Workers' Compensation IBNR After a Claims-Handling Consolidation
A monoline workers' compensation insurer writes in three states. For accident year 2023 at 12 months, paid losses are $12.4M, case reserves are $18.1M, and reported claims are 1,150. Compared with accident year 2022 at the same age, reported counts are down 8%, but average case reserves are up 22%. The claims department consolidated two regional offices in mid-2023 and assigned all new lost-time claims to a central unit. Closed counts are 610 versus 790 a year earlier. The chief actuary's prior expected loss ratio for 2023 is 80% on earned premium of $42M. The latest chain-ladder age-to-age factors show moderate development, but the case-reserve-to-paid ratio is 1.46 versus 1.18 for 2022. Management asks whether the 2023 IBNR should be reduced because fewer claims are reported, or increased because case reserves rose.
1. Diagnose the likely sources of the conflicting signals and state which reserve method you would select for accident year 2023, with reasoning.
Read the worked answer
The count decline is not credible evidence of a frequency improvement by itself. The office consolidation and central assignment of lost-time claims can delay reporting, change closing practices, and alter case-reserve philosophy. A rising average case reserve and a case-reserve-to-paid ratio of 1.46 versus 1.18 suggest either severity deterioration or deliberate case strengthening. Chain ladder is vulnerable here: if case reserves were strengthened, reported loss development is distorted upward; if reporting slowed, reported counts and paid losses are distorted downward. I would use a Bornhuetter-Ferguson or Cape Cod approach anchored to the 80% expected loss ratio, with an explicit case-reserve adequacy adjustment rather than an automatic IBNR reduction. The expected ultimate is 0.80 x $42M = $33.6M. Paid plus case is $30.5M, leaving only $3.1M of expected IBNR if current case reserves are adequate. But the BF indication may be higher if reported-to-ultimate is below 100%, so the case-reserve adequacy question is central.
2. Compute the Bornhuetter-Ferguson IBNR using an 85% reported-to-ultimate percentage and the 80% expected loss ratio. Compare it with the paid-plus-case position and explain how you would present a range.
Read the worked answer
Expected ultimate is 0.80 x $42M = $33.6M. Expected unreported is 15% x $33.6M = $5.04M. BF IBNR is $5.04M, producing total ultimate of $30.5M reported plus $5.04M = $35.54M, or an 84.6% implied loss ratio. The paid-plus-case position alone implies only $3.1M of IBNR against the 80% prior, so the BF result is more conservative because it credits reported losses at face value. I would present a range rather than a single point: a low end near $3M if the case reserves are adequate and reporting is merely delayed, and a high end near $7M to $8M if case reserves are deficient and the central unit is still building reserves. The range should be tied to explicit scenarios for case-reserve adequacy, settlement rate, and reporting lag, not to an arbitrary percentage around the point estimate.
3. Recommend additional diagnostics you would request before finalizing the reserve estimate, and explain how the claims-handling change affects the credibility of the development method.
Read the worked answer
I would request closed-without-payment counts, reopened claim counts, settlement rates by duration, average severity by state and by office, and a sample audit of case reserves in the central unit versus the former regional offices. I would also test whether the count decline is concentrated in late-reported lost-time claims, which would indicate a backlog rather than frequency improvement. The development method loses credibility because its age-to-age factors assume a stable relationship between reported losses and ultimate losses. A mid-year consolidation can change both the speed of reporting and the adequacy of case reserves, so the historical factors may no longer apply. I would use the development method as a diagnostic, not as the sole basis, and would document the claims-handling change as a material reserve risk factor.