Experience Mod versus Retrospective Rating for a Large Trucking Account

In workers compensation experience rating, losses are split at an expected-loss-dependent split point. Losses below the split are primary and receive full weight; losses above the split are excess and receive a reduced weight, or are excluded, because they are less predictable and a single claim can dominate the mod. Including a $3.1M claim at full value is not standard. First verify the split point for this risk: with $8.2M manual premium and an expected loss rate of 0.68, expected losses are $5.58M, which likely puts the split point in the $50k–$100k range depending on the state's plan. The excess portion should be weighted by an excess ratio, not 1.0. If the mod is 1.18 with the claim at full value, recompute it with the standard split; the mod may fall to 1.05–1.10. The credibility of the mod is high for a large account, but that credibility applies to the expected loss estimate, not to the tail. A single spinal-cord injury is a severity event; it is partly random and partly a signal about the account's safety program. Do not remove it entirely, but do not let it set the mod. Also check whether the claim is still open and whether the $3.1M is case reserves or paid; development could change it.

Standard premium = $8.2M × 1.18 = $9.676M. Basic premium = 0.28 × $9.676M = $2.709M. Retro premium = (B + cL) × T = (2.709 + 1.10L) × 1.03, where L is limited losses in dollars. If limited losses are 0.44 × $9.676M = $4.257M, premium = (2.709 + 4.683) × 1.03 = $7.61M. If limited losses are 1.10 × $9.676M = $10.64M, premium = (2.709 + 11.70) × 1.03 = $14.84M. The $500k per-occurrence limit is the key: the $3.1M claim is capped at $500k, so the insured does not bear the full tail. The retro plan therefore transfers less catastrophe risk to the insured than the mod does, but the basic premium (0.28) and the loss conversion factor (1.10) charge for that protection. For this account, the retro plan is likely more attractive because the large claim is limited, and the insured can control its premium by improving safety. However, the retro plan's minimum and maximum premiums must be checked; if the maximum is too low, the insurer bears the tail. Compare the expected retro premium to the mod-based premium; if the retro plan is priced to return the same expected loss ratio, it is a risk-transfer choice, not a price discount.

Removing the claim entirely would ignore real tail risk. A trucking account with a spinal-cord injury may have hours-of-service, driver-training, or maintenance issues that make another severe claim more likely than for the average risk. The correct response is not removal but proper treatment: apply the split rating so the excess portion receives its standard weight, and consider a retrospective plan with a loss limit so the insured shares in the predictable portion of losses while the insurer retains the tail. If the company has discretion, offer a deductible or a loss limit in the retro plan. Also check whether the claim is closed and whether the reserve is adequate; if the case reserve is uncertain, use a range. A credibility-weighted excess loss provision is another option: estimate the excess expected loss from the account's history and from industry data, and charge an explicit tail load rather than letting the mod absorb the full claim. Document the decision. The agent's argument is understandable but, if accepted without adjustment, it would underprice the tail and create adverse selection against the company.