Captive Insurance: A Complete Guide for Risk Managers and Actuaries

A captive insurance company is, in its simplest form, an insurance company that a non-insurance business owns and uses to insure its own risks. The parent buys coverage from a carrier it controls instead of (or alongside) a commercial market policy, and over time the captive accumulates premium, pays claims, and builds surplus the same way any insurer does. The mechanics are insurance; the economics are corporate finance.

Captives have grown from a curiosity used by a handful of large industrials in the 1950s into a mainstream risk-financing tool. As of year-end 2024 there were roughly 6,074 active captives worldwide, with Bermuda (around 19% of the global count), the Cayman Islands (around 15%), and Vermont (around 11%) hosting the largest concentrations. Premium written through captives is now estimated in the high tens of billions annually, and the structure has spread well beyond Fortune 500 manufacturers into mid-market healthcare, construction, professional services, agriculture, and even franchise systems.

This guide is written for two audiences: risk managers and CFOs evaluating whether a captive belongs in their risk-financing stack, and actuaries who price, reserve, or audit captives and need to understand both the structural choices and the technical wrinkles that make captives different from commercial insurers. It covers the major captive types, the domicile decision, the tax framework (including the controversial 831(b) micro-captive), the actuarial challenges unique to captives, and the situations where a captive genuinely earns its keep versus where it’s an expensive distraction.

A Brief History: Where the Term Comes From

The word “captive” in this context was coined by Frederic M. Reiss, an Ohio insurance broker who in the 1950s helped Youngstown Sheet & Tube Company form a wholly owned insurance subsidiary to cover its mines. Youngstown referred to certain mines as “captive mines” because they served only the parent’s production needs; Reiss extended the language to the insurance subsidiary. Reiss later moved to Bermuda in 1962 and is widely credited with building the offshore captive industry that still anchors the global market today.

The structure stayed niche through the 1960s and 1970s but exploded during the U.S. liability crises of the mid-1980s, when commercial general liability and product liability capacity dried up and rates spiked. Congress responded with the federal Liability Risk Retention Act of 1986 (LRRA), which created the Risk Retention Group — a special captive form that can write liability coverage across all U.S. states while being regulated by a single domiciliary state. RRG counts grew from about 65 in 1987 to several hundred by the late 2000s, concentrated in healthcare, education, and nonprofit sectors.

Each subsequent hard market — the post-9/11 property and aviation crunch, the 2019–2023 commercial property and casualty hardening, and the 2023–2025 nuclear-verdict-driven casualty squeeze — has pushed another wave of mid-market companies into captive structures. New domiciles have emerged regularly to capture this demand: Vermont led the U.S. wave in the 1980s, and recent additions like Tennessee, Connecticut, North Carolina, and Iowa continue to compete on regulatory speed and capital flexibility.

The Captive Family Tree

“Captive” is an umbrella term covering several distinct legal structures. The major types are worth understanding because each has different ownership rules, capital requirements, regulatory treatment, and tax consequences.

Single-Parent (Pure) Captive

The original and still most common form. A single corporate parent owns 100% of the captive and the captive insures only the parent and its affiliates. Premium dollars flow from operating subsidiaries into the captive, which then pays claims, builds surplus, and optionally cedes risk to commercial reinsurers. The arrangement is conceptually simple but it’s the one that demands the most capital and the most internal governance.

Group Captive

Multiple unrelated companies, usually in similar industries with similar loss profiles, jointly own and insure through a single captive. Members typically share underwriting results within a pool while each member retains some portion of its own loss experience through a separate account or layer. Group captives are popular among mid-market workers’ compensation buyers, trucking fleets, and contractors — companies that are too small to support a single-parent captive on their own but want better economics than the commercial market offers.

Association Captive

A captive owned by a trade or professional association and writing coverage for the association’s members. Used heavily in healthcare professional liability, where state medical societies have sponsored captives that have grown into substantial insurers over decades.

Risk Retention Group (RRG)

A creature of federal law (the 1986 LRRA), an RRG must be owned by its policyholders and can only write commercial liability coverage. Its defining advantage is single-state regulation: an RRG licensed in Vermont, for example, can write liability policies in all 50 states without separately qualifying in each. The trade-off is the homogeneity requirement (members must share comparable liability exposures), the liability-only restriction, and the absence of state guaranty fund protection — if an RRG fails, policyholders have no backstop.

