Market StructureChapter 3122 min read

Captives & Self-Insurance, why Fortune 500s buy insurance from subsidiaries they own.

A captive is an insurance company owned by its insureds. Single-parent captives, group captives, RRGs, and protected cell facilities provide structured alternatives to the commercial market. Self-insurance is the simpler version: retaining risk on the balance sheet without forming a separate entity. Both approaches work because, for stable and quantifiable risks, the commercial premium load exceeds the cost of capital required to hold the exposure internally.

§ 01

The mental model

Insurance is risk transfer for a fee. The fee includes expected losses plus a load for capital, expenses, and profit. For some insureds and some risks, the fee is greater than the cost of holding the risk on the balance sheet. Captives and self-insurance are the structures that let those insureds keep the risk internally instead of paying the load.

The structural difference between captives and self-insurance is form, not substance. A captive is a separately incorporated insurance entity, formally licensed and regulated, with its own balance sheet, its own claims operation, and its own actuarial function. Self-insurance is the absence of a transfer; the risk stays on the parent's balance sheet, often with a defined accrual mechanism but without a separate entity. Both approaches accept that the buyer is also (functionally) the underwriter for the retained portion of the risk. The buyer concludes that the long-run cost of holding the exposure is less than the long-run cost of paying premiums to a third party. For predictable, frequency-dominated risks (workers compensation, property all risk on a stable schedule, auto liability for a fleet), this calculation often favors retention.

Anchor concept

Captives are insurance companies owned by their insureds. The economics work when the parent's expected loss costs are stable and predictable, when the parent has the financial capacity to absorb timing volatility, and when the captive structure provides operational or tax benefits beyond pure risk transfer.

§ 02

Why a company forms a captive

The reasons are economic, operational, strategic, and (occasionally) tax-driven. A captive feasibility study typically examines all of them.

Economic

  • Premium savings. The captive captures the loading that commercial carriers add to expected losses (capital cost, expenses, profit). For high-frequency, low-severity exposure with stable loss patterns, this load can be substantial.
  • Retention of underwriting profit. When the captive runs at a profit (losses below expectation), the retained profit accrues to the parent rather than the commercial carrier.
  • Investment income on reserves. The captive holds reserves for unpaid losses; investment income on those reserves accrues to the captive (and ultimately the parent) rather than the commercial carrier.

Operational

  • Coverage flexibility. The captive can write coverage that the commercial market does not offer at acceptable terms, including bespoke wordings, broader trigger language, and customized retention structures.
  • Risk management discipline. The captive structure forces the parent to quantify exposure, set explicit retention levels, and integrate loss control with insurance operations.
  • Multinational coordination. A captive can act as a single risk hub for global operations, simplifying coverage gaps and overlaps across jurisdictions.

Strategic

  • Reinsurance access. The captive can buy reinsurance from the global reinsurance market, which is often more capital-efficient than purchasing primary commercial insurance.
  • Risk financing optimization. Layering, retention design, and capital deployment can be optimized across the parent's full risk profile.
  • Market cycle management. When commercial markets harden, the captive can absorb additional risk without buying expensive insurance. When markets soften, the captive can transfer risk back to the commercial market.

Tax

Properly structured captive arrangements can produce tax efficiency. Premiums paid to the captive are deductible business expenses; reserves established by the captive are deductible (subject to the rules in §07). The IRS has scrutinized captive arrangements heavily, and abusive structures (particularly micro-captives marketed as tax shelters) face significant enforcement risk. Legitimate captives focus on the economic and operational rationale, with tax treatment as a consequence rather than a primary motivation.

§ 03

Single-parent and group captives

The two dominant captive structures are single-parent and group, and the choice depends on the parent organization's profile.

Single-parent captive

Owned by a single corporate parent. The captive insures the parent and its subsidiaries. This is the structure used by virtually all large U.S. corporations with captive operations: Walmart, ExxonMobil, Microsoft, JPMorgan, the Fortune 500 generally. The single-parent structure provides full control over the captive's underwriting, claims, and capital allocation.

