When Are Premiums Paid to a Captive Insurance Company Deductible for Federal Income Tax Purposes?

Generally, premiums paid for insurance are deductible for federal income tax purposes in the year paid if the policy is an annual policy and are amortized over the policy period for a multiyear policy. In addressing the question of when premiums paid to a captive insurance company are deductible for federal income tax purposes, the key determination is whether the coverage provided by the captive will be respected as insurance for federal income tax purposes. A number of factors need to be taken into account when making that determination. The basic requirements for insurance treatment are insurance risk, risk transfer, risk distribution, and a policy that embraces common notions of insurance.

Risk Transfer

This element addresses the fact that there must be a transfer of risk from the insured to the captive insurance company.

The Internal Revenue Service (IRS) has taken the position, and courts have agreed, that if a party ("Parent") owns a captive that insures its Parent's risk, there is no risk transfer because economically Parent is in the same position as it was without the insurance (i.e., if the captive receives a premium or pays a loss, the Parent's balance sheet is unaffected).

Alternatively, case law and the IRS have recognized that, if a captive that insures Parent risk also insures sufficient "unrelated risk," the Parent risk will be shifted along with the unrelated risk. An IRS safe harbor ruling quantifies sufficient unrelated risk at 50 percent, but a Ninth Circuit case, Harper, involved just under 30 percent unrelated risk. Thus, questions arise as to what constitutes unrelated risk as there is no specific definition.  Common risks that generally are considered unrelated risks include risks of customers of the captive's corporate group, extended warranty risk, risk relating to employee benefits provided to employees of the corporate group of which the captive insurer is a part, and risks that are derived from a valid risk pool.

Also, although there is no legal authority on point, the general rule of thumb is that, in determining the percentage of unrelated risk, one should look to the net retained risk in the captive. Thus, for example, if one were to write a large premium relating to unrelated risk and reinsure it 100 percent to a third party, there would be no unrelated risk to take into account.

If a captive insures brother/sister risk (i.e., risk of companies owned by the same Parent that owns the captive), generally, that risk is shifted provided the policy does not contain features that effectively eliminate the risk. Thus, for example, if a policy provides for limits equal to premium, there will be no transfer of risk. In addition, the captive must have resources (e.g., assets or reinsurance) sufficient to pay losses covered under a policy between an insured and the captive, or transfer of risk may be found lacking. Risk mitigating policy features also can impact whether a parent-subsidiary transaction (or any insurance transaction) qualifies as insurance.

Risk Distribution

Risk distribution addresses the concept of receiving premium attributable to various unrelated risks.

Until recently, based on existing case law and IRS rulings, the focus in determining whether adequate risk distribution was present was the number of insureds and the amount of risk (generally determined by premium allocation) attributable to each. In this regard, the IRS in Rev. Rul. 2002–90 published a "safe harbor" indicating that if a captive covered the risk of 12 insureds each accounting for between 5 percent and 15 percent of the risk, there was adequate risk distribution. Various rulings have refined this test. In one ruling, the IRS noted that if a captive covers a single insured that accounts for 90 percent of the risk the captive covers, there is inadequate risk distribution.   Unrelated risk insured directly or through a valid reinsurance pool could add to the number of insureds and the spread of the risk, providing a strong risk distribution position under the IRS rulings.

At the end of 2014 and beginning of 2015, the Tax Court decided two cases, Rent-A-Center and Securitas, which involved captives that insured small numbers of brother/sister entities. The Court determined that the captive covered a significant number of independent exposure units (e.g., locations, automobiles, employees) and, accordingly, the captive has sufficient risk distribution. The Tax Court unfortunately did not provide a basis for determining when there would be a sufficient number of independent exposure units, but found the independent exposure units to be sufficient on the facts of those particular cases. This, obviously, leads to some questions about when, in any particular case going forward, the facts related to exposure units will be sufficiently different from the facts of Rent-a-Center and Securitas to prevent reliance on those cases.

