COMMUNITY FINANCIAL SYSTEM, INC. (CBU)

Item 2. Management’s Discussion and Analysis of Financial Condition and Results of Operations

​

This Management’s Discussion and Analysis of Financial Condition and Results of Operations (“MD&A”) primarily reviews the financial condition and results of operations of Community Financial System, Inc. (the “Company” or “CFSI”) as of and for the three and six months ended June 30, 2026 and 2025, although in some circumstances the first quarter of 2026 is also discussed in order to more fully explain recent trends. The following discussion and analysis should be read in conjunction with the Company's Consolidated Financial Statements and related notes that appear on pages 3 through 40. All references in the discussion of the financial condition and results of operations refer to the consolidated position and results of the Company and its subsidiaries taken as a whole. Unless otherwise noted, the term “this year” and equivalent terms refers to results in calendar year 2026, “last year” and equivalent terms refer to calendar year 2025, “second quarter” refers to the three months ended June 30, 2026, “YTD” refers to the six months ended June 30, 2026 and earnings per share (“EPS”) figures refer to diluted EPS.

​

This MD&A contains certain forward-looking statements with respect to the financial condition, results of operations, and business of the Company. These forward-looking statements involve certain risks and uncertainties. Factors that may cause actual results to differ materially from those contemplated by such forward-looking statements are set herein under the caption “Forward-Looking Statements” on page 65.

​

Critical Accounting Policies and Estimates

​

As a result of the complex and dynamic nature of the Company’s business, management must exercise judgment in selecting and applying the most appropriate accounting policies for its various areas of operations. The policy decision process not only ensures compliance with current accounting principles generally accepted in the United States of America (“GAAP”) but also reflects management’s discretion with regard to choosing the most suitable methodology for reporting the Company’s financial performance. It is management’s opinion that the accounting estimates covering certain aspects of the business have more significance than others due to the relative importance of those areas to overall performance, or the level of subjectivity in the selection process. These estimates affect the reported amounts of assets and liabilities as well as disclosures of revenues and expenses during the reporting period. Actual results could meaningfully differ from these estimates. Management considers its critical accounting estimates those that involve a significant level of estimation uncertainty and have had or are reasonably likely to have a material impact on the Company’s financial condition or results of operations. Management believes that the critical accounting estimates include the allowance for credit losses; actuarial assumptions associated with the pension, post-retirement and other employee benefit plans; and the carrying value of goodwill and other intangible assets. A summary of the critical accounting policies and estimates used by management is disclosed in the MD&A on pages 38-40 of the most recent Form 10-K (fiscal year ended December 31, 2025) filed with the Securities and Exchange Commission (“SEC”) on February 27, 2026. There have been no material changes other than those described below regarding the Allowance for Credit Losses. A summary of new accounting policies used by management is disclosed in Note C, “Accounting Policies” on page 12 of this Form 10-Q.

41

Table of Contents

Allowance for Credit Losses

​

The allowance for credit losses (“ACL”) represents management’s judgment of an estimated amount of lifetime losses on outstanding loans at the balance sheet date. This is estimated using relevant available information from internal and external sources relating to past events, current conditions and reasonable and supportable forecasts. The determination of the appropriateness of the ACL is complex and applies significant and highly subjective estimates. The ACL is measured on a collective (pooled) basis for loan segments that share similar risk characteristics, including collateral type, credit ratings/scores, size, duration, interest rate structure, origination vintage and payment structure. The Company utilizes three methods for calculating the ACL: cumulative loss, vintage loss and line loss. Historical credit loss experience provides the basis for the estimation of expected future credit losses in all three methodologies. Qualitative adjustments are made for differences in portfolio risk characteristics such as differences in underwriting standards, portfolio mix, acquisition status, current levels of delinquencies, net charge-offs and risk ratings, as well as actual and forecasted macroeconomic variables. Macroeconomic data includes unemployment rates, changes in collateral values such as home prices, commercial real estate prices including office property-specific price forecasts, office property-specific vacancy rates, automobile prices, gross domestic product, and median household income net of inflation. Management utilizes judgment in determining and applying the qualitative factors and weighting the economic scenarios used, which include baseline, upside and downside forecasts. During the first quarter of 2026, the Company updated the ACL model to add 2025 data into the historical data used for calculating the quantitative and qualitative factors as part of the annual model update procedures. With the update, the quantitative reserve in the first quarter of 2026 now includes loss history for business loans that previously was not captured in the historical quantitative loss data and was instead addressed through the use of qualitative overlays. As a result, the Company decreased the additional qualitative reserve for business loans related to size and volume of the loans in that portfolio during the first quarter of 2026 and again in the second quarter of 2026, as other components of the model were determined to have adequately captured the risk associated with these loans. The decreases in the business lending qualitative factor decreased the ACL by $7.5 million as compared to the prior factor in use at December 31, 2025. The update to the historical quantitative loss data increased the ACL by $3.5 million as compared to the quantitative loss rates in use at December 31, 2025.

