Loans and the Allowance for Credit Losses
In June 2016, the FASB issued ASU No. 2016-13, Financial Instruments – Credit Losses (“ASC 326”), which requires the measurement of all expected credit losses for financial assets held at amortized cost to be based on historical experience, current condition, and reasonable and supportable forecasts. The Company adopted this guidance effective January 1, 2023 and recorded a cumulative effect adjustment that increased the allowance for credit losses for loans and loan commitments by $3.0 million, increased deferred tax assets by $777,000 and decreased retained earnings by $2.1 million, net of tax.
The ACL for loans is measured on the loan’s amortized cost basis, excluding interest receivable, and is initially recognized upon origination or purchase of the loans, and subsequently remeasured on a recurring basis. The ACL is recognized as a contra-asset, and credit loss expense is recorded as a provision for credit losses in the consolidated statements of operation. Loan losses are charged-off against the ACL when management believes the loan is uncollectible. Subsequent recoveries, if any, are credited to the ACL. Loans are normally placed on non-accrual when a loan is determined to be impaired or when principal or interest is delinquent for 90 days or more. The Company generally does not recognize an ACL on accrued interest receivables, consistent with its policy to reverse interest income when interest is 90 days or more past due.
The Company also records an ACL on unfunded loan commitments, which is based on the same assumptions as funded loans and also considers the probability of funding. The ACL is recognized as a liability, and credit loss expense is recorded as provision for unfunded loan commitments within the provision for credit losses in the consolidated statements of operation. Upon funding of the loan, any related ACL previously recorded on the unfunded amount is reversed and an ACL is subsequently recognized on the outstanding loan.
To calculate the ACL for loans and loan commitments collectively evaluated, the Company uses models developed by a third party. The CRE, C&I, and Consumer lifetime loss rate models calculate the expected losses over the life of the loan based on exposure at default loan attributes and reasonable, supportable economic forecasts. The exposure at default considers the current unpaid balance, prepayment assumptions, and expected utilization assumptions.
Key assumptions used in the models include portfolio segmentation, prepayments, risk rating, a peer scalar, and the expected utilization of unfunded commitments among others. The portfolios are segmented by loan level attributes such as loan type, loan size, date of origination, and delinquency status to create homogenous loan pools. Pool level metrics are calculated and loss rates are subsequently applied to the pools as the loans have similar characteristics. Prepayment assumptions, if applicable, are embedded within the models and are based on the same data used for model development and incorporate adjustments for reasonable and supportable forecasts. The models employ mean reversion techniques to predict credit losses for loans that are expected to mature beyond the forecast period.
To account for economic uncertainty, the Company uses multiple economic scenarios provided by the models in determining the ACL. The forecasts include various projections based on variables such as, Gross Domestic Product, interest rates, property price indices, and employment measures, among others. The forecasts are probability-weighted based on available information at the time the calculation is conducted. Scenario weightings and model parameters are reviewed for each calculation and are subject to change.
The CRE and CRE lifetime loss rate models were developed using the historical loss experience of all banks in the model’s developmental dataset. Banks in the model’s developmental dataset may have different loss experiences due to geography and portfolio as well as variances in operational and underwriting procedures from the Company, and therefore, the Company calibrates expected losses using a peer scalar function provided by the models. The peer scalar was calculated by examining the loss rates of peer banks that have similar asset bases and that operate in similar markets as the Company and comparing these peer group loss rates to the model results.
The Company also considers qualitative adjustments to expected credit loss estimates for information not already captured in the quantitative loss estimation models. Qualitative factor adjustments may increase or decrease management’s estimate of expected credit losses. Qualitative loss factors are based on the Company’s judgment of market, industry or business specific data, changes in loan composition, performance trends, regulatory changes, uncertainty of macroeconomic forecasts, and other asset specific risk characteristics.