Case Study 1: The Large-Deductible Program and the Collateral That Wasn't There
A note on this case. This is a labeled composite, built from the real and well-documented history of large-deductible workers'-compensation programs in the United States — a structure whose mechanics, regulatory treatment, and recurring failure mode are matters of public industry record (NAIC and state guaranty-fund materials discuss them at length). No real company is named, and every figure below is illustrative. The pattern, however, is entirely real, and it has played out in actual insurer insolvencies more than once.
Background: why large deductibles took over
By the 1990s, large-deductible programs had become the dominant way that big employers bought workers' compensation. The appeal was straightforward and genuine. A large company with thousands of employees and a stable, predictable frequency of routine workplace injuries did not need an insurer to pay its working-layer losses — it could fund those itself, out of cash flow, more cheaply than it could buy them through a guaranteed-cost policy with the insurer's expense load and profit margin baked in. What such a company did need from an insurer was three things: the catastrophic layer (the rare severe claim, the multi-claimant event), the services (claims administration, regulatory compliance, the medical networks), and — non-negotiably — an admitted policy, because workers' compensation is statutory and the company's contracts, lenders, and regulators required real, filed coverage on real insurer paper.
The large-deductible structure delivered all three. The employer took a deductible far above the routine — hundreds of thousands of dollars per occurrence in many programs — and paid a dramatically reduced premium. The insurer issued a full statutory policy, paid every claim from the first dollar (because the injured worker is entitled to the full statutory benefit and is not a party to the employer's deductible), and then billed the employer back for the portion of each claim that fell within the deductible. On paper, everyone won: the employer kept its predictable losses and the cash-flow benefit of paying them over the long tail of workers'-comp claims; the insurer earned a fee for fronting and servicing; and the worker was fully protected.
The underwriting issue: the structure hides a credit risk, not an insurance risk
Read the structure the way Chapter 12 teaches you to read it, and the danger jumps out. In a large-deductible workers'-comp program, the insurer is not primarily taking an insurance risk on the working layer — it is taking a credit risk on the employer. The insurer pays the claims, including the deductible portion, now; the employer reimburses later, often over the many years it takes workers'-comp claims to run off. For the entire life of that runoff, the insurer is, in substance, lending the employer the money to pay its own losses — and like any lender, it is exposed if the borrower fails before repaying.
That is why every well-run large-deductible program is built on collateral. The employer posts security — a letter of credit, a funded trust, cash, or a surety bond — sized to the insurer's estimated future reimbursement obligation, so that if the employer becomes insolvent mid-program, the insurer can draw on the collateral instead of eating the unreimbursed deductible losses. The collateral is the underwriting. Get the collateral right — adequate in amount, secure in form, and re-evaluated as the employer's exposure and creditworthiness change — and the structure is sound. Get it wrong, and the insurer has quietly written a large unsecured loan to a company whose business it does not control.
THE LARGE-DEDUCTIBLE WC CASH FLOW [labeled composite — illustrative]
WORKER injured ──► INSURER pays full statutory benefit (FIRST DOLLAR)
│
├─► claim portion ABOVE deductible ── insurer's true insurance risk
│
└─► claim portion WITHIN deductible ── insurer FRONTS it, then
bills the EMPLOYER to reimburse
│
(reimbursement runs off over YEARS)
│
if EMPLOYER fails first ──► insurer eats the loss
UNLESS collateral covers it
The mechanics of that diagram are the whole case. The insurer's exposure is not the severity of any one injury — it is the gap between what it has already paid out and what the employer has yet to reimburse, held open across the long workers'-comp tail, secured (or not) by collateral that someone had to size and maintain.
What it shows: three ways the structure fails
The composite employer in this case — call it a large industrial contractor — entered a multi-year large-deductible program with a high per-claim deductible and posted a letter of credit as collateral. Over the following years, three things went wrong, each a textbook failure of large-deductible underwriting:
The collateral was sized to yesterday's exposure. The collateral was set at inception and not adequately re-evaluated as the program's accumulated reimbursement obligation grew. Workers'-comp losses develop upward for years after they occur (the full cost of a serious injury is not known at first report), and the stock of unreimbursed deductible losses the insurer was carrying grew steadily — but the letter of credit did not grow with it. The collateral that looked adequate in year one was a fraction of the exposure by year five.
The employer's credit deteriorated, and no one re-underwrote it. The contractor's financial condition weakened over the program's life. In a guaranteed-cost policy that would not matter much — the insurer's exposure ends when the premium is paid. In a large-deductible program it matters enormously, because the insurer's recovery of everything it has fronted depends on the employer's continued solvency. A deteriorating balance sheet in a large-deductible insured is an underwriting event, not merely a finance footnote, and it should trigger a demand for additional collateral. Here it did not, in time.
