The Federal Reserve Bank of Boston has released a working paper by Sanders Shaffer titled Fair Value Accounting: Villain or Innocent Victim: Exploring the Links between Fair Value Accounting, Bank Regulatory Capital and the Recent Financial Crisis:
There is a popular belief that the confluence of bank capital rules and fair value accounting helped trigger the recent financial crisis. The claim is that questionable valuations of long term investments based on prices obtained from illiquid markets created a pro-cyclical effect whereby mark to market adjustments reduced regulatory capital forcing banks to sell off investments which further depressed prices. This ultimately led to bank instability and the credit effects that reached a peak late in 2008. This paper analyzes a sample of large banks to attempt to measure the strength of the link between fair value accounting, regulatory capital rules, pro-cyclicality and financial contagion. The focus is on large banks because they value a significant portion of their balance sheets using fair value. They also hold investment portfolios that contain illiquid assets in large enough volumes to possibly affect the market in a pro-cyclical fashion. The analysis is based on a review of recent historical financial data. The analysis does not reveal a clear link for most banks in the sample, but rather suggests that there may have been other more significant factors putting stress on bank regulatory capital.
After a discussion of Fair Value Accounting and the criticism that has been leveled against it, the author points out:
Fair value is applied to investment securities depending on how they are classified. Investment securities classified as available for sale are measured at fair value each reporting period. The resulting adjustments are termed unrealized gains or losses. These adjustments are recorded in an equity account called Accumulated Other Comprehensive Income. An important point here is that fair value adjustments related to debt securities and unrealized gains on equity securities are excluded when computing Tier 1 regulatory capital.
The author hypothesizes:
This analysis does not address whether raising capital through the sale of investments in a distressed market would be a first choice or last resort. However, one may be able to infer that if banks were actually being forced into distressed sales, they would first try to reduce more discretionary items. Dividends on common stock are discretionary and can be reduced or suspended as a method to maintain capital ratios.
Further, it does not appear that losses were realized in practice:
To summarize, this analysis looked at the largest financial institutions. It then isolated the impacts that critics have linked to capital destruction, namely the application of fair value to banks’ investment portfolios. The analysis shows that the impact on regulatory capital was quite small and does not appear to be large enough to be considered the driver of the pro-cyclical dynamic whereby declining asset prices lead to lower capital, then on to sales of assets to replenish capital, creating further pressure on prices and so on. In addition, there was no evidence found in reported financial data which would be indicative of distressed selling activity during the crisis period of 2008.
So if mark-to-market wasn’t the villain, what was?:
Based on further analysis of 2008 financial results, it was noted that loan loss provision had a significant impact on regulatory capital for most institutions in the sample.
An example is supplied:
At the height of the crisis, State Street stock fell 59 percent in one day when it was announced that unrealized losses had doubled, and analysts noted that TCE was approaching zero based on pro-forma calculations that added in the impact of consolidating certain off-balance sheet investment conduit programs.
The Simple TCE Ratio is calculated as STCE/tangible assets. Tangible assets = total assets – goodwill – intangible assets (excluding Mortgage Servicing Rights). It is not known how much emphasis was placed on TCE versus other significant factors that were also affecting bank stocks at the same time. That being said, State Street and BNYM are two possible examples in this analysis where fair value accounting may have contributed to bank instability based on the significant affect on TCE. It should be noted though that State Street and BNYM did not sell investment assets in response to capital depletion or market stress. They were able to rely on debt and equity issuances as well as participation in government capital programs. So although fair value may have contributed to some instability, the link between fair value and pro-cyclicality did not necessarily come to fruition here, at least partially due to government intervention.
The author concludes:
Based on this simple analysis it would appear that fair value accounting had a minimal impact on the capital of most banks in the sample during the crisis period through the end of 2008. Capital destruction was due to deterioration in loan portfolios and was further depleted by items such as proprietary trading losses and common stock dividends. These are a result of lending practices and the actions of bank management, not accounting rules. Furthermore, the data suggests that banks were not raising significant capital through distressed asset sales; rather they were relying on government programs as well as debt and equity markets.
This paper is of particular interest given the recent BoC Paper on Systemic Capital Requirements and its concern regarding the contagion effect of Asset Fire Sales.
I don’t know where the previous two comments came from…
It seems to me the Boston Fed author is a little casual about the relative ease of cutting dividends vs asset fire sales. I’m sure lots of banks would much rather sell a few assets, even at a loss, before cutting the dividend. Of course, dividends were eventually cut to conserve capital, but perhaps the desire for that radical action crept up more slowly than losses got out of control.
Taking this analysis at face value, it seems to me it would make bank common dividends very volatile in the future.
I don’t know where the previous two comments came from…
Spam – now removed, the IP put on moderation watch and the linked sites blacklisted.
I’m sure lots of banks would much rather sell a few assets, even at a loss, before cutting the dividend.
Me too – with the caveat that it depends on the precise definition of “a few” and “a loss”!
I don’t like the way he slipped in that bit about losses on prop trading in the conclusion – the only time it was ever mentioned.
Taking this analysis at face value, it seems to me it would make bank common dividends very volatile in the future.
Which leads to consideration of a whole ‘nuther tug of war … reserves in the US were depleted in the run-up to the credit crunch because the SEC was concerned they were being used to smooth earnings.