LSEG Response to CESR MiFID Consultation Paper EQUITY MARKETS

MiFID REVIEW LSEG Response to CESR MiFID Consultation Paper 10-394 EQUITY MARKETS Kathleen Traynor Head of Regulatory Strategy London Stock Exchange Group 0044 (0) 20 7797 3222 ktraynor@londonstockexchange.com 28 May 2010 Submitted online at www.cesr.eu INTRODUCTION The London Stock Exchange Group (LSEG) welcomes the opportunity to respond to CESR's consultation on Equity Markets. Detailed responses to the questions posed can be found below. This submission represents the views and experience of London Stock Exchange plc, Borsa Italiana and other market operators and investment firms within the LSEG. In order for the MiFID Review to deliver the outcomes that genuinely maintain and improve market efficiency and, ultimately, deliver the desired market model, we are strongly of the view that the Commission and CESR should adopt an evidence-based approach to the development of proposals. This would require regulators to measure the adverse market effects of a particular activity, identify the cause and propose appropriate corrective measures, describing the intended outcome. Adoption of such an approach would produce comprehensive and clear proposals which would ensure buy-in from the broadest possible range of market participants. CONTEXT LSEG accounts for a significant proportion of EU cash equities trading. Trading is conducted through LSEG s Regulated Markets in the UK and Italy as well as its MTF, Turquoise Trading Limited. LSEG has already provided a high level response to the European Commission 1 with regard to those issues which need to be addressed in order 1 Please see LSEG Market Position Paper dated 27 May 2010, which can be accessed at http://www.londonstockexchange.com/about-the-exchange/regulatory/policy.htm. London Stock Exchange Group plc Registered in England & Wales No 5369106. Registered office 10 Paternoster Square, London EC4M 7LS

to preserve transparency and market quality for the benefit of investors a core aim of MiFID. In particular: Equal treatment of, and consistency of regulatory approach between, all platforms carrying out the same activities. We believe that this is essential to preserve the flexibility and competitiveness of lit order books relative to other execution mechanisms; Improved quality of post trade data, both in terms of consistency and granularity. This is essential in a fragmented marketplace to enable investors to achieve a comprehensive view of market activity, enabling them to make informed investment decisions. This paper provides more detailed responses to CESR s questions, taking account of these core values and objectives CESR SECTION 2.1 PRE-TRADE TRANSPARENCY Question 1: Do you support the generic approach described above? In general, we support the objective of achieving pre-trade transparency for organised markets. However, this approach will only work effectively to maintain the competitiveness of lit order books if it provides for sufficient flexibility to respond to market developments and recognises the importance of choice for end investors. The primacy of lit order books is an important mechanism for the preservation of the quality of price formation, providing efficient and competitive markets with tight spreads. This can only be achieved by the availability of well regulated, high quality and consistent market data, which preserves the incentives to post displayed orders into such markets, by increasing execution certainty. Whilst participants want and need sufficient transparency to create market confidence, this should not undermine their ability to deliver an investment return to end customers or to achieve execution certainty for larger orders without adverse market impact. Therefore, allowing non-displayed trading to take place within the parameters of the appropriate waivers is essential to provide choice and flexibility for end investors, without undermining the execution certainty of displayed orders and at the same time preserving the competitiveness of public order books. We believe that: The waivers should be preserved and modified as set out below (see response to questions 2 to 9); A more precise description of the waivers (a Rules based approach) is necessary to ensure clarity - in a dynamic, highly competitive environment it is key that participants are able to obtain 28 May 2010 Page 2 of 36

clarity quickly and that the approach is applied consistently across all platforms conducting the same activities. Our experience to date is that the current arrangements often result in diverging interpretations, irrespective of the work carried out by CESR to offer guidance and enhance supervisory coordination, and that the process of obtaining clarification is slow, which can impede commercial objectives; The process must retain sufficient flexibility to enable policy makers and regulators to adapt quickly to market developments, so as to preserve the ongoing competitiveness of lit order books. As the experience of the last few years has proved, markets can develop very quickly and important issues can arise which were not even on the agenda a few months previously. This links with Q46 below. Question 2: Do you have any other general comments on the MiFID pretrade transparency regime? See response to Question 1. Large In Scale Waiver Question 3: Do you consider that the current calibration for large in scale orders is appropriate (Option 1)? Please provide reasoning for your view. No, we do not consider that the current calibration is appropriate. Recalibration now and on an ongoing basis to reflect future reductions in trade sizes is essential in order to preserve the attractiveness of lit order book functionality relative to alternative options for executing business. The Large in Scale (LIS) waiver is of limited value due to falling average trade sizes in lit order books. The LIS thresholds are typically many times larger than the average trade size in order books 2, and because market participants receiving market data can determine the presence of a LIS hidden order from the first execution against it, the risk of price-impact upon discovery of a LIS order is too high. This undermines the competitiveness of lit order books compared with other trading mechanisms (broker crossing networks, price referencing non-display order books), and potentially undermines competition amongst lit order book operators. Brokers receiving orders from their customers where a particularly passive (non-displayed) execution strategy is warranted are prevented by the 2 Based on an analysis of MiFID stocks on the London Stock Exchange (LSE data), the LIS thresholds are on average 44 times larger than the average trade size in order books. For the largest, most liquid stocks the range is from 18-80 times. For less liquid stock the range is even larger. 28 May 2010 Page 3 of 36