The employer became insolvent, and the collateral fell short. When the contractor failed, the insurer was left holding years of unreimbursed deductible losses — claims it had paid to injured workers and expected to recover from an employer that no longer existed. The letter of credit covered only part of the shortfall. The rest became the insurer's loss, on a book of business it had priced as someone else's risk. And because workers'-comp claims continue to develop and pay long after the employer is gone, the loss kept growing.
The deepest lesson is the one Chapter 12 states plainly: in a large-deductible program, the structure that looks like "the insured keeps its own losses" is really "the insurer fronts the losses and lends the insured the money, secured." The word that has to carry the whole weight of that sentence is secured. When the collateral is inadequate or stale, the program is an unsecured loan wearing the costume of an insurance policy.
Outcome: the guaranty-fund backstop and the regulatory response
When an insurer is weakened or pushed toward insolvency by unrecovered large-deductible losses, the failure does not stay private. State guaranty funds — the industry-funded backstops that pay the claims of insolvent insurers — can be drawn into covering the statutory workers'-comp obligations, which means the rest of the industry, and ultimately other policyholders, can end up paying for one insurer's collateral failure. This is precisely why large-deductible programs have drawn sustained regulatory attention: the NAIC and state regulators have developed guidance on the treatment of large-deductible policies in insolvency, the handling of collateral, and the allocation of recoveries between the estate and the guaranty funds. The structure's failure mode is systemic enough that the regulators built rules around it.
For the surviving insurers, the response was to professionalize large-deductible underwriting: collateral sized to developed (not just reported) exposure, regular re-evaluation as losses develop and the insured's credit changes, secure collateral forms (clean, evergreen letters of credit; funded trusts), and a discipline of treating credit deterioration as a trigger for action. The good large-deductible book is underwritten continuously, as a lending relationship, for the whole life of the runoff — not bound once and forgotten.
Lesson for the underwriter
This case is the reason Chapter 12 insists that a deductible or retention is a structuring decision with consequences far beyond the premium credit:
- Some structures transfer a credit risk, not an insurance risk. A large deductible or SIR hands the insured a layer of loss — but where the insurer fronts statutory or third-party payments (as it must in workers' comp), it is lending, and it must be secured. Price the structure, but underwrite the credit.
- Collateral is the underwriting, and it goes stale. Sizing collateral once, at inception, against reported losses, is a slow-motion failure. Workers'-comp losses develop upward for years; the collateral must be re-evaluated against developed exposure as the program runs.
- A censored loss history can flatter a bad risk. As §12.2 warned, the loss runs on a large-deductible account show only the claims that pierced the deductible. A clean-looking run can hide a worsening frequency below the retention — and an employer whose financial condition is quietly deteriorating.
- The insured's balance sheet is part of the risk. In a guaranteed-cost policy the insurer's exposure ends with the premium; in a large-deductible program it runs as long as the reimbursement obligation does, so the insured's continued solvency is something the underwriter is exposed to and must monitor.
For Harbor Steel, the case explains a decision you have already made: at roughly \$45M in revenue with a single plant, Harbor Steel is not a large-deductible candidate on its catastrophe-exposed property, and this case is why. You do not hand a mid-size, single-site, coastal manufacturer a serious self-retained layer of correlated, high-severity risk — you share risk through deductibles and the percentage wind deductible, where there is no fronting-and-credit exposure to manage. The large deductible is the right tool for a different, larger, financially deeper insured, underwritten as the lending relationship it really is.
Discussion questions
- Explain, in your own words, why a large-deductible workers'-comp program exposes the insurer to the insured's credit rather than (mainly) to the severity of any one injury. Why does the statutory nature of workers' comp (Chapter 4, Chapter 22) make this unavoidable?
- The collateral in this composite was "sized to yesterday's exposure." Given that workers'-comp losses develop upward for years, design a collateral-review process that would have caught the growing shortfall. What would trigger a demand for more collateral?
- A pricing model reports a low loss ratio on a large-deductible account with a clean loss run. List three things the model cannot see, and explain how an underwriter would investigate each before trusting the number. (§12.2)
- When an insurer fails partly because of unrecovered large-deductible losses, the guaranty fund — and thus other policyholders — can end up paying. Connect this to the theme that insurance serves a social function (Chapter 1): why is one insurer's collateral discipline arguably a matter of the whole system's integrity?
- Make the case for large-deductible programs despite this failure mode. For the right insured, why is the structure good for both parties — and what, precisely, separates "the right insured" from the contractor in this case?