The SIR That Saves $105,000 on Paper
Cedar Valley Contracting runs 180 power units. Five years of auto liability losses: $250,000, $900,000, $1.4 million, $3.8 million (a single catastrophic spinal injury), and $600,000. The incumbent guaranteed-cost program quotes $2.95 million. The alternative is a $250,000 per-occurrence self-insured retention under a $2 million excess tower: premium $1.55 million, TPA fees $150,000, a $3 million letter of credit costing $45,000 a year, and an actuarial expected retained loss of $1.1 million inside the retention. The CFO likes the $105,000 of expected savings. The safety director notes that the company has closed two terminals and holds $4 million of available cash. An aggregate stop-loss attaching at $2.5 million is available for $180,000.
1. Quantify both programs, including the stop-loss option, and test whether the $105,000 savings is real.
Read the worked answer
Guaranteed cost. $2.95 million, fixed, budgetable.
SIR program, expected case.
- Premium $1,550,000
- TPA fees $150,000
- Letter of credit $45,000
- Expected retained losses $1,100,000
- Total = $2,845,000 → apparent savings of $105,000 (3.6%)
SIR with aggregate stop-loss. $2,845,000 + $180,000 = $3,025,000 - $75,000 more than guaranteed cost.
Test the savings.
- The $105,000 is smaller than one moderately bad claim inside the retention. A single $250,000 claim instead of an expected $60,000 claim erases it.
- Stress case: if retained losses land at the 80th percentile - say $1.9 million in a year with several severe claims, which this loss history shows is entirely possible - total cost is $3,645,000, or $695,000 worse than guaranteed cost.
- The actuarial expected loss is an estimate with a range, and it is fitted to only five years of data, one of which was a $3.8 million outlier.
Conclusion. A 3.6% expected savings does not compensate for several hundred thousand dollars of downside variance on a company that just shrank. This is a risk-tolerance decision, not a procurement decision.
2. What qualitative factors should drive the decision beyond the arithmetic?
Read the worked answer
Cash and collateral. $4 million of available cash is roughly two years of maximum retained losses. The letter of credit is priced off the collateral requirement, and collateral rises exactly when losses rise - the facility tightens in the year the company most needs liquidity.
Claims control. A SIR buys control of claims handling and safety incentives. With 180 units and a catastrophic-injury history, a TPA that reports to Cedar Valley rather than to a carrier can drive return-to-work and litigation management. That is the real value, and it is not in the $105,000.
Contractual exposure. Most general contractors and shippers require certificates showing $1 million or more with additional insured status and specific carrier ratings. A $250,000 SIR means the excess carrier issues the certificate and the insured is the primary payer for the first $250,000 - some counterparties will refuse it, and some lenders will reprice.
Regulatory and accounting. Confirm state approval of the retention level, whether the excess carrier requires a deductible rather than an SIR, and how the retained liability is booked - an incurred-but-unpaid reserve on the balance sheet, not a premium expense.
Organizational readiness. Cedar Valley just closed two terminals. Adding a captive-like risk-financing structure during a contraction spreads management attention exactly when it is scarcest.
Judgment. The savings are too small relative to the variance, the collateral burden, and the certificate friction. The safety-incentive argument is the strongest reason to move - and it does not require a $250,000 retention.
3. Propose a structure and a decision timeline that would actually fit Cedar Valley.
Read the worked answer
Recommendation for the coming year. Stay on the guaranteed-cost program. It is $105,000 more in expectation and several hundred thousand dollars cheaper in the tail, and it preserves liquidity while the company stabilizes after the terminal closures.
Build the option value instead.
- Contract the TPA on a claims-audit basis now and start receiving monthly loss runs by driver, terminal, and cause - you cannot price a retention without your own data.
- Run a three-year loss pick and a severity distribution (frequency × severity, not an average) before the next renewal.
- Ask the market for a $100,000 SIR with a $2 million limit and an aggregate attaching at, say, $1.5 million. Lower per-occurrence retention cuts variance sharply and the aggregate caps the tail; the combined premium will look far less attractive than $105,000, which is the point - a structure that is cheap is cheap because it transfers little.
- Revisit a captive only when retained losses exceed roughly $2 million a year, capital is segregated, and the parent has a multi-year commitment; frictional costs on a $1.1 million retention book swamp the benefit.
Decision rule to state to the CFO. Move to a retention only when (a) expected savings exceed 10–15% of the guaranteed-cost premium, (b) unencumbered liquid assets cover a 1-in-10-year retained loss, and (c) the largest customer contracts accept the structure. None of the three is met today.