السالم عليكم ورحمة هللا وبركاته

Ref. 1017/5117/SS 3 rd October 2017 Dr. Hamed Hassan Merah Secretary General Accounting & Auditing Organization for Islamic Financial Institutions (AAOIFI) Al Nakheel Tower 10 th Floor, Office 1001, Building 1074 Road 3622, Seef Area 436 Manama Kingdom of Bahrain Dear Dr. Merah, السالم عليكم ورحمة هللا وبركاته CIBAFI Comments on the AAOIFI Exposure Draft (ver. 5.0) on Financial Accounting Standard (FAS) No. 30 Impairment and Credit Losses The General Council for Islamic Banks and Financial Institutions (CIBAFI) compliments the Accounting & Auditing Organization for Islamic Financial Institutions (AAOIFI) and takes this opportunity to express its appreciation of the work that the AAOIFI does to promote and enhance the Islamic financial services industry (IFSI). CIBAFI is the official umbrella for all Islamic financial institutions, whose services and products comply with the Shariah rules and principles. CIBAFI acts as the voice of the Islamic finance industry, where our members comprise Islamic banks and non-bank financial institutions, both large and small, and reach more than 120 members from 32 countries and jurisdictions. We noted the request for comments on the AAOIFI s Exposure Draft (ED) on Financial Accounting Standard (FAS) No. 30 Impairment and Credit Losses and we welcome this

opportunity to offer our comments and recommendations. The comments contained in this letter represent the views of CIBAFI Secretariat and feedback received from our members. We are also attaching more detailed comments in the Appendix of this letter for AAOIFI s kind consideration. It has been difficult to evaluate this standard in isolation, and in particular without seeing the proposed standard on Reserves, which is intended to be adopted at the same time. In addition, as the standard covers some of the same topics as IFRS 9, it would have been helpful if the consultation had been accompanied by a description of the principal similarities and differences between the two in the areas where they overlap. A number of our members have to use IFRS for some or all of their financial statements, and are naturally concerned to understand where the differences lie. This is a point you may wish to consider for future consultations, whenever a proposed FAS covers similar territory to an IFRS standard. In addition, it would have been helpful to some members if the ED had been published in Arabic as well as English, and CIBAFI believes this would have increased the number of Islamic banks responses and engaged more interested parties in providing comments on the ED. Again, this is a point you may wish to consider in future consultations. As regards the substance of the Exposure Draft, CIBAFI welcomes in principle the application of an Expected Credit Loss (ECL) approach in Islamic finance, both on its merits and because of the closer alignment this will create between the Islamic and conventional standards, notably the relevant parts of IFRS 9. However, the proposed FAS 30 appears to diverge from IFRS 9 in a number of ways that do not obviously derive from the specificities of Islamic finance. It would be helpful if these two could be aligned as far as possible in order to facilitate parallel application of the two standards. We draw attention in the Appendix to some specific points where our members would like to see greater convergence. The text of the FAS 30 ED is shorter than that of IFRS 9, even for those topics covered 2

by both standards. For example, the material on determining significant increases in credit risk is much longer in IFRS 9, but this is by no means the only example. Although the FAS 30 ED occasionally borrows language from IFRS 9, the text is not closely parallel. It is unclear whether anything of importance has been lost as a result, but our members are concerned that the lack of detail will make the standard more difficult to apply. It is also unclear whether the substantive effect will be different, other than where this is necessary to reflect the specificities of Islamic finance. Therefore, CIBAFI recommends the standard to provide more elaboration and explanations, especially with regards to details on specific characteristics of Islamic banking products and practices. Finally, a number of our members have expressed concern at the exclusion of Mudaraba and Musharaka contracts from the ECL approach in respect of future cash flows. They have pointed out that in practice such contracts are often structured so that the principal risk for the Islamic Financial Institution is counterparty risk, and also that the approach proposed, of revaluing the underlying asset, can be very difficult to apply in practice. Again, these points are made in more detail in the Appendix. Ideally, AAOIFI would have aligned the date of implementation of the standard with that of IFRS 9, which comes into effect for financial years beginning on or after 1 January 2018. We recognise that it may be late to do this at this stage especially if FAS 30 is to be synchronised with the new standard on Provisions, for which no Exposure Draft has yet been published. However, we do wish to underline the importance that our members attach to as nearly as possible simultaneous implementation of new AAOIFI standards and their IFRS counterparts. We therefore see completion of this standard as a matter of some urgency. We would like to express our thanks to the AAOIFI for its great effort and commitment with respect to developing standards that accommodate the interest of the global Islamic finance industry. 3

We remain at your disposal should you need any further clarifications on the above or on the attached Appendix. Yours sincerely, Abdelilah Belatik Secretary General 4