Group captive

Owned by multiple corporate participants who share homogeneous risks. Common structures include industry-specific group captives (a hospital group captive, a trucking group captive, a contractors group captive). Group captives provide scale efficiencies that wouldn't be available to any single member, plus cross-subsidization smoothing across members' loss experiences.

Heterogeneous group captives

Less common but increasingly used: group captives where members are diverse in industry but share risk-management discipline. The diversification benefit is from the diverse loss experience patterns rather than from industry similarity. Risk Strategies, Captive Resources, and a small number of specialty captive managers operate heterogeneous group captives serving mid-market insureds.

Sponsored captives

A captive owned by a sponsor (typically a broker, MGA, or insurance company) but financially structured to serve multiple client cells, members, or programs. The sponsor handles administration; clients participate without forming their own entity.

§ 04

RRGs and protected cells

Two structures merit specific treatment.

Risk Retention Groups (RRGs)

Created by the federal Liability Risk Retention Act of 1986, RRGs are group captives chartered in one state but authorized to write liability insurance in all 50 states without obtaining licensure in each. The federal preemption is narrow (liability only, RRG members must be similar businesses), but where it applies it provides substantial regulatory efficiency.

RRG users

  • Hospitals and physician groups for medical professional liability.
  • Construction contractors for general liability.
  • Long-term care facilities for professional liability.
  • Truckers and motor carriers.
  • Other industry-specific groups with concentrated liability exposure.

Limitations

RRGs cannot write property, workers compensation, or any first-party coverage. They cannot offer guarantee fund protection (state guarantee funds do not cover RRG policies). And they face state regulatory friction in some jurisdictions despite the federal preemption.

Protected cell companies (PCCs) / Segregated Account Companies (SACs)

A single licensed insurance entity that holds multiple legally segregated cells, each with its own assets and liabilities firewalled from the others. The cell structure allows multiple insureds to participate in a captive arrangement without creating a separate entity for each.

Why this matters

  • Lower cost of entry: forming a cell within an existing PCC is much faster and cheaper than chartering a new captive.
  • Operational efficiency: the PCC handles licensing, financial reporting, and regulatory interaction at the entity level; cells handle their own underwriting and claims.
  • Bermuda, Vermont, Cayman, Guernsey, and several other jurisdictions support PCC structures.
  • Common in healthcare, sponsored arrangements, and industry-specific group programs.

Incorporated cell companies (ICCs)

A variant where each cell is a separate legal entity but operates under a common ICC umbrella. ICCs provide stronger asset segregation than PCCs at higher administrative cost. Used in jurisdictions where the legal status of unincorporated cells is contested.

§ 05

Domiciles

A captive must be domiciled somewhere; the choice of domicile is consequential.

Vermont

The largest U.S. captive domicile by entity count. Mature regulatory framework, deep service-provider ecosystem (captive managers, accountants, attorneys, actuaries), and a state insurance department with dedicated captive expertise. Vermont is the default choice for many U.S.-headquartered single-parent captives.

Other major U.S. domiciles

  • Hawaii. Established captive jurisdiction, particularly for Asia-Pacific risk and for entities with operations in Hawaii.
  • Utah. Growing rapidly, particularly for cell company structures.
  • South Carolina. Established mid-market and group captive domicile.
  • Delaware, Tennessee, North Carolina, Texas, Connecticut, Nevada, Arizona, D.C. Smaller but active domiciles, each with specific niches.

Bermuda

The largest offshore captive domicile globally. Sophisticated insurance regulator (the BMA), strong service-provider ecosystem, and recognition as a Solvency II equivalent jurisdiction. Bermuda is particularly common for property catastrophe risk, healthcare professional liability captives, large multinational corporate captives, and reinsurance-focused captives.

Cayman Islands

Particularly strong for healthcare captives (notably for U.S. hospital groups) and for segregated portfolio company structures. Cayman captives often combine onshore parent structures with Cayman risk financing entities.

Guernsey, Isle of Man, Luxembourg, Ireland

European domiciles primarily serving European-headquartered captive parents. Guernsey is the largest European captive domicile. Luxembourg and Ireland are growing for Solvency II-aligned captive structures.

