Measurement Bases for financial accounting - Measurement on initial recognition
Comments from ACCA
May 2006
ACCA is pleased to comment on the above Discussion Paper which was considered by ACCA’s Financial Reporting Committee.
Major overall comments
Our overall view is that this paper represents a thorough examination of the possible measurement bases at initial recognition and of the considerations in principle of the usefulness of those different bases. There are, however, significant issues which are not covered adequately by this paper.
There is an inadequate justification for the paper’s key conclusion that market value would be preferable to historical cost in principle.
The paper proposes a hierarchy of measurement bases – market value, fair value from models based on market inputs, current cost (historical cost as a possible surrogate) and other entity specific measures. There is insufficient exploration of the practical consequences of these chosen measurement bases. For example:
- There is no consideration of the implications of moving away from the recording of the actual value of transactions to recording them at notional values - market value or at current cost. Such a move would have a major cost implication in terms of changing existing accounting systems and the potential need to assess these values. The transactional basis is also associated with a stewardship/accountability role for financial statements and rather less with the economic decision-making objective, both of which are referred to in the Framework. It may be necessary to conclude on that part of the Framework before resolving these measurement issues.
- There is no elaboration of the impact of these measures on the reported performance of the entity for example. The paper deals only with initial recognition and so the main impact would be restricted to cases where the fair value of the consideration given is different from the fair value of what is received and the transaction gives rise to an asset or liability and potentially a Day 1 profit or loss. For those transactions that give rise immediately to an income or expense there could still be an effect if the difference between the historical cost actual transaction value and the recognition amount were to be shown separately from the main income or expense. It is not clear what the intention is on this latter issue.
- There might usefully have been some assessment of how these principles and the hierarchy might actually be translated into practice and the variation from current IFRS requirements to help readers assess the impact of what they are being asked to comment upon. The DP might helpfully have given examples of the extent to which various costs, revenues, assets or liabilities might be stated differently under these proposals.
In general terms where market value, current cost and historical cost are all available on initial recognition, one might expect them to give the same answer. The paper should explore further the reasons why differences between them might arise. This might then help to decide on how they would be best accounted for.
The paper is using definitions of fair value and of historical cost that do not coincide exactly with the way those terms are currently being used in IFRS. This makes it harder to comment finally on these proposals as they presumably constitute proposals to change these definitions as well as to choose between measurement bases.
Comments on questions raised
Q1. Possible bases
We agree that these seem a complete list for initial recognition.
Q2. Definitions
We are comfortable with the definition provided of fair value. We have noted above that the emphasis on the existence of a market and only relying on prices or inputs from an active market appears to make this meaning of fair value slightly different from that being used in IFRS which refers to arm’s length and not market transactions.
We do not agree with the definition of historical cost. The DP’s definition of historical cost as the fair value of the consideration at the time of acquisition raises a number of issues. By contrast we support the notion that historical cost means an accumulation of costs perhaps over a period of time as is referred to in paragraph 36(a). We infer that the DP’s intention that historical cost would exclude transaction costs, whereas we consider that these should be included in a cost measure.
Q3. Sources of difference between measures
We agree with the two mentioned. We suspect there may be other sources of difference and, as indicated in our overall remarks above, these deserve more thorough consideration. The effect of fixed price contracts will be a source of difference between current value and a historical cost measure.
Q4. Meaning of market value
See our response to Q2 above.
Q5. Entity specific measurement
The qualities of entity specific measures are well discussed in paragraphs 57 and 58. We note however while advantages and disadvantages are set out it is not clear how the paper then goes on to conclude on the innate superiority of market based measures for initial recognition.
Q6. Market basis to be preferred to entity specific
The conclusions on this question in paragraph 60 which is critical to the whole paper are not adequately explained even by the detail in the table in paragraph 59.
Q7. A single fair value
We accept that in theory this may be right, but in reality it turns out often not be the case in practice. The paper itself alludes to problems where there might be more than one market and where certain transaction costs are built into market prices.
Q8. Valuation includes credit risk
We accept that credit risk is a factor for valuing both assets and liabilities at initial measurement. We do not necessarily thereby accept any extension of this principle to remeasurement and particularly to the effects of changes in credit risk of an entity’s own liabilities.
