Printer Friendly

Despite IRS attempts, captive wall remains intact.

Despite recent legislative changes to the tax treatment of captive insurers and their owners that have effectively removed many of the preconceived tax abuses in the captive insurance company area, the Internal Revenue Service continues to attack the deductibility of premiums paid by U.S. insureds to captive insurers.

It should be remembered that the basic economic benefits of a captive insurer have been, and will continue to be, non-tax benefits. Captive insurers have long provided their owners with insurance at a lower cost than would otherwise be available by allowing insureds who are perceived as better risks to band together, setting their rates at an amount needed to cover their losses and not those of higher risk insureds. As owners of the captive, the insureds have a strong economic incentive to follow a risk management program that minimizes the number and severity of losses and to ensure that other insureds do likewise. In addition, captive insurers can provide their insureds with lower overall insurance costs by providing low cost access to the world reinsurance market without the intercession of, and payment of commissions to, numerous middlemen. Against this background, the tax controversy concerning captives centers around attempts by insureds to keep tax treatment of payments to captives the same as payments to unrelated insurers and attempts by the IRS to treat payments to captives differently.

The basic standard for the characterization of an insurance transaction was set forth by the U.S. Supreme Court in 1941 in Helvering V. LeGierse. There, the court stated that insurance must incorporate elements of risk shifting and risk distribution. In that case the court held that the life insurance and annuity contracts issued to the taxpayer involved were interdependent and must be treated as a single transaction. The court held that the risks purportedly assumed under one contract were offset by the other, and that the only risk present "was an investment risk similar to the risk assumed by a bank; it was not an insurance risk..."

The IRS has argued that the requisite elements of risk shifting and risk distribution are not present when premiums are paid by an insured to a wholly owned captive, which is insuring only the parent, because the parent must bear any economic loss through a decrease in the value of its wholly owned insurance affiliate. It has been argued that the result is no different than self insurance. This is the basic premise of the "economic family" theory.

Strained Logic

As group captives and risk retention groups have evolved and risks insured by

wholly owned captives have changed, the IRS' logic for attempting to deny deductibility of premiums has been strained. For example, the IRS ruled that payments to a captive that insured only its shareholders, none of whose risks exceeded 5 percent of the total risks insured and none of whom controlled the captive, were deductible insurance premiums. The situation indicated that a certain amount of economic unity between the insurer and the insured would not be fatal. However, the IRS continued to seek expansion of the "economic family" theory to find insufficient risk shifting and risk distribution. Thus, when the shareholders of the insured rather than the insured(s) have owned the captive, the IRS argued that any loss will be borne by the same economic family, and that there is insufficient risk shifting and risk distribution to constitute insurance.

The IRS' first defeat in this area came in 1985 in Crawford Fitting Co. v. United States. In that case the insureds and the captive were controlled by the same shareholders or members of their families. Without addressing the validity of the economic family argument, the court viewed the captive as a legitimate insurance vehicle.

In 1989, in Humana Inc. v. Commissioner, the U.S. Court of Appeals for the Sixth Circuit examined a similar Educuation. In Humana, the parent company and a number of its subsidiaries were insured by Health Care Indemnity, a wholly owned captive subsidiary of the parent. Using the same rationale applied in earlier cases to single parent captives insuring only the risks of the parent, the court held that premiums paid by the parent company on its own behalf to Health Care Indemnity were not deductible because there was no risk shifting.

However, regarding payments made by the parent and charged back to the other subsidiaries, the court held that the economic family theory could not be used to ignore the separate existence of corporate entities. The court found that risk was indeed shifted to a separate entity (Health Care Indemnity) and that sufficient risk distribution existed due to the number of insureds, and accordingly held payments by the subsidiaries to be deductible.

In September 1990, in Gulf Oil Corp. v. Commissioner, the U.S. Court of Appeals for the Third Circuit addressed the issue of deductibility of insurance premiums paid by Gulf Oil and its subsidiaries to INSCO, a wholly owned captive of Gulf Oil, which also insured risks of unrelated third parties. In these transactions, Gulf Oil and its wholly owned subsidiaries paid insurance premiums to unrelated insurers which then, on a prearranged basis, reinsured those risks with INSCO. To satisfy the unrelated insurers' concerns of INSCO's ability to pay reinsurance claims, Gulf Oil agreed to indemnify the insurers if INSCO could not meet its reinsured risk obligations.

