Confirming the insurance industry's view of risk distribution in a reinsurance situation, the Internal Revenue Service recently issued Revenue Rule 2009-26, which verifies that in determining whether a policy with a reinsurance company is "insurance," and in testing whether there is "risk distribution," the relevant risks to consider are those of the ultimate insureds, not the reinsured company.
Risk distribution is required in every insurance arrangement. Under the IRS historic view, risk distribution means the insurance company must cover enough different entities so that none of those is paying its own losses in a significant way–that is, its losses are being paid from a pool of premium not dominated by that insured.
In Rev. Rul. 2005-40, the IRS made it clear it believes there can never be insurance if an insurer insures only one insured, no matter how many risks are being covered.
For instance, the IRS believes there is never insurance if there is only one insured, even if the policy is workers' compensation on one million employees spread throughout the nation. In the IRS view, this is the case even if the insured and insurer are unrelated.
The insurance industry and the risk managers they serve generally do not agree with this position, but no one has purposely challenged it.
In Rev. Rul. 2009-26, the IRS addressed the situation where a large number of insureds buy coverage from a fronting company, and the fronting company reinsures with a reinsurance company in one policy–which is the reinsurer's only policy.
One question is whether the reinsurance company will be treated as only insuring the risks of one entity (the risks of the fronting company), or whether it will be treated as insuring the risks of all the ultimate insureds.
If it is treated as only insuring one entity's risks, the question is whether Rev. Rul. 2005-40 would dictate that risk distribution is not present and thus, there is no insurance for federal income tax purposes.
In Rev. Rul 2009-26, the IRS stated that the reinsurance company is reinsuring the risks of the ultimate insureds (not the risks of the reinsuring entity) and thus, there is insurance.
This is the same view the IRS took in Rev. Rul. 77-316, the original revenue ruling in which the IRS challenged captive insurers. In 2001, the IRS abandoned its "economic family" argument against captives and it revoked Rev. Rul. 77-316, thereby potentially leaving open the question of risk distribution for reinsurance companies.
Rev. Rul. 2009-26 describes two situations and finds insurance (and risk distribution) to be present in both. In each situation, the insureds insure with Y (the fronting company), which reinsures with Z (the reinsurer). Z is adequately capitalized. It is licensed and regulated in all jurisdictions in which it does business. It acted at arm's-length in all transactions with Y.
In Situation One, Y issues 10,000 commercial multiperil policies over a 10-state area to unrelated insureds. Y reinsures 90 percent of those risks with Z, in exchange for 90 percent of the premiums in an indemnity policy, in which Y remains directly liable to the insureds.
The IRS ruled that this arrangement was insurance. It "looked through" the reinsurance contract and viewed Z, the reinsurer, as if it were bearing the risk of the ultimate insureds.
Because none of the 10,000 insureds were paying their own losses (paying almost all the premiums of the insurance company), there was risk distribution–and thus, insurance.
This conclusion is consistent with a case called Alinco, where a court found insurance existed where a reinsurance company reinsured $1 billion of premiums in a single policy from a commercial insurance company.
In Situation Two, X insured its risks with Y, which reinsured X's risks with Z. Z also reinsured risks of a number of other insurers that were also reinsuring the risks of single insureds (each unrelated to X).
The ruling states that if Z had directly insured the risks of X and the other ultimate insureds, it would have had risk distribution and insurance under the IRS's traditional tests. Z's reinsurance is treated the same way as if it had directly insured the risks of the ultimate insureds, and thus there is insurance.
The Revenue Ruling refers to unrelated insureds, but there seems to be no reason it shouldn't apply in a situation where the reinsurance company is related to one or more of the insureds, if the other IRS tests are met.
If so, the ruling would support insurance treatment for an arrangement where a sufficient number of unrelated insureds are insured by a fronting company, and a proportionate part of the pooled risks of the fronting company are reinsured with a reinsurance company owned by the owner of one of the insureds.
Similarly, Rev. Rul. 2009-26 should support insurance treatment for an arrangement whereby enough operating "brother-sister" subsidiaries insure their risks (for example, workers' comp) through a commercial carrier, which reinsures with a captive owned by the same parent that owns the operating subsidiaries.
Charles J. Lavelle is an attorney in the Louisville, Ky., office of Greenebaum Doll & McDonald PLLC. He was outside tax counsel in the Humana v. Commissioner and Ocean Drilling v. United States captive insurance tax cases, and is a member of the Captive Tax Advisory Committee to CICA and VCIA.
© Arc, All Rights Reserved. Request academic re-use from www.copyright.com. All other uses, submit a request to TMSalesOperations@arc-network.com. For more information visit Asset & Logo Licensing.