RESPONSE TO DISCUSSION PAPER ON ACCOUNTING FOR DYNAMIC RISK MANAGEMENT: A PORTFOLIO REVALUATION APPROACH TO MACRO HEDGING

Transcription

1 A S C ACCOUNTING STANDARDS COUNCIL SINGAPORE 14 November 2014 Mr Hans Hoogervorst Chairman International Accounting Standards Board 1 st Floor 30 Cannon Street London EC4M 6XH United Kingdom (By online submission) Dear Hans RESPONSE TO DISCUSSION PAPER ON ACCOUNTING FOR DYNAMIC RISK MANAGEMENT: A PORTFOLIO REVALUATION APPROACH TO MACRO HEDGING The Singapore Accounting Standards Council appreciates the opportunity to comment on the Discussion Paper on Accounting for Dynamic Risk Management: A Portfolio Revaluation Approach to Macro Hedging (the DP) issued by the International Accounting Standards Board (the IASB) in April General We are supportive of the IASB s efforts to develop an accounting approach to better reflect open portfolio hedges in entities financial statements so as to enhance the usefulness of financial information. However, we are deeply concerned that the DP has gone beyond addressing the existing issues associated with the accounting for open portfolio hedges the lack of a comprehensive set of accounting solutions resulting in patchwork presentation and accounting mismatches in the financial statements by proposing an accounting approach that attempts to represent dynamic risk management (DRM). We believe that the accounting approach for open portfolio hedges should seek to faithfully represent the effects of dynamic risk mitigation using risk management instruments, rather than the effects of DRM per se, including decisions not to mitigate the managed risk. Accordingly, we consider the portfolio revaluation approach (PRA) scope focused on risk mitigation, which is aligned with the objective of general hedge accounting (GHA) in IFRS 9 Financial Instruments, to be more promising than a scope focused on DRM as the former has the potential to address existing accounting mismatches in hedged open portfolios and result in more relevant information, if the IASB could overcome the various operational issues. Moreover, the PRA scope focused on DRM would result in the revaluation of the entire managed portfolio (including unhedged portions) in respect of the managed risk, which undermines the conclusions reached in IFRS 9 on the measurement of non-derivative financial instruments, and results in profit or loss volatility that is not representative of business performance. Page 1 of 22

2 We further note that the PRA would not work well for cash flow hedges of open portfolios of floating rate instruments since cash flow variability does not create a corresponding fair value (FV) exposure. If the PRA were to be applied to all DRM activities, there would be a need for the IASB to explore alternative solutions for interest rate risk exposure to cash flow variability in open portfolios, such as developing solutions within the cash flow hedge accounting model in IFRS 9 to overcome its inherent restrictions. Specifically on the application of the PRA, we note that it would not require the externalisation of managed risk that is transferred via internal derivatives, which would conflict with the principle of not reflecting internal transactions in the financial statements a conceptually problematic but perhaps necessary trade-off to faithfully represent how risk exposures are dynamically managed in reality. However, we are concerned with not only the grossing up of both statements of financial position and comprehensive income, but also the incomplete depiction of the effects of risk mitigation if gross presentation of internal derivatives is prohibited. Non-externalisation aside, we recognise there are some merits in the accounting mechanics and underlying concepts, and the possibility of the resulting revaluation adjustment as a faithful representation of the effects of dynamic risk mitigation. In particular, we see merits in the concept of behaviouralisation of cash flows and the inclusion of pipeline transactions that are akin to constructive obligations for the purposes of applying the PRA, with appropriate disclosures regarding their estimations. Furthermore, we appreciate the inclusion of subbenchmark instruments using an approach that reflects both the embedded floor and inherent hedge ineffectiveness, but not the equity model book (EMB) primarily due to conceptual and verifiability concerns. That said, we consider that the application of the PRA, especially under a scope focused on risk mitigation, would be fraught with operational difficulties. For example, applying the PRA to one-sided risk mitigation, and inclusions or removals of exposures after initial recognition or before expiry, would necessitate tracking and/or amortisation, which would be operationally challenging in light of the dynamic nature of open portfolio hedges. We are also particularly concerned about whether transfer pricing (TP) transactions could be meaningfully included in the PRA to achieve a reasonable representation of the effects of risk mitigation, having regard to the different DRM strategies that exist in practice, the use of TP mechanisms as a policy signal to the business units to re-shape existing profiles, and the difficulty in developing robust eligibility criteria for the components of TP. Furthermore, these operational difficulties might not result in commensurate benefits, particularly for entities that do not consider the issues associated with existing GHA requirements to be a significant concern. If the IASB decides to further develop the PRA as an accounting approach for open portfolio hedges, we believe that the PRA should be optional, which is consistent with the current requirements under IFRS 9, in view of the difficulties with mandatory application and the concerns that application efforts might not result in commensurate benefits. We further think that the PRA and the GHA should represent a comprehensive set of solutions for risk management activities and that a free choice of approaches should be precluded. We consider that a well-developed description of dynamic risk mitigation has the potential to distinguish its application from that of GHA and to instil discipline in the application of both models. Page 2 of 22