high LIS threshold from representing the order in multiple lit order books (weakening competition amongst them), and must instead choose between a single lit order book or using multiple alternative facilities, where additional options for avoiding market impact are available. For the same reason, we agree with the suggestion made in Pierre Fleuriot s February 2010 report on MiFID with regard to stubs (and by CESR in Option 1 in paragraph 33), namely that it should be confirmed that the LIS thresholds also apply to residual orders in the event of partial execution and not just to the initial order. This will prevent different treatment of two orders of the same size when one is a residual order and the other is a new order entered in the system s order book. We consider this will help to keep lit order books competitive. Question 4: Do you consider that the current calibration for large in scale orders should be changed? If so, please provide a specific proposal in terms of reduction of minimum order sizes and articulate the rationale for your proposal? We consider that, at the very least, a recalibration of the minimum order size threshold for LIS eligibility by a reduction of 75% from the current level should be conducted on the basis suggested below. Given the rapid changes taking place in the marketplace 3, in order for the LIS waiver to remain relevant, we suggest that a review/recalibration exercise should be carried out every 3 months. Our rationale is set out below. In the consultation paper, CESR option 2 proposes that the current thresholds be reduced by 25%. We have assessed the extent to which the current LIS parameters would apply to the profile of orders (limit and iceberg orders only) entered in to SETS (essentially FTSE 100, 250 and SmallCap stocks) 4. The scenarios run were: 1. September 2006 we sought to establish the level of orders that would be LIS based on order profiles (and ADTs - approx) at the time MiFID was agreed (using current LIS parameters that have not changed). 2. April 2010 - LIS orders using current parameters and current ADTs. 3. CESR Option 2 - reduction of minimum order sizes for each liquidity band by 25% - data for April 2010 used. 4. CESR Option 2 flexed - reduction of minimum order sizes for each liquidity band by 50% - data for April 2010 used. 5. CESR Option 2 flexed - reduction of minimum order sizes for each liquidity band by 75%. Again, data for April 2010 used. 3 On the LSE, the average trade size has fallen from 61,751 in 2000 to 8,398 in Q1 2010 (it has nearly halved since 2007, when it was 15,459). For Borsa Italiana, the average trade size was 14,842 in 2000 but this had reduced to 10,939 in Q1 2010 (again, it has nearly halved since 2007, when it was 21,718). 4 Our approach was to calculate the proportion of total orders that would have been treated as LIS under the thresholds applied in 2006 and to seek to achieve the same level in 2010. 28 May 2010 Page 4 of 36

Conclusions The results of recreating the regime introduced in September 2006 suggest that for the core MiFID securities (those in continuous trading), 1.7% of orders would have been classified as LIS based on their size on order entry. Applying the current regime, this level has reduced to 0.2% in April 2010. Applying the calibration suggested by CESR in option 2 raises this level to only 0.4%. Extending this to a reduction in minimum order size of 50% of current calibration still does not bring the level of LIS orders up to the September 2006 level. However, if a 75% reduction to the current regime is applied we find that this would mean 1.9% of (limit and iceberg) orders would be LIS, compared with 1.7% in 2006, and although there is a disproportionate effect on smaller securities, we consider that this indicates broadly the level of reduction required if CESR is to maintain the level of trades that would be classified as LIS at around 2%. Index Sep-06 Apr-10 CESR Option 2 (25%) Reduction (-50%) Reduction (-75%) FTSE 100 2.2% 0.2% 0.3% 0.6% 1.6% FTSE 250 0.6% 0.2% 0.3% 0.6% 1.8% FTSE SmallCap 2.0% 1.9% 4.9% 8.8% 17.2% Total 1.7% 0.2% 0.4% 0.7% 1.9% At a 75% reduction the minimum order sizes would now be: ADT Band Min Value Max Value Min Order Size 1 0 500,000 12,500 2 500,000 1,000,000 25,000 3 1,000,000 25,000,000 62,500 4 25,000,000 50,000,000 100,000 5 50,000,000 999,999,999 125,000 Question 5: Which scope of the large in scale waiver do you believe is more appropriate considering the overall rationale for its application (i.e. Option 1 or 2)? Please provide reasoning for your views. For the reasons set out in Question 4 above, we do not believe either option 1 or 2 is appropriate. We believe the thresholds should be reduced by 75% and a review/recalibration exercise should be carried out every 3 months. Please also see our response to Question 3 (on stubs). 28 May 2010 Page 5 of 36

