Insurance by BogdanovaThe Encyclopedia of Insurance and Risk Science

RESEARCH DOSSIER

Claims Adjustment Expense Provisions in Ratemaking

Abstract. Claims adjustment expense is an essential part of the prospective cost of insurance. Its estimation requires linking an insurer's future claim handling activities, direct settlement expenditures, claims organization, legal and medical service purchases, and operational capacity to the risk portfolio being priced. This monograph examines the theoretical and practical foundations of prospective claims adjustment expense provisions in property and casualty ratemaking, with internationally relevant comparisons to technical-provision valuation and insurance-contract accounting. It separates the economic function of claims handling from the reporting categories through which expenses are observed; distinguishes claim-specific from shared costs and accounting classifications from actuarial cost drivers; derives frequency–severity, development-based, activity-based, and aggregate provision methods; and analyzes effects of inflation, coverage design, litigation, claims transformation, catastrophes, reinsurance, and portfolio change. Reproducible synthetic calculations demonstrate the sensitivity of indicated premium to differing expense bases and the risk of omissions or double counting. Primary reference points include the Casualty Actuarial Society's ratemaking literature, current U.S. actuarial standards, U.S. statutory accounting, IFRS 17, European solvency requirements, and scholarly investigations of unallocated expense reserving. The central conclusion is that a defensible rate indication treats claims adjustment expense as a prospective, explicitly reconciled system of heterogeneous cash costs with identifiable timing, activity, and uncertainty. Historical allocation ratios become useful estimators when their assumptions are demonstrated against the future claim process. This is an analytical synthesis and set of worked examples, not a new actuarial theory or a product-specific rate filing.

Keywords: Claims adjustment expense; loss adjustment expense; ALAE; ULAE; defense and cost containment; adjusting and other; ratemaking; technical premium; claim handling; claims frequency; claims severity; expense allocation; loss development; inflation; expense reserving; rate adequacy; property and casualty insurance; insurance regulation.

Introduction

A policy premium finances the promised transfer of risk together with the operations necessary to honor that promise. Those operations include recognizing an insured event, validating coverage, evaluating the evidence, paying a covered loss, defending contested demands, managing experts and legal counsel, maintaining claim records, and closing the claim. Costs arise at different stages, with different degrees of dependence on the number of claims, the size of settlements, the duration of open files, the composition of the portfolio, and the organization of the insurer. Treating them as one arbitrary percentage of either premium or loss expenditure obscures their causes.

Werner and Modlin (2016), in Basic Ratemaking, set out the fundamental equation in which projected premium covers losses, loss adjustment expenses, underwriting expenses, and the targeted underwriting result. The CAS ratemaking principles articulate a prospective cost-estimation orientation; their present institutional history is material because the statement originated in 1988, was rescinded in 2020, and was reinstated in 2021 as a reference for U.S.-regulated ratemaking. A principle of prospective cost estimation is an analytical starting point, while the binding professional and legal framework depends on the assignment and jurisdiction.

The ASOP 53 future-cost standard (Actuarial Standards Board [ASB], 2017) specifically instructs actuaries to examine the relevance of historical loss and loss adjustment expense experience, potential differences between their development, and changes in the future environment. The ASOP 29 expense standard (ASB, 2023; effective July 1, 2024) addresses expenses in prospective risk transfer and retention, including adjustment expenses when these have not already been combined with losses. That scope distinction determines the accounting logic of an indicated rate: an expense included with projected losses must not be added again as an independent loading.

The research object is the prospective cost of administering and settling claims attributable to policies written at the indicated rate, including an explicit view of shared operational resources. The analysis follows six stages: classification, attribution, statistical estimation, development to ultimate, translation into premiums, and validation of a cost-control decision. The calculations are transparent synthetic examples; they are neither estimates from a real insurer's portfolio nor empirical evidence of market-wide performance. U.S. sources anchor the technical terminology, and IFRS and European regulatory materials establish important differences in valuation purpose. The work's contribution is the integrated comparison and the derivation of practical consequences from documented sources; no priority or novel-theory claim is made.

Definitions of LAE and Loss Costs

The economic boundary of claims adjustment

Loss costs are expected amounts paid or owed under insured coverage. Claims adjustment expenses are costs of administering, investigating, validating, estimating, settling, and defending claims. The definition in ASOP 29, section 2.5 is deliberately functional. It directs attention to the activities creating cost rather than to whether an amount happened to be recorded as external counsel fees, internal payroll, a technology subscription, or a payment to a third-party administrator. The same technology platform can support both underwriting and claims. Its recoverable claims-related cost needs a justified allocation; the mere existence of the expense does not determine the appropriate share.