3 In principle, the PRA should be available for DRM of different types of risks that are represented by well-established benchmark indexes and for which risk mitigation could be undertaken, including foreign exchange risk and commodity price risks. We would encourage the IASB to reach out to other industries to identify any other issues that are not faced in interest rate risk management and to consider whether and how these issues could be addressed by the PRA model. With regard to alternative approaches, we do not support a PRA model that involves the use of other comprehensive income (OCI) because there is no conceptual basis to mitigate profit or loss volatility simply by replacing it with volatility in OCI, particularly when the nature and purpose of OCI is unclear. There are also additional issues associated with this approach, such as recycling to profit or loss when managed exposures are sold or risk management instruments are terminated, and addressing how ineffectiveness ought to be reflected in profit or loss, or the appropriateness of not doing so. However, the IASB should explore the alternative of developing targeted solutions within the GHA requirements in IFRS 9 to address specific operational challenges when applying the GHA model to open portfolio hedges. Such an approach this has the potential to result in a more holistic approach to accounting for all types of hedges and to produce more comparable outcomes for similar types of hedges, whether or not the exposures are managed dynamically. That said, we acknowledge that that there might be no easy solutions to some of these operational challenges, and because of the dynamic nature of open portfolio hedges, many of the tracking and amortisation issues that work against the PRA would also exist under this approach. Our comments on the specific questions in the DP are as follows: Section 1 Background and Introduction to the PRA Section 5 Scope Question 1 Need for an accounting approach for dynamic risk management Do you think that there is a need for a specific accounting approach to represent dynamic risk management in entities financial statements? Why or why not? We agree that there is a need for an accounting approach for open portfolio hedges, primarily because the current lack of a comprehensive set of accounting solutions the GHA requirements of IFRS 9 do not work well for open portfolio hedges and the FV hedge accounting for open portfolios of interest rate risk in IAS 39 Financial Instruments: Recognition and Measurement is both limited in scope and difficult to apply has resulted in patchwork presentation in financial statements. We believe that the objective of such an accounting approach should be to represent the effects of dynamic risk mitigation using risk management instruments, which is aligned with the objective of GHA in IFRS 9, rather than to represent DRM per se, including decisions not to mitigate managed risk. As such, the accounting approach should seek to address accounting mismatches in hedged open portfolios within the boundaries of the Conceptual Framework for Financial Reporting (the CF) and existing principles. Page 3 of 22

4 That said, we think that the IASB should also explore, as part of further work on this project, the alternative of developing targeted solutions within the GHA model in IFRS 9 to address specific operational challenges when applying the GHA model to hedged exposures under DRM. This has the potential to provide a more holistic approach to accounting for all types of hedges and to produce more comparable outcomes for similar types of hedges, whether or not the exposures are managed dynamically. It could also be a solution to the inherent constraints of the PRA in situations where the managed risk affects only the cash flows but not the value of managed exposures (e.g. interest rate risk exposure to cash flow variability in floating rate instruments). However, we acknowledge that there might be no easy solutions to some of these operational challenges, including (i) whether and how to modify the basis adjustment requirement of the FV hedge accounting model, and (ii) how to address ineffectiveness and recycling under the cash flow hedge accounting model. In any case, because of the dynamic nature of open portfolio hedges, many (if not most) of the tracking and amortisation issues that work against the PRA would also exist under this approach. Question 2 Current difficulties in representing dynamic risk management in entities financial statements (a) Do you think that this DP has correctly identified the main issues that entities currently face when applying the current hedge accounting requirements to dynamic risk management? Why or why not? If not, what additional issues would the IASB need to consider when developing an accounting approach for dynamic risk management? (b) Do you think that the PRA would address the issues identified? Why or why not? We think that the issues for DRM of interest rate risk in banks as identified in the DP provide a good starting point for further work on the project, and that some of these issues are also applicable to other industries or types of risk in a DRM environment. We support, in principle, an accounting model for DRM that could accommodate different types of risk, and encourage the IASB to reach out to different industries to identify any other issues that are not faced by banks in dynamic interest rate risk management and to consider whether and how these issues could be addressed by the accounting approach. Question 15 Scope (a) Do you think that the PRA should be applied to all managed portfolios included in an entity s dynamic risk management (i.e. a scope focused on dynamic risk management) or should it be restricted to circumstances in which an entity has undertaken risk mitigation through hedging (i.e. a scope focused on risk mitigation)? Why or why not? If you do not agree with either of these alternatives, what do you suggest, and why? (b) Please provide comments on the usefulness of the information that would result from the application of the PRA under each scope alternative. Do you think that a combination of the PRA limited to risk mitigation and the hedge accounting requirements in IFRS 9 would provide a faithful representation of dynamic risk management? Why or why not? (c) Please provide comments on the operational feasibility of applying the PRA for each of the scope alternatives. In the case of a scope focused on risk mitigation, how could the need for frequent changes to the identified hedged sub-portfolio and/or proportion Page 4 of 22

