What is a captive insurance company: the insurer a company owns to insure itself.

TL;DR

• A captive is an insurance company owned by the organisation it insures. Instead of paying premiums to an outside carrier, the parent pays premiums to a company it controls.
• The three main structures are single-parent (one owner), group captive (multiple owners in the same industry), and cell captive (a shared legal structure with ring-fenced cells for each participant).
• Companies use captives to control insurance costs, access the reinsurance market directly, improve risk-management discipline, and in some cases achieve tax advantages.
• The main licensing domiciles are Vermont, Bermuda, the Cayman Islands, South Carolina, Utah, and Tennessee. Vermont is the largest US domicile; Bermuda is the largest globally.
• The IRS has aggressively challenged small captives — called 831(b) captives or micro-captives — that use a tax election to shelter business income rather than to manage genuine insurance risk.

What a captive is

Most businesses buy insurance from an outside carrier. The company pays a premium. If there is a loss, the carrier pays the claim.

A captive reverses part of that. The parent company sets up its own insurance subsidiary — a licensed insurance company — and pays premiums to that subsidiary instead of to an outside carrier. The subsidiary accumulates the premium, invests it, and pays claims from that pool. If there are fewer claims than expected, the money stays inside the corporate group.

The word "captive" reflects the relationship: the insurer is captured by the insured. It has one customer — or a small number of related customers — rather than a broad public market.

A captive is a real insurance company. It must be licensed in a jurisdiction, hold reserves, and meet solvency requirements. What it is not is independent. Its raison d'être is to serve its parent.

A concrete example

Walmart's captive is licensed in Bermuda. It covers a portion of Walmart's property, workers' compensation, and general liability exposure. Rather than paying that premium slice to a third-party carrier each year, Walmart pays it internally. The captive holds those funds, pays claims from the pool, and invests the float. Over decades, a company the size of Walmart can accumulate significant capital inside its captive.

Oil and gas companies run a different model. A group of producers — too small individually to each justify a single-parent captive — may form a group captive together. Each member pays premiums into the shared vehicle, which is governed by all of them. Claims from any member are paid from the pool. A hospital group might use a similar structure for medical-malpractice coverage, a line that commercial carriers price sharply or restrict heavily.

The three main structures

Single-parent captive

One parent, one captive. The parent owns 100 percent of the subsidiary. This is the traditional model and still the most common. It suits large corporations with enough premium volume to justify the overhead of running a dedicated licensed insurer. Walmart, BP, and most Fortune 500 companies that use captives operate single-parent vehicles.

Group captive

Multiple unrelated companies — usually in the same industry — pool their risk into a single captive that all of them own. Each member contributes premium and shares in the results. Group captives work well for mid-size businesses that individually do not generate enough volume for a single-parent structure. The hospital sector, construction, and manufacturing are typical industries where group captives are common.

Cell captive

A cell captive — sometimes called a protected cell company or segregated account company — is a single legal entity divided into legally ring-fenced sections called cells. Each cell belongs to a different participant. The cells share the captive's infrastructure and licence but keep their assets and liabilities separate. A loss in one cell cannot touch the assets of another. Cell captives lower the cost of entry for smaller organisations that want the benefits of a captive without the full cost of setting one up and running it independently.

Why companies use captives

Cost control

A traditional carrier sets premiums based on industry-wide loss experience and adds a load for its own overhead, profit margin, and cost of capital. A large company with a good loss record may be cross-subsidising worse risks in its peer group. A captive lets that company retain its own risk and benefit directly from its own safety record.

Access to the reinsurance market

Retail insurers buy reinsurance — insurance for insurers — from the wholesale reinsurance market. Ordinary companies cannot access that market directly; carriers stand between them and it. A captive changes that. The captive, as a licensed insurer, can cede its excess risk to a reinsurer directly, often at better terms than a retail carrier would pass through. This is one of the most compelling economic arguments for a captive at scale.

Risk-management discipline

Running a captive forces internal discipline. The captive's management committee must quantify risks, set actuarially supported premium rates, fund reserves, and report results. Companies that go through that process often find risks that were previously invisible — uninsured gaps, accumulations of exposure, or claims patterns that were not well tracked. The captive becomes a management tool as much as a financing vehicle.

Tax treatment

Under US tax law, premiums paid to an affiliated insurer may be deductible as ordinary business expenses if the arrangement meets the requirements of a genuine insurance transaction. The IRS tests include adequate risk distribution and risk shifting. Getting those tests right can allow a company to move pre-tax dollars into a captive, where they accumulate and invest until needed for claims. The tax benefit is real but secondary to the risk-management rationale for well-structured captives.

The main domiciles

A captive must be licensed somewhere. The choice of domicile affects solvency rules, regulatory reporting requirements, available structures, and cost.

Vermont is the largest US captive domicile and has been since it passed its first captive law in 1981. It is seen as the most sophisticated US regulator in captive matters. More than 1,100 captives are licensed there.

Bermuda is the largest captive domicile globally, particularly for the biggest corporate programmes. Its regulatory framework is mature, its proximity to New York is a practical advantage, and it has handled offshore insurance business for decades.

Cayman Islands is the second-largest offshore domicile, popular for healthcare, financial services, and group captives. Its cell captive law is well developed.

South Carolina passed aggressive captive legislation in the 2000s and has grown rapidly, particularly attracting mid-market single-parent and cell structures.

Utah and Tennessee are newer US domiciles that have competed by offering flexible rules, reasonable fees, and efficient regulators. Both have grown steadily in recent years.

IRS scrutiny: the 831(b) problem

Section 831(b) of the US tax code allows a small insurance company — one with annual premiums below a threshold (currently $2.8 million) — to elect to pay tax only on investment income, not on premium income. For a legitimate small captive, this makes sense: the premium pool is modest, and taxing it twice would distort the economics.

The problem is that some promoters used this election as a pure tax shelter. The structure looked like insurance: a business owner forms a small captive, pays large premiums to it, deducts them from business income, and the captive pays no tax on the premium. But the "insurance" covered obscure, low-probability risks that never actually generated claims. The captive was a way to move money out of the operating business into a tax-sheltered vehicle, not a genuine risk-financing tool.

The IRS listed these arrangements as "listed transactions" — a formal designation meaning abusive tax shelters — and has pursued them aggressively since Notice 2016-66. The Tax Court has struck down a large number of these structures. Courts look at whether the premiums were actuarially supported, whether the risks were real and had a genuine probability of loss, and whether the claimed coverage duplicated what the business already had from commercial carriers.

The result is that "micro-captive" is now a cautionary word. Legitimate small captives structured around real risk still qualify, but the bar for demonstrating genuine insurance purpose is high, and the IRS scrutiny is ongoing.

Who actually uses captives

Large corporations with complex risk profiles account for most of the premium volume in captives. Fortune 500 companies are the obvious group. But captives also appear in:

• Hospital systems and healthcare groups, which use them primarily for medical malpractice and stop-loss coverage.
• Construction groups that pool general liability and workers' compensation risk.
• Financial institutions that retain cyber and professional liability risk in captive vehicles.
• Non-profit organisations, including universities, that cover property and liability through shared captive structures.

The threshold of economic sense for a single-parent captive is roughly $1 million in annual premium, though that number varies by domicile and structure. Below that level, the regulatory and administrative costs typically outweigh the benefits. Group and cell structures lower the entry threshold considerably.