ALLSTATE CORP

The preparation of financial statements in conformity with GAAP requires management to adopt accounting policies and make estimates and assumptions that affect amounts reported in the consolidated financial statements. The most critical estimates, presented in the order they appear in the Consolidated Statements of Financial Position, include those used in determining:

•Fair value of financial assets

•Impairment of fixed income securities with credit losses

•Evaluation of goodwill

•Reserve for property and casualty insurance claims and claims expense estimation

•Pension and other postretirement plans net costs and assumptions

In making these determinations, management makes subjective and complex judgments that frequently require estimates about matters that are inherently uncertain. Many of these policies, estimates and related judgments are common in the insurance and financial services industries; others are specific to our businesses and operations. It is reasonably likely that changes in these estimates could occur from period to period and result in a material impact on our consolidated financial statements.

A summary of each of these critical accounting estimates follows. For a more detailed discussion of the effect of these estimates on our consolidated financial statements, and the judgments and assumptions related to these estimates, see the referenced sections of this document. For a more detailed summary of our significant accounting policies, see the notes to the consolidated financial statements.

Fair value of financial assets Fair value is defined as the price that would be received to sell an asset or paid to transfer a liability in an orderly transaction between market participants at the measurement date. We are responsible for the determination of fair value of financial assets and the supporting assumptions and methodologies. We use independent third-party valuation service providers, broker quotes and internal pricing methods to determine fair values. We utilize only one single quote or price to value each financial instrument in our financial statements.

Valuation service providers typically obtain data about market transactions and other key valuation model inputs from multiple sources and, through the use of proprietary models, produce valuation information in the form of a single fair value for individual fixed income and other securities for which a fair value has been requested under the terms of our agreements. The inputs used by the valuation service providers include, but are not limited to, market prices from recently completed transactions and transactions of comparable securities, interest rate yield curves, credit spreads, liquidity spreads, currency rates, and

other information, as applicable. Credit and liquidity spreads are typically implied from completed transactions and transactions of comparable securities. Valuation service providers also use proprietary discounted cash flow models that are widely accepted in the financial services industry and similar to those used by other market participants to value the same financial instruments. The valuation models take into account, among other things, market observable information (as described above) as of the measurement date, as well as the specific attributes of the security being valued including its term, interest rate, credit rating, industry sector, and where applicable, collateral quality and other issue or issuer specific information. Executing valuation models effectively requires seasoned professional judgment and experience. For certain equity securities, valuation service providers provide market quotations for completed transactions on the measurement date. In cases where market transactions or other market observable data is limited, the extent to which judgment is applied varies inversely with the availability of market observable information.

For certain of our financial assets measured at fair value, where our valuation service providers cannot provide fair value determinations, we obtain a single non-binding price quote from a broker familiar with the security who, similar to our valuation service providers, may consider transactions or activity in similar securities among other information. The brokers providing price quotes are generally from the brokerage divisions of financial institutions with market making, underwriting and distribution expertise regarding the security subject to valuation.

The fair value of certain financial assets, including privately placed corporate fixed income securities and free-standing derivatives, for which our valuation service providers or brokers do not provide fair value determinations, is developed using valuation methods and models widely accepted in the financial services industry. Our internal pricing methods are primarily based on models using discounted cash flow methodologies that develop a single best estimate of fair value. Our models generally incorporate inputs that we believe are representative of inputs other market participants would use to determine fair value of the same instruments, including yield curves, quoted market prices of comparable securities or instruments, published credit spreads, and other applicable market data as well as instrument-specific characteristics that include, but are not limited to, coupon rates, expected cash flows, sector of the issuer, and call provisions. Because judgment is required in developing the fair values of these financial assets, they may differ from the amount actually received to sell an asset in an orderly transaction between market participants at the measurement date. Moreover, the use of different valuation assumptions may have a material effect on the financial assets’ fair values.

For most of our financial assets measured at fair value, all significant inputs are based on or corroborated by market observable data, and

The Allstate Corporation 85 significant management judgment does not affect the periodic determination of fair value. The determination of fair value using discounted cash flow models involves management judgment when significant model inputs are not based on or corroborated by market observable data. However, where market observable data is available, it takes precedence, and as a result, no range of reasonably likely inputs exists from which the basis of a sensitivity analysis could be constructed.

We gain assurance that our financial assets are appropriately valued through the execution of various processes and controls designed to ensure the overall reasonableness and consistent application of valuation methodologies, including inputs and assumptions, and compliance with accounting standards. For fair values received from third parties or internally estimated, our processes and controls are designed to ensure that the valuation methodologies are appropriate and consistently applied, the inputs and assumptions are reasonable and consistent with the objective of determining fair value, and the fair values are accurately recorded. For example, on a continuing basis, we assess the reasonableness of individual fair values that have stale security prices or that exceed certain thresholds as compared to previous fair values received from valuation service providers or brokers or derived from internal models. We perform procedures to understand and assess the methodologies, processes and controls of valuation service providers. In addition, we may validate the reasonableness of fair values by comparing information obtained from

valuation service providers or brokers to other third-party valuation sources for selected securities. We perform ongoing price validation procedures such as back-testing of actual sales, which corroborate the various inputs used in internal models to market observable data. When fair value determinations are expected to be more variable, we validate them through reviews by members of management who have relevant expertise and who are independent of those charged with executing investment transactions.