5 be accommodated? (d) Would the answers provided in questions (a) (c) change when considering risks other than interest rate risk (for example, commodity price risk, FX risk)? If yes, how would those answers change, and why? If not, why not? We can better appreciate a scope focused on dynamic risk mitigation (henceforth known as the narrow scope ), which is aligned with our aforementioned thinking in that the accounting approach for open portfolio hedges should seek to address accounting mismatches between the hedging instruments and the managed exposures that do not represent the economics of their relationships, while preserving the measurement basis for the remaining exposures for which the risk is dynamically managed but not mitigated. In addition, we consider the narrow scope to the PRA to be a more targeted response to specific DRM strategies. For example, we understand that some banks operate in an environment that enables them to mitigate interest rate exposure from mismatched portfolios using a structural balance sheet approach. Specifically, these banks tend to have more insubstance fixed rate deposits than fixed rate loans. As such, they could manage mismatch positions by issuing new fixed rate loans via TP mechanism without significant difficulty, leaving them with net exposures that are not extensively mitigated using risk management instruments. Applying the PRA only to those exposures that are mitigated using risk management instruments would reflect the DRM strategy, while a scope focused on DRM (henceforth known as the broad scope ) would require costs to be incurred on revaluing all managed exposures, even those for which risk have been naturally mitigated. Nonetheless, we acknowledge that the narrow scope involves significant practical challenges as outlined in the DP. We urge the IASB to have a more targeted outreach to identify the depth and pervasiveness of these challenges and seek inputs on potential solutions, including practical expedients on tracking and amortisation. The narrow scope aside, we question whether the broad scope to the PRA would produce more relevant information. In particular, the broad scope would result in the revaluation of the entire managed portfolio (including unhedged portions) in respect of the managed risk. This would undermine the conclusion reached in IFRS 9 on the measurement of nonderivative financial instruments, which is that amortised cost is the most relevant measurement basis for qualifying debt financial assets that are held within a business model whose objective is to hold financial assets in order to collect contractual cash flows and most financial liabilities. While one may argue that a differential measurement basis to reflect risk management activities justifies a departure from the conclusion in IFRS 9, it would beg the question as to whether reflection of an entity s risk management activities should take primacy over reflection of its business model in the financial statements. We are also concerned about the volatility created in profit or loss under the broad scope to the PRA. While the DP suggests that such volatility would be reflective of changes in an entity s economic position and results from its decision not to hedge, we think that such volatility would not represent business performance since it merely reflects changes in pointin-time values, instead of results from decisions on how and when cash flows of managed exposures are realised. This issue is of particular concern when the revaluations have no relevance on future cash flows, such as the case where managed exposures comprise fixed rate financial instruments that are held for the collection of contractual cash flows. Such Page 5 of 22

6 volatility also appears counter-intuitive, given that the typical objective of DRM is to manage profit or loss and/or cash flows volatility, and hence, should not lead to greater volatility in profit or loss as compared to situations where DRM has not been undertaken. If it is determined that the broad scope would provide useful information, we believe that such information could be provided through expanded disclosures on an entity s risk management activities instead. However, the IASB should conduct further cost-benefit analysis particularly with preparers and users of financial statements before deciding whether, and the extent to which, such disclosures should be required. Question 16 Mandatory or optional application of the PRA (a) Do you think that the application of the PRA should be mandatory if the scope of application of the PRA were focused on dynamic risk management? Why or why not? (b) Do you think that the application of the PRA should be mandatory if the scope of application of the PRA were focused on risk mitigation? Why or why not? Building on the narrow scope to the PRA, we believe that the application of the PRA model should be optional. We do not think there are compelling reasons for mandatory application of the PRA when application of the IFRS 9 GHA requirements is optional. Furthermore, there are conceptual and practical difficulties associated with the mandatory application of the PRA, even under the narrow scope. For example, the alternative of applying cash flow hedge accounting under IFRS 9 would no longer be available for a portfolio of fixed and variable rate financial assets and financial liabilities. Similarly, an entity could no longer apply the GHA requirements of IFRS 9 so long as it has undertaken risk mitigation on a dynamically managed portfolio, even if risk mitigation was only undertaken on specific exposures within that portfolio. Furthermore, the costs of mandatory application might not always be justified in reality. Notwithstanding our support for the optional application of the PRA, we believe that voluntary discontinuation of the PRA should ordinarily not be permitted, once an entity opts to apply the PRA. However, we note that there would be situations where discontinuation of the PRA could be required, such as the termination or occurrence of significant changes to the DRM approach for a particular portfolio. The IASB should consider developing guidance on when discontinuation of the PRA should be required as well as the accounting treatment for the revaluation adjustment on discontinuation, and associated disclosure requirements. Question 17 Other eligibility criteria (a) Do you think that if the scope of the application of the PRA were focused on dynamic risk management, then no additional criterion would be required to qualify for applying the PRA? Why or why not? (i) Would your answer change depending on whether the application of the PRA was mandatory or not? Please explain your reasons. (ii) If the application of the PRA were optional, but with a focus on dynamic risk management, what criteria regarding starting and stopping the application of the PRA would you propose? Please explain your reasons. Page 6 of 22

