Menu

Avrahami v. Commissioner: What’s Next for Captive Insurance?

miller steveAvrahami v. Commissioner: What’s Next for Captive Insurance? On August 21, the Tax Court issued its long awaited opinion in Avrahami v. Commissioner. Many in the micro-captive industry had been waiting for Judge Holmes to issue this opinion, in part, due to the IRS’ own anticipation. What many had hoped for was bright line guidance from the Court on an area of law highly scrutinized by the IRS with little prior IRS, Treasury or judicial direction. What we received fell short of our aspirations, containing some guidance but raising many questions. What we will try to do in this article is walk you through what the court found and what it portends for the future. And we need to remember that this is part one of a trilogy of cases that have been tried before Judge Holmes and await his irascible pen.

The Avrahamis ran a successful jewelry and real estate business. They created a captive insurance company named Feedback, which made the section 831(b) election. While they were unlucky enough to be early in line for IRS attention, the issues raised by the IRS are the same as have been raised now hundreds of times since—that is, was Feedback an insurance company and were the policies it issued insurance. In addition, there was the issue of whether Pan American, which was the reinsurance “pool” in this case, was a successful mechanism to distribute insurable risk amongst unrelated parties.

Judge Holmes did not spare words and the opinion ran 105 pages. Of those, 90 were not so good for the taxpayer and the captive manager. The issue the Court considered in this case is whether amounts paid to Feedback (the captive) and Pan American Reinsurance Company were deductible as insurance payments for Federal Tax Law purposes. The Government argued that Feedback’s policies included uninsurable risks, it failed to distribute risk because it had an insufficient pool of insureds, risk was not shifted and that the arrangements did not embody common notions of insurance because Feedback and Pan American did not operate like insurance companies and their premiums were not determined at arm’s length. The Court agreed with some of these arguments, but did so in a very fact specific manner. Ultimately, the Court did not make findings as to risk shifting or what constitutes insurable risk. It decided it did not need to in order to find for the government.

First, the Court opined on whether risk distribution was satisfied. The Avrahamis had argued they met this requirement in two ways. First, they argued that they had enough related party risk. Second, they contended that the risk pool for unrelated risk (Pan American) was sufficient. The Court dealt with each in turn. As to insuring related entities, while the government’s expert stated 35 related entities are needed to meet risk distribution and the plaintiff’s expert, citing Rev. Rul. 2002-90, stated only 12 is necessary (and argued maybe as few as seven), the Court did not opine on the total number of related affiliates it believed would actually be needed. Rather, it held that because the captive only insured three to four related entities, it did not meet the minimum stated by either expert. Therefore, we are not sure of whether the Tax Court will respect the 12 insured rule created by Rev. Rul. 2002-90, or adopt its own rule when it comes to risk distribution. Judge Holmes’ next case, Caylor, may provide us with a better sense in this regard as it relies solely on affiliate risk. However, the Court here was very clear that the number of affiliates insured was not a magic number. Instead, the Court indicated that it wanted to see diverse and numerous “independent risk exposures.” What are these? Think cars, employees, properties, etc. Judge Holmes cited to the recent Rent-A-Center case in this regard, noting he recognized that fewer such exposure units must exist in the micro-captive setting but he needed more than what the Avrahamis had.

Second, the Court decided whether risk distribution was met through the Pan American program. Under the Pan American program, an insured would pay a premium for terrorism coverage in a pool arrangement, and the captive would reinsure the pool for all the risks assumed by the pool for that insured. The Court held that risk distribution was not met due to a series of adverse factual findings, including: the seemingly circular flow of funds that left Pan American with no funds and all obligations, “utterly unreasonable” premiums, and a lack of arm’s length dealings. It found for these reasons, Pan American was not a bona fide insurance company, so the policies it was issuing were not insurance, and therefore the captive’s reinsurance of those policies did not distribute risk. But let’s think about what the Court did not say. The Court’s analysis here is telling in that it found that it had to look at the actual facts of the pool – the Court did not prohibit or even speak negatively of risk pools, but instead focused on the totality of the facts.

What we find lacking here – and not in any way to the Court’s fault – is clarity on what the Court would do with a more diverse and well operated risk pool. As in, just what is necessary between Pan American and the size and breadth of the risks in Rent-A-Center?