Price Reference Waiver Question 6: Should the waiver be amended to include minimum thresholds for orders submitted to reference price systems? Please provide your rationale and, if appropriate, suggestions for minimum order thresholds. In our view, a more important revision than introducing a minimum threshold for orders submitted to a reference price system would be to ensure consistency of approach between different types of venue in particular: All venues should have the flexibility to execute at any point within the spread (delivering price improvement benefits to investors); and MiFID should no longer require that venues have another (i.e. separate) system in place to take advantage of this waiver. We believe the current inconsistency limits the potential to deliver price improvement to investors and further fragments liquidity, and that enabling lit systems to operate in the same way as other venues would encourage greater transparency given that lit systems prioritise displayed orders across the board. We have set out our position on this issue in answer to Question 7 below. A minimum size threshold could be considered, to avoid retail flow in particular being included, but unless it applied to all execution venues, (ie. Regulated Markets, MTFs and broker crossing networks) and not only to RMs and MTFs, such a move would undermine the competitiveness of RMs and MTFs relative to others and be detrimental to efficient lit order books which provide open access to all participants. Question 7: Do you have other specific comments on the reference price waiver, or the clarifications suggested in Annex I? Price improvement options CESR has adopted a narrow interpretation of this waiver for Regulated Markets and MTFs, and has defined reference price to mean execution only at mid-point, best bid or best offer, with an explicit prohibition against offering customers price improvement. Conversely, broker crossing networks (BCNs) 5 are able to offer execution at any point within the spread and thus give small amounts of price improvement. Whilst there are good reasons for maintaining different regulatory regimes for Regulated Markets, MTFs and brokers respectively, we believe that such material differences in approach across 5 We define BCNs an internal system operated by a broker-dealer that allows the crossing of the business of one client of the firm (whether internal or external) with that of another, without the need to show the order to the market as a whole. 28 May 2010 Page 6 of 36

these venues undermine market efficiency and competition, to the detriment of investors. In our view, Regulated Markets, MTFs and BCNs should all be able to undertake the full range of price referencing, matching anywhere within the relevant spread 6. The Commission should take account of the unwritten principles of Regulated Markets and displayed MTFs: that orders are generally prioritised according to price, display type, time (in that order). Allowing Regulated Markets and MTFs to match anywhere within the spread in the same way as BCNs would be the pro-transparency approach as it would allow lit orders to be prioritised ahead of dark orders at a given price-point (rewarding transparency), whilst at the same time encouraging use of lit, liquid market places to execute business, preserving the competitiveness of these venues at the same time as delivering choice for investors. We suggest the following amendment to MiFID: Commission Regulation (EC) No 1287/2006 Article 18 Text in MiFID 1. Waivers in accordance with Article 29(2) and 44(2) of Directive 2004/39/EC may be granted by the competent authorities for systems operated by an MTF or a regulated market, if those systems satisfy one of the following criteria: (a) they must be based on a trading methodology by which the price is determined in accordance with a reference price generated by another system, where that reference price is widely published and is regarded generally by market participants as a reliable reference price; Amendment 1. Waivers in accordance with Article 29(2) and 44(2) of Directive 2004/39/EC may be granted by the competent authorities for systems operated by an MTF or a regulated market, if those systems satisfy one of the following criteria: (a) they must be based on a trading methodology by which the price is determined in accordance with a reference price generated by a designated third party, Regulated Market or MTF, and any trade may be at any price within the best bid or offer of that designated Regulated Market or MTF; (b) the system itself publishes its best bid and offer price and trades are based on a trading methodology by which the price is determined in accordance with the reference price generated by the system itself. Trades may take 6 The Commission may want to consider some minimum level of price improvement (e.g. a tick or a portion thereof). 28 May 2010 Page 7 of 36

