Spotted on the Financial Accounting Standards Board website:
FASB may have backed off mark-to-market accounting for banking assets, but the US accounting body might still decide to make some significant changes to the way banks provision for loan losses (you know, after that whole financial crisis thing). The above supplementary document was issued jointly with the International Accounting Standards Board, and is also another bid for converging standards between the two.
Specifically, FASB looks to be moving towards the so-called ‘expected loss’ model for bank loan impairments, which IASB has been considering since 2009. This (in contrast to the current ‘incurred loss’ method) would see financials having to forecast credit losses on their loans, and make provisions before there’s even any evidence of impairment. The idea is that loan impairments will be accounted for way in advance, with sufficient (and smooth) provisions made ahead of any losses.
There are, of course, some criticisms of the approach — including that it’s pretty dependent on banks’ expectations of future losses, or their own subjectivity. And to make those forecasts they often use financial models based on historical data (ahem).
FASB, incidentally, wants banks to estimate those loan losses over the “foreseeable future” — which we think means two to three years. IASB is aiming for banks to amortise expected losses into the reserve over the expected life of the portfolio.
Anyway, here’s Barclays’ Jason Goldberg with some comment:
While we are still in the early stages of interpreting these documents, and they are still subject to comment, we agree that accounting standards for loan loss provisioning needs to be reconsidered. Still, we believe this shift could result in the earlier recognition of net charge-offs and could result in banks having higher loan loss reserves than they have had historically. While reserve levels may still be lower than where banks sit today, it could mean that loan loss reserve releases could be less than banks/investors realize, impacting both capital ratios and capital redeployment. Note these higher reserves would come on top of increased capital levels tied to Basel III. At first glance, banks that have relatively lower loan loss reserve ratios, higher amounts of non-accrual/restructured/delinquent loans, and are potentially less further along this loss cycle could be more at risk than those banks at the opposite end. Based on these figures, BAC, HBA, JPM and PNC may screen as better positioned than BBT, CMA, RF and STI under this proposal.
So, says Goldberg, meet the differentiated bank book(s):
The proposal requires the reporting entity to differentiate its treatment of assets based on the certainty of recovery. For assets where the risk management objective is receiving the regular payments, the above approach is appropriate. However, for assets where the recovery of the asset is the main objective, the bank is required to recognize the entire expected loss in the current period. In essence, this creates two portfolios of loans – a “good book” and a “bad book” using two different accounting methods for loss recognition … Still, this introduces a degree of subjectivity as to the determination of when an asset needs to be transitioned into the “bad book”, but the IASB/FASB are comfortable that it is a similar level of subjectivity as in current practice.
FASB’s proposal is open for comment until April.
Related links:
FASB softening stance in bid for convergence, say analysts - Risk
Provisioning for losses the Spanish way – FT Alphaville