment firms and non-credit institutions should be subject to more stringent requirements to become member or participants, which type of additional information should they provide to trading venues? <ESMA_QUESTION_231> LSEG does not agree that non-investment firms and non-credit institutions should be subject to more stringent requirements as ESMA's proposal covers the key factors for consideration in relation to automated trading, so long as the firm is considered as being sufficiently regulated in respect of capital adequacy, and fitness and probity as mentioned in our response to Q250. <ESMA_QUESTION_231> Q232: Do you agree with the list of parameters to be monitored in real time by trading venues? Would you add/delete/redefine any of them? In particular, are there any trading models permitting algorithmic trading through their systems for which that list would be inadequate? Please elaborate. <ESMA_QUESTION_232> Yes, LSEG agrees. <ESMA_QUESTION_232> Q233: Regarding the periodic review of the systems, is there any element that has not been considered and/or needs to be further clarified in the ESMA Guidelines that should be included? <ESMA_QUESTION_233> LSEG: In our view, all relevant elements are considered in the list proposed by ESMA. However, LSEG disagrees with the requirement in paragraph 13(i) that trading venues should calculate the median lifetime of orders of modified or cancelled. For details, see our response to Q167 of the Consultation Paper on High Frequency Trading Techniques (HFTT), to which this metric is linked. In summary, LSEG s view is that this metric is not suitable to define HFTT and has a number of flaws, most notably that it does not consider any aggressive HFTT strategies and is based on the activity of other trading members and the market model of the trading venue, rather than real/ absolute parameters (it is not an independent assessment of the techniques used by a trading firm). It may also be complicated to calculate and does not convey any meaningful information, so should be removed from the list. <ESMA_QUESTION_233> Q234: Do you agree with the above approach? <ESMA_QUESTION_234> Yes, LSEG agrees with the approach. Please also note our response to Q236 in this regard. LSEG asks ESMA to clarify what it means when it refers to duplicated hardware components to allow for business continuity in paragraph 21 on p.248 of the Discussion Paper. We assume that ESMA is suggesting that business continuity infrastructure is not used in a venue s assessment of its capacity. We also assume that ESMA means to say that there should be no single point-of-failure between the primary and 64

65 secondary trading environments (in terms of shared hardware). We would welcome further detail from ESMA in this regard. Further, the last two sentences in paragraph 23(ii) seem to imply that ESMA will require trading venues to inform their NCAs immediately of any changes in capacity, and NCAs will have final discretion in the matter of (any) capacity changes. It is LSEG s view that this requirement is too onerous venues regularly make changes to their infrastructure and hardware, and requiring such changes to be communicated (and in some instances, approved) is not required. LSEG notes here that in order to meet this requirement, investment firms will also need to have in place sufficient capacity in relation to their technical connectivity and downstream systems. <ESMA_QUESTION_234> Q235: Do you think ESMA should determine minimum standards in terms of latency or is it preferable to consider as a benchmark of performance the principle no order lost, no transaction lost? <ESMA_QUESTION_235> LSEG prefers ESMA to consider no order lost, no transaction lost as the benchmark of performance. <ESMA_QUESTION_235> Q236: Do you agree with requiring trading venues to be able to accommodate at least twice the historical peak of messages? <ESMA_QUESTION_236> In principle LSEG agrees, and this is our current approach to our trading systems (including our downstream functionalities). In our experience, trading venues operate with sufficient capacity to ensure that they can deal with periods of significant market activity and volatility. The purpose of the headroom is to deal with any outlier or anomalous event. The actual sizing criteria should also take into account whether the venue adopts flow mechanisms which do not affect the market structure and whether historical peaks are isolated or occurred a long time ago. In case an anomalous event occurs but there is still sufficient spare capacity, we suggest there should be some tolerance to judge if this event is truly a new outlier or an isolated incident. In particular, the venue should have the discretion to observe trading activity over a reasonable reference period to judge whether rescaled capacity is required to deal with increased activity. If activity returns to normal in a short time frame, an increase in capacity may not be required. However if the increase is systematic, then an upgrade to the new benchmark is merited within an appropriate time scale. Frequent capacity increases for isolated incidents that do not threaten the stability or resilience of the trading venue would be an inefficient solution. <ESMA_QUESTION_236> Q237: Do you agree with the list of abilities that trading venues should have to ensure the resilience of the market? <ESMA_QUESTION_237> Yes, LSEG agrees. <ESMA_QUESTION_237> Q238: Do you agree with the publication of the general framework by the trading venues? Where would it be necessary to have more/less granularity? <ESMA_QUESTION_238> 65

66 Yes, LSEG agrees that a general framework should be published. In some cases, trading venues may not specify the value of certain parameters specific to a member/participant or market (e.g. message limits) to avoid market manipulation. <ESMA_QUESTION_238> Q239: Which in your opinion is the degree of discretion that trading venues should have when deciding to cancel, vary or correct orders and transactions? <ESMA_QUESTION_239> LSEG suggests that orders and transactions should be dealt with separately, since the circumstances in which they are cancelled are usually different. Transactions Cancelling of transactions should be a rare occurrence performed in exceptional circumstances. Wherever possible, venues should seek approval from both counterparties. Trading venues should have rules regarding cancellations/variations/corrections of transactions, but because of the case by case nature of each circumstance, venues should publish the general framework for cancellation/variations/corrections procedures (either triggered by erroneous input of price/size or due to system failures), but should retain ultimate discretion in order to guarantee orderly and fair management of the market. Orders A trading venue should be able to cancel orders when instructed by the investment firm; this is usually in a situation where a firm is unable to connect to the system to modify or cancel their orders, or when their trading system is entering clearly erroneous orders. For some scenarios, trading venues trading venues have published procedures to cancel orders without instruction of investment firm (e.g. on corporate actions, moving a security from one segment to another, currency changes etc.). It should also be noted a venue s functionality may alter some attributes of the original order from those entered by the user (when a user chooses to do so) in very limited, prescribed circumstances (e.g. a change from continuous trading to auction/post-close trading session), so that its orders can execute from a technical perspective. We note here that trading venues have no discretion on how the orders attributes are changed (it is programmed in advance). The identical functionality is offered to all members on a non-discriminatory basis. The attributes are changed to a quantity (e.g. price) which is known publicly and in advance of the change (e.g. the closing price). For example, on Borsa Italiana and London Stock Exchange, the price of an order is restated in the event that it is entered in advance specifically for the Closing Price Crossing ( CPX ) Session (this is a session where orders execute ONLY at the closing price), if the order and has a price better than the closing price but still wants it to execute in the session. For example, if the closing price to buy an instrument is 14.85, but a user s order is at a better price (e.g ), the user may use the functionality where LSE/BIt restates the price of its order at the lower closing price of so it can execute in the CPX session. <ESMA_QUESTION_239> Q240: Do you agree with the above principles for halting or constraining trading? <ESMA_QUESTION_240> Yes, LSEG agrees with ESMA. <ESMA_QUESTION_240> Q241: Do you agree that trading venues should make the operating mode of their trading halts public? 66

67 <ESMA_QUESTION_241> Yes, LSEG agrees. <ESMA_QUESTION_241> Q242: Should trading venues also make the actual thresholds in place public? In your view, would this publication offer market participants the necessary predictability and certainty, or would it entail risks? Please elaborate. <ESMA_QUESTION_242> LSEG: Analysis from our own markets suggests that best approach would be to give flexibility to the trading venue. Some venues publish actual thresholds; others prefer that some parameters remain unpublished to prevent the risk of market manipulation or gaming. LSEG does not believe there should be a standardised approach as to the degree of information provided in this matter, as long as participants are clearly aware of the operating model and broad parameters. <ESMA_QUESTION_242> Q243: Do you agree with the proposal above? <ESMA_QUESTION_243> Yes, LSEG agrees. We note some additional considerations below. A trading venue s conformance and/or non-live testing environment should not be an exact mirror of the live production environment. There is typically a different market structure on testing environments to allow for fuller testing, and whilst it is functionally equivalent it may not have the entire sets of instruments or market segments. Another factor to consider is that a venue s trading software releases may be ahead of that in the production environment for some part of the year. Additionally, there may be clients who are neither taking market data nor order-book replay and recovery services. Further, in paragraph 39(ii) on p.251 of the Discussion Paper, it is not clear how often would it be expected that the list of instruments in production is synchronised with that of the testing environment. There are several reasons why the market structures are not aligned and it is not possible to maintain an identical copy of instrument data. Additionally there are numerous test instruments which are only for use in the test environment. It is reasonable to expect the instruments to reflect production in general terms, but unrealistic to propose an identical match. We suggest a clarification from ESMA on this basis. In our view, it would be substantial administrative requirement for venues to maintain an exact replica of the live environment, and we suggest it would be disproportionate for the purpose, and counter intuitive to the type of testing that is performed. Finally, paragraph 39(iii) suggests that trading venues must have a self-certification front end for member testing of market scenarios. Whilst we agree with ESMA s view in paragraph 9 on p.214 of the Discussion Paper where it states that trading venues do not have to and should not have to provide any ex-ante signoff or authorisation of algorithms, we do not think there should be a prescribed approach to certification. This could be either through a front-end, or on member-by-member basis, and venues should have the flexibility to choose the most suitable option that meets the regulatory end of adequate and complete testing. <ESMA_QUESTION_243> Q244: Should trading venues have the ability to impose the process, content and timing of conformance tests? If yes, should they charge for this service separately? <ESMA_QUESTION_244> LSEG: Yes, they should have the ability to impose process, content and timing of conformance tests, if required, to ensure that all members (and their clients) reach a minimum benchmark. It should be decided by the trading venue whether this is charged for separately. <ESMA_QUESTION_244> 67

68 Q245: Should alternative means of conformance testing be permitted? <ESMA_QUESTION_245> Yes, LSEG agrees; whilst clients should usually conform only on the test environment of the trading venue itself, if approved by the trading venue, there should be flexibility to allow third party support within the conformance regime. <ESMA_QUESTION_245> Q246: Could alternative means of testing substitute testing scenarios provided by trading venues to avoid disorderly trading conditions? Do you consider that a certificate from an external IT audit would be also sufficient for these purposes? <ESMA_QUESTION_246> Yes, LSEG agrees. External providers could provide alternative environment and/or simulators, should they use a functional mirror of the trading venue s system, as approved by the venue. The use of an external IT audit may in theory provide a level of comfort to NCAs, but standards may vary on this and risk management may not be as robust as a common minimum standard set by the trading venue. Whilst there may be products offered by external IT suppliers which allow clients to test disorderly trading conditions (as approved by the trading venue), for the purpose of testing core functionality and conformance we suggest that clients should usually conform on the test environment of the trading venue itself, unless third party support is approved by the trading venue. <ESMA_QUESTION_246> Q247: What are the minimum capabilities that testing environments should meet to avoid disorderly trading conditions? <ESMA_QUESTION_247> LSEG: Testing has traditionally been used for certification purposes, as this is its core function. There are several 'scenario' based features available from trading venues, such as testing peak message flows and service disruptions. However, trading venues are in a good position to introduce more functionality and tools which allow members to test against 'disorderly' market conditions. Testing environments for disorderly conditions should replicate real market scenarios as closely as possibly. From this perspective, members should be allowed to receive real market feeds and enter orders on trading venues test environment or internal market simulators. The minimum standards that should be met should be specified by trading venues, and include specifications on certification, conformance, and failovers test scenarios. The test environments should support similar functionalities, protocols and structure (customer facing structure) as the live environments, but should not be expected to identically mirror the production environment s performance or capacities. Finally, we note whilst test environments may be a copy of the real market structure, it may not be possible to exactly reproduce a usual trading day or identically replicate real life disorderly trading conditions (because there will be fewer participants active at any particular point of time, and as firms are testing and not actually trading for commercial purposes, it is unrealistic to say that a test environment closely simulates a real market s message flows). We reiterate, therefore, that the results of the test should not prejudice the investment firm s final responsibility towards the appropriateness and fitness of its algorithm. As these services are seen as value-add to the current set of core testing requirements, it may be suitable for trading venues to offer this environment on a fair and commercial basis. <ESMA_QUESTION_247> Q248: Do you agree with the proposed approach? <ESMA_QUESTION_248> 68

69 LSEG notes that the purpose of this list is to require proper pre-trade controls at trading venues to ensure resilience of the trading system. For this regulatory objective to be fulfilled, we believe that controls around price, size, value, kill functionality and the number of orders are a sufficient set. We note here that, the number of orders is usually already controlled at an aggregate level keeping in mind overall capacity, message limits and throttling mechanisms. Other controls, e.g. maximum long/short instrument or overall strategy positions are internal considerations for an investment firm and depend on its risk appetite and on information available only to the firm and its DEA client they cannot be known to the trading venue, and should be removed from this list. In addition, a market impact assessment control is too subject, and in our view, not a relevant consideration when considering systems resilience; we suggest it also is removed. <ESMA_QUESTION_248> Q249: In particular, should trading venues require any other pre-trade controls? <ESMA_QUESTION_249> LSEG: No. Please also refer to our response to Q248. <ESMA_QUESTION_249> Q250: Do you agree that for the purposes of Article 48(5) the relevant market in terms of liquidity should be determined according to the approach described above? If, not, please state your reasons. <ESMA_QUESTION_250> LSEG: Yes, we agree. Please also see our response to the reference price waiver in Q61 of the Discussion Paper. <ESMA_QUESTION_250> Q251: Are there any other markets that should be considered material in terms of liquidity for a particular instrument? Please elaborate. <ESMA_QUESTION_251> LSEG: No. <ESMA_QUESTION_251> Q252: Which of the above mentioned approaches is the most adequate to fulfil the goals of Article 48? Please elaborate <ESMA_QUESTION_252> LSEG suggests that Option A should be the de minimis standard across all trading venues. If venues so choose, they may also go for a stricter regime based on Option B, although this may duplicate some of the activities and assessments performed by the NCA. In practice, venues may choose a combination of Option A and B for different types of DEA. For instance, trading venues may wish to review every application for Sponsored Access (i.e. Option B), as the client would maintain a direct connection to the venue's trading system. This would also enable the trading venue to ensure inter alia that the venue did not contravene any local conditions relating to direct trading connections in the jurisdiction of the Sponsored Access client. Trading venues may wish to maintain a general framework approach for other forms of DEA access which route orders via a member's trading systems, e.g. DMA (i.e. Option A). <ESMA_QUESTION_252> Q253: Do you envisage any other approach to this matter? 69

70 <ESMA_QUESTION_253> LSEG: Please see our response to Q252. <ESMA_QUESTION_253> Q254: Do you agree with the list of elements that should be published by trading venues to permit the provision of DEA to its members or participants? <ESMA_QUESTION_254> Yes, LSEG agrees. We suggest a further consideration below. The ESMA proposal makes no reference to "sub-delegation" of DEA access as per the Organisational Requirements for Investment Firms and the Proportionality Principle (refer to paragraph 51(1)(i) on page 210). It may be helpful to include an additional condition that trading venues make it explicit that in the event that the member permits its DEA customers to sub-delegate its access, the member should put in place effective measures to ensure it retains effective control of DEA access at all times. <ESMA_QUESTION_254> Q255: Do you agree with the list of systems and effective controls that at least DEA providers should have in place? <ESMA_QUESTION_255> Yes, LSEG agrees. In addition, ESMA may also wish to consider including a provision for the DEA provider to ensure there are effective security controls upon any network infrastructure between the DEA client and the DEA provider so that order instructions can only be submitted by the DEA client. <ESMA_QUESTION_255> Q256: Do you consider it is necessary to clarify anything in relation to the description of the responsibility regime? <ESMA_QUESTION_256> LSEG: The responsibility of the DEA provider is already clearly specified; the DEA providers are responsible as regards trading venues for all trades using their market participant ID code. The regime should be updated with reference to sub-delegation of DEA access - where this is permitted, it is the responsibility of the DEA provider to ensure that appropriate due diligence is conducted on any subdelegatee of DEA access. <ESMA_QUESTION_256> Q257: Do you consider necessary for trading venues to have any other additional power with respect of the provision of DEA? <ESMA_QUESTION_257> LSEG: No; ESMA s list seems appropriate. We suggest a further consideration below. The ESMA proposal makes no reference to "sub-delegation" of DEA access as per the Organisational Requirements for Investment Firms and the Proportionality Principle (refer to paragraph 51(1)(i) on page 210). It may be helpful to include an additional condition that trading venues make it explicit that in the event that the member permits its DEA customers to sub-delegate its access, the member should put in place effective measures to ensure it retains effective control of DEA access at all times. <ESMA_QUESTION_257> 70

71 4.4. Market making strategies, market making agreements and market making schemes Q258: Do you agree with the previous assessment? If not, please elaborate. <ESMA_QUESTION_258> LSEG: In general, yes, however there are several considerations here that ESMA should consider further, as follows: General comments on market making LSEG is supportive of strengthening the role of market makers in trading venues, by ensuring that appropriate requirements and obligations apply on firms that enter into market making agreements. Various uses of market maker / market making strategy / market making agreement in MiFID II and the need for regulatory clarity We note that MiFID makes use of the similar terms in several instances for example, a market maker is defined in MiFID Article 4(1)(7), which is the basis of its use in Article 2(1)(d)(i) (requirement for market makers to be authorised). It also uses the phrase market making strategy in Articles 17 and 48, which can appear to be independent from the definition of market maker. In our view, all these definitions are similar and cover the same type of activity the use of risk capital to provide two-sided liquidity in markets (whether through algorithmic or other types of trading). Separately MiFID also uses the terms market maker agreement and market making schemes interchangeably in Articles 17 and 48. In paragraph 2 on p.259 of the Discussion Paper and later in paragraph 8 on p.261, ESMA implies that these are only to be used in conjunction with the algorithmic trading form of market making strategies (but not with the definition of market maker?), and thus section 4.4 and its requirements should only apply to algorithmic market making strategies, and eventually only for liquid securities. LSEG feels such a link is tenuous and often gives rise to very confusing interpretations. For example, does this mean that some firms can consider themselves market makers, but not be required to sign up to market making agreements, and thus not be subject to any obligations? LSEG s view is that market making should, as has been traditional, be the submission to a set of obligations imposed by the trading venue for the benefit of providing two-way liquidity to the market more generally. This is especially relevant as the concept of market making is adopted in other legislations, such as Short Selling and various financial activity/transaction taxation scenarios. LSEG suggested approach: Our suggestion is that ESMA should simplify its treatment of market makers and market making under MiFID/MiFIR. The easiest way to do so is to define a framework based on the market making agreement, which is the one link connecting each of the terms. LSEG proposes that a market maker is any firm which either: Voluntarily enters into a market making agreement/arrangement with a trading venue (to provide two-sided liquidity in comparable sizes with obligations), OR Uses algorithmic techniques that fall under the category of market making strategies, and hence is required to enter into a written binding market making agreement/arrangement with a trading venue. This suggestion is driven by our view of the market consequence of the requirements in MiFID II. In practice, to avoid confusion for its members, a trading venue is likely to make use of a common framework for market making arrangements to define its market makers regardless of whether a firm signs up voluntarily or whether it is required to do so in the event that it operates a market making strategy. This common basic framework may extend across both liquid and not liquid instruments, algorithmic and other forms of trading, a range of asset classes (e.g. equities, bonds, derivatives etc.) and a range of regulatory 71

72 requirements e.g. MiFID II, Short Selling Regulation etc., with differences in the detail to account for different characteristics; so ESMA should be keep this in mind when designing its regime for market making. An example where this approach simplifies confusion is in the case of most European government bonds, where market making agreements exists between a sovereign borrower and a group of firms (Primary Dealers). In such cases, a trading venue is often formally designated by the borrower as the place where Primary Dealers perform their market making activity and there is an agreement between the trading venue, the borrower and the market maker. In our view, if each of these terms (i.e. market maker, market making agreement, market making strategies) is not linked, then there may be some ambiguity in how this structure operates, especially in trading venues which necessitate the use of algorithmic techniques between dealers, and between dealers and customers. An approach based on the market making agreement would take into account this fundamental feature of the European government bonds market, and also works for other asset classes. ESMA should also be careful not to define the way to determine whether a strategy is a market making strategy too strictly this may lead participants to avoid the definition, for example by only making one-way quotes in small sizes, or by preferring to become indirect/dea clients rather than members of the trading venue, if the assessment is only done at a member level. The role of market making/ avoiding flash crash scenarios Some commentators have suggested that the withdrawal of liquidity providers during volatile markets has a negative feedback effect, and that a commitment to provide liquidity regardless of market conditions would prevent the occurrence of flash crash scenarios. In paragraph 2 on p.259 of the Discussion Paper, ESMA notes that the regulatory role of such market making arrangements is to mitigate the risks of a disorderly market (e.g. flash crash scenarios) and (by implication) ensure that trading liquidity does not vanish in such situations. In LSEG s view, the reality is that in normal situations, registered market makers may be active on the order book voluntarily (i.e. without being compelled by their mandatory obligations). However, in the face of a market that is moving rapidly against them, market makers are likely withdraw from the market and any quoting obligations, regardless of regulatory requirements on them. Under the current structure, they may even fulfil their quoting obligations (e.g. be present for 70%-90% of the trading day) and withdraw in such disorderly situations. The risk of a censure or fine is likely to seem small to them compared to the risk of potentially significant losses or insolvency from meeting mandatory obligations in very fast-moving and volatile markets. In this regard, it may not be suitable to require a minimum number of market makers irrespective to the nature of the market, even for liquid securities. We suggest that it is the role of other microstructure tools e.g. volatility halts, message thresholds, tick sizes and throttles etc to manage the transition from disorderly to orderly markets and ESMA must not put too much emphasis on market making requirements in such scenarios. In reality, LSEG s experience suggests that market making is more important for less liquid securities (e.g. SME shares, illiquid bonds and derivatives) which may not be traded on an order driven model. Such markets depend on the ability of one or more registered market makers to commit risk capital and trade with investors; this secondary market activity is intricately linked to the cost of capital for the issuer and its ability to access public markets for capital. ESMA should be mindful of these considerations in this section to avoid any unintended consequences. <ESMA_QUESTION_258> Q259: Do you agree with the preliminary assessments above? What practical consequences would it have if firms would also be captured by Article 17(4) MiFID II when posting only one-way quotes, but doing so in different trading venues on different sides of the order book (i.e. posting buy quotes in venue A and sell quotes in venue B for the same instrument)? 72

73 <ESMA_QUESTION_259> Please see LSEG s response to Q258 for views on ESMA s approach. In summary, whilst we understand why ESMA wishes to consider the terms in Article 17 and 48 independently from other references to market makers and market making, we feel that such an approach may lead to confusion as the terms market maker, market making, market making agreement and market making strategy often mean the same thing and in practice are used interchangeably. In particular we feel that the focus should be on the market making agreement and any member who voluntarily signs up to one, or is required to sign up to one (as it operates a market making strategy) should be considered a market maker, and the framework in this section should apply. LSEG agrees that for the purpose of determination of a market making strategy only external and objective elements should be considered (firm, simultaneous two-way quotes at competitive prices). LSEG strongly feels that only two-way activity on a single trading venue should be considered for the purpose of this determination. This is because: Firms which operate one sided quotes on different a venue may decide that the obligations of becoming a registered maker are not suitable for them, which could result in them withdrawing their liquidity; Market participants on different trading venues will not have access to two-sided liquidity from such activity, unless they are members of both trading venues, which may not be the case; Such data will be challenging to collect and monitor, and gives rise to uncertainty e.g. with which venue will the participant need to enter into a market making agreement? How will the obligations be monitored? <ESMA_QUESTION_259> Q260: For how long should the performance of a certain strategy be monitored to determine whether it meets the requirements of Article 17(4) of MiFID II? <ESMA_QUESTION_260> Evidence from LSEG s trading venues suggests that there can be different approaches, depending on the market, asset class or geography. From our experience, we suggest an observation period of 1-3 months for the performance of a strategy is suitable, with flexibility for the trading venue to decide the most suitable period and methodology for its market model. For example, a venue may assess a strategy s performance for a certain number of hours daily, and then aggregate this over a period of 1-3 months. <ESMA_QUESTION_260> Q261: What percentage of the observation period should a strategy meet with regard to the requirements of Article 17(4) of MiFID II so as to consider that it should be captured by the obligation to enter into a market making agreement? <ESMA_QUESTION_261> LSEG believes that ESMA should require the minimum principle that it should observe strategies under the same criteria and equal thresholds that a registered market maker is obliged to fulfil once it enters into a written agreement (e.g. percentage of activity, proportion of trading hours active, pricing, sizes, quotes etc.). Setting lower thresholds for the observation period and then requiring firms to be subject to more stringent requirements once regulation requires them to sign up to a binding agreement would be a disproportionate outcome, in our view. In terms of the actual percentage, this may depend on the trading venue. Evidence from our markets suggests that a period of 70%-90% should be appropriate, although different approaches may be required for different asset classes (e.g. performance in European rates markets is 5 hours per day, assessed over a monthly period; for shares, 80% would align MiFID with the requirements in the Short Selling Regulation). <ESMA_QUESTION_261> 73

74 Q262: Do you agree with the above assessment? <ESMA_QUESTION_262> Yes, LSEG agrees that the observation of strategies and requirement to sign up to a market making agreement should be for members of the trading venue. Only principal flow should be considered in this assessment. However, we are conscious that this should not create an imbalance and an incentive for firms to become DEA clients rather than being direct members of a trading venue. This may potentially be an unintended consequence, with members choosing to become DEA clients to avoid being subject to a requirement to register as market makers. Thus, we suggest that this requirement should not preclude the possibility for a trading venue to create market making arrangements suitable for DEA clients. This could be through a tri-party agreement (between the venue, DEA provider and the underlying client) or exclusively with the DEA provider, whereby a DEA client that assumes the role of a market maker will be required to deal on third account in the instruments whose liquidity it undertakes to support. <ESMA_QUESTION_262> Q263: Do you agree with this interpretation? <ESMA_QUESTION_263> Yes, LSEG agrees. <ESMA_QUESTION_263> Q264: Do you agree with the above assessment? If not, please elaborate. <ESMA_QUESTION_264> Yes, LSEG agrees with ESMA s view in paragraph 16 on p.262 of the Discussion Paper. <ESMA_QUESTION_264> Q265: Do you agree with the above interpretation? <ESMA_QUESTION_265> Yes, LSEG agrees. The two sides of the market maker quotes should be simultaneous. On venues, market makers usually submit a two-sided quote through a single message. For the purposes of assessing if a strategy is market making, a per second basis seems appropriate. <ESMA_QUESTION_265> Q266: Do you agree with the above proposal? <ESMA_QUESTION_266> No. LSEG is of the view that whilst the overall exposure of the firm may be correlated to market makerlike behaviour, it is not a consequence of market making activity, and hence should not be the metric to judge if a firm posts orders in comparable sizes. It is also inconsistent with ESMA s approach in the rest of the section, where it seeks to identify market making strategies on an individual basis, rather than on aggregate criteria. LSEG suggests that as an alternative NCAs or trading venues should monitor for simultaneous quotes at comparable sizes, where the size criteria is at least equal to the minimum sizes at which registered market makers are obliged to quote on that trading venues. <ESMA_QUESTION_266> Q267: Do you agree with the above proposal? <ESMA_QUESTION_267> Yes, LSEG agrees. 74

75 <ESMA_QUESTION_267> Q268: Do you agree with the approach described (non-exhaustive list of quoting parameters)? <ESMA_QUESTION_268> Yes, LSEG agrees. We note that market maker quotes are subject to minimum size obligations (which may be expressed in terms of volume or value). In paragraph 29 on p.265 of the Discussion Paper, ESMA only refers to quotation value; we suggest that ESMA clarifies that quotation volume is also considered here. <ESMA_QUESTION_268> Q269: What should be the parameters to assess whether the market making schemes under Article 48 of MiFID II have effectively contributed to more orderly markets? <ESMA_QUESTION_269> LSEG agrees with ESMA s assessment in paragraph 29 on p.265 of the Discussion Paper that the parameters should be the proportion of time present during the trading day, size of quotes, spread of quotes and firm/executable prices. Similar to our response in Q261 and Q268, ESMA s minimum principle should be that the level of the thresholds for assessing whether a strategy is market making should be at least equal to the minimum quoting requirements for registered market makers at that trading venue. <ESMA_QUESTION_269> Q270: Do you agree with the list of requirements set out above? Is there any requirement that should be added / removed and if so why? <ESMA_QUESTION_270> In principle, LSEG agrees with ESMA s approach in paragraph 32 on p.266 of the Discussion Paper. From a trading venue s perspective we suggest that it may be less relevant for criteria (vii) remuneration schemes to be included in a market making agreement. <ESMA_QUESTION_270> Q271: Please provide views, with reasons, on what would be an adequate presence of market making strategies during trading hours? <ESMA_QUESTION_271> LSEG: The actual proportion may be different for different trading venues, depending on the asset class, the type of users, liquidity, market model etc. Evidence from LSEG s markets suggests that a period of 70%-90% is appropriate for shares, ETFs, DRs and exchange traded derivatives, and 80% would align MiFID with the requirements in the Short Selling Regulation for shares. However, for European rates markets, the standard daily performance requirement is expressed as 5 hours per day, assessed over a monthly period, which is a different method to equity markets. <ESMA_QUESTION_271> Q272: Do you consider that the average presence time under a market making strategy should be the same as the presence time required under a market making agreement? <ESMA_QUESTION_272> Yes, LSEG agrees, given that a strategy is assessed at an instrument level, and should it meet the required thresholds a market making agreement would apply to the firm as a whole. Please also see our response to Q258. ESMA should consider the market making agreement as the driver for this section as this is the regulatory/ legal outcome (i.e. you are market maker because you sign up to 75