During periods of high volatility or market disruption, we may perform an analysis to determine whether there has been a significant decrease in the volume and level of activity for the asset when compared to normal market activity, and if so, whether transactions may not be orderly. Among the indicators we consider in determining whether a significant decrease in the volume and level of market activity for a specific asset has occurred include the level of new issuances in the primary market, trading volume in the secondary market, level of credit spreads over historical levels, bid-ask spread, and price consensuses among market participants and sources. If evidence indicates that prices are based on transactions that are not orderly, we place little, if any, weight on the transaction price and will estimate fair value using an internal model. As of December 31, 2023 and 2022, we did not adjust fair values provided by our valuation service providers or brokers or substitute them with an internal model for such securities.

(1)Includes $21 million that are valued using broker quotes and $354 million that are valued using quoted prices or quoted net asset values from deal sponsors.

For additional detail on fair value measurements, see Note 6 of the consolidated financial statements.

Impairment of fixed income securities with credit losses For fixed income securities classified as available-for-sale, the difference between amortized cost, net of credit loss allowance (“amortized cost, net”) and fair value, net of certain other items and deferred income taxes (as disclosed in Note 5 of the consolidated financial statements), is reported as a component of AOCI on the Consolidated Statements of Financial Position and is not reflected in the operating results of any period until reclassified to net income upon the consummation of a transaction with an unrelated third party or when a credit loss allowance is recorded. We have a comprehensive portfolio monitoring process to identify and evaluate each fixed income security that may require a credit loss allowance.

For each fixed income security in an unrealized loss position, we assess whether management with the

appropriate authority has made the decision to sell or whether it is more likely than not we will be required to sell the security before recovery of the amortized cost basis for reasons such as liquidity, contractual or regulatory purposes. If a security meets either of these criteria, any existing credit loss allowance would be written-off against the amortized cost basis of the asset along with any remaining unrealized losses, with the incremental losses recorded in earnings.

If we have not made the decision to sell the fixed income security and it is not more likely than not we will be required to sell the fixed income security before recovery of its amortized cost basis, we evaluate whether we expect to receive cash flows sufficient to recover the entire amortized cost basis of the security. We calculate the estimated recovery value based on the best estimate of future cash flows considering past events, current conditions and reasonable and supportable forecasts. The estimated future cash flows are discounted at the security’s current effective rate, and are compared to the amortized cost of the

86 www.allstate.com security. The determination of cash flow estimates is inherently subjective, and methodologies may vary depending on facts and circumstances specific to the security. All reasonably available information relevant to the collectability of the security are considered when developing the estimate of cash flows expected to be collected. That information generally includes, but is not limited to, as applicable, the remaining payment terms of the security, prepayment speeds, the financial condition and future earnings potential of the issue or issuer, expected defaults, expected recoveries, the value of underlying collateral, origination vintage year, geographic concentration of underlying collateral, available reserves or escrows, current subordination levels, third-party guarantees and other credit enhancements. Other information, such as industry analyst reports and forecasts, sector credit ratings, financial condition of the bond insurer for insured fixed income securities, and other market data relevant to the realizability of contractual cash flows, may also be considered. The estimated fair value of collateral will be used to estimate recovery value if we determine that the security is dependent on the liquidation of collateral for ultimate settlement.

If we do not expect to receive cash flows sufficient to recover the entire amortized cost basis of the fixed income security, a credit loss allowance is recorded in earnings for the shortfall in expected cash flows; however, the amortized cost, net of the credit loss allowance, may not be lower than the fair value of the security. The portion of the unrealized loss related to factors other than credit remains classified in AOCI. If we determine that the fixed income security does not have sufficient cash flow or other information to estimate a recovery value for the security, we may conclude that the entire decline in fair value is deemed to be credit related and the loss is recorded in earnings.

When a security is sold or otherwise disposed or the security is deemed uncollectible and written off, we remove amounts previously recognized in the credit loss allowance. Recoveries after write-offs are recognized when received.

For additional detail on investment impairments, see Note 5 of the consolidated financial statements.

Evaluation of goodwill Goodwill represents the excess of amounts paid for acquiring businesses over the fair value of the net assets acquired, less any impairment of goodwill recognized. Goodwill is recognized when acquired and allocated to reporting units based on which unit is expected to benefit from the synergies of the business combination. Our goodwill reporting units are equivalent to our reportable segments to which goodwill has been assigned: Allstate Protection, Protection Services, and Allstate Health and Benefits.

The goodwill balance was $3.50 billion at both December 31, 2023 and 2022.

Upon acquisition, the purchase price of the acquired business is assumed to be its fair value. Subsequently, we estimate the fair value of our businesses in each goodwill reporting unit, utilizing a

combination of widely accepted valuation techniques including a stock price and market capitalization analysis, discounted cash flow (“DCF”) calculations and an estimate of a business’s fair value using market to book multiples derived from peer company analysis. The stock price and market capitalization analysis takes into consideration the quoted market price of our outstanding common stock and includes a control premium, derived from relevant historical acquisition activity, in determining the estimated fair value of the consolidated entity before allocating that fair value to individual reporting units. The DCF analysis utilizes long term assumptions for revenues, investment income, benefits, claims, other operating expenses and income taxes to produce projections of both income and cash flows available for dividends that are present valued using the weighted average cost of capital. Market to book multiples represent the mean market to book multiple for selected peer companies with operations similar to each goodwill reporting unit to which the multiple is applied. The outputs from these methods are weighted based on the nature of the business and the relative amount of market observable assumptions supporting the estimates. The computed values are then weighted to reflect the fair value estimate based on the specific attributes of each goodwill reporting unit.