Protected Cell Captive / Segregated Cell / Sponsored Captive

A Protected Cell Company (PCC) is a single legal entity divided into legally segregated “cells,” each with its own assets and liabilities. A sponsor (typically a captive manager or a commercial insurer) holds the core capital that satisfies regulatory minimums, and individual participants rent a cell to write their own program. The structural promise is that one cell’s liabilities cannot reach another cell’s assets, even in insolvency — the segregation is statutory, not merely contractual. Bermuda, Cayman, Guernsey, Vermont, South Carolina, and Washington D.C. all have well-tested protected cell statutes. Cells are the fastest way to put a captive structure to work because the sponsor’s core absorbs the regulatory overhead.

Rent-a-Captive

Predecessor to and conceptually similar to the cell. The participant rents underwriting capacity from a sponsor-owned captive and shares in the underwriting results, but without the statutory ring-fencing that a true PCC provides. The participant is exposed to the sponsor’s and other tenants’ credit risk in a way that cell participants are not. Rent-a-captives still exist but cell structures have largely replaced them in jurisdictions that offer PCC legislation.

Micro-Captive (831(b) Captive)

Not a separate legal structure but a tax election available to small captives writing under a statutory premium ceiling (currently around $2.85 million indexed annually). Section 831(b) of the Internal Revenue Code allows a qualifying small insurer to be taxed only on its investment income rather than on underwriting profit. The election has legitimate uses for genuinely small captives, but it has also become the most aggressively marketed and most heavily IRS-scrutinized corner of the captive market — covered in detail below.

Branch Captive

A captive licensed in one jurisdiction but operating a branch in another, typically to access the second jurisdiction’s tax or regulatory advantages without re-domesticating. Most often seen with offshore-domiciled captives operating U.S. branches under section 953(d) elections.

Why Companies Form Captives

The honest answer is: for a few different reasons that overlap in any given case. The reasons that hold up over a multi-year horizon usually include some combination of the following.

Cost reduction in the long run. A captive eliminates the commercial insurer’s acquisition expense, profit margin, and risk load from the portion of risk the parent retains. For a company with predictable loss experience that historically pays a commercial insurer 1.4–1.6x its expected losses, that gap is the captive’s structural arbitrage. The math only works at meaningful premium volumes (usually north of about $1 million in ceded premium for a single-parent captive to clear setup and ongoing operating cost) and only if the company actually has predictable, frequency-driven losses to capture.

Premium stability through hard markets. Commercial pricing moves in cycles. A captive lets the parent insulate at least the predictable layer of its program from market swings — it pays itself a stable premium based on its own loss experience rather than absorbing a 30% increase because a different industry had a bad year.

Access to reinsurance markets. Reinsurance is generally cheaper than retail insurance at the same attachment point because reinsurers don’t carry commercial acquisition costs. A captive can buy reinsurance directly and pass the savings to the parent rather than buying retail insurance whose price embeds those costs.

Coverage customization. Captives can write coverage on terms the commercial market won’t offer: deductible buy-downs, difference-in-conditions wrappers, employee benefits cells, cyber, reputational risk, supply chain disruption, pandemic interruption. The captive doesn’t need a standard ISO form — the policy is whatever the parent and its actuary agree it should be.

Cash flow and investment income. Premium paid to the captive stays inside the consolidated group and earns investment income until claims are paid. For long-tail lines like workers’ compensation, general liability, or medical professional liability, the float can be substantial.

Tax treatment. Premium paid to a captive that qualifies as an “insurance company” for federal income tax purposes is deductible by the parent the same way premium paid to a commercial insurer would be. Reserves held inside the captive grow tax-deferred. By contrast, an internal self-insurance reserve sitting on the parent’s balance sheet is not deductible until paid out as a claim. This single difference is enough to drive many captive decisions — though, as the IRS controversies below show, the “qualifies as insurance” test has teeth.

Risk management discipline. Less quantifiable but often cited: the act of pricing the company’s own risk through its own insurer surfaces loss patterns and exposure data the parent didn’t previously analyze rigorously. Many captive parents report that the loss-control investments triggered by captive formation paid for themselves independent of any premium savings.