76 the agreement the reason, whether voluntary or as a consequence of your operating a market making strategy assessed as per the criteria in Q is irrelevant). Based on our interpretation of the level I text of MiFID II, once you are a registered market maker, the thresholds in the relevant market making agreement (including time present) will be the same as the thresholds that were used to assess if your strategy is market making. It should also be recognised that an individual trading venue may offer a number of different market maker schemes (which may be more restrictive than the most basic agreement), with different levels of obligations for an individual instrument. The thresholds in the relevant agreement/scheme should apply equally to firms who take up those market making agreements/schemes voluntarily. This would ensure an equal treatment of all market makers subscribing a specific agreement/scheme and prevent any disparity or confusion. <ESMA_QUESTION_272> Q273: Should the presence of market making strategies during trading hours be the same across instruments and trading models? If you think it should not, please indicate how this requirement should be specified by different products or market models? <ESMA_QUESTION_273> LSEG: No. Evidence from our markets suggests that different methods apply depending on the asset class, the type of users, liquidity, market model etc. For e.g. a period of 70%-90% is appropriate for shares ETFs and DRs, and 80% would align MiFID with the requirements in the Short Selling Regulation for shares. For European rates markets, the standard daily performance requirement is expressed as 5 hours per day, assessed over a monthly period, which is a different method to equity markets. <ESMA_QUESTION_273> Q274: Article 48(3) of MiFID II states that the market making agreement should reflect where applicable any other obligation arising from participation in the scheme. What in your opinion are the additional areas that that agreement should cover? <ESMA_QUESTION_274> LSEG: In general, the additional areas may include an obligation for fair and honest behaviour that ensures market integrity and adequate systems and controls. Arrangements may also cover the type of quote or flag or message that market makers need to use when performing their activity. In rates and government bond markets, for Primary Dealers there may also be minimum capital requirements. <ESMA_QUESTION_274> Q275: Do you disagree with any of the events that would qualify as exceptional circumstances? Please elaborate. <ESMA_QUESTION_275> No, LSEG does not disagree. <ESMA_QUESTION_275> Q276: Are there any additional exceptional circumstances (e.g. reporting events or new fundamental information becoming available) that should be considered by ESMA? Please elaborate. <ESMA_QUESTION_276> For securities based on an underlying index or basket, e.g. ETFs or derivatives, in addition to the list, LSEG s view is that market making obligations may be suspended if no firm price is available for 10% or more (by weighting) of the underlying index. This may be through either regulatory suspension or public holiday. Should a significant event occur such that prices are not available for any underlying constituent, then a registered market maker can request to have the maximum spread regime temporarily suspended. Once notified, the trading venue will issue a message to the market. 76

77 Trading venues may have the ability to relax market making obligations where an instrument is subject to wide price movements ( fast market conditions ), when there are technical errors in the trading systems of the venue or the market maker that have an effect on the ability of market participants to enter or receive messages in proper time, or in circumstances linked to corporate actions. <ESMA_QUESTION_276> Q277: What type of events might be considered under the definition of political and macroeconomic issues? <ESMA_QUESTION_277> LSEG: Events may include the default of a large or systemic participant, economic sanctions or exogenous unexpected events that have macroeconomic impact (e.g. Acts of God such as natural disasters, political or martial conflict, acts of terrorism etc). <ESMA_QUESTION_277> Q278: What is an appropriate timeframe for determining whether exceptional circumstances no longer apply? <ESMA_QUESTION_278> Based on our experience, LSEG s view is that it is not possible to provide a precise timeframe. This will depend on the circumstances of each situation, which are likely to be exceptional; so each should be assessed on a case-by-case basis at each end-of-day, with a decision communicated to the market sufficiently in advance before the open <ESMA_QUESTION_278> Q279: What would be an appropriate procedure to restart normal trading activities (e.g. auction periods, notifications, timeframe)? <ESMA_QUESTION_279> LSEG: Once again, this depends on the circumstances of each exceptional situation. Ideally, the venue should resume operation as it would in normal circumstances. Notification to the market and market makers is important. A market restart may be through an auction (if that is its mechanism), and consideration may be given if gradual re-instatement of market making obligations is required (with progressive tightening of maximum spreads and size criteria until a normal level is reached). <ESMA_QUESTION_279> Q280: Do you agree with this approach? If not, please elaborate. <ESMA_QUESTION_280> No. LSEG disagrees with certain elements, as below. Here we note, similar to our General Comments in our responses to Q258 and Q272, that ESMA risks confusion when linking the requirements in Article 17 and Article 48 only to algorithmic trading and to instruments with liquid markets. Market making extends across all instruments (both liquid and not liquid) and a narrow approach to only market making strategies and agreements without considering all registered market makers (i.e. all participants under market making agreements) risks inconsistency and disparity between different types of market makers. In particular, ESMA s comment in paragraph 43(i) on p.270 of the Discussion Paper that trading venues should ensure that for the type of instruments addressed by Articles 48 and 17 of MiFID II, the liquidity existing in their market does not result exclusively or quasi-exclusively from market making activity, as far as liquid instruments are concerned is slightly misguided. Some instruments in quote driven markets, e.g. SME equity securities, which may or may not be liquid, are driven by substantial activity by market makers. Even in sovereign bond markets on electronic platforms, some venues are exclusively dedicated to Primary Dealers, who are all required to act as market makers, and exist as the result of the agreement between a sovereign borrower and its Primary Dealers. Activity on these markets is driven by these Dealers. These 77

78 systems are in scope of algorithmic trading requirements, so should be relevant to the requirements in Article 17 and 48; however, ESMA s approach which is more linked to very liquid securities, is not correct in this regard. In paragraph 43(iii), we agree that trading venues should have publicly available rules in place in relation to how market makers are expected to comply with market making arrangements and the penalties for not being compliant. LSEG does not believe ESMA requires venues to publish the level of compliance for each of its market makers, and if it does we would not support such a provision. This would disincentivise participants from registering as market makers and may be detrimental to the liquidity that they may otherwise be willing to provide. Finally, on paragraph 43(iv), LSEG notes that trading venues should only make public the identities of those market makers who agree that their names can be disclosed. <ESMA_QUESTION_280> Q281: Would further clarification be necessary regarding what is fair and nondiscriminatory? In particular, are there any cases of discriminatory access that should be specifically addressed? <ESMA_QUESTION_281> LSEG: None, as far as we are aware. <ESMA_QUESTION_281> Q282: Would it be acceptable setting out any type of technological or informational advantages for participants in market making schemes for liquid instruments? If yes, please elaborate. <ESMA_QUESTION_282> LSEG: No; our view is that participation in a market making scheme should be available to all members according to fair, objective and non-discriminatory criteria. We do not see what technological or informational advantages market makers would or should have over other participants, although we recognise that some technical specifications (e.g. message throttle limits) may be different proportional to a market maker s obligations. In particular, from an informational standpoint, LSEG is firmly of the view that all market participants should receive market data without any arbitrage opportunities. <ESMA_QUESTION_282> Q283: In which cases should a market operator be entitled to close the number of firms taking part in a market making scheme? <ESMA_QUESTION_283> LSEG: Consideration could be given, for instance, to the possibility of limiting the number of market makers if their number posed a risk to the IT system, but this is unlikely to be the case, given the requirements on system capacity. In some cases, trading venues may also offer schemes that are objective and non-discriminatory but exclusive after a limit (e.g. scheme open to five members, applications accepted on first-come-first-served basis). For example, in the case of MTS Hungary, MTS admits only primary dealers who act on the market as market makers, the number of which is determined by the country s Debt Management Office - in this case the ÁKK. <ESMA_QUESTION_283> Q284: Do you agree that the market making requirements in Articles 17 and 48 of MiFID II are mostly relevant for liquid instruments? If not, please elaborate how you would apply the requirements in Articles 17 and 48 of MiFID II on market making schemes/agreements/strategies to illiquid instruments. <ESMA_QUESTION_284> 78

79 No, LSEG disagrees. Please also see our General Comments in response to Q258, Q272 and Q280. LSEG s view is that ESMA risks confusion when linking the requirements in Article 17 and Article 48 only to algorithmic trading and to liquid instruments. Market making extends across all instruments (both liquid and not liquid) and a narrow approach to just market making strategies and agreements without considering all registered market makers (i.e. all participants under market making agreements) risks inconsistency and disparity between different types of market makers. Indeed, it is arguable that market making provides much higher marginal benefits in illiquid instruments (across equities, fixed income and derivatives), which are more likely to be quote driven. ESMA s regulatory requirements for market making must be driven by the outcome, i.e. a market making agreement between the member and the venue, regardless of how that outcome was reached - whether through an assessment of whether the firm is operating a market making strategy, or whether the firm signs up to an agreement voluntarily. It is also worth noting that the systems covered by algorithmic trading may include both continuous orderdriven systems (usually for liquid), quote driven systems (less liquid and more reliant on market makers) and some hybrid systems or RFQ systems operated by trading venues particularly in the non-equity space (e.g. hybrid systems that require price streaming from market makers or SI quoting engines to facilitate execution necessitate forms of algorithmic trading). So the regime should extend across all liquidity classes and market models. LSEG does not see any major difference in application to less liquid instruments. In practice we think that, to avoid confusion for its members, a trading venue is likely to make use of a common framework for market making arrangements to define its market makers regardless of whether a firm signs up voluntarily or whether it is required to do so in case it operates a market making strategy. This common basic framework may extend across both liquid and not liquid instruments, algorithmic and other forms of trading, a range of asset classes (e.g. equities, bonds, derivatives etc.) and a range of regulatory requirements e.g. MiFID II, Short Selling Regulation etc., with differences in the detail to account for different market characteristics. LSEG does not see any particular challenge to adapting the framework as appropriate. <ESMA_QUESTION_284> Q285: Would you support any other assessment of liquidity different to the one under Article 2(1)(17) of MiFIR? Please elaborate. <ESMA_QUESTION_285> No; LSEG agrees this is the right definition to use. <ESMA_QUESTION_285> Q286: What should be deemed as a sufficient number of investment firms participating in a market making agreement? <ESMA_QUESTION_286> LSEG disagrees with ESMA s approach and its reasoning in paragraph 52 on p.272. We explain why below: As noted in LSEG s General Comments in Response to Q258, Q272, Q282 and Q284, we disagree with ESMA s approach to limit its treatment of market making to liquid securities. It is arguable that market making provides much higher benefits to trading in illiquid instruments (across equities, fixed income and derivatives), which are much more likely to be quote driven in our experience. LSEG also suggests that ESMA risks inconsistency when linking the requirements in Article 17 and Article 48 only to algorithmic trading and to instruments with liquid markets. A narrow approach to just market making strategies and agreements without considering all registered market makers (i.e. all participants under market making agreements) may create disparity between different types of market makers. 79

80 With regards to a requirement for a minimum number of market makers for liquid securities to provide sufficient coverage against disorderly trading conditions, this approach has a number of flaws. In our view, the reality is that in normal situations registered market makers may be active on the order book voluntarily, i.e. without being compelled by their mandatory obligations. However, in the face of a market that is moving rapidly against them, market makers are likely withdraw from the market and any quoting obligations, regardless of the regulatory requirements on them. Under the proposed structure for market making, they may even fulfil their quoting obligations (e.g. be present for 70%-90% of the trading day) and withdraw in such disorderly situations. The risk of a censure or fine is likely to seem small to them compared to the risk of potentially significant losses or insolvency from meeting mandatory obligations in very fastmoving and volatile markets. Thus LSEG s view is that it is not suitable to require a minimum number of market makers even for liquid securities. This should be left to the trading venue s discretion to ensure sufficient control at the security level. LSEG suggests that it is the role of other microstructure tools e.g. volatility halts, message thresholds, tick sizes and throttles etc. to manage the transition from disorderly to orderly markets and ESMA must not put too much emphasis on market making requirements in such scenarios. <ESMA_QUESTION_286> Q287: What would be an appropriate market share for those firms participating in a market making agreement? <ESMA_QUESTION_287> LSEG does not believe this question is relevant to the discussion. In order driven securities, the market share of any individual participant will be low. In quote driven securities supported primarily by market makers, the share will be high. Indeed, when there is only one market maker supporting liquidity (e.g. in some cases for ETFs), the share of the market maker will be 100%. In our view, ESMA does not need to define any thresholds for an appropriate market share of a market maker, given the wide range of instruments for which market making arrangements are in place. <ESMA_QUESTION_287> Q288: Do you agree that market making schemes are not required when trading in the market via a market making agreement exceeds this market share? <ESMA_QUESTION_288> No, LSEG disagrees. Please see our responses to Q286 and Q287 for reasons. LSEG s view is that ESMA s approach to this issue is over complicated and risks confusion. Please also see our response to Q258. <ESMA_QUESTION_288> Q289: In which cases should a market operator be entitled to close the number of firms taking part in a market making scheme? <ESMA_QUESTION_289> LSEG: Market making agreements should be offered on a fair and non-discriminatory basis. We suggest the possibility of permitting a limitation on the number of market makers, if their number posed a risk to the IT system. This is unlikely to be the case, given the requirements on system capacity but worth mentioning. In some cases, trading venues may also offer schemes that are objective and non-discriminatory but exclusive after a limit (e.g. scheme open to five members, applications accepted on first-come-first-served basis). For example, in the case of MTS Hungary, MTS admits only primary dealers who act on the market 80

81 as market makers, the number of which is determined by the country s Debt Management Office - in this case the ÁKK. This should not be based on any market share metrics. Please also see our answers to Q286 and Q287. <ESMA_QUESTION_289> 81

82 4.5. Order-to-transaction ratio (Article 48 of MiFID II) Q290: Do you agree with the types of messages to be taken into account by any OTR? <ESMA_QUESTION_290> Yes, LSEG agrees. LSEG supports ESMA s interpretation that OTRs should be a tariff based mechanism to avoid excessive capacity burden on the system due to user generated messages. A description LSE s 5 and Borsa Italiana s 6 tariffs based on OTRs can be found on its public price lists (please see footnotes below). Only user generated messages (submission, modifications, deletions) should count in the calculation. Users do not have control on messages generated by a trading system, so it would be disproportionate to include this in the calculation. In this regard, an IOC order must be counted as ONE message, because from a technical perspective it is only a single user-generated message (order entry), like a limit order. The cancellation is programmed in the system itself as an order expiry if the execution condition is not filled. This is similar to how limit orders are programmed the system will expire the order under the specified condition e.g. next auction period, end of day etc. We therefore do not believe that IOC should be treated any differently. In theory, OTRs should only apply to liquid shares and DRs, and we welcome that ESMA reflects this in its approach. LSEG notes here that it is not in favour of a mandatory OTR regime applying to ETFs. The number of orders in a share and those instrument based on a basket of underlyings (e.g. ETFs) is likely to be very different; as a market participant providing liquidity in ETFs may need to update its orders more frequently due to movements in the underlying. As we note in our answer to Q54, ETFs are generally traded in large quantities, but are not traded as frequently as shares. A mandatory OTR for ETFs may unintentionally suppress liquidity further. We note here that other tools, e.g. message throttle limits, may be used to ensure system capacity. We also note that this approach does not prevent a trading venue from applying an OTR tariff for ETFs on a voluntary basis. LSEG also suggests that a mandatory OTR is not relevant to quote-driven or RFQ trading models, particularly for bonds. <ESMA_QUESTION_290> Q291: What is your view in taking into account the value and/or volume of orders in the OTRs calculations? Please provide: <ESMA_QUESTION_291> LSEG: There seems to be no obvious purpose, benefit or ease of introducing volume/value into the OTR calculation. Reasoning 5 LSE Price List, Page 7, note 2: 20December% pdf 6 Borsa Italiana Price List, Page 21, paragraph 10: 82

83 The primary purpose of the OTR is to disincentivise users from entering too many messages that may place an excessive burden on the system, or that may be evidenced to be excessive. In our view, this is best reflected using system message ratios. Cons of volume/value approaches OTRs based on volumes and values may have undesirable and unintended consequences. For example, many would suggest that large displayed orders add value to the market place even if they are unexecuted and later cancelled. Value/ volume based OTRs may disincentivise the posting of such high value/ volume orders, especially if the member is not convinced that his order will execute fully. This is quite opposite to the regulatory push to incentivise trading in higher sizes. Further, accounting for value and volume risks penalising participants trading frequently, but at low value or low volume (e.g. retail brokers). Finally, the definition complicates the concept of OTR which is currently well understood throughout the industry as applying to order messages. Methodology No comments, as we disagree with the use of value or volume metrics. <ESMA_QUESTION_291> Q292: Should any other additional elements be taken into account to calibrate OTRs? If yes, please provide an explanation of why these variables are important. <ESMA_QUESTION_292> LSEG: No. A message-based OTR is a simple, complete and well understood quantity. <ESMA_QUESTION_292> Q293: Do you agree with the proposed scope of the OTR regime under MiFID II (liquid cash instruments traded on electronic trading systems)? <ESMA_QUESTION_293> LSEG: Yes, we agree for shares and DRs. We disagree for ETFs and bonds. ETFs: LSEG notes here that it is not in favour of a mandatory OTR regime applying to ETFs. The number of orders in a share and those in instrument based on a basket of underlyings (e.g. ETFs) is likely to be very different; as a market participant providing liquidity in ETFs may need to update its orders more frequently due to movements in the underlying. As we note in our answer to Q54, ETFs are generally traded in large quantities, but are not traded as frequently as shares. A mandatory OTR for ETFs may unintentionally suppress liquidity further. We note here that other tools, e.g. message throttle limits, may be used to ensure system capacity. We also note that this approach does not prevent a trading venue from applying an OTR tariff for ETFs on a voluntary basis. Bonds: Paragraph 6 on p.275 of the Discussion Paper notes that Empirical work indicates that high frequency traders tend to trade liquid stocks with high market value ( blue chips ). There is no such empirical evidence relating to either sovereign bonds or corporate bonds, and no evidence that OTR regimes are applicable to such instruments or likely to yield benefits. We are unaware of any issue on any of the fixed income markets operated by LSEG which an OTR regime would resolve. We would accordingly propose that the initial scope is restricted to shares and DRs, and if felt appropriate this scope is reviewed in the future. In theory though, if the OTR is appropriately calibrated, it should take into account the different nature of the bond market and leave it undisturbed. However, given the support for electronification and liquidity of ETF and fixed income markets, we suggest it is appropriate to start off with shares and DRs first. 83

84 <ESMA_QUESTION_293> Q294: Do you consider that financial instruments which reference a cash instrument(s) as underlying could be excluded from the scope of the OTR regime? <ESMA_QUESTION_294> Yes, LSEG agrees. Please also see our response to Q293. <ESMA_QUESTION_294> Q295: Would you make any distinction between instruments which have a single instrument as underlying and those that have as underlying a basket of instruments? Please elaborate. <ESMA_QUESTION_295> LSEG: Yes. An ETF/ETP which is based on a basket of securities is likely to require more message updates than a single instrument. Please refer to our response to Q292 and Q293 LSEG does not support the inclusion of ETFs in the scope of the mandatory OTR. <ESMA_QUESTION_295> Q296: Do you agree with considering within the scope of a future OTR regime only trading venues which have been operational for a sufficient period in the market? <ESMA_QUESTION_296> LSEG: This question rather underlines that the regulatory rationale for the mandatory OTR is not clear if it is intended to be a throttling mechanism to ensure market resilience and adequate capacity, this issue will apply, regardless of the period for which the trading venue has been operating. Similarly, if the OTR is designed to apply in a way that counters perceived abusive or excessive order entry and cancellation, this should be driven by empirical evidence of such abuse on that venue. This is why ESMA s statement that one size does not fit all in paragraph 12 on p.276 of the Discussion Paper is particularly true, and LSEG welcomes such an approach. On that basis, LSEG suggests a review of the use/operation of the system when deciding whether a mandatory OTR should apply the approach should be reviewed by the trading venue and its NCA. <ESMA_QUESTION_296> Q297: If yes, what would be the sufficient period for these purposes? <ESMA_QUESTION_297> Please see LSEG s response to Q296. <ESMA_QUESTION_297> Q298: What is your view regarding an activity floor under which the OTR regime would not apply and where could this floor be established? <ESMA_QUESTION_298> LSEG believes that members with the lowest levels of messaging are unlikely by themselves to pose a threat to the capacity of a trading venue. However, we consider that a floor should not be defined by regulators and trading venues should be allowed to apply a floor at their own discretion. <ESMA_QUESTION_298> Q299: Do you agree with the proposal above as regards the method of determining the OTR threshold? <ESMA_QUESTION_299> No, LSEG disagrees with ESMA s proposed approach. We explain why below. 84

85 In our view, the purpose of introducing OTR is twofold: (1) to ensure orderly trading conditions on trading venues by controlling the number of orders members may send to the trading system, thereby ensuring system capacity is not overburdened; (2) to disincentivise excessive order entry and subsequent cancellation that may give the perception of liquidity when it does not really exist (in our view, this should be driven by empirical evidence of such abuse on a venue). LSEG s view is that using average observed OTRs and then basing with respect to the highest OTR observed does not work to fulfil either of these purposes. The average OTR may as it may be significantly lower than the total capacity of the trading system, in which case the OTR cap would be applied at a level that is inconsistent with the capacity made available in the trading system. An average metric (which is based on message rates of other participants) may also not be an appropriate proxy for a level of excessive orders to trades if appropriate for that trading venue. LSEG s suggestion is that any mandatory OTR cap should be set at a level taking consideration of the total capacity of the trading system to manage sustained message flow at that rate, and an evidence based level of orders-to-trades which is considered excessive for a particular asset class. In line with our response to Q 296, LSEG suggests that the approach to OTR should be reviewed by a venue and its NCA. <ESMA_QUESTION_299> Q300: In particular, do you consider the approach to base the OTR regime on the average observed OTR of a venue appropriate in all circumstances? If not, please elaborate. <ESMA_QUESTION_300> No, LSEG disagrees with ESMA s proposed approach. Please see our response to Q299 for further detail. <ESMA_QUESTION_300> Q301: Do you believe the multiplier x should be capped at the highest member s OTR observed in the preceding period? <ESMA_QUESTION_301> No, LSEG disagree with ESMA s proposed approach. Please see our response to Q299 for further detail. Further, the volatility may be low in a period so that every OTR is abnormally lower than previous periods. The OTR threshold should be set at a level taking consideration of the total capacity of the trading system to manage sustained message flow at that rate. <ESMA_QUESTION_301> Q302: In particular, what would be in your opinion an adequate multiplier x? Does this multiplier have to be adapted according to the (group of) instrument(s) traded? If yes, please specify in your response the financial instruments/market segments you refer to. <ESMA_QUESTION_302> LSEG disagrees with ESMA s proposed approach. Please see our response to Q299 for further detail. <ESMA_QUESTION_302> Q303: What is your view with respect to the time intervals/frequency for the assessment and review of the OTR threshold (annually, twice a year, other)? <ESMA_QUESTION_303> LSEG: A minimum of annually would be an appropriate timeframe for review because this is consistent with existing best practice. Monthly may be more suitable for the inclusion of new venues <ESMA_QUESTION_303> Q304: What are your views in this regard? Please explain. <ESMA_QUESTION_304> 85

86 LSEG supports the approach in paragraph 20(ii), whereby the continuous two-way quoting activity of registered market makers i.e. those market makers that are subject to a written agreement, under their agreed obligations is exempt from the OTR. <ESMA_QUESTION_304> 4.6. Co-location (Article 48(8) of MiFID II) Q305: What factors should ESMA be considering in ensuring that co-location services are provided in a transparent, fair and non-discriminatory manner? <ESMA_QUESTION_305> LSEG: Co-location and close proximity access services, whether provided directly by a trading venues or indirectly via a reciprocal/or other arrangement with a 3rd party data centre provider, should be provided in a transparent, fair and non-discriminatory manner to investment firms. Pricing of colocation/ proximity hosting services should be consistent, with standard pricing based on objective factors such as contract term, size of space required, connectivity specifications etc. <ESMA_QUESTION_305> 4.7. Fee structures (Article 48 (9) of MiFID II) Q306: Do you agree with the approach described above? <ESMA_QUESTION_306> Yes, LSEG agrees with ESMA s approach. Fees should be transparent, fair and non-discriminatory and not encourage improper trading practices or disorderly markets. The same requirements are applied in the same way to all trading venues, including regulated markets, MTFs and OTFs, within the same context, i.e. asset classes and client base (wholesale trading vs nonwholesale trading). <ESMA_QUESTION_306> Q307: Can you identify any practice that would need regulatory action in terms of transparency or predictability of trading fees? <ESMA_QUESTION_307> Transparency is fundamental to guarantee non-discriminatory fee structures. Here, ESMA should be mindful that equal requirements should apply for services provided by trading venues and investment firms within the scope of MiFID II, and providers outside the scope of MiFID II who offer the same services (e.g some third party providers that are not regulated entities such as data centres). Regulatory action should be careful not to create an unlevel playing field or the potential of a competitive advantage through regulatory arbitrage. <ESMA_QUESTION_307> Q308: Can you identify any specific difficulties in obtaining adequate information in relation to fees and rebates that would need regulatory action? <ESMA_QUESTION_308> LSEG notes here that ESMA should be aware that some trading venues provide information on discounts/rebates only on request then only to a restricted category of customer or potential customer. For example, fee incentives connected to liquidity provision schemes that are not disclosed on the fee structure and that are not made public. We would discourage such practices. 86

87 <ESMA_QUESTION_308> Q309: Can you identify cases of discriminatory access that would need regulatory action? <ESMA_QUESTION_309> No, for published fee structures, we are not aware of any discriminatory access to fees. This is not possible to say, of course, in relation to incentives and schemes made available only on request. <ESMA_QUESTION_309> Q310: Are there other incentives and disincentives that should be considered? <ESMA_QUESTION_310> The scope of fee structure should include all services related to membership or participation in a trading system. ESMA should be conscious of trading like services offered outside trading venues; if a similar multilateral activity is performed through other mechanisms, then there is a risk of an unlevel playing field and bias towards firms outside the scope of MiFID II. With regard to the list set in paragraph 13 on p.282 of the Discussion Paper, we would suggest adding: introductory offers, promotions, cross-selling and responding to ad hoc requests from customers on a transparent and non-discriminatory basis. <ESMA_QUESTION_310> Q311: Do any of the parameters referred to above contribute to increasing the probability of trading behaviour that may lead to disorderly and unfair trading conditions? <ESMA_QUESTION_311> It is important that trading venues retain the flexibility to operate price lists that are innovative and adaptable: we assume that when ESMA says in paragraph 13 on p.282 of the Discussion Paper that there are three types of incentives/disincentives in MiFID II, this is not an exhaustive list. In that regard, it is correct when ESMA says that the parameters set out in paragraph 15 cannot of themselves incentivise improper trading behaviour. Moreover, even when provided with liquidity obligations, rebates should not be so high as that it encourages improper trading. <ESMA_QUESTION_311> Q312: When designing a fee structure, is there any structure that would foster a trading behaviour leading to disorderly trading conditions? Please elaborate. <ESMA_QUESTION_312> Fee structures can incentivise improper trading practices when fees become the sole or the prevalent reason for trading and/or for becoming member of a market, such as the example in response to Q311. <ESMA_QUESTION_312> Q313: Do you agree that any fee structure where, upon reaching a certain threshold of trading by a trader, a discount is applied on all his trades (including those already done) as opposed to just the marginal trade executed subsequent to reaching the threshold should be banned? <ESMA_QUESTION_313> LSEG is not opposed to the banning of such cliff-edge schemes, if the threshold is fixed and known. However, the scheme might be acceptable if the threshold for reaching the discount is not known in advance, and if the trading activity enhances the liquidity of the market. 87

88 A ban should not outlaw other schemes that may encourage meeting a certain other measure in order to gain a benefit to be met over a defined period, e.g. a discount for certain ratio of passive to aggressive trading. <ESMA_QUESTION_313> Q314: Can you identify any potential risks from charging differently the submission of orders to the successive trading phases? <ESMA_QUESTION_314> Yes. The risks are (i) many venues do not charge a fee for order entry, and charge for executions ad valorem, therefore charging for orders on a basis relative to the execution fee would render ad valorem price lists obsolete. Also, (ii) order entry does not guarantee execution, therefore there is a legitimate reason not to charge for order entry if the participant is not guaranteed execution. The same principles to orders/ executions apply to activity in the different trading phases identified in Q317. <ESMA_QUESTION_314> Q315: Are there any other types of fee structures, including execution fees, ancillary fees and any rebates, that may distort competition by providing certain market participants with more favourable trading conditions than their competitors or pose a risk to orderly trading and that should be considered here? <ESMA_QUESTION_315> LSEG suggests that ESMA may look at the effect that the following practices have: Offering free admission to trading to certain instruments (like ETPs), if the issuer de-lists the same instrument from the other markets. This practice may distort and/or reduce competition because trading would be forced to one trading venue only and traders would be obliged to switch their flow to that venue. In addition to this, the competing trading venue would benefit from the listing procedures already managed by the other primary venues of listing. Payments to firms for them to become members of markets, without any related obligations. <ESMA_QUESTION_315> Q316: Are there any discount structures which might lead to a situation where the trading cost is borne disproportionately by certain trading participants? <ESMA_QUESTION_316> As long as fee structures are transparent and applied according to objective criteria, the risk highlighted in the question is reduced. When there are bundled services and/or pricing packages, any component of the offering should be clarified and users should be offered the chance to buy each service separately. <ESMA_QUESTION_316> Q317: For trading venues charging different trading fees for participation in different trading phases (i.e. different fees for opening and closing auctions versus continuous trading period), might this lead to disorderly trading and if so, under which circumstances would such conditions occur? <ESMA_QUESTION_317> No, LSEG is not aware of any evidence that this might lead to disorderly trading. <ESMA_QUESTION_317> Q318: Should conformance testing be charged? <ESMA_QUESTION_318> 88