Estimating the fair value of reporting units is a subjective process that involves the use of significant estimates by management. Changes in market inputs or other events impacting the fair value of these businesses, including discount rates, operating results, investment returns, strategies and growth rate assumptions, among other factors, could result in goodwill impairments, resulting in a charge to income. Certain of our goodwill reporting units are comprised of a combination of legacy and acquired businesses and as a result have substantial internally generated and unrecognized intangibles and fair values that significantly exceed their carrying values.

The most significant assumptions utilized in the determination of the estimated fair value of the Protection Services reporting unit are the earnings growth rate and discount rate. The earnings growth rate in the DCF projection period is based on the business strategy for the reporting unit.

The discount rate, which is consistent with the weighted average cost of capital expected by a market participant, is based upon industry specific required rates of return, including consideration of both debt and equity components of the capital structure. Our discount rate may be impacted by changes in the risk-free rate, cost of debt, equity risk premium and entity specific risks.

Changes in our growth assumptions, including the risk of loss of key customers, or adverse changes in the discount rates could result in a decline in fair value and result in a goodwill impairment charge.

The Allstate Corporation 87

Reserve for property and casualty insurance claims and claims expense estimation Reserves are established to provide for the estimated costs of paying claims and claims expenses under insurance policies we have issued. These reserves are an estimate of amounts necessary to settle all outstanding claims, including estimates of all expenses associated with processing and settling incurred claims and IBNR, less the amount of paid losses, as of the financial statement date. Total reserves comprise three major components: case (established by claim adjusters), supplemental, and IBNR. Estimating the ultimate cost of claims and claims expenses is an inherently uncertain and complex process involving a high degree of judgment and is subject to the evaluation of numerous variables. Underwriting results are significantly influenced by estimates of property and casualty insurance claims and claims expense reserves.

Reserves are established for each business segment and line of business, independently of business segment management. The significant lines of business are auto, homeowners, and other personal lines for Allstate Protection, and asbestos, environmental, and other run-off lines for Run-off Property-Liability. Allstate Protection’s current period claims are typically reported promptly with relatively little reporting lag between the date of occurrence and the date the loss is reported. Auto and homeowners liability losses generally take an average of about two years to settle, while auto physical damage, homeowners property and other personal lines generally have an average settlement time of less than one year. Liability losses, especially those involving litigation, can take several years to resolve. Run-off Property-Liability involves long-tail losses, such as those related to asbestos and environmental claims, which often involve substantial reporting lags and extended times to settle.

We update most of our reserve estimates quarterly and as new data and information become available or as events emerge that may affect the resolution of unsettled claims. Changes in prior reserve estimates (reserve reestimates), which may be material, are determined by comparing updated estimates of ultimate losses to prior estimates, with the differences recorded as property and casualty insurance claims and claims expense in the Consolidated Statements of Operations in the period such changes are determined.

The actuarial methods used to develop reserve estimates Reserve estimates are derived by using several different actuarial estimation methods that are variations on one primary actuarial technique. The actuarial technique is known as a “chain ladder” estimation process in which historical loss patterns are applied to actual paid losses and reported losses (paid losses plus individual case reserves established by claim adjusters) for an accident year or a report year to create an estimate of how losses are likely to develop over time. An accident year refers to classifying claims based on the year in which the claims occurred. A report year refers to classifying claims based on the year in which the claims are reported. Both

classifications are used to prepare estimates of required reserves for payments to be made in the future. The key assumptions affecting our reserve estimates comprise data elements including claim counts, paid losses, case reserves, and development factors calculated with this data.

In the chain ladder estimation technique, a ratio (development factor) is calculated which compares current period results to results in the prior period for each accident year. The effects of inflation are implicitly considered in the reserving process, as development factors use historic data that incorporates inflation from recent prior periods in estimating future loss costs. The development factor estimation methodology may require modification when data changes due to changing claim reporting practices, changing claim settlement patterns, external regulatory or financial influences, or contractual coverage changes. Changes in such items and inflation can result in increased variability in loss costs and reserve estimates. Actuarial judgment is then applied to develop a best estimate of gross ultimate losses. These developments are discussed further in the loss ratio disclosures within the Allstate Protection Segment and the Property and Casualty Insurance Claims and Claims Expense Reserves sections of the MD&A. See the Run-off Property-Liability reserve estimates section for specific disclosures of industry and actuarial best practices for this segment.

How reserve estimates are established and updated Reserve estimates are developed at a detailed level, and the results are aggregated to form a consolidated reserve estimate. The detailed estimates include each line of insurance, major components of losses (such as coverages and perils), major states or groups of states and for reported losses and IBNR. Development factors are calculated for data elements such as claim counts reported and settled, paid losses, and paid losses combined with case reserves. The historical development patterns for these data elements are used to calculate reserve estimates. Based on this review, our best estimate of required reserves is recorded.