Historically, ratemaking distinguished allocated loss adjustment expenses (ALAE), directly assignable to particular claims, from unallocated loss adjustment expenses (ULAE), which are maintained and estimated in aggregate. Werner and Modlin (2016) explain the ALAE–ULAE distinction and the related difficulties of different reporting and data structures. The familiar formula LAE = ALAE + ULAE is an exhaustive economic partition only after definitions, accounting populations, and exclusions are aligned. A claims adjuster's salary could be identified to files through a reliable time record, or initially recorded within an aggregate cost center; the ultimate mathematical provision should reflect the same service once under either convention.

U.S. statutory classification: DCC and AO

The National Association of Insurance Commissioners (NAIC), in SSAP No. 55 within its Accounting Practices and Procedures Manual (as of March 2026), distinguishes defense and cost containment (DCC) from adjusting and other (AO). DCC encompasses defense, litigation, and medical cost containment, whether services are internal or external; AO encompasses other adjustment activities, including adjuster services and certain coverage-related legal work. Therefore DCC is not equivalent to ALAE, and AO is not automatically identical to ULAE. One classification follows the function of expenditure and the other conventionally considers assignment to claims. The mismatch is a two-axis classification problem.

For analytical reconciliation, a claims expense ledger can be represented by four cells: directly attributable DCC, shared DCC, directly attributable AO, and shared AO. If the relevant accounting population is exhaustive and mutually exclusive, their sum equals total claims adjustment expense, independently of the route used to produce the ratemaking provision. This representation is a bookkeeping crosswalk, not a new actuarial model. It permits reported totals to be reconciled while maintaining different forecasting methods for legal defense and claim administration. A coding migration that moves spending between cells does not by itself imply an economic change in expected future expense. An observed increase in DCC percentage may result entirely from a revised expense coding practice.

The 2026 NAIC accounting manual also establishes that liabilities for unpaid adjustment expenses arise from insured events already incurred; it addresses reported claims, incurred-but-not-reported claims, relevant inflation, and changes in estimation. Those liabilities answer a valuation question about existing obligations. Ratemaking concerns the prospective expenses created by future insured exposure. A recent reserve estimate can inform the latter, but a reserve dollar is not itself an appropriate rate unit without exposure, accident period, and future-process adjustments.

Cash costs, provisions, and reporting presentations

Let an insurer's prospective earned exposure count be E and let its ultimate claim losses for that exposure be L. Let D denote prospective directly attributed claims handling costs, U denote shared claims handling costs causally allocated to that exposure, F other fixed underwriting costs, v the premium-proportional underwriting expense rate, and p the target underwriting profit/contingency rate expressed as a fraction of premium in a simplified nondiscounted illustration. The indicated total premium is

[ P = \frac{L+D+U+F}{1-v-p}, \qquad 0 \leq v+p < 1. ]

The indicated average rate is P/E. This identity requires disjoint component definitions: if L already includes D, then D must be zero in the separately displayed addend or the combined loss figure must be separated before application. If U has been included in F under an accounting convention, the appropriate adjustment is to reclassify rather than add a second charge. Premium-related taxes, commission structures, and capital costs may call for additional terms; the formula above intentionally isolates the claims expense decision.

Other useful forms follow from the same identity. If predicted losses per exposure are l and claim adjustment expense per exposure is a, with fixed underwriting cost f per exposure, then indicated rate r = (l+a+f)/(1-v-p). If adjustment expense is represented by k times loss cost, the numerator becomes (1+k)l+f. If part of the claim expense instead varies with policy volume, severity, open duration, or claim count, it should enter through an appropriate forecast of those variables. The mathematical equivalence of a correctly calibrated ratio and a directly forecast cash cost does not make their assumptions equivalent across changing portfolios.

Boundary with premiums, reserves, and accounting profit

The IFRS Foundation's IFRS 17 overview explains a reporting framework that recognizes a risk-adjusted present value of insurance fulfillment cash flows alongside, where applicable, a contractual service margin. Its fulfillment cash flow terminology covers expected premiums, claims, benefits, expenses, timing, and risk. These are measurement concepts for accounting groups of contracts. A technical price indication can incorporate different commercial objectives and regulatory restrictions while drawing on compatible cash-flow projections. It is methodologically useful to reconcile the forecast inputs and explicitly document why the final measures differ.

The European Solvency II Article 77 best-estimate requirement is a probability-weighted present value of required future settlement cash flows. The delegated regulation's Article 31 expense categories include claims management expenses and the associated overhead, which must be allocated consistently and objectively for the valuation purpose. These provisions support the inclusion of claim management costs in technical-provision cash flows; they do not impose a single global pricing ratio or substitute reserve valuation for competitive premium setting.