89 In line with our response to Q244, LSEG is of the view that a trading venue should have the ability to impose process, content and timing of conformance tests, if required, to ensure that all members (and their clients) reach a minimum benchmark. It should be decided by the trading venue whether this is charged for separately on a fair, transparent and non-discriminatory basis. <ESMA_QUESTION_318> Q319: Should testing of algorithms in relation to the creation or contribution of disorderly markets be charged? <ESMA_QUESTION_319> Please see our response to Q318. Trading venues are in a good position to introduce functionality and tools to test algorithms against disorderly trading environments; which are currently seen as value-add rather than as part of the core testing services. It is reasonable to expect that the costs of development and maintenance of testing environments may be offered on a fair and commercial basis. In LSEG s view, it should be decided by the trading venue whether this is charged for separately or as part of a trading venue s membership, but separated pricing provides additional transparency. <ESMA_QUESTION_319> Q320: Do you envisage any scenario where charging for conformance testing and/or testing in relation to disorderly trading conditions might discourage firms from investing sufficiently in testing their algorithms? <ESMA_QUESTION_320> No. LSEG s view is that as long as it is offered on a fair and commercial basis and if transparency and objective criteria are applied to fee structures, there is no risk of such an outcome. We note here that firms are also separately required to test their algorithms separately (as in ESMA s analysis in Section 4.2. of the Discussion Paper) and ensure that they are fit and suitable. <ESMA_QUESTION_320> Q321: Do you agree with the approach described above? <ESMA_QUESTION_321> Not entirely. For further detail, see our response to section 4.4 on market making for our thoughts on the application to liquid instruments, the various use of terms market making or similar in MiFID II and the requirement to have a sufficient number of market makers (particularly to Q286). Further it is not clear from ESMA s analysis if rebates will only be allowed for registered market makers, i.e. those participants who are subject to a written market making agreement; or whether it is possible for trading venues to offer other market making schemes to participants who may not be registered market makers (e.g. if they fulfil certain liquidity provision criteria of the scheme?). We would welcome clarification on how ESMA expects such schemes to be different from market making arrangements. Please also see our response to Q282 of the Discussion Paper. <ESMA_QUESTION_321> Q322: How could the principles described above be further clarified? <ESMA_QUESTION_322> Please see LSEG s response to section 4.4 and to Q321. <ESMA_QUESTION_322> Q323: Do you agree that and OTR must be complemented with a penalty fee? <ESMA_QUESTION_323> LSEG agrees with penalty fee as a useful tool for discouraging excessive OTR. 89

90 However, LSEG does not support mandatorily linking exceeding the OTR to fee structures. Each venue should decide if fees should be charged in case members cross the OTR threshold, taking into consideration the total capacity of the trading system to manage sustained message flow at high rates. A description LSE s 7 and Borsa Italiana s 8 order-to-trade tariffs can be found on its public price list (please see footnotes below). We note here that, in line with our response to Q290, LSEG is not in favour of an OTR regime applying to ETFs, and an OTR is not relevant to quote-driven or RFQ trading models, particularly for bond markets. <ESMA_QUESTION_323> Q324: In terms of the approach to determine the penalty fee for breaching the OTR, which approach would you prefer? If neither of them are satisfactory for you, please elaborate what alternative you would envisage. <ESMA_QUESTION_324> Both options have disadvantages. Instruments, trading platforms and trading roles are so different that a homogenous approach, such as that proposed by Option B, could be difficult to design and implement. In addition to this, the usage of historical data could divert HFTs to venues with higher historical OTR. Although Option A might be a better approach, it would be difficult to implement it as it is problematic to demonstrate that fees effectively discourage members or participants from systematically breaching the OTR threshold. Please also see LSEG s response to Q296 and Q298. <ESMA_QUESTION_324> Q325: Do you agree that the observation period should be the same as the billing period? <ESMA_QUESTION_325> The observation period should be daily as a longer period could smooth spikes too much. The billing system can then charge according to a different regular time-period, on the basis of daily data. We note here that, in line with our response to Q290, LSEG is not in favour of an OTR regime applying to ETFs, or to quote-driven or RFQ trading models operated by trading venues particularly for bond markets. <ESMA_QUESTION_325> Q326: Would you apply economic penalties only when the OTR is systematically breached? If yes, how would you define systematic breaches of the OTR? <ESMA_QUESTION_326> In case the OTR is accompanied by fees, it should be the trading venue to decide the fee structure, i.e. charging each time the ratio has been breached and/or add additional fees if this happens on a regular basis, and/or waive the fee in case of sporadic breach of the threshold. <ESMA_QUESTION_326> Q327: Do you consider that market makers should have a less stringent approach in terms of penalties for breaching the OTR? 7 LSE Price List, Page 7, note 2: 20December% pdf 8 Borsa Italiana Price List, Page 21, paragraph 10: 90

91 <ESMA_QUESTION_327> Yes, those market makers subject to a market making agreement are required to make continuous twoway quotes; for that reason, they should be exempt from the OTR. <ESMA_QUESTION_327> Q328: Please indicate which fee structure could incentivise abusive trading behaviour. <ESMA_QUESTION_328> Beyond the matters discussed above, we are not aware of any practice that needs regulatory action. <ESMA_QUESTION_328> Q329: In your opinion, are there any current fee structures providing these types of incentives? Please elaborate. <ESMA_QUESTION_329> Beyond the matters discussed above, we are not aware of any practice that needs regulatory action. <ESMA_QUESTION_329> 4.8. Tick sizes (Article 48(6) and Article 49 of MiFID II) Q330: Do you agree with the general approach ESMA has suggested? <ESMA_QUESTION_330> Yes, LSEG agrees with ESMA s approach. We believe that there are benefits to a harmonised regime across the EU; and in our view, an objective regime would set tick sizes to minimise transaction costs borne by investors, giving due consideration to spread-costs and price impact. <ESMA_QUESTION_330> Q331: Do you agree with adopting the average number of daily trades as an indicator for liquidity to satisfy the liquidity requirement of Article 49 of MiFID II? Are there any other methods/liquidity proxies that allow comparable granularity and that should be considered? <ESMA_QUESTION_331> No, LSEG disagrees with this approach where average number of daily trades is the metric used to model liquidity. Analysis reveals that the number of trades and the tick size are not independent variables; they are correlated, as a more granular tick size will result in a higher number of trades with smaller average trade sizes, whilst a less granular tick will lead to lesser trades at larger sizes. We strongly suggest that an independent variable should be used as an indicator for the liquidity. LSEG s view is that this should be the turnover, which is a measure for the level of trading demand for a particular instrument (and the product of number of trades and average trade sizes) as the appropriate metric that should be used, as it is observed as being unaffected by tick size change. Accordingly, LSEG strongly suggests that the turnover (measured as ADT) is used as the liquidity indicator. <ESMA_QUESTION_331> Q332: In your view, what granularity should be used to determine the liquidity profile of financial instruments? As a result, what would be a proper number of liquidity bands? <ESMA_QUESTION_332> 91

92 LSEG s views on the appropriate granularity of the liquidity profile and bands are developing subject to further analysis; however we suggest some considerations below. At minimum, there should be at least three liquidity bands: illiquid, liquid and the most liquid super liquid for a sufficient level of granularity; this would ensure that the blue chip end of liquid instruments have an appropriate calibrated tick regime, whilst mid-caps which are also liquid have a suitable table to cater to their particular characteristics. If this approach is taken, then MiFIR s definition of liquid instruments in Article 2(1)(17) can be a starting point to set these thresholds. More bands may be appropriate for further granularity. One possible approach could be for ESMA to consider the same liquidity bands it uses for calibrating the LIS thresholds and post-trade deferred publication regime for shares, ETFs and depositary receipts (as per section 3.1 of the Discussion Paper). This would lead to a harmonised approach across cash securities in the equity and equity-like world, and we can see the operational merits (in terms of simpler instrument reference data tables) if the same ADT bands are used. <ESMA_QUESTION_332> Q333: What is your view on defining the trade-off between constraining the spread without increasing viscosity too much on the basis of a floor-ceiling mechanism? <ESMA_QUESTION_333> The proposed approach seems acceptable. LSEG agrees that a theoretical floor-ceiling mechanism is helpful to define the appropriate viscosity for an instrument. However, we suggest that the floor-ceiling mechanism should be used as an initial assessment tool to assist in the definition of liquidity bands, but not employed on an ongoing systematic basis. In this regard, the use of ADT to model liquidity is a suitable proxy to use in the tick table. If a floor-ceiling is used on an on-going basis, there could be unnecessary complexity, opacity and administration in case trying to make adjustments related to the observed spread on one or more lit venues on an ongoing basis. <ESMA_QUESTION_333> Q334: What do you think of the proposed spread to tick ratio range? <ESMA_QUESTION_334> It is agreed that there should be a theoretical floor and ceiling, and intuitively the range proposed (between 1.4 and 2.5 for liquid instruments and 1.4 to 5 for less liquid instruments) appears reasonable. LSEG suggests, however, that further empirical analysis should be done in this regard. The academic literature on the appropriate spread-to-tick ratios for different securities is sparse. A high spread-to-tick ratio would point to a difficulty in price formation due to an unnaturally small tick size, while a low value would indicate that prevailing tick size is larger than the ideal spread. In addition to this, the market depth and historical volatility of a security would also need to be taken account to obtain a holistic picture of optimal spreads. LSEG notes that in the US, the SEC has adopted what is perceived to be a useful practice of creating and running Pilot Studies, where, after appropriate consultation and review/design, a proposed measure is introduced on a trial basis to determine effect and gather evidence of its impact. In particular, the SEC has used a pilot plan for the introduction of a tick size regime for small cap securities 9. LSEG also suggests a possible approach for ESMA to consider the implementation of the harmonised tick regime could through a pilot programme, at least initially. Such a pilot study should only be undertaken on 9 SEC Announces Order for Tick Size Pilot Plan: Pilot to Assess Impact of Tick Size on Market Quality for Small Cap Companies. 25 June 2014: 92

93 the basis of a clear and agreed proposition as to the measures to be applied (e.g. instruments to be considered and the exact regime), the outcomes to be analysed (impact on liquidity, prices and spreads), expected or required success factors (based on market quality and impact on participants). We understand the challenge of introducing pilots through secondary legislation in the EU; so it may be appropriate for ESMA to consider a limited set of instruments to initially include in the tick regime, and review the impact of the regime within a year of implementation to assess if the regime is fit to extend across a wider range of instruments or if any changes are needed. <ESMA_QUESTION_334> Q335: In your view, for the tick size regime to be efficient and appropriate, should it rely on the spread to tick ratio range, the evolution of liquidity bands, a combination of the two or none of the above? <ESMA_QUESTION_335> In LSEG s view, price and liquidity, with annual revisions (or semi annual revisions for ETFs) based on the liquidity of the instrument represent two criteria that (jointly) are able to provide a reasonable trade-off between simplicity of application of the new regime and robustness of the solution against excessive viscosity or narrow ticks. Principally, the liquidity bands in the tick size regime should be based on some quantitative assessment that confirms the apparent negative correlation between liquidity and spread, and therefore the regime should be able to rely solely on the liquidity bands. The addition of a third criteria based on the spread to tick ratio, while theoretically intuitive, may increase the complexity, opacity and administrative needs of the regime, and to ex-ante determination of acceptable thresholds. Thus, LSEG favours the approach defined under Option 1 based on a liquidity (ADT)/ price tick table, where such ratio is embedded in the initial determination and methodology to set the table. <ESMA_QUESTION_335> Q336: What is your view regarding the common tick size table proposed under Option 1? Do you consider it easy to read, implement and monitor? Does the proposed two dimensional tick size table (based on both the liquidity profile and price) allow applying a tick size to a homogeneous class of stocks given its clear-cut price and liquidity classes? <ESMA_QUESTION_336> LSEG: The table is easy to read, implement and monitor. As per our response to Q331, average daily turnover (ADT) should be used as liquidity metric as opposed to number of trades. Turnover is observed to be independent to the tick size (whilst number of trades is not), and is unaffected by a change in the tick regime. <ESMA_QUESTION_336> Q337: What is your view regarding the determination of the liquidity and price classes? <ESMA_QUESTION_337> LSEG s view is that the price classes are well defined and sufficiently granular. As per our response to Q331, average daily turnover (ADT) should be used as liquidity metric as opposed to number of trades. Turnover is observed to be independent to the tick size (whilst number of trades is not), and is unaffected by a change in the tick regime. As per our response to Q332, our views on the appropriate granularity of the liquidity profile and bands are developing subject to further analysis. In our view, there should be at least three liquidity bands: illiquid, liquid and the most liquid super liquid ; using the same ADT classes as for the determination of LIS may also be considered a possible approach. <ESMA_QUESTION_337> 93

94 Q338: Considering that market microstructure may evolve, would you favour a regime that allows further calibration of the tick size on the basis of the observed market microstructure? <ESMA_QUESTION_338> LSEG: As with any such innovations and reform, it is reasonable to expect a review process that would check the impact and effectiveness of the changes introduced. Therefore, potential for further calibration should be maintained and re-calibration made in case rigorous analysis and sentiment indicates such a requirement. In any case, the liquidity of an instrument should be reviewed annually (or semi-annually for ETFs) to determine its appropriate regime. Please also see our response to Q334 on the possibility for a pilot regime before wider implementation. <ESMA_QUESTION_338> Q339: In your view, does the tick size regime proposed under Option 1 offer sufficient predictability and certainty to market participants in a context where markets are constantly evolving (notably given its calibration and monitoring mechanisms)? <ESMA_QUESTION_339> LSEG: At this stage, it seems that the mechanism proposed in Option 1 is sufficient to adapt to evolving markets, subject to the liquidity measure being changed to ADT and a commitment to review the impact and effectiveness of the changes. Regarding newly admitted instruments, there should be sufficient evidence from similar free float market capitalisation instruments that once established, turnover velocity within a market is typically consistent and therefore the liquidity band can be easily assigned based on the observations of a number of similar instruments. Furthermore, the initial period of trading in a newly admitted instrument is typically more active than subsequent periods and therefore would be inappropriate to assign a tick size based on a relatively abnormal trading period. <ESMA_QUESTION_339> Q340: The common tick size table proposed under Option 1 provides for re-calibration while constantly maintaining a control sample. In your view, what frequency would be appropriate for the revision of the figures (e.g., yearly)? <ESMA_QUESTION_340> LSEG: An annual review is appropriate for shares; a shorter period may be required for ETFs, consistent with the period used to determine the liquidity, LIS thresholds and post-trade transparency regime (6 months, based on previous 6-12 months data). <ESMA_QUESTION_340> Q341: In your view, what is the impact of Option 1 on the activity of market participants, including trading venue operators? To what extent, would it require adjustments? <ESMA_QUESTION_341> Impact on tick sizes of individual instruments Based on the experience of the FESE tables and consideration of Option 1 as proposed, there would seem to be a suitable balance of instruments that will stay on the same tick size and others that would change tick sizes in order to establish rigorous quantitative analysis of the impact and effectiveness of the new regime. In line with our response to Q331, we note again that the appropriate metric for liquidity should be the turnover (based on ADT), which is a measure for the level of trading demand for a particular instrument 94

95 (and the product of number of trades and average trade sizes) as the appropriate metric that should be used, as it is observed as being unaffected by tick size change. Impact on market participants Given that Option 1 has been created based on the principles and mechanism already established with the FESE tables mechanism, it would seem that there should be a minimal impact on the market. LSEG notes here that a trading venue will usually consult with its members prior to implementing any significant changes to tick sizes to allow all market participants to provide feedback and also indicate any development and testing timelines. Impact on trading venues From a technical perspective, LSEG does not envisage any problems in implementing the proposal in Option 1 for LSEG s trading venues for the instruments considered (shares, DRs and ETFs). From an operational perspective, the introduction of additional criteria (e.g. differentiation of tick table based on the level of liquidity of individual shares) may require some adjustments to the trading systems currently in use. However, so long as the tick table: i) is fixed in advance: ii) is not subject to excessive frequent changes, and; iii) not subject of ad hoc adjustments for each individual share (e.g. the solution proposed under Option 2), we believe that current systems can adapt to the new regime, and that this is indeed the preferable option. LSEG also notes that venues would require such market structure changes and these would have to be planned in advance into the development cycle. LSEG is of the view that all trading venues (and those firms trading on SIs/OTC) should follow the tick regime precisely so that market participants can be confident that, for any individual tradable instrument, every venue will offer the same price increments. Some participants have argued that it should be the primary listing venue that picks an appropriate spread-tick ratio and tick table that suits their markets (and instruments) the best. Our initial view on this is that it is potentially confusing and not consistent with the uniformity and harmonisation sought through MiFID II. <ESMA_QUESTION_341> Q342: Do you agree that some equity-like instruments require an equivalent regulation of tick sizes as equities so as to ensure the orderly functioning of markets and to avoid the migration of trading across instrument types based on tick size? If not, please outline why this would not be the case. <ESMA_QUESTION_342> In principle, yes. LSEG agree that ETFs (and other ETP products, including ETCs, ETNs etc.) and depositary receipts should be subject to the same regulation. In our view, there should also be parity across the trading venues of types in these instruments (i.e. RMs and MTFs for equity like instruments) and Systematic Internalisers and OTC; to ensure that the model for price formation across all platforms is consistent, especially if SIs can offer price improvement in certain situations. <ESMA_QUESTION_342> Q343: Are there any other similar equity-like instruments that should be added / removed from the scope of tick size regulation? Please outline the reasons why such instruments should be added / removed? <ESMA_QUESTION_343> LSEG: Temporary equity lines should be assigned the same tick table as the related primary instrument, so even if the price of the temporary instrument is different, the tick as a proportion of the price will be broadly equivalent. Please also see our response to Q

96 <ESMA_QUESTION_343> Q344: Do you agree that depositary receipts require the same tick size regime as equities? <ESMA_QUESTION_344> Yes, LSEG agrees. <ESMA_QUESTION_344> Q345: If you think that for certain equity-like instruments (e.g. ETFs) the spread-based tick size regime 10 would be more appropriate, please specify your reasons and provide a detailed description of the methodology and technical specifications of this alternative concept. <ESMA_QUESTION_345> In LSEG s view Option 1 is the preferable approach (with ADT as the appropriate metric for liquidity), so long as the tick table: i) is fixed in advance: ii) is not subject to excessive frequent changes, and; iii) not subject of ad hoc adjustments for each individual share (e.g. the solution proposed under Option 2). <ESMA_QUESTION_345> Q346: If you generally (also for liquid and illiquid shares as well as other equity-like financial instruments) prefer a spread-based tick size regime11 vis-à-vis the regime as proposed under Option 1 and tested by ESMA, please specify the reasons and provide the following information: <ESMA_QUESTION_346> In LSEG s view, Option 1 is the preferable approach (with ADT as the appropriate metric for liquidity). <ESMA_QUESTION_346> Q347: Given the different tick sizes currently in operation, please explain what your preferred type of tick size regulation would be, giving reasons why this is the case. <ESMA_QUESTION_347> LSEG s preferred approach is specified in paragraph 48(ii) on p.299 of the Discussion Paper (i.e. a tick size regime, where a tick size is formed based on price and liquidity, with similarities to the regime for equities ). For DRs, the same regime as shares should apply, i.e. Option 1 with ADT as the appropriate liquidity metric. For ETFs (including ETCs and ETNs), ESMA should develop a similar regime, based on similar price bands and ADT thresholds for liquidity. For ETFs in particular, we note that some fixed income ETFs are generally traded in large quantities, but are not traded as frequently as shares (e.g. some may not be traded daily). This may result in an artificially low ADT, but does not necessarily mean that they are not liquid or trade only in small sizes. We suggest that ESMA should be mindful of this when determining the tick regime and may allow flexibility to trading venues to set customized ticks for those particular fixed income ETFs (for instance liquidity ETFs on overnight interest rate) LSEG believes that all venues should follow the tick regime precisely so that market participants can be confident that, for any individual tradable instrument, every venue will offer the same price 10 Please see the description of Option 2 regarding tick sizes below. 11 Please see the description of Option 2 regarding tick sizes below. 96

97 increments. Some participants have argued that it should be the primary listing venue that picks an appropriate spread-tick ratio and tick table that suits their markets (and instruments) the best. Our initial view on this is that it is potentially confusing and inconsistent with the uniformity and harmonisation sought through MiFID II. <ESMA_QUESTION_347> Q348: Do you see a need to develop a tick size regime for any non-equity financial instrument? If yes, please elaborate, indicating in particular which approach you would follow to determine that regime. <ESMA_QUESTION_348> In LSEG s view, it would be premature to develop a tick size regime for non-equity instruments, and we agree with ESMA s approach. <ESMA_QUESTION_348> Q349: Do you agree with assessing the liquidity of a share for the purposes of the tick size regime, using the rule described above? If not, please elaborate what criteria you would apply to distinguish between liquid and illiquid instruments. <ESMA_QUESTION_349> Whilst LSEG s preferred approach is Option 1, in our view the concept of "liquid instrument" should reflect the new definition in MiFIR Article 2(1)(17) which ESMA defines further in Section 3.1 of the Consultation Paper, instead of linking this to the definition of liquidity under MiFID I. We understand that ESMA conducted its impact assessment based on the MiFID I definition, but for consistency it would be suitable if the MiFID II definition is used. LSEG also suggests that two classes of liquidity are unlikely to be sufficient to give the appropriate variation of tick size required within a price range. For example, the most liquid instruments in the liquid classification are likely to find themselves significantly constrained if employing the current FESE 2 tick sizes as proposed. The liquidity bands used in Option 1 (using ADT as the suitable proxy) offers a smoother transition for an instrument changing from illiquid and liquid and vice versa <ESMA_QUESTION_349> Q350: Do you agree with the tick sizes proposed under Option 2? In particular, should a different tick size be used for the largest band, taking into account the size of the tick relative to the price? Please elaborate. <ESMA_QUESTION_350> LSEG: The principles behind this approach are consistent and valid, but a modified Option 1 (to assess liquidity based on ADT) is our preferred approach. <ESMA_QUESTION_350> Q351: Should the tick size be calibrated in a more granular manner to that proposed above, namely by shifting a band which results in a large step-wise change? <ESMA_QUESTION_351> The granularity proposed would be suitable, but a modified Option 1 (to assess liquidity based on ADT) is LSEG s preferred approach. <ESMA_QUESTION_351> Q352: Do you agree with the above treatment for a newly admitted instrument? Would this affect the subsequent trading in a negative way? <ESMA_QUESTION_352> LSEG disagrees with the approach. The initial period of trading in a newly admitted instrument is typically more active than subsequent periods and therefore would be inappropriate to assign a tick size based on a relatively atypical trading period. 97

98 As an alternative, a trading venue should consider sufficient evidence from a similar free float market capitalisation instruments on the same market. Once established, a turnover velocity within a market is typically consistent and therefore the liquidity band can be easily assigned based on the observations of a number of similar instruments. <ESMA_QUESTION_352> Q353: Do you agree that a period of six weeks is appropriate for the purpose of initial calibration for all instruments admitted to the pan-european tick size regime under Option 2? If not, what would be the appropriate period for the initial calibration? <ESMA_QUESTION_353> Please see our response to Q352. LSEG disagrees with this approach. Trading is typically more active in the initial period after admission and therefore would not be appropriate for calibration. A more appropriate method would be for the venue/issuer to assume a proper level of liquidity based on the level of liquidity of comparable shares by free float on the same market. <ESMA_QUESTION_353> Q354: Do you agree with the proposal of factoring the bid-ask spread into tick size regime through SAF? If not, what would you consider as the appropriate method? <ESMA_QUESTION_354> No, LSEG disagrees. The additional adjustment - on top of the liquidity criteria and price - based on the spread to tick ratio appears unnecessarily complex and potentially opaque. It would appear to be much more suitable to define a distinct set of tables from which to select for each individual instrument based on some robust analysis of the observed relationship between liquidity and spread from existing data. This would be similar to Option 1 proposed by ESMA, although, as per our response to previous questions, the liquidity should be based on ADT and not the average daily number of trades. <ESMA_QUESTION_354> Q355: Do you agree with the proposal to take an average bid-ask spread of less than two ticks as being too narrow? If not, what level of spread to ticks would you consider to be too narrow? <ESMA_QUESTION_355> LSEG: Whilst it is difficult to propose a target value for the spread to tick ratio at this stage, without having performed specific simulations, LSEG supports a ratio of between proposed under Option 1. We suggest that at the lower level (i.e. 1.4), a spread equal to 1 tick can be observed for 60% of the time; this seems sufficient to minimise risks of excessive 'overbidding' and ensuring a sufficient degree of viscosity. We are keen to understand why some market participants believe higher spread to tick ratios, e.g. 4-20, are important for efficient pricing for certain instruments or the characteristics of certain markets. However, rather than constantly assess the spread on an ongoing basis, it is advisable to instead agree the broad relationship between liquidity (ADT) and spread at the outset and monitor liquidity only. This would greatly reduce the potential for unnecessary complexity, uncertainty and opacity. <ESMA_QUESTION_355> Q356: Under the current proposal, it is not considered necessary to set an upper ceiling to the bid-ask spread, as the preliminary view under Option 2 is that under normal conditions the risk of the spread widening indefinitely is limited (and in any event a regulator may amend SAF manually if required). Do you agree with this view? If not, how would you propose to set an upper ceiling applicable across markets in the EU? <ESMA_QUESTION_356> No, LSEG does not agree with the Option 2 approach of SAF maintenance. 98

99 There are a number of drawbacks to an SAF approach. Apart from the being contrary to the objective of harmonisation that MiFID II seeks to achieve, assigning a SAF value by instrument seems a complex approach given its limited value as an attribute and the need to maintain it as a key reference data field. For some securities, the SAF approach is to likely result in an inefficient and disruptive annual reduction in tick size for a number of years. <ESMA_QUESTION_356> Q357: Do you have any concerns of a possible disruption which may materialise in implementing a review cycle as envisioned above? <ESMA_QUESTION_357> Please see our response to Q356. LSEG does not agree with the Option 2 approach. <ESMA_QUESTION_357> Q358: Do you agree that illiquid instruments, excluding illiquid cash equities, should be excluded from the scope of a pan-european tick size regime under Option 2 until such time that definitions for these instruments become available? If not, please explain why. If there are any equity-like instruments per Article 49(3) of MiFID II that you feel should be included in the pan-european tick size regime at the same time as for cash equities, please list these instruments together with a brief reason for doing so. <ESMA_QUESTION_358> LSEG: Re illiquid cash equities, Option 1 proposes an approach which can take these into account, so there is no obvious reason to exclude these securities. Depositary receipts should be subject to the same regime as shares, and ETFs should be subject to a similar regime. <ESMA_QUESTION_358> Q359: Do you agree that financial instruments, other than those listed in Article 49(3) of MiFID II should be excluded from the scope of the pan-european tick size regime under Option 2 at least for the time being? If not, please explain why and which specific instruments do you consider necessary to be included in the regime. <ESMA_QUESTION_359> LSEG: Shares, ETFs (including ETNs and ETCs) and DRs should be in scope. In our view, it would be premature to develop a tick size regime for non equity instruments, and we agree with ESMA s approach. <ESMA_QUESTION_359> Q360: What views do you have on whether tick sizes should be revised on a dynamic or periodic basis? What role do you perceive for an automated mechanism for doing this versus review by the NCA responsible for the instrument in question? If you prefer periodic review, how frequently should reviews be undertaken (e.g. quarterly, annually)? <ESMA_QUESTION_360> As in our responses above, LSEG disagrees with the approach in Option 2 and prefer a modified Option 1 where ADT is used as measure of liquidity. However, if Option 2 is considered, we believe that: a dynamic adjustment of tick in relation to the price movements should be applicable. In LSEG s views, this is not problematic to implement from a technical perspective; attribution of instruments to a specific liquidity category and application of SAF adjustment (under option 2) should be performed on a periodic basis. A yearly review should be acceptable, aligned with the yearly review of the liquidity definition, and assignment of instruments to the appropriate LIS liquidity band. A review of liquidity classes should in accordance to the liquidity definition in MiFIR Article 2(1)(17). 99

100 <ESMA_QUESTION_360> 100

101 5. Data publication and access 5.1. General authorisation and organisational requirements for data reporting services (Article 61(4), MiFID II) Q361: Do you agree that the guidance produced by CESR in 2010 is broadly appropriate for all three types of DRS providers? <ESMA_QUESTION_361> LSEG: Not for all 3 types of DRS providers. The CESR guidance was specifically designed for APAs and certain sections are inappropriate for ARMs as the roles of these two DRS are entirely different. Compatible with ARMs: Security, Operational Hours, Resources and contact arrangements, Recovery provisions, Correction of trade information, Monitoring and Conflict of interest However, we suggest separate and specific guidance for ARMs on: Dissemination, Identification of incomplete or potentially erroneous information and Regulatory reporting responsibilities. We believe that many of the responsibilities proposed for APAs in terms of monitoring accuracy of data submitted would be inappropriate for ARMs as the NCAs would have a greater role in policing the quality of transaction reporting data submitted by participants <ESMA_QUESTION_361> Q362: Do you agree that there should also be a requirement for notification of significant system changes? <ESMA_QUESTION_362> LSEG: Yes- it is sensible to notify the relevant NCA of significant system changes to DRS providers core systems. Additional guidance on what constitutes significant. We suggest significant system changes should mean changes that may impact output in a manner that is visible to users or user firms machines which may read the content. <ESMA_QUESTION_362> Q363: Are there any other general elements that should be considered in the NCAs assessment of whether to authorise a DRS provider? <ESMA_QUESTION_363> LSEG: Yes we suggest that part of the assessment should be that the NCA is satisfied that the staff of the CTP and APA can demonstrate appropriate experience of operating in a regulated environment <ESMA_QUESTION_363> 101

102 5.2. Additional requirements for particular types of Data Reporting Services Providers Q364: Do you agree with the identified differences regarding the regulatory treatment of ARMs. <ESMA_QUESTION_364> LSEG: Not for all 3 types of DRS providers. The CESR guidance was specifically designed for APAs and certain sections are inappropriate for ARMs as the roles of these two DRS are entirely different. Compatible with ARMs: Security, Operational Hours, Resources and contact arrangements, Recovery provisions, Correction of trade information, Monitoring and Conflict of interest However, we suggest separate and specific guidance for ARMs on: Dissemination, Identification of incomplete or potentially erroneous information and Regulatory reporting responsibilities. We believe that many of the responsibilities proposed for APAs in terms of monitoring accuracy of data submitted would be inappropriate for ARMs as the NCAs would have a greater role in policing the quality of transaction reporting data submitted by participants. <ESMA_QUESTION_364> Q365: What other significant differences will there have to be in the standards for APAs, CTPs and ARMs? <ESMA_QUESTION_365> LSEG: For CTPs and APAs: It will be important that they (and all trading/execution venues and investment firms, including Systematic Internalisers and firms trading OTC) are required to follow the harmonised and standardised approach to trade reporting in the standards mandated under MiFIR Article 7(2) and the market Model Typology (MMT) mentioned by ESMA in relation to Question 80 of this DP. Only adoption of these standards will allow any Consolidated Tape to be developed. For ARMs: Dissemination ARMs must be able to ensure they can route the transaction reports to the correct NCA Identification of incomplete or potentially erroneous information ARMs should ensure they have appropriate validation of data. In particular, any Trade Repository that wishes to become an Approved Reporting Mechanism should be able to demonstrate appropriate controls to differentiate between EMIR reporting requirements and MiFIR reporting requirements, where they differ. Regulatory reporting responsibilities <ESMA_QUESTION_365> 102

