5. General Insurance Claims and Reserve Accounting

Introduction and Context

General insurance accounting is shaped by a fundamental timing problem: premium is often received before the insurer has provided the complete period of protection, while claims may be reported, assessed, and settled long after the insured event occurs. Financial statements must therefore distinguish between revenue earned and revenue merely received, and between claims already paid and the insurer’s best estimate of claims that will ultimately be paid.

Reserve accounting addresses this timing and uncertainty. Unearned premium reserves represent future insurance service still owed to policyholders. Claims reserves represent expected future payments relating to insured events that have already occurred. The quality of these estimates affects profitability, solvency, pricing, reinsurance decisions, capital management, and the credibility of financial reporting.

For a postgraduate insurance manager, reserve accounting is not merely a bookkeeping exercise. It is an actuarial, financial, operational, and governance process involving claims data, legal interpretation, statistical modelling, inflation assumptions, discount rates, internal controls, and professional judgement.

Learning Objectives

After studying this chapter, the reader should be able to:

1. Explain earned and unearned premium and their relationship with insurance service.

2. Distinguish reported claims, case reserves, incurred-but-not-reported claims, and incurred claims.

3. Apply major claims reserving methods and interpret claims development triangles.

4. Explain frequency-severity models and the effects of inflation, discounting, and changing claims conditions.

5. Evaluate reserve uncertainty, adequacy, governance, and control mechanisms.

6. Assess the managerial and regulatory implications of reserve estimates.

7. Use analytical evidence to identify reserve deterioration and potential under-reserving.

The Accounting Logic of Premium and Claims

Unearned premium and the passage of risk

A policy premium is normally collected at or near inception, but the insurer earns it as coverage is provided. If a twelve-month policy begins on 1 October and the reporting date is 31 December, only three months of the insurance service have generally been provided. The remaining premium relates to future coverage and is therefore unearned.

The unearned premium reserve (UPR) is the portion of written premium allocated to future periods. Under a simple daily pro-rata method:

UPR = Written premium × Unexpired coverage days / Total coverage days

For example, if a policy premium of ₹120,000 covers 365 days and 274 days remain at year-end, the unearned amount is approximately ₹90,082. The earned premium is the written premium less the closing UPR, subject to adjustments for cancellations, endorsements, acquisition costs, and the applicable financial reporting framework.

The pro-rata method is convenient but not always appropriate. Risks such as crop, construction, marine cargo, or catastrophe-exposed property may not be uniform throughout the policy period. In those circumstances, an exposure-based or risk-profile approach may better represent the pattern of service and expected claims.

Claims liabilities

Claims accounting begins with the distinction between paid and unpaid claims. Paid claims are cash amounts already disbursed. Unpaid claims are obligations arising from insured events that remain to be settled. The unpaid claims reserve is commonly divided into:

• Case reserves, established for claims that have been reported and individually assessed.
• IBNR reserves, relating to claims incurred but not reported, together with possible development on reported claims not fully captured in case estimates.

The total incurred claim cost for a period can be expressed as:

Incurred claims = Paid claims + Closing outstanding claims − Opening outstanding claims

This formula explains why an insurer may report high incurred claims even when cash payments are modest. A significant increase in outstanding reserves produces an expense in the current period because the insurer has recognized additional obligations.

Important Concepts and Terminology

A claim occurrence date is the date on which the insured event happens. The reporting date is when the insurer receives notification. The settlement date is when payment is made or the claim is otherwise closed. The period between these dates creates reporting and development uncertainty.

A reported claim is known to the insurer, whether or not it has yet been fully investigated. A closed claim has been settled, rejected, withdrawn, or otherwise finalized. An open claim remains subject to further payment or adjustment. Case outstanding is the individual estimate recorded against an open reported claim.

IBNR conventionally means incurred but not reported claims. In practice, a broader statistical reserve may also include IBNER, incurred but not enough reported, representing future development on claims already reported but currently under-reserved. Accordingly, reserve specialists often distinguish pure IBNR from the total reserve required beyond recorded case estimates.

Ultimate claims are the final expected costs of claims arising from a defined accident or underwriting period. Loss development is the change in cumulative reported or paid losses as claims mature. Loss ratios compare claims with premium, usually on an earned or written basis. Claims frequency is the number of claims per exposure unit, while claims severity is the average cost per claim.

Reserve adequacy means that the recorded liability is sufficient, on the selected accounting and valuation basis, to meet expected obligations and associated expenses. Reserve deficiency occurs when later experience indicates that earlier reserves were insufficient. Reserve redundancy occurs when reserves prove higher than eventual requirements, although redundancy is not automatically evidence of superior management because it may reflect excessive conservatism or changing conditions.