Reserves are reestimated quarterly and periodically throughout the year, by combining historical results with current actual results to calculate new development factors. This process continuously incorporates the historic and latest actual trends, and other underlying changes in the data elements used to calculate reserve estimates. New development factors are likely to differ from previous development factors used in prior reserve estimates because actual results occur differently than the assumptions contained in the previous development factor calculations. When actual development of these data elements is different than the historical development pattern used in a prior period reserve estimate, a new reserve is determined. The difference between indicated reserves based on new reserve estimates and recorded reserves (the previous estimate) is the amount of reserve reestimate. This amount, which could be material and vary significantly from period to period is recognized as an increase or decrease in Property and casualty

88 www.allstate.com insurance claims and claims expense in the Consolidated Statements of Operations. A more detailed discussion of reserve reestimates is presented

in the Property and Casualty Insurance Claims and Claims Expense Reserves section of the MD&A.

(1)Includes the unamortized fair value adjustment related to the acquisition of National General.

(2)Each of these results is consistent within a reasonable actuarial tolerance for the respective businesses.

Allstate Protection reserve estimate

Factors affecting reserve estimates Reserve estimates are developed based on the processes and historical development trends described above. These estimates are considered in conjunction with known facts and interpretations of circumstances and factors including our experience with similar cases, actual claims paid, historical trends involving claim payment patterns and pending levels of unpaid claims, loss management programs, product mix and contractual terms, changes in laws and regulations, judicial decisions, and economic conditions. When we experience changes of the type previously mentioned, we apply actuarial judgment in the determination and selection of development factors considered more reflective of the new trends, such as combining shorter or longer periods of historical results with current actual results to produce development factors used to reestimate our reserves. For example:

•Supply chain disruptions and labor shortages, higher used car prices, labor and part cost increases, unemployment levels, changes in commuting activity and driving behavior have and may continue to lead to historical development trends being less predictive of future loss development, potentially creating additional reserve variability.

•If a legal change is expected to have a significant impact on the development of claim severity for a coverage which is part of a particular line of insurance in a specific state, judgment is applied to determine appropriate development factors that will most accurately reflect the expected impact on that specific estimate.

•A change in economic conditions is expected to affect the cost of repairs to damaged autos or property for a particular line, coverage, or state, actuarial judgment is applied to determine appropriate development factors to use in the reserve estimate that will most accurately reflect the expected impacts on severity development.

Case and supplemental reserves

•As claims are reported, for certain liability claims of sufficient size and complexity, the field adjusting staff establishes case reserve estimates of ultimate cost, based on their assessment of facts and circumstances related to each individual claim.

•For other claims which occur in large volumes and settle in a relatively short time frame, it is not practical or efficient to set case reserves for each claim, and a statistical case reserve is set for these claims based on estimation techniques described above.

•In the normal course of business, we may also supplement our claims processes by utilizing third-party adjusters, appraisers, engineers, inspectors, and other professionals and information sources to assess and settle catastrophe and non-catastrophe related claims.

The Allstate Corporation 89

•Historically, the case reserves set by the field adjusting staff have not proven to be an entirely accurate estimate of the ultimate cost of claims. To provide for this, a development reserve is estimated using the processes described above and allocated to pending claims as a supplement to case reserves.

•Typically, the case, including statistical case, and supplemental development reserves comprise about 90% of total reserves.

Incurred but not reported (“IBNR”)

•IBNR can be a small percentage of reserves for relatively short-term claims, such as auto physical damage claims, or a large percentage of reserves for claims that have uncertain payout requirements over a long period of time, such as auto injury and MCCA claims.

•Comprises about 10% of total reserves.

Generally, the initial reserves for a new accident year are established based on actual claim frequency and severity assumptions for different business segments, lines and coverages based on historical relationships to relevant inflation indicators. Reserves for prior accident years are statistically determined using processes described above. Changes in auto claim frequency may result from changes in mix of business, driving behaviors, miles driven or other macroeconomic factors. Changes in auto current year claim severity are generally influenced by inflation in the medical and auto repair sectors of the economy, the effectiveness and efficiency of our claim settlements and changes in mix of claim type. Injury claims are affected largely by medical inflation, treatment trends, attorney representation and litigation costs while physical damage claims are affected largely by auto repair cost inflation, used car prices, length of claim resolution and the timing of receipt of third-party carrier claims.

We mitigate these effects through various loss management programs. For auto physical damage coverages, we monitor our rate of increase in average cost per claim against the auto maintenance, repair, parts and equipment price indices. We believe our claim settlement initiatives, such as improvements to the claim review and settlement process, the use of special investigative units to detect fraud and handle suspect claims, litigation management and defense strategies, as well as various other loss management initiatives underway, contribute to the mitigation of injury and physical damage severity trends.

Changes in homeowners current year claim severity are generally influenced by inflation in the cost of building materials, the cost of construction and property repair services, the cost of replacing home furnishings and other contents, the types of claims that qualify for coverage, deductibles, other economic and environmental factors and the effectiveness of our claim practices.

As loss experience for the current year develops for each type of loss, it is monitored relative to initial assumptions until it is judged to have sufficient statistical credibility. From that point in time forward,

reserves are reestimated using statistical actuarial processes to reflect the impact actual loss trends have on development factors incorporated into the actuarial estimation processes.

