Construction Captive Insurance Companies: Navigating Tax Considerations

When commercial insurance enters a hard market, customers naturally seek to control or minimize cost increases, often considering captive insurance arrangements as a potential tool. As noted in a recent Construction Executive article, the rate increases for Owner Controlled Insurance Programs (OCIPs) and Contractor Controlled Insurance Programs (CCIPs) are expected to range from 25-50% or more for 2021.1 “[D]ue to increasing losses and poor investment returns, rates for most lines of coverage are increasing, capacity is declining, and underwriters are being more selective on which accounts they will even consider.”2 This level of rate increase, combined with the advertised risk control and economic benefits of captive insurance companies, presents a compelling case for establishing a captive insurance company.

On the other hand, the IRS recently published IR-2021-82, which urges participants of abusive microcaptive insurance arrangements to exit from these arrangements.3 A microcaptive is a captive insurance company that has elected to be taxed under IRC §831(b).4 IR-2021-82 specifically notes the recent Tax Court decision in Caylor Land & Development Inc., et. al., v. Commissioner of Internal Revenue (the Caylor case), in which the Tax Court held for the IRS.

The Court determined that the arrangement in question did not qualify for the claimed tax benefits and its decision sustained the accuracy-related penalties.5 IR-2021-82 also noted that, in 2020, the IRS deployed 12 new microcaptive examiner teams to increase the number of exams of what the IRS perceives as abusive microcaptive transactions.6

Does this focus of the IRS against abusive microcaptives mean that a taxpayer should not consider forming a captive insurance company despite the potential economic benefits?

The answer is complex and requires some background. State laws generally govern whether an entity is regulated as an insurance company. But just because a company is regulated as an insurance company for state purposes does not mean that it will be taxed as an insurance company for federal income tax purposes. Although microcaptives are popular, not all captive insurance companies that are taxed as insurance companies elect to be taxed under IRC §831(b). Licensed insurance companies must meet certain criteria annually to be taxed as an insurance company for federal income tax purposes, and additional requirements exist to be taxed as a microinsurance company.

This article will outline the annual requirements to be taxed as an insurance company for federal income tax purposes, discuss why the taxpayer failed to meet these requirements in the Caylor case, briefly discuss the additional microcaptive requirements, and suggest a framework to determine whether to use a captive insurance company and if it should be taxed as an insurance company. It will also cover some state tax issues that companies should include in their annual determination.

Federal Taxation as an Insurance Company

According to IRC §831(c), the term “insurance company” is defined as noted in §816(a). Under IRC §816(a) and from other sources discussing IRC §816(a), the term applies to any company for which more than half of its business during the taxable year involves issuing insurance or annuity contracts or reinsuring risks underwritten by insurance companies. Additionally, since the code does not define “insurance” or “insurance contract” for federal income tax purposes, the definitions of these terms have evolved through case law and rulings issued by the IRS. For instance, the courts and the IRS consistently cite the U.S. Supreme Court case Helvering v. Le Gierse, which held that an arrangement must involve insurance risk, risk shifting, and risk distribution to qualify as insurance for federal income tax purposes.

In the Caylor case, the Tax Court indicated that it used four nonexclusive criteria to distinguish nondeductible self-insurance from deductible insurance: 1) risk shifting; 2) risk distribution; 3) insurance risk; and 4) whether an arrangement looks like commonly accepted notions of insurance.7

The courts and IRS consider or emphasize other factors when determining whether an entity should be taxed as an insurance company, including whether there is a parental guarantee, but these other factors often relate to one of the four criteria (e.g., a parental guarantee might indicate that risk has not been shifted to the captive).

Risk Shifting

In Revenue Ruling 88-72, the IRS indicated that risk shifting occurs when economic risk of loss transfers from the insured to the insurer.8 In Humana Inc. v. Commissioner of Internal Revenue, the Sixth Circuit Court of Appeals used a balance sheet approach to affirm the Tax Court’s holding that a parent corporation could not transfer its risk to a captive insurance subsidiary that only insured related entities.

However, when a captive insures parent company risks as well as a large enough pool of risks from unrelated third parties, guidance from Revenue Ruling 2002-89 has indicated that the parent can shift its risk of loss to a wholly-owned captive insurer. According to Revenue Ruling 2002-89, if greater than 50% of the risk insured by the captive insurance subsidiary represented third-party risk, then risk shifting and risk distribution tests were met even though the captive insured its parent.

According to The Harper Group v. Commissioner of Internal Revenue, risk shifting can be met with a lower percentage of outside risk in some instances. In addition, according to Malone & Hyde Inc. v. Commissioner, facts such as related-party loans or parental guarantees can jeopardize whether a court will find that the risk shifting criteria has been met.

Risk Distribution

The position of the IRS with respect to risk distribution can be found in Revenue Ruling 2002-90, which describes risk distribution as spreading the insured risks among enough small independent risks so the insured entity, in effect, is not paying for its own claims. Revenue Ruling 2002-90 holds that risk distribution was met when the captive insurance company insured 12 separate taxable entities and the risk insured by each entity ranged from 5-15%. Historically, the IRS looked to the number of separate taxable entities that were insured by a captive and the amount of risk each insured entity represented to determine whether risk distribution had been achieved.9

While the IRS continues to focus on the number of insured entities, in more recent cases, the courts have considered the number of independent risks rather than the actual number of insureds. In Rent-A-Center Inc. and Affiliated Subsidiaries v. Commissioner of Internal Revenue, the court concluded that risk distribution existed if multiple entities (in this case, 3,000 stores, more than 14,000 employees, and more than 7,000 vehicles) were insured.

More recently, in the Caylor case, which involved 34 independent exposures, the Tax Court held that risk distribution had not been met.

The Tax Court did not explain how it derived the number of independent exposures, but it appeared to be negatively influenced by its findings that most of the insured entities other than the main company derived the majority of their income to pay the insurance premium from undocumented fees from the main company.10

In addition, as discussed previously, the guidance available indicates that risk distribution can be met when a sufficient third-party premium exists.11 As previously noted, Revenue Ruling 2002-89 held that risk distribution was met if there was sufficient outside premium (greater than 50%). Certain cases, such as The Harper Group v. Commissioner of Internal Revenue, indicate that a lower percentage of outside risk might still result in risk distribution under some circumstances.

No specific guidance exists about whether OCIPs, CCIPs, or subcontractor default insurance (SDI) premiums would qualify as third-party risk, but depending on the specific facts and circumstances, however, it could be argued that some OCIPs, CCIPs, and SDI programs are analogous to other situations that have been determined to be third-party risks.12 This type of determination is fact-dependent, and a risk exists that not all OCIP, CCIP, and SDI programs may qualify as third-party risk.

If you are a CFMA member login to continue reading this article. If you aren't a member yet and would like unlimited access to all of the content on, plus a variety of other benefits, join CFMA today!

About the Author

William F. Buechler, Jr,

Read full bio