Current Expected Credit Loss: Lessons from 2007-2009

Abstract

We use a top-down approach to estimate the amount of credit loss allowances under the current expected credit loss (CECL) methodology during the 2007-2009 financial crisis. The new standard will replace the incurred loss methodology that is used nowadays by banks. We find that CECL would have been highly procyclical had it been in place during the past crisis, amplifying the contraction in bank lending and the severity of the crisis. This procyclicality would have occurred because macroeconomic models (and macroeconomic forecasters) are generally unable to predict turning points in the business cycle. As a result, CECL allowances generated using real-time forecasts of the economy would not have increased significantly until the beginning of 2007. As the problems in the housing sector gained steam in early 2007, credit loss allowances under CECL would have started to rise rapidly and would have caused a sharp decline in banks’ regulatory capital ratios. In addition, the trough in banks’ regulatory capital ratios would have occurred around the time of the failure of Lehman Brothers. Lastly, we estimate bank lending would have fallen by an additional 9 percentage points during 2009 as it would have been very difficult for banks to raise capital.

Keywords: current expected credit loss approach, loan loss provisions, capital requirements, bank lending, procyclicality

JEL Classification: G18, G21, G28