The court held that premium payments by Gulf Oil and its subsidiaries were not deductible. While the subsidiaries initially appeared to be in the same position as the subsidiaries of Humana, the court in Gulf Oil distinguished Humana, noting that in Humana, the captive had been adequately capitalized initially, and that there were no agreements to contribute additional amounts to the captive to fund losses. However, in Gulf Oil, the captive was not initially adequately funded and the parent, Gulf Oil, agreed to bear all losses.

One interesting aspect of the Gulf Oil case was that, unlike many other captives, INSCO also wrote some insurance for unrelated third parties. The majority of the Tax Court held that, since net premiums charged to unrelated third parties amounted to only 2 percent of net premiums, the third party business was de minimis and did not indicate the existence of true risk transfer. This obviously inferred that some level of third party business would "cleanse" a transaction in which sufficient risk transfer might not otherwise exist. While the Tax Court would not enunciate a standard in this area without expert testimony, it stated in dictum that if at least 50 percent of the premiums were for unrelated business it could not believe sufficient risk transfer would not exist. The Third Circuit avoided this issue entirely, focusing instead on Gulf Oil's guarantee of INSCO's liabilities.

In response to the dictum contained in the Tax Court's opinion in Gulf Oil, the IRS issued Revenue Ruling 88-72, stating its position on the effect of unrelated business in a captive. The IRS took the position that the level of unrelated business written by a wholly owned insurer, no matter how large in relation to the business written for its affiliates, would not affect whether the transactions with related parties have sufficient risk transfer to constitute insurance.

Three Landmark Tax Court Cases

The IRS position on unrelated business set forth in Revenue Ruling 88-72 and the Tax Court's position on the same issue set forth in the dictum in Gulf Oil set the stage for three 1991 Tax Court cases: AMERCO v. Commissioner, The Harper Group v. Commissioner and Sears Roebuck and Co. v. Commissioner. Each of these cases involved payments to a wholly owned captive that also wrote significant unrelated business. The Tax Court, in opinions released Jan. 24, determined that in each case insurance premiums paid by the parent and the sister companies of the captive insurer were indeed deductible.

In reaching its conclusion, the Tax Court provided a basic framework for determining whether insurance exists for federal income tax purposes. The court held that an insurance transaction must involve "insurance risk" and elements of risk shifting and risk distribution. In addition, without a statutory definition, insurance is to be defined in its commonly accepted sense and federal income tax matters must be resolved with principles of federal income taxation in mind.

In AMERCO v. Commissioner, AMERCO, a holding company, was the parent of an affiliated group of corporations that generally comprised the U-Haul rental system. Republic Western Insurance Co. was a wholly owned insurance subsidiary of AMERCO but did not join in AMERCO's consolidated tax return. Republic Western wrote policies covering various risks of AMERCO and its subsidiaries. They included policies concerning risks of independent fleet owners (owners of U-Haul vehicles who leased those vehicles to the U-Haul system) and customers of U-Haul, and policies covering risks of unrelated third parties. After resolving that risks of independent fleet owners and U-Haul customers constituted unrelated business, the Tax Court found that gross premiums written for unrelated business comprised from 52 percent to 74 percent of the total gross premiums written by Republic Western.

The Tax Court reviewed the insurance industry perspectives of the transactions, economic theories of insurance and financial accounting principles, as presented by the party experts. The court then addressed whether there was existing precedent for deciding the issues. While noting that both the Tax Court and the Ninth Circuit (the circuit where the AMERCO case may be appealed) had addressed captive issues, the court found no controlling authority It felt the AMERCO case was factually distinguishable from earlier cases based on who Republic Western insured and how it was licensed and regulated.