Causes of reserve estimate uncertainty Since reserves are estimates of unpaid portions of claims and claims expenses for claims that have occurred, including IBNR losses, the establishment of appropriate reserves, including reserves for catastrophe losses, requires regular reevaluation and refinement of estimates to determine our ultimate loss estimate.

At each reporting date, the highest degree of uncertainty in estimates for most of our losses from ongoing businesses arise from claims remaining to be settled for the current accident year and the most recent preceding accident year. The greatest degree of uncertainty exists in the current accident year because the current accident year contains the greatest proportion of losses that have not been reported or settled as well as heightened uncertainty for claims that involve litigation or take longer to settle during periods of rapidly increasing loss costs, but must be estimated as of the current reporting date. Most of these losses relate to damaged property such as automobiles and homes, and medical care for injuries from accidents. During the first year after the end of an accident year, a large portion of the total losses for that accident year are settled. When accident year losses paid through the end of the first year following the initial accident year are incorporated into updated actuarial estimates, the trends inherent in the settlement of claims emerge more clearly. After the second year, the losses that we pay for an accident year typically relate to claims that are more difficult to settle, such as those involving serious injuries or litigation. Private passenger auto insurance provides a good illustration of the uncertainty of future loss estimates: our typical annual percentage payout of reserves remaining at December 31 for an accident year is approximately 50% in the first year after the end of the accident year, 20% in the second year, 10% in the third year, 10% in the fourth year, and the remaining 10% thereafter.

Reserves for catastrophe losses Catastrophe losses are an inherent risk of the property and casualty insurance industry that have contributed, and will continue to contribute, to potentially material year-to-year fluctuations in our results of operations and financial position. We define a “catastrophe” as an event that produces pre-tax losses before reinsurance in excess of $1 million and involves multiple first party policyholders, or a winter weather event that produces a number of claims in excess of a preset, per-event threshold of average claims in a specific area, occurring within a certain amount of time following the event. Catastrophes are caused by various natural events including high winds, winter storms and freezes, tornadoes, hailstorms, wildfires, tropical storms, tsunamis, hurricanes, earthquakes and volcanoes.

We are also exposed to man-made catastrophic events, such as certain types of terrorism, civil unrest, wildfires or industrial accidents. The nature and level of

90 www.allstate.com catastrophes in any period cannot be reliably predicted.

The estimation of claims and claims expense reserves for catastrophe losses also comprises estimates of losses from reported claims and IBNR, primarily for damage to property. In general, our estimates for catastrophe reserves are based on claim adjuster inspections and the application of historical loss development factors as described above. However, depending on the nature of the catastrophe, the estimation process can be further complicated. For example, for hurricanes, complications could include the inability of insureds to promptly report losses, limitations placed on claims adjusting staff affecting their ability to inspect losses, determining whether losses are covered by our homeowners policy (generally for damage caused by wind or wind driven rain) or specifically excluded coverage caused by flood, exposure to mold damage, and the effects of numerous other considerations, including the timing of a catastrophe in relation to other events, such as at or near the end of a financial reporting period, which can affect the availability of information needed to estimate reserves for that reporting period. In these situations, we may need to adapt our practices to accommodate these circumstances in order to determine a best estimate of our losses from a catastrophe. For example, to complete estimates for certain areas affected by catastrophes not yet inspected by our claims adjusting staff, or where we believed our historical loss development factors were not predictive, we rely on analysis of actual claim notices received compared to total PIF, as well as visual, governmental and third-party information, including aerial photos, using satellites, aircrafts and drones, area observations, and data on wind speed and flood depth to the extent available.

Potential variability in reserve estimates Reserve estimates, by their nature, are very complex to determine, subject to significant judgment, and represent approximations rather than an exact determination for each outstanding claim, including claims incurred but not reported. Accordingly, as actual claims, paid losses, and case reserve results emerge, our estimate of the ultimate cost to settle will differ from previous estimates.

The reserve liability recorded in the Consolidated Statements of Financial Position represents the aggregation of numerous analyses by each business segment, line of insurance, major types of losses (such as coverages and perils), and individual states or groups of states for reported losses and IBNR. Because of this detailed approach to our reserve analysis, there is not a single set of assumptions that determines our reserve estimates at the consolidated level or that management believes can produce a statistically credible or reliable actuarial reserve range that would be meaningful.

To develop a statistical indication of potential reserve variability within reasonably likely possible outcomes, an actuarial technique (stochastic modeling) is applied to the countrywide consolidated data elements for paid losses and paid losses combined

with case reserves separately for injury losses, auto physical damage losses, and homeowners insurance losses excluding catastrophe losses. Based on the combined historical variability of the development factors calculated for these data elements, an estimate of the standard error or standard deviation around these reserve estimates is calculated within each accident year for the last twelve years for each type of loss. The variability of these reserve estimates within one standard deviation of the mean (a measure of frequency of dispersion often viewed to be an acceptable level of accuracy) is believed by management to represent a reasonable and statistically probable measure of potential variability. Based on our products and coverages, historical experience, the statistical credibility of our extensive data and stochastic modeling of actuarial chain ladder methodologies used to develop reserve estimates, we have derived the potential standard deviation and impact to net income to our Allstate Protection reserves, excluding catastrophe losses, as shown below. Catastrophe losses are also an inherent risk of the property and casualty insurance industry that have contributed, and will continue to contribute, to potentially material year-to-year fluctuations in our results of operations and financial position.

