On June 16, the Financial Accounting Standards Board issued its long-awaited Current Expected Credit Loss impairment standard, or CECL. The standard, which requires "life of loan" estimates of losses to be recorded for loans at origination or purchase, has drawn significant concern from community bankers over the new rule's complexity and implementation costs. Regulators, on the other hand, have indicated their intent to reasonably apply the accounting standard based on the size of the institution and in consideration of the nature, scope and risk of the traditional lending activities of community banks.
Should bankers believe the regulatory assertions?
One way to predict what mayhappen under a new standard is to analyze the extent to which regulators tailored implementation of existing accounting standards for smaller institutions, and then to extrapolate from there. With this in mind, economists at the Federal Reserve Bank of St. Louis conducted empirical tests to determine how, under generally accepted accounting principles, regulators have evaluated the methodologies used by banks to manage their allowances for loan losses.
Using data from thousands of commercial banks in varying size categories over the last 20 years, the St. Louis Fed economists applied the regulatory expectation that changes in the level of the reserve for loan and lease losses should "be directionally consistent with" concurrent changes in credit quality. This was based on the 2006 Interagency Policy Statement on the Allowance for Loan and Lease Losses. That guidance stated that changes in the level of the allowance should be "directionally consistent" with changes in factors that, taken as a whole, evidence credit losses while also keeping in mind the characteristics of an institution's loan portfolio. For example, if declining credit quality trends relevant to the types of loans in an institution's portfolio are evident, the allowance as a percentage of the portfolio should generally increase.
The study found that annual changes in allowances for smaller banks, relative to larger banks, are less closely correlated with concurrent changes in nonaccrual loans. This is consistent with the idea that, while regulators expect all institutions to comply with GAAP, regulators have scaled their expectations, allowing smaller banks to operate within a broad "reasonable range of probable credit losses" as described by FASB. The study authors used the level of nonaccrual loans as a benchmark for changes in factors that evidence credit losses. Nonaccrual loans can generally be described as those with observable and significant deterioration in the financial condition of the borrower. They were chosen as a metric, in part, because they are simple to understand, intuitive and representative of other benchmarks used in the industry.
In addition to this empirical test, qualitative evidence exists that regulators currently scale their expectations regarding the sophistication and complexity of methodologies utilized at banks of different sizes. Larger more complex banks routinely use methods such as probability of default and loss given default, while smaller community banks utilize much simpler methods such as loss rate multiplied by principal balance. Regulators have not in the past required smaller community banks to switch methodologies to these more sophisticated and complex methods, and we do not expect to in the future.
To be sure, assuming that past approaches will be carried into the future is a speculative exercise. And the methodology used by St. Louis Fed economists to make the empirical connection is subject to potentially important caveats. Nonetheless, the main conclusion of the study is encouraging. Quantifiable and qualitative evidence of historical scaling in approach may offer some encouragement to bankers as we begin this important transition.
Julie Stackhouse is the executive vice president and managing officer of Banking Supervision, Credit, Community Development and Learning Innovation for the Federal Reserve Bank of St. Louis.