The choice criteria

  • Regulatory sophistication and reputation.
  • Capital and surplus requirements.
  • Tax and accounting treatment.
  • Service-provider ecosystem.
  • Reinsurance arrangements (recognized credit-for-reinsurance status).
  • Geographic and time-zone alignment with the parent's operations.
§ 06

Fronting and reinsurance

Most captives operate within an integrated structure that includes a fronting carrier and external reinsurance. Pure self-insurance through a captive is rare.

Fronting

A licensed admitted carrier (the fronting carrier) issues policies on its paper to the parent and its subsidiaries. The fronting carrier then cedes the underwriting risk to the captive through reinsurance. The fronting carrier earns a fronting fee for its role; the captive bears the underlying loss exposure.

Why fronting is needed

  • Admitted paper. Many insureds need policies issued by admitted carriers for compliance, contractual, or financial reasons. Bermuda or Vermont captives are not admitted carriers in most U.S. states.
  • Filings. Workers compensation, auto liability, and several other lines must be filed with state insurance departments. Captive-issued policies cannot satisfy these filings.
  • Counterparty acceptance. Lenders, contractual counterparties, and certificate holders often require admitted-carrier paper.

Fronting fees

Typically 2-6% of premium, depending on the fronting carrier's role, the captive's financial strength, and the lines involved. The fronting carrier provides credit support (the carrier is on the hook to claimants if the captive fails to pay) and assumes residual claims-handling responsibility.

Reinsurance

Most captives buy excess reinsurance to limit their exposure to severe losses. The captive retains a primary layer (often $1M-$10M per loss), with reinsurance covering excess. Reinsurance is typically purchased through a London-based broker who places the program with Lloyd's syndicates and the global reinsurance market.

Common structures

  • Per-occurrence excess. Reinsurance attaches above a per-loss retention, providing protection against severity.
  • Aggregate stop-loss. Reinsurance attaches above a defined aggregate loss for the year, providing protection against frequency.
  • Quota share. The captive retains a defined percentage of every loss, with the reinsurer assuming the balance.

The integrated structure

A typical captive program looks like: parent pays a premium to the fronting carrier, fronting carrier issues admitted policies and cedes to the captive (less the fronting fee), captive holds the primary risk and buys reinsurance for severity protection, captive's financial results consolidate into the parent's results.

§ 07

Tax considerations

The tax treatment of captives is consequential and heavily regulated.

The deduction question

For premiums paid to the captive to be deductible business expenses (and for reserves established by the captive to be deductible), the arrangement must qualify as insurance for federal tax purposes. The IRS and the courts have developed a body of doctrine on what constitutes insurance for these purposes.

Risk shifting and risk distribution

Two requirements that the IRS treats as essential to insurance:

  • Risk shifting. The economic risk of loss is transferred from the insured to the captive.
  • Risk distribution. The captive insures a sufficiently broad pool of risks that statistical predictability applies. The IRS has historically required at least 12 unrelated insureds or the equivalent through brother-sister arrangements.

The micro-captive issue

Section 831(b) of the Internal Revenue Code provides favorable tax treatment for small captives (annual premiums up to a defined threshold, currently around $2.4M). The treatment elects taxation only on investment income, not on underwriting income. Properly structured 831(b) captives serve legitimate purposes for mid-market insureds. Improperly structured 831(b) captives have been heavily marketed as tax shelters; the IRS has identified abusive 831(b) structures as a "listed transaction" requiring disclosure and has pursued enforcement aggressively.

Bermuda and offshore captive taxation

U.S. parent companies can elect to treat their offshore captives as U.S. taxpayers (Section 953(d) election), which is the standard structure for Bermuda-domiciled U.S.-parent captives. Without the election, the captive is a controlled foreign corporation subject to Subpart F rules, which generally produces unfavorable tax outcomes.

Tax structure should be a consequence, not the purpose

Captive arrangements driven primarily by tax efficiency face IRS scrutiny and often unwind under audit. Captives with substantial economic and operational rationale (premium savings, coverage flexibility, risk management discipline) generally hold up. The captive feasibility study should focus on the economic case; the tax treatment follows from the structure.

§ 08

When self-insurance is the better answer

Captives are not always the right structure for retained risk. Sometimes simple self-insurance, with no separate entity, makes more sense.

