top stories

Share

Comments from ACCA
May 2007

The Association of Chartered Certified Accountants (ACCA) is pleased to have this opportunity to comment on the discussion paper of the International Accounting Standards Board (IASB) on fair value measurements, which was considered by ACCA's Financial Reporting Committee . I am writing to give you their views.

Overall comments

The IASB should be careful in the development of the proposals contained in the Discussion Paper (DP). The whole concept of fair value (FV) is a difficult one with many parties who believe that fair values are going to be evermore widely required in IFRS. We are aware that the DP concerns how fair values should be reached and not when FV should be required. IASB have also stated that any application of this DP would be preceded by a complete review of cases where IFRS require fair value. These limitations are however easily overlooked in the debate on this DP. It is unsettling to many that the bigger question of the extent to which fair values should be used appears to be being ignored as seemingly already decided, in favour of the more limited (and one might have thought subsequent) consideration of exactly how to reach fair values. We are aware that the bigger question is being addressed in the review of the conceptual framework and IASB should progress this in step with the subject of this DP.

The DP promotes one particular FV measure that might be better called current market exit value (CMEV). CMEV in our view will not be suitable for every case where the term ‘fair value' is used currently in IFRS, let alone to further extensions of FV. The conceptual framework project seems to be commencing with an inventory of measurement bases on a more precise basis. The debate on appropriate measurement bases in IFRS will be facilitated by using terms of greater precision (such as depreciated replacement cost or CMEV) instead of the vaguer terminology (for example historical cost or FV). Generally we consider that the DP sets out a good, if lengthy, exposition of how to reach this one measure CMEV, but does not set out how to reach other sorts of ‘fair value' measurements, despite its title referring to the plural.

ACCA's responses to questions raised for comment by IASB

Q1. Would a single source of guidance for all FV measurements in IFRS both reduce complexity and improve consistency in measuring FV?

We would like to see terms used consistently within IFRS and that as much as possible any guidance is converged with practice under US GAAP. Whether a single source of guidance in the form a new IFRS is needed to achieve this is not clear to us. There would be some guidance which would be specific to individual IFRS, for example for fair values in agriculture or insurance, even if the general principles were the same. The ‘unit of account' issue might also need different answers in different cases.

In addition we assume that considerations of verifiability/reliability of the FVs and of practicability or cost/benefit exceptions would have to be made in individual standards.

We are not convinced that the CMEV promoted by this DP would be appropriate wherever the phrase ‘fair value' is currently used in IFRS and there would have to be a reconsideration of where CMEV should apply (as noted in paragraph 17).

Q2. Is there FV measurement guidance in IFRS that you believe is preferable to the provisions of SFAS157?

Yes. IFRS are addressing fair value as a more widely applied term than SFAS157 and so some of its existing guidance is more appropriate to those uses. Even for CMEV sometimes the guidance in IFRS is better, as we note in our answer to Q21 below.

Q3. Should fair value (FV) be defined as an exit price from the perspective of a market participant that holds the asset or owes the liability?

The definition proposed we consider is that for CMEV and as noted above more precise definitions of measurement bases would be helpful. At present in IAS39 fair value for financial instruments seems part way between an exit and an entry value. In that standard CMEV should be used in place of the term fair value and the greater precision in these proposals would be helpful. However as noted above, not all cases where the term FV is currently used in IFRS should be switched to CMEV.

Q4. Do you believe an entry price also reflects current market-based expectations of flows of economic benefits? Do you agree that, excluding transaction costs, entry and exit prices will differ only when they occur in different markets?

Yes in principle to both these questions. The entry prices reflect the expectations of the cash flows of buyers. To the second question, however, the answer in practice must be no. While entry and exit prices should in theory be the same, in many markets it can be observed there is in practice a dealing margin or profit between buying and selling – certainly outside deep and liquid markets for commodities or financial instruments. It is also observable that transaction costs can be rolled up into that spread and that it can be difficult to identify these separately.

Q5. Would fair value be better replaced by some other term?

Yes. Current market exit value would be a much more exact term to cover the model promoted in this DP. It would also be a more neutral term. ‘Fair value' as a term also has implicit connotations of superiority and honesty which detract from its neutrality.

Q6. Does the exit price measurement objective in SFAS157 differ from FV measurements in IFRS as applied in practice?