When a Captive Is the Wrong Tool

Captives are not universally a good idea. The structure makes sense when the parent has enough predictable risk to fund a captive at scale and enough capital to back it. It does not make sense when:

The risk being insured is genuinely catastrophic or low-frequency/high-severity. A captive can’t efficiently retain a one-in-fifty-year casualty event — the capital required would dwarf the premium savings. Commercial reinsurance is still the right answer for the tail.
The parent doesn’t have sufficient premium volume. A captive’s annual operating cost (captive manager fees, audit, actuarial, legal, regulatory filings, premium taxes) typically runs $75,000–$200,000 for a straightforward single-parent captive and more for group captives or PCCs. Below roughly $1 million of ceded premium, the overhead usually swamps the savings.
The parent lacks the governance bandwidth. A captive needs a board, an underwriting and claims framework, an actuary, an auditor, and meaningful management attention. Treating it as a passive tax shelter is the fastest path to IRS trouble.
The primary motivation is tax savings. The IRS has won every major recent micro-captive case on substance-over-form grounds. If the captive doesn’t pass a market-standard insurance test — risk transfer, risk distribution, fortuitous loss, common notions of insurance — the tax benefit reverses and penalties follow.

Domicile Selection: The Decision Behind the Decision

Where a captive is licensed shapes its capital requirements, regulatory environment, tax exposure, and operating cost. There are now roughly 70 captive domiciles worldwide. The major ones cluster into a few categories:

Offshore Centers

Bermuda remains the largest single domicile by captive count (around 958 active captives in recent counts). Its appeal is mature infrastructure (captive managers, lawyers, accountants, actuaries all on-island), a regulator that genuinely understands the industry, geographic proximity to the U.S. East Coast, and the option for a U.S.-parented captive to elect under IRC section 953(d) to be taxed as a U.S. insurer while operating in Bermuda. Bermuda’s tier system (Class 1 through Class 4) lets very small captives operate with proportionate oversight while larger reinsurers face full Solvency II-equivalent regulation.

Cayman Islands ranks second globally (around 765 captives), with particular strength in healthcare and group captives. The Cayman Monetary Authority has a similar tier system and a regulatory ethos slightly more flexible than Bermuda’s for some structures.

Guernsey dominates the European captive market and pioneered the protected cell concept in the 1990s. Guernsey is particularly common for UK and continental European parents seeking proximity and Solvency II-aligned regulation.

Other offshore centers — Barbados, the British Virgin Islands, Isle of Man, Anguilla, Gibraltar, Luxembourg — serve specific national or industry niches but together hold a much smaller share than the top three.

U.S. Onshore Domiciles

Vermont is the dominant U.S. domicile (around 567 captives) and has been since enacting its captive law in 1981. Vermont’s edge is a long-tenured, well-staffed Department of Financial Regulation that knows the industry intimately, plus geographic and time-zone convenience for East Coast parents. Vermont licenses every major captive type including PCCs, RRGs, and sponsored captives.

Utah and Delaware have built large captive books with regulatory speed and competitive capital requirements. Utah grew quickly through the 2010s before tightening its review process in response to micro-captive concerns; Delaware emphasizes its corporate-law expertise and series LLC structures.

South Carolina, Tennessee, North Carolina, Hawaii, Nevada, Connecticut, Washington D.C., and Montana all have meaningful captive sectors, each with their own combination of capital minimums, premium tax rates, and licensing speed. Tennessee in particular has marketed aggressively to mid-market captive formations.

Iowa enacted modern captive legislation in 2025, joining a small number of states that have entered the captive market recently with redomestication tax provisions designed to attract captives from older domiciles.

The choice between offshore and onshore typically reduces to three questions: how much regulatory oversight does the parent want, how much premium tax friction can it absorb, and how exposed is it to controlled-foreign-corporation rules. Offshore domiciles often have lower headline capital requirements and no premium tax, but the U.S. parent has to navigate Subpart F, BEAT, and CFC rules unless it makes the 953(d) election (which moves the captive’s tax residence to the U.S. while keeping its legal domicile offshore). Onshore captives avoid the international tax overhead but pay state premium tax and face U.S. statutory accounting and regulatory examination.