The existence of unrelated business constituting more than 50 percent of Republic Western's gross premiums written for each of the years at issue distinguished the cases from Carnation, Clougherty Packing Co., Humana and Gulf. In each of the first three cases no unrelated insurance was written, and in Gulf Oil no unrelated insurance was written in the first year under audit and 2 percent was written in the second. Further, the court noted that by the end of 1985 Republic Western was licensed and regulated under standard state insurance laws in 45 states and the District of Columbia. The court contrasted this with Carnation and Gulf Oil, in which the insurers were Bermuda corporations and Clougherty Packing Co. and Humana, in which the insurers were licensed under state captive insurance laws.

On the question of determining the existence of insurance risk, the court differentiated between investment risk, which it concluded did not support a finding of insurance, and insurance risk. Insurance risk exists when the insured faces a potential loss and the insurer accepts a premium and agrees to perform some act if or when the loss occurs. According to the court, when the parties structure their transaction so that the only risk is similar to that assumed by a bank, insurance risk is not present. The court then held that each of the insureds faced some potential hazard that constituted insurance risks, and that no additional agreements counteracted the insurance policy risks.

Finding that insurance risk existed, the Tax Court then addressed the issue of whether risk shifting existed. For risk shifting to exist, the court found that the risk of loss must be shifted away from the taxpayer seeking to deduct insurance premiums. The existence of risk shifting was apparent in light of written insurance contracts, transferred premiums and paid losses. Republic Western was found to be a separate viable entity, financially capable of meeting its obligations as evidenced by the A.M. Best rating and continued permission to operate as a fully licensed insurer under state insurance laws. The court noted that insurance written for AMERCO itself totaled less than I percent of the transactions at issue, and that Republic Western wrote other business, including insurance for subsidiaries of AMERCO related to Republic Western as brother-sisters. Accordingly, the court found that the substance of the transaction matched their form and that risk shifting occurred between Republic Western and all of its insureds.

On the question of risk distribution, the court noted that an insurance contract is part of a larger collection of coverages combined to distribute the risk between insureds. Noting that the economic family approach to risk distribution had been uniformly rejected by the Tax Court, the court found that Republic Western's insurance business was diverse, involving substantial amounts of outside risks. In addition, more of the money in the pool supporting the risks came from outside unrelated insureds than from AMERCO and its subsidiaries. Accordingly, the court found sufficient risk distribution for insurance. In light of all the findings, the court held that insurance, for federal tax purposes, was present in all transactions.

Having found in favor of the taxpayer in AMERCO, the court then addressed similar issues in The Harper Group. Harper was also a holding company, and the common parent of domestic and foreign subsidiaries engaged in businesses related to international shipping. Harper organized Rampart Insurance Co. Ltd., an insurance subsidiary, to insure Harper's subsidiaries and customers using the services of Harper's subsidiaries. Premiums from such customers totaled from 29 percent to 33 percent of Rampart's gross premium revenue for the years in question.

Using the criteria first set forth in AMERCO, the Tax Court in the Harper Group again found that insurance existed for federal income tax purposes. However, the facts of Harper differed significantly from those in AMERCO. First, although the court found Rampart to be adequately capitalized and that Harper and its subsidiaries provided no guaranty, security or other financial arrangement regarding Rampart's obligations, Rampart was not subject to state insurance regulation. Rather, Rampart was a Hong Kong insurer and was subject to Hong Kong insurance regulation. In addition, Rampart clearly had been formed to provide insurance for Harper, its subsidiaries and their customers. Rampart also shared employees and facilities with WFNE, an operating subsidiary of Harper, and reimbursed WFNE directly. Finally, not only were unrelated premiums as low as 29 percent of gross premiums, but the unrelated and related risks were highly correlated because the same event, a loss at sea, would trigger both the related and unrelated insureds' risks.

While the court's conclusion in The Harper Group was grounded in the same analysis set forth in AMERCO, certain aspects of the Harper opinion should be noted. First, the court found that when 30 percent of premiums are received from unrelated insureds there is a sufficient pool of funds for risk distribution. It also noted that brother-sister companies might be considered unrelated, but concluded that it was not necessary to address this issue because 30 percent of the premiums were received from the customers of Harper's subsidiaries. Also, in addition to rejecting the economic family theory, the court rejected the IRS theory that a parent corporation can never obtain insurance from a wholly owned subsidiary.