Allocated Claims Expenses

From claim events to expected direct expenditures

The basic exposure-level direct expense forecast can be separated into claim occurrence frequency, probability that a claim requires a specific handling service, and the conditional cost of that service. For claim types j and direct activities k,

[ \mathbb E[D]/E = \sum_j \lambda_j \sum_k q_{jk} c_{jk}, ]

where λ_j is expected reported claim count of type j per unit exposure, q_jk is the expected frequency of activity k per reported claim of type j (or an expected activity count), and c_jk is its expected unit cost. For example, legal defense may occur in a minority of liability claims but dominate direct expenses for litigated files. Medical review can have a materially different pattern. The activity approach preserves those distinctions while allowing a transparent sum.

The expected unit costs should incorporate future professional fees, jurisdiction, settlement timing, technology, case complexity, and the insurer's chosen claims strategy. Multiplying total losses by a historical ALAE-to-loss ratio is a compact approximation if the future mix of direct expenses remains proportionate to the future loss dollars. The approximation may deteriorate when repair inflation moves claim severity without equivalent growth in legal or administrative expense, when litigation frequency changes independently of indemnity size, or when deductibles and coverage limits affect the settlement amounts but leave substantial investigation costs unchanged.

Case-level development and ultimate direct expense

Case-specific allocated expenses can be paid or reserved by claim, allowing triangles by accident, report, or policy period. It is useful to separate paid direct costs, case estimates of future direct costs, and incremental adjustment for unreported or inadequately reserved expenses. Developments of losses and expenses may be estimated jointly where credible association exists, or separately when their time patterns diverge. The ASOP 53 treatment of historical LAE explicitly recognizes that expense analysis can require methods and assumptions different from those used for losses. It is inappropriate to apply a paid-loss development factor automatically to attorney costs when the attorney payment tail differs from the underlying compensation tail.

Consider 1,000 synthetic liability claims with $9 million in current indemnity incurred and $1.2 million in paid or case-estimated direct expense. Suppose credible analysis projects ultimate indemnity of $10 million and ultimate direct expense of $1.6 million. The observed 13.33% direct-expense-to-current-loss ratio understates the prospective direct-expense percentage of ultimate indemnity, 16.0%, if the same claim mix and prospective adjustments apply. Conversely, applying a 16.0% ratio to an aggregate loss forecast that already includes $1.6 million of these expenses double counts them. The correct numerator is either $10.0 million + $1.6 million or $11.6 million as one combined loss-and-expense amount.

A claim-level forecast should distinguish ultimate indemnity from external and internal legal defense, investigation, medical containment, independent adjusting, and other directly incurred costs. Reserving methods for reported claims and for incurred-but-not-reported exposures are relevant inputs, but claims handling costs also arise after a case is nominally closed, through re-openings, contested recoveries, and administrative follow-up. Those channels require explicit consideration where significant.

Coverage features and legal organization

Deductibles, policy limits, defense obligations, claims-made notice conditions, arbitration clauses, and third-party administrator contracts can influence direct expense without causing commensurate changes in insured indemnity. The econometric and economic literature on claims settlement provides a reason to treat claims handling as an endogenous activity rather than a fixed cost scalar. Crocker and Tennyson (2002), in their study of fraud and settlement strategies, analyze how anticipated litigation and bad-faith costs constrain claim settlement incentives; the publisher abstract and bibliographic record were reviewed, while the complete article text was not independently inspected. This is relevant evidence of strategic interaction, not a coefficient ready to apply to an unrelated market.

Direct expense estimation should therefore assess a change in litigation propensity, claims counsel contracting, loss control practice, and provider networks separately from expected indemnity trend. A hypothetical 10% increase in court-related costs coupled with unchanged claim frequency can produce an expense movement that an indemnity-only trend misses. Where domestic laws require particular claim investigation or dispute-resolution steps, those costs enter the relevant jurisdiction's projected claims process; international comparisons must maintain jurisdiction-specific legal and contractual boundaries.

Unallocated Claim Expenses

Shared claims operations as a production cost

Shared claims expense includes salaries and benefits of claims personnel, claims supervision, call centers, common case-management platforms, physical facilities, quality review, records management, and other common costs to the extent their function is claims adjustment. The challenge is determining which future costs are required to service the priced claim population and how that activity varies with volume and duration. Operating budgets provide a starting point; they are not automatically a loss-period expense estimate.

A useful cost decomposition is

[ U_t = C_t^{\mathrm{capacity}}+c_{o,t}N_t^{\mathrm{opened}}+c_{c,t}N_t^{\mathrm{closed}}+c_{p,t}W_t+H_t, ]

where capacity represents staffing and systems held for the period, the N terms count opening and closure work, W_t denotes pending claim workload or weighted open-file time, and H_t captures additional shared tasks. This is an activity-accounting identity for planning purposes; the split is organization-specific and should avoid double counting capacity costs embedded in the unit activity coefficients. The coefficients should be based on identifiable spending, time studies, activity data, or budget decomposition. Capacity can create thresholds: a small change in claim volume may have almost no short-run cash effect until a new team is required, after which cost changes discontinuously.