7 (b) Do you think that if the scope of the application of the PRA were to be focused on risk mitigation, additional eligibility criteria would be needed regarding what is considered as risk mitigation through hedging under dynamic risk management? Why or why not? If your answer is yes, please explain what eligibility criteria you would suggest and, why. (i) Would your answer change depending on whether the application of the PRA was mandatory or not? Please explain your reasons. (ii) If the application of the PRA were optional, but with a focus on risk mitigation, what criteria regarding starting and stopping the application of the PRA would you propose? Please explain your reasons. Building on an optional PRA model based on the narrow scope, we believe that eligibility criteria, as well as disclosures on how these criteria have been met, should be required, in order to instil discipline in the application of the PRA. We think that possible eligibility criteria, in addition to the requirement for the risk to be managed dynamically, could include those related to the documentation of DRM objectives and strategies, existence of economic relationship between managed exposures and risk management instruments, clear identification of the portfolio (e.g. at the sub-portfolio level or a proportion of the portfolio in notional terms) for which risk is mitigated as well as the corresponding risk management instruments, and the conditions under which exposures and risks management instruments could be added to or removed from the hedging relationships. As the PRA would require all revaluation adjustments, and any resulting ineffectiveness, to be recognised in profit or loss, we are less concerned about the need for effectiveness tests, provided that an economic relationship exists between managed exposures and risk management instruments. Furthermore, any requirements for effectiveness tests would introduce significant complexity to the PRA. Section 2 Overview Question 3 Dynamic Risk Management Do you think that the description of dynamic risk management in paragraphs is accurate and complete? Why or why not? If not, what changes do you suggest, and why? Description of DRM We believe that a robust description of DRM is important because it is a pre-requisite for the application of the PRA and establishes the boundary within which specific additional disclosures would apply. If the IASB decides on the broad scope to the PRA, a precise definition of DRM (including what constitutes risk management and qualifying exposures as entities could identify and analyse more types of notional exposures than those that they undertake risk mitigation) becomes all the more important because it alone determines the population of exposures that would be revalued for the managed risk under the PRA. We broadly agree with the notion of DRM in the DP, but think that its description needs to be clarified for the following: (i) the types of qualifying risk; (ii) whether risk management Page 7 of 22

8 needs to be updated for changes to items within the managed portfolio only or for changes in market conditions in respect of the managed risk as well; (iii) whether DRM must necessarily include open portfolios (e.g. would a portfolio fall within the scope of DRM if it does not change frequently but is otherwise actively monitored for changes in market conditions in respect of the managed risk and for which risk management is updated on a timely basis in reaction to those changes); and (iv) whether DRM must necessarily involve net risk position (e.g. could DRM be undertaken for a portfolio of only assets or only liabilities). In addition, we understand that the term frequently in the phrase open portfolio(s), to which new exposures are frequently added and existing exposures mature may be subject to different interpretations depending on the context. For instance, for an entity that manages equity price risks of investments that are held for a longer horizon, it may have interpreted exposures that are added or removed only once every few months to have met the term frequently in the context of the holding period for those investments. We think that a contextual assessment as to whether exposures are frequently added or removed from a portfolio might not be appropriate and might lead to non-intuitive conclusions. If the IASB decides to retain the description of DRM, further guidance should be developed on how open portfolios should be determined. Description of dynamic risk mitigation Building on the narrow scope to the PRA, we believe that a robust description for dynamic risk mitigation is as important as it is for DRM. It would determine whether the PRA could be applied to an open portfolio that is managed on a dynamic basis, and if so, which portion of the portfolio should be revalued for the managed risk. We further believe that dynamic risk mitigation should be confined strictly to the hedging of the managed risk using risk management instruments. Practically, risk mitigation could take different forms, including active structural balance sheet management, which ought not to create accounting mismatches due to different measurement bases. However, the PRA should not be applied to exposures managed using the structural balance sheet approach since costs would be incurred to revalue a potentially large population of exposures in respect of the hedged risk without commensurate benefits. We note that one of the issues with an optional PRA model based on the narrow scope is its interaction with the GHA model in IFRS 9. Specifically, there is a need to determine whether the application of the PRA and the GHA should be (i) by way of choice, (ii) based on specified hierarchy, or (iii) mutually exclusive. We think that the PRA and the GHA should represent a comprehensive set of solutions for risk management activities, and that a free choice of approaches should be precluded on the basis that it might compromise the faithful representation of an entity s risk management activities, result in transactions of similar substance being accounted for differently and undermine the GHA requirements. We consider that a well-developed description of dynamic risk mitigation has the potential to distinguish its application from that of GHA and to instil discipline in the application of both models, e.g. to prevent the structuring of general hedge relationships into open portfolio hedges to take advantage of the naturally less stringent requirements of the PRA. We think that the IASB should perform analysis, as part of further work on this project, on whether it would be appropriate to require, if an entity elects to apply hedge accounting, either the application of (i) the GHA only, or (ii) the PRA to any exposures that are dynamically managed and risk mitigated and the GHA to any exposures that are managed and risk Page 8 of 22