​

One of the most significant estimates and judgments influencing the results of the ACL calculation is the macroeconomic forecasts. Changes in these economic forecasts could significantly affect the estimated expected credit losses and lead to materially different amounts from one period to the next. To illustrate the sensitivity of the ACL calculation to these economic forecasts, management performed a hypothetical sensitivity analysis using a weighting of 100% to the downside forecast, rather than the existing weighting of baseline, upside and downside of 40%, 20% and 40%, respectively. The scenario-weighted average unemployment rate and GDP growth forecasts used in the ACL model at June 30, 2026 were 5.4% and 1.3%, respectively, compared to 5.5% and 1.3%, respectively, at December 31, 2025. The hypothetical downside forecast includes assumptions of a weakening economy represented by a cumulative decline in real GDP of 2.6%, enhanced geopolitical tensions, rising energy prices, a peak unemployment rate of 8.5% and an average unemployment rate of 7.1%. The Company calculated that this hypothetical scenario would increase the ACL and provision for credit losses as of and for the three and six months ended June 30, 2026 by approximately $4.5 million, and decrease net income by $3.4 million (net of tax). This change is reflective of the sensitivity of the various economic factors used in the ACL model. The resulting difference is not intended to represent an expected increase in allowance levels, as future conditions are uncertain and there are several other quantitative and qualitative factors that will also fluctuate at the same time that economic conditions are changing, which would affect the results of the ACL calculation. The impact that the economic factors have on the model is affected by the upside or downside severity of the scenarios used, the product type mix, and the interaction of the economic factors with other quantitative and qualitative factors in the model, as changes in any particular factor or input may not occur at the same rate or be directionally consistent across all loan segments. Improvements in one factor may offset deterioration in other factors, both qualitative and quantitative. The third-party downside economic forecast used in the hypothetical scenario described does not predict a severe economic downturn, but rather a moderate recessionary environment. The Company’s geographic distribution of loans being primarily outside of major metropolitan areas, combined with low statistical correlation between its historical losses and national economic indicators, is reflected in the current methodology that would produce changes to the allowance that are less significant as compared to economic metric-based modeling that is more directly correlated, and therefore sensitive, to fluctuations in historical and projected national economic activity.

42

Table of Contents

Supplemental Reporting of Non-GAAP Results of Operations

​

The Company also provides supplemental reporting of its results on an “operating” or “tangible” basis. Results on an “operating” basis exclude the after-tax effects of acquisition expenses, acquisition-related contingent consideration adjustments, litigation accrual, gain (loss) on equity securities, restructuring expenses and amortization of intangible assets. Results on a “tangible” basis exclude goodwill and intangible asset balances, net of accumulated amortization and applicable deferred tax amounts. Although these items are non-GAAP measures, the Company’s management believes this information helps investors and analysts measure underlying core performance and improves comparability to other organizations that have not engaged in acquisition or restructuring activities. In addition, the Company provides supplemental reporting for “operating pre-tax, pre-provision net revenues,” which excludes the provision for credit losses, acquisition expenses, acquisition-related contingent consideration adjustments, litigation accrual, gain (loss) on equity securities, restructuring expenses and amortization of intangible assets from income before income taxes. Although operating pre-tax, pre-provision net revenue is a non-GAAP measure, the Company’s management believes this information helps investors and analysts measure and compare the Company’s performance through a credit cycle by excluding the volatility in the provision for credit losses associated with Current Expected Credit Loss (“CECL”) allowance methods, helps investors and analysts measure underlying core performance and improves comparability to other organizations that have not engaged in acquisition or restructuring activities. The Company also provides supplemental reporting of its interest income, net interest income and net interest margin on a fully tax-equivalent (“FTE”) basis, which includes an adjustment to interest income and net interest income that represents taxes that would have been paid had nontaxable investment securities and loans been taxable. Although fully tax-equivalent interest income, net interest income and net interest margin are non-GAAP measures, the Company’s management believes this information helps enhance comparability of the performance of earning assets that have different tax profiles. Reconciliations of GAAP amounts with corresponding non-GAAP amounts are presented in

[Excerpt truncated for page length; source filing is linked above.]