Ohlsson's 2016 ULAE reserving study describes individual-claim fixed and variable expenses and their aggregation into line-of-business provisions, based on an example from a Swedish insurer. The article's journal bibliographic data and abstract were checked; a paywalled complete text was not independently available. The study reinforces the logical link between operational cost and claim state, with an important distinction: a reserve projects the run-off of existing claim obligations, whereas ratemaking estimates the claim workload associated with future insured exposure.

Attribution bases and allocation integrity

Common allocation bases include expected reported claim counts, weighted open claims, claim handling hours, expected closure events, and activity-specific cost drivers. A percentage of indemnity loss is a coarse proxy when the claims process differs greatly by settlement size. Premium volume is often a still weaker proxy for allocating the actual cost of adjusting claims, since rate increases can change premium without changing activity. A transparent allocation should sum back to the total supported operating budget, reproduce the period and portfolio definitions, and distinguish required future capacity from sunk costs and transition projects.

Let total attributable projected shared cost across business segments be U and let weights w_i be nonnegative measures of expected workload. A simple allocation is U_i = U w_i/(Σw_j), provided the weights capture the causal use of capacity and all relevant activities are within the same pool. For example, suppose two lines produce 30,000 and 10,000 claims respectively, but their average complexity weights are 1 and 3. Their weighted claims are both 30,000, implying an equal share of the common claim handling pool under this chosen criterion. Allocation based on raw claim counts would instead give 75% and 25%. Neither rule is justified solely by the fact that it reconciles to 100%; the claim complexity evidence is decisive.

The ASOP 29 expense allocation guidance supports methods based on exposures, policies, loss percentages, or other drivers when they reflect how costs arise. Its section 3.6 also requires a separate provision for claim adjustment expenses omitted from the loss component and allows distinct methods for subcategories. The intended measure of cost—mean, percentile, or another supported measure—should be documented, as should the reason a business segment receives its share. Allocation fairness and actuarial adequacy meet at this point: the cross-subsidies implied by administrative convenience should be visible to governance and rate reviewers.

The reserving literature and its limits for prospective pricing

Classical unallocated expense reserve calculations often use ratios of calendar-year paid ULAE to paid loss with an assumed relationship to unpaid losses. Johnson (1989), in Determination of Outstanding Liabilities for ULAE, develops a claim reporting and closure approach that directly considers claims department cost. Allen and Mango (1999), in Two Alternative Methods for ULAE Reserves, assess weaknesses of paid-to-paid and count-based methods and discuss alternative approaches reflecting expected loss bases and staff workloads. Kittel (1981), in ULAE Reserves in Inflationary Conditions, explores the distortions produced by growth and inflation. These CAS archival pages were checked for their abstracts and bibliographic identification; conclusions are confined to that accessible evidence.

The research relevance extends beyond reserve calibration. A growing portfolio may show rising opening workload while settlement expenses for old cohorts remain high; an aggregate paid-to-paid ratio blends these processes across accident and accounting periods. A claim-count projection exposes the resulting timing mismatch. Conversely, a workload system with inadequate coding may give false precision. The strongest estimate is selected by direct prediction performance, reconciled accounting totals, and coherent assumptions about future settlement operations, rather than by preferred historical tradition.

Claims Handling Frequency and Severity

Claim occurrence, service incidence, and expenditure

A claim is an event generating a possible payment obligation, but not every reported event requires the same amount of investigative or settlement work. For ratemaking, three frequencies must be distinguished: insured claim incidence per exposure, the incidence of particular claims-handling activities conditional on a claim, and the frequency of repeated activities while a file remains open. In long-tailed claims, the third can be driven by duration and successive procedural stages rather than simply by the original occurrence count. The predictive object is the future expenditure distribution induced by the portfolio's expected claims and operations.

A decomposition based on loss occurrence and service intensity permits separate scenario analysis. Let N be claim count for E exposures. For each claim i, let K_i count chargeable service tasks and let C_ik be the corresponding cost. Then total claim-specific expenses are D = Σ_i Σ_k C_ik, with expected cost E[D]/E determined by frequency and conditional activity. Shared expense U additionally reflects common processing intensity and available capacity. The total expected expense per exposure is E[(D+U)]/E. The last equality is merely a ratio of expectations with exposure treated as planned and fixed; where earned exposures are themselves random, choosing whether the rate targets an expectation of total cost per expected exposure or the expectation of a random cost-to-exposure ratio requires care.

Reported claims frequency is also sensitive to claims definitions, zero-payment claims, catastrophe coding, reopening, and the decision to aggregate multiple coverages into a single file. A payment-count model cannot be substituted for a reported-file-count model without explicit conversion. A file opened for investigation and denied may incur cost without indemnity. Conversely, automated handling may settle some claims with minimal incremental labor. Frequency modeling that conditions only on positive indemnity will systematically omit the first group and may underrepresent low-cost handling in the second. A sound data design retains zero-payment claims and distinct handling outcomes.

