XML 36 R26.htm IDEA: XBRL DOCUMENT v3.23.3
Allowance for Credit Losses and Credit Quality of Loans (Policies)
 9 Months Ended
Sep. 30, 2023
Allowance for Credit Losses and Credit Quality of Loans [Abstract]  
Allowance for Credit Losses
As described in Note 3 Recent Accounting Pronouncements, the Company’s adoption of ASU 2022-02 resulted in an insignificant change to its methodology for estimating the allowance for credit losses on TDRs. The Day 1 decrease in allowance for credit loss on TDR loans relating to adoption of ASU 2022-02 was $0.6 million.

The allowance for credit losses totaled $114.6 million at September 30, 2023, compared to $100.8 million at December 31, 2022. The allowance for credit losses as a percentage of loans was 1.19% at September 30, 2023, compared to 1.24% at December 31, 2022.

During the first quarter of 2023, the Company made adjustments to the class segments within the portfolios to better align risk characteristics and reflect the monitoring and assessment of risks as the portfolios continue to evolve. Paycheck Protection Program was consolidated with Commercial & Industrial, as the portfolio had decreased to less than $1 million and no longer warranted a material class segment. The Other Consumer class segment was further separated into Residential Solar and Other Consumer. The growth in our Residential Solar loans warranted evaluation of this class separately from the Other Consumer class segments. The change to the class segments was applied retrospectively and did not have a significant impact on the allowance for loan losses. The following table illustrates the portfolio and class segments for the Company’s loan portfolio:

Portfolio Segment
Class
Commercial Loans
Commercial & Industrial
 
Commercial Real Estate
Consumer Loans
Auto
 
Residential Solar
 
Other Consumer
Residential Loans

The allowance for credit losses calculation incorporated a 6-quarter forecast period to account for forecast economic conditions under each scenario utilized in the measurement. For periods beyond the 6-quarter forecast, the model reverts to long-term economic conditions over a 4-quarter reversion period on a straight-line basis. The Company considers a baseline, upside and downside economic forecast in measuring the allowance.

The quantitative model as of September 30, 2023 incorporated a baseline economic outlook along with an alternative downside scenario sourced from a reputable third-party to accommodate other potential economic conditions in the model, particularly significant unknowns relating to downside risks as of the measurement date. The baseline outlook projected a relatively low unemployment rate of 3.7%, rising modestly to 4.13% by the end of the forecast period. Northeast GDP’s annualized growth (on a quarterly basis) was expected to start the fourth quarter of 2023 at 3.5%, decreasing slightly to 2.9%, and then increasing to 3.8% by the end of the forecast period. Other utilized economic variables remained stable overall, with outlooks for annualized growth in retail sales improving, business output slightly deteriorating, and housing starts staying relatively unchanged. Key assumptions in the baseline economic outlook include the economy being at full employment in the coming quarters, recent bank failures not being symptomatic of broader issues in the financial system, and the reduction in the Federal Reserve’s balance sheet remaining on autopilot. The alternative downside scenario assumes deteriorated economic conditions from the baseline outlook. Under this scenario, northeast unemployment increases to 5.1% in the fourth quarter of 2023 and rises to a peak of 6.9% in the fourth quarter of 2024. These scenarios and their respective weightings are evaluated at each measurement date and reflect management’s expectations as of September 30, 2023. At September 30, 2023, the weightings were 70%, 0%, and 30% for the baseline, upside and downside economic forecasts, respectively. Qualitative adjustments were made for isolated model limitations related to modeled inputs given abnormally high historical retail sales and business output growth rates in prior quarters along with low unemployment. Additionally, a qualitative adjustment for inflation is applied to adjust for potential model limitations arising from the use of a macroeconomic variable denominated in nominal dollar terms as well as a qualitative adjustment for potential weakness in auto and commercial real estate collateral values. These factors were considered through separate quantitative processes and incorporated into the estimate of current expected credit losses at September 30, 2023.

