Spotlight On Captive Insurance

Spotlight On Captive Insurance | Norton Rose Fulbright

December 01, 2001

By Heléna Klumpp

More and more companies appear to be setting up captive insurance subsidiaries. In 2000, the number of captives increased worldwide by 31.6% over the prior year, climbing to more than 3,400. Insurance industry specialist A.M. Best predicts that the 2001 figures will show continued strong growth in the number of captives.

Captives became popular as companies sought ways around the skyrocketing insurance premiums that resulted from a tightening of the insurance market in the mid-1980’s. Although at first glance self insurance appears to be the best way to cut costs, amounts set aside in a self-insurance reserve are not deductible for tax purposes (as compared to traditional insurance premiums, which are). The captive market boomed as companies figured out arrangements that allowed them to enjoy the best of both worlds: lower insurance costs plus deductibility of premiums.


In the simplest, or “single parent,” captive structure, a parent company establishes a wholly-owned subsidiary to insure or reinsure the parent company or other companies in the parent company’s group. The subsidiary, known as a “captive insurance company” or simply a “captive,” is incorporated in a US state or foreign country that has a captive insurance statute. Nearly a third of the world’s captive insurance companies are located in Bermuda. The next most popular locations are the Cayman Islands, Vermont, Guernsey, Luxembourg, Barbados, the British Virgin Islands, Ireland, the Isle of Man and Hawaii. Washington, DC established a captive insurance regime in 2000.

Offshore jurisdictions are popular for several reasons, most notably for the fact that some maintain less stringent regulatory requirements for insurers and reinsurers. For example, many US states strictly regulate the type of investments an insurance company can make. Many offshore captive jurisdictions have no such restrictions. Also, when they first came onto the insurance scene, captives that were established in tax havens like Bermuda or the Cayman Islands were able to avoid income taxes on their underwriting and investment profits. The US tax laws have since evened this playing field, requiring current taxation in the US of the income of most offshore captive insurance companies that are owned by US parents. (There are some exceptions to this rule for offshore captives of US parents that insure risks located in the captive’s home country or in any other country but the US. Those exceptions are set to expire at the end of this year, but may be extended by Congress.) Some state premium taxes may still be avoided by incorporating a captive in an offshore jurisdiction.

The captive can be used to insure against the same types of risks for which a corporation would use a third-party insurer or self-insurance program: property and casualty, general and product liability, workers’ compensation, etc.

In a different twist on the same basic idea, the captive may reinsure primary liability coverage written for other members of its group by a third-party insurer. Most captive insurance companies are reinsurers. This is true for a number of reasons, especially the fact that a third party who deals with the insured party — for example, a lender — may be more comfortable if a recognizable insurance name is the company’s primary insurer. Also,in many jurisdictions the reporting and capitalization requirements are less onerous for reinsurers than for primary insurers.

Other variations on the basic captive structure include several different types of “group captives,” in which a single captive is owned by a number of unrelated corporate parents. The parents are typically similar in size or in the same industry or both.


Insuring through a captive may be attractive for a number of reasons.

First, a captive may be able to charge lower premiums based on a company’s loss experience. A third-party insurer bases the premiums it charges on industry loss averages. A company whose loss experience is lower than the norm may thus pay a rate that takes into account the higher average. When a company in that position insures through a captive, its loss experience will be viewed in isolation and the premiums will be priced lower accordingly. In other words, the insured’s premiums will be based on its particular level of risk instead of an insurance carrier’s general view of the market for that type of coverage. Thus, a company whose loss experience is lower than average may benefit from using a captive.

In addition, a captive insurer — particularly a reinsurer — will save on overhead costs associated with commissions, marketing, claims handling and regulatory compliance. These costs can be high for an insurance company and may represent close to half of an insured’s premium. By contrast, the overhead of a captive is insignificant. This is especially true in the case of an offshore captive, which is less likely to be burdened with extensive regulatory requirements. However, in many cases, marketing, commission and other acquisition costs will be low or nonexistent for a captive.

Another benefit is that the captive can earn investment income on its premiums and thus potentially reap financial gains for the corporate group that would otherwise be lost to a third-party insurer. This investment income is typically subject to income tax in the US, either directly (where the captive is a US entity) or indirectly through anti-deferral rules that apply to certain offshore entities (where the captive is a foreign entity).

A captive may be able to provide coverage that is unavailable or prohibitively expensive if purchased from third-party insurers. The insurance market is cyclical. Low premiums one year may lead to high premiums the following year to make up for the insurance companies’ losses. A captive’s risks are more isolated than a traditional insurance company and thus it is less subject to the market’s volatility. As a result, it may be able to provide coverage at more stable rates than are typically available in the insurance market, depending on the loss experience of the insured party.

Finally, if the parent company chooses to establish a licensed insurance company (as opposed to a reinsurance captive), that captive can purchase reinsurance in the wholesale reinsurance market directly. It can sometimes be cheaper for a captive to obtain reinsurance than for a traditional insurance company. In addition, a captive can control the extent to which it reinsures its risks. When reinsurance is cheap, the captive can take advantage of the market and buy more coverage; when reinsurance is expensive, the captive can choose to limit its coverage and retain more of the risk itself. There are downsides to this approach, though, as reinsurers may prefer to maintain consistent relationships with insurers.