(1)Excludes reserves related to catastrophes.

Although this evaluation reflects most reasonably likely outcomes, it is possible the final outcome may fall below or above these amounts. Historical variability of reserve estimates is reported in the Property and Casualty Insurance Claims and Claims Expense Reserves section of the MD&A.

The Company regularly updates its reserve estimates as new information becomes available and as events unfold that may affect the resolution of unsettled claims. Changes in reserve estimates, which may be material, are reported in property and casualty insurance claims and claims expense in the Consolidated Statements of Operations in the period such changes are determined. Management believes that the reserve for property and casualty insurance claims and claims expense, net of recoverables, is appropriately established in the aggregate and adequate to cover the ultimate net cost of reported and unreported claims arising from losses which had occurred by the date of the Consolidated Statements of Financial Position based on available facts, laws and regulations.

The Allstate Corporation 91

Reserves for Michigan and New Jersey unlimited personal injury protection Claims and claims expense reserves include reserves for Michigan mandatory unlimited personal injury protection coverage to insureds involved in qualifying motor vehicle accidents. The administration of this program is through the MCCA, a state-mandated, non-profit association of which all insurers actively writing automobile coverage in Michigan are members.

The process employed to estimate MCCA covered losses involves a number of activities including the comprehensive review and interpretation of MCCA actuarial reports, other MCCA members’ reports and our personal injury protection loss trends which have increased in severity over time. A significant portion of incurred claim reserves can be attributed to a small number of catastrophic claims and thus a large portion of the recoverable is similarly concentrated. We conduct comprehensive claim file reviews to develop case reserve type estimates of specific claims, which inform our view of future claim development and longevity of claimants. Each year, we update the actuarial estimate of our ultimate reserves and recoverables. We report our paid and unpaid claims based on MCCA requirements. The MCCA develops its own reserving estimates based on its own reserve methodologies, which may not align with our estimations. The MCCA does not provide member companies with its estimate of a company’s claim costs.

We provide similar personal injury protection coverage in New Jersey for auto policies issued or renewed in New Jersey prior to 1991 that is administered by PLIGA. We use similar actuarial estimating techniques as for the MCCA exposures to estimate loss reserves for unlimited personal injury protection coverage for policies covered by PLIGA. Unlimited coverage was not offered after 1991; therefore, no new claimants are being added.

Reserve estimates are confidential and proprietary and by their nature are very complex to determine and subject to significant judgments. Reserve estimates do not represent an exact determination for each outstanding claim. Claims may be subject to litigation. As actual claims, paid losses and case reserve results emerge, our estimate of the ultimate cost to settle may be materially greater or less than previously estimated amounts.

For additional information related to indemnification recoverables, see Item 1 - Regulation, Indemnification Programs and Note 11 of the consolidated financial statements.

Adequacy of reserve estimates We believe our net claims and claims expense reserves are appropriately established based on available methodologies, facts, laws and regulations. Additionally, we rely on historical claims experience to inform the level of the recorded reserve. We calculate and record a single best reserve estimate, in conformance with generally accepted actuarial standards and practices, for each line of insurance, its components (coverages and perils) and state, for reported losses and for IBNR losses, and as a

result we believe that no other estimate is better than our recorded amount. Due to the uncertainties involved, the ultimate cost of losses may vary materially from recorded amounts, which are based on our best estimates.

Run-off Property-Liability reserve estimates

Characteristics of Run-off exposure Our exposure to asbestos, environmental and other run-off claims arise principally from assumed reinsurance coverage written during the 1960s through the mid-1980s, including reinsurance on primary insurance written on large U.S. companies, and from direct excess commercial insurance written from 1972 through 1985, including substantial excess general liability coverages on large U.S. companies. Additional exposure stems from direct primary commercial insurance written during the 1960s through the mid-1980s. Asbestos claims relate primarily to bodily injuries asserted by claimants who were exposed to asbestos or products containing asbestos. Environmental claims relate primarily to pollution and related clean-up costs. Other run-off claims exposures primarily relate to general liability and product liability mass tort claims, such as those for medical devices and other products, workers’ compensation claims and claims for various other coverage exposures other than asbestos and environmental.

In 1986, the general liability policy form used by us and others in the property and casualty industry was amended to introduce an “absolute pollution exclusion,” which excluded coverage for environmental damage claims, and to add an asbestos exclusion. Most general liability policies issued prior to 1987 contain annual aggregate limits for product liability coverage. General liability policies issued in 1987 and thereafter contain annual aggregate limits for product liability coverage and annual aggregate limits for all coverages. Our experience to date is that these policy form changes have limited the extent of our exposure to environmental and asbestos claim risks.

Our exposure to liability for asbestos, environmental and other run-off claims losses manifests differently depending on whether it arises from assumed reinsurance coverage, direct excess commercial insurance or direct primary commercial insurance.

Direct excess commercial insurance and reinsurance involve coverage written by us for specific layers of protection above retentions and other insurance plans and largely has resulted in asbestos, environmental and mass tort claims. The nature of excess coverage and reinsurance provided to other insurers limits our exposure to loss to specific layers of protection in excess of policyholder retention on their primary insurance plans. Our exposure is further limited by the significant reinsurance that we had purchased on our direct excess business.

