The U.S. Financial Accounting Standards Board has just announced a proposal to change the way that American banks account for potential loan losses -- that is, losses that have not yet been incurred. See here. Some banks estimate that they would have to increase their reserves by 50%. That would mean that they would have to reduce their reported earnings over a period of a few quarters by a significant amount. The banks are fighting the proposal, of course. They trot out their usual spiel: Make us set aside more reserves and we will lend less -- that is the public bogey man.
Please remember that in 2010 and 2011, the banks recorded profits as they declared that they did not need all the loan loss reserves they had built up in 2008-2009. I commented on that in February. Loan loss reserves always are a source either of pain or of joy in bank earnings reports. As investors, the only thing we know about them for sure is that they are wrong; we do not even know in which direction they are wrong. The proposed changes would make it more likely that the reserves would be overstated than understated, which might be a good thing.
But please also recognize that we are talking here about accounting estimates, not reality. Reality is the stress tests that the Fed administers twice a year. The stress tests leapfrog over the loan loss reserve issue by forcing banks to value their loans based on a specific economic scenario. I explained this in March. Banks' capital requirements now are governed not by their reported balance sheets, but by their stress-tested balanced sheets. Thus, whether or not a bank has to increase its capital, may buy back shares or may pay dividends is no longer based on its reported balance sheet, which now is functionally irrelevant. It is that functionally irrelevant balance sheet that the proposed rule would impact.
What about earnings, you say? Earnings matter, and they are not governed by the stress-tested balance sheet. True enough. But I submit that anyone who invests based on a bank's quarterly earnings is a foolish investor. (One might trade on quarterly earnings, but that is not investing.) Unless you are a great trader, bank stocks should be bought for the long term based on the earning capacity of the bank. That may be difficult to discern for the average investor. But that is what will pay off.
I am disappointed that big American banks fight every effort to make them better. But as investors, the banks' intransigence only should make us wary. A regulated industry that fights efforts to make it improve itself is likely to be over-regulated unless it can capture the regulators. At the moment, regulatory capture is not working. Banks potentially seriously affected by the regulatory backlash include JPMorgan (JPM), Morgan Stanley (MS) and Goldman Sachs (GS). Caveat emptor.
Disclosure: I am long GS.