Captive insurance is a licensed insurance company that a business, or a group of businesses, creates and owns specifically to insure its own risks. Instead of paying premiums to an outside commercial carrier and losing the surplus when claims come in low, the insured business sits on both sides of the transaction: it pays premiums into its captive, the captive holds reserves and pays claims, and any underwriting profit stays inside the business. The idea is straightforward. The execution involves real capital, real regulation, and real tax rules that decide whether the arrangement works or collapses under scrutiny.
How a Captive Actually Works
In a traditional policy, a business pays premiums to an unrelated insurer that pools those premiums with thousands of other policyholders and keeps whatever is left after paying claims. A captive flips that model. The business forms its own licensed insurer, pays premiums to it, and the captive holds those funds to pay the parent’s claims. If losses come in below expectations, the surplus stays within the captive rather than enriching an outside carrier.
The captive has to operate as a real insurance company, not a bookkeeping entry. It needs a license from its domicile jurisdiction, adequate capitalization, actuarially determined premiums, and the ability to pay claims when they arise. It files financial statements, submits to regulatory examinations, and maintains reserves like any commercial insurer. The distinguishing feature is ownership: the insured business, or the group of businesses, controls the insurer.
The Main Types of Captives
Structure depends on size, risk profile, and how much administrative responsibility the owner wants to carry.
- Single-parent captive. One company owns the captive and insures its own risks. This is the most common form and gives the parent full control over underwriting, coverage design, and investment of reserves. It suits large organizations with enough risk volume to absorb setup and operating costs.
- Group captive. Multiple unrelated companies, often in the same industry, jointly own a single captive. Members pool premiums, share governance, and spread fixed costs. Workers’ compensation and general liability are common lines.
- Rent-a-captive. A company rents capacity from an existing captive rather than forming its own, avoiding the cost and complexity of licensing a new entity. It works for businesses with predictable, moderate-severity risks that don’t justify a standalone captive.
- Protected cell company. A PCC operates like a rent-a-captive with a legal upgrade: each participant’s assets and liabilities sit in a legally separate cell that cannot be reached by claims against other cells or the captive’s core. Even in a cell liquidation, other participants’ assets are shielded.
What It Costs to Form and Run a Captive
Forming a captive is not cheap, and the numbers tend to surprise business owners hearing them for the first time. Initial startup typically runs $50,000 to over $100,000. A feasibility study alone runs $15,000 to $25,000; legal and formation work starts around $10,000 and rises with complexity; licensing and domicile fees run $5,000 to $15,000 depending on the jurisdiction. The feasibility study is the gatekeeper, analyzing the company’s loss history, risk profile, and projected premiums to decide whether a captive makes financial sense at all.
Ongoing costs continue every year. Captive management fees commonly run 15% to 35% of annual written premiums, or a flat fee of $36,000 to $100,000 or more. Actuarial opinions cost $5,000 to $15,000. Independent audits and tax preparation add another $10,000 to $20,000. Premium taxes vary by domicile. A small captive can expect to spend six figures annually just on administration. That fixed-cost floor is why a captive only pencils out when annual premium volume is large enough, and loss experience favorable enough, to justify it.
Where Captives Are Licensed
Every captive is licensed in a specific jurisdiction, and the choice shapes capital requirements, reporting obligations, investment restrictions, and premium taxes. More than 30 U.S. states have captive insurance statutes. Vermont has long been the dominant domestic domicile, with roughly 680 captives as of 2024, followed by Utah, North Carolina, Delaware, and Hawaii. Bermuda and the Cayman Islands remain common offshore choices for larger or multinational programs.
Minimum capital and surplus for a single-parent captive in most U.S. states sits around $250,000, though group captives and reinsurance captives often face higher thresholds, and some offshore domiciles set significantly lower floors. Most owners work with a captive manager or consultant who knows the regulatory landscape well enough to match the business’s needs to the right jurisdiction.
Why Businesses Form Captives: The Tax Picture
Tax treatment is the primary financial driver behind most captive programs, and it is also where the biggest legal risk lives. When a captive arrangement qualifies as genuine insurance for federal tax purposes, the parent can deduct premiums paid to the captive as ordinary business expenses. The captive can deduct losses and loss reserves when calculating its taxable income, which defers tax compared with simply setting aside money to self-insure.
The 831(b) Micro-Captive Election
Small captives can elect under Section 831(b) of the Internal Revenue Code to be taxed only on investment income, effectively excluding all premium income from tax. The trade-off is a premium cap. For 2026, the captive’s net written premiums cannot exceed $2.9 million, inflation-adjusted from a statutory base of $2.2 million. The captive also has to meet diversification requirements; generally, no single policyholder can account for more than 20% of written premiums.
On paper this is a powerful election. A qualifying captive pays federal income tax only on interest, dividends, and capital gains from its investment portfolio, while premiums pass through tax-free. The same feature has drawn aggressive tax planning, which has in turn drawn heavy IRS scrutiny.