The Tax Framework

Federal income tax treatment is one of the most consequential captive design choices and one of the most heavily litigated. The relevant Code sections and concepts are worth knowing in outline.

Section 831(a): The Default Insurance Company Tax

A captive that qualifies as an insurance company for federal tax purposes is taxed under subchapter L like any commercial P&C carrier — on underwriting income plus investment income, with deductions for loss reserves (subject to discounting) and unearned premium reserves. This is the baseline for any captive writing more than the 831(b) premium ceiling.

Section 831(b): The Micro-Captive Election

A captive writing under the annual premium ceiling (currently around $2.85 million, indexed for inflation) can elect to be taxed only on its investment income, with underwriting income excluded from federal income tax. The election has legitimate uses — genuine small-business captives covering niche or hard-to-place risks can benefit meaningfully — but it has also been the structure of choice for aggressive tax shelters marketed to high-income business owners who park deductible premium in a captive and extract it later as long-term capital gains or estate-planning vehicles.

The IRS has listed certain 831(b) structures as “transactions of interest” (most recently under proposed regulations in 2023–2024) and has won a string of Tax Court cases against micro-captive promoters, including Avrahami v. Commissioner (2017), Reserve Mechanical Corp. v. Commissioner (2018), Syzygy Insurance Co. v. Commissioner (2019), and several later decisions. The common factors that sink these cases are the same: artificial premium pricing untethered to actuarial analysis, no real risk distribution, circular reinsurance pools designed to satisfy a numerical risk-distribution test without genuine independent risk, claims payment patterns that don’t look like insurance, and captives that exist primarily to move deductible dollars into a low-tax vehicle. The CIC Services case (2021), separately, addressed the IRS’s ability to require disclosure of micro-captive arrangements; the Supreme Court ruled the disclosure rule could be challenged pre-enforcement, which shifted the procedural landscape but did not weaken IRS substantive authority.

A well-structured 831(b) captive that writes genuine third-party-comparable premium for genuine risks, with a real actuarial pricing study, a real claims history, real risk distribution across enough independent insureds (or through a properly structured reinsurance pool), and arms-length governance can survive scrutiny. The cottage industry of promoters peddling cookie-cutter 831(b) structures to dentists, doctors, and small business owners cannot.

Section 953(d): The Offshore-to-Domestic Election

A foreign-domiciled captive can elect to be treated as a U.S. corporation for federal income tax purposes. The election lets a U.S.-parented captive operate in Bermuda or another offshore center for regulatory and operational reasons while paying U.S. federal income tax at the captive level (avoiding Subpart F and BEAT complications). It’s effectively the bridge that lets most large U.S. corporate captives sit in Bermuda without triggering the controlled-foreign-corporation regime.

State Premium Tax

Onshore captives pay state premium tax in their domicile, usually at a rate well below the commercial premium tax rate (Vermont, for example, taxes captive premium on a sliding scale from 0.214% down to 0.048%, capped at $200,000 annually). Offshore captives pay no state premium tax in their domicile but may face federal excise tax (1%–4% depending on the line of business) on premium paid by U.S. insureds to a non-U.S. insurer that hasn’t elected under 953(d).

Actuarial Considerations: Why Captives Are Harder Than They Look

From an actuary’s perspective, a captive is not just a small commercial insurer. The features that make captives economically attractive also make them harder to price and reserve than the commercial portfolios most actuaries learn on.

Data Sparsity

A single-parent captive has, by construction, exposure to one company’s loss experience. Credibility is limited; the chain-ladder factors derived from the parent’s own triangles may be based on so few claims that the standard error swamps the central estimate. Actuaries reserving for single-parent captives typically rely on Bornhuetter-Ferguson or Cape Cod methods that blend the captive’s own experience with industry benchmark patterns — the question of which benchmark and what credibility weight is itself a major judgment call.

Tail Factor Uncertainty

Many captives write long-tail coverage (workers’ compensation, medical professional liability, general liability) where the tail beyond the captive’s development history materially affects the reserve estimate. A captive with five years of data writing thirty-year-tail medical professional liability faces an enormous tail-factor problem: the answer is almost entirely judgment-driven, and small changes in the assumed tail can move the reserve estimate by 20% or more.

