Comments from ACCA
ACCA is pleased to have this opportunity to comment on the proposed accounting standards update (PASU) from the Financial Accounting Standards board (FASB) on the above subject, which was considered by ACCA's Financial Reporting Committee.
The financial crisis has highlighted the significance of the issue and it needs to be addressed by FASB and other accounting standard setters. The G20 meeting of the leaders of the most important countries has repeatedly called for the improvement of accounting standards in this area by the development of single high quality standard for use across the world.
ACCA has supported the development of a single set of global accounting standards and in particular the joint development of new standards between International Accounting Standards Board (IASB) and FASB to help achieve that. We therefore regret the lack of co-ordination on this the most critical project for a new accounting standard. Appendix A sets out the significant differences between this PASU and the standards and EDs issued by IASB in their project to replace IAS39. These are highly significant differences which cannot be skated over. We urge FASB and IASB to try to achieve a similar standard nonetheless. Unless they do a key plank of the convergence programme will have been lost. The demands will continue for a 'level playing field', which have prompted some adverse developments in financial reporting over the past two years in both US GAAP and IFRS.
In our view the proposals in the PASU are moving in the wrong direction on two key issues. We broadly support the IASB's proposals that some financial instruments are better stated at amortised cost and some are better at fair value as compared to the largely fair value model in the PASU. We have also supported the principle that the impairment of financial assets should be on the basis of expected losses requiring earlier recognition of impairments and not restricting losses to those incurred, the approach in this PASU.
We have responded to FASB's request for comments because we would like to contribute to the development of a new converged standard in this area given the importance we attach to global standards. In our responses we have addressed only those questions where we feel we can usefully contribute and have not covered for example those which relate principally to US specific issues or to those which deal with how the PASU would integrate with the rest of US GAAP.
ACCA's responses to FASB's specific questions
Q8. We agree with the main proposals for initial measurement.
Q9. We think that there are some financial instruments that are more usefully stated at amortised cost. The sort of instruments that the PASU would allow to have qualifying changes in fair value recognised as other comprehensive income (QCOCI) are included in the category that should in our view be using amortised cost. The effect of the QCOCI treatment seems to be to try to have an income statement which provides a measure of performance based on amortised cost. Given these we would support the use of the transaction price as the initial measurement unless it could be shown that the transaction was not at arm's length.
Q10 and 11. We agree that the subsequent treatment should affect the initial recognition amount. Treatment at fair value through the income statement means that initial acquisition fees and costs should be expensed. Given the apparent objective of the QCOCI category then the fees and costs should in this case be spread as part of the amortisation.
Q13 and 15. We do not agree that the default measurement basis for all financial instruments should be fair value. Financial assets that are held to collect their contractual cash flows and where these consist of interest and capital repayments should be stated at amortised cost. Where fair values are not practicable to obtain, for example many unlisted equity investments, then amortised cost should be used. Financial liabilities should be stated at amortised cost except where specific criteria are met – they are derivatives, held for trading or there would otherwise be an accounting mismatch. In all these cases amortised cost is a better indicator of expected future cashflows from these instruments than fair value and so we believe will be more helpful to users of the financial statements.
Q14. Given the apparent objective of the QCOCI model we agree with the proposed treatment.
Q16. The different treatments under this PASU and the classifications in IFRS9 depend on the business model within which the instrument is held. Business models are not matters that change very often if at all, but when they do they should have an impact on the accounting treatment. We support mandatory reclassifications with appropriate disclosures of the effect and the reasons, but only if the business model has changed.
Q17 and Q31. We do not support the proposed measurement basis for core deposit liabilities. As noted above the default measurement for liabilities should in our view be amortised cost. The proposed measurement basis would show these liabilities at less than their contractual value which are repayable on demand or at short notice. This seems inconsistent with how other obligations are shown in financial statements where the contractual basis is the basic principle, subject to a recognition of the time value of money. The only justification for the proposed re-measurement basis can be that the contractual liability is reduced by the recognition of an intangible asset for the customer relationship and this is inconsistent with how other internally generated intangibles are accounted for.
We also think that there will be practical difficulties with definitions, and so applications, of the alternative funds rate and the all-in-cost-to-service-rate.
Q18. We agree that in these cases financial liabilities should be at amortised cost, but as already noted we would prefer amortised cost be the normal method for measuring liabilities and a more restricted group should be at fair value.
Q24. We do not think that the dual presentation on the face of the balance sheet of both the amortised cost and fair value for a range of instruments will be very helpful to users. This will tend to produce a clutter of figures on the primary statement that will tend to inhibit good communication. Our preference would be for different instruments to be stated at either amortised cost or at fair value as is most appropriate and for the alternative valuation to be disclosed by way of a note to the financial statements.
Q32. We do not agree with the proposed presentation of changes in own credit standing. These gains or losses have only dubious value in terms of the performance of the business and could be misleading for users. We would prefer for the few liabilities (other than derivatives) shown at fair value that the component of the fair value change which derives from changes on own credit risk, should be part of other comprehensive income.
Q33 and 34. We find that neither of these methods is a convincing way of measuring the fair value of liabilities. Method 1 has the problem of application to unrated entities and method 2 assumes that the sector's credit will be priced similarly which seems unlikely to be true in all cases.
Q37 and 38. There are advantages and disadvantages to both the incurred loss (that is included in the PASU) and to the expected loss model proposed by IASB. On balance we think in principle the expected loss model is superior, especially for banks and other lenders, because the reported interest income will be more predictive of the future cash flows and make the initial rate of return reported more helpful to users.
Q39. We agree that these changes should not produce a credit impairment.
Q40. No methodology should be specified.
Q41. No. There should be a reversal of the credit impairment in net income and this should not be spread forward via the effective interest rate. We see no reason why there should not be symmetrical treatments between impairments and reversals of impairments in this regard. Immediate recognition records in income the economic event causing the reversal at the time when it happens.
Q48 to 50. In line with the expected loss model the effective interest rate should be set at initial recognition and any changes in impairment expectations being recognised as they occur rather than being spread forward over the remaining life of the asset. We therefore answer no to Q48 when it refers to QCOCI assets and yes to Q49. Q50 refers to proposals that for FVTPL assets there should be an option that interest income could be separately identified from other changes in fair value. Given the different business strategies which might be in place for these assets an option would be sensible. However if interest income is separately identified it would make sense for this to be done on a consistent basis with QCOCI assets.
Q56 to 58. We see this as principally an area of interface with US GAAP. Hedge accounting has been identified in surveys we have done of our members as a key area of complexity in IFRS. Therefore the opportunity of this revision of the accounting standards should be used to try to reduce that complexity. The proposals in this PASU should be a key input to the thinking of IASB on this subject as maintaining convergence between the two systems should be a priority.