From Individual Files to Portfolio Estimates

Case reserving

Case reserving is the primary operational method for reported claims. Claims handlers review policy coverage, liability, injury or damage, legal costs, recovery prospects, and likely settlement value. The reserve is then updated as evidence develops.

Case reserves are valuable because they use claim-specific information. They are particularly important for large or unusual claims. However, they can be inconsistent across handlers, influenced by settlement expectations, affected by performance targets, or slow to reflect systemic changes. A case reserve also cannot capture claims that have not yet been reported.

Strong case reserving requires documented authority levels, peer review for material claims, reserve update triggers, separation of claims handling and financial approval where appropriate, and clear treatment of salvage, subrogation, deductibles, and reinsurance recoveries.

Statistical reserving

Statistical methods estimate future development from historical patterns. They are most reliable where data is sufficiently homogeneous and operational conditions are reasonably stable. No method eliminates judgement; rather, the method makes assumptions explicit and provides a disciplined challenge to individual estimates.

#### Chain-ladder method

The chain-ladder method projects cumulative claims using historical development factors. Claims are arranged by accident year, underwriting year, or report year and by development age. A simplified cumulative paid triangle may appear as follows:

| Accident year | 12 months | 24 months | 36 months |

|---|---:|---:|---:|

| 2021 | 40 | 70 | 80 |

| 2022 | 45 | 78 | — |

| 2023 | 50 | — | — |

The selected factor from 12 to 24 months and from 24 to 36 months is applied to immature observations to estimate ultimate claims. The method assumes that historical development patterns remain relevant. It can be distorted by inflation, claims process changes, portfolio growth, exceptional large losses, or changes in case reserving philosophy.

#### Bornhuetter–Ferguson method

The Bornhuetter–Ferguson approach combines an a priori expected loss ratio with the proportion of claims expected to emerge by each development age. It is less sensitive than chain-ladder to unusually low early reported claims and is often useful for immature accident years or portfolios with limited credible experience.

Conceptually:

Ultimate loss = Reported loss + Expected future loss

Expected future loss is calculated from the expected ultimate loss and the proportion not yet reported or paid. The method introduces external underwriting information, such as pricing assumptions, exposure growth, and expected loss ratios.

#### Expected loss ratio method

Under this method:

Expected ultimate claims = Earned premium × Expected loss ratio

It is useful when claims data is immature or unstable, but it depends heavily on the quality of the expected loss ratio. It should generally be reviewed against emerging experience rather than applied mechanically.

#### Frequency-severity approach

A frequency-severity model estimates the number and cost of claims separately:

Expected claims cost = Expected claim frequency × Expected average severity

Frequency may be modelled using exposure measures such as vehicles, policies, payroll, or insured values. Severity may be adjusted for inflation, limits, deductibles, litigation, and claim-size distribution. The approach is particularly useful for pricing, scenario analysis, and portfolios in which claim counts and average costs respond differently to economic or operational conditions.

Large claims often require a separate treatment because the average severity can be highly volatile. A portfolio may therefore be divided into attritional, large, and catastrophe claims, with different assumptions and governance.

Claims Development Triangles and Interpretation

A development triangle is a matrix in which rows represent origin periods and columns represent development ages. It may contain paid claims, reported claims, claim counts, or incurred claims. Triangles can be organized by accident year, underwriting year, or report year. The selection must reflect the business and the purpose of analysis.

Paid triangles are less affected by case reserve philosophy but may develop slowly. Incurred triangles provide earlier information but inherit changes in case reserving practice. Reported claim-count triangles can reveal frequency and reporting delays, while paid-count triangles can assist in studying closure rates.

The analyst should examine not only selected development factors but also:

• changes in claim reporting speed;
• shifts in claim closure rates;
• changes in policy limits and deductibles;
• legal and regulatory developments;
• inflation and repair-cost trends;
• catastrophe or large-loss distortions;
• changes in claims systems or coding;
• reinsurance and commutation effects.

A triangle is not a physical law. It is a historical record whose usefulness depends on comparability. Mature data may still be misleading if the underlying process has changed.

Inflation, Discounting, and the Time Value of Claims

Claims reserves must distinguish between nominal future payments and their value at the reporting date. Inflation may affect medical costs, wages, vehicle parts, construction materials, legal awards, and claims-handling expenses. Social inflation—expanded liability interpretations, litigation funding, higher settlement expectations, or broader judicial awards—may be difficult to observe directly but can materially affect long-tail liabilities.