Appendix Comments on Exposure Draft of AAOIFI Financial Accounting Standard (FAS) No. 30 Impairment and Credit Losses CIBAFI s analysis has identified the following comments with regards to the specificities in the FAS 30 Exposure Draft. Comments follow the structure of the ED. 1. Definitions: The ED states in definition 6(f) that fair value is the amount for which an asset could be exchanged, or an obligation settled, between well informed, willing parties (seller and buyer) in an arm s length transaction. However, it may be noted that IFRS 13 defines fair value as the price that would be received to sell an asset or paid to transfer a liability in an orderly transaction between market participants at the measurement date (an exit price). It is not clear why the FAS definition differs from the IFRS one and we recommend that the two be aligned. The ED defines Lifetime expected credit losses in 6(k) as the expected credit losses that result from all possible default events over the expected life of a receivable or irrevocable unutilised commitment component. However, under IFRS 9, lifetime ECL is the expected present value of losses that arise if borrowers default on their obligations at some time during the life of the financial asset. For a portfolio, ECL is the weighted average credit losses (loss-given-default) with the probability of default as the weight. Again we recommend that the two be aligned; inclusion of the probability weighting in the case of a portfolio is particularly important. The definition of 12-month expected credit loss in 6(w) is the portion of lifetime expected credit losses that represent the expected credit losses that result from default events on a receivable that are possible within the 12 month after the reporting date. IFRS 9 refers to a financial instrument rather than a receivable, and we consider this better since the scope of the ED also includes loan commitments and financial 5

guarantees. We therefore suggest that AAOIFI have a single definition of a financial instrument or financial asset and reference it in the document instead of receivable. We also recommend that AAOIFI add additional definitions of: derecognition, writeoffs and originated credit-impaired financial assets. 2. Para 7b identifies assets and exposures subject to an impairment approach. There is some ambiguity in the drafting here, which led one of our members to believe that the bracket (subject to impairment approach) was intended to apply to the previous word inventories. We do not believe that was the intention, but you may wish to look again at the drafting. 3. In para 7 more generally, members indicated that it would also be helpful if AAOIFI could provide guidelines and examples of the types of financial instruments to which each of the approaches should be applied. 4. With regard to Para 8, determination of credit deterioration, we consider that (as in IFRS 9) an entity should also determine impact of impairment when it becomes aware of any incident which indicates credit deterioration. 5. Paras 15-17 The ED deals relatively briefly with how to determine whether there has been a significant increase in credit risk, and how to measure credit losses. Islamic banks face some difficult issues here, sometimes relating to Shariah and supervisory issues. For example, they may have to deal with apparently solvent customers procrastinating or delaying payments. Banking supervisors may also have their own classifications of nonperforming financing. While a difference between accounting and regulatory standards can never be excluded, it would be helpful if the ED could provide more detail on how to assess the increase in the expected risk and the resultant losses. 6. Para 16 could helpfully provide guidance for low credit risk similar to that in paras B5.5.22-B5.5.24 of IFRS 9. 6

7. Para 17 might also mention a rating downgrade as a factor that might cause an institution to identify a significant increase in credit risk. 8. The ED deals to some extent with modified receivables and restructurings (paras 18 and 19). More details in this area would be welcome. We note that, while IFRS 9 deals in some detail with how the modified cash flows should be evaluated, and whether a financial asset should be derecognised and replaced by a new one, the ED deals only with considering whether there is an increase in credit risk. 9. Para 19 states that if the proceeds of a new contract have been applied for settlement of an earlier receivable with the same customer, there is a rebuttable assumption that there has been a significant increase in the credit risk or risk of impairment in respect of the new contract. We suggest that the above only applies if there is objective evidence of deterioration of the financial soundness of the client. It is common for Islamic banks to roll over Murabaha contracts for genuine business purposes where the mechanics of the transaction is to execute a new Murabaha contract to settle the previous Murabaha contract. 10. In para 20 b the reference to the allocation of relevant income applying the effective rate of return method is not clear. Further, the reference here is to contractual maturity, whereas in para 24, the reference is to maximum contractual period. The two may differ if the contract has extension options. We suggest keeping both paragraphs consistent by using maximum contractual period. 11. Additionally, in para 20 c, we suggest replacing the wording of with reasonable to without undue cost or effort, in line with the IFRS 9 requirements. 12. Para 21 of the ED refers in general terms to other AAOIFI standards, the text being: The computation of the credit losses shall be based on the amount of receivable net of any deferred profits accounted for under the requirements of respective AAOIFI FAS. ; Unless contrary to a specific requirement of respective AAOIFI FAS (para 46). Mentions of other AAOIFI standards should be made more specific by citing the standard 7