The quantitative model as of June 30, 2023 incorporated a baseline economic outlook along with an alternative downside scenario sourced from a reputable third-party to accommodate other potential economic conditions in the model, particularly significant unknowns relating to downside risks as of the measurement date. The baseline outlook reflected an unemployment rate environment at pre-coronavirus (“COVID-19”) pandemic levels of 3.8% and increasing gradually to 4.2% by the end of the forecast period. Northeast GDP’s annualized growth (on a quarterly basis) was expected to start the third quarter of 2023 at about 3.6% and reach 4.1% by the end of the forecast period. Other utilized economic variables worsened overall, with outlooks for annualized growth in retail sales and business output declining from the prior quarter and housing starts staying relatively unchanged. Key assumptions in the baseline economic outlook include the economy being at full employment in the coming quarters, recent bank failures not being symptomatic of broader issues in the financial system, and the reduction in the Federal Reserve’s balance sheet remaining on autopilot. The alternative downside scenario assumes deteriorated economic conditions from the baseline outlook. Under this scenario, northeast unemployment increases to 5.2% in the third quarter of 2023 and rises to a peak of 7.0% in the third quarter of 2024. These scenarios and their respective weightings are evaluated at each measurement date and reflect management’s expectations as of June 30, 2023. At June 30, 2023, the weightings were 60%, 0% and 40% for the baseline, upside and downside economic forecasts, respectively. Qualitative adjustments were made for isolated model limitations related to modeled inputs given abnormally high retail sales and business output growth rates in prior quarters along with low unemployment. Additionally, a qualitative adjustment for inflation is applied to adjust for potential model limitations arising from the use of a macroeconomic variable denominated in nominal dollar terms as well as a qualitative adjustment for potential weakness in auto collateral values. These factors were considered through separate quantitative processes and incorporated into the estimate of current expected credit losses at June 30, 2023.

The quantitative model as of December 31, 2022 incorporated a baseline economic outlook along with an alternative downside scenario sourced from a reputable third-party to accommodate other potential economic conditions in the model. At December 31, 2022, the weightings were 50%, 0% and 50% for the baseline, upside and downside economic forecasts, respectively. The baseline outlook reflected an unemployment rate environment initially around pre-COVID-19 levels at 3.9% that increases slightly during the forecast period to 4.0%. Northeast GDP’s annualized growth (on a quarterly basis) was expected to start the first quarter of 2023 at approximately 3.9% and hovering around 4.6% by the end of the forecast period. Other utilized economic variables have generally deteriorated in their respective forecasts, with retail sales and housing starts forecasts declining from the prior year. Key assumptions in the baseline economic outlook included a full employment economy being realized in the near future, continued tapering of the Federal Reserve balance sheet, an increasing yield on ten-year treasury securities, and a gradual decline in global oil prices. The alternative downside scenario assumed deteriorated economic and pandemic related conditions from the baseline outlook. Under this scenario, northeast unemployment rises from 3.9% in the fourth quarter of 2022 to a peak of 6.9% in the first quarter of 2024. These scenarios and their respective weightings are evaluated at each measurement date and reflect management’s expectations as of December 31, 2022. Additional adjustments were made for factors not incorporated in the forecasts or the model, such as loss rate expectations for certain loan pools, considerations for inflation, and recent trends in asset value indices. Additional monitoring for industry concentrations, loan growth, and policy exceptions was also conducted. All these factors were considered through separate quantitative processes and incorporated when applicable into the estimate of current expected credit losses at December 31, 2022.

There were $219.5 million of PCD loans acquired from Salisbury during the three and nine months ended September 30, 2023 which resulted in an allowance for credit losses at acquisition of $5.8 million. There were no loans purchased with credit deterioration during the year ended December 31, 2022. During 2023, the Company purchased $2.4 million of residential loans at a 7.01% premium with a $17 thousand allowance for credit losses recorded for these loans. During 2022, the Company purchased $11.5 million of residential loans at a 1.53% premium and $50.1 million in consumer loans at par with an allowance for credit losses recorded on the purchase date of $3.2 million.

The Company made a policy election to report AIR in the other assets line item on the balance sheet. AIR on loans totaled $32.5 million at September 30, 2023 and $25.0 million at December 31, 2022 and there was no estimated allowance for credit losses related to AIR as of September 30, 2023 and December 31, 2022.

Individually Evaluated Loans

As of September 30, 2023, there was one relationship identified to be evaluated for loss on an individual basis which had an amortized cost basis of $1.9 million, with no allowance for credit loss. As of December 31, 2022, two different relationships were identified to be evaluated for loss on an individual basis, which in aggregate, had an amortized cost basis of $2.4 million, with no allowance for credit loss.