Synthetic prospective rate indication

Consider a one-year illustrative portfolio of 10,000 earned exposures, with 0.075 reported claims per exposure and expected indemnity severity of $4,000. Expected indemnity is 750 × $4,000 = $3,000,000. Every claim incurs an expected $180 of direct basic handling, while 12% of claims are expected to require an additional $900 of external legal/technical services. The direct average is $180 + 0.12 × $900 = $288 per claim, yielding D = $216,000. Shared cost comprises $420,000 of annual claims capacity plus $55 of incremental administration per new claim, producing U = $420,000 + 750 × $55 = $461,250. Other fixed underwriting expenses are $250,000. A variable expense provision of 16% of premium and an underwriting target of 4% of premium are assumed solely for the illustration.

Prospective component Assumed cost, dollars Cost per earned exposure, dollars
Ultimate claim indemnity 3,000,000 300.0000
Direct claims adjustment 216,000 21.6000
Shared claims adjustment 461,250 46.1250
Other fixed underwriting expense 250,000 25.0000
Sum before proportional costs and underwriting target 3,927,250 392.7250
Indicated premium after 20% total proportional allowance 4,909,062.50 490.90625

A 16% + 4% denominator allowance gives an indicated premium of $3,927,250 / 0.80 = $4,909,062.50, or $490.91 per earned exposure when rounded to cents. The loss and adjustment expense component is $3,677,250, equivalent to $367.725 per exposure, and the projected combined claims-adjustment-to-indemnity ratio is $677,250 / $3,000,000 = 22.575%. That ratio is a description of this prospective cost configuration, not an independently justified factor for another risk portfolio.

If the shared claims provision were omitted from the same assumptions, the indicated rate would fall to ($3,000,000 + $216,000 + $250,000)/(0.80 × 10,000) = $433.25. The resulting understatement of $57.65625 per exposure is the shared claims cost of $46.125 per exposure amplified by the inverse of the 0.80 denominator. This illustrates why a genuine claims cost omitted from both the numerator and loss data creates an inadequacy greater than the original omitted expense when premium-proportional costs and targets are applied.

Stress tests and elasticities

Holding all else fixed, an increase in the probability of extra legal/technical service from 12% to 20% raises the direct average from $288 to $360 and total direct expense from $216,000 to $270,000. The indicated rate increases by ($54,000)/(0.80 × 10,000) = $6.75, to $497.65625. A 10% increase in the capacity component raises shared expense by $42,000 and indicated rate by $5.25, to $496.15625. If both changes occur simultaneously and are independent, the combined uplift is $12.00 and the indicated rate is $502.90625. These are conditional sensitivities, not statistical confidence limits.

A 12% increase in indemnity severity with the same number of files and service activities increases indemnity cost by $360,000, but does not mechanically increase the modeled direct or shared adjustment expense. The indicated rate becomes $535.90625. If the initial 22.575% expense-to-loss ratio had instead been applied uncritically to the higher indemnity, it would project an additional $81,270 of claims expense despite the held-constant service assumptions. This counterfactual isolates the cost-driver error in a loss-proportional formula. Different circumstances, particularly fee schedules indexed to damages or more complex large-loss litigation, could justify a positive cost response; the relationship must be supported by experience.

For model calibration and practical governance, it is useful to compute the derivatives of the simplified indicated rate: ∂r/∂D = ∂r/∂U = 1/[E(1-v-p)] when D and U denote total portfolio dollars, and ∂r/∂v = (L+D+U+F)/[E(1-v-p)^2] for a change in the premium-proportional loading. The equations show that erroneous attribution between two separately identified claims expense components does not alter the aggregate indication if their sum is preserved, yet can materially distort segment relativities or future projected trends. Classification accuracy is therefore both a financial reporting and a distributional pricing issue.

Frequency segmentation, credibility, and tail service costs

A portfolio can be segmented by jurisdiction, coverage, claim type, adjuster channel, litigation status, severity class, or catastrophe versus non-catastrophe handling. A direct fee component may show high skewness from occasional multi-year defense cases. Its mean is sensitive to large observations, and independent sampling of open-claim expense can overrepresent long-running complex cases. Count-based estimates should account for censoring of ultimate expense and selection of claims still pending. Statistical models can use log or other appropriate links, two-part models for a positive expense probability and positive severity, or hierarchical partial pooling across segments. Whatever the method, its validation target remains future per-exposure expenditure and calibration to actual aggregate financial cost.

A claims department may change settlement behavior in response to litigation incentives, audit technology, or the contractual control of defense. The settlement relationship studied by Crocker and Tennyson (2002) illustrates why frequency and severity can interact with economic behavior. It does not establish a universally beneficial claims settlement strategy. Actuarial expense predictions should evaluate both external handling costs and any induced changes in indemnity, delay, complaints, or litigation incidence. Cost reduction that increases later losses can be economically adverse even when the direct expense ledger falls.

