OCTAVE SPECIALTY GROUP INC

CRITICAL ACCOUNTING POLICIES AND ESTIMATES ($ in thousands)

Octave's Consolidated Financial Statements have been prepared in accordance with accounting principles generally accepted in

the U.S. ("GAAP"). This section highlights accounting estimates management views as critical because they are most important to the portrayal of the Company's financial condition; and require management to make difficult and subjective judgments regarding matters that are inherently uncertain and subject to change. These estimates are evaluated on an ongoing basis considering historical developments, political events, market conditions, industry trends and other information. There can be no assurance that actual results will conform to estimates and that reported results of operations will not be materially adversely affected by the need to make future accounting adjustments to reflect changes in these estimates from time to time.

Management has identified the following critical accounting policies and estimates: (i) valuation of specialty property and casualty losses and loss adjustment expense reserves, (ii) business combinations including identification and valuation of intangible assets, and (iii) goodwill and intangible asset impairment analysis. Management has discussed each of these critical accounting policies and estimates with the Audit Committee, including the reasons why they are considered critical and how current and anticipated future events impact those determinations. Additional information about these policies can be found in Note 2. Basis of Presentation and Significant Accounting Policies to the Consolidated Financial Statements included in Part II, Item 8 in this Form 10-K.

SPECIALTY PROPERTY AND CASUALTY LOSSES AND LOSS EXPENSE RESERVES

The specialty property and casualty insurance segment consists of Everspan-affiliated carriers. Loss and loss adjustment expense reserves represent management's estimate of the ultimate liability for unpaid losses and loss expenses for claims that have been reported and incurred but not yet reported ("IBNR") as of the balance sheet date.

Loss and loss adjustment expense reserves by line of business were as follows as of December 31, 2025 and 2024:

(1) Unallocated loss adjustment expenses.

(2) Includes $23,530 and $0 total loss and loss expense reserves on a gross and net of reinsurance basis related to legacy liabilities obtained from the acquisitions of Providence Washington Insurance Company, Greenwood Insurance Company, and Consolidated Specialty Insurance Company. All legacy liabilities remain obligations of affiliates of the sellers through reinsurance and contractual indemnities.

(1) Unallocated loss adjustment expenses.

(2) Includes $35,146 and $0 total loss and loss expense reserves on a gross and net of reinsurance basis related to legacy liabilities obtained from the acquisitions of Providence Washington Insurance Company, Greenwood Insurance Company, and Consolidated Specialty Insurance Company. All legacy liabilities remain obligations of affiliates of the sellers through reinsurance and contractual indemnities.

Loss and loss adjustment expense reserves, evaluated at a program and line of business level, are estimated based upon experience and using a variety of actuarial methods and are subject to the impact of future changes in factors such as claim severity and frequency, underwriting and claims practices, changes in social and economic conditions, including the impact of inflation, legal and judicial developments, medical cost trends and upward trends in damage awards. The key assumptions used to arrive at the best estimate of loss reserves are the expected loss ratios and loss development factors representing reported and paid loss emergence patterns. Our actuarial methods may also rely on external data, such as industry loss ratios, loss development factors, or trend factors. The initial estimate for an accident year is generally based on an exposure-based method using the loss ratio projection method. The loss ratio projection method develops an initial estimate of ultimate claims and claim adjustment expenses for an accident year by multiplying earned premium for the accident year by a projected loss ratio. The projected loss ratio is determined by analyzing prior period experience, and adjusting for loss cost trends, rate level differences, a mix of business changes and industry loss ratios and other known or observed factors influencing the accident year relative to prior accident years.

The loss and loss adjustment expense reserves estimate may be based on a judgmental weighting of estimates produced from multiple estimation and analysis methods considered. The method(s) selected and weighted are those that are believed to produce the most accurate estimate at that particular evaluation date. The following estimation and analysis methods are principally used by the Company’s engaged independent actuarial specialists to estimate the ultimate cost of loss and loss adjustment expenses. These estimation and analysis methods are typically referred to as conventional actuarial methods.

•The paid loss development method assumes that the future change (positive or negative) in cumulative paid losses for a given cohort of claims will occur in a stable, predictable

pattern from year-to-year, consistent with the pattern observed in past cohorts.