103 5.3. Technical arrangements promoting an efficient and consistent dissemination of information Machine readability Article 64(6), MiFID II Q366: Do you agree with the proposal to define machine-readability in this way? If not, what would you prefer? <ESMA_QUESTION_366> LSEG: Yes. <ESMA_QUESTION_366> 5.4. Consolidated tape providers Q367: Should the tapes be offered to users on an instrument-by-instrument basis, or as a single comprehensive tape, or at some intermediate level of disaggregation? Do you think that transparency information should be available without the need for value-added products to be purchased alongside? <ESMA_QUESTION_367> LSEG: General Comments on the Consolidated Tape Under the terms of the MiFID level I text, the CT is mandated, but with the exception of Recital 117 in MiFID, there is no practical description of what the CT is and how it operates in a market context. The architecture of how the industry or any other potential provider might move from the concept of the CT to the final delivery is missing and this makes any constructive input problematic. In that respect, we make the following points for consideration: LSE has always strongly maintained its commitment to the dissemination of high-quality, standardised and harmonised post-trade data to the markets, covering both the on-venue and offvenue/otc business- only then will investors be able to see what is trading where and regulators will understand the true dynamics of the market. High-quality and robust post-trade data is necessary for our customers for a range of essential activities. LSEG has always operated to the highest standards in producing its data using rigorous checking processes and the latest technology. Recital 117 of MiFID makes clear that the policy objective is to deliver a method for investors to see a consolidated view of the post trade activity in the markets, and across all markets and OTC. There has been debate on whether the policy objective requires that the Consolidated tape be provided by a single entity. Given that across the EU there are, on any reasonable estimate 100,000 reportable securities (and when non-equities are added, this will increase by many factors), we suggest it is difficult to see that a single entity will have the practical resources or capability to provide such a mechanism. It is more likely, as suggested by Recital 117 and the legislative text, to be made up of a number of CT providers operating to the same standards, (including all venues and investment firms adopting the MMT for all business and adopting the universal time stamping conventions for orders and trades) but operating in different sectors (or the same, where competition allows) to provide what investors can see as a quality marked feed of regulated post trade data- the tape of record as some would have it. The sectoral division of the CT providers could be by asset classes and/or by geography, liquidity of trading, asset classes or types, market capitalisation- these are some of the same concepts that underlie the disaggregation discussions in Questions of the DP. It is important to set out this approach, because it will inform how industry and providers should approach delivery of the CT in good time before the review by the Commission required under MiFID 90(2) in August

104 These comments inform our approach to the data consolidation and dissemination questions LSEG: General Comments on disaggregation by CTP and by venues: 1. There is an assumption implicit in the ESMA DP that disaggregation will necessarily reduce costs to users. We would caution against creating an expectation that disaggregating, whether at CTP or trading venue/investment firm level will always lead to a similar reduction in charges. This is because there will be a number of factors that drive the charges of the trading venues and investment firms/mtfs, relating to the cost of collection, production, dissemination and distribution at the level of the relevant trading venue, investment firm or distribution mechanism. Volume of information provided is, therefore, not the only driver for fee levels. 2. For all but the most basic views on a terminal, the exercise of disaggregation is not a simple matter of switch a receiver or filter to increase or reduce the size of the messages- systems will need to be reengineered. Many of the costs incurred in accessing and subscribing to data would continue or could increase these costs would include the costs of providing additional technology architecture to disseminate granular sets of data: Additional feed handlers to be built and maintained by directly connected end users and information vendors Additional permissioning systems and controls to be maintained by end users and information vendors Additional product and administration fees levied by information sources to provide and support multiple new data sets For instance, many of the larger or more complex customers of venues and data vendors will already have feed handlers, systems and arrangements set up to receive and manage established feeds, and there will be a cost to their revision. Similarly, data vendors may have systems that have been established to manage particular types and content of feed and these systems may not, without costly reengineering, readily accommodate a revised data feed or series of feeds. 3. As client requirements develop, venues should be encouraged to provide solutions to changing data consumption patterns, but in obliging venues to do so, ESMA risks inadvertently increasing the overall cost of data to users. The potential events described above reflect the reality of an environment where exchange data fees account for only a small proportion of overall data access and subscription costs (8-15% according to Oxera) and that the largest costs for end users are actually the products and systems that they subscribe to from data vendors/aggregators or build themselves. 4. It is also important to bear in mind that exchanges in most Member States, in fulfilling their important role of bringing together investors and issuers, provide an infrastructure that offers services to a wide variety of investor types (from private individuals through to pension funds and insurance companies) and for a broad range of issuers, from the largest/ highly liquid to the small cap/sme that are less liquid. 5. We are concerned that disaggregation of, for instance, data services, should not reduce the effectiveness of that infrastructure or the visibility that exchanges provide in relation to the small cap and SME sectors of the market, where the trading volumes are lower so the profile they would receive would be otherwise reduced. The small cap and SME markets are especially important in the current post-crisis agenda of supporting jobs and growth. Specific response to Q367 Instrument by Instrument /Single Tape or Intermediate Level? LSEG: In our view, disaggregation at some intermediate level. LSEG believes that the packaging of content should be informed by what customers need, and it is these needs that should drive the level of disaggregation of the CT; apart perhaps from regulators, it is difficult to envisage any users who will want, or have the technical capacity or appetite, to take the complete CT. 104

105 Similarly, offering a tape on an instrument-by-instrument basis will present very significant technical challenges for the largest exchanges, requiring major system re-configuration, with potential negative impacts on system performance. For instance, although some 3500 UK instruments may be traded on the LSE systems, there are some 30,000 instruments available for on-exchange trade reporting- replicating this scale for the EU would create a significant technical challenge if individual securities feeds had to be managed. Not all customers have a single outlook that drives their investment decisions. Customers may have geographic, market sector, liquidity, market capitalisation or product based outlooks. For this reason, we suggest that the CT will need to be flexibly organised, so that a consolidated tape could be available on the following bases (all these options assume the adoption of the harmonised and standardised approach to trade reporting in the standards mandated under MiFIR Article 7(2) and the market Model Typology (MMT) mentioned by ESMA in relation to Question 80 of this DP): country-by-country market model market capitalisation groupings liquidity bands (ADT) market segments CT information as stand-alone? LSEG: Again, this should be driven by customers, but in any event, it is difficult to see how CT rules could be permitted that mandated the bundling of non-transparency information with the transparency information- it is not consistent with the policy objectives of MiFIR. Post-trade data should be made available as a stand alone content set with no obligation to take (either technically or commercially) any other added-value content. We dis-aggregated this content in 2010, recognising the need to provide high quality real time post-trade content as a separate data package for end users. However, nothing should prevent the CT offering additional services alongside. <ESMA_QUESTION_367> Q368: Are there other factors or considerations regarding data publication by the CTP that are not covered in the standards for data publication by APAs and trading venues and that should be taken into account by ESMA? <ESMA_QUESTION_368> LSEG: Yes- although the standards deal with the costs of dissemination, the standards should deal with the allocation of the cost that will fall on all levels of users. This will be particularly important given the lack of detailed specification of the way in which a CT will operate commercially. We suggest that the cost allocation should be a mix of the CESR standard (namely that the CT and APAs, (and market operators/mtfs) charge on a non-discriminatory, reasonable commercial basis) but also that the principle of reporter pays is also applied, as it is currently. Thus, a trading venue, APA or CTP should be entitled to charge the entity that is sending the trade report a fee for the processing of the trade report. This will ensure an appropriate allocation of the costs of the various services across the users in terms of the market participants/investors trading and reporting and the investors and other data users who view and use the CT data. <ESMA_QUESTION_368> Q369: Do you agree that CTPs should be able to provide the services listed above? Are there any others that you think should be specified? <ESMA_QUESTION_369> LSEG: Yes. We do not believe that the CT should be restricted in the additional services it is able to provide, subject to the following conditions: 105

106 (a) provision of any additional services does not inhibit its ability to provide the CTP service or present it with a conflict of interest that it cannot effectively manage and (b) it is clear to which of the services the requirement to charge on a reasonable commercial basis is applied- it would be appropriate for that to be applied to the non-ctp activities. On that basis, we do not think it is necessary to set an exhaustive list of services <ESMA_QUESTION_369> 5.5. Data disaggregation Q370: Do you agree that venues should not be required to disaggregate by individual instrument? <ESMA_QUESTION_370> LSEG: We repeat our General Comments on disaggregation by CTP and venues given in response to Q367 above. By Individual Instrument? LSEG: We agree that venues should not be required to disaggregate by individual instrument, for the following reasons. LSEG believes that the packaging of content should be informed by what customers need, and it is these needs that should drive the level of disaggregation of the CT; apart perhaps from regulators, it is difficult to envisage any users who will want, or have the technical capacity or appetite, to take the complete CT. Similarly, offering a tape on an instrument-by-instrument basis will present very significant technical challenges for the largest exchanges and customers, requiring major system re-configuration, with potential negative impacts on system performance. For instance, although some 5000 instruments may be traded on the LSE systems, there are some 27,400 instruments available for on-exchange trade reporting- replicating this scale for the EU would create a significant technical challenge if individual securities feeds had to be managed. <ESMA_QUESTION_370> Q371: Do you agree that venues should be obliged to disaggregate their pre-trade and posttrade data by asset class? <ESMA_QUESTION_371> LSEG: See our General Comments on disaggregation in the response to Question 367. LSEG believes that the packaging of content should be informed by what customers need, and it is these needs that should drive the level of disaggregation of the venues data if customers want pre and post trade data disaggregated by assets class, venues should be prepared to explore that service. <ESMA_QUESTION_371> Q372: Do you believe the list of asset classes proposed in the previous paragraph is appropriate for this purpose? If not, what would you propose? <ESMA_QUESTION_372> LSEG: See our General Comments on disaggregation in the response to Question 367. As we believe this should be user driver it depends what they want, but if it becomes an obligation, it is an appropriate list. <ESMA_QUESTION_372> 106

107 Q373: Do you agree that venues should be under an obligation to disaggregate according to the listed criteria unless they can demonstrate that there is insufficient customer interest? <ESMA_QUESTION_373> LSEG: See our General Comments on disaggregation in the response to Question 367. Again, or views on disaggregation are that it should be driven by customer needs, so in principle we can accept an approach that would entail explaining why there was insufficient customer demand for further disaggregation. <ESMA_QUESTION_373> Q374: Are there any other criteria according to which it would be useful for venues to disaggregate their data, and if so do you think there should be a mandatory or comply-orexplain requirement for them to do so? <ESMA_QUESTION_374> LSEG: See our General Comments on disaggregation in the response to Question 367. LSEG believes that the packaging of content should be informed by what customers need, and it is these needs that should drive the level of disaggregation of the venues data <ESMA_QUESTION_374> Q375: What impact do you think greater disaggregation will have in practice for overall costs faced by customers? <ESMA_QUESTION_375> LSEG: See our General Comments on disaggregation in the response to Question 367 This issue is also related to the requirement for various entities to be required to charge on a reasonable commercial basis, discussed by ESMA in section 4.3 of the CP. Requirements for disaggregation and cost controls will only be effective if applied on a uniform basis, and it is clear to us that they are not- the Level I MiFID and MiFIR create an unfair and uncompetitive environment in this sensitive area. As the recent Oxera report (referenced by ESMA on page 219 of the CP) highlights the cost of exchange trading data licence fees represents between 8% and 15% of the total cost of accessing and subscribing to data, with data vendor services estimated to account for the remaining 65% to 80% - (page v) suggesting that the vast majority of the cost sits outside the scope of the charges levied by trading venues, MTFs, APAs or CTPs, all of whom are subject to the reasonable commercial basis requirements; data vendors are subject to no such requirement. In this context, vendors could include firms and banks who redistribute data to their customers. Without effectively addressing the other aspects of the cost chain (vendors charges, cost of IT connectivity infrastructure, cost of purchasing a display terminal), LSEG suggests this policy initiative may only be partially successful, leaving only additional costs for clients to access and manage the data, additional costs for trading venues to recover due to increased technical controls and technology infrastructure and potential reduction in the ability of exchanges to support their broader role. <ESMA_QUESTION_375> 5.6. Identification of the investment firm responsible for making public the volume and price transparency of a transaction (Articles 20(3) (c) and 21(5)(c), MiFIR) Q376: Please describe your views about how to improve the current trade reporting system under Article 27(4) of MiFID Implementing Regulation. 107

108 <ESMA_QUESTION_376> LSEG: We agree that a harmonised approach should be adopted. It is important that the same approach is adopted for the on-book, on venue trading as for non systematised, off-book and OTC reporting and in all instruments. <ESMA_QUESTION_376> 5.7. Access to CCPs and trading venues (Articles 35-36, MiFIR) Q377: Do you agree that exceeding the planned capacity of the CCP is grounds to deny access? <ESMA_QUESTION_377> Introduction: LSEG s general comments to section 5.7 on access to CCPs and trading venues (Articles 35-36, MiFIR) In the following questions, LSEG presents its expertise as an operator of open/ horizontal trading venues, CCPs and CSDs across the range of asset classes covered by MiFIR, EMIR and the CSD Regulation. Our views are common across all entities in the Group. The Group s trading venues and CCPs offer the choice to market participants to trade and clear products on the following platforms across different asset classes (both transferable securities and derivatives): Equities: o London Stock Exchange clears its order book securities (on SETS) through LCH.Clearnet Ltd and SIX X-Clear Ltd, and announced an agreement to clear through EuroCCP 12. Borsa Italiana (MTA) clears through CC&G. o Turquoise, the pan-european MTF clears equities through LCH.Clearnet, X-Clear and EuroCCP, and offers participants the ability to choose different CCPs in different countries. Fixed income: CC&G clears for MTS, MOT, EuroTLX, himtf and BrokerTech in a fully fungible manner. LCH.Clearnet SA clears for MTS, EuroMTS, BrokerTech and Tullett Prebon. In addition, LCH.Clearnet SA and CC&G operate an interoperability agreement for the clearing of Italian government bonds executed on MTS S.p.a. and Broker Tech and settled on Monte Titoli. Exchange traded derivatives (ETDs): o LSE Derivatives clears through LCH Clearnet Ltd and operates a linked order book for trading of Norwegian equity derivatives with Oslo Børs. Clearing is offered by an interoperability arrangement for members through both Oslo Clearing and LCH.Clearnet Ltd. o LCH.Clearnet SA also clears equity derivatives for Euronext and LCH.Clearnet Ltd clears interest rate derivatives for Nasdaq OMX NLX. o CC&G clears exchange traded derivatives for the IDEM, IDEX and AGREX Markets. OTC derivatives: LCH.Clearnet s OTC derivative services SwapClear, CDS.Clear and Forex- Clear comply with the open access provisions in Article 7 and 8 in EMIR. LSEG continues to support open access. Participant access and competition concerns have been recurrent themes in Europe, many of which were highlighted in the EU Commission s decision on the proposed merger between Deutsche Börse and NYSE-Euronext in February The purpose of open access is summarised in Recital 40 of MiFIR to empower EU investors through enhanced choice, better service, 12 EuroCCP to clear London Stock Exchange trades: 13 Commissioner for Competition Almunia stated that there are major barriers to entry in [exchange traded derivatives] markets due to the closed vertical silo. There is ample evidence that the opening of markets and an increase in competition has improved liquidity and has led to a drastic decrease in the cost of trading. 108

109 lower costs, greater capital efficiency, innovation, and to avoid the concentration of risk in closed Financial Market Infrastructures (FMIs). LSEG supports the general principles behind ESMA s approach in the Discussion Paper to define a regime that recognises the benefits of the choice that access brings to investors, whilst mitigating any undue risks in the process. In answering the questions in this section, LSEG s main suggestion is that concerns around capacity, cost, on-boarding of new trading and clearing members, conflicts of laws etc. are, in our experience, in the natural course of business for commercially operating trading venues and CCPs with horizontal models. The prescribed template for open access should be built on current open arrangements where they work. -- Specific response to Q377 LSEG s view is that for all financial instruments covered under MiFIR, exceeding planned capacity (to do with volumes) should only be a ground to deny access if accepting the request would lead to a CCP operating manifestly above its planned and expected future capacity in a way that poses a threat to its viability, based on fair and reasonable criteria. In LSEG s experience, dealing with additional volumes is in within the normal of business of operating a CCP. So it should not normally be a problem. Further, LSEG agrees with ESMA s assessment in paragraph 15 on p.344 of the Discussion Paper, that CCPs authorised under EMIR should ordinarily have scalable capacity. A CCP s robustness and scalability is essential in performing the clearing functions with a trading venue and its clearing members. Finally, LSEG notes a CCP has a commercial incentive to attract greater clearing volumes (and therefore, greater revenues), subject to its regulatory requirements and risk management framework. If this requires the addition of further capacity (in the form of storage, servers etc.), it is not difficult for a CCP to do so. Any reason to deny access to do with capacity should also apply in a non-discriminatory and fair manner to all trading venues currently accessing a CCP or wishing to access it for clearing in the future (including how planned and/or future capacity is allocated). A CCP should also be able to prove to a requesting party why an access request would manifestly exceed its planned and future capacity, when denying access on this basis. Further, a requesting trading venue should have the opportunity to pay for the costs of adding capacity (if evidenced and fair), should the cost of adding capacity be the reason to deny access. <ESMA_QUESTION_377> Q378: How would a CCP assess that the anticipated volume of transactions would exceed its capacity planning? <ESMA_QUESTION_378> LSEG: There are a number of ways CCPs can assess the anticipated volume of transactions: If the trading venue operates a pre-existing market in the financial instrument it would like to clear at the CCP, the CCP will conduct a forecast as to the potential share of the pre-existing market at the trading venue that may be routed to the CCP in normal and peak markets, and identify any factors that may cause significant growth in the volume of the financial instrument (such as, for example, portfolio margining at the CCP). In the event that the trading venue is looking to facilitate trading for a financial instrument for which the trading venue has not previously facilitated, the CCP will look to anticipate volume by identifying the nearest comparable instruments/marketplaces, liquidity of the instrument and past trading volumes under similar market conditions. A CCP would also gauge the likely demand for the new service by asking its existing clearing members; which can help it make an assessment of the new volumes likely to be cleared through its systems. 109

110 The requesting trading venue would also make an assessment of its anticipated volumes, which a CCP would take into account. For transferable securities, if additional interoperability links are being considered as part of an access request, a CCP would assess projected volumes from each link. ESMA and NCAs should be mindful that when a new trading venue wishes to offer a service in competition to an incumbent trading venue by clearing through a common CCP, this situation may lead to a shift of demand from the incumbent venue to the competitor, rather than an increase in total demand. In this case, capacity itself will not be a reason to deny access, as the CCP will be clearing the same or similar volumes, just fragmented between 2 or more trading venues. <ESMA_QUESTION_378> Q379: Are there other risks related to the anticipated volume of transactions that should be considered? If so, how would such risks arise from the provision of access? <ESMA_QUESTION_379> LSEG: A CCP needs to ensure mechanisms to ensure settlement for those financial instruments with physical deliveries, e.g. links to CSDs for relevant equity, fixed income and physically settled ETDs, and warehouse arrangements for physically settled commodity derivatives. Once again, this is within the normal course of operation of a CCP, as it would already have the required have arrangements (e.g. links to CSDs) and contracts (e.g. with warehouses to store commodities) in place for clearing physically settled products, with adequate spare capacity. For commodity derivatives in particular we suggest that, in general, positions are usually closed during their lifetime. This would mitigate any risk of physical settlement from being an undue risk. <ESMA_QUESTION_379> Q380: Do you agree that exceeding the planned capacity of the CCP is grounds to deny access? <ESMA_QUESTION_380> ESMA s use of the term users needs to be clarified. If ESMA is referring to clearing members, then LSEG believes that exceeding capacity on the basis of new CCP members (in terms of a greater quantity) is not a valid ground to deny access to a trading venue, for both transferable securities and derivatives. In our view, CCPs should be scalable in terms of their ability to offer ISA or omnibus accounts to new clearing members. Certainly, this is not an issue for LSEG s CCPs. <ESMA_QUESTION_380> Q381: How would a CCP assess that the number of users expected to access its systems would exceed its capacity planning? <ESMA_QUESTION_381> As indicated in our comments to Q380, LSEG does not believe an additional number of users represents a material or undue risk to a CCP. A CCP would assess its anticipated new users in a similar way it would assess anticipated volumes. Please see our comments to Q378 for further details. <ESMA_QUESTION_381> Q382: Are there other risks related to number of users that should be considered? If so, how would such risks arise from the provision of access? <ESMA_QUESTION_382> 110

111 LSEG: No. From the perspective of a CCP, as indicated in our comments to Q380, LSEG does not believe that an additional number of users represents a material or undue risk to a CCP (in terms of scalability of account structures e.g. ISA or omnibus, or automation processes e.g. STP). <ESMA_QUESTION_382> Q383: In what way could granting access to a trading venue expose a CCP to risks associated with a change in the type of users accessing the CCP? Are there any additional risks that could be relevant in this situation? <ESMA_QUESTION_383> In LSEG s view, none; there are no additional risks. A CCP would, as in its normal course of operation, evaluate any new clearing members according to its eligibility and risk management criteria; as ESMA notes in paragraphs 18 and 19 of the Discussion Paper. As ESMA also notes in paragraph 20, members of the new trading venue connecting to a CCP through an access request could either clear directly, through an existing clearing member of the CCP or indirectly. In practice, a trading venue will have rules (when on-boarding new members or offering new markets/ clearing solutions) that require firms to have the appropriate clearing and settlement arrangements in place. As we also note in our comments to Q380, in our view, CCPs should be scalable in terms of their ability to offer ISA or omnibus accounts to new clearing members. Certainly, this is not an issue for LSEG s CCPs. <ESMA_QUESTION_383> Q384: How would a CCP establish that the anticipated operational risk would exceed its operational risk management design? <ESMA_QUESTION_384> A CCP s operational risk structure would require a requesting trading venue to meet certain operational criteria (for example, straight-through-processing, connectivity etc.); if an access request does not fulfil these criteria, it may be a reason for a CCP to deny access. Within these criteria, a CCP will look at the specifications of the market offered by the trading venue requesting access, including number and complexity of products; product delivery/settlement mechanism; post trade events such as corporate actions; trading hours and trading calendar. In practice, a CCP establishes a certification protocol that a trading venue must satisfy prior to being operationally live with the CCP. This includes, amongst other criteria, a certification test will ensure that connectivity has been established, that the trading venue has the proper technological infrastructure to support the instruments it wishes to clear at the CCP and that the trading venue has the appropriate processes to facilitate straight-through-processing with the CCP. The trading venue would also be required to enter in an operational contract with the CCP. <ESMA_QUESTION_384> Q385: Are there other risks related to arrangements for managing operational risk that should be considered? If so, how would such risks arise from the provision of access? <ESMA_QUESTION_385> LSEG s view is that both the CCP and the requesting trading venue should have sufficient and qualified staff, and management oversight, and be able to demonstrate this. <ESMA_QUESTION_385> Q386: Given there will be costs to meeting an access request, what regard should be given to those costs that would create significant undue risk? <ESMA_QUESTION_386> 111

112 LSEG would ask ESMA to bear in mind here that, in general, access arrangements are usually commercially viable to both the receiving party (in this case the CCP) and the party requesting access (in this case, the trading venue). Thus, in our view, it is unlikely that the costs to meet an access request would be so substantial to cause significant or undue risk to either party. On the separate question of which party should bear the cost, LSEG believes that if the CCP incurs cost and such costs can be evidenced and non-discriminatory, then it should be the responsibility of the requesting trading venue to bear a material percentage of the cost. This is in line with industry practice, and the mutually adopted European Code of Conduct on Clearing and Settlement 14 jointly agreed by the Federation of European Securities Exchanges (FESE), the European Association of Clearing Houses (EACH) and the European Central Securities Depositories Association (ECSDA). Of course, this does not preclude the possibility that that the cost of meeting an access request would be commercially recoverable for the CCP (potentially to be partly or fully compensated if the new link results in additional revenues) or a revenue share agreement could be reached. <ESMA_QUESTION_386> Q387: To what extent could a lack of harmonization in certain areas of law constitute a relevant risk in the context of granting or denying access? <ESMA_QUESTION_387> In LSEG s view, this risk is not significant in an EEA context, and we agree with ESMA s assessment in paragraph 28. In the EEA, member states should have already applied the Settlement Finality Directive (SFD). CCPs will apply the rules of their home country jurisdiction, and would have already received legal opinions on applicability and enforceability of their rules. Thus, we do not envisage any problem. <ESMA_QUESTION_387> Q388: Do you agree with the risks identified above in relation to complexity and other factors creating significant undue risks? <ESMA_QUESTION_388> Yes, LSEG agrees with ESMA s analysis in paragraph 24 of the Discussion Paper, and note some additional considerations. From a practitioner s perspective, it is unlikely that a trading venue would use its MiFIR access rights to request access to a CCP for a product that a CCP is not authorised to clear. However, there may be some cases, particularly for commodities, where a trading venue and a CCP may jointly develop a clearing solution for a new product; for which a CCP would need to ask for authorisation. LSEG also agrees with ESMA s comment at the Open Hearing on Markets Issues on 07 July 2014, that if a trading venue requests a CCP to clear a product that it is authorised to clear, but does not currently clear, then that is not a valid reason for the CCP to deny access. This is particularly important to ensure that the clearing and trading obligations for derivatives are implemented effectively. Finally, we note that since it is within the commercial incentive of the requesting trading venue that its access request is agreed, in practice we believe that the venue will ensure that any factor related to complexity is addressed adequately for both its members, the CCP and the relevant National Competent Authorities. <ESMA_QUESTION_388>

113 Q389: Q: Are there other risks related to complexity and other factors creating significant undue risks that should be considered? If so, how would such risks arise from the provision of access? <ESMA_QUESTION_389> Multiple access requests/ resource constraints: An additional complexity may be a situation where a CCP has to deal with multiple access requests simultaneously; especially if a CCP is engaged in the delivery of other services or products and can prove the lack of dedicated resource to deal with the access requests for an appropriate (and justifiable) time frame. In such scenarios, there is some risk that the multiple access requests may become an administrative and operational burden. In particular, CCPs may not be sufficiently staffed to manage the changes required for the approval of the requests or manage the on boarding process for the new trading venue and clearing members. LSEG s suggestion is that a CCP should have agreed and transparent guidelines on how it would allocate resources to deal with multiple requests. One possible approach to mitigate any administrative burden is a controlled phasing approach for applications (e.g. through waves of applications/go-lives on a monthly/ quarterly/ semi-annual cycle). Other factors: Additionally, where access would require a simultaneous change (i.e. changing the netting model from Continuous Net Settlement (CNS) to Trade Date Netting (TDN)) it might result in some processing constraints. <ESMA_QUESTION_389> Q390: Do you agree with the analysis above and the conclusion specified in the previous paragraph? <ESMA_QUESTION_390> Yes, LSEG agrees with ESMA s analysis in paragraph 30 of the Discussion Paper. If the trading venue is a MiFID authorised trading venue (RM, MTF and OTF) and complies with the relevant organisational requirements and/or requirements for electronic trading (e.g. MiFID article 48), then there should be no risk. <ESMA_QUESTION_390> Q391: To what extent would a trading venue granting access give rise to material risks because of anticipated volume of transactions and the number of users? Can you evidence that access will materially change volumes and the number of users? <ESMA_QUESTION_391> In LSEG s view, there are unlikely to be any material risks. From a trading venue s perspective, setting up an additional CCP for its members does not necessarily mean additional members (direct or DEA) or traded volume. Of course, mirroring our comments to Q384, a trading venue should assess if the CCP requesting access is operationally fit to clear its products. But again, this is unlikely to be a risk, since the CCP is likely to be EMIR authorised. Finally, if the trading venue believes the connection to a new CCP will give rise to substantially increased volumes or users, it should be scalable to deal with the additional flow. But of course, this is the preferred outcome for the venue and beneficial to it from a commercial perspective. So in practice, the trading venue should be able to address any capacity shortfall as soon as practicably possible.<esma_question_391> 113