If future payments are expected to be ₹10 million in three years and the applicable annual discount rate is 6%, the present value is approximately:

₹10,000,000 / (1.06³) = ₹8,396,000

The selected discount rate must be consistent with the applicable accounting framework, liability characteristics, currency, duration, and risk treatment. Discounting may be inappropriate or restricted for short-tail claims where the time value is immaterial, or where standards require a particular presentation.

Inflation and discounting operate in opposite directions but should not be casually netted. A reserve model should first project future payments in an appropriate inflationary basis and then apply a consistent discounting methodology. Double-counting inflation or applying a nominal rate to real cash flows can materially distort the liability.

Reserve Uncertainty and Adequacy Testing

A reserve estimate is not a single certain number. It is an estimate surrounded by process risk, parameter risk, model risk, and external uncertainty. Process risk arises from inaccurate claim coding or delayed reporting. Parameter risk concerns selected development factors, inflation rates, or expected loss ratios. Model risk occurs when a method fails to represent the portfolio. External uncertainty includes court decisions, catastrophes, economic changes, and emerging claims patterns.

Adequacy testing should combine quantitative and qualitative evidence. Common techniques include:

1. Back-testing: comparing prior selected reserves with actual subsequent development.

2. Actual-versus-expected analysis: identifying unexpected claim frequency, severity, or reporting patterns.

3. Method comparison: comparing chain-ladder, Bornhuetter–Ferguson, frequency-severity, and case-based estimates.

4. Sensitivity testing: measuring the effect of changes in development factors, inflation, discount rates, or claim assumptions.

5. Scenario analysis: considering adverse legal, economic, catastrophe, or operational events.

6. Range analysis: presenting a central estimate together with reasonable lower and upper outcomes.

A reserve range is not necessarily a probability distribution unless the methodology supports that interpretation. Management should avoid presenting false precision. Material uncertainty should be communicated to the board, audit committee, regulator, and other authorized users.

An Illustrative Case Study: Motor Claims Reserve Deterioration

Background

Illustrative insurer Meridian General writes private motor insurance. At the end of Year 1, it records earned premium of ₹500 million and an outstanding claims reserve of ₹210 million. The portfolio has historically been short-tail, but recent supply-chain disruptions have increased vehicle repair times and parts prices.

Situation

During Year 2, paid claims are higher than expected. Claim frequency is broadly stable, but average severity has increased by 14%. The claims department has also introduced a new digital settlement process, accelerating the closure of straightforward claims while leaving complex bodily injury claims open for longer. A simple chain-ladder analysis suggests a reserve of ₹235 million, while the Bornhuetter–Ferguson method indicates ₹248 million. The case reserve inventory totals ₹225 million, but the actuarial review identifies potential IBNER of ₹28 million.

Analysis

The difference between the methods is not merely a technical disagreement. The chain-ladder method may understate the effect of the recent severity trend because historical development factors reflect lower repair inflation. The case inventory may also appear stronger because simple claims are closing quickly, whereas complex claims are becoming more concentrated among the open files. The frequency-severity analysis confirms stable frequency but materially higher severity.

The reserve committee reviews repair-cost indices, claim closure patterns, bodily injury settlement durations, large-loss files, and salvage recoveries. It separates attritional claims from large bodily injury claims and tests a higher inflation assumption. It also checks whether the new digital process has changed the meaning of development ages in the triangle.

Approach and solution

Management selects a central reserve of ₹253 million, supported by a documented rationale rather than by an automatic average of competing methods. It records an additional provision over the case inventory, commissions a focused review of bodily injury claims, and introduces quarterly severity monitoring. The actuarial function prepares a sensitivity showing the effect of a further 5% increase in bodily injury severity. The board risk committee receives the reserve range and the main drivers of uncertainty.

Outcome

In the following year, the reserve develops adversely by ₹9 million, but the deterioration is within the previously communicated range. The insurer revises its repair-cost assumptions, improves claim segmentation, and incorporates operational changes into future development selections.

Learning points

The case demonstrates that reserve adequacy cannot be judged by one method or by comparison with case reserves alone. Operational changes can invalidate historical development patterns. Stable claim frequency does not imply stable claims cost. Finally, governance improves when assumptions, uncertainty, and management actions are documented together.

Governance, Controls, and Professional Responsibilities

Reserve governance should assign clear responsibility among claims management, actuarial functions, finance, risk management, internal audit, senior management, and the board. The first line creates and maintains claim information; the actuarial or specialist function independently estimates liabilities; finance ensures appropriate accounting; risk management evaluates uncertainty and capital implications; internal audit tests control design and operation.