Premium and Loss Mismatching

A captive’s premium reflects what the parent decided to charge itself for the coverage. If the captive is being capitalized initially with under-priced premium — or, in micro-captive abuse cases, over-priced premium — the underwriting income signal is distorted. Actuaries reserving for captives have to separate true loss experience from pricing artifacts in a way that’s rarely an issue at a commercial insurer with arms-length pricing.

Reinsurance and Net Retention

Most captives buy reinsurance to cap their net retention. The captive’s reserve estimate has to reflect the reinsurance program: gross loss reserves, ceded reserves, net reserves, and the credit risk associated with reinsurance recoverables. For captives that participate in pooling arrangements (common in group captives and association captives), the actuary has to reserve both the captive’s own retained layer and its share of the pool.

Capital Adequacy

Captive regulators generally require capital adequacy analysis — some form of risk-based capital, often a variant of NAIC RBC scaled for the captive’s premium volume and risk profile. Funding levels are commonly set at a target confidence level on the loss distribution (the 75th, 85th, or 90th percentile is common, depending on the line). The actuary’s loss distribution — particularly the tail of the aggregate distribution — drives both the capital requirement and the discussion with management about how much capital is enough.

Reserve Discounting

Different accounting bases produce different reserve numbers from the same underlying loss data. Statutory reserves are typically undiscounted (with limited exceptions); GAAP reserves are usually undiscounted for short-duration contracts but may be discounted for long-duration contracts (and now reflect LDTI for life captives); tax reserves are discounted using prescribed IRS interest rates and patterns; IFRS 17 introduces fulfillment cash flows and risk adjustment. A captive’s actuary frequently has to produce all four numbers from a single dataset and explain to the board why they differ.

Loss Portfolio Transfers, ADCs, and Runoff

When a captive is being wound down — because the parent has been acquired, the program is being restructured, or the captive is no longer economically viable — the board has three structural options for the existing book. Each has a different risk, accounting, and tax profile, and the choice frequently turns on actuarial inputs that the board itself is not equipped to evaluate without help.

Loss Portfolio Transfer (LPT). The captive pays a single premium to a third-party reinsurer or specialty runoff company, which assumes the closed accident years and pays claims as they emerge. Legal ownership transfers. The captive achieves finality on the transferred portfolio — subject to one important caveat: if the assuming party fails before all claims are paid, the captive may face a contingent liability through uncollectible reinsurance recoverables. The canonical cautionary case is the asbestos runoff books of the 1990s, where some cedents booked net IBNR exceeding gross because so much of the reinsurance had become uncollectible. Counterparty due diligence, collateral trusts, parental guarantees, and downgrade triggers are standard protections.

Adverse Development Cover (ADC). The captive keeps the liability but buys reinsurance protection above an attachment point typically set at 105–110% of the central reserve estimate. The captive retains the central estimate but caps downside. ADCs usually cost less than LPTs for an equivalent reduction in worst-case exposure, which makes them attractive when the parent has reasonable confidence in the central estimate but wants to insulate against tail surprises. The risk-transfer analysis is sharper than for an LPT: aggregate caps that look more like a financing arrangement than insurance can fail the §832 risk-transfer tests, with the premium getting recharacterized as a deposit and the deduction disallowed.

Runoff. The captive simply stops writing new business and pays existing claims as they emerge, eventually dissolving once the book is exhausted or commuted. For long-tail lines this means continuing to operate the captive — annual statements, audit, statement of actuarial opinion, capital adequacy filings — for ten to twenty years. Runoff is often the default when no LPT market exists at acceptable pricing or when the regulator prefers a gradual wind-down to a transaction-based exit.

Hybrid structures are common — an ADC layered over a runoff to cap the tail while keeping central exposure, an LPT covering only closed accident years while open years stay on the captive’s books, or commutation of specific reinsurance contracts to clean up the ceded position without touching the gross book. The decision is rarely a clean choice among three options at a single board meeting; it is more often a sequence of structural moves refined over years as the loss experience develops and the market for assumption capacity shifts.