place at any price within the best bid or offer of that system. Another System The price reference waiver is currently the only waiver that forces venues to have another (i.e. separate) system in place 7. This approach artificially fragments liquidity between venues controlled by the same operator, and restricts the flexibility of lit order books, relative to other execution venues, undermining opportunities for price improvement. It is hard to see any justification for this approach. In our view: firstly, it is not clear what harm would flow from the amalgamation of these functions into a single book. If the concern is that this would encourage greater use of dark execution methods, disadvantaging lit venues, where lit liquidity takes priority is only likely to exacerbate this concern; secondly, the artificial fragmentation of liquidity undermines price formation and market efficiency to the detriment of end investors. Further, the potential for price improvement as against lit orders would benefit investors 8. thirdly, by disadvantaging lit venues, the approach undermines the competitiveness of lit book offerings (which are unable to offer these services as efficiently) rather than encouraging flow to them. We therefore propose that the PRW is revised to allow lit systems to use it, but on the basis that they prioritise displayed orders across the board. We believe that such changes would allow Regulated Markets and MTFs to continue to operate in a way that remains attractive and competitive. It has been suggested (in Pierre Fleuriot s report for the AMF) that consideration could be given to suspending systems that use the PRW when the transaction volume under this waiver exceeds a threshold set by the regulator at the European level, in order to protect the quality of the price formation mechanism. This proposal would suggest that there is some level at which price referencing becomes a problem we do not believe that any such steps should be taken unless there is a clear, evidence-based rationale for doing so. Widely Published 7 See CESR/10-394 Annex I, paragraph 5. 8 Please see Appendix I for details of the trading activity in hidden orders on our trading system, called SETS. Whilst this is taking place under the LIS waiver, the figures show there is demand for interaction with dark liquidity within the framework of public order books. It also shows that the average price improvement relative to the best displayed price at the time of execution was 8.1 basis points with the greatest improvement on a single trade being equivalent to 320 basis points. 28 May 2010 Page 8 of 36

In our view, CESR has adopted an unnecessarily narrow interpretation of what is meant by widely published, which we consider is not helpful or required. If users of a price-referencing platform are aware of, and accept that execution will be based on, the basis of calculation of a reference price, this should be sufficient. Negotiated Trade Waiver Question 8: Do you have any specific comments on the waiver for negotiated trades? The negotiated trade waiver has proved useful to the market and should be retained. In particular negotiated trades are being relied on as a means of reducing the systematic risk of OTC trades, by passing these via a public order book to a central counterparty for clearing. In 2009, 10.3% of LSE trades were reported under this waiver (see Appendix I). It is also worth noting that under the current London Stock Exchange Rules, a negotiated trade must be on terms that are no worse than those that could be achieved on our order or quote book, after taking into account any relevant trading, settlement and clearing costs. It therefore ensures that a trade is on terms that are at least as good as those which would have been achieved on a lit system. Order Management Waiver Question 9: Do you have any specific comments on the waiver for order management facilities, or the clarifications provided in Annex I? The order management system waiver was originally designed to cater for iceberg orders on Regulated Markets and MTFs, but to do so in a generic way that allowed for further development of such systems within public markets. Iceberg orders have improved liquidity to the benefit of those using them and those trading against them. However, the restriction on further developments has prevented Regulated Markets and MTFs from providing further enhancements to the benefit of potential users of these orders and to those who might trade against them. We urge CESR to allow greater flexibility with respect to the order management system waiver as set out below, so that Regulated Markets and MTFs are able to deliver better executions and more liquidity to customers, to the benefit of end investors. We do not think it is feasible to address this by levelling up, as CESR has suggested, as this would not only deprive public markets of this liquidity, but also prevent the brokers from offering this valuable service to customers. History of the iceberg order 28 May 2010 Page 9 of 36

Prior to the implementation of iceberg orders within public markets, brokers who received a large order from a customer developed automated order management systems to split these orders up into smaller chunks and place them into the market. For example, a customer who had an order to buy 100,000 shares at 99p might have that order split into ten tranches on 10,000 share orders that the broker s system would feed into the public market for execution, refreshing the 10,000 shares each time the order was filled. This allowed the customer to trade a large block at 99p without ever displaying the full size of the order. However, there was a disadvantage that the order could be traded through. Take the scenario where the market price is 99 100 and the customer has a 10,000 share buy order at 99p, if a seller comes in to sell 20,000 shares aggressively, he might trade down to 98p, thereby depriving the buyer of the opportunity to buy at 99p and the seller of getting a better price. The iceberg order was an innovation that overcame this by enabling the broker to hold the full 100,000 customer order within the public order book whilst only displaying the peak of 10,000 shares. In this scenario, if a 20,000 share aggressive sell order is submitted, the full 20,000 shares executes at 99p, benefitting the buyer by getting more of his business done, the seller by achieving a better price and the market as a whole by having more liquidity. Current practice The old tranching model that customers used in the past has evolved over the years into a more sophisticated order management system in response to customer requirements and the increasing ability of other market participants to discover the presence of simple iceberg orders and trade to move the price away from them. Hence, brokers will now offer the customer the ability to place the order of 10,000 share tranches (often now with a randomised size element) at the best bid, moving up or down as the best bid moves. With a moving priced order such as this, the customer will also want to place a maximum price that he will not pay more than. He may also wish to aggressively buy stock if it is offered at or below a certain price and in greater than a specified quantity. As an illustration, in our example above with the market price at 99 100, the customer may submit the 100,000 share order to buy at the bid but not to pay more than 102p, and to aggressively buy if the offer price ever drops to 99.5p or below. This sort of enhanced iceberg sophistication (and more) is readily available from broker systems, but it is not clear that the current order management system waiver allows it. This disadvantages not only public execution venues, but also the end customer; in the same case as with the original simple iceberg order above, the inability to place the whole order in the public market deprives: 1) the customer of the opportunity of getting his business done if a large sell order is placed, 2) the seller of the opportunity to sell stock at a better price and 3) the public market of liquidity. 28 May 2010 Page 10 of 36