Item 7. Management’s Discussion and Analysis of Financial Condition and Results of Operations

​

This Management’s Discussion and Analysis of Financial Condition and Results of Operations (“MD&A”) primarily reviews the financial condition and results of operations of the Company for the past two years, although in some circumstances a period longer than two years is covered in order to comply with SEC disclosure requirements or to more fully explain long-term trends. The following discussion and analysis should be read in conjunction with the Company’s Consolidated Financial Statements and related notes that appear on pages 77 through 148. All references in the discussion to the financial condition and results of operations refer to the consolidated position and results of the Company and its subsidiaries taken as a whole.

​

Unless otherwise noted, all earnings per share (“EPS”) figures disclosed in the MD&A refer to diluted EPS. The term “this year” and equivalent terms refer to results in calendar year 2025, “last year” and equivalent terms refer to calendar year 2024, and all references to income statement results correspond to full-year activity unless otherwise noted. For a discussion of 2024 results as compared with 2023 results, see Part II, Item 7, “Management’s Discussion and Analysis of Financial Condition and Results of Operations” in the Annual Report on Form 10-K for the year ended December 31, 2024.

​

This MD&A contains certain forward-looking statements with respect to the financial condition, results of operations, and business of the Company. These forward-looking statements involve certain risks and uncertainties. Factors that may cause actual results to differ materially from those contemplated by such forward-looking statements are provided under the caption “Forward-Looking Statements” beginning on page 70.

​

​

37

Table of Contents

Critical Accounting Policies and Estimates

​

As a result of the complex and dynamic nature of the Company’s business, management must exercise judgment in selecting and applying the most appropriate accounting policies for its various areas of operations. The policy decision process not only ensures compliance with the current accounting principles generally accepted in the United States of America (“GAAP”), but also reflects management’s discretion with regard to choosing the most suitable methodology for reporting the Company’s financial performance. It is management’s opinion that the accounting estimates covering certain aspects of the business have more significance than others due to the relative importance of those areas to overall performance, or the level of subjectivity in the selection process. These estimates affect the reported amounts of assets and liabilities as well as disclosures of revenues and expenses during the reporting period. Actual results could meaningfully differ from these estimates. Management considers its critical accounting estimates those that involve a significant level of estimation uncertainty and have had or are reasonably likely to have a material impact on the Company’s financial condition or results of operations. Management believes that the critical accounting estimates include the allowance for credit losses; actuarial assumptions associated with the pension, post-retirement, and other employee benefit plans; and the carrying value of goodwill and other intangible assets. A summary of the significant accounting policies used by management is disclosed in Note A, “Summary of Significant Accounting Policies,” starting on page 83.

​

Allowance for Credit Losses

​

The allowance for credit losses (“ACL”) represents management’s judgment of an estimated amount of lifetime losses on outstanding loans at the balance sheet date. This is estimated using relevant available information from internal and external sources relating to past events, current conditions, and reasonable and supportable forecasts. The determination of the appropriateness of the ACL is complex and applies significant and highly subjective estimates. The ACL is measured on a collective (pooled) basis for loan segments that share similar risk characteristics, including collateral type, credit ratings/scores, size, duration, interest rate structure, origination vintage, and payment structure. The Company utilizes three methods for calculating the ACL: cumulative loss, vintage loss, and line loss. Historical credit loss experience provides the basis for the estimation of expected future credit losses in all three methodologies. Qualitative adjustments are made for differences in portfolio risk characteristics such as differences in underwriting standards, portfolio mix, acquisition status, current levels of delinquencies, net charge-offs, and risk ratings, as well as actual and forecasted macroeconomic variables. Macroeconomic data includes unemployment rates, changes in collateral values such as home prices, commercial real estate prices including office property-specific price forecasts, office property-specific vacancy rates, automobile prices, gross domestic product, and median household income net of inflation. Management utilizes judgment in determining and applying the qualitative factors and weighting the economic scenarios used, which include baseline, upside, and downside forecasts. During 2025, the Company updated the ACL model to add 2024 loss history results into the historical data used for calculating the quantitative and qualitative factors as part of the annual model update procedures; adjusted the weighting of the three economic scenarios, increasing the weight for the downside scenario from 30% to 40% and decreasing the weight of the upside scenario from 30% to 20%, to capture additional economic uncertainty; and increased reserves for business lending to capture additional risk in the portfolio due to the increase in larger individual exposures in the business lending portfolio. The change in weighting of the economic scenarios increased the ACL by $0.7 million as compared to the prior weighting in use at December 31, 2024. The additional business lending qualitative factor increased the ACL by $9.3 million as compared to the prior factor in use at December 31, 2024.