Pricing any of these structures depends on actuarial work that goes well beyond a routine year-end reserve review. The actuary needs a central estimate and a credible distribution around it (the distribution above the central estimate is what an ADC is actually pricing), discounted and undiscounted views at multiple interest rates, a multi-year cash flow projection by line and accident year, and an explicit view of the counterparty’s likely pricing perspective. As LossReserves.com’s walkthrough of LPT, ADC, and captive runoff argues, the most useful thing the captive’s actuary can do for the board is decompose an incoming LPT quote into its three components — discounted reserves, risk margin for variability, and the assuming party’s profit margin — so the board can see how much of the quoted price is each, and how much of the discount benefit the assuming party is retaining rather than passing through.

A common pitfall: captive owners often expect an LPT premium to reflect heavily discounted reserves with most of the discount benefit accruing to them. In practice the assuming party prices the discount in but adds a risk margin and a profit margin on top, and for a long-tail program with an eight-year average payment duration discounted at 4%, the undiscounted reserve is roughly 1.35x the discounted reserve — a large enough gap that the assuming party can retain most of the discount benefit while still quoting a price the cedent finds attractive relative to its undiscounted statutory carry value. The actuary’s job is to make the decomposition visible so the negotiation is informed.

A second pitfall is the statement of actuarial opinion. A captive in runoff still needs an annual SAO — domicile regulators don’t waive it — and after an LPT or ADC the opinion has to specifically address the post-transaction balance sheet: pre- and post-transaction gross and net liability, the portfolio transferred or capped, retained contingent liability, and the reasonableness of the captive’s remaining net position. The signing actuary’s view of the post-transaction reserve has to be defensible to the regulator, and any disagreement with the pricing actuary tends to surface only at close. Bringing the SAO actuary into the structuring conversation early — not at year-end after the deal is signed — is the standard fix.

This corner of captive actuarial work has grown rapidly as legacy captives have been simplified or exited following corporate M&A, and it is one of the areas where the deepest pricing work happens because both sides of the transaction have their own actuaries, both with strong views, and the gap between them is real money.

Common Use Cases by Industry

Captives have spread broadly but certain industries have been particularly heavy users:

Healthcare systems use captives extensively for medical professional liability, where commercial capacity has been historically volatile and homogeneous risk distribution lends itself to RRG and group captive structures. Many of the largest U.S. health systems run multiple captives covering professional liability, general liability, workers’ compensation, and employee health benefits.
Trucking and transportation have driven significant group captive formation in commercial auto liability, where the commercial market has been hardening for years and pricing has not reliably tracked individual fleet experience.
Construction uses captives for workers’ compensation, general liability, builders risk, and subcontractor default. The contractor-controlled insurance program (CCIP) and owner-controlled insurance program (OCIP) wrap-up structures often involve a captive.
Manufacturing — the original captive use case — remains a major user across products liability, workers’ compensation, and property programs.
Energy companies operate captives for well-control coverage, environmental liability, and business interruption that the commercial market either won’t cover or prices punitively.
Higher education and nonprofits often participate in association-sponsored captives or RRGs for educators’ professional liability, directors-and-officers, and general liability.
Franchise systems and retail have grown into mid-market captive use for cyber liability, employment practices liability, and customized commercial property programs.

The Formation Process

Spinning up a captive is a multi-month exercise. The standard sequence runs roughly as follows.

Feasibility study (months 1–3). A captive actuary or consulting firm models the parent’s loss history, projects the captive’s expected underwriting result, models capital requirements at various confidence levels, compares the captive’s expected economics to the commercial market alternative, and stress-tests the conclusion under reasonable adverse scenarios. The feasibility study is the document that decides whether the captive proceeds and shapes its structure.

Domicile selection (month 2–3). Driven by the feasibility study’s capital and structural conclusions plus the parent’s tax posture. The decision typically narrows quickly to two or three candidates and is finalized after meetings with the relevant regulators.

Application and licensing (months 3–6). The captive submits a business plan, capital plan, actuarial opinion on initial funding, biographical affidavits for directors and officers, and reinsurance contracts to the chosen regulator. Most established domiciles can license a straightforward single-parent captive in three to four months; complex structures or first-time-of-their-kind risks take longer.

Capitalization and policy issuance (month 5–7). The parent contributes initial capital (usually a mix of cash, letters of credit, and admitted assets), the captive issues its first policies, and premium flows. From this point the captive operates as an insurer: it collects premium, pays claims, books reserves, files annual statements, and undergoes annual audit and actuarial opinion review.