Why self-insurance can win

  • Lower administrative cost. No separate entity to charter, no captive manager, no audit, no actuarial study, no domicile fees. The accrual sits on the parent's balance sheet.
  • Simpler tax treatment. Self-insurance reserves are not deductible until losses are actually paid (with limited exceptions for defined statutory frameworks). The captive deduction structure is more efficient, but the simplicity of self-insurance can offset the tax disadvantage at smaller scale.
  • Direct cash flow management. Self-insurance funds remain in the parent's working capital. Captive funds are committed to the captive entity and have constraints on use.

Statutory self-insurance frameworks

  • Workers compensation self-insurance. Most states allow large employers to self-insure their workers comp obligation, subject to financial qualification, security deposits, and ongoing reporting. The largest self-insured WC employers are entire industries (Walmart, the federal government, large hospital systems).
  • Auto liability. Self-insurance certifications are available in most states for fleet operators.
  • Health benefits. ERISA self-insurance for health benefits is the dominant structure for U.S. mid-to-large employers, often through a third-party administrator with a stop-loss insurance overlay.

The hybrid model

Many large insureds combine self-insurance for a primary layer with captive participation for an excess layer. The simplest layer of frequency exposure stays on the parent's books; the more volatile excess layer goes through the captive structure for risk financing efficiency.

§ 09

Where IDP earns its keep

Captive operations generate substantial documentation: actuarial reports, audited financials, regulatory filings, claims summaries, underwriting files, fronting carrier reports, reinsurance documentation. The captive manager handles most of this directly, but the parent's risk management and finance functions consume the outputs and use them in broader planning.

1
Intake
Captive financials, actuarial reports, fronting reports, claims data, reinsurance documentation.
2
Classify
Single-parent vs group vs cell; line of business; domicile-specific reporting requirements.
3
Extract
Reserves by line and accident year, paid losses, IBNR development, reinsurance recoverables.
4
Validate
Reconcile actuarial development with prior year reports; flag adverse or favorable trends.
5
Triage
Generate parent-level risk-financing dashboard. Flag retention adequacy, reinsurance gaps, capital deployment.
6
Underwriter
Pre-built captive performance summary feeding parent risk management committee review.
Indico use cases for captives

The captive operation is information-rich and analysis-poor. Actuarial reports, fronting carrier reports, and audited financials all contain similar data presented in different formats. An extraction agent that reads these documents into a normalized captive performance schema (loss triangles, reserves, paid losses, premiums, expenses) supports both the parent's risk management function and the broader corporate reporting requirements. Secondary extractions: reinsurance contract data extraction, fronting fee reconciliation, multi-cell captive aggregation, parent-to-captive reconciliation.

Chapter 31 · Market Structure · 22 min read

Captives & Self-Insurance — Cheat Sheet

A captive is an insurance company owned by its insureds. Single-parent captives, group captives, RRGs, and protected cell facilities provide structured alternatives to the commercial market. Self-insurance is the simpler version: retaining risk on the balance sheet without forming a separate entity. Both approaches work because, for stable and quantifiable risks, the commercial premium load exceeds the cost of capital required to hold the exposure internally.

The mental model: Captives are insurance companies owned by their insureds. The economics work when the parent's expected loss costs are stable and predictable, when the parent has the financial capacity to absorb timing volatility, and when the captive structure provides operational or tax benefits beyond pure risk transfer.

Key terms

Single-parent captive · One-parent captive insurer
RRG · Risk Retention Group, federally chartered
PCC · Protected Cell Company
Fronting · Admitted-paper issuer with cession
§ 831(b) · Small captive tax election
§ 953(d) · Offshore captive U.S. taxpayer election
Domicile · Captive's licensing jurisdiction

If you remember three things

Captives are insurance companies owned by their insureds, formed when the cost of holding risk internally is less than the commercial premium load. The captive structure is almost always integrated with a fronting carrier and excess reinsurance, not a pure stand-alone risk financing vehicle. Tax considerations are real but should follow from economic substance, not drive the structure, since IRS scrutiny of tax-motivated captives is heavy and growing.