Insurance Risk and Common Notions of Insurance

The IRS has also raised issues regarding whether certain types of coverage should be treated as insurance for federal income tax purposes. Thus, for example, imbedded warranty, retroactive coverage (i.e., coverage for an event that has already happened), and currency value protection have been questioned. In the R.V.I. Guaranty case, the IRS litigated, and lost, the question of whether residual value insurance was insurance for tax purposes.

In addition, the IRS has questioned other elements of captive formation that to some degree overlap with the risk transfer and risk distribution categories, for example, whether loanbacks to insureds or guarantees provided by the parent or another member of the consolidated group affect insurance status, whether unrelated business must be "homogeneous" or similar to the related business written by the captive, and whether the method of calculating premium (e.g., actuarially or by some other method), and its allocation among subsidiaries or members of a corporate group was appropriate. The IRS often challenges these elements by arguing that they are contrary to common notions of insurance.

Many of the cases concluding insurance status was applicable have found that the transactions at issue embrace common notions of insurance, identifying certain elements without developing a specific definition of the concept. In the R.V.I. Guaranty case, the Tax Court synthesized the analyses in a number of prior cases and applied the following factors to determine that the arrangement at issue involved insurance in its commonly accepted sense: (1) whether the insurance company was organized, operated and regulated as an insurance company by the states in which it did business, (2) whether the insurance company was adequately capitalized, (3) whether the residual value insurance policies were valid and binding, (4) whether the premiums were reasonable in relation to the risk of loss covered under the policies, and (5) whether the premiums were duly paid and the loss claims were duly satisfied.

The Small Captive Cases

There are two small insurance company tax regimes (IRC 831(b) and IRC 501(c)(15)) designed to give small insurance companies the benefit of alternative tax regimes that may provide for lesser overall federal tax payments. Taxpayers and the IRS have litigated several cases considering whether the transactions of the small captive insurance companies in question were insurance and each was decided against the taxpayers. These cases generally focused on risk distribution and common notions of insurance. Each of the litigated cases had many facts that the Tax Court found to be problematic for insurance characterization.

In those cases, when analyzing risk distribution, the courts have considered both the number of insureds and the number of exposure units covered by the captive. In most cases, the courts have found risk distribution to be lacking under both measures. The facts of those cases generally involved exposure unit numbers that were substantially smaller than the numbers of exposure units in the Rent-a-Center and Securitas cases. Thus, there remains a gray area between the very large numbers of exposure units in Rent-a-Center and Securitas and the very small numbers of exposure units in the small captive cases.

In many of the small captive cases, the captive sought to increase its risk distribution by participating in a risk pool. However, the courts have found the specific risk pools at issue lacking. In particular, the courts concluded that the transactions with the risk pools failed to qualify as insurance because they lacked meaningful risk transfer and they operated in a way that did not reflect common notions of insurance. In an appeal of one of the cases, the Tenth Circuit emphasized that it was not opining on risk pools generally, just the one in the case before it. If correctly designed, a pool can help a captive achieve greater risk distribution.

In the cases involving small captives, the courts have identified a number of issues that led the court to conclude that the captives in the cases did not operate in a way that reflected common notions of insurance. Common issues included inappropriate pricing methods, lack of claims and inadequate claims handling, coverages that were not reasonably suited to the need of the insureds, and illiquid assets. In other cases, the courts identified other issues that were not consistent with common notions of insurance, such as policies that were not timely issued, payment of premiums before the policies were written and the premiums were finalized, manipulation of premiums to achieve the desired tax result, and failure to follow policy formalities, among others.

Although this article does not provide a complete discussion of all of the issues described, it provides a meaningful start to an analysis. For much greater detail, please see "Tax Implications of Risk Financing," by P. Bruce Wright, Esq., and Saren Goldner, Esq., in the section titled "Insurance Characterization for US Federal Income Tax Purposes" of the IRMI reference service Risk Financing. If you subscribe in IRMI Online, you will find the discussion here. If you subscribe in ReferenceConnect, you will find the discussion here.

Ms. Goldner is a partner in the New York office of Eversheds Sutherland, Ms. Rizzolo is a Senior Counsel in the Washington, DC office of Eversheds Sutherland, and Mr. Wright is Of Counsel in the New York office of Eversheds Sutherland.