Our assumed reinsurance business involved writing generally small participations in other insurers’ reinsurance programs. The reinsured losses in which we participate may be a proportion of all eligible losses or eligible losses in excess of defined retentions. Of the

92 www.allstate.com majority of our assumed reinsurance exposure, approximately 85%, is for excess of loss coverage, while the remaining 15% is for pro-rata coverage.

Our direct primary commercial insurance business comprises a cross section of policyholders engaged in many diverse business sectors throughout the country and did not include coverage to large asbestos manufacturers.

How reserve estimates are established and updated We conduct an annual review in the third quarter to evaluate, establish and adjust as necessary, asbestos, environmental and other run-off claims reserves. Changes to reserves are recorded in the reporting period in which they are determined. Using established industry and actuarial best practices and assuming no change in the regulatory or economic environment, this detailed and comprehensive methodology determines asbestos reserves based on assessments of the characteristics of exposure (i.e. claim activity, potential liability, jurisdiction, products versus non-products exposure) presented by individual policyholders, and determines environmental reserves based on assessments of the characteristics of exposure (i.e. environmental damages, respective shares of liability of potentially responsible parties, appropriateness and cost of remediation) to pollution and related clean-up costs. The number and cost of these claims are affected by advertising by trial lawyers seeking asbestos plaintiffs, and entities with asbestos exposure seeking bankruptcy protection as a result of asbestos liabilities, initially causing a delay in the reporting of claims, often followed by an acceleration and an increase in claims and claims expenses as settlements occur.

After evaluating our insureds’ probable liabilities for asbestos, environmental and other run-off claims, we evaluate our insureds’ coverage programs for such claims. We consider our insureds’ total available insurance coverage, including the coverage we issued. We also consider relevant judicial interpretations of policy language and applicable coverage defenses or determinations, if any.

Evaluation of both the insureds’ estimated liabilities and our exposure to the insureds depends heavily on an analysis of the relevant legal issues and litigation environment. This analysis is conducted by our specialized claims adjusting staff and legal counsel. Based on these evaluations, case reserves are established by claims adjusting staff and actuarial analysis is employed to develop an IBNR reserve, which includes estimated potential reserve development and claims that have occurred but have not been reported. As of December 31, 2023 and 2022, IBNR was 55.7% and 55.9%, respectively, of combined net asbestos and environmental reserves.

For both asbestos and environmental reserves, we also evaluate our historical direct net loss and expense paid and incurred experience to assess any emerging trends, fluctuations or characteristics suggested by the aggregate paid and incurred activity. Other run-off claims reserves are based on considerations similar to those described above, as they relate to the

characteristics of specific individual coverage exposures.

Potential reserve estimate variability Establishing net loss reserves for asbestos, environmental and other run-off claims is subject to uncertainties that are much greater than those presented by other types of property and casualty claims. Among the complications are lack of historical data, long reporting delays, uncertainty as to the number and identity of insureds with potential exposure and unresolved legal issues regarding policy coverage; unresolved legal issues regarding the determination, availability and timing of exhaustion of policy limits; plaintiffs’ evolving and expanding theories of liability; availability and collectability of recoveries from reinsurance; retrospectively determined premiums and other contractual agreements; estimates of the extent and timing of any contractual liability; the impact of bankruptcy protection sought by various asbestos producers and other asbestos defendants; and other uncertainties. There are also complex legal issues concerning the interpretation of various insurance policy provisions and whether those losses are covered, or were ever intended to be covered, and could be recoverable through retrospectively determined premium, reinsurance or other contractual agreements. Courts have reached different and sometimes inconsistent conclusions as to when losses are deemed to have occurred and which policies provide coverage; what types of losses are covered; whether there is an insurer obligation to defend; how policy limits are determined; how policy exclusions and conditions are applied and interpreted; and whether clean-up costs represent insured property damage. Our reserves for asbestos, environmental and other run-off exposures could be affected by tort reform, class action litigation, and other potential legislation and judicial decisions. Environmental exposures could also be affected by a change in the existing federal Superfund law and similar state statutes. There can be no assurance that any reform legislation will be enacted or that any such legislation will provide for a fair, effective and cost-efficient system for settlement of asbestos or environmental claims. We believe these issues are not likely to be resolved in the near future, and the ultimate costs may vary materially from the amounts currently recorded resulting in material changes in loss reserves. Historical variability of reserve estimates is demonstrated in the Property and Casualty Insurance Claims and Claims Expense Reserves section of the MD&A.

Adequacy of reserve estimates Management believes its net loss reserves for asbestos, environmental and other run-off claims exposures are appropriately established based on available facts, laws, regulations, and assessments of other pertinent factors and characteristics of exposure (i.e. claim activity, potential liability, jurisdiction, products versus non-products exposure) presented by individual policyholders, assuming no change in the legal, legislative or economic environment. Due to the uncertainties and factors described above, management believes it is not practicable to develop a

The Allstate Corporation 93 meaningful range for any such additional net loss reserves that may be required.

For further discussion of these estimates and quantification of the impact of reserve estimates, reserve reestimates and assumptions, see Note 9 and Note 15 of the consolidated financial statements and the Property and Casualty Insurance Claims and Claims Expense Reserves section of the MD&A.

