The Need for a CECL Quantitative Impact Study

Executive Summary

Preliminary testing of the current expected credit loss (CECL) estimation models by larger banks indicate that, while CECL should theoretically reduce procyclicality in the banking industry by providing earlier loss recognition, practical application of CECL will likely increase procyclicality. This is primarily due to the inability to forecast turns in the economy, especially in the long-term. The result also particularly impacts the potential cost and availability of credit to consumers and to loans with longer tenors. Both the increased procyclicality and the uneven impact have public policy concerns. Indeed, due to the fact that a change in loan impairment accounting was initially requested by the Financial Stability Board in order to reduce procyclicality, the American Bankers Association (ABA) believes it is unlikely that neither banking regulators nor bank investors would have supported the change to CECL if they thought CECL would do the opposite.

There have been several papers that have estimated the impacts of CECL, the majority of those focusing on how the implementation of CECL would impact bank capital had CECL been in effect prior to the financial crisis. Each of these studies has strengths and can be instructive, particularly those that test allowable CECL techniques that banks may consider deploying. However, they all have critical weaknesses that can include (among other things) limited scope, inaccurate assumptions of current accounting, the lack of analysis of loan-level credit risk metrics at origination, and the unrealistic application of perfect foresight to forecasting the economy and estimating loss given default. How “procyclicality” is assessed can also be questioned, as a study may conclude that procyclicality is decreased by CECL, even though credit loss allowances are forecast to spike higher than they actually did during the financial crisis time period. Bankers believe such spikes would provide further incremental disincentive to lending during an economic downturn, increasing the procyclicality in the industry.

As a result, the banking industry has requested that a quantitative impact study be performed prior to the 2020 implementation of CECL. The study would include the top ten lenders in each major lending product, plus Fannie Mae and Freddie Mac, as well as a range of midsize and community banks, using actual bank modeling, forecasting results as they would likely have occurred during the financial crisis time period (December 2005 through 2010). Including these lenders would cover the majority of loans held in the regulated lending industry. The results can be summarized, extrapolated to an impact on lending as well as on individual institutions throughout the remainder of the industry.

In addition to quantitative impact on capital and lending, surveys of midsize and community banks would be performed to assess the expected costs in implementing CECL on an ongoing basis.

As part of the study, changes to regulatory capital requirements, regulatory guidance, and to the CECL standard itself will be considered to mitigate any identified unintended consequences.