•The case incurred development method is the same as the paid loss development method but is based on cumulative case-incurred losses rather than paid losses.

•The Bornhuetter-Ferguson method uses an initial estimate of ultimate losses for a given product line reserve component, typically expressed as a ratio to earned premium. The method assumes that the ratio of additional claim activity to earned premium for that component is relatively stable and predictable over time and that actual claim activity to date is not a credible predictor of further activity for that component. The method is used most often for more recent accident years where claim data is sparse and/or volatile, with a transition to other methods as the underlying claim data becomes more voluminous and therefore more credible.

The actuarial results provide a range of estimated losses by program and line of business including a low, central and high estimate of losses and loss expenses. At December 31, 2025, management selected the respective midpoint loss ratio between the actuarially determined central and high estimate for active and runoff programs and lines of business for each respective accident year when recording loss and loss adjustment expense reserves. In the prior year, management selected the high end of the respective actuarial range for programs in runoff, but due to (i) increased claims and actuarial oversight, (ii) further run-off of these programs and (iii) further evaluation of Everspan's aggregate reserve position it was determined to record both active and runoff programs at the midpoint loss ratio between the actuarial central and high estimate.

Since the reserves are based on estimates, the ultimate liability may be more or less than such reserves. Octave's actuarial evaluation at December 31, 2025, provided a range of losses incurred. Losses at the low end of the range would be below our recorded gross and net loss expense reserves by approximately $49,900 and $12,500, respectively at December 31, 2025, and losses at the high end of the range would exceed our recorded gross and net loss and loss adjustment expense reserve by approximately $14,600 and $2,200, respectively, at December 31, 2025. This range reflects low and high reasonable reserve estimates determined after using judgment to adjust the methods, factors, and assumptions selected within the internal reserve review. This approach produces a range of reasonable reserve estimates but does not represent a distribution of all possible outcomes.

Additionally, changes in assumptions such as loss development patterns and expected loss ratios can result in variability in actuarial estimates.

•For the loss development pattern we considered the impact of the reported incurred losses developing faster or slower than expected in our projections. For every 1.0% slower or faster the losses develop, we would expect our net indicated reserves to increase or decrease, respectively, by a range of 0.8% to 1.2%. If our reported loss development pattern was 5% slower, the net indicated reserves would be approximately 5% higher. If our reported loss development pattern was 5% faster, the net indicated reserves would be approximately 4.3% lower.

•For the expected losses we utilize industry benchmark loss ratios and internal pricing loss ratios applied to earned premium. For every 1.0% higher or lower the expected losses are, we would expect our net indicated reserves to increase or decrease by approximately 0.6%. If our expected losses were 5% higher, the net indicated reserves would be approximately 3.2% higher. If our expected losses were 5% lower, the net indicated reserves would be approximately 3.2% lower.

Consequently, final outcomes may be greater or less than the estimates. The extent of the range and variability of loss and loss adjustment expense reserves could be further impacted by future changes in factors discussed above. See “Risk Factors” in Part I, Item 1A in this Annual Report on Form 10-K.

BUSINESS COMBINATIONS

The acquired entities comprising the ID segment primarily represent business combinations that were accounted for under the acquisition method of accounting. The acquisition method requires us to allocate the total consideration transferred for each acquisition to the assets acquired, liabilities assumed and noncontrolling interests ("NCI") based on their fair values as of the date of acquisition, including identifiable intangible assets. The allocation of the consideration utilizes significant estimates in determining the fair values of net assets acquired, which primarily consist of customer relationship intangible assets, but may include other finite-lived intangible assets including trade names or non-compete agreements. Measurement of the purchase balance sheet also requires valuation of redeemable NCI interests and nonredeemable NCI interests when applicable.

The valuation method used to determine customer relationship intangible assets was the multi-period excess earnings method "(MPEEM"), which quantifies the residual (or excess) cash flows generated by the intangible asset and discounts those cash flows to their present value. The significant assumptions used in determining the fair value of customer relationships include estimated revenue growth, customer attrition rates, operating margins, and discount rate. These estimates directly impact the amount of identified intangible assets recognized and the related amortization expense in future periods. As of December 31, 2025 and 2024, an aggregate of $463,785 and $333,562, respectively, of acquired intangible assets, net of accumulated amortization, was recorded on the Consolidated Balance Sheets, of which $446,835 and $323,720, respectively, represented customer relationships.