Important controls include data reconciliation between claims systems and the general ledger, authorization of reserve changes, review of material claims, access controls, validation of model code, version control, documented assumption approval, and post-period reserve monitoring. A reserve committee should record the basis for method selection, changes from prior periods, rejected alternatives, and unresolved disagreements.

Ethically, reserve estimates must not be manipulated to smooth earnings, meet bonus targets, defer recognition of deterioration, or create hidden cushions without transparent justification. Excessive optimism may mislead policyholders, investors, regulators, and reinsurers; excessive conservatism may distort pricing and unfairly transfer performance between periods. Professional judgement must be evidence-based, consistent, and open to challenge.

Regulatory and accounting requirements vary by jurisdiction. Insurers must distinguish the purpose of a regulatory technical provision, a financial reporting liability, an internal economic estimate, and a pricing provision. These bases may differ, but reconciliations and explanations should be available. Reinsurance recoveries should also be assessed separately for collectability and counterparty risk; a gross claims liability cannot automatically be reduced merely because a contractual recovery is expected.

Technology, Analytics, and Emerging Practice

Modern claims platforms can link policy, exposure, payment, litigation, repair, and customer data. Automated alerts can identify claims with reserve changes outside tolerance, unusually long reporting delays, repeated payments, or inconsistent coding. Predictive models may assist triage and severity estimation, but they do not remove the need for actuarial oversight.

Artificial intelligence can support document classification, fraud screening, liability assessment, and claims summarization. Its use creates risks involving biased training data, lack of explainability, privacy, cyber security, and inappropriate automation. A model that performs well on historical data may fail after a legal change or catastrophe. Human review, model validation, audit trails, and controlled override procedures are therefore essential.

Emerging reserving practice increasingly uses granular data, claim-level models, stochastic simulation, and scenario-based capital analysis. These tools can improve insight, but sophisticated modelling is not automatically superior. A transparent, well-governed model using reliable data is more useful than a complex model whose assumptions users cannot understand.

Best-Practice Principles

Effective reserve accounting generally follows these principles:

• define the valuation objective and accounting basis before selecting a method;
• reconcile premium, exposure, claims, and payment data;
• segment portfolios into reasonably homogeneous risk groups;
• use more than one method where uncertainty is material;
• investigate changes in claims operations before relying on historical triangles;
• model inflation, discounting, recoveries, and expenses consistently;
• distinguish central estimates from uncertainty ranges;
• back-test estimates and document lessons from adverse development;
• maintain independent challenge and appropriate approval levels;
• communicate material assumptions and limitations clearly.

Chapter Summary

General insurance reserve accounting reflects the mismatch between the timing of premium collection, insurance service, claim reporting, and claim settlement. Unearned premium reserves defer revenue relating to future coverage, while case reserves and IBNR estimates recognize obligations arising from insured events.

Claims reserving combines individual claim assessment with statistical methods. Chain-ladder relies on historical development, Bornhuetter–Ferguson combines emerging experience with an expected loss view, and frequency-severity models separate claim counts from claim costs. Claims development triangles are powerful analytical tools but must be interpreted in light of inflation, legal change, operational developments, and data quality.

Reserve adequacy requires back-testing, sensitivity analysis, method comparison, scenario analysis, and professional judgement. Discounting and inflation must be applied consistently, and reserve uncertainty should be communicated rather than hidden behind false precision. Strong governance, independent challenge, reliable systems, ethical conduct, and controlled use of analytics are essential to credible reserve accounting.

Key Terms

Unearned Premium Reserve: The portion of written premium relating to insurance coverage not yet provided.

Earned Premium: Premium allocated to the period during which insurance protection has been provided.

Case Reserve: An individual estimate of the unpaid cost of a reported claim.

IBNR: Claims incurred but not reported to the insurer at the valuation date.

IBNER: Future development on claims already reported but inadequately reserved.

Outstanding Claims: Unpaid obligations relating to claims that have occurred.

Ultimate Claims: The final expected cost of claims for a defined origin period.

Claims Development Triangle: A matrix showing cumulative claims or counts by origin period and development age.

Chain-Ladder Method: A reserving method projecting immature claims using historical development factors.

Bornhuetter–Ferguson Method: A method combining reported experience with an a priori expected loss estimate.

Expected Loss Ratio Method: A method estimating ultimate claims from earned premium and an expected loss ratio.

Frequency: The number of claims relative to an exposure measure.

Severity: The average cost per claim.

Discounting: Converting expected future payments into present value.

Social Inflation: Increases in liability costs associated with legal, social, and claims-environment changes.

Reserve Adequacy: The sufficiency of recorded reserves to meet expected obligations on the relevant basis.

Reserve Deficiency: A shortfall arising when recorded reserves are below eventual required payments.