Development and Expense Projections

Historical periods and the prospective matching problem

An insurer's accounting-year cash outflow for claims handling can relate to a wide span of accident and policy years. A rate effective next year must instead provide for future policy exposure and the associated ultimate cost of handling claims throughout their lifetime. When historical accounting-year expenses are divided by historical premium, paid loss, or closed counts, the resulting ratio mixes different settlement cohorts and varying durations. The ratio is easiest to interpret in a stable, stationary portfolio; it requires explicit corrections when claim frequency, settlement speed, inflation, outsourcing, reserving practice, or portfolio mix change.

The data plan must reconcile claim transaction dates, accident and report dates, policy year, line of business, incurred and paid expense, cost-center allocations, and exposure units. The ASOP 23 data quality standard (ASB, 2016) provides a framework for selecting, reviewing, and disclosing material limitations of data. Claims and financial ledgers often use different identifiers and cutoffs. The objective is an explainable bridge between their aggregate dollars, not a spurious appearance of perfect individual allocation. Data limitations should be quantified when possible, including unmatched records, delayed postings, retroactive reclassifications, and unavailable expense categories.

A practical retrospective series may preserve three separate components: ultimate direct expenses by claim cohort, allocated shared operating costs by the workload generated by those cohorts, and truly prospective infrastructure changes supported by an approved operating plan. One may develop direct expenses from paid and case data, estimate shared unit costs from projected staff hours, and attach a catastrophe handling allowance where normal experience excludes surge capacity. The choice is case-specific. In the CAS manual's discussion of catastrophe LAE, exceptional catastrophe settlement operations can distort normal expense ratios, so an explicit catastrophe component can be more faithful than indiscriminate inclusion in a routine estimate.

Development methods and claims settlement states

A paid-expense triangle accumulates observed ALAE or DCC payments for successive claim cohorts. Chain-ladder techniques develop observed expenditures to ultimate under relatively stable development patterns. A separate expected-cost method may be preferable for immature cohorts or structural changes. Count-based models measure claims reported, still open, reopened, and closed, weighted by activity and expected remaining duration. Under a transparent workload approach, projected future annual shared cost can be assigned to future policies according to the settlement activity their claims create over future years. The gross total must include settlement cost after expiry of the coverage period when it is part of the expected obligation.

The ASOP 43 unpaid-claim standard (ASB, 2007) is relevant to the methodology for claims arising from past events, explicitly including adjustment expenses in the unpaid claim estimate, but excludes estimates developed solely for ratemaking. It should therefore inform the calibration of unpaid claims obligations without being described as the direct professional standard governing a prospective rate indication. ASOP 53 supplies the forward-looking professional context. This distinction resolves a common misreading of actuarial standards and preserves the separate objectives of reserving and pricing.

The research by Johnson (1989), Allen and Mango (1999), and Ohlsson (2016) offers alternative ways to relate claim state and department workload to unallocated expense liability. Their reserving models provide useful conceptual ingredients, while prospective ratemaking requires a further transition to expected future written or earned exposure and the operating environment in which its claims will be handled. Importing a reserve methodology without this transition risks systematically misplacing the cost of current versus future cohorts.

Expense trends, social inflation, and operational changes

Claim adjustment expenditure can be exposed to wage inflation, legal fee schedules, medical cost increases, court congestion, claims handling regulation, cyber incidents, catastrophe response charges, and technology investment. Each has different effect on activity counts and unit costs. The ASOP 13 trending guidance (ASB, 2009, revised 2011) supports consideration of trends in the context of the relevant actuarial purpose. The ASOP 29 expense trend provisions explicitly contemplate expense trends that differ from trends in premium, insured loss, and exposure.

Suppose the synthetic baseline has $216,000 direct expense and $461,250 shared expense. If direct service fees rise 7% while shared costs rise 3%, with all service volumes unchanged, projected expense becomes $231,120 + $475,087.50 = $706,207.50. This adds $28,957.50 to the claims expense forecast and raises the indicated rate by $3.6196875. In a rate filing, the analyst must state whether the base period expense figures have already been trended. Applying inflation to both a developed-to-future expense estimate and its source activity rates would double adjust the same cost. Trend and development periods should be explicit and coherent with expected settlement dates.

Changes to claims technology merit an economic rather than promotional treatment. An automated intake system may lower the variable time per ordinary claim, increase fixed software costs, create migration expenditure, or improve fraud detection and alter indemnity. A business case should distinguish implementation costs, ongoing software fees, staff productivity, claims accuracy, customer outcomes, and the timing of savings. A premium indication is a forward-looking cash cost estimate, so estimated benefits require an operationally feasible rollout, not merely a budget aspiration. For short-run pricing the capacity floor may prevent immediate savings despite improved processing speed.