In order to maximise flexibility and execution opportunities for customers, whilst at the same time ensuring the attractiveness of public order books, the order management waiver must therefore be clarified to allow for enhancements. Systematic Internalisers Question 10: Do you consider the SI definition could be made clearer by: i) removing the reference to non-discretionary rules and procedures in Article 21(1)(a) of the MiFID Implementing Regulation Without further explanation of the benefit CESR seeks to achieve by adjusting the SI regime, it is not possible properly to assess all of CESR s proposals in relation to SIs. Although registration as an SI may bring greater transparency through reporting requirements, which is in the interests of investors and overall market quality and acts to further level the playing field between different trading venues, it is not clear why CESR has suggested the adjustments to the definition. There is no evidence that the internalisation of trades currently performed by market participants who are not registered as SIs actually has any adverse market impact, particularly given CESR s own acknowledgment that the current number of SIs may in fact be the correct number. We would encourage CESR to provide more detail on its regulatory objectives in relation to systematic internalisation of trades, in order to provide market participants and infrastructure providers with the clarity they need to assess CESR s SI proposals effectively. We note the concerns of other market operators, both in Europe and the US, that too much internalisation degrades the quality of liquidity reaching public order books, ultimately weakening incentives to post visible bids and offers and consequently leading to reduced liquidity and less effective price formation. They argue that brokers only route unwanted or exhaust liquidity to public markets, which increases the likelihood of public order book participants suffering from adverse selection. They further argue that SIs, by making execution at the BBO available without accessing the public order book, undermine incentives to post displayed limit orders, as the certainty of execution for such limit orders is reduced when potential contra-orders are internalised by an SI, rather than interacting with the order setting the public BBO. We accept that there is some validity in these concerns. However, we also believe that there is a legitimate role for brokers to provide liquidity to their customers away from public order books. Many clients, including asset managers and retail brokers, are able to trade effectively and at lower cost that might be the case if they were forced to use public order books. LSEG believes that choice and competition are essential to ensuring our markets remain vibrant and efficient. 28 May 2010 Page 11 of 36

Our proposals above for dealing with the MiFID waivers are intended to address these without undermining choice and flexibility for investors. We have set out below some suggestions for enhancing choice and competition on which we believe it would be appropriate for CESR to consult further with a diverse group of market participants: Should clients relying on SIs receive regular assurances that the quality of execution being achieved matches that available on public order books; Should SIs publish information about the proportion of orders internalised against the operator or an affiliate of the operator, the proportion matched against other client orders, and the proportion routed to public order books; and Should SIs be obliged to route to public order books when matching a client order at the BBO price, as a means of increasing the certainty of execution for publicly posted limit orders, and should price improvement offered by SIs be subject to a de minimus threshold. ii) providing quantitative thresholds of significance of the business for the market to determine what constitutes a material commercial role for the firm under Article 21(1)(a) of the MiFID Implementing Regulation. As we consider that CESR has not clearly articulated the purpose behind the systematic internaliser regime, it is difficult to comment on any method for establishing the level at which any quantitative thresholds might be set, or indeed whether they would serve an effective purpose. CESR s proposal of thresholds suggests they consider there to be a point at which greater transparency will have a benefit to the wider market which outweighs the benefits customers receive from flexibility and choice. If thresholds are intended to provide an indication of when systematic internalisation of orders by a firm reaches such a point, there needs to be clear evidence of the impact that will be mitigated through the transparency requirements to which a firm would become subject as an SI. Such an explanation would require a clear understanding of the precise market impact the SI regime is designed to manage and which firms or types of business it is intended to capture. Question 11: Do you agree with the proposal that SIs should be required to maintain quotes in a size that better reflects the size of business they are prepared to undertake? Again, it is difficult to assess this proposal without a clear rationale for the SI regime. In theory, minimum quote sizes for SIs will aid transparency by enabling market participants to better compare the prices available on different systematic internalisers and understand how the prices available on SIs relate to those on other trading venues. We are aware that some SIs who meet the current quoting obligations feel the non-addressable SI quotes do 28 May 2010 Page 12 of 36

