Most contractors buy insurance on a guaranteed cost basis. You pay a fixed premium. The carrier pays the claims. Your cost is set before the year begins.
That structure works well early on. But it stops working once your losses run better than your rate assumes.
At that point, you are funding someone else’s claims. Your good years subsidize the carrier’s pool, and you never see the difference. A loss-sensitive program changes that. It ties what you pay to what your own losses actually cost.
The trade-off is real, though. You take on risk, you post collateral, and you commit for years rather than months. This guide explains how each structure works, when the switch makes sense, and the collateral issue that catches contractors by surprise.
What “loss-sensitive” actually means
In a guaranteed cost program, the carrier accepts your risk for a fixed price. Your premium reflects expected losses for a contractor of your size and class. It does not reflect your actual results.
In a loss-sensitive program, your final cost moves with your claims. Good years cost less. Bad years cost more.
That is the entire concept. Everything else is a question of how much risk you keep and how you fund it.
The spectrum of program structures
Loss-sensitive programs are not one product. They sit on a spectrum, and each step increases both your potential savings and your obligations.
| Structure | Who funds losses | Collateral required | Cash flow benefit | Admin burden |
| Guaranteed cost | Carrier | None | None | Low |
| Dividend plan | Carrier, with a return if losses run well | None | Minimal | Low |
| Large deductible | You, up to a per-claim amount | Yes | High | Moderate |
| Retrospective rating | You, through premium adjustment | Usually | Moderate to high | Moderate |
| Self-insurance | You, directly | Yes, plus state security | Highest | High |
| Captive | You and fellow members | Yes | High | High |
Group captives sit at the far end of that spectrum. They deserve their own discussion, so we cover them separately.
Dividend plans
A dividend plan is the gentlest entry point. You buy a guaranteed cost policy. If losses run favorably, the carrier returns a portion of premium after the period closes.
Dividends are never guaranteed. Still, they introduce loss sensitivity without collateral or risk retention.
Large deductible programs
This is the most common structure for growing contractors, and it works differently than people assume.
The carrier still issues the policy. That matters for compliance, because workers’ compensation must satisfy statutory requirements, and your certificates still show a licensed carrier. Owners and general contractors see a normal policy.
Behind that policy, the economics change. The carrier pays each claim from the first dollar. Then it bills you back, up to your deductible amount per occurrence.
You pay the carrier a smaller premium covering expenses, taxes, and the risk above your deductible. Meanwhile, you fund your own losses as they are paid.
Retrospective rating
Under a retro plan, you pay a standard premium during the year. Afterward, the carrier recalculates your premium using your actual losses.
The formula includes a basic premium for carrier expenses, your converted losses, and a tax multiplier. A minimum and maximum premium bracket the result, so your downside has a defined limit.
Two versions exist. An incurred loss retro adjusts using reserves, which means open reserves drive your cost. A paid loss retro adjusts using amounts actually paid, which improves cash flow considerably.
Self-insurance
At the far end, you become a state-approved self-insured employer. You handle claims through a third-party administrator and buy excess coverage above a retention.
This route demands state approval, a security deposit, and genuine internal capability. Few contractors need it, but it exists.
The cash flow advantage nobody explains well
Here is where loss-sensitive programs create value beyond pure savings.
In a guaranteed cost program, you pay premium up front. The carrier holds that money and pays claims over several years. The carrier earns the investment income on your dollars.
In a large deductible or paid loss retro program, you keep the money until claims are actually paid. Workers’ compensation claims pay out slowly, often over many years. As a result, you hold and use that cash in the meantime.
For a contractor managing working capital across multiple projects, this benefit alone can justify the switch.
The collateral problem — and why it matters more than you think
Now for the issue that surprises contractors most.
When you retain risk, the carrier needs assurance that you will pay the losses you promised to fund. So it requires collateral. Usually that means a letter of credit from your bank, though cash or a surety bond can sometimes substitute.