After finding that there was no risk distribution, the Court began to disassemble Feedback, the Avrahami’s captive. The Court found Feedback was not an insurance company and the arrangement was not insurance in the commonly accepted sense. Why? Again, let us count the ways. While the Court found no quarrel with St. Kitts as a domicile for organizing a captive (though it did make a few snarky comments), the Court found the operation of Feedback sadly lacking. The key reasons include: no claims until initiation of the IRS audit, and when they showed up they were dealt with on an ad hoc basis, claimed late and without substantiation; operations and formalities were lacking; policy terms were unclear; investments were in illiquid assets such as long term unsecured loans to related parties/insureds; and failure to receive regulatory approval for loans. One line the Court did draw for us: do not have 65% of your assets in related party loans that are unsecured.

The Court was somewhat confusing on the level of capitalization required. The Court chose to follow precedent and interestingly sided with the Avrahamis in holding that Feedback was adequately capitalized because it met the minimum requirements of St. Kitts. However, it held that Pan American—even though it met the minimum capitalization requirements of its jurisdiction—was not adequately capitalized, because it would not be able to cover the nearly $308 million in potential losses and that it would likely not be able to collect payment from the captives.

Regarding premiums, in holding that the arrangement did not constitute insurance in the commonly accepted sense, the Court spent many pages on the work of the actuary for the captive and Pan American. It found that his explanations “were often incomprehensible” and that “the premiums were utterly unreasonable.” The Court spent time in its opinion analyzing the details of the actuary’s methodology, and found error in the way he selected and utilized variables to artificially inflate premiums. We think that the methodology employed by actuaries will continue to be an issue raised by the IRS and Courts. We have, in fact, attended an interview where an actuary spent hours with an IRS team walking them though the actuarial science. Not an easy task and not an easy interview. As a result, it is important to make sure that the actuary pricing premiums for a captive is well qualified and relies on actuarially sound principles.

In the last pages of the opinion, the Court did award the Avrahamis a valuable win in granting relief on most penalties. The Court found that the Avrahamis had relied on a tax advisor and had relied in good faith. What this relief tells us is that the Court will continue to respect reasonable reliance on a professional advisor, even in a micro-captive context. What is less clear, is whether the Court will also find that reliance to be in good faith. The Court was sympathetic to the Avrahamis, finding good faith due to the lack of IRS or judicial guidance on micro-captives. However, now with this opinion available, it may be slightly more difficult for newly formed captives to assert good faith, especially in instances where facts are in line with those of Avrahami. Arguably, this good faith should apply for all years before the opinion but we shall see.

The two major issues the Court did not address are risk shifting and what constitutes an insurance risk. While Judge Holmes did not opine on these issues, he and the IRS did offer a little insight on insurance risks. In the facts section of his opinion, it was made clear that the IRS agreed that coverage for Administrative Actions and Employee Fidelity were insurance risks. On the flip side, the Court stated Tax Indemnity is the “most peculiar policy of all.” Further guidance will be needed on other types of coverages common to micro-captive arrangements.

And further guidance on insurance risks, risks shifting and for other fact specific issues may be coming soon. Judge Holmes has two fully tried captive cases on his docket – Caylor v. Commissioner and Wilson v. Commissioner. Caylor is similar in some ways to Avrahami, specifically in that it involves terrorism insurance through Pan American. Caylor, on the other hand, does not involve a risk pool, but instead is an arrangement where a captive provides insurance to 12 related construction/real estate entities. Caylor may offer valuable insight as to whether Rev. Rul. 2002-90 will be respected by the Court.

What Avrahami provides is a set of facts where the Court will not find insurance present. What it does not provide is guidance on what it would find acceptable. We have but one point of data in many areas. While Avrahami only answers a few questions due to the fact specific nature of the opinion, we sincerely hope that further guidance comes from Caylor and Wilson. Further, guidance from the IRS would go a long way. In an ideal world, the IRS would take the general findings that come from these cases to create both an audit resolution program for those already under audit and also a safe-harbor describing permissible behavior concerning major issues. Time will tell.


Steven T. Miller is alliantgroup’s National Director of Tax and former IRS Acting Commissioner. He spent 25 years with the IRS serving the agency in a number of diverse and increasingly important roles.