Pension and other postretirement plans net costs and assumptions Our defined benefit pension plans cover most U.S. employees. Benefits are provided to plan participants based on a cash balance formula. Certain participants also have a significant portion of their benefits attributable to a former final average pay formula. 79% of the projected benefit obligation (“PBO”) of our primary qualified employee plan is related to the former final average pay formula. See Note 18 of the consolidated financial statements for a discussion of these plans and their effect on the consolidated financial statements.

Our pension and other postretirement benefit costs are calculated using various actuarial assumptions and methodologies. These assumptions include discount rates, health care cost trend rates, inflation, expected returns on plan assets, mortality and other factors. The assumptions utilized in recording the obligations under our pension plans represent our best estimates and we believe they are reasonable based on information as to historical experience and performance as well as other factors that might cause future expectations to differ from past trends.

Net costs for our defined benefit plans are recognized on the Consolidated Statements of Operations and consist of two elements: 1) costs comprised of service and interest costs, expected return of plan assets, amortization of prior service credit and curtailment gains and losses which are reported in property and casualty claims and claims expense, operating costs and expenses, net

investment income and, if applicable, restructuring and related charges and 2) remeasurement gains and losses comprised of changes in actuarial assumptions and the difference between actual and expected returns on plan assets which are recognized immediately in earnings as part of pension and other postretirement remeasurement gains and losses.

We recognize expected returns on plan assets using an unadjusted fair value method. Our policy is to remeasure our pension and postretirement plans on a quarterly basis. We immediately recognize the remeasurement of the projected benefit obligation and plan assets in earnings as it provides greater transparency of our economic obligations in accounting results and better aligns the recognition of the effects of economic and interest rate changes on pension and other postretirement plan assets and liabilities in the year in which the gains and losses are incurred.

Differences in actual experience and changes in other assumptions affect our pension and other postretirement obligations and expenses. Differences between expected and actual returns on plan assets affect remeasurement gains and losses. The primary factors contributing to pension and postretirement remeasurement gains and losses are: 1) changes in the discount rate used to value pension and postretirement obligations as of the measurement date; 2) differences between the expected and the actual return on plan assets; 3) changes in demographic assumptions, including mortality and participant experience; and 4) changes in lump sum interest rates used to value pension obligations as of the measurement date.

Pension and other postretirement service cost, interest cost, expected return on plan assets and amortization of prior service credits are allocated to our reportable segments. The pension and other postretirement remeasurement gains and losses are reported in the Corporate and Other segment.

Impact of assumption changes to net cost for pension and other postretirement plans Remeasurement losses in 2023 primarily related to losses on the remeasurement of the projected benefit obligation driven by a decrease in the liability discount rate, partially offset by gains due to favorable asset performance compared to expected return on plan assets. Remeasurement losses in 2022 primarily related to unfavorable asset performance compared to the expected return on plan assets, partially offset by a reduction in the projected benefit obligation due to an increase in the liability discount rate and changes in

other assumptions, primarily related to an increase in the long-term lump sum interest rate.

The discount rate is based on rates at which expected pension benefits attributable to past employee service could effectively be settled on a present value basis at the measurement date. We develop the assumed discount rate by utilizing the weighted average yield of a theoretical dedicated portfolio derived from non-callable bonds and callable bonds with a make-whole provision available in the Bloomberg corporate bond universe having ratings of at least “AA” by S&P or at least “Aa” by Moody’s on the

94 www.allstate.com measurement date with cash flows that match expected plan benefit requirements. Significant changes in discount rates, such as those caused by changes in the credit spreads, yield curve, the mix of bonds available in the market, the duration of selected bonds and expected benefit payments, may result in volatility in pension cost. The weighted average discount rate used to measure the benefit obligation decreased to 5.35% at December 31, 2023 compared to 5.64% at December 31, 2022, resulting in remeasurement losses for 2023.

The expected long-term rate of return on plan assets reflects the average rate of earnings expected on plan assets. While this rate reflects long-term assumptions and is consistent with long-term historical returns, sustained changes in the market or changes in the mix of plan assets may lead to revisions in the assumed long-term rate of return on plan assets that may result in variability of pension cost. Differences between the actual return on plan assets and the expected long-term rate of return on plan assets are immediately recognized through earnings upon remeasurement. Short-term asset performance can differ significantly from the expected rate of return, especially in volatile markets. In 2023, the actual return on plan assets was higher than the expected return

primarily due to positive equity returns and higher fixed income valuations from lower interest rates and tighter credit spreads. In 2022, the actual return on plan assets was lower than the expected return primarily due to higher interest rates, widening credit spreads and weak equity market performance.

We complete periodic evaluations of demographic information and historical experience that affects our pension and other postretirement obligations to identify any required changes to long-term actuarial assumptions and methodologies. Demographic assumptions affect both our pension and postretirement plans and include elements such as retirement rates and participation rates in our postretirement programs, among other factors.

These actuarial assumption updates affect our pension and other postretirement obligations and are incorporated into our best estimates of these assumptions. Remeasurement losses for other assumptions in 2023 primarily related to a decrease in the long-term lump sum interest rate, partially offset by gains from mortality updates. Remeasurement gains for other assumptions in 2022 primarily related to an increase in the long-term lump sum interest rate.

The Allstate Corporation 95