and, where appropriate, the paragraph(s) to make it easier to understand this standard fully. 13. Para 25 states that the total exposure of a transaction includes the receivable and an irrevocable unutilized commitment component. We suggest that the AAOIFI consider defining the period under consideration to be based on the possible maximum period (i.e. including the unutilized commitment period). But we recommend that the definition of total exposure in such cases might be kept out of scope of the document as it entails modelling how the unutilized commitment will convert to credit exposures. 14. Para 28 sets out which receivables should be subject to 12-month expected losses, and which to lifetime expected losses. But it also says whether these allowances should be collective or specific. We suggest that this latter element should be removed from this standard, partly because it is more appropriate to the new standard on Provisions, but also because it requires a more nuanced treatment, which may depend on the modelling approach used by the bank. Sophisticated banks may model stage 1 allowances on a customer/asset specific approach. We also note that IFRS 9 does not make reference to collective and specific provisions. See also para 32 where a segregation between incurred losses (specific) and expected losses (collective) is maintained in FAS-30 while IFRS-9 has moved away from such segregation and all losses are treated as Expected Credit Losses. Since the basis of calculation of losses is same it is not clear why this segregation is retained. 15. Also in para 28 it would be helpful to have guidance on how movements between the three stages should be handled, though we recognise that this may be more appropriate to the standard on Provisions. 16. Para 30 says that the calculation of expected credit losses has to consider the risk weightage. Risk weighting is normally a capital adequacy concept, and we are unclear how and why it is being used here. (We did not find a comparable concept used in IFRS 9). We also suggest that the list of factors to be considered should be subject to availability of information without undue cost or effort. 8

17. Para 32 states that An allowance for credit losses duly distinguishing between incurred credit losses and expected credit losses shall be shown as a deduction from the gross carrying value of the respective asset. However, it is not always practically possible to disclose both elements (incurred and expected losses) separately in the financial statements. Hence para 58 is more reasonable with regard to disclosure (total provision per each stage to be disclosed). We also suspect that the term expected incurred credit loss is a typo and the reference should be simply to incurred credit loss. 18. In para 34 it is not clear why provision against off-balance sheet exposures should be allocated to shareholders only. Where depositors (IAH) participate in fee and commission income on off-balance sheet exposures, any resultant provisions on off balance sheet exposure should also be allocated to depositors (IAH). 19. Para 36 concerns the way that impairments are to be treated in the case of Profit Sharing Investment Accounts (PSIAs). The ED notes at a number of points (including para 36) that impairments need to be properly attributed to other participatory stakeholders such as unrestricted investment account holders. However, AAOIFI has taken the view that this requires the consent of those stakeholders (para BCS2). Our members are concerned that seeking retrospective consent would be impracticable in practice. We note, however, that this view does not form part of the standard proper. 20. Para 45 restricts the range of assets to which the ECL is to be applied, and is supplemented in particular by para BCA4. The ED considers that it should be applied strictly to credit receivables, such as those resulting from Murabaha transactions or Ijarah rentals. These restrictions have given rise to very considerable concern among our members. Some at least consider that Ijarah Muntahiya Bitamleek, in substance, is a financial lease and should be treated as a credit receivable and calculate the provisions based on the expected credit loss approach. Their concerns are still greater in relation to Mudaraba or Musharaka contracts. Although in form these contracts are quasi equity-like, some of our banks see them as, in substance, 9

financing transactions, where the counterparty risk is the most material and would expect to treat such contracts using the expected credit loss approach since the substance of the transaction is a receivable. The changes in the asset value are mainly subject to market risk and other factors which can be modelled when calibrating the Loss Given Default (LGD) or the cashflows as a part of the expected credit loss calculation. Furthermore, an approach which depends dominantly on a regular revaluation of the underlying asset is difficult enough in a consumer transaction such as a Diminishing Musharaka mortgage, but is still more difficult in a business transaction, where the asset may be highly specific to a particular business or project. Our members have also raised issues as to how any impairment related to asset value should be allocated, and how changes in this value should be treated. We consider that AAOIFI needs to consider again how these contracts should be treated, and also give much more detail of how its proposed approach should be applied. 21. As regards onerous contracts or commitments (paras 51-55), the ED mentions two examples (Ijarah and Salam). The concept of an onerous contract is one that does not seem to be present in IFRS 9, and we should welcome more examples and discussion of how this concept is to be applied in practice. 22. Under para 63 the applicability of the standard on retrospective basis without appropriate restatement of comparative figures could be misleading. 23. Para 64 states that in case of a shortfall (of allocated provisions from IRR to unrestricted investment account holders), a temporary transfer with due Shariah approval may be made from shareholders equity to be recoverable from income of such class of stakeholders within a period in line with the maturity of respective assets to be duly specified in the approval. We request AAOIFI to clarify whether the temporary transfer is applicable during transitional periods, i.e. up to 1 January 2020. Also, please clarify the mechanics and timing of crediting the provisions back to the shareholder account. Islamic banks may have already allocated previous provisions to unrestricted investment account 10

holders and having the cumulative charge allocated once again may create a duplicate charge on unrestricted investment account holders. 24. Para BCA5 comments that the general provision under the existing standard is meant for the past and present losses only and does not take care of the future losses. This will therefore need to be an issue to be considered when drafting the revised standard on Provisions. In this context we note that the definition of Tier 2 capital under the BCBS Basel III regime treats as eligible (up to a certain limit) Provisions or loan-loss reserves held against future, presently unidentified losses [which] are freely available to meet losses which subsequently materialise Provisions ascribed to identified deterioration of particular assets or known liabilities, whether individual or grouped, should be excluded. We suggest AAOIFI should consider whether the BCBS s definition would be a helpful reference drafting its revised standard on provisions. 11