Having determined in AMERCO and The Harper Group that risk distribution could be shown by sufficient unrelated business, the Tax Court then addressed what many believed was the most favorable fact pattern for the taxpayer in Sears, Roebuck and Co. v. Commissioner. In Sears, the parent company, Sears, Roebuck and Co., was the parent of a consolidated group of corporations which included Allstate.

Moreover, during the years in question, premiums for unrelated risks totaled 99.75 percent of all premiums, and unrelated risks were not correlated with related risks. However, while the overwhelming amount of unrelated business, coupled with Allstate's high standing in the insurance industry, made it difficult to argue that insurance did not exist, it was not necessary for the IRS to rely solely on an economic family or "balance sheet" argument that the arrangement had the same effect as self insurance. Rather, because Sears and Allstate filed a consolidated tax return, the arrangement had the same tax effect as allowing Sears a deduction for self-insurance reserves. Nevertheless, after examining the economic theory underlying insurance and the regulatory framework of the insurance industry, and applying the criteria set forth in AMERCO and The Harper Group, the Tax Court again found that insurance existed for federal tax purposes.

In reaching its conclusion in Sears, the Tax Court found that it is the premium pool, not the balance sheet of the parent or the net worth of the economic family, which bears the insurance risk. If the pool includes sufficient amounts of premium from unrelated parties, risk distribution will exist. While the percentage of unrelated business in Sears was overwhelming, the Tax Court again handed down a standard first enunciated in its Gulf Oil decision.

Under that standard, when the aggregate premiums paid by the captive's affiliated group is substantially insufficient to pay the aggregate anticipated losses of the entire group, the premiums paid by the affiliated group should be deductible. In addition, they should no longer be characterized as a self-insurance mechanism. While this standard does not provide a "brightline" test for an acceptable level of unrelated business, it raises the possibility that, in certain cases, a level of unrelated business below the 30 percent found in The Harper Group would support a finding of risk distribution.

Moreover, in discussing the presence of insurance risk in Sears, the Tax Court focused on the type of losses covered by the policies and the designated responsibility for payment of losses. The fact that Sear's total financial uncertainty was not altered by the Allstate arrangements, and that the ultimate profits or losses from Allstate's operations would inure to the benefit or detriment of Sears, did not significantly effect the analysis of whether or not insurance risk was present.

The Tax Court in Sears also accepted the validity of retrospectively rated premiums, even when the retrospective portion of the insured's premium was based solely on its own loss experience. Although the court's analysis of this issue was not detailed, it apparently relied on the fact that the retrospective premiums had an absolute maximum, and that in certain cases the insured losses could be borne by the premium pool rather than by the insured.

IRS Appeal Is Almost Certain

While AMERCO, The Harper Group and Sears represent clear victories for taxpayers in the area of wholly owned captives, they are virtually certain to be appealed by the IRS. However, by formally adopting a pooling theory of insurance, rather than the economic family or balance sheet approach, the Tax Court has provided a viable rationale for deducting premiums paid to a related insurer when sufficient unrelated business exists.

The IRS has also shown a willingness to challenge the deductibility of premiums when the affiliation between the insured and the insurer fits within the guidelines announced in Revenue Ruling 78-338. In the recently issued Coordinated Issues List on Captive Insurance Companies, the IRS insurance industry specialist advised examining the structure of insurance contracts to ensure that retrospective rate credits do no result in risks remaining with the insured. Additionally, the nature of the business itself is to be examined to determine if there is actual risk transfer. The Coordinated Issues List cited as an example Revenue Ruling 60-275 in which all the members of a captive insuring against flood loss are located in the same valley. Because a flood will result in losses to all insureds, the IRS' position is that no risk distribution exists. In light of the Tax Court's holding regarding correlated risks in The Harper Group, this rationale may also be questionable.

The IRS has also recently attacked the deductibility of insurance premiums, in a non-captive situation, based on the type of risk transferred. In Revenue Ruling 89-96, a fire insurance policy was purchased from an unrelated insurer after a fire, and the losses at the time the policy was purchased were known by both parties to exceed the policy limits. The IRS held that no insurance relationship existed, because the only transferred risks were the timing of claims payments and the investment rate to be earned by the insurer in the interim.