9 mitigated on a static basis, and whether there are any unintended consequences from such an approach. Section 3 The Managed Portfolio Question 4 Pipeline transactions, EMB and behaviouralisation Pipeline transactions (a) Do you think that pipeline transactions should be included in the PRA if they are considered by an entity as part of its dynamic risk management? Why or why not? Please explain your reasons, taking into consideration operational feasibility, usefulness of the information provided in the financial statements and consistency with the Conceptual Framework for Financial Reporting (the Conceptual Framework). EMB (b) Do you think that EMB should be included in the PRA if it is considered by an entity as part of its dynamic risk management? Why or why not? Please explain your reasons, taking into consideration operational feasibility, usefulness of the information provided in the financial statements and consistency with the Conceptual Framework. Behaviouralisation (c) For the purposes of applying the PRA, should the cash flows be based on a behaviouralised rather than on a contractual basis (for example, after considering prepayment expectations), when the risk is managed on a behaviouralised basis? Please explain your reasons, taking into consideration operational feasibility, usefulness of the information provided in the financial statements and consistency with the Conceptual Framework. Question 9 Core demand deposits (a) Do you think that core demand deposits should be included in the managed portfolio on a behaviouralised basis when applying the PRA if that is how an entity would consider them for dynamic risk management purposes? Why or why not? (b) Do you think that guidance would be necessary for entities to determine the behaviouralised profile of core demand deposits? Why or why not? Behaviouralisation We see merits in the concept of behaviouralisation of cash flows in the PRA because it would be aligned with how an entity undertakes DRM activities. Furthermore, we do not think that it would be inconsistent with the CF, which includes a notion of behaviouralisation in cashflow-based measurements, and existing requirements such as the amortised cost concept and the expected loss impairment model in IFRS 9, which require measurements based on expected cash flows. Moreover, a portfolio FV hedge of interest rate risk under IAS 39 already allows prepayment risk to be incorporated by behaviouralisation. Page 9 of 22

10 Pipeline transactions, EMB and core demand deposits If pipeline transactions, EMB and core demand deposits are included in an entity s DRM activities, we acknowledge the argument that the inclusion of these exposures in the PRA could provide useful and relevant information because of its consistency with the entity s risk management perspective. Otherwise, accounting mismatches would persist, and financial statements would fail to completely portray the effects of risk mitigation efforts, if entities were to enter into derivatives to hedge the exposures from these items. However, we note that including EMB, and to a certain extent pipeline transactions, in the PRA would be inconsistent with both the CF and accounting conventions, which begs a fundamental question as to whether complete alignment between financial reporting and risk management is appropriate or even possible. We believe that a departure from the CF in order to reflect DRM activities in financial statements should be avoided as far as possible. Furthermore, including these exposures in the PRA would raise concerns on not only estimation uncertainties and subjectivity, but also whether converting a range of estimates into a single reportable number for each exposure would result in meaningful representation of risk management activities. Taking into consideration the above, we are not convinced by the argument supporting the inclusion of EMB in the PRA. We consider its conflict with the CF to be particularly severe because the recognition of gains or losses and resulting assets or liabilities on an entity s equity, which is the residual interest, runs contrary to the reporting entity concept, distorts performance reporting and affects the understandability of financial statements. We are also concerned about the verifiability of the revaluation adjustment on the EMB since it is premised on an entity s intention and projection, instead of historical data and forwardlooking market information. Besides, we are also less concerned about the possible nonalignment with DRM activities because EMB appears to be a less common exposure that is considered in DRM. We appreciate that pipeline transactions are commonly included in DRM and thinks that there are some merits in their inclusion in the PRA, to the extent that they have been marketed to the public and the entity has historically been honouring its commitment under such pipeline transactions. Such pipelines transactions could be viewed as behaviourally binding and akin to constructive obligations. Furthermore, there might be fewer concerns about estimation uncertainties and subjectivity if the PRA is based on the narrow scope, since entities typically undertake hedging only for those portions of exposures on which they have some level of reasonable estimates. In addition, we see merits in the inclusion of behaviouralised core demand deposits, which is routinely considered in DRM, in the managed portfolio when applying the PRA. We think that, given the sticky nature of core demand deposits, the uncertainties involved in the estimation of the deemed tenure and volume for the purposes of applying the PRA might not be significantly more than those related to other behaviouralised cash flows, such as expected credit losses in IFRS 9. In order to enhance the usefulness of information resulting from the revaluation adjustment, we believe that the PRA must be augmented by appropriate disclosures about how Page 10 of 22

