Captives, Protected Cells, and Mutuals: Organized Self-Insurance
Book: Structured Finance and Insurance: The ART of Managing Capital and Risk Author: Christopher L. Culp Publisher: Wiley Finance, 2006 ISBN: 978-0-471-70631-1
Chapter 23 takes us into the world of organized self-insurance. When a company decides to retain a risk rather than transfer it, it still needs a way to finance potential losses from that risk. You can’t just say “we’ll deal with it when it happens” and expect investors to be happy. There needs to be a credible, organized plan for funding those losses before they occur. That’s what captives, protected cell companies, and mutuals are all about.
The Cookie Jar Problem
Before we get to captives, Culp explains why simple self-insurance is harder than it sounds. The most obvious way to prefund a retained risk is to set up a reserve. Put money aside and call it your insurance fund. Problem is, nobody trusts reserves.
Reserves can be reversed too easily. A company can raid the reserve for unrelated purposes. Late in 2004, a major U.S. manufacturing firm reversed a product liability reserve with no explanation and its stock price immediately dropped. Investors are suspicious of reserves for good reason.
Culp calls this the “cookie jar” problem. And it’s not just an accounting or perception issue. The temptation to reverse a reserve is a real risk. Remember the underinvestment problem: a company strapped for cash might be genuinely tempted to grab money from a reserve earmarked for future losses. So the cookie jar problem is about credibility and about actual behavior.
Even accounting rules can work against self-insurance. If a loss is probable but impossible to quantify, the firm can’t take an accounting reserve for it. The money might be set aside internally, but it won’t show up in the financial statements as being committed to anything.
There are some workarounds. Increasing the deductible on an existing insurance policy creates a retention. Self-insurance pools let multiple firms pool their self-insurance funds (not their risks, just the funds) to smooth the timing of loss payouts. But these are limited solutions.
Enter the Captive
A captive insurance company is a wholly owned subsidiary set up by a firm specifically to insure its own retained risks. The sponsor capitalizes the captive by buying its equity. The captive invests those funds in low-risk marketable securities that serve as reserves. The sponsor then pays insurance premiums to the captive in exchange for coverage.
Think of it as moving the cookie jar into a separate room with its own lock and its own management team. The money is still the company’s money. But now it’s governed by an actual insurance company with proper reserves, proper licensing, and proper accounting. It’s much harder to raid.
Here’s a concrete example. BigChip, a silicon chip manufacturer, faces three risks: damage to chips in transit, California earthquake damage to its headquarters, and yen/dollar exchange rate fluctuations. BigChip sets up BigChip Insurance Co. and capitalizes it by buying 100 percent of the stock.
For chip damage (high frequency, low severity), BigChip buys per-occurrence insurance from its own captive. The captive retains 100 percent of this risk and funds losses from its equity and reserves. For earthquake risk (low frequency, high severity), BigChip Insurance buys reinsurance from external reinsurers. The captive passes through the risk. For currency risk, BigChip and its captive execute mirroring derivatives, with the captive passing the risk to swap dealers.
The captive manages all three risks. It retains the ones the parent wants to keep and transfers the ones it doesn’t. This is enterprise-wide risk management through a dedicated subsidiary.
The Benefits of Captives
Underwriting profits and investment income stay in-house. If actual losses are lower than expected, the sponsor gets the savings back as dividends.
Access to the reinsurance market. Companies can’t easily buy reinsurance directly. But a captive, as a licensed insurer, can. Reinsurance tends to be less regulated, more sophisticated, and more flexible than primary insurance. For many firms, this alone justifies setting up a captive.
Lower adverse selection costs. When you insure yourself, there’s no lemons problem. You know your own risks perfectly. Traditional insurers charge a premium for the uncertainty they face about your risk profile. Self-insurance avoids that.
Cash flow flexibility. Internal insurance premiums can be timed to suit the sponsor. Traditional insurance requires upfront annual premium payment.
Reduced agency costs. The captive solves the cookie jar problem by creating a transparent, purpose-built entity. Monitoring a captive is straightforward because it does one thing: manage the sponsor’s retained risks.
Captive Variations
Single-parent captive insurer is the basic version described above. The sponsor buys 100 percent of the captive’s equity.
Single-parent captive reinsurer with a fronting insurer is used when local laws require local insurance coverage. The sponsor buys a policy from a local insurer, which then cedes the risk to the captive reinsurer. The fronting insurer charges 5 to 30 percent of premiums for this service. The vast majority of captives today are actually reinsurance companies because reinsurance regulation is lighter.
Multibranch captive reinsurers serve multinational companies that need local insurance in multiple jurisdictions. Each branch pays premium through its own fronting insurer, and everything flows back to a single captive reinsurer.