not serve a useful purpose within the market, given the nature of the business they carry out. Question 12: Do you agree with the proposed minimum quote size? If you have a different suggestion, please set out your reasoning. A minimum quote size would bring a degree of harmonisation across systematic internalisers but, consistent with our responses to Q10 & 11 above, it is not clear what CESR seeks to achieve by implementing minimum quote sizes. Minimum quote sizes and two-sided quoting are longestablished requirements when acting as a market maker on the London Stock Exchange, with the express purpose of providing reliable prices and allowing investors to compare the prices offered by different market makers in each security. However, such requirements may not be suitable for SIs given they do not operate in the same manner as market makers and because their quotes do not necessarily serve the same purpose. Question 13: Do you consider that removing the SI price improvement restrictions for orders up to retail size would be beneficial/not beneficial? Please provide reasons for your views. This question is particularly difficult to respond to without more detailed explanation of the specific market impact CESR wishes to address through regulation of systematic internalisers, and the aspect of their activity which have these effects. It is a standard feature of acting as a market maker on the London Stock Exchange that a market maker provides firm quotes. As with minimum quote sizes, this provides a degree of certainty for investors when making trading decisions and allows comparison between different prices. However, given that systematic internalisers operate on a different basis to market makers, we cannot comment on whether a similar effect results from the current SI price improvement restrictions or is in fact intended at all. It is thus equally difficult to determine the potential effect of removing the price improvement restrictions. However, in general we believe that SIs deliver greater value to their customers when price improvement vs. the public order book BBO is provided, and that where price improvement is provided (in excess of some de minimus threshold) that there is less concern that execution certainty is being undermined for publicly displayed limit orders. As discussed above, there may be an argument for requiring SIs either to deliver price improvement or to interact with the public limit order books. These potential benefits of such an arrangement would need to be assessed against the higher costs that SIs would incur in servicing their customers. Question 14: Do you agree with the proposal to require SIs to identify themselves where they publish post-trade information? Should they only identify themselves when dealing in shares for which they are 28 May 2010 Page 13 of 36

acting as SIs up to standard market size (where they are subject to quoting obligations) or should all trades of SIs be identified? We suggest that the proposal to rescind the provision exempting systematic internalisers from identifying themselves in post-trade reports may give rise to difficulties. Whilst transparency is clearly important, a post-trade tape which does not identify specific counterparties is a core feature of current effective market functioning; it allows firms to take on orders and risks in a principal capacity with the certainty their positions, strategies and clients will not be exposed to the market. SIs should be subject to post-trade transparency obligations, such that all executions conducted on their automated systems are reported immediately, but we suggest that it is a matter for the CESR/Industry Working Group to determine whether, for instance, the use of the SI marker (on all SI trades) is sufficient to indicate to the market that such trades were executed on systematic internalisers. Please also see our responses to Section 2.2 on post-trade transparency. Question 15: Have you experienced difficulties with the application of Standard Market Size as defined in Table 3 of Annex II of the MiFID Implementing Regulation? If yes, please specify. The LSEG does not currently make use of Standard Market Size in our trading system and we have no plans to do so. Question 16: Do you have any comments on other aspects of the SI regime? No further comments. CESR SECTION 2.2 POST-TRADE TRANSPARENCY Question 17: Do you agree with this multi-pronged approach? Yes, we agree with CESR s multi-pronged approach to improving the quality of post trade information, as set out in para 69 of the CP. In particular, we support the proposal to establish a joint CESR/Industry Working Group to finalise the development of standards by July 2010. For the full benefits of broad-based transparency to be realised, trade information published through different sources must be available in a format that allows for comparison between the different trading venues; this requires a much greater degree of harmonisation of standards than currently applies. The task facing this Working Group should not be underestimated. Harmonised standards should cover at least the following: Transaction types; Instrument identifier; 28 May 2010 Page 14 of 36

Price notation; Transaction Venue identifier, including the OTC venue; Conditions (e.g.: negotiated trades, risk trades, delayed trades). Taking transaction types as an example, there is currently a lack of harmonisation across countries and venues. For instance, volume weighted average price (VWAP) transactions, portfolio transactions, ex/cum trades, give-up trades, OTC hedges, and transfers are all transaction types (or categories) that venues choose to publish alongside the other details of the individual trade in order to indicate this detail, but there is no requirement to identify these transactions in a standard way. As a result, there is no consistency in the way these transactions are published. This is an example of why post-trade data has been described as lacking in quality and reliability. We also agree with CESR that a common approach is required as to what is reportable to avoid multiple reporting of trades, and to avoid both market and client side legs of the same trade being reported and/or published. Greater clarity is also required in terms of which of the investment firms involved has the obligation to publish an OTC transaction. In our view, these standards should be implemented in a market-centred way through Level 3 guidelines to ensure consistent adoption and application by all investment firms, venues, data disseminators and information vendors. We further consider that the adoption of such standards would allow the effective and economic consolidation of post-trade data by the multiple approve publication arrangements proposed by CESR and would facilitate commercial development of one or more effective consolidated post-trade feeds without the need for regulatory intervention or building of any separate Mandated Consolidated Tape arrangements. Timing of post trade publication Question 18: Do you agree with CESR s proposals outlined above to address concerns about real-time publication of post-trade transparency information? If not, please specify your reasons and include examples of situations where you may face difficulties fulfilling this proposed requirement. We agree with CESR that the timeliness of post-trade transparency information could be improved. However, introducing a new reduced time limit may not necessarily be the answer. MiFID is clear in saying that trades must be reported as soon as possible. The current 3 minute deadline is not the period within which most trades must be reported, it is the exceptional limit for trades where there are system or other issues preventing immediate reporting. It follows that, in our view, the fact that the 3-minute reporting deadline is being used much more frequently than intended (not in exceptional 28 May 2010 Page 15 of 36

