Analyzing Loss Reserves

Analyzing Loss Reserves

Evaluates loss reserve adequacy using development triangle analysis, actuarial projection methods, and benchmark comparisons to determine whether carried reserves are reasonable, deficient, or redundant.

When To Use

• Reviewing an insurer's or reinsurer's Statement of Actuarial Opinion and supporting exhibits
• Assessing reserve adequacy during due diligence for M&A, commutation, or loss portfolio transfer
• Interpreting loss development triangles provided in Schedule P, statutory filings, or internal actuarial reports
• Comparing carried reserves against independent point estimates or range estimates
• Evaluating reserve changes period-over-period (favorable/adverse development) and their drivers
• Supporting reinsurance treaty pricing or reserve credit analysis

Inputs To Gather

• Loss development triangles: Paid and incurred triangles by accident year (or underwriting year/report year), ideally at 12-month intervals; note whether triangles are cumulative or incremental
• Earned premium and exposure data: By corresponding year and line of business
• Carried reserve balances: Case reserves, IBNR, and total by accident year and LOB
• Actuarial report or opinion: Including selected methods, assumptions, selected loss development factors (LDFs), and tail factors
• Line of business and claim type: Workers' comp, general liability, professional liability, auto, property, etc. — development patterns vary dramatically
• Benchmark data: Industry LDFs from sources such as AM Best, ISO/Verisk, or NAIC Schedule P industry aggregates [VERIFY: confirm which benchmark source is available and appropriate for the line]
• Prior analyses: Previous reserve studies or external auditor findings for trend comparison

Workflow

1. Validate triangle integrity

   • Confirm triangles are on a consistent basis (paid vs. incurred, cumulative vs. incremental)
   • Check that the latest diagonal ties to the balance sheet carried amounts
   • Identify any triangle adjustments (large-loss caps, commutation removals, currency conversions) and note their impact

2. Calculate age-to-age development factors

   • Compute link ratios for each development period across all accident years
   • Examine simple average, volume-weighted average, and medial (excluding high/low) selections
   • Identify accident years with unusually high or low factors — flag potential large-loss distortion, reserve strengthening, or claim settlement pattern changes

3. Apply standard actuarial projection methods

   • Chain Ladder (Development): Apply selected LDFs to the latest cumulative values; most reliable for mature, stable lines
   • Bornhuetter-Ferguson (BF): Blend development projection with an a priori expected loss ratio; preferred for immature accident years or volatile lines
   • Cape Cod (Stanard-Buhlmann): Use exposure-weighted expected losses; useful when loss ratios are expected to be stable across years
   • Frequency-Severity: Where claim count and average severity data are available, project separately; valuable for lines with known count trends
   • Select tail factors for development beyond the triangle's observed maturity [VERIFY: tail factor assumptions are highly judgment-dependent — confirm basis and reasonableness]

4. Develop ultimate loss estimates and ranges

   • Produce point estimates from each method by accident year
   • Weight or select among methods based on data credibility, line characteristics, and maturity
   • Construct a reasonable range (e.g., low/central/high) reflecting parameter uncertainty
   • Compare to the company's carried reserves — quantify redundancy or deficiency by accident year and in total

5. Analyze reserve development trends

   • Track prior-year development (favorable or adverse) over multiple calendar periods
   • Identify whether development is concentrated in specific accident years or lines
   • Assess whether development patterns indicate systematic under- or over-reserving
   • Consider external drivers: legal environment changes, inflation, claim handling practice shifts

6. Benchmark and stress test

   • Compare company LDFs and ultimate loss ratios to industry benchmarks for the same line and maturity
   • Test sensitivity to alternative tail factors, LDF selections, and expected loss ratio assumptions
   • Quantify the reserve impact of plausible adverse scenarios (e.g., social inflation, latent exposure emergence)

Output

Structure the analysis report as follows:

• Executive Summary: Overall reserve adequacy conclusion (adequate / likely deficient / likely redundant), magnitude of estimated surplus or shortfall, key risk factors
• Data and Scope: Lines of business, accident years, triangle basis, and any data limitations
• Development Factor Analysis: Table of age-to-age factors with selected factors and basis for selection; highlight anomalies
• Ultimate Loss Projections: Table by accident year showing results from each method, selected ultimates, and comparison to carried reserves
• Reserve Adequacy Assessment: Quantified redundancy/deficiency by year and total; confidence range; discussion of key judgment areas
• Development History: Summary of favorable/adverse development trends and their implications
• Sensitivity Analysis: Impact of alternative assumptions on the adequacy conclusion
• Limitations and Caveats: Data gaps, areas of high uncertainty, reliance on third-party information

Quality Checks

• Confirm all triangle arithmetic is internally consistent (incremental values sum to cumulative; latest diagonal matches reported data)
• Verify that selected LDFs fall within a reasonable range relative to historical averages and industry benchmarks — outlier selections require explicit justification
• Ensure BF and Cape Cod a priori loss ratios are sourced and documented, not assumed without basis
• Check that tail factors are reasonable for the line's claim closure characteristics [VERIFY: long-tail lines like workers' comp or environmental liability may need tails extending 20+ years]
• Confirm the analysis addresses both paid and incurred bases — significant divergence between paid and incurred projections should be explained
• Validate that large losses or one-time events are identified and their treatment (capped, excluded, separately developed) is clearly stated
• Ensure the adequacy conclusion accounts for discount effects if reserves are on a present-value basis [VERIFY: confirm whether reserves are discounted and the applicable discount rate/standard]
• Flag any areas where actuarial judgment materially drives the result and a qualified actuary should review