While Revenue Ruling 89-96 is itself very fact specific, the principles it enunciates, that investment and timing risk, in the view of the IRS, are not sufficient to constitute insurance, are troublesome as applied to other factual situations. For example, the IRS could argue that a historic self insurer that buys insurance for known losses for which there is a stable and predictable loss development pattern, and for whom incurred but not reported losses (IBNR) are not significant, has transferred only investment and timing risk. Conversely, it would be difficult for the IRS to make this argument where the loss development pattern is volatile or where IBNRs are significant. In this regard the Tax Court's analysis in AMERCO, The Harper Group and Sears, that the transfer of insurance risk is essential to an insurance transaction and investment risk alone is insufficient, is favorable to the IRS.

Another area of potential attack by the IRS on the deductibility of premiums arises from the recently issued proposed regulations under Section 461(h) of the Internal Revenue Code. Section 461(h) is generally designed to prevent accrual basis taxpayers from claiming deductions for expenses prior to "economic performance." Section 461(h)(2)(A)(i) provides that, except as provided in Treasury regulations, if the liability of the taxpayer arises out of the providing of services to the taxpayer by another person, economic performance occurs as such person provides services. The proposed regulations provide that, in the case of property/casualty insurance, economic performance occurs as premium payments are made by the insured.

However, in the case of "contingent payments"-such as when the taxpayer may be or become entitled to a refund-payment is not considered to be made until the contingency is removed. While the regulations are only in proposed form, and have been the subject of numerous negative comments, their effect, right now, would be to deny a deduction for all premiums for which there may be a retrospective credit or refund, until the actual amount of the credit or refund has been determined. Because eligibility for the retrospective credit often depends on one year's loss development, which often takes several years, this provision could significantly defer deductions.

Although the proposed regulations are not specifically aimed at captives, the use of retrospective rate credits or refunds is consistent with the goal of providing lowest cost insurance and is not uncommon in the captive area. While the tax years in question in Sears predate the adoption of Section 461(h), Section 461(h) itself does not necessarily require the restrictive view of payment apparently taken by the proposed regulations. Acceptance by the courts of the commercial legitimacy of retrospectively rated premiums will provide taxpayers with added support against IRS attacks in this area.

Captives Still Advantageous

Captive insurers will continue to provide a viable lower cost alternative to the traditional insurance market for sophisticated insurance purchasers. As the IRS continues to attack the deductibility of premiums paid to captives from several directions, greater care will be necessary in forming captives and in structuring insurance policies. While the recent decisions in AMERCO, Harper and Sears represent significant taxpayer victories, the appeal of one or all by the IRS is likely, and uncertainty will remain in this area for some time to come. However, by careful planning and attention to detail, insureds may find significant advantages in the use of captives.

For example, to enhance the argument that premiums paid by a related party are deductible, many captives may once again seek to underwrite unrelated business. When captives sought this type of business in the past, many suffered significant losses due to their unfamiliarity with the type of business involved. Accordingly, captives seeking unrelated business may be more likely to look for the type of unrelated business underwritten in Harper, where even though the insureds were unrelated, the insurer and its parent were very familiar with the risks involved and could institute a viable, ongoing risk management program. P. Bruce Wright is a partner and F Roy Sedore is an associate in the New York law firm LeBoeuf, Lamb, Leiby & MacRae.
COPYRIGHT 1991 Risk Management Society Publishing, Inc.
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1991 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Title Annotation:Internal Revenue Service; captive insurance companies
Author:Wright, P. Bruce; Sedore, F. Roy
Publication:Risk Management
Date:Sep 1, 1991
Previous Article:Taxation and the future of offshore insurers.
Next Article:The alternative market: shopping on a two-way street.

Related Articles
Vermont enacts legislation affecting captive insurers.
Captives maintain allure in a soft market.
IRS issues captive insurer guidelines.
Captive premiums deductible.
The truth behind captives and taxation.
Insurance in the captive setting.
The consolidated captive.
The IRS and captives.
The Humana case and other tax issues.
A unified approach to captive insurance tax policy.

Terms of use | Copyright © 2017 Farlex, Inc. | Feedback | For webmasters