Credit Quality Indicators

The Company has developed an internal loan grading system to evaluate and quantify the Company’s loan portfolio with respect to quality and risk. The system focuses on, among other things, financial strength of borrowers, experience and depth of borrower’s management, primary and secondary sources of repayment, payment history, nature of the business and outlook on particular industries. The internal grading system enables the Company to monitor the quality of the entire loan portfolio on a consistent basis and provide management with an early warning system, enabling timely recognition and response to problem loans and potential problem loans.

Commercial Grading System

For Commercial and Industrial (“C&I”) and Commercial Real Estate (“CRE”) loans, the Company uses a grading system that relies on quantifiable and measurable characteristics when available. This includes comparison of financial strength to available industry averages, comparison of transaction factors (loan terms and conditions) to loan policy and comparison of credit history to stated repayment terms and industry averages. Some grading factors are necessarily more subjective such as economic and industry factors, regulatory environment and management. C&I and CRE loans are graded Doubtful, Substandard, Special Mention and Pass.

Doubtful

A Doubtful loan has a high probability of total or substantial loss, but because of specific pending events that may strengthen the asset, its classification as a loss is deferred. Doubtful borrowers are usually in default, lack adequate liquidity or capital and lack the resources necessary to remain an operating entity. Pending events can include mergers, acquisitions, liquidations, capital injections, the perfection of liens on additional collateral, the valuation of collateral and refinancing. Generally, pending events should be resolved within a relatively short period and the ratings will be adjusted based on the new information. Nonaccrual treatment is required for Doubtful assets because of the high probability of loss.

Substandard

Substandard loans have a high probability of payment default or they have other well-defined weaknesses. They require more intensive supervision by bank management. Substandard loans are generally characterized by current or expected unprofitable operations, inadequate debt service coverage, inadequate liquidity or marginal capitalization. Repayment may depend on collateral or other credit risk mitigants. For some Substandard loans, the likelihood of full collection of interest and principal may be in doubt and those loans should be placed on nonaccrual. Although Substandard assets in the aggregate will have a distinct potential for loss, an individual asset’s loss potential does not have to be distinct for the asset to be rated Substandard.

Special Mention

Special Mention loans have potential weaknesses that may, if not checked or corrected, weaken the asset or inadequately protect the Company’s position at some future date. These loans pose elevated risk, but their weakness does not yet justify a Substandard classification. Borrowers may be experiencing adverse operating trends (i.e., declining revenues or margins) or may be struggling with an ill-proportioned balance sheet (i.e., increasing inventory without an increase in sales, high leverage, and/or tight liquidity). Adverse economic or market conditions, such as interest rate increases or the entry of a new competitor, may also support a Special Mention rating. Although a Special Mention loan has a higher probability of default than a Pass asset, its default is not imminent.

Pass

Loans graded as Pass encompass all loans not graded as Doubtful, Substandard or Special Mention. Pass loans are in compliance with loan covenants and payments are generally made as agreed. Pass loans range from superior quality to fair quality. Pass loans also include any portion of a government guaranteed loan, including Paycheck Protection Program loans.

Consumer and Residential Grading System

Consumer and Residential loans are graded as either Nonperforming or Performing.

Nonperforming

Nonperforming loans are loans that are (1) over 90 days past due and interest is still accruing or (2) on nonaccrual status.

Performing


All loans not meeting any of the above criteria are considered Performing.
 
Allowance for Credit Losses on Off-Balance Sheet Credit Exposures

The allowance for losses on unfunded commitments totaled $4.8 million as of September 30, 2023, which included $0.8 million of acquisition-related provision for unfunded loan commitments, compared to $5.1 million as of December 31, 2022.
Loan Modifications to Borrowers Experiencing Financial Difficulties
Loan Modifications to Borrowers Experiencing Financial Difficulties

As discussed in Note 3 Recent Accounting Pronouncements, the Company’s January 1, 2023 adoption of ASU 2022-02 eliminates the recognition and measurement of TDRs. Upon adoption of this guidance, the Company no longer recognizes an allowance for credit losses for the economic concession granted to a borrower for changes in the timing and amount of contractual cash flows when a loan is restructured. The adoption of ASU 2022-02 resulted in a change to reporting for loan modifications to borrowers experiencing financial difficulties. With the adoption of ASU 2022-02 these modifications required enhanced reporting on the type of modifications granted and the financial magnitude of the concessions granted.

When the Company modifies a loan with financial difficulty, such modifications generally include one or a combination of the following: an extension of the maturity date at a stated rate of interest lower than the current market rate for new debt with similar risk; a change in scheduled payment amount; or principal forgiveness.