Back-Testing: Comparing earlier reserve estimates with subsequent claims development.

Reserve Governance: The structures, controls, responsibilities, and oversight governing reserve estimates.

Review Questions

Short Answer Questions

1. Why is written premium not necessarily equal to earned premium?

2. Distinguish between a case reserve and an IBNR reserve.

3. What information is presented in a claims development triangle?

4. State the basic relationship between claim frequency, severity, and expected claims cost.

5. Why can historical development factors become unreliable?

Long Answer Questions

1. Explain the accounting treatment and managerial importance of unearned premium reserves.

2. Compare the chain-ladder, Bornhuetter–Ferguson, and expected loss ratio methods.

3. Discuss the principal sources of uncertainty in general insurance claims reserves.

4. Explain how inflation and discounting should be incorporated into long-tail claims estimates.

5. Describe an effective governance framework for claims reserve estimation and approval.

Analytical / Application Questions

1. An insurer has written premium of ₹24 million for a twelve-month policy portfolio, with six months of coverage remaining at the reporting date. Calculate the approximate unearned premium under a pro-rata method and explain the assumptions involved.

2. A motor insurer reports stable claim frequency but a 12% increase in average severity, while its historical triangle indicates improving loss development. Which evidence should management prioritize, and what additional tests should be performed?

3. A reserve committee receives a case reserve estimate of ₹100 million, a chain-ladder estimate of ₹106 million, and a Bornhuetter–Ferguson estimate of ₹118 million for an immature portfolio. Explain why selecting the simple average may be inappropriate and recommend a decision process.

22. Multiple Choice Questions with Answers

Question 1

A policy has a written premium of ₹120,000 for 365 days, with 274 days of coverage remaining at the reporting date. What is the approximate unearned premium under a daily pro-rata method?

A. ₹30,000

B. ₹60,000

C. ₹90,082

D. ₹120,000

Correct Answer: C. ₹90,082

Question 2

Which item primarily represents future development on claims already reported but inadequately reserved?

A. Pure IBNR

B. IBNER

C. Unearned premium

D. Earned premium

Correct Answer: B. IBNER

Question 3

Under the claims accounting relationship, incurred claims equal:

A. Paid claims plus opening outstanding claims minus closing outstanding claims

B. Paid claims plus closing outstanding claims minus opening outstanding claims

C. Written premium minus paid claims

D. Closing outstanding claims minus earned premium

Correct Answer: B. Paid claims plus closing outstanding claims minus opening outstanding claims

Question 4

Which reserving method combines reported claims experience with an a priori expected loss estimate?

A. Chain-ladder

B. Bornhuetter–Ferguson

C. Pure case reserving

D. Payment allocation method

Correct Answer: B. Bornhuetter–Ferguson

Question 5

A portfolio has stable claim frequency but sharply rising average claim cost. Which model decomposition best highlights this development?

A. Frequency-severity analysis

B. Unearned premium analysis only

C. Claim closure analysis only

D. Written-premium reconciliation only

Correct Answer: A. Frequency-severity analysis

Question 6

What is a major weakness of relying mechanically on historical chain-ladder factors?

A. They cannot use cumulative data

B. They ignore all reported claims

C. They may be distorted by changes in claims processes or inflation

D. They can only be applied to paid claims

Correct Answer: C. They may be distorted by changes in claims processes or inflation

Question 7

If ₹10 million is payable in three years and the annual discount rate is 6%, its approximate present value is:

A. ₹10.60 million

B. ₹10.00 million

C. ₹9.40 million

D. ₹8.40 million

Correct Answer: D. ₹8.40 million

Question 8

Which activity is an example of reserve back-testing?

A. Comparing current premium with written premium

B. Comparing previous reserve selections with subsequent actual development

C. Estimating policy acquisition costs

D. Calculating the number of policies issued

Correct Answer: B. Comparing previous reserve selections with subsequent actual development

Question 9

Why may rapid closure of simple claims create a misleading impression in a development triangle?

A. It can leave a concentration of complex claims developing more slowly

B. It eliminates all reserve uncertainty

C. It converts earned premium into written premium

D. It prevents inflation from affecting claims

Correct Answer: A. It can leave a concentration of complex claims developing more slowly

Question 10

Which governance practice best supports credible reserve accounting?

A. Allowing claims handlers to approve all final reserves without review

B. Selecting the lowest available estimate to improve reported profit

C. Documenting assumptions, applying independent challenge, and monitoring subsequent development

D. Replacing all professional judgement with an automated model

Correct Answer: C. Documenting assumptions, applying independent challenge, and monitoring subsequent development