Here is the catch. That letter of credit uses your bank credit capacity. And your bank credit capacity is exactly what supports your bonding program.
Sureties evaluate your working capital, your borrowing capacity, and your overall balance sheet strength. A large letter of credit sitting against your line reduces what remains available. Consequently, a decision made to save on insurance can quietly shrink your bonding capacity.
For a contractor whose growth depends on bonded work, that trade-off can outweigh the premium savings entirely.
Collateral does not release quickly
There is a second dimension. Collateral does not disappear when the policy year ends.
Claims develop for years afterward. Carriers therefore hold collateral until those claims close or reserves drop materially. In practice, you carry collateral for multiple overlapping policy years at once.
Contractors often model the first year correctly and miss the stacking effect. Ask your broker to project collateral across a full five-year horizon before you commit.
What you must control to succeed
A loss-sensitive program rewards discipline and punishes drift. Three capabilities matter most.
Claims management. You now pay your own losses, so how claims get handled directly affects your cost. Report immediately. Investigate properly. Push for closure on stale files. Review loss runs quarterly rather than annually.
Reserve accuracy. Under an incurred loss retro, reserves drive your premium adjustment. Under a large deductible, reserves drive your collateral. Inflated reserves therefore cost you real money in both structures.
Safety performance. Frequency drives loss-sensitive economics more than severity does. A steady trickle of small claims will erode your savings faster than one serious event.
You should also consider unbundling. Many programs let you select your own third-party administrator instead of using the carrier’s claims unit. When you fund the losses, controlling who adjusts them becomes genuinely valuable.
Are you actually a candidate?
Carriers set their own thresholds, and appetite varies. However, most contractors who succeed with these programs share a consistent profile.
Premium size. Loss-sensitive structures carry fixed costs and administrative overhead. Below a certain premium volume, those costs consume the savings. Generally, contractors need substantial annual premium — often several hundred thousand dollars across workers’ compensation, general liability, and auto — before the math works.
Better-than-average losses. This is the core test. If your loss ratio consistently beats what your rate assumes, you are subsidizing others under guaranteed cost. If it does not, you would simply be buying your own bad results at retail.
Predictable, controllable claims. Frequency you can manage through safety programs works well. Volatile severity does not.
Balance sheet strength. You need working capital to fund losses and credit capacity to post collateral without damaging your surety relationship.
Genuine claims discipline. A program only pays off if someone inside the company owns claims management actively.
A multi-year horizon. These structures need three to five years to demonstrate value. You cannot enter and exit annually, because collateral and loss development follow you either way.
When you should stay guaranteed cost
Sometimes the answer is no, and that answer is correct.
Stay put if your losses are volatile or a single claim could strain you. Stay put if your working capital is thin. Stay put if collateral would compete with bonding capacity you need for growth.
Also stay put if nobody internally will own claims management. The savings assume active involvement, and without it, a loss-sensitive program simply transfers risk to a company that is not managing it.
Finally, avoid the switch if you are considering a sale within a few years. Collateral obligations and loss development complicate transactions.
How to evaluate the decision properly
Do not compare a guaranteed cost premium against a loss-sensitive premium. Those numbers are not comparable, because one includes loss funding and the other does not.
Instead, model total cost under three scenarios.
Expected losses. What does the program cost if losses run as projected?
Adverse losses. What happens at a loss ratio meaningfully worse than expected?
Stress case. What happens in a genuinely bad year, and could you fund it?
Then add the costs that never appear on a premium quote. Include letter of credit fees, the opportunity cost of tied-up capacity, TPA fees, program administration, and the effect on your bonding line.
That total is your real comparison. It also connects directly to how sophisticated contractors evaluate insurance generally, which is by total cost of risk rather than premium alone.
A realistic example
Consider two contractors with identical revenue and similar operations. Both run roughly the same guaranteed cost premium.