circumstances only), may reflect poor monitoring, supervision and enforcement by relevant supervisors. In that event, the key remedy is for the immediate reporting requirement to be enforced across the EU. We also suggest that information relating to portfolio trades would need to be exempted from any 1-minute reporting deadline requirement (i.e., essentially that Article 29(3) of MiFID should not change). In the London Stock Exchange Rules for members, we say that in relation to a portfolio trade, due to the need to allocate prices to particular securities, the Exchange recognises that the process to allocate prices to each share of the portfolio trade may not be instantaneous, which is aligned with Article 29(3) of MiFID. This recognition would have to be maintained if the time limit for reports is reduced. Question 19: In your view, would a 1-minute deadline lead to additional costs (e.g. in terms of systems and restructuring of processes within firms)? If so, please provide quantitative estimates of one-off and ongoing costs. What would be the impact on smaller firms? We anticipate that moving to more immediate reporting will result in an increase in our market operator functions having to undertake more trade report corrections, as our member firms back-office/end-of-day processing functions will probably find more errors, due to the shorter reporting deadline. This likely will result in an increase in the number of late correction reports we receive, so we will have some increased processing costs. We will also have some systems changes to implement as a result of any change, but we would not expect these to result in significant costs on the basis of the current proposal. It will be for our member firms to comment on any increased costs or other burden they will face as a result of any such changes. Question 20: Do you support CESR proposal to maintain the existing deferred publication framework whereby delays for large trades are set out on the basis of the liquidity of the share and the size of the transaction? Yes, we believe that an approach that uses liquidity and transaction size is a logical and straightforward basis for setting the thresholds for publication of post-trade data. The potential market impact of a transaction is central to the need for a deferred publication framework because the ability for market participants to trade a large order without the market price moving significantly against them is a vital aspect of any vibrant and liquid marketplace. Basing the deferred publication framework on the potential market impact of a trade in the relevant security, according to liquidity and transaction size, is sensible given the purpose of the framework. Question 21: Do you agree with the proposal to shorten delays for publication of trades that are large in scale? If not, please clarify 28 May 2010 Page 16 of 36

whether you support certain proposed changes but not others, and explain why. We consider the deferred publication framework necessary to ensure firms have adequate time to unwind certain types of position, but we also acknowledge the concerns noted by CESR regarding firms using the full delay when not actually required for unwinding their positions. It may be appropriate to consider a reduction in some of the permitted delays but we believe the CESR proposal does not provide adequate justification for the removal of all publication delays extending into the next trading day. LSEG analysis shows the trade reports currently eligible under the current MiFID framework for delays of 1-3 days beyond the end of the trading day constitute approximately 1% of total trade reports, but this represents approximately 10-12% of value traded. 9 This illustrates that the trades captured by these delays are already very low in number but, given their size, each trade represents a substantial risk position taken on by the relevant market participant. In a properly functioning market, it is essential that firms taking on such risk positions are permitted adequate time in which to unwind them. The CESR proposal would result in all the trades that currently receive a publication delay of 1-3 days only receiving a maximum delay of 120 minutes beyond the end of the trading day. We consider that a greater degree of granularity in the regime is necessary to provide adequately for firms taking on such sizeable trades. LSEG is of the view that limiting publication delays to 120 minutes beyond the end of the trading day, regardless of when the trade was agreed, will significantly disadvantage firms executing very large trades, particularly in illiquid stocks. It may well not be possible for firms to unwind such illiquid positions before the end of the day, particularly if the order is received in the last 1-2 hours of trading. As a result, investors are likely to receive inferior risk prices because firms will know their reports will be published sooner than would be desirable, thereby reducing the efficiency of the market and negatively affecting trading behaviour and investors. Furthermore, capping publication delays at the end of the trading day is likely to encourage firms to split large trades or hold back final agreement of the trade (price & size) with their client/customer until after the end of the trading day, thereby ensuring publication the following morning. Introducing an increased incentive for firms to breach trade reporting requirements should be avoided by CESR. Accordingly, we suggest a realistic compromise is necessary in order that the concerns raised by CESR are balanced with the potential for adverse market impact if market participants are not given enough time to unwind large trades before they are made public. 9 Based on all trades in securities admitted to trading on a regulated market which are traded off order book but under the rules of the London Stock Exchange. Please also see Appendix II for information on the proportion of trades eligible for each level of delay under the current deferred publication framework and CESR s proposed framework. 28 May 2010 Page 17 of 36

