Is a captive right for your business? A feasibility checklist

A captive is worth a feasibility study when you pay enough premium in lines with steady, measurable losses, your own loss record beats the market's pricing, you can fund capital and collateral without strain, and you are willing to stay in the program for several years. If any of those four is missing, a large deductible, a group captive or a rented cell usually fits better than a company of your own. Taxes should never be the reason on their own.

This guide is the screening step. How a captive works is in what is a captive insurance company and how does it work?, and the structures are compared in single-parent vs group captives: which structure fits?.

Start with what a regulator would ask

A captive has to be licensed, so the regulator's questions are a fair preview of the feasibility work. Vermont's statute on licensing, 8 V.S.A. § 6002(c), has the applicant show the amount and liquidity of its assets relative to the risks it will assume, the expertise, experience and character of the people who will manage it, the soundness of its plan of operation and the adequacy of its insureds' loss prevention programs. Every one of those is a question about you, not about the insurance market.

The Utah Insurance Department's captive formation page lists the screening questions it expects an organization to answer before it applies. Our checklist below follows that list and adds the practical detail brokers see.

The screening checklist

The Utah dollar figures are one regulator's benchmarks, and other domiciles and consultants cite different numbers. They are starting points for a conversation, not cutoffs. Premium needed also depends on the line: a workers' compensation layer with long-tail claims needs more history to price credibly than a property deductible layer.

Premium volume: why it matters

Premium volume matters because fixed costs do not shrink with the program. The Utah page says operating costs, including fronting and reinsurance, typically run about 35 to 40 percent of captive premium. A captive with very little premium spends so much of it on audit, actuarial, management and fronting that it cannot beat the commercial price. It also has too few claims for the results to mean anything.

Count premium only for lines you would actually move, and use the retained layer rather than the full policy. If your auto liability costs $900,000 and the captive will take only the first $250,000 of each claim, the captive's premium is the price of that layer, not the whole $900,000.

Loss history: the test most programs fail

Pull five to ten years of loss runs: what they are and how to request them by line, valued at the same date, along with your exposure bases for each year (payroll, vehicles, revenue, locations). An actuary will develop open claims to ultimate cost, because a liability claim reported this year may not close for years.

Look for three things before paying for a study:

• Frequency versus severity. A good loss ratio built on a lucky stretch with no large claims will not repeat. A bad ratio caused by one claim in ten years may still suit a captive.
• Your record against the market. If commercial carriers already price you at a discount for your safety record, a captive gains less. If they price you at the class average while your losses run far below it, the gap is the opportunity.
• Claim reporting and reserving habits. Captive financials depend on case reserves. If your current carrier reserves conservatively and you have not tracked what open claims will really cost, the actuary will find the gap.

Capital, collateral and the worst year

The statutory capital minimum is the smallest number in the discussion. Vermont's 8 V.S.A. § 6004 sets $250,000 of paid-in capital and surplus for a pure captive, $500,000 for an association captive, $100,000 for a sponsored captive and $1,000,000 for a risk retention group, and lets the commissioner require more depending on the type, volume and nature of the business. Capital can be cash, marketable securities, an approved trust or an irrevocable letter of credit from an approved bank.

Then comes the money the statute does not mention. A fronting carrier will want collateral, usually a letter of credit, trust or cash, behind the captive's share of the losses, and the fronting arrangements and fronting carriers: how they work guide covers how those terms are set. Reinsurers will look at the captive's capital before they quote. Ask your CFO how a letter of credit that stays outstanding for the life of the longest claim would affect your borrowing capacity.

Then model the bad year, not the average one. If losses in the captive's layer run double the expected amount, who writes the check, and is the answer acceptable to your lender and your board?

What a feasibility study contains

The format varies by provider, but a competent study covers the same ground:

1. Data review. Loss runs, exposures and current program terms.
2. Actuarial projection. Expected losses by line and layer, with a confidence level and best-case and worst-case scenarios.
3. Program design. Which lines, which retention, what reinsurance above the captive, and whether a fronting carrier is needed.
4. Structure and domicile. Pure, group, sponsored cell or risk retention group; see Captive domiciles compared: Vermont, Utah, Delaware and offshore for the state and offshore trade-offs.
5. Pro forma financials. At least five years, including a scenario worse than expected.
6. Tax and legal review. Whether the arrangement will qualify as insurance and how the captive itself is taxed. See how captives are taxed: premiums, 831(b), excise tax and distributions.
7. Cost budget. Management, audit, actuary, premium tax, fronting, claims administration and reinsurance.

The study should say "do not proceed" when the numbers do not work. If the provider who prepares the study is paid mainly when you form the captive, weigh the recommendation accordingly.

Alternatives to compare first

Many businesses get most of the benefit from something simpler.

• Large deductible or self-insured retention. You keep the first layer of each claim without a new insurer. See large deductible workers' comp programs: how they work and what they cost and deductibles vs self-insured retentions: who pays first. The deductible and retention break-even calculator shows the break-even.
• Retrospective rating. Your premium adjusts to your own losses within a corridor.
• A group captive or rented cell. Lower entry cost, shared pooling; see cell captives and protected cell companies: how renting a cell works.
• A risk retention group. For liability exposures of one industry, under federal law; see Risk retention groups (RRGs): how they work and what to check.

Tax: what to expect and what to avoid

Premium paid to a captive is deductible only if the arrangement is insurance for federal tax purposes, with genuine risk shifting and risk distribution. The captive is then taxed as an insurance company. Small captives that elect Section 831(b) are taxed only on investment income if net written premiums (or direct written premiums, if greater) do not exceed $2,900,000 for taxable years beginning in 2026, under IRS Rev. Proc. 2025-32. Those structures are the ones the IRS targets: final regulations effective January 14, 2025 require disclosure of listed transactions and transactions of interest involving 831(b) captives. If a promoter leads with the deduction, that is your answer. The details are in Micro-captives (831(b)) and the IRS: rules, red flags and reporting and how captives are taxed: premiums, 831(b), excise tax and distributions.

A hypothetical screening

Say a 400-employee regional distributor with 60 delivery trucks pays commercial carriers $1.8 million a year for auto liability, general liability and workers' compensation. Its losses within the first $250,000 of each claim have been below 55 percent of premium for seven years. It has a safety director, telematics in every truck and an owner who is willing to commit for five years. It could fund $400,000 of capital and a letter of credit without touching its operating line.

That profile passes most of the checklist, so a feasibility study is worth the fee. Change two facts and the answer flips: if one year had a $1.5 million claim and the finance team cannot spare the letter of credit, the same distributor should look at a group captive or a large deductible plan first. These figures are invented to show the screening, not typical results.

Key takeaways

• Screen on four things before paying for a study: premium volume in steady lines, a loss record better than your price, capital and collateral you can afford, and a multi-year commitment.
• Benchmarks vary: Utah's regulator cites $750,000 of annual premium for a pure captive and $1,000,000 for a group captive, and about 35 to 40 percent of captive premium going to operating costs.
• Run the worst year in the model, and ask what your lender and board would accept.
• Compare a large deductible, retrospective rating, a group captive and a rented cell before forming your own company.
• If the main pitch is a tax deduction, stop. Your CPA and a tax attorney should review any structure, especially an 831(b) election.