The first has invested in safety for years. Claim frequency runs well below class average, a full-time safety director manages the program, and open claims get reviewed monthly. Working capital is strong, and the bonding line has room.
For this contractor, a large deductible program makes sense. Good performance finally translates into lower cost, and the cash flow benefit strengthens working capital further.
The second contractor has similar revenue but inconsistent losses. Two serious claims hit in the last four years. The bank line is largely committed to bonding capacity for a growing backlog.
The same program would be a poor fit here. Collateral would compete with surety credit, and loss volatility could produce a year the company cannot comfortably absorb.
Same premium. Same trade. Opposite answers.
Making the transition
If the analysis supports a switch, plan the timing carefully.
Start early. These placements require underwriting data, financial statements, and negotiation. Allow at least 120 days, not the 90 you might use for a standard renewal.
Prepare a strong package. Underwriters want loss runs going back five years, financial statements, your safety program with training documentation, claims handling procedures, and a clear explanation of any significant losses.
Negotiate the terms that matter beyond price. Focus on the collateral formula, the aggregate stop-loss protecting you against a catastrophic year, TPA selection rights, and the schedule for releasing collateral as claims close.
Finally, build the internal capability before you need it. Assign claims ownership. Set a quarterly loss review. Establish a return-to-work program if you lack one.
Frequently asked questions
What is a loss-sensitive insurance program?
It is a program where your final cost depends on your actual claims rather than a fixed premium set in advance. Common structures include dividend plans, large deductible programs, retrospective rating, self-insurance, and captives.
How does a large deductible program work?
The carrier issues a standard policy and pays claims from the first dollar. It then bills you back up to your per-occurrence deductible. You pay a reduced premium covering carrier expenses, taxes, and coverage above the deductible, while funding your own losses as they are paid.
Why do carriers require collateral?
Because you have promised to reimburse losses that the carrier pays on your behalf. Collateral, usually a letter of credit, secures that obligation. Carriers hold it until claims close, which frequently means carrying collateral across several policy years at once.
Does insurance collateral affect my bonding capacity?
Frequently, yes. A letter of credit consumes bank credit capacity, and sureties evaluate that capacity when setting bonding limits. Contractors who depend on bonded work should model this effect before committing, since reduced surety capacity can outweigh premium savings.
What is the difference between incurred loss and paid loss retro?
An incurred loss retro adjusts your premium using reserves, so open reserves affect your cost. A paid loss retro adjusts using amounts actually paid, which improves cash flow but extends the adjustment period.
How much premium do I need before this makes sense?
Carriers set their own minimums, and the threshold varies by line and market conditions. Generally, contractors need substantial annual premium before fixed program costs stop consuming the savings. Your broker should model the breakeven point using your actual figures.
What happens if I have a bad year?
Your cost rises, subject to whatever protections you negotiated. An aggregate stop-loss caps total retained losses, and retro plans include a maximum premium. Confirm both before binding, because they define your worst case.
Can I switch back to guaranteed cost?
Yes, but obligations from prior years continue. You still fund claims from those years as they develop, and collateral remains until reserves decline. That is why these programs suit contractors with a multi-year commitment.
Make the decision on total cost, not premium
Loss-sensitive programs reward contractors who genuinely control their losses. They punish contractors who assume they do.
The analysis is not complicated, but it must be honest. Look at five years of loss history rather than one good year. Model an adverse scenario alongside the expected one. Count the collateral cost, including what it does to your bonding line. Then decide.
If the numbers support it, the savings are real and they compound over time. If they do not, guaranteed cost remains a perfectly sound answer.
Would a loss-sensitive structure work for your program? We can model it against your actual loss history and show you the breakeven point, the collateral requirement, and the effect on your surety capacity before you commit to anything.
This article is general information, not legal, tax, or actuarial advice. Program structures, collateral requirements, and regulatory treatment vary by carrier and state. Review any specific program with your broker, accountant, and surety.