I entered banking almost three decades ago. My mentors then followed a 10-to-4 work schedule with a two-hour lunch in between. They were outstanding bankers, but they didn't keep too many loan files. Their greatest strength was intuition. For a moment, go back to the 1990s and pretend you are a regulator, and you walk into a bank asking for a file on any loan, and the folder is empty. The bank would say, “Oh, that loan was made by our star banker and this is a good loan.” This would actually happen.
That star banker may have had a pretty good gut feeling about a borrower. But based on past crises, was our industry better off with that gut feeling and a bunch of empty file folders? Even if good bankers can rely on intuition, we still need something to quantify that intuition. That is why I am strongly in favor of the Financial Accounting Standards Board's Current Expected Credit Loss model. And, yes, I know I am in the minority.
I believe the CECL initiative will, to a small degree, help to bring a positive change by providing a simple and methodical forward-looking tool to accompany gut feeling.
Since the crisis, how to account for troubled loans has been a significant item on the agenda for both FASB and international accounting regulators. Indeed much of the carnage from the 2008 meltdown was attributed to institutions delaying for too long the recognition of credit losses. The basic concept of the CECL model is that waiting until losses from a loan are probable to add to reserves is insufficient. The accounting board's proposal would essentially require bankers to make an accounting projection for losses as soon as a loan is made.
Many bankers are focused on one side of CECL: the cost of implementation. But I believe the industry is not recognizing the potential benefits, in terms of easing the effect of losses and enhancing overall performance, of having a more precise measure of credit losses. That is what the proposed FASB model really tries to achieve.
But I also see room for compromise, and a potential opening for FASB to lighten banks' regulatory burden.
The accounting board has stated that banks would have ready sources of data to build more forward-looking models. This is true. For example, a preliminary CECL model could already be built rather soon using data related to loss projections from the Federal Deposit Insurance. I have also seen an excellent 2016 economic forecast from Nathan Kauffman of the Federal Reserve Bank of Kansas City, which may be considered an unbiased input for modeling.
Yet FASB could also consider delaying the release of the final CECL rule to allow for more modeling data on 2016 to become available. That would only expand the resources at banks' disposal to comply with the new accounting standard. I have asked the FDIC to add discussion of CECL models to the agenda for the next meeting of the agency's Advisory Committee on Community Banking. The federal regulatory agency could even consider releasing potential examples of models that banks could use. Delaying the final rule until more data is available for banks to start building their own models would be a morale boost for the industry.
There are other reasons to think that complying with the CECL model will not be as burdensome as institutions think. For one thing, banks have already been looking at their information systems and data modeling since the crisis, meaning institutions will not be starting from scratch in adapting to the new FASB standard. And past implementation of the Sarbanes-Oxley law is another cause for optimism. The first to comply with SOX had a tremendous expense. But the later adopters had the benefit of simple and more cost effective solutions. I expect there would be a similar cost experience with implementing CECL.
As an industry can we yell, scream, run in circles and panic. Or, we can try something different. Even if it's just a draft, with the first quarter of 2016, we can try modeling now.
Tim Alexander is a managing director and economist at Triune Global Financial Services, a consulting firm for banks and regulators.