Recommendations 1) We recommend allowing the publication of a trade undertaken in the final two hours of the trading day to be delayed until noon the following day (as provided for in the current framework). 2) We also advocate permitting publication of particularly large trades and trades in illiquid shares (i.e. those in the ADT<EUR 100,000 band) to be delayed until the end of the next trading day. We consider that this will go some way to providing sufficient allowance for the additional difficulty of unwinding positions in such shares, and ensuring that firms with specific business need for delayed trade report publication are able to offer services and prices that meet the requirements of investors. Our proposals will lead to a general reduction in publication delay periods down to end of day, as proposed by CESR, but with exceptions for trades in the last two hours of the trading day and very large trades/ trades in illiquid securities to be delayed until the end of the second day. We also wish to highlight the need for frequent reviews of ADT levels in order to ensure they are accurate and fit for purpose; if the permitted publication delays are reduced there is an even greater need for ADT calculations to be as up to date as possible. As we argue elsewhere, we believe ADT should be reviewed on a quarterly basis, thereby providing a compromise between practicality and accuracy such that the levels are more suitable for meeting the needs of investors and market participants. Question 22: Should CESR consider other changes to the deferred publication thresholds so as to bring greater consistency between transaction thresholds across categories of shares? If so, what changes should be considered and for what reasons? Basing criteria on ADT is a means of making the thresholds specific to different categories of shares and reflects the direct link between market impact and the liquidity of the relevant share. It is this market impact that the deferred publication regime is designed to mitigate and, as such, a degree of differentiation between shares with different levels of liquidity is a sensible feature of the framework. There is wide variation in the daily volume traded across shares admitted to trading on EU Regulated Markets and, thus, wide variation in the market impact that a trade of a particular size or value will have. In light of this, we consider that using four liquidity bands is a sensible compromise providing granularity without excessive complexity within the framework. In our view, the removal of any distinction between less liquid and more liquid shares would be significantly detrimental to the functioning of the market and the interests of investors. Question 23: In your view, would i) a reduction of the deferred publication delays and ii) an increase in the intraday transaction size thresholds lead to additional costs (e.g. in ability to unwind large 28 May 2010 Page 18 of 36

positions and systems costs)? If so, please provide quantitative estimates of one-off and ongoing costs. LSEG has no comment to make regarding additional costs from changes to the deferred publication framework; investment firms will no doubt be able to provide input on the implications of these changes. Application of transparency obligations for equity-like instruments Question 24: Do you agree with the CESR proposal to apply transparency requirements to each of the following (as defined above): - DRs (whether or not the underlying financial instrument is an EEA share); - ETFs (whether or not the underlying is an EEA share); - ETFs where the underlying is a fixed income instrument; - ETCs; and - Certificates If you do not agree with this proposal for all or some of the instruments listed above, please articulate reasons. We agree with the CESR proposal to apply transparency requirements to equity-like instruments, including all of those identified in the question. We suggest that it be clarified that the regime also covers Exchange Traded Notes (ETNs). Extending MiFID post trade transparency to equity-like instruments would have significant benefits for investors by providing full transparency of trading activity in such instruments 10. Requiring greater transparency would likely attract more users to these instruments, further developing the market and enhancing liquidity. ETFs behave very much like equities and are beneficial for investors because they are a low cost, easy way for investors to track indices across a number of markets. Unlike equities, there is currently no requirement under MiFID to trade report executions in ETFs or ETNs, even where admitted to an EU Regulated Market. The regulation proposed at national level does not allow market participants to get a complete picture of OTC trades because it is addressed only to investment firms under the supervision of the national competent authorities. Trading venues can require trades in ETFs to be reported (for example, as under the Rules of the London Stock Exchange) but this applies only to exchange members. Question 25: If transparency requirements were applied, would it be appropriate to use the same MiFID equity transparency regime for each of the equity-like financial instruments (e.g. pre- and post-trade, timing 10 Currently, trading venues can require trades (for example, in ETFs) to be reported under their rules, but this applies only to members of that venue. The current situation does not allow market participants to get a complete picture of OTC trades. 28 May 2010 Page 19 of 36