114 Q392: To what extent would a trading venue granting access give rise to material risks because of arrangements for managing operational risk? <ESMA_QUESTION_392> In LSEG s view, there are unlikely to be any material risks, so long as stable procedures governing trade execution and novation of trades to CCPs are in place. These are usually covered in a trading venue s contract with the CCP to provide clearing services. Please also see our comments to Q391. If establishing a connection to new CCP leads to the unlikely scenario where the trading venue s members or volumes substantially increase, this would be commercially beneficial to the venue (and indeed, the CCP requesting access), and the venue would address any capacity shortfall as a priority. <ESMA_QUESTION_392> Q393: Given there will be costs to meeting an access request, what regard should be given to those costs that would create significant undue risk? <ESMA_QUESTION_393> In LSEG s view, if the trading venue incurs cost and such costs can be evidenced and non-discriminatory, then it should be the responsibility of the requesting CCP to bear a material percentage of the cost. The requesting CCP should be aware of legal, operational and technical costs associated to the request, and the analysis and development required to implement the request. This is in line with industry practice, and the mutually adopted European Code of Conduct on Clearing and Settlement jointly agreed by the Federation of European Securities Exchanges (FESE), the European Association of Clearing Houses (EACH) and the European Central Securities Depositories Association (ECSDA). Of course, this does not preclude the possibility that that the cost of meeting an access request would be commercially recoverable for the trading venue (potentially to be partly or fully compensated if the new link results in additional revenues) or a revenue share agreement could be reached. For example, in cash equities, LCH.Clearnet Ltd has agreements with 9 trading venues that all address a commercial cost in the provision and acceptance of the trade feed. LCH.Clearnet has not encountered any situation where the cost of such a trade feed generated undue significant risk. <ESMA_QUESTION_393> Q394: Do you believe a CCP s model regarding the acceptance of trades may create risks to a trading venue if access is provided? If so, please explain in which cases and how. <ESMA_QUESTION_394> It is possible that risks could arise where the CCP s trade registration rules are inconsistent with the rulebook of the trading venue, and trades sent for clearing could thus be rejected. However, in our view, it is unlikely that this risk will materialise. In practice, LSEG believes that the CCP (as the requesting party) would ensure, in cooperation with the venue to which it is requesting access, that it can provide an operational model compatible with that of the venue, and harmonised with its incumbent CCP links. In particular for cash equities, ESMA s guidelines on interoperability under article 52(1)(a) and (b) of EMIR (Legal Requirements, Guideline One) require that the CCP has assessed with a high degree of confidence that its rules and procedures concerning the moment of entry of transfer orders into its systems and the moment of irrevocability have been defined in accordance with EMIR. As part of CCP interoperability procedures there is an agreed procedure between CCPs on acceptance of trades, and so this is unlikely to be a material risk. 114

115 Finally, even if a CCP offers a different trade acceptance or clearing model to its incumbent link (e.g. novation v.s. open offer), the final decision on how to deal with this should be left to the trading venue to which access is requested and the requesting CCP (i.e. whether a change of solution is needed to facilitate access). It should not be an automatic ground for a trading venue to deny access to a CCP, and we agree with ESMA s view in paragraph 32 that these are risks unrelated to the provision of access. In time, we believe that for particular product classes, CCPs may develop harmonised trade acceptance standards to facilitate an open access world. This would benefit the market. <ESMA_QUESTION_394> Q395: Could granting access create unmanageable risks for trading venues due to conflicts of law arising from the involvement of different legal regimes? <ESMA_QUESTION_395> No. In LSEG s view, this is not significant in an EEA context. <ESMA_QUESTION_395> Q396: Are there other risks related to complexity and other factors creating significant undue risks that should be considered? If so, how would such risks arise from the provision of access? <ESMA_QUESTION_396> No, in LSEG s view there are no other significant risks. Similar to our comments to Q379, we note here that CCP access to a new trading venue may introduce new markets and/or assets which could require new or enhanced downstream post trade processes for the CCP. This could include, for example, new arrangements to a CSD or new contracts to store commodities, to ensure physical settlement of trades. The CCP should be able to operate its post trade downstream systems in an orderly and robust manner in that there are no adverse effects on the trading venue s liquidity and reputation. <ESMA_QUESTION_396> Q397: Do you agree with the conditions set out above? If you do not, please state why not. <ESMA_QUESTION_397> In principle, yes, although please note LSEG s specific comment to paragraph 35(iii)(e)(bullet 3) below. We envisage that these terms will be covered in a Service Level Agreement or contract once the access agreement has been agreed and implemented. With regards to conditions on termination, LSEG notes that these are generally set out and agreed between the parties in the contract of services (in this case, the Clearing Agreement). In our view, the condition in 35(iii)(e)(bullet 3) [ termination should be allowed if risks increase in a way that would have justified denial of access in the first instance ] is too wide and subjective. It is not clear what purpose ESMA intends it to serve, and which party could use this condition to terminate the agreement. It is also not clear how this is assessed, and how a disagreement between the two parties would be dealt with. As an alternative, LSEG proposes that in exceptional circumstances, for example, a change in the authorisation of one of the entities in the access arrangement, or when one or both of the parties to the access arrangement repeatedly fail to provide service as per the agreed KPIs in the contract of services, the relevant competent authority may decide that risks have increased in a material manner to terminate the original access arrangement. <ESMA_QUESTION_397> Q398: Are there any are other conditions CCPs and trading venues should include in their terms for agreeing access? 115

116 <ESMA_QUESTION_398> No, LSEG does not believe there are any other conditions. <ESMA_QUESTION_398> Q399: Are there any other fees that are relevant in the context of Articles 35 and 36 of MiFIR that should be analysed? <ESMA_QUESTION_399> No. LSEG agrees with ESMA s analysis in paragraphs We particularly agree with the point made by ESMA in paragraph 39, i.e. all clearing members should be subject to the same schedule of fees and rebates, not just a subset of them. Please also note our response to Q400. <ESMA_QUESTION_399> Q400: Are there other considerations that need to be made in respect of transparent and non-discriminatory fees? <ESMA_QUESTION_400> LSEG believes that, in the context of fees and access arrangements, an important issue for ESMA to consider is the pricing approach of trading venues and CCPs in a closed vertical silo (i.e. where trading venues and CCPs refuse to have any arrangements outside their vertically integrated group), and how this can potentially be used to prevent the benefits of access being passed to investors. For example, in a scenario where a vertically integrated closed silo group maintains its overall transaction fee level for trading and clearing, but loads the vast majority of its fee onto the clearing element and reduces the trading fee element to a negligible amount. This would suggest an assumption that trading fee competition will inhibit or prevent a competitor from taking trading away from its venue but that the clearing element is safe and can therefore carry the bulk of the fee. We suggest this would be use of a fee structure to prevent the benefits of competition through access (at the trading level) from accruing to investors and market participants, by loading the cost of the contract at the clearing level and keeping fees at the same level. (There is evidence that this trend has already started. [1] ) <ESMA_QUESTION_400> Q401: Do you consider that the proposed approach adequately reflects the need to ensure that the CCP does not apply discriminatory collateral requirements? What alternative approach would you consider? <ESMA_QUESTION_401> LSEG s first observation is that the term economically equivalent contracts is linked to the outcome of non-discriminatory collateral requirements and netting of such contracts by the CCP. We urge ESMA that, in the absence of the definition of economically equivalent in the level 1 text, it must provide RTS or guidance on how a CCP should determine contracts traded on different trading venues to be economically equivalent for harmonised treatment in MiFIR and EMIR. LSEG provides its considerations on the definition and its impact below. LSEG also suggests that this discussion is more relevant when multiple trading venues connect to a single CCP, for the purposes of clearing a similar type of product. [1] For example, in May 2014, ICE-Liffe reduced the fee for trading Liffe interest rate futures to 0.08 from an average of 0.25 and increased the clearing fee to 0.20 from its previous level of about 0.03, which means the all-in cost of trading and clearing on Liffe is the same for most customers. Source: ICE tweaks Liffe fees amid ongoing changes, Futures and Options Weekly: 116

117 For exchange-traded and OTC derivatives: In paragraph 44 of the Discussion Paper, ESMA implies that it considers economically equivalent contracts to be those that present the same market risk to contracts already cleared by the CCP. If that is the case, then we suggest that: A CCP may justifiably require different collateral and margining requirements for such contracts, since the basis of the contracts being deemed to be economically equivalence is simply the market risk (i.e. the volatility) and not the characteristics of the contract, i.e. the underlying constituents of the contract, the weights of different constituents, the contract specifications (including the strike, expiry, tenor etc.), treatment of corporate actions by contracts traded on different venues, manifestly different liquidities for different contracts etc. For such contracts (i.e. same market risk, but not the same underlying characteristics), LSEG is of the view that a CCP will offset some risk (through some margin reductions/ cross-margining), but it should be allowed to apply different collateral and margin requirements. However, a CCP must apply its margin and collateral methodologies, and risk management standards in a non-discriminatory manner. For the same collateral and margin requirements, and for the validity of netting procedures, LSEG suggests that the contracts on different trading venues cleared by the CCP should be identical and treated as fungible for the purposes of portfolio margining. In LSEG s view, this is not an overly restrictive interpretation our suggestion is that when a new trading venue wants to offer a competing offering to an incumbent trading venue which clears through its affiliated CCP, then it would be in its incentive to build a market which (as far as possible) offers identical contracts for its customers to achieve margin efficiencies at the clearing level. If it is not possible for the venue to design identical contracts, then it will design contracts to highest degree of similarity as possible. The same principles apply for transferable securities, however since fungibility and/or identical contracts are easier to assess, this process is straight forward. <ESMA_QUESTION_401> Q402: Do you see other conditions under which netting of economically equivalent contracts would be enforceable and ensure non-discriminatory treatment for the prospective trading venue in line with all the conditions of Article 35(1)(a)? <ESMA_QUESTION_402> No, LSEG agrees with ESMA s analysis. With regards to ESMA s analysis paragraph 46, we suggest that cross-border issues are less relevant in an EEA context, where the Settlement Finality Directive (SFD) applies. In this case, the jurisdiction of the CCP is paramount, as its home jurisdiction s laws will apply. In the case of voluntary interoperability, where more than one CCP in multiple jurisdictions may clear the same product, there are usually agreed procedures to deal with any conflict of laws. Examples of these arrangements include those between LCH.Clearnet SA and CC&G (for fixed income securities) and LCH.Clearnet Ltd, SIX X-Clear and EuroCCP (for cash equities). <ESMA_QUESTION_402> Q403: The approach above relies on the CCP s model compliance with Article 27 of Regulation (EU) No 153/2013, do you see any other circumstances for a CCP to cross margin correlated contracts? Do you see other conditions under which cross margining of correlated contracts would be enforceable and ensure non-discriminatory treatment for the prospective trading venue? <ESMA_QUESTION_403> 117

118 LSEG agrees with ESMA s analysis; cross-margining should be done in accordance with a CCP s risk management framework and in compliance with Article 27 of Regulation 153/2013. <ESMA_QUESTION_403> Q404: Do you agree with ESMA that the two considerations that could justify a national competent authority in denying access are (a) knowledge it has about the trading venue or CCP being at risk of not meeting its legal obligations, and (b) liquidity fragmentation? If not, please explain why. <ESMA_QUESTION_404> LSEG agrees that (a) if an NCA has knowledge that a trading venue or CCP is at risk of not meeting its legal obligations is a valid consideration, but does not necessarily agree with (b) liquidity fragmentation being valid in all circumstances. Both conditions should be defined further. We provide some additional thoughts below and in our answer to Q405. Re (a) if an NCA has knowledge that a trading venue or CCP is at risk of not meeting its legal obligations LSEG notes that ESMA asks this question with regards to its task in defining the conditions under which granting access will threaten the smooth and orderly functioning of markets. In the case where a trading venue or a CCP is at a risk of not meeting its legal obligations on an ongoing basis and an incoming access request may make this even more risky, we suggest this is more of an issue of the receiving trading venue/ CCP s ongoing authorisation status, rather than the access request itself. We suggest that the EU rulebook consisting of MiFID II/MiFIR, EMIR and CSD-R requires FMIs to be robust and resilient, so it should be unlikely that an access request pushes a trading venue/ccp over an NCA s risk thresholds. Re (b) liquidity fragmentation LSEG agrees with ESMA that the Level 1 requires a NCA to assess if liquidity fragmentation threatens the smooth and orderly functioning of the market or adversely affects systemic risk. However, in our view, ESMA should specify clearly the conditions under which it views liquidity fragmentation to be a valid consideration, for a harmonised application across NCAs. In LSEG s view, we are yet to encounter a situation where liquidity fragmentation (as defined by MiFIR Article 2(1)(45)) adversely affects smooth and orderly functioning of markets. Both legs of the definition are relevant when a trading venue clears its products through more than one CCP (and not when multiple trading venues clear through a single CCP). We provide our analysis of this circumstance below: For transferable securities and money market instruments, LSEG s view is that this condition is entirely irrelevant. EMIR allows and specifies the framework for interoperating which prevents either of the conditions in MiFIR Article 2(1)(45) from arising. For derivatives: If a voluntary interoperability agreement can be agreed by the relevant parties and agreed by the NCAs, there is no liquidity fragmentation. If voluntary interoperability agreement cannot be agreed, then a preferred clearing arrangement (i.e. where all trading members have access to a default clearer, and a subset of members have access to other clearing solutions) would prevent liquidity fragmentation from arising and being a consideration. Condition (a) or (b) may only arise when trading venues split their order books to allow for more than one clearing solution to which all trading participants may not have access, or if a preferred clearing arrangement is not reached (and indirect clearing arrangements cannot be facilitated). In our view, this scenario is very unlikely to arise, because there has to be sufficient reason for participants (trading and clearing members) to support this solution, as it would limit the potential for netting at the clearing level. However, it should be up to the participants to agree and support this 118

119 solution, and if all parties (participants, trading venue and CCPs) agree that there is sufficient reason to do so, then we don t feel this should be an automatic criteria for NCAs to reject access. LSEG also recommends that when an NCA uses the reason of granting access would threaten to the smooth and orderly functioning of markets or would adversely affect system risk to deny access, it should explain why to at least the requesting party, on the basis of fair and non-discriminatory criteria to prevent any misuse of this consideration. We also suggest that the requesting party should have a right to appeal. In the case the receiving party is: A trading venue: the appeal should be considered by the college of the CCP, and the final decision subject to a vote of the college on a majority basis. A CCP: an appeal should be considered by ESMA, the NCA of the trading venue and the NCA of the CCP. The final decision should be a vote of the three relevant parties. <ESMA_QUESTION_404> Q405: How could the above mentioned considerations be further specified? <ESMA_QUESTION_405> Please see LSEG s response to Q405. In summary, our key suggestions are: 1. General considerations LSEG recommends that when an NCA uses the reason of granting access would threaten to the smooth and orderly functioning of markets or would adversely affect system risk to deny access, it should explain why to at least the requesting party, on the basis of fair and non-discriminatory criteria to prevent any misuse of this consideration. LSEG also suggest that the requesting party, or both the requesting and receiving party, should have a right to appeal. In the case the receiving party is: A trading venue: the appeal should be considered by the college of the CCP and the NCA of the trading venue. The final decision subject to a vote of the college and trading venue s NCA on a majority basis. A CCP: an appeal should be considered by ESMA, the NCA of the trading venue and the NCA of the CCP. The final decision should be a vote of the three relevant parties. 2. Liquidity fragmentation ESMA should specify that this condition is only relevant when a trading venue clears its products through more than one CCP (and not when multiple trading venues clear through a single CCP). For transferable securities and money market instruments, our view is that this condition is entirely irrelevant. EMIR allows and specifies the framework for interoperating which prevents either of the conditions in MiFIR Article 2(1)(45) from arising. For derivatives: o If a voluntary interoperability agreement can be reached by the relevant parties and agreed by the NCAs, there is no liquidity fragmentation. o If voluntary interoperability agreement cannot be agreed, then a preferred clearing arrangement (i.e. where all trading members have access to a default clearer, and a subset of members have access to other clearing solutions) would prevent liquidity fragmentation from arising and being a consideration. o Condition (a) or (b) may only arise when trading venues split their order books to allow for more than one clearing solution to which all trading participants may not have access, or if a preferred clearing arrangement is not reached (and indirect clearing arrangements cannot be facilitated). In our view, this scenario is very unlikely to arise, because there has to be sufficient reason for participants (trading and clearing members) to support this solution, as it would limit the potential for netting at the clearing level. However, it should be up to the participants to agree and support this solution, and if all parties (participants, trading venue and CCPs) agree that there is sufficient reason to do so, then we do not feel this should be an automatic criteria for NCAs to reject access. 119

120 <ESMA_QUESTION_405> Q406: Are there other conditions that may threaten the smooth and orderly functioning of the markets or adversely affect systemic risk? If so, how would such risks arise from the provision of access? <ESMA_QUESTION_406> No, LSEG does not consider there are any other conditions. <ESMA_QUESTION_406> Q407: Do you agree with ESMA s proposed approach that where there are equally accepted alternative approaches to calculating notional amount, but there are notable differences in the value to which these calculation methods give rise, ESMA should specify the method that should be used? <ESMA_QUESTION_407> Yes, LSEG agrees this is a sensible approach. It should also be noted that in some cases, notional values are less meaningful. For example a 30 year bond contract with notional EUR 100,000 leads to a greater exposure to a given shift in interest rates than a short term interest rate contract of notional EUR 1 million. For interest rate instruments, notional could be calculated on a 10 year equivalent basis or similar which would more accurately reflect risk than pure notional. <ESMA_QUESTION_407> Q408: Do you agree that the examples provided above are appropriate for ESMA to adopt given the purpose for which the opt-out mechanism was introduced? If not, why, and what alternative(s) would you propose? <ESMA_QUESTION_408> LSEG: Yes, we agree. Please also see our comment to Q407. <ESMA_QUESTION_408> Q409: For which types of exchange traded derivative instruments do you consider there to be notable differences in the way the notional amount is calculated? How should the notional amount for these particular instruments be calculated? <ESMA_QUESTION_409> LSEG: No particular comments. It should also be noted that in some cases, notional values are less meaningful. For example a 30 year bond contract with notional EUR 100,000 leads to a greater exposure to a given shift in interest rates than a short term interest rate contract of notional EUR 1 million. For interest rate instruments, notional could be calculated on a 10 year equivalent basis or similar which would more accurately reflect risk than pure notional. <ESMA_QUESTION_409> Q410: Are there any other considerations ESMA should take into account when further specifying how notional amount should be calculated? In particular, how should technical transactions be treated for the purposes of Article 36(5), MiFIR? <ESMA_QUESTION_410> LSEG: No, in our view there are no other considerations <ESMA_QUESTION_410> 120

121 5.8. Non- discriminatory access to and obligation to license benchmarks Q411: Do you agree that trading venues require the relevant information mentioned above? If not, why? <ESMA_QUESTION_411> Specific response to Q411 Yes, LSEG agrees. A feed of the benchmark values on an end of day, periodic or real time basis according to the type of index and frequency of its calculation and publication is appropriate. In principle, a trading venue needs access to the same set of data and information that is commercially available to market participants and sufficient to perform proper assessment of the marketability and development of the product, as well as controls required under the relevant regulatory framework, including market operations and surveillance. -- Introduction: LSEG s general comments on non-discriminatory access and obligation to license benchmarks (MiFIR article 37) LSEG uses its experience of operating its trading venues, CCPs and benchmark businesses in answering the questions in this section. LSEG operates FTSE International (FTSE) a global leader in indexing and analytical solutions. FTSE calculates thousands of unique indices that measure and benchmark markets and asset classes in more than 80 countries around the world. FTSE indices are used extensively by market participants worldwide for investment analysis, performance measurement, asset allocation and portfolio hedging. FTSE s transparent rules-based methodology is overseen by independent committees of leading market participants, focused on applying the highest industry standards in index design and governance. As of 8 April 2014, indices run by MTS joined the FTSE family of indices. MTS runs a number of indices that are independent total-return indices measuring the performance of the largest and most widely-traded securities in the Eurozone government bond market. These include the EuroMTS Index Range (ex-cno Etrix), EuroMTS Inflation-Linked Indices, MTS Italy and ex-bank of Italy index families. In summary, our key views on the topics in section 5.8 of the Discussion Paper are: Purpose of Article 37: LSEG agrees with ESMA s assessment in paragraph 41(ii) that Article 37 is part of a legal solution encompassing Articles 35 and 36, to empower EU investors in derivatives markets (both ETD and OTC derivatives (OTCDs)) through enhanced choice, better service, lower costs, greater capital efficiency, innovation, and to avoid the concentration of risk in closed FMI. The need for this is also noted in EMIR recital and in the CSFI report on EU Post-trade infrastructure. Information required by TVs/ CCPs: Trading venues/ CCPs need access to the same set of data and information that is commercially available to market participants and sufficient to provide a well functioning market, surveillance and risk management (e.g. calculation of margins). The frequency depends on the nature of the index and the frequency of calculation and publication and the type of information/data concerned. The method of access should be reliable (preferably, nonproprietary), and data should be made available in a widely used, standard format. 15 EMIR recital 36 states that: In certain areas within financial services and trading of derivative contracts, commercial and intellectual property rights may also exist. In instances where such property rights relate to products or services which have become, or impact upon, industry standards, licences should be available on proportionate, fair, reasonable and non- discriminatory terms

122 IP rights/ confidentiality: A standard licence agreement usually contains a series of restrictions and practical measures in place to protect the IPRs and any trademark. Standard confidentiality provision around licensor s data and information would be included, although such a provision would be less relevant where a provider s index data and the methodologies are publicly available (i.e. for benchmarks based on post-trade transparent data, e.g. equity indices). Benchmark providers should not be required to provide proprietary information provided to them as part of the index determination process, such as the underlying data that gives rise to constituent weights. However, if data is provided by the benchmark provider to its clients in client files, that information should also be made available to a trading venue or a CCP on a non-discriminatory basis. Licence provisions: In general, LSEG supports an approach where users should be provided with the relevant information through agreements with the person with proprietary rights to the benchmark (the Benchmark Administrator) and not from the trading venue/ccp. For exchange traded contracts, a CCP would generally expect the venue to obtain a licence for the benchmarks on terms that enable the CCP to clear the relevant product, and for the venue to provide the benchmark to the CCP. For OTCDs, the CCP will typically obtain a licence itself. Licences should be available for individual indices, i.e. a CCP should be entitled to obtain an index licence covering a single index, rather than only for bundles or families of benchmarks. A reference to compliance with the IOSCO Principles for Benchmarks would also be expected in the licence, once in force. FTSE issued a statement of compliance with the IOSCO principles in July Conflicts of interest: Licences should be at an arms length and on the same basis as similar commercial arrangements made by independent providers. Where the owner of proprietary rights to a benchmark is integrated (i.e. as part of a Financial Market Infrastructure (FMI) group along with another trading venue or CCP), it is important that the trading venue/ccp accesses the benchmark on an unbundled basis with no cross-subsidy, so as to enable third party trading venues / CCPs to obtain access on comparable commercial terms. Benchmarks Regulation: It will also be important for ESMA to take account of how any obligation to grant a non-exclusive licence of a benchmark will interact with the terms under which a benchmark provider would deal with a request to permit use of its benchmark in relation to a financial instrument or contract under the Commission s proposed benchmark regulation and the extent to which they are compatible. <ESMA_QUESTION_411> Q412: Is there any other additional information in respect of price and data feeds that a trading venue would need for the purposes of trading? <ESMA_QUESTION_412> LSEG: Yes. A trading venue would also need: A feed of and/or access to any corrections/recalculations/restatements of a benchmark value. Index expiry values, where available. <ESMA_QUESTION_412> Q413: Do you agree that CCPs require the relevant information mentioned above? If not, why? <ESMA_QUESTION_413> 16 FTSE issues Statement of Compliance with respect to IOSCO Principles: 122

123 Yes, LSEG agrees with ESMA. <ESMA_QUESTION_413> Q414: Is there any other additional information in respect of price and data feeds that a CCP would need for the purposes of clearing? <ESMA_QUESTION_414> LSEG: Yes. A CCP would also need the following information directly from the benchmark provider or from the trading venue (if it is agreed in the licence that the venue will provide the CCP with the feeds): A feed of and/or access to any corrections/recalculations/restatements of a benchmark value. Index expiry values, where available. <ESMA_QUESTION_414> Q415: Do you agree that trading venues should have access to benchmark values as soon as they are calculated? If not, why? <ESMA_QUESTION_415> Yes, LSEG agrees that benchmark values and any information concerning any issue in calculating and disseminating the index value should be promptly notified to the trading venue or CCP We agree with ESMA that the frequency in part depends on the nature of the benchmark and the frequency of calculation and publication and the type of information/data concerned. Some will be required real time, or periodically during the day or at end of day. Others according to the expiry cycle. Real time should be without any delay in publication. Standing information on the methodology, governance and composition should be available permanently, via the index provider s website. Up-to-date benchmark values must be available to the trading venue, up to real-time values, where such data are calculated by the index provider. Where the benchmark is mainly used for settlement purposes, a daily data feed of index values could be considered sufficient. <ESMA_QUESTION_415> Q416: Do you agree that CCPs should have access to benchmark values as soon as they are calculated? If not, why? <ESMA_QUESTION_416> Yes. Please also see LSEG s response to Q415. Benchmark values must be available as soon as they are calculated to a CCP for it to calculate margins with reference to certain products (such as options). However where the benchmark is mainly used for settlement purposes, a daily feed of index values could be considered sufficient. <ESMA_QUESTION_416> Q417: Do you agree that trading venues require the relevant information mentioned above? If not, why? <ESMA_QUESTION_417> Yes, LSEG agrees with ESMA. For a trading venue, information around benchmark composition is particularly relevant for the initial assessment of the characteristics of the benchmark. Availability of such information is also useful to support on-going market surveillance activities by the trading venue, especially where it also manages the market in the underlying for the relevant derivatives contract. <ESMA_QUESTION_417> Q418: Is there any other additional information in respect of composition that a trading venue would need for the purposes of trading? 123

124 <ESMA_QUESTION_418> LSEG: Yes. A trading venue will also need data files to facilitate the expiry calculation on every expiry <ESMA_QUESTION_418> Q419: Do you agree that CCPs require the relevant information mentioned above? If not, why? <ESMA_QUESTION_419> Yes, LSEG agrees with ESMA. A CCP needs information relating to benchmark composition to perform its stress-tests. <ESMA_QUESTION_419> Q420: Is there any other additional information in respect of composition that a CCP would need for the purposes of clearing? <ESMA_QUESTION_420> LSEG: Yes. In addition, a CCP should have access to any published changes to the composition <ESMA_QUESTION_420> Q421: Do you agree that trading venues and CCPs should be notified of any planned changes to the composition of the benchmark in advance? And that where this is not possible, notification should be given as soon as the change is made? If not, why? <ESMA_QUESTION_421> Yes, LSEG agrees with ESMA. <ESMA_QUESTION_421> Q422: Do you agree that trading venues need the relevant information mentioned above? If not, why? <ESMA_QUESTION_422> Yes, LSEG agrees. A trading venue should have full visibility of the methodology in order to assess whether it represents a suitable benchmark to develop derivatives contracts. Availability of such information is also required to support on-going market monitoring and surveillance activities that a trading venue will need to have in place, especially where it also manages the market in the underlying for the relevant derivatives contract. This will be required as part of the implementation of the IOSCO Principles by Benchmark Administrators. <ESMA_QUESTION_422> Q423: Is there any other additional information in respect of methodology that a trading venue would need for the purposes of trading? <ESMA_QUESTION_423> LSEG: Yes, some additional information may be required by trading venues. For example, treatment of corporate actions, taxes and the relevant exchange rate data used by the benchmarks provider, as provided to its clients in the client files. <ESMA_QUESTION_423> Q424: Do you agree that CCPs require the relevant information mentioned above? If not, why? <ESMA_QUESTION_424> Yes, LSEG agrees with ESMA. A CCP needs full visibility of the methodology to the extent it shall be able to assess whether the benchmark can be used for clearing purposes, and to manage stress tests. <ESMA_QUESTION_424> 124

125 Q425: Is there any other additional information in respect of methodology that a CCP would need for the purposes of clearing? <ESMA_QUESTION_425> LSEG: Yes, some additional information may be required by CCPs. For example, treatment of corporate actions, taxes and the relevant exchange rate data used by the benchmarks provider, as provided to its clients in the client files. <ESMA_QUESTION_425> Q426: Is there any information is respect of the methodology of a benchmark that a person with proprietary rights to a benchmark should not be required to provide to a trading venue or a CCP? <ESMA_QUESTION_426> LSEG: No. We suggest that the principle should be that whatever a benchmark IPR owner would license to a relevant third party to allow it to trade and/or clear products based on that benchmark, that is what should be provided. It is not obvious to us what methodology information might be withheld. We note that benchmark providers should not be required to provide proprietary information provided to them as part of the index determination process, such as the underlying data that gives rise to constituent weights. However, if data provided by the benchmark provider to its clients in client files, or to another trading venue or a CCP for trading or clearing, that information should also be made available to a trading venue or a CCP requesting access on a non-discriminatory basis. <ESMA_QUESTION_426> Q427: Do you agree that trading venues require the relevant information mentioned above (values, types and sources of inputs, used to develop benchmark values)? If not, why? <ESMA_QUESTION_427> LSEG: Yes. Types and sources of inputs should be included in the Benchmark Administrator s methodology documents and therefore should be publicly available to the trading venues. Constituent values should be available in the Benchmark data files. Where an index is based on submissions, LSEG agrees that individual submissions should not be available if they are not published. LSEG also notes that it agrees with ESMA s interpretation of the Level I text that pricing covers the input data, i.e. the prices covered by the first section on Relevant Price and data feeds. <ESMA_QUESTION_427> Q428: Is there any other additional information in respect of pricing that a trading venue would need for the purposes of trading? <ESMA_QUESTION_428> LSEG: No. <ESMA_QUESTION_428> Q429: In what other circumstances should a trading venue not be able to require the values of the constituents of a benchmark? <ESMA_QUESTION_429> LSEG: Where the values are derived from underlying proprietary information, such as in the case of individual submissions from banks or dealers used in the construction of fixed income or reference rate indices, the proprietary information should not be required by the trading venue. To require such information could be a breach of the contractual terms between the information provider and the Benchmark Administrator. However, if data provided by the benchmark provider to its clients in client data files, that information should also be made available to a trading venue or a CCP. 125