Discounting and cross-system consistency

Claims adjustment expenses are paid at different future dates. In a present-value calculation for a defined purpose, a payment vector C_t may be valued as PV(C) = Σ_t C_t/(1+i_t)^t using purpose-appropriate discount factors. The revised ASOP 20 (ASB, December 2025; effective June 1, 2026) covers property/casualty cash flow analysis, including ratemaking studies when a discounted cash-flow analysis is performed. A present-value method must respect the adopted accounting, statutory, and pricing purpose; the risk margin, financing assumptions, and claim-payment timing cannot simply be imported from a different measure.

For a synthetic nominal $100,000 of future expense payable 60% at the end of year one and 40% at the end of year two, a flat 4% annual discount rate produces $60,000/1.04 + $40,000/(1.04)^2, or approximately $94,674.56. The difference from $100,000 reflects the specified timing and rate assumption, not a universal permission to discount insurance rates. A U.S. statutory unpaid liability has its own applicable recognition and discounting restrictions; SSAP No. 55's provisions should be read in that context. In European technical provisions, the Solvency II best-estimate rule instead expressly uses present-value future cash flows. IFRS 17 separately combines current expected cash flows with the relevant risk adjustment and measurement structure. The consistent method is to prepare one reliable schedule of expected cash expenses, and then derive each required valuation basis without blending their discounting and margin rules.

Validation and uncertainty

An actuarial projection should be checked through development backtests, emerging paid and incurred expense comparisons, claims volume and duration studies, and reconciliation with actual department payroll and third-party expenditure. Diagnostic tests can distinguish pure volume error from activity intensity error, claim severity-linked cost error, and budget variance. The intended measure should be stated: expected mean expense for prospective cost estimation is different from the total premium that also includes a chosen profit or capital allowance. An unfavorable realization of expense after pricing does not necessarily mean the original expectation was unreasonable; systematic bias in repeated forecasts does.

Sources of uncertainty include incomplete claims expense coding, shifts in claim mix, legal regime, inflation, technology performance, catastrophe clustering, tail development, model parameter estimates, and behavioral response to settlement policy. A defensible study records the exposure and claim periods, data exclusions, development horizon, and quantitative sensitivity to assumptions. Material model limitations should be conveyed in human-readable terms. The synthetic scenarios in this monograph are deterministic demonstrations of formulas; they cannot estimate confidence intervals or judge an actual insurer's rate adequacy.

Rate Adequacy and Cost Controls

Rate indication as a reconciliation of obligations and service production

A technical indication has two related tasks: predict future risk transfer payments and predict the operational activities necessary to perform the insurer's obligations. Premium adequacy depends on the combined prospective cost and target, while rate fairness across classes depends on credible differentiation of cost drivers. Allocating shared claims salaries by last year's premium may preserve an enterprise total, but it can assign more shared handling cost to a high-value, low-frequency class than to a high-frequency, modest-premium class without an activity justification. Such a practice may satisfy a bookkeeping allocation but not the actuarial question about prospective exposure-related expense.

Rate analysis should disclose whether ALAE or DCC is embedded in the loss estimate, whether shared AO is in a separate provision, and which expense amounts are genuinely additional. Under ASOP 29's claims expense treatment, the residual expense provision addresses all relevant adjustment costs that have not already entered the loss component. This is an accounting conservation rule with substantive economic implications. It protects aggregate adequacy against omissions and double counting while giving analysts freedom to select a mixture of ratios, activity drivers, and cohort projections supported by data.

The relationship to reinsurance depends on contract terms. A reinsurer may reimburse some indemnity but not equivalent defense or settlement cost; alternatively, a loss corridor or quota share arrangement may share claims expenses in a specified manner. The gross economic cost of handling claims, ceded recovery, reinsurance-related servicing expense, and net retained cost should be displayed separately. Otherwise, a net indemnity ratio applied to gross claims expense can distort the primary price. Tax and assessment rules, including treatment of defense costs, are jurisdiction-specific and require local checking at the actual filing stage.

Cross-jurisdictional interpretation

The U.S. actuarial language supplies convenient categories and methods, but its exact filing conventions do not constitute a universal legal accounting framework. The NAIC's DCC/AO classification and SSAP 55 serve a statutory reporting role. U.S. rate regulation is state-specific, and the use of actuarial standards is governed by their scope and applicable law. In Solvency II valuation, claims management and overhead expenses are components of the best estimate under the delegated regulation. IFRS 17's insurance-contract measurement principles connect expected settlement and operating cash flows to the measurement of insurance contract groups. None of these sources prescribes a single numerical cost ratio for all markets. Their common analytical lesson is the importance of completeness, traceability, and purpose-specific measurement.