The valuation method used to determine the fair value of redeemable NCI interests and related put and call options was the Monte Carlo Simulation. The significant fair value assumptions used in the simulation include the exercise thresholds, EBITDA forecasts, discount rate and long-term growth rate. The valuation method to determine the fair value of nonredeemable NCI interests, which do not contain put or call options, was the discounted cash flow approach. The significant fair value assumptions used in the model include estimated long term revenue and expense forecasts and the discount rate.

The excess of purchase price over the fair value of assets acquired, liabilities assumed, and NCI interests (both redeemable and nonredeemable) is recorded as goodwill. We may refine our estimates and make adjustments to the assets acquired and liabilities assumed over a measurement period, not to exceed one year from the date of acquisition.

GOODWILL AND INTANGIBLE IMPAIRMENT ANALYSIS

Goodwill impairment evaluation

We perform the impairment assessment of goodwill at the reporting unit level within our ID segment on an annual basis or more frequently if circumstances indicate a possible impairment. We have determined that each of the entities acquired in the ID segment represent an individual reporting unit. The impairment test may first consider qualitative factors to determine whether it is more likely than not that the fair value of a reporting unit is less than its carrying amount. Examples of qualitative factors include, macroeconomic conditions, industry and market considerations, cost factors, overall financial performance, entity-specific events, events affecting reporting units and sustained changes in our stock price. If results of the qualitative assessment indicate a more likely than not determination or if we elect not to perform a qualitative assessment, a quantitative test is performed by comparing the estimated fair value using a weighted average of an income approach and market approach for each reporting unit to its carrying value. For the 2025 annual impairment evaluation, we elected to bypass the qualitative evaluation and perform quantitative tests on four reporting units containing $432,728 or 80% of the aggregate balance of goodwill as of December 31, 2025. For the remainder of our goodwill balance, including related to ArmadaCorp at year end, we performed a qualitative assessment. There was no goodwill impairment for any of the reporting units.

Under the quantitative assessment, the determination of fair value includes assumptions, which are considered Level 3 inputs, that are subject to risk and uncertainty. We consider different valuation approaches in the quantitative assessment. The income approach uses discounted cash flows which are dependent on subjective factors including the timing of future cash flows, the underlying margin projection assumptions, future growth rates and the discount rate. The market approach uses valuation multiples and is dependent on subjective factors including the determination of industry market multiples and EBITDA forecasts. These results are weighted to arrive at management's estimate of fair value for the reporting unit. Weight applied to each approach is based on the judgments about the extent that they reasonably reflect the value of the particular reporting unit that would be considered by a market participant. Additionally, to corroborate our estimated fair value, we perform a market capitalization reconciliation to determine if the implied control premium is reasonable.

For reporting units evaluated under the quantitative assessment, estimated fair values exceeded carrying values by between 1% and 19%. Market and income approaches were weighted 50% each, except for Octave Ventures where we applied 80% weight to the income approach, considering that the selected public companies used in the market approach do not adequately represent Octave Ventures's business as an MGA incubator under a Lloyds consortium structure. Fair values are particularly sensitive to the discount rate used in the income approach and the multiple used in the market approach. With no other changes to inputs or assumptions, a 1% increase in the discount rate would have produced an aggregate 11% reduction to the fair value of the tested reporting units. With no other changes to inputs or assumptions, a 20% decrease in the market multiples used would have produced an aggregate 4% reduction to the fair value of the tested reporting units. Changes to these or other assumptions or estimates in our fair value calculations or variances to any of the above subjective factors from what was expected, could impact our impairment analysis and result in a decline in fair value that may trigger future impairment charges.

Intangible asset impairment and useful life evaluation

We review acquired finite-lived intangible assets that are being amortized for impairment whenever events or changes in circumstance indicate that their carrying amount may not be recoverable. Qualitative factors considered include any adverse developments in regulation, unfavorable market conditions, or the extent to which an asset will be utilized. We do not believe there will be a material change in the estimates or assumptions used to calculate impairments or useful lives of amortizable intangible assets. However, if actual results are not consistent with our estimates and assumptions, we may be exposed to an acceleration of amortization or impairment losses that could be material.