126 This is not usually a problem for benchmarks based on regulated data, i.e. post-trade data from transparent markets. <ESMA_QUESTION_429> Q430: Do you agree that CCPs require the relevant information mentioned above? If not, why? <ESMA_QUESTION_430> LSEG: Yes. CCPs should require the same information and be treated in the same way as trading venues and other clients of the Benchmark. Types and sources of inputs should be included in the Benchmark Administrator s methodology documents and therefore should be publicly available to the trading venues. <ESMA_QUESTION_430> Q431: Is there any other additional information in respect of pricing that a CCP would need for the purposes of clearing? <ESMA_QUESTION_431> LSEG: No. <ESMA_QUESTION_431> Q432: In what other circumstances should a CCP not be able to require the values of the constituents of a benchmark? <ESMA_QUESTION_432> LSEG: Where the values are proprietary information, such as in the case of fixed income or reference rate index submissions, they should not be required by the trading venue. To require such information could be a breach of the contractual terms between the information provider and the Benchmark Administrator. However, if data provided by the benchmark provider to its clients in client files, that information should also be made available to a trading venue or a CCP. <ESMA_QUESTION_432> Q433: Do you agree that trading venues require the additional information mentioned above? If not, why? <ESMA_QUESTION_433> Yes, LSEG agrees with ESMA. <ESMA_QUESTION_433> Q434: Do you agree that CCPs require the additional information mentioned above? If not, why? <ESMA_QUESTION_434> Yes, LSEG agrees with ESMA. <ESMA_QUESTION_434> Q435: Is there any other information that a trading venue would need for the purposes of trading? <ESMA_QUESTION_435> LSEG: Yes. This includes: 1. The technical characteristics of the data feed (e.g. data dissemination protocol and technology used, format of data, backup solution) should be agreed in advance, and any change of the technical features of such data feed should be disclosed to the trading venue with proper advance. A reliable (preferably, non-proprietary) method of access is preferable, and data should be made available in a widely used, standard format. 126

127 2. Where relevant, a trading venue should be given access to any proprietary software that is typically distributed in conjunction with the benchmark and that is necessary to use the benchmark (or understand its constituents) effectively. 3. A reference to compliance with the IOSCO Principles for Benchmarks would also be expected, once in force. <ESMA_QUESTION_435> Q436: Is there any other information that a CCP would need for the purposes of clearing? <ESMA_QUESTION_436> LSEG: Yes. Where relevant, a CCP should be given access to any proprietary software that is typically distributed in conjunction with the benchmark and that is necessary to use the benchmark (or understand its constituents) effectively. A reference to compliance with the IOSCO Principles for Benchmarks would also be expected, once in force. <ESMA_QUESTION_436> Q437: Do you agree with the principles described above? If not, why? <ESMA_QUESTION_437> In principle, LSEG agrees. In particular, LSEG agrees with the approach in paragraph 43 of the DP and this is the model FTSE follows rather than that in paragraph 44. FTSE licences the trading venue/ccp directly and also licences the user directly <ESMA_QUESTION_437> Q438: Do users of trading venues need non-publicly disclosed information on benchmarks? <ESMA_QUESTION_438> LSEG: No, but if they would like such information they may obtain this directly from the benchmark administrator on a non-discriminatory licence basis. <ESMA_QUESTION_438> Q439: Do users of CCPs need non-publicly disclosed information on benchmarks? <ESMA_QUESTION_439> LSEG: Members of a CCP need the ability to access sufficient information to understand how the CCP calculates margin requirements. For example, if the CCP relies on historical data to calculate margin, then users should be able to access the historical data necessary to replicate margin calculations. Clearing members may access this directly from the benchmark administrator on a non-discriminatory basis. <ESMA_QUESTION_439> Q440: Where information is not available publicly should users be provided with the relevant information through agreements with the person with proprietary rights to the benchmark or with its trading venue / CCP? <ESMA_QUESTION_440> In general, LSEG supports an approach where users should be provided with the relevant information through agreements with the person with proprietary rights to the benchmark (the Benchmark Administrator) and not from the trading venue/ccp. <ESMA_QUESTION_440> Q441: Do you agree with the conditions set out above? If not, please state why not. 127

128 <ESMA_QUESTION_441> Yes, LSEG agrees. The trading venue/ccp and the benchmark administrator should enter into a licence agreement to cover the terms of provision of the benchmark. FTSE considers that the parties should be free to agree the terms of the licence agreement in line with their standard licensing terms, with the important condition that the licence terms should be non discriminatory between all users of the benchmark (including trading venues/ccps and product issuers). <ESMA_QUESTION_441> Q442: Are there any are other conditions persons with proprietary rights to a benchmark and trading venues should include in their terms for agreeing access? <ESMA_QUESTION_442> LSEG: No. <ESMA_QUESTION_442> Q443: Are there any are other conditions persons with proprietary rights to a benchmark and CCPs should include in their terms for agreeing access? <ESMA_QUESTION_443> LSEG: No. <ESMA_QUESTION_443> Q444: Which specific terms/conditions currently included in licensing agreements might be discriminatory/give rise to preventing access? <ESMA_QUESTION_444> LSEG: We can only speculate as to what the range of existing arrangements covers. Such conditions would have to go beyond the standard scope of the terms of a licence (e.g. scope, control and termination) An example could be an explicit restriction of an index provider from licensing an index to other trading venue or CCPs. This should not be a problem in circumstances where there is a new product, as a licensing trading venue may typically request a period of exclusivity from the Benchmark Administrator, as in MiFIR Article 37(2). However, outside this exclusivity period, such a provision would need to be specifically added to the license agreement, and could call into question issues of competition law. It is also our view that licences between proprietary rights holders and associated trading venues and/or CCPs should be at arms length and on the same basis as similar commercial arrangements made by independent providers. Internal cross charging/allocation arrangements should reflect these terms and prices and be transparent to trading venues/ccps seeking access to a benchmark under this provision. Where there is a benchmark provider integrated with a trading venue, it is important that the trading venue access the benchmark on an unbundled basis with no cross-subsidy, so as to enable third party trading venues / CCPs to obtain access on comparable commercial terms. <ESMA_QUESTION_444> Q445: Do you have views on how termination should be handled in relation to outstanding/significant cases of breach? <ESMA_QUESTION_445> LSEG: FTSE includes termination provisions in its standard licence agreements that typically provide for: Termination for convenience after a specified notice period (typically 3 months but, on occasion, only exercisable after the anniversary of the licence agreement in the case of trading venues) Termination for cause (un-remedied breach, change of control, insolvency, etc) Termination for specific events such as: o Licensee in breach of applicable securities laws o Licensee failed to build liquidity in the derivative contract 128

129 Typically the termination provisions will include confirmation that permissions granted in respect of any Exchange Traded Products already in issue do not terminate provision, effectively creating a run off period following termination until the last of the derivatives expire. FTSE considers that termination in cases of outstanding/significant cases of breach of the licensing terms should be handled in this manner. <ESMA_QUESTION_445> Q446: Do you agree with the approach ESMA has taken regarding the assessment of a benchmark s novelty, i.e., to balance/weight certain factors against one another? If not, how do you think the assessment should be carried out? <ESMA_QUESTION_446> LSEG: Yes - although we would stress the importance of methodologies (and their differences), compared to the potential correlation to existing benchmarks (it is important to bear in mind that correlation does not always mean causality.) <ESMA_QUESTION_446> Q447: Do you agree that each newly released series of a benchmark should not be considered a new benchmark? <ESMA_QUESTION_447> Yes, LSEG agrees. <ESMA_QUESTION_447> Q448: Do you agree that the factors mentioned above could be considered when assessing whether a benchmark is new? If not, why? <ESMA_QUESTION_448> Yes, LSEG agrees. The real test of a new benchmark is whether it is seen by the market and the users as sufficiently novel and related to its stated objective to be viable; there is unlikely to be interest in trading a new benchmark if it is not sufficiently different from an existing index. Some of such factors could be considered further. Taking them in turn: 56(i) Our view is that fungible and capable of being netted are different things. If the benchmark is fungible, we agree that the benchmark is less likely to be new. On the other hand, if the benchmark is capable of being netted then it is more likely to be considered a new benchmark. 56(ii) We agree this category means the benchmark is less likely to be new. 56(iii) We agree with footnote 170 to this factor and do not think this is a meaningful test in determining whether or not a benchmark is new. 56(iv) We consider this to be very important and agree that in the circumstances set out, the benchmark is unlikely to be new. 56(v) We agree with this criterion. 57(i) This view is that this is not an important factor. 57(ii) We agree with footnote 174 and consider that the user group is not important in deciding the novelty of the benchmark. <ESMA_QUESTION_448> 129

130 Q449: Are there any factors that would determine that a benchmark is not new? <ESMA_QUESTION_449> LSEG: No. <ESMA_QUESTION_449> 130

131 6. Requirements applying on and to trading venues 6.1. Admission to Trading Q450: What are your views regarding the conditions that have to be satisfied in order for a financial instrument to be admitted to trading? <ESMA_QUESTION_450> LSEG: The conditions that are already in place have worked satisfactorily and ensure a level playing field across trading venues. Under the UK framework the Financial Conduct Authority (FCA) is the national competent authority responsible for operating the UK listing regime and oversight of UK regulated markets, including approving prospectuses and overseeing continuing obligations such as the Disclosure and Transparency Rules (DTRs) applicable to issuers on the London Stock Exchange s regulated markets. Admission to trading is subject to London Stock Exchange s rules through which the Exchange retains the right to refuse an admission to trading operating under its obligations under the FCA s Recognised Investment Exchange Handbook. However, beyond fair and orderly trading or specific financial crime issues that have come to the Exchange s attention, it is difficult to envisage the circumstances where the power to refuse admission might be necessary or rightfully exercised post the approval of a security or issuer by a national competent authority. Admission to trading in Italy is subject to Borsa Italiana s rules through which the Borsa retains the right to refuse an admission to trading operating. <ESMA_QUESTION_450> Q451: In your experience, do you consider that the requirements being in place since 2007 have worked satisfactorily or do they require updating? If the latter, which additional requirements should be imposed? <ESMA_QUESTION_451> LSEG: We are of the view that the existing implementing provisions on the characteristics of the financial instruments are appropriate and should be maintained. <ESMA_QUESTION_451> Q452: More specifically, do you think that the requirements for transferable securities, units in collective investment undertakings and/or derivatives need to be amended or updated? What is your proposal? <ESMA_QUESTION_452> LSEG: Overall LSE and Borsa Italiana agree with both arrangements and do not think any update is required, except to require the same standards for non-ucits ETFs. On LSEG s markets, each listed ETF must be supported by at least one official market maker. The fulfilment of market maker obligations is essential for the provision of liquid, transparent and orderly secondary markets. The London Stock Exchange Group s Market Supervision team closely monitors market makers performance against these obligations to ensure that market makers are continuously displaying two-way prices within the applicable maximum spread and minimum quote size throughout the trading day <ESMA_QUESTION_452> 131

132 Q453: How do you assess the proposal in respect of requiring ETFs to offer market making arrangements and direct redemption facilities at least in cases where the regulated market value of units or shares significantly varies from the net asset value? <ESMA_QUESTION_453> LSEG: Please see response to Q452 Overall we agree with both arrangements and do not think that any update is required except requiring the same standards for non-ucits ETFs. With regard to secondary market investor, we welcome the provision of giving all investors the possibility of seeking redemption directly from the fund (or an appointed agent) for non-ucits ETFs. By definition, it is likely that non UCITS ETFs will not meet the UCITS ETFs standards. In Borsa a experience, investors that have bought an ETF on the secondary market rarely request the reimbursement from the issuer, even when this facility is available. This is mainly due to the difficulties of the administrative process and, consequently, to the related costs. We agree that in times of market disruption, it might be useful for unit/share holders to be able to redeem in the primary market. This provision already exists for ETFs distributed in Italy and is overseen by the Bank of Italy. <ESMA_QUESTION_453> Q454: Which arrangements are currently in place at European markets to verify compliance of issuers with initial, on-going and ad hoc disclosure obligations? <ESMA_QUESTION_454> LSEG: In the UK, the Financial Conduct Authority (acting in its capacity as the UK Listing Authority) is the national competent authority responsible for approving prospectuses for admissions to a Regulated Market as well as on-going compliance with issuers disclosure obligations such as the obligations arising from the Transparency Directive. In Italy, Borsa Italiana verifies the respect of the obligations through the monitoring of the storage of documents in the NIS system (Network Information System for the storage and disclosure of regulated information) and also through the control of information disclosed via media. <ESMA_QUESTION_454> Q455: What are your experiences in respect of such arrangements? <ESMA_QUESTION_455> LSEG: In the UK, the FCA is the body that has greatest experience of these arrangements as national competent authority. London Stock Exchange s experiences principally relate to receiving confirmation from the FCA that a prospectus has been approved, prior to listing of the security to its regulated market. In Italy, Borsa Italiana verifies the respect of the obligations through the monitoring of the storage of documents in the NIS system (Network Information System for the storage and disclosure of regulated information) and also through the control of information disclosed via media. <ESMA_QUESTION_455> Q456: What is your view on how effective these arrangements are in performing verification checks? <ESMA_QUESTION_456> LSEG: In the UK, the FCA is the body that has greatest experience of these arrangements as national competent authority. In Italy, the current monitoring arrangements are effective for the purpose. <ESMA_QUESTION_456> 132

133 Q457: What arrangements are currently in place on European regulated markets to facilitate access of members or participants to information being made public under Union law? <ESMA_QUESTION_457> LSEG: Currently, the Transparency Directive and its implementing measures require regulated information for issuers on regulated markets to be disseminated through an information service provider that complies with specified minimum standards. Specifically, regulated information must be disseminated in a manner ensuring that it is capable of being disseminated to as wide a public as possible. In addition, each Member State has a repository to act as a storage mechanism for information on all issuers on the regulated markets for that Member State. Many market operators across the EU, including London Stock Exchange Group, also publish key information and issuer announcements on their websites, thus further facilitating public access to the information. For instance, Borsa Italiana, through the publication of Notices on its website makes available to the market the information published by issuers. At the same time, Borsa Italiana arranges the NIS service for the storage and transmission of the regulated information to different persons (issuers, information providers, press agency, etc.), Moreover, issuers information (balance sheet, shareholder meetings report, etc) are also disclosed via Borsa s website. <ESMA_QUESTION_457> Q458: What are your experiences in respect of such arrangements? <ESMA_QUESTION_458> LSEG: In the UK, the FCA is the body that has greatest experience of these arrangements as national competent authority. For Italy, see Response to Question 457 <ESMA_QUESTION_458> Q459: How do you assess the effectiveness of these arrangements in achieving their goals? <ESMA_QUESTION_459> TYPE YOUR TEXT HERE <ESMA_QUESTION_459> Q460: Do you agree with that, for the purpose of Article 51 (3) (2) of MiFID II, the arrangements for facilitating access to information shall encompass the Prospectus, Transparency and Market Abuse Directives (in the future the Market Abuse Regulation)? Do you consider that this should also include MiFIR trade transparency obligations? <ESMA_QUESTION_460> LSEG: We agree that the arrangements should encompass TD, PD and MAD. Anything that would improve transparency and free flow of public information would help the market and ensure a level playing field across exchanges and markets. <ESMA_QUESTION_460> 6.2. Suspension and Removal of Financial Instruments from Trading - connection between a derivative and the underlying financial instrument and standards for determining formats and timings of communications and publications 133

134 Q461: Do you agree with the specifications outlined above for the suspension or removal from trading of derivatives which are related to financial instruments that are suspended or removed? <ESMA_QUESTION_461> LSEG: Agree. We support ESMA s analysis and note that, in the case of a derivative that is completely dependent on the underlying instrument, a national competent authority/market operator is not required to extend the suspension to a derivative if one of the following three reasons do not exist: (i) suspected market abuse; (ii) a take over bid; or (iii) the non-disclosure of inside information about the issuer or financial instrument in breach of MAR articles 7 and 17 <ESMA_QUESTION_461> Q462: Do you think that any derivatives with indices or a basket of financial instruments as an underlying the pricing of which depends on multiple price inputs should be suspended if one or more of the instruments composing the index or the basket are suspended on the basis that they are sufficiently related? If so, what methodology would you propose for determining whether they are sufficiently related? Please explain. <ESMA_QUESTION_462> LSEG: We consider that in general derivatives that relate to indices, baskets or other tradable financial instruments that consist of multiple price input should be excluded from an automatic suspension regime, because the halted underlying instrument will only represent a fraction of the contract value, and suspend an index contract this manner would contribute to further market disruption. If ESMA was to introduce a threshold for sufficiently related securities, we might expect that this should relate to at least 20% of the value of the contract. The measure should be more precise, i.e. where a single or group of instruments comprise of a defined percentage or weighting of the overall index. Indices are widely use as benchmarks for the asset management industry, so their suspension should be reserved for situations where the true valuation is clearly compromised. <ESMA_QUESTION_462> Q463: Do you agree with the principles outlined above for the timing and format of communications and publications to be effected by trading venue operators? <ESMA_QUESTION_463> LSEG: We broadly agree with the principles outlined for timing and format of communications and publications to be effected by trading venue operators. We support the notification requirement from the market operator to the competent authority (in relation to a MTF) being in a format which ensures a quick and seamless processing by competent authorities/market operators so they can without difficulty take their decision in respect of the suspension, removal or lifting of suspensions. <ESMA_QUESTION_463> 134

135 7. Commodity derivatives 7.1. Ancillary Activity Q464: Do you see any difficulties in defining the term group as proposed above? <ESMA_QUESTION_464> TYPE YOUR TEXT HERE <ESMA_QUESTION_464> Q465: What are the advantages and disadvantages of the two alternative approaches mentioned above (taking into account non-eu activities versus taking into account only EU activities of a group)? Please provide reasons for your answer. <ESMA_QUESTION_465> TYPE YOUR TEXT HERE <ESMA_QUESTION_465> Q466: What are the main challenges in relation to both approaches and how could they be addressed? <ESMA_QUESTION_466> TYPE YOUR TEXT HERE <ESMA_QUESTION_466> Q467: Do you consider there are any difficulties concerning the suggested approach for assessing whether the ancillary activities constitute a minority of activities at group level? Do you consider that the proposed calculations appropriately factor in activity which is subject to the permitted exemptions under Article 2(4) MiFID II? If no, please explain why and provide an alternative proposal. <ESMA_QUESTION_467> TYPE YOUR TEXT HERE <ESMA_QUESTION_467> Q468: Are there other approaches for assessing whether the ancillary activities constitute a minority of activities at group level that you would like to suggest? Please provide details and reasons. <ESMA_QUESTION_468> TYPE YOUR TEXT HERE <ESMA_QUESTION_468> Q469: How should minority of activities be defined? Should minority be less than 50% or less (50 - x)%? Please provide reasons. <ESMA_QUESTION_469> TYPE YOUR TEXT HERE <ESMA_QUESTION_469> 135

136 Q470: Do you have a view on whether economic or accounting capital should be used in order to define the elements triggering the exemption from authorisation under MiFID II, available under Article 2(1)(j)? Please provide reasons. <ESMA_QUESTION_470> TYPE YOUR TEXT HERE <ESMA_QUESTION_470> Q471: If economic capital were to be used as a measure, what do you understand to be encompassed by this term? <ESMA_QUESTION_471> TYPE YOUR TEXT HERE <ESMA_QUESTION_471> Q472: Do you agree with the above assessment that the data available in the TRs will enable entities to perform the necessary calculations? <ESMA_QUESTION_472> LSEG: This will be dependent upon the quality of the data submitted to the trade repository. There is a limit to the verification (point 46) that a TR can perform and this is often limited to simple validation. Additionally, the aggregations made public by the trade repositories are by high level asset class and appear not to have the granularity required. It is also questionable what data the trade repositories hold that the submitting firms are not also required to hold. <ESMA_QUESTION_472> Q473: What difficulties do you consider entities may encounter in obtaining the information that is necessary to define the size of their own trading activity and the size of the overall market trading activity from TRs? How could the identified difficulties be addressed? <ESMA_QUESTION_473> LSEG: The aggregations made public by the trade repositories are by high level asset class and appear not to have the granularity required. It is possible for the Trade Repositories to request further information from reporting counterparties, but they could not make population of these additional fields mandatory. <ESMA_QUESTION_473> Q474: What do you consider to be the difficulties in defining the volume of the transactions entered into to fulfil liquidity obligations? <ESMA_QUESTION_474> TYPE YOUR TEXT HERE <ESMA_QUESTION_474> Q475: How should the volume of the overall trading activity of the firm at group level and the volume of the transactions entered into in order to hedge physical activities be measured? (Number of contracts or nominal value? Period of time to be considered?) <ESMA_QUESTION_475> TYPE YOUR TEXT HERE <ESMA_QUESTION_475> Q476: Do you agree with the level of granularity of asset classes suggested in order to provide for relative comparison between market participants? <ESMA_QUESTION_476> TYPE YOUR TEXT HERE 136

137 <ESMA_QUESTION_476> Q477: What difficulties could there be regarding the aggregation of TR data in order to obtain information on the size of the overall market trading activity? How could these difficulties be addressed? <ESMA_QUESTION_477> LSEG: It is possible that the TRs may be using different methodologies to aggregate the data and there may be inconsistencies in the level of asset class granularity. <ESMA_QUESTION_477> Q478: How should ESMA set the threshold above which persons fall within MiFID II s scope? At what percentage should the threshold be set? Please provide reasons for your response. <ESMA_QUESTION_478> TYPE YOUR TEXT HERE <ESMA_QUESTION_478> Q479: Are there other approaches for determining the size of the trading activity that you would like to suggest? <ESMA_QUESTION_479> TYPE YOUR TEXT HERE <ESMA_QUESTION_479> Q480: Are there other elements apart from the need for ancillary activities to constitute a minority of activities and the comparison between the size of the trading activity and size of the overall market trading activity that ESMA should take into account when defining whether an activity is ancillary to the main business? <ESMA_QUESTION_480> TYPE YOUR TEXT HERE <ESMA_QUESTION_480> Q481: Do you see any difficulties with the interpretation of the hedging exemptions mentioned above under Article 2(4)(a) and (c) of MiFID II? How could potential difficulties be addressed? <ESMA_QUESTION_481> TYPE YOUR TEXT HERE <ESMA_QUESTION_481> Q482: Do you agree with ESMA s proposal to take into account Article 10 of the Commission Delegated Regulation (EU) No 149/2013 supplementing EMIR in specifying the application of the hedging exemption under Article 2(4)(b) of MiFID II? How could any potential difficulties be addressed? <ESMA_QUESTION_482> TYPE YOUR TEXT HERE <ESMA_QUESTION_482> Q483: Do you agree that the obligations to provide liquidity under Article 17(3) and Article 57(8)(d) of MiFID II should not be taken into account as an obligation triggering the hedging exemption mentioned above under Article 2(4)(c)? <ESMA_QUESTION_483> 137

138 TYPE YOUR TEXT HERE <ESMA_QUESTION_483> Q484: Could you provide any other specific examples of obligations of transactions in commodity derivatives and emission allowances entered into to fulfil obligations to provide liquidity on a trading venue which ESMA should take into account? <ESMA_QUESTION_484> TYPE YOUR TEXT HERE <ESMA_QUESTION_484> Q485: Should the (timeframe for) assessment be linked to audit processes? <ESMA_QUESTION_485> TYPE YOUR TEXT HERE <ESMA_QUESTION_485> Q486: How should seasonal variations be taken into account (for instance, if a firm puts on a maximum position at one point in the year and sells that down through the following twelve months should the calculation be taken at the maximum point or on average)? <ESMA_QUESTION_486> TYPE YOUR TEXT HERE <ESMA_QUESTION_486> Q487: Which approach would be practical in relation to firms that may fall within the scope of MiFID in one year but qualify for exemption in another year? <ESMA_QUESTION_487> TYPE YOUR TEXT HERE <ESMA_QUESTION_487> Q488: Do you see difficulties with regard to the two approaches suggested above? <ESMA_QUESTION_488> TYPE YOUR TEXT HERE <ESMA_QUESTION_488> Q489: How could a possible interim approach be defined with regard to the suggestion mentioned above (i.e. annual notification but calculation on a three years rolling basis)? <ESMA_QUESTION_489> TYPE YOUR TEXT HERE <ESMA_QUESTION_489> Q490: Do you agree that the competent authority to which the notification has to be made should be the one of the place of incorporation? <ESMA_QUESTION_490> TYPE YOUR TEXT HERE <ESMA_QUESTION_490> 7.2. Position Limits 138

139 Q491: Do you agree with ESMA s proposal to link the definition of a risk-reducing trade under MiFID II to the definition applicable under EMIR? If you do not agree, what alternative definition do you believe is appropriate? <ESMA_QUESTION_491> LSEG: Yes, we agree with ESMA s proposal. <ESMA_QUESTION_491> Q492: Do you agree with ESMA s proposed definition of a non-financial entity? If you do not agree, what alternative definition do you believe is appropriate? <ESMA_QUESTION_492> LSEG: Yes, we agree with ESMA s proposal. <ESMA_QUESTION_492> Q493: Should the regime for subsidiaries of a person other than entities that are wholly owned look to aggregate on the basis of a discrete percentage threshold or on a more subjective basis? What are the advantages and risks of either approach? Do you agree with the proposal that where the positions of an entity that is subject to substantial control by a person are aggregated, they are included in their entirety? <ESMA_QUESTION_493> LSEG: Yes, we agree with ESMA s proposal. <ESMA_QUESTION_493> Q494: Should the regime apply to the positions held by unconnected persons where they are acting together with a common purpose (for example, concert party arrangements where different market participants collude to act for common purpose)? <ESMA_QUESTION_494> TYPE YOUR TEXT HERE <ESMA_QUESTION_494> Q495: Do you agree with the approach to link the definition of economically equivalent OTC contract, for the purpose of position limits, with the definitions used in other parts of Mi- FID II? If you do not agree, what alternative definition do you believe is appropriate? <ESMA_QUESTION_495> LSEG: Yes, we agree with ESMA s proposal. <ESMA_QUESTION_495> Q496: Do you agree that even where a contract is, or may be, cash-settled it is appropriate to base its equivalence on the substitutability of the underlying physical commodity that it is referenced to? If you do not agree, what alternative measures of equivalence could be used? <ESMA_QUESTION_496> LSEG: We agree that even where a contract is, or may be, cash-settled, it is appropriate to base its equivalence on the substitutability of the underlying physical commodity that it is referenced to. <ESMA_QUESTION_496> Q497: Do you believe that the definition of economically equivalent that is used by the CFTC is appropriate for the purpose of defining the contracts that are not traded on a trading venue for the position limits regime of MiFID II? Give reasons to support your views as well as any suggested amendments or additions to this definition. <ESMA_QUESTION_497> 139

140 LSEG: Yes, we believe that the definition of economically equivalent that is used by the CFTC is appropriate for the purpose of defining the contracts that are not traded on a trading venue for the position limits regime of MiFID II. <ESMA_QUESTION_497> Q498: What arrangements could be put in place to support competent authorities identifying what OTC contracts are considered to be economically equivalent to listed contracts traded on a trading venue?? <ESMA_QUESTION_498> LSEG: We believe that Exchanges and clearing houses should provide competent authorities with detailed description of the listed/cleared contracts. Counterparties trading OTC derivatives should provide equivalent definitions to the competent authorities. <ESMA_QUESTION_498> Q499: Do you agree with ESMA s proposal that the same derivative contract occurs where an identical contract is listed independently on two or more different trading venues? What other alternative definitions of same could be applied to commodity derivatives? <ESMA_QUESTION_499> LSEG: We do not agree that a same derivative contract occurs where an identical contract is listed independently on two or more different trading venues. An OTC contract can be economically equivalent to a listed contract on a trading venue. Only the method of trading is different, which doesn't remove the equivalence. Therefore, a same derivative contract may occur when an identical contract is traded on a trading venue and Over the Counter. <ESMA_QUESTION_499> Q500: Do you agree with ESMA s proposals on aggregation and netting? How should ESMA address the practical obstacles to including within the assessment positions entered into OTC or on third country venues? Should ESMA adopt a model for pooling related contracts and should this extend to closely correlated contracts? How should equivalent contracts be converted into a similar metric to the exchange traded contract they are deemed equivalent to? <ESMA_QUESTION_500> LSEG: Yes, we agree with ESMA s proposal. <ESMA_QUESTION_500> Q501: Do you agree with ESMA s approach to defining market size for physically settled contracts? Is it appropriate for cash settled contracts to set position limits without taking into account the underlying physical market? <ESMA_QUESTION_501> LSEG: Yes, we agree with ESMA s approach to defining market size for physically settled contracts. <ESMA_QUESTION_501> Q502: Do you agree that it is preferable to set the position limit on a contract for a fixed (excluding exceptional circumstances) period rather than amending it on a real-time basis? What period do you believe is appropriate, considering in particular the factors of market evolution and operational efficiency? <ESMA_QUESTION_502> LSEG: We agree that it is preferable to set the position limit on a contract for a fixed (excluding exceptional circumstances) period rather than amending it on a real-time basis. We believe that the appropriate fixed period is three months. <ESMA_QUESTION_502> 140

141 Q503: Once the position limits regime is implemented, what period do you feel is appropriate to give sufficient notice to persons of the subsequent adjustment of position limits? <ESMA_QUESTION_503> LSEG: We believe that a period of one month is sufficient notice. <ESMA_QUESTION_503> Q504: Should positions based on contracts entered into before the revision of position limits be grandfathered and if so how? <ESMA_QUESTION_504> TYPE YOUR TEXT HERE <ESMA_QUESTION_504> Q505: Do you agree with ESMA s proposals for the determination of a central or primary trading venue for the purpose of establishing position limits in the same derivative contracts? If you do not agree, what practical alternative method should be used? <ESMA_QUESTION_505> LSEG: We do not believe in the determination of a central primary trading venue for the purpose of establishing position limits. We believe that the open interest of identical contracts on all trading venues listing those contacts should be added together and considered. <ESMA_QUESTION_505> Q506: Should the level of significant volume be set at a different level to that proposed above? If yes, please explain what level should be applied, and how it may be determined on an ongoing basis? <ESMA_QUESTION_506> Per our response to Q505, we do not believe in the determination of a central primary trading venue. <ESMA_QUESTION_506> Q507: In using the maturity of commodity contracts as a factor, do you agree that competent authorities apply the methodology in a different way for the spot month and for the aggregate of all other months along the curve? <ESMA_QUESTION_507> LSEG: In using the maturity of commodity contracts as a factor, we agree that competent authorities may apply the methodology in a different way for the spot month and for the aggregate of all other months along the curve. <ESMA_QUESTION_507> Q508: What factors do you believe should be applied to reflect the differences in the nature of trading activity between the spot month and the forward months? <ESMA_QUESTION_508> LSEG: We believe that deliverable supply, volatility and liquidity of different contract periods as well as open interest in each period should be the factors applied to reflect the differences in the nature of trading activity between the spot month and the forward months. <ESMA_QUESTION_508> Q509: Do you agree with ESMA s proposal for trading venues to provide data on the deliverable supply underlying their contracts? If you do not agree, what considerations should be given to determining the deliverable supply for a contract? 141