Ongoing operations require a captive manager (the third-party administrator that handles regulatory filings, premium and claims accounting, and board support), an auditor, a qualified actuary, legal counsel familiar with the domicile, and active board governance. Total annual operating cost for a straightforward single-parent captive typically runs $75,000–$200,000; group captives, PCCs, and complex multi-cell structures cost more.

What Could Change the Captive Landscape Next

Three forces are worth watching for any practitioner active in captive work:

IRS micro-captive enforcement. The IRS has been consistently expanding its disclosure rules and audit posture on 831(b) captives, and recent Tax Court decisions have given the agency strong precedent on substance-over-form challenges. Promoters of cookie-cutter structures continue to lose. Legitimate small captives can still elect 831(b) but need to demonstrate genuine insurance substance, with documentation to match.

Hard-market migration. Each commercial hard market — in commercial auto, cyber, D&O, medical professional liability, and most recently nuclear-verdict-affected casualty lines — pushes more mid-market companies to consider captives. The 2023–2025 casualty pricing cycle, in particular, has driven significant new captive formation in trucking, healthcare, and contractor segments. If commercial casualty rates stabilize or soften, some of that demand returns to the standard market; if pricing continues to harden, the captive sector continues to grow.

Regulatory convergence and Solvency II equivalence. Bermuda achieved Solvency II equivalence in 2016 and the European-Union-aligned regulatory standard has gradually become the global benchmark for larger captives. Smaller captives still operate under proportionate regimes, but the long-term trend is toward more standardized solvency oversight, more standardized own-risk-and-solvency assessment requirements, and more standardized governance. For captive actuaries the implication is that the role is becoming more demanding: more capital modeling, more stress testing, more board-level reporting.

The Bottom Line

A captive is not a tax trick and it is not a substitute for risk management. It is an insurance company — a regulated, capitalized, governed entity that takes years to operate well and demands real actuarial, legal, and operational discipline. The companies that get the most out of captives treat them as long-term strategic risk-financing infrastructure: they fund them adequately, govern them seriously, price coverage with real actuarial analysis, and refresh the strategy as their risk profile evolves.

For actuaries, captives offer some of the most intellectually engaging work in the profession. The data is messier than at a commercial insurer, the judgment calls are larger, the audience is closer to the decision (you’re often presenting directly to the parent’s CFO or risk committee), and the structural choices — domicile, tax election, capital target, reinsurance program, cell vs. wholly-owned — all turn on actuarial inputs. Whether you’re building the feasibility study for a first-time captive, signing the annual statement of actuarial opinion for a fifteen-year-old single-parent program, or pricing a loss portfolio transfer for a captive in runoff, the work sits at the intersection of insurance, corporate finance, tax, and regulation. It is one of the corners of the actuarial profession where being a generalist with depth in each of those areas pays off.

Sources and Further Reading

Vermont Department of Financial Regulation, Captive Insurance Division — dfr.vermont.gov
Bermuda Monetary Authority, Insurance Sector pages — bma.bm
Cayman Islands Monetary Authority, Insurance Supervision pages — cima.ky
Internal Revenue Service, “Abusive Tax Shelters and Transactions” — irs.gov
U.S. Tax Court, Avrahami v. Commissioner, 149 T.C. No. 7 (2017)
U.S. Tax Court, Reserve Mechanical Corp. v. Commissioner, T.C. Memo 2018-86
U.S. Supreme Court, CIC Services, LLC v. IRS, 593 U.S. ___ (2021)
Federal Liability Risk Retention Act of 1986, 15 U.S.C. ch. 65
NAIC, “Captive Insurance Companies” CIPR topic page — content.naic.org
Captive.com industry news and learning center — captive.com
Captive Insurance Companies Association (CICA) — cicaworld.com
LossReserves.com, captive lifecycle and reserving series — lossreserves.com/learn/captive
LossReserves.com, “LPT, ADC, and Captive Runoff: A Board-Ready Walkthrough” — lossreserves.com/learn/lpt-adc-captive-runoff
Marsh, “Captive Landscape Report” (annual)
Aon, “Global Risk Management Survey” (biennial) — captive prevalence and use case data