Thanks to financialart for the point at the BCBS’s response to the partial draft of the replacement to IAS 39.

Frankly, though – I am disappointed by their (the BCBS’s) response – which I shall refer to as “the document” from this point on. Overall it is a long wish-list which is in places self-contradictory and makes a couple of evident misunderstandings on the role of IAS 39. For a start, the document makes it look like IAS 39 is just for banks, when it applies to all IFRS reporting entities – and so it has to be able to work for everyone, not just banks.

Perhaps using the old “Good, Bad and Ugly” criteria would be useful – but in reverse order.

The Ugly (the Contradictory or Confusing Parts)

To make one thing clear up front – IAS 39 is an accounting standard, not a prudential standard. Its role is to provide accounting information to external users. Accounting, by its very nature, is (and should be) concerned with facts and (where possible) should avoid as far as possible the use of models to provide accounting information. At times this is not possible (for example in attempting to determine how much of a loan portfolio meets an IBNR classification) but where models must be used they should be very restricted in scope and the accompanying disclosures should be clear. Prudential standards, however, need to be forward looking to a much greater extent as they are designed to answer very different questions.

Accountants are meant to answer questions like “How much was that position worth on balance date?” and “What was the net profit for the year?”. Prudential rules need to be able to answer questions like “What is the probability of this bank failing over the next 12 months?” and “How much liquidity is enough?”. These are very different questions – but the BCBS seems to be trying to have an accounting standard that does both.

The document is also asking (and I agree with it) for a simpler set of standards – they devote a whole section to this request – and then in subsequent sections they ask for what could only be provided through increased complexity. For example, a move from an incurred loss model to a more prudential style one would effectively mandate that every single publisher of financial information using IFRS (i.e. every company in Australia) would need a prudential loss model for their creditors. The new standard would also have to spell out how that was to be done and this would represent an enormous increase in complexity, particularly for smaller banks and non-banks. If this is not to apply to everyone then we would need at least two (differing) impairment models, increasing complexity even further.

The document is also calling for the new Standard to be introduced with a timetable consistent with “financial stability”. This is nonsense. A new standard takes years to write and consult about – this one up until 2012 – so trying to time its introduction is just silly. If it is a better standard it will improve reporting, and so will enhance stability. If it is not, it will hurt and so should not be introduced. Either way, trying to time the end of a multi-year process to enhance “financial stability” is just silly.

I would also add the last sentence of an otherwise Good first section into here – the main point of accounting standards is to provide (as far as possible) “truth and fairness”. Increasing “market confidence” should only come from being respected, not from attempts to tweak the standards to hide the truth.

The Bad

The “tweaks” referred to above are in para 10 of the document, where there is a call to “de-link” the valuation from “certain aspects of income and profit recognition”. The call here is not clear at all and could result (if incorporated in the Standard) a significant role for management judgement in the valuation of difficult to value items. This element will always have to be there, but the sorts of language incorporated here is just dangerous.

Probably the worst misunderstandings in the document are in the area of fair value – with the writer of this letter mixing up the two related (but not identical) concepts of fair and market value. IAS 39 is pretty clear that there are differing ways to measure fair value and only the first (market value in an orderly, liquid market) is really discussed here. The author clearly confuses the two – witness point 4(b) in the document, where the author says that “the new standard should…recognise that fair value is not effective when markets become dislocated or are illiquid”. Ahem – the current standard does that, as does the new draft.

Another Bad bit is directly above – an accounting standard should not “…reflect the need for earlier recognition of loan losses…” than the current standard does – that is a prudential standards’ role. Doing it here would just go back to the old system where management were (to an extent) allowed to put whatever they liked into the provision for doubtful debts. The current method, while at times complex, at least is less susceptible to manipulation.

The Good

To be fair, though, I should recognise the excellent parts of the document. The first section is a combination of motherhood and useful statements (except for the last sentence). The call for more meaningful disclosures is good, and one I have made before. IFRS 7 is both overly complex and too simplistic simultaneously – a good trick at the best of times. The call in the letter to have it revised is timely – the suggestion to make the disclosures in a more standardised format is a good one.

Section 2 and 3 are good – but are contradicted (as covered before) by (inter alia) paras 12 onwards. The rest is broadly good – but mostly just covers areas that are self-evident anyway.

Overall, if this were given to me by a new graduate attempting to write a wishlist for the new IAS 39 I would hand it back with a few complimentary notes on it applauding their efforts. Coming from a major regulator (never mind the peak banking body) I can only

shake my head.