142 <ESMA_QUESTION_509> LSEG: We agree with ESMA s proposal for trading venues to provide data on the deliverable supply underlying their contracts. <ESMA_QUESTION_509> Q510: In the light of the fact that some commodity markets are truly global, do you consider that open interest in similar or identical contracts in non-eea jurisdictions should be taken into account? If so, how do you propose doing this, given that data from some trading venues may not be available on the same basis or in the same timeframe as that from other trading venues? <ESMA_QUESTION_510> LSEG: We believe that all trading venues, whether in EEA jurisdiction or not, should have their volumes in identical contracts considered. <ESMA_QUESTION_510> Q511: In the absence of published or easily obtained information on volatility in derivative and physical commodity markets, in what ways should ESMA reflect this factor in its methodology? Are there any alternative measures that may be obtained by ESMA for use in the methodology? <ESMA_QUESTION_511> TYPE YOUR TEXT HERE <ESMA_QUESTION_511> Q512: Are there any other considerations related to the number and size of market participants that ESMA should consider in its methodology? <ESMA_QUESTION_512> TYPE YOUR TEXT HERE <ESMA_QUESTION_512> Q513: Are there any other considerations related to the characteristics of the underlying commodity market that ESMA should consider in its methodology? <ESMA_QUESTION_513> LSEG: We believe that all considerations related to the characteristics of the underlying commodity market have already been taken into account by ESMA. <ESMA_QUESTION_513> Q514: For new contracts, what approach should ESMA take in establishing a regime that facilitates continued market evolution within the framework of Article 57? <ESMA_QUESTION_514> LSEG: We believe that ESMA should look at how other similar markets have evolved, with respect to growth in number of participants and open interest. <ESMA_QUESTION_514> Q515: The interpretation of the factors in the paragraphs above will be significant in applying ESMA s methodology; do you agree with ESMA s interpretation? If you do not agree with ESMA s interpretation, what aspects require amendment? <ESMA_QUESTION_515> LSEG: We agree with ESMA s proposal. <ESMA_QUESTION_515> 142

143 Q516: Are there any other factors which should be included in the methodology for determining position limits? If so, state in which way (with reference to the proposed methodology explained below) they should be incorporated. <ESMA_QUESTION_516> TYPE YOUR TEXT HERE <ESMA_QUESTION_516> Q517: What do you consider to be the risks and/or the advantages of applying a different methodology for determining position limits for prompt reference contracts compared to the methodology used for the position limit on forward maturities? <ESMA_QUESTION_517> LSEG: We believe that position limits should reflect the open interest in the forward periods and the volatility of the contract. Prompt position limits should reflect the supply picture of the underlying commodity. <ESMA_QUESTION_517> Q518: How should the position limits regime reflect the specific risks present in the run up to contract expiry? <ESMA_QUESTION_518> TYPE YOUR TEXT HERE <ESMA_QUESTION_518> Q519: If a different methodology is set for the prompt reference contract, would it be appropriate to make an exception where a contract other than the prompt is the key benchmark used by the market? <ESMA_QUESTION_519> TYPE YOUR TEXT HERE <ESMA_QUESTION_519> Q520: Do you agree that the baseline for the methodology of setting a position limit should be the deliverable supply? What concrete examples of issues do you foresee in obtaining or using the measure? <ESMA_QUESTION_520> LSEG: We agree that the baseline for the methodology of setting a position limit should be the deliverable supply for the prompt month. <ESMA_QUESTION_520> Q521: If you consider that a more appropriate measure exists to form the baseline of the methodology, please explain the measure and why it is more appropriate. Consideration should be given to the reliability and availability of such a measure in order to provide certainty to market participants. <ESMA_QUESTION_521> TYPE YOUR TEXT HERE <ESMA_QUESTION_521> Q522: Do you agree with this approach for the proposed methodology? If you do not agree, what alternative methodology do you propose, considering the full scope of the requirements of Article 57 MiFID II? <ESMA_QUESTION_522> TYPE YOUR TEXT HERE <ESMA_QUESTION_522> 143

144 Q523: Do you have any views on the level at which the baseline (if relevant, for each different asset class) should be set, and the size of the adjustment numbers for each separate factor that ESMA must consider in the methodology defined by Article 57 MiFID II? <ESMA_QUESTION_523> TYPE YOUR TEXT HERE <ESMA_QUESTION_523> Q524: Does the approach to asset classes have the right level of granularity to take into account market characteristics? Are the key characteristics the right ones to take into account? Are the conclusions by asset class appropriate? <ESMA_QUESTION_524> LSEG: We believe that ESMA s approach has right level of granularity to take into account market characteristics. <ESMA_QUESTION_524> Q525: What trading venues or jurisdictions should ESMA take into consideration in defining its position limits methodology? What particular aspects of these experiences should be included within ESMA s work? <ESMA_QUESTION_525> LSEG: We believe that ESMA should take into consideration all trading venues involved in the asset class should be considered, as well as physical market participants. <ESMA_QUESTION_525> Q526: Do you agree that the RTS should accommodate the flexibility to express position limits in the units appropriate to the individual market? Are there any other alternative measures or mechanisms by which position limits could be expressed? <ESMA_QUESTION_526> LSEG: We agree with ESMA that position limits should be stated both as: i) a percentage relationship between a position and some measure of absolute deliverable supply or overall market size measure; ii) an amount of lots of a specific contract on a trading venue; or expressed as a quantity of the underlying in the contract (which may be tonnes, barrels, MWh, etc). The method should be defined by what is appropriate to the individual market. <ESMA_QUESTION_526> Q527: How should the methodology for setting limits take account of a daily contract structure, where this exists? <ESMA_QUESTION_527> TYPE YOUR TEXT HERE <ESMA_QUESTION_527> Q528: Do you agree that limits for option positions should be set on the basis of delta equivalent values? What processes should be put in place to avoid manipulation of the process? <ESMA_QUESTION_528> LSEG: We agree that limits for option positions should be set on the basis of delta equivalent values. <ESMA_QUESTION_528> 144

145 Q529: Do you agree that the preferred methodology for the calculation of delta-equivalent futures positions is the use of the delta value that is published by trading venues? If you do not, please explain what methodology you prefer, and the reasons in favour of it? <ESMA_QUESTION_529> We agree with ESMA s approach that the preferred methodology for the calculation of delta-equivalent futures positions is the use of the delta value that is published by trading venues. <ESMA_QUESTION_529> Q530: Do you agree that the description of the approach outlined above, combined with the publication of limits under Article 57(9), would fulfil the requirement to be transparent and non-discriminatory? <ESMA_QUESTION_530> TYPE YOUR TEXT HERE <ESMA_QUESTION_530> Q531: What challenges are posed by transition and what areas of guidance should be provided on implementation? What transitional arrangements would be considered to be appropriate? <ESMA_QUESTION_531> TYPE YOUR TEXT HERE <ESMA_QUESTION_531> 7.3. Position Reporting Q532: Do you agree that, in the interest of efficient reporting, the data requirements for position reporting required by Article 58 should contain elements to enable competent authorities and ESMA to monitor effectively position limits? If you do not agree, what alternative approach do you propose for the collection of information in order to efficiently and with the minimum of duplication meet the requirements of Article 57? <ESMA_QUESTION_532> TYPE YOUR TEXT HERE <ESMA_QUESTION_532> Q533: Do you agree with ESMA s definition of a position for the purpose of Article 58? Do you agree that the same definition of position should be used for the purpose of Article 57? If you do not agree with either proposition, please provide details of a viable alternative definition. <ESMA_QUESTION_533> LSEG: Yes, we agree with ESMA s definition. <ESMA_QUESTION_533> Q534: Do you agree with ESMA s approach to the reporting of spread and other strategy trades? If you do not agree, what approach can be practically implemented for the definition and reporting of these trades? <ESMA_QUESTION_534> LSEG: Yes, we agree with ESMA s approach. <ESMA_QUESTION_534> 145

146 Q535: Do you agree with ESMA s proposed approach to use reporting protocols used by other market and regulatory initiatives, in particular, those being considered for transaction reporting under MiFID II? <ESMA_QUESTION_535> LSEG: Yes, we agree with ESMA s proposed approach. <ESMA_QUESTION_535> Q536: Do you have any specific comments on the proposed identification of legal persons and/or natural persons? Do you consider there are any practical challenges to ESMA s proposals? If yes, please explain them and propose solutions to resolve them. <ESMA_QUESTION_536> TYPE YOUR TEXT HERE <ESMA_QUESTION_536> Q537: What are your views on these three alternative approaches for reporting the positions of an end client where there are multiple parties involved in the transaction chain? Do you have a preferred solution from the three alternatives that are described? <ESMA_QUESTION_537> LSEG: We support the hybrid solution (point 32), however as a trading venue we wouldn't have the required information. <ESMA_QUESTION_537> Q538: What alternative structures or solutions are possible to meet the obligations under Article 58 to identify the positions of end clients? What are the advantages or disadvantages of these structures? <ESMA_QUESTION_538> TYPE YOUR TEXT HERE <ESMA_QUESTION_538> Q539: Do you agree with ESMA s proposal that only volumes traded on-exchange should be used to determine the central competent authority to which reports are made? If you do not agree, what alternative structure may be used to determine the destination of position reports? <ESMA_QUESTION_539> TYPE YOUR TEXT HERE <ESMA_QUESTION_539> Q540: Do you agree that position reporting requirements should seek to use reporting formats from other market or regulatory initiatives? If not mentioned above, what formats and initiatives should ESMA consider? <ESMA_QUESTION_540> LSEG: Yes, we agree. <ESMA_QUESTION_540> Q541: Do you agree that ESMA should require reference data from trading venues and investment firms on commodity derivatives, emission allowances, and derivatives thereof in order to increase the efficiency of trade reporting? <ESMA_QUESTION_541> LSEG: Yes, we agree. 146

147 For instrument identification, we believe that ISINs should be adopted as the sole instrument identifier. The product identifier is one of the key fields in any trade or transaction report. The main principle behind a product code is that it both uniquely identifies the instrument and enables participants to understand the attributes of that instrument through the associated reference data. ISINs are assigned for exchange traded derivatives and some OTC derivatives, such as cleared only and flex contracts, and there is no reason why the ISIN cannot be extended by the national numbering agencies to all derivatives. ISO standards were created to promote efficiency and understanding, using alternatives or multiple identifiers would result in inefficiencies. The dangers of using a product identifier that is not subject to an appropriate governance framework have been highlighted by the use of the Alternative Instrument Identifier for the MiFID transaction reporting of transactions on some derivative exchanges. We believe that the Aii has a number of weaknesses as an instrument identifier; instead of the identifier being used to understand the attributes of the instrument through associated reference data, the reference data elements are used to synthesise the identifier itself. Where firms make a mistake in reporting the reference data in a report, an incorrect identifier is produced. This limits the ability of CAs to carry out their surveillance activities and, if copied into trade repository reporting, will make it extremely difficult for regulators to determine positions in a particular derivative and to assess systemic risk. We would urge ESMA to recognise the importance of adopting a robust, established international standard for the instrument identifier. We believe the ISIN, as the ISO standard, should be this standard. <ESMA_QUESTION_541> Q542: What is your view on the use of existing elements of the market infrastructure for position reporting of both on-venue and economically equivalent OTC contracts? If you have any comments on how firms and trading venues may efficiently create a reporting infrastructure, please give details in your explanation. <ESMA_QUESTION_542> LSEG: We support using existing infrastructure used for EMIR trade reporting. <ESMA_QUESTION_542> Q543: For what reasons may it be appropriate to require the reporting of option positions on a delta-equivalent basis? If an additional requirement to report delta-equivalent positions is established, how should the relevant delta value be determined? <ESMA_QUESTION_543> LSEG: We believe the best, most objective measures are: premium and notional. We believe using delta equivalent basis would add unnecessary complexity. <ESMA_QUESTION_543> Q544: Does the proposed set of data fields capture all necessary information to meet the requirements of Article 58(1)(b) MiFID II? If not, do you have any proposals for amendments, deletions or additional data fields to add the list above? <ESMA_QUESTION_544> TYPE YOUR TEXT HERE <ESMA_QUESTION_544> Q545: Are there any other fields that should be included in the Commitment of Traders Report published each week by trading venues other than those shown above? <ESMA_QUESTION_545> TYPE YOUR TEXT HERE <ESMA_QUESTION_545> 147

148 8. Market data reporting 8.1. Obligation to report transactions Q546: Do you agree with ESMA s proposal for what constitutes a transaction and execution of a transaction for the purposes of Article 26 of MiFIR? If not, please provide reasons. <ESMA_QUESTION_546> LSEG: General Comments on Transaction reporting We support the extended scope of transaction reporting and ESMA s intention to clarify what constitutes reportable transactions. While we recognise that transaction reports need to contain sufficient data to enable NCAs to meet their obligations, we would question the inclusion of some of the proposed additional fields given the potential impact of implementing the required changes. We also raise the issue of non- MiFID firms (including from third countries), their possible non-compliance with new mandated fields (e.g. algo IDs and short selling flags) and the impact on trading venues who will have the responsibility for transaction reporting on their behalf. Specific response to Q546 LSEG: We welcome the additional focus in MiFID II on clarifying the definition of execution and transaction for the purposes of Article 26 MiFIR. We believe the definition provided is clear for all participants. <ESMA_QUESTION_546> Q547: Do you anticipate any difficulties in identifying when your investment firm has executed a transaction in accordance with the above principles? <ESMA_QUESTION_547> TYPE YOUR TEXT HERE <ESMA_QUESTION_547> Q548: Is there any other activity that should not be reportable under Article 26 of MiFIR? <ESMA_QUESTION_548> LSEG: We are not aware of any other activity that should not be reportable. <ESMA_QUESTION_548> Q549: Do you foresee any difficulties with the suggested approach? Please elaborate. <ESMA_QUESTION_549> LSEG: We do not provide specific comments on detailed conditions proposed by ESMA. However, we believe that the option of the transmitting firm to pass on the required information to the receiving firm will lead to additional confusion for reporting firms (and regulators). In addition, as we highlight in our responses to questions , there are a number of issues regarding the new data required for transaction reports (such as short selling flags, aggregated client orders) and these will also lead to implementation challenges in respect of transmitted orders. <ESMA_QUESTION_549> Q550: We invite your comments on the proposed fields and population of the fields. Please provide specific references to the fields which you are discussing in your response. <ESMA_QUESTION_550> LSEG: We fully support the concept that the transaction reports need to contain sufficient data to enable NCAs to meet the objectives outlined in Article 24. However, given the significant impact on reporting firms 148

149 and trading venues that we believe many of the proposed additional fields will have, we question whether the inclusion of all of these can be fully justified. In addition, we would highlight that firms not authorised under MiFID, per the exemptions of Article 2 MiFID II, and for whom the trading venue has the obligation to transaction report under article 26(5) MiFIR, are not subject to MiFIR and it may therefore be difficult for trading venues to impose on them the necessary changes/development required to comply with all of the new requirements. This issue is particularly relevant, and pressing, when third-country firms are involved. We would also note that, as trading venues, we can only be held responsible for processing and transmitting correctly the data provided to us by a participant, as we have no control over the accuracy and correctness of the data provided, nor can we have such control. In this context we comment below on a number of these proposed fields. Where a field is subject to a further question in section 8.1, we have not included our response here. 9. Trading Capacity. We have noticed that there is a degree of confusion on what constitutes an agency trade and how principal is defined across participants in different Member States. We believe there needs to be greater clarity on the definition of these capacities. 17. Consideration We would question the addition of this field when it can easily be calculated by the recipients of the data through other fields present in the report. 18. Venue of execution we believe additional clarification is required on which level MIC should be used (operational or segment). We believe it should be the segment level MIC , 30 38, 40 46, 49 55, Natural Person details. We would question the importance of collecting and storing names, addresses and dates of birth of individuals when identifiers are already collected. It results in significant data protection issues, significant duplication of data within reports, but brings, we believe, limited additional value. Counterparty and Client There needs to be further information on how the fields counterparty and client should be populated in relation to the trading capacity field. We believe the current suggestion that client identification is the same as the beneficiary can be dangerously misleading. In a principal transaction, the beneficiary can be the counterparty. Unless the definition of these fields becomes more precise (see trading capacity above), the reporting entities may not populate these fields in ways expected by the NCAs. 56 & 57 Instrument Identification Code We believe that the ISIN code has a robust and proven methodology, is internationally accepted and should be used for instrument identification; the Aii is a collection of reference data rather than an identification code, has no owner or governing body and does not, in our view, have a sufficiently robust methodology for this purpose. 60 & 61 Underlying Codes We would encourage ESMA to mandate use of the ISIN for underlying codes. ISINs are available for referential instruments and have been assigned to indices, for example. 91 Report Matching Number. We believe this will be a particularly difficult field for reporting firms to populate. Whilst we recognise the value of this field, the infrastructure for the communication of the appropriate codes across a chain of reporting entities is not currently adequate and we believe that it would be difficult for firms to implement this by the anticipated entry into force date. 93 Status. We believe values for new, cancel and update would be useful for reporting entities and NCAs alike. <ESMA_QUESTION_550> 149

150 Q551: Do you have any comments on the designation to identify the client and the client information and details that are to be included in transaction reports? <ESMA_QUESTION_551> LSEG: For legal persons, we see no reason why the LEI should not be used as the single permissible standard as we believe all legal persons are eligible for a LEI. However, we believe that there should be some form of temporary arrangement whereby new participants/clients awaiting allocation of a LEI can still be identified on a transaction report, pending that allocation For natural persons, we appreciate the value in having a single national level identifier. A number of natural persons will have dual nationality so, as is the case for residency, an alphabetical approach to determine which country s national coding scheme could be adopted. <ESMA_QUESTION_551> Q552: What are your views on the general approach to determining the relevant trader to be identified? <ESMA_QUESTION_552> LSEG: We appreciate the value this information will provide to regulators. However, we also recognise that the delivery of accurate information through the proposed framework will be complex and difficult to implement by reporting firms, particularly since the distinction between the trader that executes, as opposed to makes the investment decision, is unlikely to be made currently. The additional systems and controls required would be significant. In addition, where trading venues are reporting on behalf of non- MiFID authorised firms, we anticipate difficulties imposing these requirements on these firms and validating any values provided. <ESMA_QUESTION_552> Q553: In particular, do you agree with ESMA s proposed approach to assigning a trader ID designation for committee decisions? If not, what do you think is the best way for NCAs to obtain accurate information about committee decisions? <ESMA_QUESTION_553> LSEG: See also our response to Q552. It is not clear to us how investment decisions made by a committee(s) can be accurately and reliably captured in all cases and we would question the value of this information compared to the cost required to generate it accurately. <ESMA_QUESTION_553> Q554: Do you have any views on how to identify the relevant trader in the cases of Direct Market Access and Sponsored Access? <ESMA_QUESTION_554> LSEG: We believe that the firm offering DMA or sponsored access should not be required to identify the trader involved as it may not be able to identify the trader with any certainty this should be the requirement of the end-user firm. Where reporting responsibility rests with the DMA provider, they should ensure that the appropriate information is received from the DMA user. <ESMA_QUESTION_554> Q555: Do you believe that the approach outlined above is appropriate for identifying the computer algorithm within the investment firm responsible for the investment decision and the execution of the transaction? If not, what difficulties do you see with the approach and what do you believe should be an alternative approach? <ESMA_QUESTION_555> LSEG: See also response to Q552. We recognise the value of the information, but we believe it will be very difficult for the investment firm to assign the proper algo code to each single order/transaction; reference data for algo codes would be required to implement this requirement. Although the DP outlines criteria for the algo ID attribution, the parameters and configurations of the algo may also be relevant for 150

151 the behaviour of the algo itself. As such the algo ID information may, in such circumstances, be of limited value. <ESMA_QUESTION_555> Q556: Do you foresee any problem with identifying the specific waiver(s) under which the trade took place in a transaction report? If so, please provide details. <ESMA_QUESTION_556> LSEG: In many cases (e.g. where firm trades on own account with an agency broker, or directly with a client) we would question whether the counterparty would be aware of whether a trade was executed under a particular waiver. We would, therefore, expect in these circumstances that the identification would be reported by one party to a trade (i.e. the firm dealing on own account). <ESMA_QUESTION_556> Q557: Do you agree with ESMA s proposed approach to adopt a simple short sale flagging approach for transaction reports? If not, what other approaches do you believe ESMA should consider and why? <ESMA_QUESTION_557> LSEG: We believe the simple short sale flagging approach, as outlined in para 94 of the DP, is the most appropriate way to identify that a transaction represents a short sale of shares or sovereign debt. The introduction of a partial short flag or the use of a % for partial shorts would, we believe, be extremely complicated to administer. While we do not believe that the technical changes required to incorporate this flag are complex, we would again highlight the issue of whether this field will be correctly populated by non-mifid authorised firms, for whom this requirement may not apply, and that trading venues transaction reporting on their behalf will only be able to include values they have received. <ESMA_QUESTION_557> Q558: Which option do you believe is most appropriate for flagging short sales? Alternatively, what other approaches do you think ESMA should consider and why? <ESMA_QUESTION_558> LSEG: ESMA recognises the problems associated with both options. Of the two, Option 1 would be more appropriate. A requirement for investment firms to calculate the holdings of their clients to determine whether the transaction represents a short sale would be complex (particularly if the firm is not a custodian or the client uses a number of custodians) and the development required to implement would outweigh any suggested benefits. See also our response to Q557. <ESMA_QUESTION_558> Q559: What are your views regarding the two options above? <ESMA_QUESTION_559> LSEG: Option 1 would be our preferred option. Option 2, would, we believe, be unworkable in practice. See also our response to Q557. <ESMA_QUESTION_559> Q560: Do you agree with ESMA s proposed approach in relation to reporting aggregated transactions? If not, what other alternative approaches do you think ESMA should consider and why? <ESMA_QUESTION_560> LSEG: See our response to Q557. <ESMA_QUESTION_560> 151

152 Q561: Are there any other particular issues or trading scenarios that ESMA should consider in light of the short selling flag? <ESMA_QUESTION_561> LSEG: See our response to Q557. <ESMA_QUESTION_561> Q562: Do you agree with ESMA s proposed approach for reporting financial instruments over baskets? If not, what other approaches do you believe ESMA should consider and why? <ESMA_QUESTION_562> LSEG: We believe there are considerable reference data issues involved in determining whether the index/basket falls into the three options. To do this accurately, firms would not only need to know the current composition and weightings of the indices but also need to react to future changes to the composition and weightings on the indices. This process will also be complicated by the fact that the constituent financial instruments of the indices will de-list or list on trading venues over time (the same will be true for the underlying indices in option 3). The only solution might be to require all index derivatives to be reported. If reporting entities consider this more onerous, then maybe a reference data provider (or ESMA) can supply details on which indices are reportable. <ESMA_QUESTION_562> Q563: Which option is preferable for reporting financial instruments over indices? Would you have any difficulty in applying any of the three approaches, such as determining the weighting of the index or determining whether the index is the underlying in another financial instrument? Alternatively, are there any other approaches which you believe ESMA should consider? <ESMA_QUESTION_563> LSEG: See our response to Q562. <ESMA_QUESTION_563> Q564: Do you think the current MiFID approach to branch reporting should be maintained? <ESMA_QUESTION_564> LSEG: We believe that the current MiFID approach to branch reporting across the EEA is not well understood by firms and we recognise the issue that ESMA raises whereby firms may report the same transaction to more than one NCA, or to none, resulting in misleading information for regulators. <ESMA_QUESTION_564> Q565: Do you anticipate any difficulties in implementing the branch reporting requirement proposed above? <ESMA_QUESTION_565> LSEG: The head office of the branch may not always have the required information as it is held at the branch, so this could add significant overheads for firms. The systems and controls over transaction reporting are often held at the branch rather than the head office, so this could also be a major disruption. It appears that the proposal aims to provide certainty to firms over where they should report. However, we would argue that this can be achieved without requiring head office reporting. It would be simple for branches to continue to report through ARMs and the ARMs can report to the correct home state NCA without disturbing firms infrastructures. So long as there are harmonised delivery/connectivity standards across all NCAs and no NCA unfairly favours a particularly delivery route, there should be no difficulties in adopting this approach. <ESMA_QUESTION_565> 152

153 Q566: Is the proposed list of criteria sufficient, or should ESMA consider other/extra criteria? <ESMA_QUESTION_566> LSEG: The new criteria appear over-complicated. Criteria ii and iii may perhaps be better replaced with a more simple and potentially more accurate member state of branch. <ESMA_QUESTION_566> Q567: Which format, not limited to the ones above, do you think is most suitable for the purposes of transaction reporting under Article 26 of MiFIR? Please provide a detailed explanation including cost-benefit considerations. <ESMA_QUESTION_567> LSEG: Of the formats identified, our preferred format would be (iii) CSV. It is simple, easy to generate, easy to compress, has no licence costs and is easy for the recipient to process and understand. <ESMA_QUESTION_567> 8.2. Obligation to supply financial instrument reference data Q568: Do you anticipate any difficulties in providing, at least daily, a delta file which only includes updates? <ESMA_QUESTION_568> LSEG: No, we do not anticipate any difficulties in providing this. <ESMA_QUESTION_568> Q569: Do you anticipate any difficulties in providing, at least daily, a full file containing all the financial instruments? <ESMA_QUESTION_569> LSEG: No, we do not anticipate any difficulties in providing this. <ESMA_QUESTION_569> Q570: Do you anticipate any difficulties in providing a combination of delta files and full files? <ESMA_QUESTION_570> LSEG: No, we do not see any issues with a combined approach. We currently prepare both types of file. However, we assume that any proposed timetable for such a combined approach would relate to the provision of this data to NCAs only and not to the provision of reference data to other users. <ESMA_QUESTION_570> Q571: Do you anticipate any difficulties in providing details of financial instruments twice per day? <ESMA_QUESTION_571> LSEG: Yes - distribution once a day would be more appropriate. Notification s regarding suspensions/halts would still be provided intra day. In addition, it would be useful for ESMA to clarify whether activities by Debt Management Offices in relation to sovereign bonds (e.g. buy-backs) should be considered as events triggering instrument reference data changes. <ESMA_QUESTION_571> 153

154 Q572: What other aspects should ESMA consider when determining a suitable solution for the timeframes of the notifications? Please include in your response any foreseen technical limitations. <ESMA_QUESTION_572> LSEG: See response to Q571. Notifications should be once a day and should fit in with the processes to update reference data run by the trading venues. Notifications would be made at the end of the trading day or at the start of the following day in both cases they would be received by the CA before the start of trading relevant to that updated reference data.<esma_question_572> Q573: Do you agree with the proposed fields? Do trading venues and investment firms have access to the specified reference data elements in order to populate the proposed fields? <ESMA_QUESTION_573> LSEG: We broadly agree with the proposed fields as detailed, although a number of fields proposed would require systems development to implement. However, we would note that this is still very securities focussed and we believe that there are a number of non-securities derivative reference data that are not included. In addition, a number of the fields may not be currently available and therefore the impact of including them must me assessed, where this is the case. In particular, we would note the following: As a trading venue, we do not possess details of the depository bank issuer of GDRs and ADRs; only the issuer of the underlying instrument. We are unsure what value details of the depository bank will provide. We presume trading venue/systematic internaliser only refers to our own trading venue. We may not know all the other trading venues / SIs where the instruments may trade. (Debt) Issuer Type we do not currently hold this information. (Debt) Ultimate Issuer Name/Country Code it is not clear to us how ultimate issuer differs from issuer unless this is also intended for ADR/GDRs on debt. (Debt) it is not clear how the seniority of the bond should be identified. Issuer s Group clarification is required as to what is meant by this. Rest of debt fields Lei of guarantor, Guarantor s group, Fixed rate (coupon?), Index Benchmark (for FRNs), seniority of bond, issuance price, reimbursement price (is this not par value?). These are not captured as part of existing reference data, therefore the impact of costs/feasibility of capturing this data would be required. Country of the underlying Index It is unclear how this is to be defined and whether this relates to the country of the index owner/calculating agent or criteria on which the index is based. There appears to be nothing on convertible equity or debt (e.g. what security is it convertible to?) Derivatives how is the full instrument name defined? Options we feel additional provision of a put/call identifier (not included in the list) could be useful for regulators. Entitlements (Rights) - In relation to warrants, an additional field for covered/uncovered could be useful for regulators. Instrument Identifiers we believe that ISINs should be adopted as the sole instrument identifier. The product identifier is one of the key fields in any trade or transaction report. The main principle behind a product code is that it both uniquely identifies the instrument and enables participants to understand the attributes of that instrument through the associated reference data. ISINs are assigned for exchange traded derivatives and some OTC derivatives, such as cleared only and flex contracts, and there is no reason why the ISIN cannot be extended by the national numbering agencies to all derivatives. ISO standards were created to promote efficiency and understanding, using alternatives or multiple identifiers would result in inefficiencies. The dangers of using a product identifier that is not subject to an appropriate governance framework have been highlight- 154

