This is the third article in a three-part series. For more, read "Leveraging quality control sampling for your business," (Part I of III) and "Using data to manage repurchase risk and loan quality" (Part II of III).
On December 20, 2012, the Financial Accounting Standards Board (FASB) issued a proposed Accounting Standards Update (ASU) that discusses changes to the ways banks recognize and account for potential credit losses (the ASU is "Financial Instruments—Credit Losses," Subtopic 825-15). A simple summary of the update is that the FASB proposes that banks and other financial institutions modify recognition of impairment from a “probable loss” to a “lifetime of loss” estimate. For mortgages, this means changing the base of the impairment provision from a provision for losses arising from the current delinquency inventory to a provision for all mortgages, recognized at origination. Impairment provisions for delinquent loans are typically estimated using a roll-rate model based on recent experience. This article provides a methodology to estimate credit losses (including losses on loan repurchases) for the lifetime of a loan.
FASB proposal background
Prior to the financial crisis, the United States and international accounting boards began projects to revise and improve accounting standards and treatments. One of the goals of the projects was to harmonize accounting methodologies in response to a global banking system. During the financial crisis, a weakness in the current accounting methodology was highlighted through delayed credit loss recognition. Credit losses were not recorded until the loss was classified as a “probable loss.” Therefore, a bank with a pool of loans that were originated to high-risk borrowers (e.g., with low FICO scores, minimal down payments, and other exotic loan features) would not recognize losses on the pool until the loans actually missed payments, even though available data would indicate future losses would develop on them.
Therefore, the FASB began researching alternative methodologies for accounting for credit losses that would resolve the “delayed recognition” issue. The proposed methodology, termed the current expected credit loss (CECL) model, is to estimate and develop a provision for credit losses at origination for the lifetime of the loan.
Industry responses to the proposal have been mixed. In general, investors welcome additional insight into the exposure of banking institutions to future credit losses. The banking industry also welcomes the guidance and initiative but is quick to point out the difficulty and potential volatility in adopting the proposed methodology. A response from the American Bankers Association (ABA) highlights the following complications with the proposed methodology:
- Implementation may be difficult and require significant investment by banks.
- Current models and databases may not collect the information required to perform these estimates.
- It is inconsistent with the way banks monitor credit exposure.
- It is is subject to a high degree of uncertainty, which may result in large swings in reserve estimates.
- It requires economic and borrower behavior forecasts past a reasonable time frame for high-confidence results.
The ABA agrees with the initiative but thinks the current approach needs refinement before a final ruling should be issued. As there has yet to be a final ruling on the matter, the ABA is currently pushing for an alternative solution, termed the Banking Industry Model (BIM). However, the BIM isn’t the current methodology of choice for the FASB, so we will not discuss it further in this article.
Proposed CECL approach
Adopting the CECL model could have a significant impact on lenders both in terms of technological requirements and in changes in the reserve estimate. The CECL model requires that a reserve be set at origination for expected future losses on every loan. Expected future losses include exposure to both credit losses for loans held on balance sheet and to repurchase losses for loans sold to a third party.
To estimate such a reserve amount, we discuss a frequency/severity approach that is common within the insurance industry. For loans that are held on balance sheet, the estimated reserve amount is calculated as:
CECL reserve = loan amount * default frequency * loss given default
For loans that are sold to government-sponsored enterprises (GSEs) or the Federal Housing Administration (FHA), the estimated reserve amount is generally calculated as:
CECL reserve = loan amount * repurchase/indemnification probability * (default frequency * loss given default + [1 – default frequency * scratch-and-dent market price])
There are several industry tools available that estimate the frequency of a default or a repurchase event. For the CECL reserve, these models should reflect key characteristics of the risk of the loan such as, but not limited to:
- Borrower credit quality
- Borrower attributes and potential for underwriting risk (e.g., self-employed borrower)
- Loan attributes
- Economic consideration
- Loan age
Generally, as a loan ages, the risk of repurchase and default decreases. This should be explicitly modeled in determining the remaining reserve liability.
Please contact Jonathan Glowacki for more information.