11 behaviouralised core demand deposits and pipeline transactions are estimated, including the assumptions, methodologies, judgements and estimation uncertainties. Question 5 Prepayment risk When risk management instruments with optionality are used to manage prepayment risk as part of dynamic risk management, how do you think the PRA should consider this dynamic risk management activity? Please explain your reasons. When options are used to manage prepayment risk, we think that the PRA should, in principle, reflect the mitigation of a one-sided risk exposure. However, we acknowledge that there are significant operational difficulties when one-sided risk exposures are included in the PRA, e.g. determining the level of the one-sided risk when it is managed using multiple options, and distinguishing managed exposures for which one-sided risk was managed from those for which all changes in risk were managed. Although the DP discusses the operational issues related to instruments with optionality only in the context of prepayment risk, we note that, in a DRM environment, such instruments could be used more commonly to manage other risks, such as interest rate repricing risk and foreign exchange risk. The IASB would have to consider how these operational issues could be addressed, including the possible use of practical expedients. Question 6 Recognition of changes in customer behaviour Do you think that the impact of changes in past assumptions of customer behaviour captured in the cash flow profile of behaviouralised portfolios should be recognised in profit or loss through the application of the PRA when and to the extent they occur? Why or why not? We think that the impact of changes in past assumptions of customer behaviour captured in the cash flow profile of behaviouralised portfolios should be recognised in profit or loss prospectively, through the application of the PRA when and to the extent they occur. Such an approach would be consistent with the general requirements in IFRS related to the accounting for changes in estimates and the accounting for changes in hedge effectiveness under the IFRS 9 GHA model. Question 7 Bottom layers and proportions of managed exposures If a bottom layer or a proportion approach is taken for dynamic risk management purposes, do you think that it should be permitted or required within the PRA? Why or why not? If yes, how would you suggest overcoming the conceptual and operational difficulties identified? Please explain your reasons. We can appreciate permitting both the bottom layer approach and the proportion approach within the PRA as long as they are consistent with an entity s risk management strategy. However, we acknowledge the operational issues related to the bottom layer approach, i.e. identifying items that fall within the bottom layer of a non-homogeneous portfolio, and tracking and amortisation of changes in exposures within the bottom layer. We think that the operational issues for the proportion approach might be more generic, particularly if this Page 11 of 22

12 approach permits the PRA to be applied to a proportion of the managed portfolio, without specifically identifying the exposures that make up that proportion. We think that the IASB should perform further analysis on these approaches and consider developing practical expedients to address these operational issues. Question 8 Risk limits Do you think that risk limits should be reflected in the application of the PRA? Why or why not? We think that risk limits are inherent in the narrow scope to the PRA, i.e. they are the trigger point for the need to undertake risk mitigation, and would naturally be reflected in the application of the PRA. That said, we acknowledge that the act of risk mitigation may not always coincide with when an entity s net open risk position falls outside its risk limits (i.e. an entity can choose not to risk mitigate despite having breached its risk limits, and vice versa). Nonetheless, we are concerned that overlaying the narrow scope to the PRA with risk limits would introduce additional complexity without commensurate benefits. Under the broad scope to the PRA, the incorporation of risk limits in the PRA would lead to counterintuitive outcomes (i.e. wider risk limits lead to less volatility in profit or loss) that would not align well with the objective of the broad scope to the PRA, which is to represent DRM activities, including the effects of an entity s decision not to undertake risk mitigation for the managed risk. Question 10 Sub-benchmark rate managed risk instruments (a) Do you think that sub-benchmark instruments should be included within the managed portfolio as benchmark instruments if it is consistent with an entity s dynamic risk management approach (i.e. Approach 3 in Section 3.10)? Why or why not? If not, do you think that the alternatives presented in the DP (i.e. Approaches 1 and 2 in Section 3.10) for calculating the revaluation adjustment for sub-benchmark instruments provide an appropriate reflection of the risk attached to sub-benchmark instruments? Why or why not? (b) If sub-benchmark variable interest rate financial instruments have an embedded floor that is not included in dynamic risk management because it remains with the business unit, do you think that it is appropriate not to reflect the floor within the managed portfolio? Why or why not? We prefer an approach that includes contractual cash flows (including effects of the embedded floor if the coupon cannot be negative despite the negative margin) discounted at the benchmark rate, notwithstanding that the managed risk is typically the benchmark rate. Such an approach would better reflect the inherent hedge ineffectiveness for sub-benchmark instruments that are dynamically managed only for the benchmark risk. However, we acknowledge that this approach would add operational complexity to the PRA (e.g. tracking and amortisation of Day 1 revaluation) and recommend that the IASB should perform further work on possible practical expedients to address the operational challenges of tracking and amortisation. Page 12 of 22