Yes. We agree with the cases identified in the DP of the use of fair value in IFRS as entry values

• IFRS3 – where the costs of an acquisition have to be allocated among acquired and recognised assets and liabilities using FV
• IAS17 – in lease accounting the FV of the leased asset may be used and this will often be the price if acquired for cash.

In addition we note that

• FV in IAS16 is used for non-cash consideration to acquire property or other assets which is again an entry price
• Also for revaluation IAS16 refers to replacement and therefore entry value, and so for instance specialised plant should be valued at depreciated replacement cost.

Q7. Do you agree with how the market participant view is articulated in SFAS157?

No. We note that SFAS157 aims at a value a market participant might use and this depends on the existence of a market to participate in. A market implies trading on a regular basis between a number of possible buyers and sellers. For most items in a balance sheet there may be trades but in some cases these are very infrequent and between a restricted number of players.

SFAS157 sets out a hierarchy of different levels of FV information. Level 3 inputs (which still can be used for reaching a FV) consist of unobservable market inputs put into a FV model. We do not agree that unobservable inputs can produce a market value.

Q8. Do you agree that the market participant view in SFAS157 is consistent with the existing definition in IFRS?

Yes. They would seem to be much the same, with the exception of the specific references to markets and the implications of that noted above in our answer to

Q9. Do you agree that the FV of a liability should be based on the price that would be paid to transfer the liability to a market participant?

Yes. If FV is to represent a market value then for a liability it should be a transfer value and not a settlement value.

That said we would expect transfer values to be of very restricted application to the measurement of liabilities. They would not seem appropriate for the measurement of most provisions under IAS37 for instance and may well not be suitable for many claims under insurance contracts either.

Q10. Does the transfer measurement objective for liabilities in SFAS157 differ from FV measurements by IFRS as applied in practice?

We agree with the restricted cases identified in the DP where under IFRS liabilities are to be at FV – IAS39 for traded financial liabilities and derivatives with negative values, contingent liabilities when acquired in a business combination – both of which potentially represent transfers of liabilities. The value, however, of contingent liabilities cannot generally be derived from market values as no such markets exist. In practice these have to be the subjective specific assessments made by the acquirer in the business combination.

Q11. In your view is it appropriate to use a measurement that includes inputs that are not observable in a market as FV at initial recognition even if this differs from the transaction price?

No. We prefer the existing treatment in IAS39 as this would produce more reliable information. As noted above we do not agree that a value based on level 3 information would constitute a market value. The actual transaction price is at least an observable traded price and is therefore more reliable.

Q12. Do you believe that the provisions of SFAS 157, considered in conjunction with the unit of account guidance in IAS39, would result in a portfolio-based valuation of identifiable risks of instruments considered in aggregate, or an in-exchange exit price for the individual instruments?

We note that SFAS157 does not appear to have much to say on this matter, but paragraph 27 would indicate that generally it is requiring an item by item approach. In developing an IFRS on this subject IASB should clarify this matter.

As noted however in our answer to Q1 above we consider that the right ‘unit of account' is perhaps a matter that should be left to individual standards and not set in an overall FV measurement standard. See also our answer to Q20 below.

Q13. Do you agree that a FV measurement should be based on the principal market for the asset, or in the absence of a principal market, the most advantageous market?

Yes. This seems a reasonable approach, where such multiple markets exist.

Q14. Do you agree that a FV measurement should consider attributes specific to an asset or liability that market participants would in pricing an asset or liability?

Yes.

Q15. Do you agree that transaction costs that would be incurred to sell an asset or transfer a liability are an attribute of the transaction and not of the assets or liability?

Yes. We agree that such transaction costs should be excluded from the measurement of FV of assets and liabilities and therefore be expensed when the sale or transfer occurs. Costs needed to get the item to market could be deducted in the valuation model, such as preparation and transport costs, but the selling costs and commission could not.

Certain IFRS require measurements at fair value less cost to sell (in IAS36 for impairment calculations and in IFRS5 for assets held for disposal). This would be a different basis from FV, but might well remain appropriate for such instances.

Q16. Do you agree that the risk of non-performance, including credit risk, should be considered in measuring the FV of a liability?

Yes. There is no doubt that the entity's credit standing, and any special variations attaching to a particular liability (e.g. collateral, security), affect any initial valuation of the liability.