Q9. Unit of account issues
We accept on both of these issues that the approaches suggested may be pragmatic solutions on initial recognition. We also note that these are cases in effect where the market measures are reverting thereby to essentially entity-specific cost measures.
(a) We agree with the portfolio approach at initial acquisition at least. We note that this issue might not necessarily be able to apply to subsequent remeasurement if for instance there is a partial disposal.
(b) We also agree with the proposals on the level of aggregation for valuation on initial measurement. We note that this is an area little explored by the existing IASB framework at least. Subsequently the levels of aggregation may have to be different for the various bases. For example individual pieces of plant in a factory could be considered separately when assessing net realisable value. For the value in use measurement the level would have to be the factory as a whole to measure the expected future cash flows.
Q10. Choice of markets
We agree that this is an issue and the implications of multiple markets needs more research. The proposal on the best source for market values being the market in which the asset was acquired seems a reasonable, if pragmatic, solution to us.
Q11. Transaction costs
For market values it is not straightforward to distinguish transaction costs between those that can be recovered in the market price and those that cannot. Any market value must depend on a specific agreement about which party bears which transaction costs. Market participants undoubtedly consider the transaction costs when agreeing prices. We would favour fair values including transaction costs.
When determining costs whether historical or current, the case for including transaction costs seems clearer still. Cost is an entry value measure and so should include all elements payable to the various third parties that would have to be incurred to obtain the asset.
Q12. Balance of relevance and reliability
We agree with the concept of reliability as a threshold quality which must be met and then relevance as the key criterion for selecting among the various reliable measures.
Q13. Limitations on measurement reliability
We agree with these.
Q14. Fair value as the most relevant measure on initial recognition
As noted in our overall comments and in our answer to Q6 above, we find that the DP has failed to prove this case. Significant factors are ignored, including the various assumptions about the objective of financial reports whether for accountability or economic decision making. The cost-benefit considerations of moving away from a transactional basis for accounting to reporting notional values have not been considered. It is difficult to decide on the initial recognition question when the subsequent remeasurement is not addressed and how differences between the transaction value and the fair value would be reported. For any items that will be remeasured at fair value, the initial measurement basis is of little practical consequence unless the components of the change from transaction value are reported separately.
Q15. Fair values sometimes not reliably measurable
We agree with the conclusions in this regard, including those on the more specific propositions in (a) and (b) concerning a single transaction price and on measurement models and consistency with market expectations.
Q16. Comparative relevance and reliability of other measurement bases
The DP principally presents these in comparison with fair value. As noted in Q14 above we do not accept as proved at all the premise on which this has been done.
Net realisable value, value in use and deprival value are essentially bases for subsequent rather than initial measurement in our view.
Our comments above in terms of the choice between historical cost and fair value in answer to Q14 apply largely to the relative merits of current cost and historical cost as well. There would seem little point in measuring at current cost at initial recognition and then not doing so at subsequent remeasurement. Such remeasurement would need an implicit capital maintenance model which is not an issue properly dealt with in the IASB framework as yet. The meaning and relevance of performance based on current costs was debated thoroughly in the 1970’s in the UK for one and it did not seem entirely convincing even at a time of significant inflation.
Q17. Other measures as consistent with fair value as possible
As already noted we do not find the case proved for the preference of fair values. We do not agree with the suggested treatment of transaction costs.
Q18. Proposed hierarchy of valuation bases
As noted above we do not find that convincing cases have been made for preferring in practice either fair value or current cost over historical cost for initial recognition.
We agree that there needs to be a fourth level of entity specific measures because of the unavailability of any other measures in some cases.
We would therefore prefer the presumption at initial recognition to be the other way around. Initial recognition should be at historical cost including transaction costs. Where a meaningful cost is not available fair value should be used (for example in barter transactions or share-based payments) or failing that the fourth level measures (for provisions or other liabilities with no initial consideration). This largely represents the existing position in IFRS. This basis for initial recognition would leave entirely open the question of subsequent remeasurement.