13 Section 4 Revaluing the Managed Portfolio Question 11 Revaluation of the managed exposures (a) Do you think that the revaluation calculations outlined in this Section provide a faithful representation of dynamic risk management? Why or why not? (b) When the dynamic risk management objective is to manage net interest income with respect to the funding curve of a bank, do you think that it is appropriate for the managed risk to be the funding rate? Why or why not? If not, what changes do you suggest, and why? We recognise there are some merits in the revaluation calculations outlined in the DP and the possibility of the resulting revaluation adjustment as a faithful representation of the effects of dynamic risk mitigation. However, we understand that DRM activities could be undertaken to manage exposure to variability in cash flows instead of FV risk. For example, some banks may dynamically manage interest rate risk to protect future net interest income. From a conceptual perspective, the PRA could arguably be applied to cash flow hedges under DRM because measuring both managed exposures and hedging instruments at FV is likely to provide useful information about the residual exposure to volatility in cash flows due to ineffectiveness. However, the PRA would not work well for cash flow hedges of open portfolios of floating rate instruments since cash flow variability does not create a corresponding FV exposure. If the PRA were to be applied to all DRM activities, there would be a need for the IASB to explore alternative solutions for interest rate risk exposure to cash flow variability in open portfolios, such as developing solutions within the cash flow hedge accounting model in IFRS 9 to overcome its inherent restrictions. In addition, we think that the treatment of cost of hedging that is applicable to the GHA requirements of IFRS 9 should be considered for inclusion in the PRA, particularly if the IASB is to develop targeted solutions within the GHA model in IFRS 9 for open portfolios. However, we acknowledge that, because of the dynamic nature of open portfolio hedges, including the cost of hedging concept in the PRA would introduce additional complexity (e.g. determining the amortisation period, and developing accounting mechanics for subsequent over-hedges and reversal of over-hedges). We think that when a bank s risk management objective is to manage net interest income with respect to its funding curve, it would be appropriate for the managed risk to be the funding rate. Question 12 Transfer pricing transactions (a) Do you think that transfer pricing transactions would provide a good representation of the managed risk in the managed portfolio for the purposes of applying the PRA? To what extent do you think that the risk transferred to ALM via transfer pricing is representative of the risk that exists in the managed portfolio (see paragraphs )? (b) If the managed risk is a funding rate and is represented via transfer pricing transactions, which of the approaches discussed in paragraph do you think provides the most Page 13 of 22

14 faithful representation of dynamic risk management? If you consider none of the approaches to be appropriate, what alternatives do you suggest? In your answer please consider both representational faithfulness and operational feasibility. (c) Do you think restrictions are required on the eligibility of the indexes and spreads that can be used in transfer pricing as a basis for applying the PRA? Why or why not? If not, what changes do you recommend, and why? (d) If transfer pricing were to be used as a practical expedient, how would you resolve the issues identified paragraphs concerning ongoing linkage? Fundamentally, we believe that, for the purposes of applying the PRA, the managed risk should be represented by well-established benchmark indexes (e.g. LIBOR) and for which risk mitigation could be undertaken. Otherwise, the application of the PRA could result in information that would not faithfully represent the effects of risk mitigation, or in some situations, would be practically impossible. The PRA should also capture actual risk exposures and any hedge ineffectiveness that exists between the managed exposures and the risk management instruments. Notwithstanding so, we appreciate that the PRA should be operationally flexible to accommodate various DRM strategies that exist in practice, whilst preserving the principles underlying the PRA. Following from the above position, we can appreciate the merits of using TP transactions, both in terms of the TP rate as well as the exposures in the TP transactions, as a proxy for revaluing the managed exposures, if TP transactions are representative of the managed risk exposures. This would provide much operational relief because managed exposures are transacted over time and in varying interest rate environments, and hence, determining for fixed rate instruments the level of managed rate at the time when they first give rise to the exposure could be difficult, if the risk management and business unit systems are not integrated. However, restricting its application only to TP transactions that are representative of the managed risk exposures would be particularly important because giving entities a free reign to use TP transactions as a proxy for managed exposures, without requiring some tests for representativeness, provides opportunities for abuse. Notwithstanding the above, we have some reservations about whether such an operational expedient would be effective in providing operational relief in practice. While the operational expedient might work for the simplest TP arrangements (e.g. TP rate comprises only the benchmark rate, the entity s external funding credit spread and/or normal profit margin, while the DRM strategy requires all external exposures to be backed by corresponding TP transactions), it is unclear whether it would be appropriate or how much flexibility it should be allowed as proxies, in light of the different DRM strategies that exist in practice. These issues are compounded by the fact that TP is commonly used as a tool to encourage particular behaviours in business units, which has little relevance to the managed risk and would be adjusted over time in response to changes in market conditions and business strategies. Even in situations where the TP rate includes only the entity s external funding credit spread and normal profit margin, we understand that banks would ordinarily manage the TP transactions without considering these spreads/margins, even though these spreads/margins are components of actual exposures in a managed portfolio. Furthermore, the risk management and business unit systems are often not integrated, and DRM had to be undertaken for practical reasons using TP transactions as a proxy for actual exposures. The Page 14 of 22

