Is applying SCALE to the CECL problem the right thing to do for small banks and credit unions?

CECL allows banks significant flexibility for their choice of method(s) for estimating potential credit losses for capital provisioning. Banks can apply different ‘Expected Credit Loss’ (ECL) methods to specific pools of loans, to optimally incorporate their reflect their credit risk and manage the resultant provision most effectively.

The methodological flexibility in CECL regulation is extremely helpful to banks. However, it can lead to decisioning complexity with the underlying options. To address this, The Federal Reserve created the ‘Scaled CECL Allowance for Loss Estimator’ or SCALE – a simpler spreadsheet-based approach aimed at smaller community banks and credit unions.

This was recently confirmed as an acceptable ECL method, allowing banks to adjust scaling factors according to local conditions. However, this approach does not fully address the complexity and requires banks to fully understand the proxy data being used and the underlying qualitative adjustment factors.

The key to the SCALE approach is that it was designed for smaller community banks with relatively simple portfolios. It was not conceived to be a general ‘go-to’ and, crucially, does not claim to optimize the final CECL provision. While SCALE represents a significant simplification of the CECL computation process for smaller banks, it will likely not create the most efficient provision framework.  

This raises the question as to who exactly should use this CECL methodology created by The Fed. 

If a bank is contemplating running its entire CECL computation on spreadsheets, the SCALE method certainly simplifies this work, making use of publicly available data to create proxy losses. However, bank managements should be cognizant that SCALE is just one of several computation methodologies. They still will have to justify their choices and selections vis-à-vis balance sheet inventory, pooling, and their appropriateness, specifically around the qualitative adjustment factors. These qualifications and justifications for the use of SCALE will need to be created and presented to regulators. 

Banks should also keep in mind that generating a valid CECL result is just the starting point for their CECL computation – a minimum bar to be cleared. Of equal importance is minimizing the institutions initial and ongoing P&L impact.

The most appropriate ECL method should accurately reflect banks’ credit profiles and optimize capital provision, in turn lowering the cost of funding their balance sheets. Selected CECL methods also have to be stress-tested to ensure that the capital provision does not vary excessively across different economic regimes. Reducing the ‘lumpiness’ of the provisions, and therefore the P&Ls, must be a key goal to avoid the need for over buffering the CECL allowance.

In conclusion, if a bank is looking for the most optimum solution for CECL quantification, SCALE becomes just one more method that may be used. For each pool of loans comparisons of SCALE with other methods, including Vintage, WARM, Discounted Cashflow, and PD/LGD should be conducted to ensure the most efficient CECL charge balance the lowest level of future P&L turbulence. 

GreenPoint’s ‘CECL Express” is a streamlined solution to test all computation methodologies   including SCALE. It is designed to enable banks to be prudent and efficient in their computation of CECL charge and optimization of capital and P&L variability.