We, however, are unconvinced that changes in a business' credit ratings should affect the measurement of most of its liabilities. We would restrict the concept of FV as set out in SFAS 157 and this DP (that is CMEV) to the few tradable liabilities, such as derivatives with negative values, and measure all other liabilities at either amortised consideration or at the present value of the expected future cash flows to settle it.

Q17. Is it clear that the in-use valuation premise used to measure FV of an asset in SFAS157 is different from value in use (VIU) in IAS36?

Yes. The in-use premise for FV in SFAS157 would be when an asset is valued by a market participant in its present use, possibly in combination with other assets. VIU is different, it is an entity-specific valuation of the future cash flows expected from holding on to the asset.

We would note furthermore that

• we support the continued application of VIU and would not want impairment measured solely by reference to FV less cost to sell, even on an in-use basis.
• the terminology of the two could make for confusion and so some other term should be used by IASB for FV.

Q18 and 19. Do you agree with the FV hierarchy in SFAS157 and are the differences between the levels clear?

Yes. The differences between the levels are based on the information bases of each. These seem clear and very similar for instance to those already incorporated in IAS39

• Quoted prices from an active market for identical items
• Market prices adjusted for different items and markets and applied in a model
• Unobservable inputs

We consider that a FV using solely level 3 unobservable inputs should not qualify as a FV at all, because it fails the market participant part of the definition.

Q20. Do you agree with SFAS157 that a blockage adjustment should be prohibited for financial instruments when there is a price in an active market (level 1)? In addition do you agree that this provision should apply as a principle to all levels of the hierarchy?

We are not convinced by the approach suggested. It would raise problems for example

• in some active, but thin, markets where a holding might be too big to be absorbed without affecting the price severely. The holding might arguably not be realisable in a reasonable time-frame at the FV in the balance sheet.
• Equally, large holdings may gain extra value – blocking stakes etc. So a unit price times quantity may understate investments in associates that may also be listed companies.

It seems wrong in principle to ignore these concerns. There may need to be provisos to downgrade inputs to level 2 where the market is not deep and liquid. The associates example might indicate that a consistent definition of fair value for all standards is not desirable.

Q21. Do you agree that FV measurements should be determined using the price within the bid-ask spread that is most representative of FV in the circumstances as in SFAS157?

Yes. SFAS157, however, lacks explanation of how to determine the price most representative of FV within the spread. If FV is to be more specifically CMEV then the exit price should be the bid price for assets as in IAS39. Bid-ask spreads could also represent both transaction costs and dealers' margins and in principle the transaction costs should be excluded. Any guidance in IFRS might point preparers in the direction of the existing IAS39 pricing convention, as achieving clarity and comparability, while noting other factors that could adjust that to some other price.

Q22. Should a pricing convention be allowed even when another price within the bid-ask spread might be more representative of FV?

No. See our answer to Q21 above .

Q23. Should the bid ask pricing guidance apply to all input levels of the hierarchy?

A bid-ask spread on inputs that are not observable in a market is an impractical concept. Level 2 and 3 inputs should probably have in mind the exit value objective in the proposed definition and apply that as far as possible.

Q24. Do the disclosure requirements of SFAS157 provide sufficient information?

No. They would need to be supplemented in particular standards with specific disclosures (for agriculture or insurance for example). Also there would need to be more disclosures about key assumptions when models depend on using level 3 inputs, as is currently the case with some IFRS.

In SFAS157 for non-recurring FV exercises (presumably mostly instances of impairment) the disclosures are not required on a continuing basis. For the revaluation of assets under IAS16 and IAS28, the basis of valuation needs to be disclosed in the years between valuations as this continues to be significant to the financial statements.

Q25 and 26. Is the SFAS157 application guidance sufficient, including for emerging economies?

No. The examples cover financial and non-financial assets for instance, but either in a general fair value IFRS or specific standards using FV/CMEV, the examples and guidance might need to be extended to cover agriculture, investment property and insurance for instance.

IASB should provide further guidance for developing countries or emerging economies, where a key aspect is the issue of a lack of active markets, let alone deep and liquid ones. It would be important to recognise in such guidance that in these countries and economies that the inputs might be predominantly level 3 and therefore in our view ruling these out as FV/CMEV as explained in our answer to Q19 above.

Back to top