155 ed by the use of the Alternative Instrument Identifier for the MiFID transaction reporting of transactions on some derivative exchanges. We believe that the Aii has a number of weaknesses, has no owner or governing body and does not, in our view, have a sufficiently robust methodology for this purpose. As an instrument identifier; instead of the identifier being used to understand the attributes of the instrument through associated reference data, the reference data elements are used to synthesise the identifier itself. Where firms make a mistake in reporting the reference data in a report, an incorrect identifier is produced. This limits the ability of CAs to carry out their surveillance activities and, if copied into trade repository reporting, will make it extremely difficult for regulators to determine positions in a particular derivative and to assess systemic risk. We would urge ESMA to recognise the importance of adopting a robust, established international standard for the instrument identifier. We believe the ISIN, as the ISO standard, should be the sole standard adopted for MiFIR reporting. It is not clear from the DP whether reporting firms need to populate reference data associated with the instrument identifier on transaction reports. We would encourage ESMA to adopt the current MiFID practice of not requiring this data to be populated by firms as errors do occur. It would be preferable for the regulators to use the reference data they collect from trading venues for this purpose. <ESMA_QUESTION_573> Q574: Are you aware of any available industry classification standards you would consider appropriate? <ESMA_QUESTION_574> LSEG: We believe that ISO standards should be used as far as possible. The ISO CFI standard should prove suitable, but please note that this is undergoing a fundamental review, rather than the simple expansion alluded to in footnote 219. For instrument identification (see also Q573) we believe that ISINs should be adopted as the sole instrument identifier. The product identifier is one of the key fields in any trade or transaction report. The main principle behind a product code is that it both uniquely identifies the instrument and enables participants to understand the attributes of that instrument through the associated reference data. ISINs are assigned for exchange traded derivatives and some OTC derivatives, such as cleared only and flex contracts, and there is no reason why the ISIN cannot be extended by the national numbering agencies to all derivatives. ISO standards were created to promote efficiency and understanding, using alternatives or multiple identifiers would result in inefficiencies. The dangers of using a product identifier that is not subject to an appropriate governance framework have been highlighted by the use of the Alternative Instrument Identifier for the MiFID transaction reporting of transactions on some derivative exchanges. We believe that the Aii has a number of weaknesses, has no owner or governing body and does not, in our view, have a sufficiently robust methodology for this purpose. As an instrument identifier; instead of the identifier being used to understand the attributes of the instrument through associated reference data, the reference data elements are used to synthesise the identifier itself. Where firms make a mistake in reporting the reference data in a report, an incorrect identifier is produced. This limits the ability of CAs to carry out their surveillance activities and, if copied into trade repository reporting, will make it extremely difficult for regulators to determine positions in a particular derivative and to assess systemic risk. We would urge ESMA to recognise the importance of adopting a robust, established international standard for the instrument identifier. We believe the ISIN, as the ISO standard, should be the sole standard adopted for MiFIR reporting. <ESMA_QUESTION_574> Q575: For both MiFID and MAR (OTC) derivatives based on indexes are in scope. Therefore it could be helpful to publish a list of relevant indexes. Do you foresee any difficulties in providing reference data for indexes listed on your trading venue? Furthermore, what reference data could you provide on indexes? <ESMA_QUESTION_575> 155

156 LSEG: Indices do not list on trading venues; but there are derivatives and ETFs that are based on indices. We believe it is possible to produce a list of indices that are the basis for ETFs and derivatives trading on our exchanges. ISINs are assigned for indices, but it is the responsibility of the index owner/calculating agent to provide the relevant reference data, such as constituency and weighting information. The relevant NNA based on the jurisdiction of the index owner/calculating agent assigns an ISIN. The reference data readily available for indices would be ISIN as an identifier, CFI for classification, index name, index owner/calculating agent and country of index owner/calculating agent. A country of index for non-securities indices might prove difficult without precise guidance. <ESMA_QUESTION_575> Q576: Do you agree with ESMA s intention to maintain the current RCA determination rules? <ESMA_QUESTION_576> LSEG: We agree with the intention to maintain the current RCA determination rules. <ESMA_QUESTION_576> Q577: What criteria would you consider appropriate to establish the RCA for instruments that are currently not covered by the RCA rule? <ESMA_QUESTION_577> TYPE YOUR TEXT HERE <ESMA_QUESTION_577> <ESMA_QUESTION_1> NOT A VALID QUESTION HERE? <ESMA_QUESTION_1> 8.3. Obligation to maintain records of orders Q578: In your view, which option (and, where relevant, methodology) is more appropriate for implementation? Please elaborate. <ESMA_QUESTION_578> LSEG: Option 1 is our preferred option. However, it is likely that further development would still be required by trading venues (and therefore costs incurred) to capture the additional data required by MiFIR. We agree with ESMA that this would remove the risk of losing important granular information and this option would also avoid losing relevant information, for example where trading venues provide specific functionality or complex negotiation mechanisms (e.g. RFQ-like). Where there is a need for data collection by a CA, then they would probably be in a better position to convert the date. They have the overall view of the market and they might be in a better position to transform and reconcile the data. <ESMA_QUESTION_578> Q579: In your view, what are the data elements that cannot be harmonised? Please elaborate. <ESMA_QUESTION_579> LSEG: Examples of data elements that cannot be harmonised would be order types and options that may vary across different venues. Users can set a minimum quantity to be executed with their orders, order cancellation/suspension could work differently across trading venues, there may be specific options available on certain trading venues only (e.g. parallel quotation mechanisms). In addition, changing data ele- 156

157 ments and their construction/formats may be incompatible with the other data structures in complex data warehouse data models of trading venues. <ESMA_QUESTION_579> Q580: For those elements that would have to be harmonised under Option 2 or under Option 3, do you think industry standards/protocols could be utilised? Please elaborate. <ESMA_QUESTION_580> TYPE YOUR TEXT HERE <ESMA_QUESTION_580> Q581: Do you foresee any difficulties with the proposed approach for the use of LEI? <ESMA_QUESTION_581> LSEG: We strongly support the extended use of LEI for identification of any member or participant of the trading platform. This will ensure transparency and consistency across different trading venues. However, this will require system changes for many trading venues and therefore has cost implications. <ESMA_QUESTION_581> Q582: Do you foresee any difficulties maintaining records of the Client IDs related with the orders submitted by their members/participants? If so, please elaborate. <ESMA_QUESTION_582> LSEG: We do not foresee particular difficulties with ESMA s proposed approach, although trading venues may not currently receive this information from trading participants and would therefore need to plan the necessary changes to the trading system to capture any additional fields. However, ESMA will need to consider the sensitivity of client information, whether the passing of this information breaches any client/broker confidentiality agreements and the extent to which this creates new security/data protection issues for venues. In addition, we agree that ESMA needs to consider the case of aggregated client orders (para 35, page 503 DP), where a firm would not necessarily know the identity of individual clients behind an order they have aggregated to manage their flow efficiently. We suggest that a suitable client group identifier specific to that firm would be required. <ESMA_QUESTION_582> Q583: Are there any other solutions you would consider as appropriate to track clients order flows through member firms/participants of trading venues and to link orders and transactions coming from the same member firm/participant? <ESMA_QUESTION_583> LSEG: No. The use of client/trader IDs by members/participants on order messages is an important component of maintaining continuity of this information through the trade cycle, e.g. from firm to trading venue to CCP. However, note our response to Q582. <ESMA_QUESTION_583> Q584: Do you believe that this approach allows the order to be uniquely identified If not, please elaborate <ESMA_QUESTION_584> LSEG: Yes, if item iii. in para 45 of the DP (date of receipt) includes a full timestamp (i.e. date and time). In addition, we would note that, further to comments made to Q573, there is increased risk in the ability to ensure the uniqueness of the identification code of the order where the inclusion of the Aii product code is permissible. We believe ISIN, as an ISO standard with robust governance and assignment principles, should be the sole standard adopted for MiFIR reporting. 157

158 <ESMA_QUESTION_584> Q585: Do you foresee any difficulties with the implementation of this approach? Please elaborate <ESMA_QUESTION_585> LSEG: We do not see any specific difficulties; however these requirements will have IT impacts and potentially significant cost and development time impact. It is unclear from the DP whether ESMA recommends keeping the same order number when an order has been modified or cancelled or whether it recommends generating new order codes every time the order is amended. It may not be able to link a modified/cancelled order back to the original order. <ESMA_QUESTION_585> Q586: Do you foresee any difficulties with the proposed approach? Please elaborate <ESMA_QUESTION_586> LSEG: We do not see any specific difficulties but these requirements will have an IT impact; some development work may be needed and coordination and accurate time synchronisation across venues may prove challenging. <ESMA_QUESTION_586> Q587: Do you foresee any difficulties with the proposed approach? Please elaborate. <ESMA_QUESTION_587> LSEG: We do not see any specific difficulties using a sequence number across different types of messages. However, this may require significant changes to existing systems. <ESMA_QUESTION_587> Q588: Would the breakdown in the two categories of order types create major issues in terms of mapping of the orders by the Trading Venues and IT developments? Please elaborate <ESMA_QUESTION_588> LSEG: Yes, complexity is introduced in producing this kind of report from a trading system IT perspective. Management of order status is specific to each trading system and its harmonization is likely to impact every system and would create additional cost and operational issues for trading venues to implement this mapping. We therefore question whether any suggested benefits would outweigh the development required to implement these changes. Order types listed by a trading venue can be categorised per the ESMA requirement in static reference data. There appears to be no reason to require that these categories are recognised and implemented within the trading system itself. For the inclusion of any new order types introduced by a venue, these should be acknowledged by the NCA and allocated one of the categories. <ESMA_QUESTION_588> Q589: Do you foresee any problems with the proposed approach? <ESMA_QUESTION_589> LSEG: See response to Q588. Further complexity is introduced and would create additional cost and operational issues for trading venues to implement this approach. <ESMA_QUESTION_589> Q590: Are the proposed validity periods relevant and complete? Should additional validity period(s) be provided? Please elaborate. <ESMA_QUESTION_590> 158

159 LSEG: We agree with the principle of allowing traders to input the exact time / date until which the order will be valid. We broadly agree with the validity periods highlighted by ESMA but would add: We also have a Good for Closing Price Crossing Session validity period, where an order is injected at start of the Closing Price Crossing Session with any remaining volume expired after the end of that session or on confirmation that there will be no Closing Price Crossing Session in that instrument that day; Good-till-time: Additionally, any GTT orders with an expiry time during any auction call phase will not be deleted until after uncrossing has completed. <ESMA_QUESTION_590> Q591: Do you agree that standardised default time stamps regarding the date and time at which the order shall automatically and ultimately be removed from the order book relevantly supplements the validity period flags? <ESMA_QUESTION_591> LSEG: We do not believe that this will be practical due to the existence of validity periods (such as Good for Closing Price Crossing Session see response to Q590) that are linked to the end of an auction period, which have random period ends. We believe that this should be left for trading venues to decide/calibrate. <ESMA_QUESTION_591> Q592: Do venues use a priority number to determine execution priority or a combination of priority time stamp and sequence number? <ESMA_QUESTION_592> LSEG: We currently do not give priority numbers to each order as it arrives and then moves within the order book as other orders around it move. For price-time priority venues, the time stamp defines order priority at any point in time. <ESMA_QUESTION_592> Q593: Do you foresee any difficulties with the three options described above? Please elaborate. <ESMA_QUESTION_593> LSEG: Introducing an integer based priority and moving away from time/price based priority would be a major change. Such a move would require a dynamic order book to be maintained, prioritised to allocate an explicit priority, on our systems, since the priority of an order may change over its lifetime. This will introduce latency and would require the retransmission of all orders on the book with updated priorities upon any execution or new order addition - this will impact bandwidth requirements and would represent an inefficient mechanism for data dissemination. It is important to note also that different venues can operate different matching logics, and not all use price/time priority. This means that comparison of priority between orders will need to be understood in the context of the matching logic of the trading system. This also applies if comparing orders across trading systems. Existing market data customers, surveillance systems, analytics providers and others are all currently able to rebuild an accurate view of the full order book real-time and retrospectively based on the audit trail of order events within a security up to the point in time to be assessed without the need for order book priority at the point in time. All that is required is the sequence number of the event and/or an accurate internal timestamp. For regulatory purposes, this includes non-displayed order volumes (e.g. iceberg reserve sizes) that are not published to market. The need to re-allocate an order priority integer every time an order priority changes would create significant additional development, management and storage of order data with no benefit. 159

160 Furthermore, orders are not always managed on a time-priority basis. For example, if more than one iceberg order was resting at a single price point, once the visible peaks of these icebergs and any other visible orders are exhausted by an incoming liquidity-taking order, any iceberg reserve sizes are allocated volume on a pro-rata basis and not in turn. Therefore, the two reserve sizes effectively the same priority since some of each will be allocated but must be identified with a different order priority type since they must be considered differently from visible orders. Also, other venues may operate a size-priority basis whose orders can clearly be allocated a priority number but can also be easily rebuilt based on the order attributes of the orders present without the explicit priority presently assigned. Other factors affecting the order priority include such features employed variously across European trading venues as self-execution prevention and member priority matching, each of which could result in the order priority being different for one participant compared to another. If any one of the three options would be required, only option (i) would be in any way logical or manageable but as described above, this still is unnecessary since an order book can be rebuilt for any instrument based on the order event time, sequence number and other relevant order attributes such as order size as is relevant for the specific matching logic. Overall, we believe that rather than try to standardise a method of order priority across disparate matching logics and rules, an NCA should understand the matching logic and rules employed by its regulated venues and the order data maintained in order to rebuild order books in the appropriate manner in each case, noting that most logics will be very similar. <ESMA_QUESTION_593> Q594: Is the list of specific order instructions provided above relevant? Should this list be supplemented? Please elaborate. <ESMA_QUESTION_594> LSEG: On a number of our trading systems we do not currently have some of the proposed characteristics (e.g. MES, MAQ, modification number, passive only indicator, passive or aggressive indicator). There would therefore be a significant impact and cost to develop the trading system to incorporate the new data. See also our response to Q593 on order modification/priority. <ESMA_QUESTION_594> Q595: Are there any other type of events that should be considered? <ESMA_QUESTION_595> LSEG: We believe the list looks comprehensive and relevant. We would add that in the case where last look or manual acceptance mechanisms are in place, events could also include rejection by the counterparty. <ESMA_QUESTION_595> Q596: Do you foresee any difficulties with the proposed approach? Please elaborate. <ESMA_QUESTION_596> LSEG: In some of the markets operated by LSEG, a value of S is also used to denote riskless principal orders. For these markets, therefore, this should also be a valid value. Note that for the purposes of data passed to the relevant CCPs, the values of A and P only are used (the value S is converted to P). <ESMA_QUESTION_596> Q597: Do you foresee any problems with the proposed approach? Do you consider any other alternative in order to inform about orders placed by market makers and other liquidity providers? <ESMA_QUESTION_597> 160

161 LSEG: Where orders entered are in the form of executable quotes, we do not foresee any issues. In the case of other orders entered by a MM or liquidity provider, a combination of fields/values can be used to flag such orders. We would like clarification, however, in the case of liquidity provision on behalf of an issuer, since all market making activity is proprietary. In addition, the existence of a liquidity provider on behalf of an issuer does not mean that the market maker is trading with issuer as client. The wording in the DP is confusing. <ESMA_QUESTION_597> Q598: Do you foresee any difficulties in generating a transaction ID code that links the order with the executed transaction that stems from that order in the information that has to be kept at the disposal of the CAs? Please elaborate. <ESMA_QUESTION_598> LSEG: We do not see difficulties with this approach, although this will require system changes in many cases. <ESMA_QUESTION_598> Q599: Do you foresee any difficulties with maintaining this information? Please elaborate. <ESMA_QUESTION_599> LSEG: We do not foresee any difficulties with maintaining this information. <ESMA_QUESTION_599> 8.4. Requirement to maintain records of orders for firms engaging in highfrequency algorithmic trading techniques (Art. 17(7) of MIFID II)17 Q600: Do you foresee any difficulties with the elements of data to be stored proposed in the above paragraph? If so, please elaborate. <ESMA_QUESTION_600> TYPE YOUR TEXT HERE. <ESMA_QUESTION_600> Q601: Do you foresee any difficulties in complying with the proposed timeframe? <ESMA_QUESTION_601> TYPE YOUR TEXT HERE <ESMA_QUESTION_601> 8.5. Synchronisation of business clocks Q602: Would you prefer a synchronisation at a national or at a pan-european level? Please elaborate. If you would prefer synchronisation to a single source, please indicate which would be the reference clock for those purposes. <ESMA_QUESTION_602> 17 Please note that this section has to be read in conjunction with the section on the Record keeping and co-operation with national competent authorities in this DP. 161

162 National v. pan-european Level LSEG s view is that synchronisation at a pan-european level under UTC notation should be the preferred approach. For trading systems of venues and investment firms, our suggestion is that synchronisation should be achieved through the introduction of the Network Time Protocol (NTP). General comments on synchronisation and precision In general, LSEG supports improvements to time precision and clock synchronisation to achieve an accurate, transparent and comparable view of trading environments across the EU; although there may be technical challenges in achieving the lowest granularity by regulatory mandate. LSEG recognises here that that upgrading live trading systems to achieve a higher degree of precision is not a trivial or inexpensive proposition for investment firms and trading venues, and usually requires a significant upgrade of the technology infrastructure. LSEG venues already operate to millisecond accuracy across both trading and data systems, with plans to introduce microsecond accuracy through PTP synchronisation as platforms evolve. LSEG also recognises that PTP synchronisation may be considered more difficult to implement across the EU Single Market, and potentially may be disproportionate to the requirement in MiFID Article 50, as different market structures operate at varying degrees of automation and firms operate with different degrees of technical expertise. We note that this should not prevent firms from adopting a lower level of granularity, if they deem it suitable for their businesses. However, as a regulatory requirement, NTP with millisecond accuracy is suitable, and hence the Group s recommendation. <ESMA_QUESTION_602> Q603: Do you agree with the requirement to synchronise clocks to the microsecond level? <ESMA_QUESTION_603> Please see our response to Q602. LSEG suggests that NTP synchronisation with millisecond accuracy should be ESMA s proposed approach. LSEG recognises here that that upgrading live trading systems to achieve microsecond precision is not a trivial or inexpensive proposition for investment firms and trading venues, and usually requires a significant upgrade of the technology infrastructure. LSEG venues already operate to millisecond accuracy across both trading and data systems, with plans to introduce microsecond accuracy through PTP synchronisation as platforms evolve. LSEG also recognises that PTP synchronisation at microsecond level will be far more difficult to implement across the EU Single Market, and potentially may be disproportionate to the requirement in MiFID Article 50, as different market structures operate at varying degrees of automation and firms operate with different degrees of technical expertise. We note that this should not prevent firms from adopting a lower level of granularity, if they deem it suitable for their businesses. However, as a regulatory requirement, NTP with millisecond accuracy is suitable, and hence the Group s recommendation <ESMA_QUESTION_603> Q604: Which would be the maximum divergence that should be permitted with respect to the reference clock? How often should any divergence be corrected? <ESMA_QUESTION_604> A maximum divergence of a millisecond from the masterclock should permissible. <ESMA_QUESTION_604> 162

163 9. Post-trading issues 9.1. Obligation to clear derivatives traded on regulated markets and timing of acceptance for clearing (STP) Q605: What are your views generally on (1) the systems, procedures, arrangements supporting the flow of information to the CCP, (2) the operational process that should be in place to perform the transfer of margins, (3) the relevant parties involved these processes and the time required for each of the steps? <ESMA_QUESTION_605> LSEG General Comments to section 9 on post-trading issues In answering this section, LSEG draws on its experience of operating multiple trading venues for the trading of exchange traded derivatives (ETDs) e.g. LSE Derivatives, IDEM, IDEX and AGREX, and OTC derivatives (OTCDs), including MTS Swaps. In addition, we offer clearing for ETDs through the Group s EMIR authorised CCPs, including CC&G, LCH.Clearnet Ltd and LCH.Clearnet SA, and for OTCDs, through the LCH.Clearnet Ltd s SwapClear, ForexClear and EnClear services, and LCH.Clearnet SA s CDSClear service. LCH.Clearnet Ltd and LCH.Clearnet SA are dual authorised, in the EU under EMIR and in the US under the Commodity Exchange Act as Derivatives Clearing Organisations (DCOs) regulated by the CFTC. As a regulatory aim, we support cross border harmonisation and where appropriate, we use our experience of operating in multiple regulatory jurisdictions to comment on these issues. -- Specific response to Q605 In general, LSEG supports greater automation and certainty in the clearing process and the use of straight through processing (STP) in this regard. In ETDs and OTCDs, a high degree of STP exists, and the current model works well. For products executed fully electronically on trading venues, whether as ETDs or OTCDs, trading venues communicate transaction flows to the CCPs involved on a real time basis, on behalf of their market participants. The CCPs calculate the net position and communicate them to the clearing members. This model requires for the operational processes for the transfer of margin to be defined in the rulebook and procedures of the CCPs, and the ability to comply with this processes is a condition for membership. The operational process starts when CCP registers the trades and calls for margins, which frequently occurs on an intraday basis. For OTCDs, there are 3 areas that we wish to emphasise when ESMA develops an EEA regime for STP: (1) We believe that trading venues should not be obliged to void trades that are rejected for clearing (please see our answer to Q613 for further details); (2) We believe that a post-execution affirmation process should continue to be allowed for, at a minimum, trades that are not executed fully electronically; (3) We believe that CCPs should not be obliged to facilitate pre-execution clearing guarantees; and (4) To the extent ESMA wishes to require clearing members to conduct limit checks on the orders of their customers prior to execution (i.e., not a guarantee that a CCP will clear a trade, but a guarantee that the clearing member will accept the trade should the trade be cleared), we request that ESMA consider exceptions where pre-execution checks are not currently practicable, including for trades that are not executed fully electronically, trades that are executed initially as uncleared trades, but are subsequently 163

164 submitted for clearing, and bilateral trades that exist prior to the clearing mandate (including those caught under the front-loading requirement). For at least these trades, we wish for the CCP to continue to be allowed to facilitated a post-execution, pre-clearing process for the clearing member to accept or reject said trades. With regards to the time required for a CCP to clear a trade once submitted, we suggest that a CCP should accept (or reject) a trade submitted to clearing within 10 seconds of receipt. This is the same time as under CFTC requirements in the US. As mentioned above, for both ETDs and OTCDs, a trade should be considered submitted to clearing after a CCP has received a trade that is accepted for clearing by both clearing members (and the CCP has communicated the information to the relevant clearing members, as per its rulebook). <ESMA_QUESTION_605> Q606: In particular, who are currently responsible, in the ETD and OTC context, for obtaining the information required for clearing and for submitting the transaction to a CCP for clearing? Do you consider that anything should be changed in this respect? What are the current timeframes, in the ETD and OTC context, between the conclusion of the contract and the exchange of information required for clearing on one hand and on the other hand between the exchange of information and the submission of the transaction to the CPP? <ESMA_QUESTION_606> LSEG: In both ETD and OTC, the trade source (trading venue or affirmation provider, depending upon execution model) is responsible for the submission to the CCP. For products executed fully electronically on trading venues, whether as ETDs or OTCDs, a trading venue generally defines the specifications of the contract and shares it with the CCP. The CCP then defines the clearing and risk model for that contract. A trading venue routes a transaction to the CCP through STP, and these requirements are clearly defined in the rules of the trading venue and that of the CCP. This process works well and does not require changes. Different models exist for products that are not executed fully electronically, especially for OTCDs. For products that are not executed fully electronically, the responsibility for submission normally belongs to the counterparties, and is generally facilitated via an affirmation platform. We believe that market participants should be able to continue to use affirmation providers to facilitate the submission of transactions that are not executed fully electronically. For products that are not executed fully electronically, the CCP is responsible for facilitating the Clearing Member approval or rejection of a client trade. We believe that it is appropriate to continue to allow the CCP to facilitate the Clearing Member approval/rejection for trades that are not executed fully electronically. <ESMA_QUESTION_606> Q607: What are your views on the balance of these risks against the benefits of STP for the derivatives market and on the manner to mitigate such risks at the different levels of the clearing chain? <ESMA_QUESTION_607> LSEG: Our focus is on the relationship between the CCP and its clearing members. A CCP does not generally regulate the relationships between a clearing member and its clients. In any case, LSEG does not think that STP can alter the risk position of a CCP or of a client. The position risk of a CCP depends on when the CCP assumes the responsibility (the novation or open offer model) and when collateral is collected. In many cases, but not mandatorily, a CCP will require prefunded resources in advance to the start of activity (for example, a minimum contribution to the default fund). 164

165 Accordingly, LSEG s view is that it is not practical for clients and/or clearing members to prefund trading activity in all circumstances. As such, CCPs should evaluate the risk of transactions prior to clearing a trade and ensure that members are in good standing in advance of clearing a trade. Our view is that ETDs and OTC products should be treated on the same playing field as it relates to any prefunding obligations. <ESMA_QUESTION_607> Q608: When does the CM assume the responsibility of the transactions? At the time when the CCP accepts the transaction or at a different moment in time? <ESMA_QUESTION_608> LSEG: For ETDs and OTCDs, the CM assumes responsibility for the transaction the moment the CCP accepts the transaction. This is in line CPSS-IOSCO s Principles for Financial Market Infrastructures, which prescribes 2 models for clearing: (i) open offer, where the CCP takes the responsibility at the moment of trading, and (ii) novation where CCP takes responsibility when it accepts the trade. According to the models, the Clearing Member assumes responsibility against the CCP simultaneously. <ESMA_QUESTION_608> Q609: What are your views on how practicable it would be for CM to validate the transaction before their submission to the CCP? What would the CM require for this purpose and the timeframe required? How would this validation process fit with STP? <ESMA_QUESTION_609> LSEG: It is common for ETDs and for OTCDs executed on swap execution facilities (SEFs) for the CM to validate an order prior to the order becoming a transaction and being submitted for clearing. It is important to note that for OTCDs this infrastructure is not completely developed and there are a number of cases, especially with trades that are not executed fully electronically, where pre-execution validation by the CM is not yet appropriate. <ESMA_QUESTION_609> Q610: What are your views on the manner to determine the timeframe for (1) the exchange of information required for clearing, (2) the submission of a transaction to the CCP, and the constraints and requirements to consider for parties involved in both the ETD and OTC contexts? <ESMA_QUESTION_610> LSEG: From the perspective of a CCP, both ETD and OTC markets now employ a high degree of STP and fully automated registration processes. In our view, the current practices of submission from the execution venue to the CCP works adequately for the market. When determining the EEA regime, we suggest that ESMA: (a) recognises the benefits of affirmation for products executed not fully electronically and the additional time requirements necessary to conduct affirmation on these types of trades; and (b) do not require CCPs to guarantee clearing of orders priors to execution (i.e. pre-execution clearing certainty). <ESMA_QUESTION_610> Q611: What are your views on the systems, procedures, arrangements and timeframe for (1) the submission of a transaction to the CCP and (2) the acceptance or rejection of a transaction by the CCP in view of the operational process required for a strong product validation in the context of ETD and OTC? How should it compare with the current process and timeframe? Does the current practice envisage a product validation? <ESMA_QUESTION_611> LSEG: We suggest that a CCP should accept (or reject) a trade submitted to clearing within 10 seconds of receipt. This is the same time as the CFTC requirements in the US. 165

166 For ETDs, in a STP process, if the transaction is concluded according to the rules of the trading venue, it is pre-validated and there is no scope for a CCP to reject the transactions on product eligibility grounds. The compliance of the trades with contract specifications is directly made by the trading venue preexecution. The STP process is usually in real time. Due to the bespoke nature of some OTCDs, it may not be prudent for the CCP to outsource product validation to an execution venue. However, where product validation is conducted upon CCP submission, 10 seconds should generally be sufficient for the CCP to register the trade. <ESMA_QUESTION_611> Q612: What should be the degree of flexibility for CM, its timeframe, and the characteristics of the systems, procedures and arrangements required to supporting that flexibility? How should it compare to the current practices and timeframe? <ESMA_QUESTION_612> TYPE YOUR TEXT HERE <ESMA_QUESTION_612> Q613: What are your views on the treatment of rejected transactions for transactions subject to the clearing requirement and those cleared on a voluntary basis? Do you agree that the framework should be set in advance? <ESMA_QUESTION_613> LSEG agrees that the framework for treatment of transactions rejected for clearing should be set in advance. Transactions can be rejected for clearing for limit breaches or operational reasons. Rejections by a clearing house as a result of a limit breach is uncommon, and limited to ETD and OTCD clearing models that operate novation models, as opposed to open offer models. Rejections due to operational reasons, such as a product that is not eligible for clearing (product eligibility) or where a participant s account is not correctly set-up for clearing (account eligibility), are also uncommon, but are issues primarily for OTCD products. For ETDs, the treatment of trades that are rejected for clearing (assuming the exchange and CCP do not operate an open offer clearing model) is determined under the rules of a trading venue. This model appears to work well. Under the open offer model, a CCP takes the responsibility at the moment of trading and a clearing member usually performs limit checks pre-execution, so if the trade is executed under the rules of a trading venue, it is highly unlikely that a CCP will reject the trade. For OTCDs executed on Swap Execution Facilities (SEFs) or Derivative Contract Markets (DCMs) operating under the jurisdiction of the CFTC, trades that are rejected for clearing are required to be voided ab initio by the execution venue. For OTCDs executed off SEF/DCM, the handling of trades rejected for clearing is managed bilaterally between the two executing counterparties. We do not believe that a void ab initio provision for OTCDs is appropriate in all circumstances. Our view is that trading venues should not be obliged to void (ab-initio) the contracts that are rejected for clearing. Instead, these frameworks should allow flexibility depending upon the execution model and allow for trading venues, market participants, and CCPs, to partner to specify how rejections will be handled for any given market. <ESMA_QUESTION_613> 9.2. Indirect Clearing Arrangements 166

167 Q614: Is there any reason for ESMA to adopt a different approach (1) from the one under EMIR, (2) for OTC and ETD? If so, please explain your reasons. <ESMA_QUESTION_614> LSEG: We agree with ESMA s regulatory objective to extend to MiFIR the indirect clearing arrangements established in EMIR to accommodate and protect a set of clients for whom direct access to a clearing member may be neither possible nor attractive. In our view, for clients who are active in both ETD and OTCD markets, harmonisation of approach in this regard would be welcome. We are, however, keen that the indirect clearing regime applies with the appropriate proportionality, particularly for indirect clients of clearing members that trade only a small range of ETD products and do not clear as frequently or in significant volumes. An unintended consequence of applying a one-size-fits-all regime may be that it would deny access to markets for smaller firms. Since it would be contrary to the purpose of EMIR and MiFIR for this outcome to occur, we would urge for ESMA to be proportional in its approach in terms of requirements for different clients, whilst ensuring that all indirect clearing arrangements appropriately protect the interests of the client. <ESMA_QUESTION_614> Q615: In your view, how should it compare with current practice? <ESMA_QUESTION_615> TYPE YOUR TEXT HERE <ESMA_QUESTION_615> 167