The political pressure resulting from the financial crisis, transitioned to an investor information argument and now a bank-saving argument, was a dominant driver of the new standard. In December, the FASB Chairman hinted that FASB would not consider exemptions to CECL (speech attached), since 168 community banks had failed during the financial crisis (2007 to 2009). From this statistic, he indicts community banks stating that “clearly, community banks have been a major part of the problem.” This is quite a leap that reflects posturing against opposition instead of rational conclusion, which will likely increase animosity in the February public roundtable. Community bank losses from the financial crisis, for example, partly due to plummeting real estate collateral values, were a result of the financial crisis, not a cause. He also seems to imply that if these banks had been required to apply CECL then, they would not have failed. This logic reveals a major CECL flaw, which is that people can reliably predict the future. If bankers had been able to reliably predict the financial crisis, certainly they would not have made the loans that turned to losses as borrower cash flows dried up and collateral values collapsed. Obviously they could not, and obviously, people’s ability to reliably predict the future will remain a dangerous illusion. If people could reliably predict the future, would CECL be needed?
Similarly, the FASB Chairman noted in relation to the “true economies” of bank loans that “We’re just trying to get them recognized today before the bank goes “belly up” tomorrow.” He doubles down on the assertion that CECL will prevent bank failures. It will not, but it might hasten or even inappropriately cause them due to the apparently inconceivable notion that predictions of the future might be wrong and cause devastating recognition of losses that never would occur, or conversely that rosy “expectations” for timing and magnitude of improving conditions could deny losses that currently exist.
Incredibly, the FASB Chairman states that “We are not asking banks to change their methods of estimating losses, but we are requiring banks to change the assumptions used.” In other words, if you like your credit loss estimation methodology, you can keep your credit loss estimation methodology. With CECL, FASB obviously requires banks to change their methods of estimating losses, that is what the new standard does.
Clearly FASB is invested and entrenched and has moved to the “sales” phase of standard-setting.
Another CECL illusion is that it moves away from the accounting foundation of recognizing losses (i.e., expenses) when they are incurred. If this were true, then CECL would require banks to recognize losses that have not been incurred (as a side note, existing accounting standards require that a loss/expense be “probable” before it is “incurred” and recognized, which by definition means “likely”). This would mean that the provision for loan loss expense would have a different theoretical foundation than any other expense in the income statement. What CECL actually does is change the definition of what an incurred loan loss is, or at least how it is determined. CECL requires predictions of the future where existing accounting standards preclude predictions of the future. Which method seems more reliable? Which method affords greater latitude to determine current earnings? The FASB Chairman also noted “…we are requiring assumptions only as long as there exists reasonable and supportable data about the future, generally two to three years.” This seems to be misguided misdirection. You don’t really have to predict the future. You just have to make reasonable and supportable assumptions within a model. Then if you inaccurately predicted the future, you were not wrong, you just have to make new assumptions going forward (i.e., a change in accounting estimate handled currently and prospectively). It seems that hedging has become part of the theory of accounting standard setting. That is sad commentary on the state of accounting standard-setting.
Consider the stock market for a moment. Most transactions that occur reflect a different expectation for the future by the buyer and the seller. Two opposing predictions, at least somewhat influenced by emotion. The prevailing prediction is revealed by the direction of the market, but which one is right is determined over time. That is a risk inherent in investing that is assumed by the investor. Should accounting seek to do this for the investor? Are accountants really better at predicting the future than investors? If they are not, will they become purveyors of misinformation? At what cost?
Now consider whether anyone who can reliably predict the future (if that person exists, particularly over time), even for only two or three years, will be spending their time estimating loan losses or building their fortune in the stock market.
CECL makes the gambles bigger. If bankers inaccurately predict the future, then they inappropriately influence the perceived value of their stock by an increased magnitude, at least for those that rely on the financial statements. How will investors (and regulators) react to “oops” when they have incurred inevitable magnified losses?