BNP Paribas has reclassified almost all its peripheral Euro-zone sovereign exposures, taking advantage of an accounting provision brought in after the sub-prime crisis that allows banks to move assets into their loans and receivables portfolio if they are deemed to be illiquid. — Daily News, 2 August 2011
The combination of elective measurement and financial asset impairment rules, under an amortised cost accounting framework, provides an egregious example of how accounting standards can contribute to the disconnection between financial statements’ reported information and economic reality — as exemplified by the European sovereign debt crisis and in the manner in which loss recognition is either delayed or at the discretion of banks known to have exposures to the Euro-periphery states.
There is ample evidence that these banks’ balance sheets are economically impaired, yet the accounting rules allow a deferred recognition of related losses. This is likely contributing to the price-to-book ratios of less than one as many of these banks with sovereign exposure to the Euro-periphery have not yet adjusted the values of their lending exposures.
The current financial instrument impairment model requires losses only to be recognized when evidenced by a specific past or current trigger event (i.e. incurred loss). Under this approach, the impairment recognition involves revising the expected future cash flows, at the occurrence of a trigger event, and discounting these cash flows based on the expected yield at issuance (i.e. what is described as effective interest rate). Unlike a fair value approach, amortised cost impairment does not allow a contemporaneous adjustment of the value of exposures corresponding to signals from the economic environment nor does it enable timely market correction. This begs the question: Could an impairment model based on amortised cost remedy the problem of ‘too-little, too-late’ recognition of losses, which also has pro-cyclical impacts as it can potentially concentrate the recognition of losses to crisis periods?
Is There a Meaningful Way Forward?
Both the IASB and FASB have grappled with the search for a workable amortised cost impairment model that is both operational and would also allow the timely recognition of credit risk losses in a manner that conforms to economic reality. However, the quest for such a model has proven to be an intractable problem. Though a number of refinements have been proposed, these are increasingly looking like futile thought experiments. The gridlock in developing a model in part is due to it being influenced by a hodgepodge of objectives, including accommodating the desire by regulators to have countercyclical provisioning occurring through the financial statements. Thus far on the table, from the IASB and FASB, are:
- an expected loss model by the IASB
- a modified, more forward-looking incurred loss by the FASB
- modified expected loss
- mooting of the concept of differentiated treatment of open portfolio good book and bad book
- a three-bucket partitioning and differentiated impairment
It is also noteworthy that all the conceptualized impairment models are significantly more complex relative to the determination of fair value. Admittedly, even under a fair-value regime, there would still be the challenge of decomposing the credit-risk component from other components of fair value such as liquidity risk or prepayment risk. But this pales in comparison to the difficulties arising from the complex internal calculation and allocation of gains or losses, of an expected loss model, or any other variant of amortised cost impairment. All the mooted impairment models are based on very complex building blocks and difficult-to-ascertain estimates. Examples of such estimates include an estimate of the foreseeable future for purposes of predicting credit-related losses, or having to determine time-proportionate losses. Debate also has focused on whether to immediately recognize catch-up adjustments or to smooth such losses over time. Putting it mildly, it is all a morass of arcane complexity.
Fair Value Is Superior to Amortized Cost Impairment
In sum, it is increasingly clear that amortised cost and its required impairment approaches is second best relative to fair value. Indeed, compared to fair-value measurement, amortised-cost-impairment approaches are too complex, while failing to provide comparable and relevant economic information. Furthermore, it simply grants financial institutions’ managers discretion on the timing of their recognition of losses.
In addition there is compounded complexity arising from the impairment approach overlaying the mixed-measurement classification and measurement approaches, and this will lead to very complex decisions when accounting for financial instruments. This is notwithstanding the intended goal of reducing financial instrument accounting complexity by the standard setters.