​

​

38

Table of Contents

One of the most significant estimates and judgments influencing the results of the ACL calculation is the macroeconomic forecasts. Changes in these economic forecasts could significantly affect the estimated expected credit losses and lead to materially different amounts from one period to the next. To illustrate the sensitivity of the ACL calculation to these economic forecasts, management performed a hypothetical sensitivity analysis using a weighting of 100% to the downside forecast, rather than the existing weighting of baseline, upside, and downside of 40%, 20%, and 40%, respectively. The scenario-weighted average unemployment rate and gross domestic product (“GDP”) growth forecasts used in the ACL model at December 31, 2025 were 5.5% and 1.3%, respectively, compared to 4.8% and 1.8% at December 31, 2024, respectively. The hypothetical downside forecast includes assumptions of a weakening economy represented by a cumulative decline in real GDP of 2.6%, enhanced geopolitical tensions, elevated inflation, a peak unemployment rate of 8.4% and an average unemployment rate of 7.1%. The Company calculated that this hypothetical scenario would increase the ACL and provision for credit losses as of and for the year ended December 31, 2025 by approximately $3.8 million, and decrease net income by $2.8 million (net of tax). This change is reflective of the sensitivity of the various economic factors used in the ACL model. The resulting difference is not intended to represent an expected increase in allowance levels, as future conditions are uncertain and there are several other quantitative and qualitative factors that will also fluctuate at the same time that economic conditions are changing, which would affect the results of the ACL calculation. The impact that the economic factors have on the model is affected by the upside or downside severity of the scenarios used, the product type mix, and the interaction of the economic factors with other quantitative and qualitative factors in the model, as changes in any particular factor or input may not occur at the same rate or be directionally consistent across all loan segments. Improvements in one factor may offset deterioration in other factors, both qualitative and quantitative. The third-party downside economic forecast used in the hypothetical scenario described does not predict a severe economic downturn, but rather a moderate recessionary environment. The Company’s geographic distribution of loans being primarily outside of major metropolitan areas, combined with low statistical correlation between its historical losses and national economic indicators, is reflected in the current methodology that would produce changes to the allowance that are less significant as compared to economic metric-based modeling that is more directly correlated, and therefore sensitive, to fluctuations in historical and projected national economic activity. Further details regarding the methodologies applied to estimate the various components of the ACL are provided in Note A, “Summary of Significant Accounting Policies,” starting on page 83.

​

Pension, Post-Retirement and Other Employee Benefit Plans

​

The Company provides a qualified defined benefit pension to eligible employees and retirees, other post-retirement health and life insurance benefits to certain retirees, an unfunded supplemental pension plan for certain key executives and an unfunded stock balance plan for certain of its nonemployee directors. The benefit obligations for the pension and post-retirement benefits plans require significant management judgment. The assumptions used in calculating the benefit obligation include the discount rate, expected return on plan assets, rate of compensation increase and interest crediting rates. The discount rate was determined based upon the yield on high-quality fixed income investments expected to be available during the period to maturity of the pension benefits. The expected long-term rate of return was estimated by taking into consideration asset allocation, long-term capital market assumptions, reviewing historical returns on the type of assets held and current economic factors. Mortality tables are also utilized in calculating the benefit obligation, the selection of which is based on management judgment. The Company analyzed the sensitivity of the discount rate and the expected long-term rate of return on plan assets on the pension benefit obligation and net periodic pension cost. At December 31, 2025, a decrease in the discount rate of 100 basis points would increase the pension benefit obligation by $11.3 million, while an increase in the discount rate of 100 basis points would decrease the pension benefit obligation by $9.5 million. For the year ended December 31, 2025, a decrease in the discount rate of 100 basis points would reduce the net periodic pension income by $0.7 million, while an increase in the discount rate of 100 basis points would decrease the net periodic pension income by $0.2 million. A decrease in the expected long-term rate of return on plan assets of 100 basis points would reduce the net periodic pension income by $2.7 million, while an increase of 100 basis points would increase net periodic pension income by $2.7 million. Further detail on the assumptions used and a comparison between 2025 and 2024 assumptions is included in Note K, “Pension and Other Benefit Plans”, starting on page 119.

​

​

39

Table of Contents

Goodwill and Other Intangible Assets

​

The initial carrying value of goodwill is impacted by the initial carrying value of intangible assets including core deposit intangibles, customer relationship intangibles

[Excerpt truncated for page length; the complete text is on the linked full-MD&A page.]