Excise Tax on Offshore Captives
Businesses that domicile offshore face an added cost. Section 4371 of the Internal Revenue Code imposes a federal excise tax of 4% on casualty insurance premiums and 1% on reinsurance premiums paid to foreign insurers. A captive can avoid the excise tax by electing to be treated as a U.S. corporation for tax purposes, but that election pulls the captive fully into the U.S. tax system. Offshore owners have to weigh the excise tax against whatever they gain from the foreign domicile.
What the IRS Requires for the Arrangement to Count as Insurance
Not every captive arrangement qualifies as “insurance” for federal tax purposes. If the IRS decides the captive is not real insurance, the premium deductions disappear. Courts have settled on three requirements every captive must satisfy.
Risk shifting means the insured company genuinely transfers the financial consequences of a potential loss to the captive. If the parent keeps the economic risk through guarantees, side agreements, or circular fund flows that return premiums to the parent, risk has not actually shifted.
Risk distribution means the captive spreads risk across enough independent exposure units. A captive insuring a single risk for a single entity may fail. Courts have held that insuring three or four affiliated entities is not enough. Group captives satisfy this naturally through their membership. Single-parent captives usually address it by insuring multiple subsidiaries or by writing a meaningful volume of unrelated third-party business. The IRS safe harbor asks for at least 50% of premiums from unrelated parties, though courts have accepted as little as 30%. A “brother-sister” structure, where the captive and the insured companies share a common parent but the captive is not directly owned by the insureds, can also satisfy the test under the analysis in the Humana line of cases.
Insurance in the commonly accepted sense is the broader inquiry. The Tax Court has looked at whether the captive was created for legitimate nontax reasons, whether premiums were actuarially determined at arm’s length, whether the captive was adequately capitalized, whether policies had clear terms, and whether claims were paid from a separately maintained fund. A captive that checks the technical boxes but charges inflated premiums or invests its reserves in illiquid loans to its own parent will fail.
The Avrahami Warning
The Tax Court’s decision in Avrahami v. Commissioner shows what happens when a captive falls apart under examination. The court found the taxpayers’ arrangement was not insurance for federal tax purposes because it lacked genuine risk distribution and did not operate as insurance in the commonly accepted sense. Premiums were “utterly unreasonable.” Actuarial justifications were unpersuasive. The captive invested 65% of its assets in illiquid long-term loans to related parties, investments the court said “only an unthinking insurance company would make.” Funds moved in a suspicious circle among the related entities. The premium deductions were denied in full.
The lesson is direct. A captive that exists mostly to generate deductions will not survive IRS examination. Arm’s-length premiums, real claims activity, legitimate business purpose, and sound investment practices are the minimum for the arrangement to work.
Micro-Captive Scrutiny and Mandatory Disclosure
The IRS has made micro-captive arrangements one of its highest enforcement priorities. In January 2025, the IRS issued final regulations classifying certain 831(b) micro-captive transactions as “listed transactions” and “transactions of interest.” Both categories trigger mandatory disclosure. In March 2026, a federal court upheld those regulations, confirming they are enforceable.
The regulations target arrangements with specific features: ownership linkage between the insured and the captive above a 20% threshold, low loss ratios suggesting premiums far exceed actual claims, and premium funds that cycle back to the insured through loans, guarantees, or investments. Where those features are present, disclosure is not optional.
Participants in reportable transactions must file Form 8886 with their tax return and send a copy to the IRS Office of Tax Shelter Analysis. Material advisors who promoted or assisted with the arrangement have separate filing obligations. Penalties for failing to disclose are steep: 75% of the tax benefit claimed from the transaction, with a minimum penalty of $5,000 for individuals and $10,000 for entities. For listed transactions, the maximum penalty reaches $100,000 for individuals and $200,000 for other persons.
Reinsurance
Most captives buy reinsurance to protect against catastrophic losses that could exceed their reserves. The captive retains a manageable layer of risk and transfers the excess to a third-party reinsurer, which stabilizes financial results and protects capital.
Two structures are common. Proportional reinsurance gives the reinsurer a fixed percentage of every premium and every loss; it is predictable, but less efficient for low-frequency risks. Excess-of-loss reinsurance pays only when a single claim or aggregate claims exceed a specified threshold, and it is the more typical choice for captives managing large but infrequent exposures. Reinsurance also serves a tax function: for single-parent captives that need to demonstrate risk distribution, ceding a portion of risk to an unrelated reinsurer brings independent exposure into the captive’s book of business.
Captive insurance rewards businesses that treat it as a real insurance operation and punishes those that treat it as a tax shelter. If the loss history, premium volume, and management capacity are there, the arrangement can genuinely lower long-term risk-financing costs and keep underwriting profit inside the business. If they are not, the setup costs, ongoing expenses, and IRS exposure add up to a problem, not a solution. A feasibility study built on actual loss data is the honest test of whether a captive is the right answer.