A coherent multinational insurer may maintain one underlying operational expense dataset while reconciling it to distinct local statutory classifications, actuarial pricing components, and IFRS reporting allocations. The bridge should record changes in expenses and changes in classification separately. An accounting-standard change is a reason to recast the historical comparison before drawing a trend conclusion. An operational change is a reason to revisit projected activity and capacity. A legal change is a reason to update the claim process and any contractual obligation. These distinctions are especially important when combining entities or replacing claims platforms.

Evaluating cost-control initiatives

Reducing claims expense is economically desirable when the total expected cost of meeting insurance obligations falls without inducing disproportionate adverse claim outcomes. Outsourcing can convert fixed salary costs into per-claim fees and change the volatility of expense without necessarily reducing its expectation. Early settlement can reduce open-file administration and defense cost, but may alter indemnity, fraud incentives, or claims fairness. Medical cost containment can be an expense that generates a larger reduction in indemnity; its evaluation must incorporate both sides. Fraud investigation increases immediate adjustment costs for selected claims while possibly changing payments and future behavior. Cost-accounting metrics should therefore be examined jointly with ultimate indemnity, reopen rates, litigation, complaints, and timeliness.

The optimality problem is empirical and jurisdiction-dependent. The Crocker–Tennyson claims study demonstrates, within the authors' setting, why loss exaggeration incentives and expected litigation expenses affect settlement decisions. It does not provide authority for underpayment or establish a general claims policy. The responsible inference is to evaluate expected total claims outcomes, contractual obligations, and consumer protection alongside operational savings. A short-run reduction in recorded AO can represent deferred work rather than a genuine improvement if the pending backlog increases.

A practical evaluation may estimate the change ΔT = ΔL + ΔD + ΔU + ΔF for total expected cost, with separate analysis of implementation cost and any altered uncertainty. A cost-control program that reduces U by $100,000 but increases expected ultimate L by $120,000 has a $20,000 adverse economic effect before changes in other costs. Conversely, an investment in independent medical review that raises D by $40,000 and credibly lowers expected insured payments by $90,000 has a favorable $50,000 expected net effect, subject to lawful and contractually sound execution. The equations assign no value judgment to the means; their admissibility must be assessed separately under the applicable legal framework.

Scientific synthesis and research boundaries

The examined literature supports five connected conclusions. First, the economic definition of claims adjustment expense is based on the activity of meeting claims obligations, while ALAE/ULAE and DCC/AO provide different classification perspectives. Second, the forecasting base should reflect expected future activity: loss dollars, claim counts, exposure, file duration, and capacity are suitable under different operational conditions. Third, reserve estimation contributes data and methods but cannot replace the future-period transformation required by ratemaking. Fourth, claims expenses can respond to inflation, legal incentives, portfolio shifts, and technical changes differently from indemnity. Fifth, coherent accounting reconciliation and prospectively validated activity models are mutually reinforcing rather than competing methods.

These conclusions synthesize established actuarial and accounting ideas. They do not claim new theoretical laws. They also specify important research needs: internationally comparable expense classifications; privacy-preserving measurement of claim-level handling effort; objective evaluation of technology's effect on ultimate cost; and longitudinal estimation of defense and settlement expense elasticity under changes in legal systems. Future empirical work can test such questions with insurer-level data and independent replication, subject to appropriate access and confidentiality constraints.

Conclusion

Claims adjustment expense provisions are a necessary and independently analyzable component of an adequate technical insurance premium. The prospective object is the full expected cost of administering and settling claims arising from the priced exposure, including claim-specific expenditures and the properly attributed cost of common operational capacity. Reliable estimation begins with a mutually reconciled expense population and proceeds through claim process measurement, cohort development, future activity projection, expense-specific trend, and a clear premium formula.

The historical ALAE/ULAE distinction organizes attributable and shared claims costs. The U.S. DCC/AO system distinguishes classes of expense activity and must be crosswalked rather than treated as a synonym. Statistical ratios are efficient approximations when the proportionality and timing assumptions are appropriate. Claim-frequency, service-intensity, duration, and operational capacity forecasts can provide superior causal alignment under changes in business mix or settlement arrangements. The worked examples show that omissions, duplicate expenses, and wrong driver assumptions change indicated rates materially even when the ledger totals appear plausible.

Regulatory and accounting comparisons establish that prospective rates, unpaid-claims provisions, IFRS 17 measurements, and Solvency II best estimates are related but distinct decisions. They can share the same underlying expected cash costs while applying different measurement conventions, discount rates, risk adjustments, and statutory constraints. For each purpose, scientific integrity requires an explicit bridge from observed claims to projected cash costs, documented assumptions, consistency tests, and disclosure of limitations.

Claims adjustment expense is a cost of keeping the insurance promise. Its disciplined inclusion in ratemaking makes the price more transparent, connects operational planning to actuarial expectations, and provides a sound basis for investigating claims management efficiency and risk-pooling fairness. The present research completes an analytical synthesis and transparent calculation set; it establishes no proprietary doctrine or market-wide empirical estimate.

References