Even prior to the global financial crisis, fair-value accounting (FVA) was a sharply divisive issue.
Supporters of fair-value accounting, including investor-focused CFA Institute, have consistently argued that there is no better alternative to FVA — especially given that all investment and lending decisions are made on a fair-value basis. For example, if one wants to renew a residential property mortgage when house prices fall, banks will only consider the fair value of the property. They will not take into account, for instance, whether house prices are likely to recover, or if the property owner does not intend to sell the house.
The viewpoint of those who support fair value is backed by a substantial body of empirical evidence showing that the fair value of financial instruments is priced by capital market participants. In addition, fair value is the only measurement method that, on an ongoing basis, can meaningfully inform readers of financial statements about company risk exposures. Another supporting argument is that FVA can enable a consistent accounting approach across all financial instruments much better than the current mixed measurement approach.
On the other hand, opponents of fair-value accounting were particularly voluble during the subprime mortgage-triggered credit crisis. These critics frequently voiced concerns about issues they considered fair-value related: excessive write-downs, regulatory capital depletion, and procyclical forced sales. However, there is mounting empirical evidence showing that these concerns were overstated. An example of such evidence is a recently published article by Brad A. Badertscher, Jeffrey J. Burks, and Peter D. Easton in the Accounting Review Journal titled “A Convenient Scapegoat: Fair Value Accounting by Commercial Banks during the Financial Crisis.” Further evidence is presented in the 2010 edition of the Journal of Economic Perspectives, in the article “Did Fair Value Accounting Contribute to the Financial Crisis?,” by Christian Laux and Christian Leuz.
There are several strands to the evidence refuting the alleged adverse effects of fair-value accounting. First, there is evidence showing that write-downs were neither excessive nor primarily attributable to FVA. Second, there is evidence showing the minimal impact of FVA on regulatory capital erosion and forced sales.
No Evidence of Excessive Write-downs Due to Fair-Value Accounting
The Laux-Leuz paper lays out several arguments and evidence about the limited impact of FVA on asset write-downs. To begin, fair value through the income statement is applied to a limited extent by a majority of financial institutions. Derivatives and trading assets that apply FVA represent approximately 15 percent of bank holding companies’ assets and are typically higher for investment banks.
Second, even where FVA is applied, banks could have (and did) ignore market prices that were considered distressed. Under fair-value accounting rules, there is discretion to ignore distressed market prices and, as an alternative, apply internal valuation models (i.e. categorise under level 3 fair values). For example, this discretion was significantly exercised by banks in relation to mortgage exposures. There was also an increase in the proportion of level 3 assets during the crisis. Indeed, the Laux-Leuz paper shows an increase from about 7 percent in 2007 to roughly 15 percent in 2009.
Third, the most significant proportion of banking balance sheets consists of loans and leases, typically around 45-60 percent; these are accounted for on an amortised cost basis. The Laux-Leuz paper provides evidence that, as a result, these banks likely had overstated balance sheet values. The inference of banks having overstated book values is derived by comparing three different independent estimates of the loan losses to accounting impairments, revealing that banks were often conservative in their write-downs during the crisis. Taken together, the evidence in the Laux-Leuz paper shows that the bulk of losses that occurred during the crisis were neither fair-value related nor excessive. Hence, the attribution of excessive write-downs to FVA is unwarranted.
No Evidence of Regulatory Capital Erosion and Forced Sales
Regulatory capital impacts arise when losses due to credit risk impact on earnings. Now, a bit of ‘accounting-speak’ is unavoidable to explain why the impact of fair value on regulatory capital is limited. Asset write-downs occur in relation to loans and leases (i.e. typically more than half of the balance sheet), and these write-downs, which are not fair-value related, are typically reflected through the bad debt expense. In the U.S. context, the link between FVA and regulatory capital arises in relation to the asset impairment category described as the ‘other than temporary impairment’ (OTTI). OTTI occurs when fair-value impairment losses are recognized for financial assets where there is neither the intention, nor the ability, to hold securities until the point where the values have recovered. OTTI can be recognised for financial instruments that are classified as ‘available for sale’ (AFS) or ‘held to maturity’ (HTM). In effect, under U.S. accounting rules, the impact of fair-value accounting on net earnings and regulatory capital would mainly arise when there are OTTI impairments related to debt securities classified as AFS.
The Badertscher-Burks-Easton paper provides empirical evidence derived from a sample of 150 U.S. commercial banks, showing that during the period between September 2007 and December 2008 the proportion of OTTI impairment was insignificant compared to the bad debt expense of loans. There were $19 billion losses of OTTI compared to $214 billion of bad debt expense during this period. Correspondingly, the OTTI impact on capital ratio is insignificant relative to the impact of incremental bad debt expense on capital ratio. Stripping out OTTI write-downs, the median capital ratio would have been 10 percent rather than 9.9 percent. In the same vein, without bad debt expense, the capital ratio would have been 10.7 percent. This shows that the capital erosion due to impairments was primarily due to the bad debt expense of loans where FVA was not applied.
The paper also shows that that there is no clearly established relationship between capital ratio and the extent to which forced sales of assets occur. For instance, banks with higher capital ratios sold more debt securities than banks with lower capital ratios. Furthermore, there was no conclusive evidence of increased levels of forced sales for the 150 commercial banks reviewed. These various pieces of evidence reveal that arguments stating that FVA leads to capital erosion and, thereafter, forced sales during crisis periods are faulty assertions.
Contribution of Accounting to Crisis Overstated
A critical examination of banking-related crises through the years shows that the role of accounting as a contributing factor to the latest crisis was overstated. Such an examination will also show that it is the uncertainty and information asymmetries across counterparties about their asset quality that leads to lending aversion, refinancing difficulties, and panic liquidations. In other words, the adverse economic consequences witnessed during the crisis are the by-product of poor economic choices by banks, including poor investments, high leverage, and short-term borrowing.
The subprime fallout would have occurred regardless of the prevailing accounting regime. Still, in light of the emerging evidence of overstated FVA concerns, one cannot help but wonder whether the concerns were nothing more than a red herring. It also leaves one wondering how the overstated FVA concerns may have unduly influenced accounting standard-setter choices whilst reforming financial instrument accounting.