Tax Deduction

The main reason companies set up separate captive insurance companies rather than simply self insuring directly is that premiums paid to a captive by a US company may be deductible for federal income tax purposes. Case law extending back to the 1920’s establishes the principle that amounts set aside in a self-insurance reserve account are not deductible. Conversely, insurance premiums paid to a captive are deductible for federal income tax purposes as ordinary and necessary business expenses if the arrangement is properly structured.


The tax laws provide a deduction for insurance premiums that are connected with a taxpayer’s trade or business. The tricky part of that sentence is the word “insurance.” If an arrangement does not meet the Internal Revenue Service’s definition of “insurance,” then a taxpayer may not deduct its premiums.

Two things must be true to have “insurance.”  The first is that the arrangement must shift risk. Self insurance is not insurance because it does not meet this part of the test — no risk shifts when a taxpayer sets aside amounts to cover its own potential losses. The second test is that risk must be distributed. This means that the insurer must spread around the risks it assumes. In cases and rulings involving captives, the IRS and the courts have tended to focus more on the first part of the test. The IRS’s main concern is that a taxpayer should not be allowed a deduction for an “expense” it pays to itself. If the risk of loss never shifts, the “premium” is not a true insurance expense.

Last summer, the IRS clarified its position on when and how it will challenge deductions claimed in connection with single-parent captive insurance arrangements. The IRS tries to attack such deductions by arguing that the underlying arrangement is not insurance. Previously, the IRS had relied on a fuzzy “economic family” theory to analyze whether such premiums should be deductible. Under this theory, the IRS looked only to the relationships between the captive and the insured parties to determine whether a captive arrangement was truly insurance or not. However, in the 24 years since the IRS first espoused the economic family theory, not a single court that analyzed a captive insurance situation has fully ascribed to it.

Going forward, the IRS said last summer that it will apply a facts-and-circumstances approach to determine whether a captive arrangement is truly insurance and not just a disguised attempt at self insurance.

A few guiding principles emerge.

First, amounts paid to insure a captive’s parent are not deductible if the parent is the captive’s only insurance customer. For every dollar the parent pays in premiums, it sees a corresponding increase in the value of its stock in the captive. No risk shifts in this scenario. If the captive pays out on a claim for a loss suffered by the parent, then the value of the parent’s stock decreases dollar-for-dollar with the amount of the payout, and thus the parent still feels the entire sting of its loss. The risk has not been distributed. This same principle applies to premiums paid to a third-party insurer by the parent where the insurer cedes the underlying risk to the captive and collects reinsurance premiums from it (where the parent’s risk is the only one the captive reinsures). However, if a fair portion of the captive’s business is written for unrelated companies, then the IRS is more likely to respect the captive’s insurance of the parent. In that case, by distributing its risks among a number of clients, the captive effectively shifts the parent’s risk of loss because the premiums the parent pays may have to be paid out to any third-party customer of the captive who suffers a loss. The case law suggests that an arrangement is acceptable if at least 30% of the captive’s annual business, measured by net premiums, is written for unrelated parties.

Second, the IRS will not challenge an arrangement where a captive insures or reinsures its sister subsidiary, unless factors indicate that either the captive itself or the overall arrangement is a “sham.” If not a sham, insuring a sister subsidiary does shift risk and thus constitutes insurance under the IRS’s test. Unlike the parent-subsidiary situation, if the captive must pay out on a claim for a loss experienced by its sister sub, then the sister sub suffers no corresponding diminution in its assets. To determine whether a “sham” exists, the IRS will look at several factors. Any evidence that the parent “propped up” the subsidiary with a guarantee or that the subsidiary was thinly capitalized points to a finding that the captive is a “sham” entity. The IRS will also look at whether the insured parties faced true, substantial risks and whether the premiums charged by the captive are based on market rates. The IRS will investigate the sub’s business practices — were its activities kept separate from its parent’s? Did it put in place appropriate claims handling procedures as opposed to just paying out on every claimed loss? If the answer is “no” to either question, then the IRS is more likely to suspect that the captive is a sham. The captive also looks more suspicious if it was formed in a jurisdiction in which its activities are loosely regulated — though clearly this factor alone is not dispositive as most offshore captives are located in such jurisdictions.

Third, the IRS said nothing new about group captives in last summer’s ruling. Past guidance suggests that the IRS will not challenge deductions for premiums paid to a group captive if the ownership is diffuse enough. For example, the IRS said in a 1978 ruling that the arrangements between a captive and its 31 unrelated owners were “insurance” for tax purposes. No shareholder owned a controlling interest in the captive and no shareholder’s individual risk coverage exceeded 5% of the total risks insured by the captive.

Speaking at a conference last summer, the primary author of the IRS’s latest ruling on captives described the IRS’s new approach to captive insurance companies as a “sliding scale,” noting that “the closer [a transaction] resembles a commercial, arm’s-length insurance transaction, the better you’ll be.”  This suggests that the IRS will not automatically attack every captive and that a well-structured, sensible arrangement should avoid challenge.