15 application of the PRA in the context of such DRM strategies is both conceptually and operationally challenging. We also share the concerns outlined in the DP in respect of ongoing linkage between TP transactions and the managed risk exposures over time. We understand that TP rates are typically more sticky than benchmark rates and some DRM strategies might give business units some latitude in how internal funding is applied, e.g. applying internal funding released from prepaid loans to new transactions, which may render TP transactions less representative of the managed risk exposure. The issue is whether it would be practicable to develop some tests of representativeness that entities would have to perform to validate the use of TP transactions as a proxy for managed exposures. In view of the above difficulties, we believe it is imperative that the IASB gives serious consideration to whether TP transactions could be meaningfully included in the PRA to achieve a reasonable representation of the effects of risk mitigation. We note that all the three approaches outlined in the DP require some form of tracking to remove the effects of spreads/margins or to amortise the Day 1 revaluation difference when applying the PRA. Arguably, in order to meaningfully reduce the need for tracking, it might be necessary to adopt an approach that requires cash flows that are based on the TP rate (including spreads/margins) to be discounted using the current TP rate, provided that the spreads/margins in the TP rate meet specified eligibility criteria. However, such an approach would require robust eligibility criteria, which in itself could be a significant challenge. For example, it is arguable whether normal profit margin, being one of the most common elements of TP rates, could be representative of the managed risk. Even if it is considered so, determining whether a particular internal profit margin includes elements that are not attributable to normal profit margin, such as adjustments made to encourage particular behaviours in business units, could be subjective or arbitrary. We recommend that the IASB should perform further analysis on whether these conceptual and operational issues could be addressed before exploring how TP transactions could be an effective operational expedient. Question 13 Selection of funding index (a) Do you think that it is acceptable to identify a single funding index for all managed portfolios if funding is based on more than one funding index? Why or why not? If yes, please explain the circumstances under which this would be appropriate. (b) Do you think that criteria for selecting a suitable funding index or indexes are necessary? Why or why not? If yes, what would those criteria be, and why? We agree that applying the PRA could cause significant operational difficulties, when banks raise funding from a variety of sources and manage their portfolios without matching particular assets as being funded by particular liabilities. Even if banks assign a particular funding index to particular portfolios via TP transactions, there is also the issue as to whether the assigned funding index is indeed representative of the risk exposures within those particular portfolios. Therefore, there is a need to explore how the PRA could be made more operational when funding for a managed portfolio is based on more than one funding index. We think that under such situations, it is possible to consider identifying a single funding index for all managed portfolios to be incorporated within the PRA, particularly when there is a predominant funding index that is used to price most of the funding. However, the Page 15 of 22

16 relevance of a single funding index diminishes significantly when the funding for the managed portfolios is well distributed over a range of sources with different funding indexes. A conceivable alternative is perhaps to permit the use of a blended rate index that is constructed based on the weighted average of actual funding indexes within the managed portfolios. The IASB should perform further analysis on whether, when and how the use of a single index for the purposes of applying the PRA could be acceptable. Irrespective of the approach adopted, additional disclosure on how the single index is selected or determined would be necessary. In addition, we note that there are further complications arising from the interaction between multiple funding indexes and TP transactions when both are used in DRM. Specifically, for some DRM strategies, the funding curve that is being managed represents blended rates that are determined based on different benchmark indexes used for various funding sources, and these blended rates are used as the TP rates, which form the basis on which the pricing for portfolio assets is determined. Risk mitigation is then undertaken using different derivatives that are based on the underlying benchmark funding indexes. The PRA requires the revaluation of cash flows that represent the exposure to the managed risk, but it is unclear how these cash flows should be determined in such situations. The IASB should consider developing further guidance on the determination of the revaluation adjustment under a DRM strategy whereby the managed risk represents a blended rate and is hedged using the benchmark rates as a proxy for the risk resulting from actual exposures. Question 14 Pricing index (a) Please provide one or more example(s) of dynamic risk management undertaken for portfolios with respect to a pricing index. (b) How is the pricing index determined for these portfolios? Do you think that this pricing index would be an appropriate basis for applying the PRA if used in dynamic risk management? Why or why not? If not, what criteria should be required? Please explain your reasons. (c) Do you think that the application of the PRA would provide useful information about these dynamic risk management activities when the pricing index is used in dynamic risk management? Why or why not? We think that the use of the pricing index for the purposes of applying the PRA would be appropriate if the exposure is managed on this basis. However, we understand that DRM based on a pricing index is not common. Page 16 of 22