Taxation
In the U.S., the IRS questioned whether premiums paid to a single-parent captive were really just capital infusions. By 1992, a practical test emerged: premiums paid by a sponsor to its captive are tax deductible if the captive writes at least 30 percent of its business to unrelated third parties. This 30 percent rule also affects the taxation of investment income. Without unrelated business, the captive’s net income gets taxed at the sponsor level.
Rent-a-Captives and Protected Cell Companies
Not every company can justify the expense of setting up its own captive. The minimum premium for a single-parent captive to make economic sense is around $700,000 per year. For smaller firms, there are alternatives.
Rent-a-captives are multi-participant structures set up and owned by (re)insurance companies or brokers. Individual firms don’t own any of the captive’s equity. Instead, they pay premiums through a fronting insurer, the rent-a-captive tracks each participant’s account, and underwriting profits and investment income are tracked and potentially returned.
The catch is insolvency risk. If one participant submits a huge claim that exceeds its account, the rent-a-captive uses general funds. If that participant can’t repay the negative balance, other participants might end up subsidizing the loss. This is called “ex post mutualization,” and it made people nervous.
Protected cell companies (PCCs) were invented around 1997 to fix this. They work like rent-a-captives but with legally ring-fenced customer accounts. Each cell is protected from the liabilities of other cells. In some cases, the PCC is set up as a master trust with separate affiliated trusts for each customer. This reduces the minimum premium to as little as $100,000 per year. Neither the insolvency risk of rent-a-captives nor the legal protection of PCCs has ever been fully tested in court, though.
Mutuals: Self-Insurance Syndicates
The last piece of the chapter covers mutuals, but not the big ones like Liberty Mutual or Mass Mutual. Culp focuses on small, specialized mutual insurance companies with fewer than a hundred members (sometimes as few as 5 to 10). He calls them “self-insurance syndicates.”
These are groups of firms that both own and purchase insurance from the same company. Each member contributes equity and pays premiums. The mutual underwrites specific risks for its members, often at high attachment points complementing existing captive programs.
Why does mutualization work? Two reasons. Portfolio diversification: if the risks of multiple firms are not perfectly correlated, combining them reduces total risk. Better statistical inference: the central limit theorem tells us that the average loss from a large group of exposures is approximately normally distributed, regardless of the individual distributions. This makes expected losses easier to predict and premiums easier to set.
The design of a good mutual requires balancing heterogeneity in risk exposures (for diversification) with homogeneity in member quality (to avoid cross-subsidies). Members should have similar size and financial strength. Industry-specific mutuals make sense when the underlying risk exposures are uncorrelated across otherwise similar firms.
Oil Insurance Limited (OIL) is a great example. A Bermuda-based mutual serving the energy industry since 1972, now with about 84 members. Members must have at least $1 billion in gross assets, derive at least 50 percent of revenues from energy operations, and have a minimum BBB-/Baa3 credit rating. OIL covers property damage, well control risks, and environmental liability up to $250 million per occurrence with a $1 billion aggregation limit for correlated events.
OIL handles the cross-subsidy problem by running two separate risk pools. The $5 million excess of $5 million layer (which is working capital for some members but not others) is pooled separately from policies attaching at $10 million and up. This prevents low-attachment-point members from being subsidized by high-attachment-point members.
In 2004, OIL had $443 million in net premiums earned, $238 million in investment income, and $777 million in underwriting losses from 12 loss events. That’s what happens when you insure the energy industry against catastrophic property damage.
Captives and Mutuals Are Starting Points, Not Endpoints
Culp ends with an important observation. Captives and mutuals are rarely the final destination for risk. They’re the first step. A captive manages the retention and the working capital layer, but it also buys reinsurance for risks above the retention. It uses derivatives for financial risks. It might participate in a mutual for catastrophic layers.
Mutuals themselves aren’t just passive pools. They buy retrocession, use finite risk products, and increasingly employ the structured insurance solutions covered in upcoming chapters. Even the mutual’s own credit risk can be managed. If a participant might default, the mutual can buy CDS or EDS protection on that member.
The takeaway: captives and mutuals are organizational structures for risk management, not products. They’re the framework within which a company deploys the full range of risk finance and risk transfer tools.
My Take
This chapter is less glamorous than cat bonds or synthetic CDOs, but it’s arguably more important for understanding how real companies actually manage risk. Most large corporations have a captive. The cookie jar problem is real. The tax considerations are real. The choice between retaining risk and transferring it is not binary. It’s a continuous spectrum, and captives are the tool that lets companies fine-tune their position on that spectrum.
I also appreciate the discussion of mutuals. There’s something elegant about a small group of similar companies pooling their risks and sharing the statistical benefits of diversification. It’s insurance stripped down to its most fundamental principle: a group of entities agreeing to share the burden of uncertain losses.
The OIL example brings this to life. Eighty-four energy companies, each too big to be insured cheaply by traditional markets, agreeing to insure each other for catastrophic property damage. And then carefully designing the structure to prevent cross-subsidies and manage correlated risk. That’s risk management as a cooperative enterprise.