September 19, 2017
by Steven Miller, former IRS Acting Commissioner and alliantgroup National Director of Tax
On Aug. 21, the U.S. Tax Court released its long-awaited opinion in Avrahami v. Commissioner — the court’s first ever section 831(b) decision. Over the past few years the U.S. Internal Revenue Service has ramped up its scrutiny of “microcaptives,” as evidenced not only by a steadily increasing number of audits, but also by 1) the IRS’ placing such arrangements on its annual “dirty dozen” list for each of the past three years, 2) classifying these arrangements as transactions of interest under Notice 2016-66 and 3) making microcaptives one of the first IRS large business division campaigns.
However, even with this level of scrutiny, the captive insurance industry has been left in the dark with little to no guidance from courts, the U.S. Treasury Department or the IRS on what an acceptable arrangement looks like. With Avrahami being the first 831(b) case tried, many have been waiting and hoping that it would provide this sort of guidance. However, even with the release of this opinion, we think there are still many unanswered questions.
More cases are coming. Avrahami is one of three cases on Judge Holmes’ docket — Caylor v. Commissioner and Wilson v. Commissioner are the two other microcaptive cases that have also been fully tried, with opinions hopefully coming soon. As will be discussed later in this article, these cases may answer a couple of large outstanding issues, but they will likely still leave a number of unanswered questions, or at least many areas where the analysis is highly dependent upon facts of various shades.
Let’s start with issues where the Tax Court provides some assistance. First, the court gave some guidance as to what it will look at in determining whether a captive is providing insurance in the common sense of that term. When considering this issue, the court placed great emphasis on the operations and procedures of the captive — as in, did it function and operate in a customary manner one would expect from an insurance company?
The court focused on whether the company was organized, operated and regulated as an insurance company, whether it was adequately capitalized, whether policies were valid and binding, whether premiums were reasonable and at arm’s length and whether claims were paid. What we learned is that the court will focus on the claims process — not only whether claims were filed, but whether they were appropriately processed under existing procedures and paid out.
In Avrahami, the court stated that claims were dealt with on an “ad-hoc basis.” The Avrahamis did not file any claims until after the IRS audit began, when they then filed 6 claims. However, with these claims, the Avrahamis did not always provide documentation to substantiate their claim, as is customary with commercial insurance. Additionally, most of the claims were filed well outside the notice window required under the policy.
In addition to claims, the court focused on policies, premiums and actuarial work. It analyzed the policies and found them to be less than a model of clarity. The court was confused as to whether the policies were claims made or occurrence policies, as they had elements of both.
It then looked at the reasonableness of premiums and seemed skeptical of how insurance premiums for the Avrahamis rose from $150,000 to $1.1 and $1.3 million. Related to premium pricing, the court spent a considerable amount of effort dissecting and attacking the actuary’s methodology. It took issue with how the actuary seemingly cherry-picked inputs to generate inflated premiums.
What this tells us is that the Tax Court will thoroughly analyze each policy and then methodically review the actuary’s work in setting premiums. As a result, we can expect the IRS to continue to scrutinize these areas heavily. It will be important for captive owners to understand their polices, ensure the policies are clear and well-drafted and confirm they are working with a well-qualified actuary. We also learned that the court will look at the level of loans to related parties. 65 percent of assets in these loans raised the court’s eyebrow to the roof.
Second, we learned a bit about how the Tax Court will look at risk distribution. Here, the court at least walked us through the fact that earlier precedent matters in the microcaptive area. For example, if one looks at affiliates, how many are necessary before sufficient risk is distributed? The court refused to tell us, but it did say that the analysis is not just about the number of entities, but also about the number of diverse independent risk exposures (think trucks in different cities).
Further, it provided a window into how it will look at the risk pools which many captives utilize to ensure risk distribution. It looked at whether Pan American, the pool in this case, was providing insurance (terrorism in this case) and whether it operated like an insurance company.
Finally, we learned some good news. We learned that the Tax Court may be sympathetic to penalty relief in the 831(b) context. The court granted relief in this case due to reasonable reliance on a professional advisor, and due to the fact that the Avrahamis acted in good faith in an area where there is a dearth of legal guidance. In a second piece of good news for the Avrahamis, the court accepted (with “jaundiced eye”) that the loans from the captive were actually loans. This, the court did so after reviewing the usual factors to distinguish loans from distributions.
So what did we not learn from the case? A whole lot. There are a great number of issues that remain undecided after Avrahami — we will discuss the major items.
First, while the Tax Court dealt at length with risk distribution and insurance in the common sense of that term, it did not deal with risk shifting or insurable risk, which constitute remaining two of the recognized four factors to determine insurance. And even in risk distribution, questions remain.
As mentioned, the court addressed risk distribution, and discussed whether it was satisfied through insuring either affiliated entities or unrelated entities through the pool. With the affiliated entities, the court did not rule on the number of entities required to satisfy risk distribution. The government’s expert claimed that 35 entities were necessary, while the plaintiff’s expert relied on Rev. Rul. 2002-90, which states that 12 related entities would create risk distribution.
Because the captive in Avrahami only insured three to four entities, the court decided that it need not opine on the number actually needed, since three to four was lower than 12 or 35. We don’t know if the court will ultimately respect Rev. Rul. 2002-90 and hold that insuring 12 affiliated entities will be enough to satisfy risk distribution requirements.
In Caylor, the captive insures 12 related entities and is not involved in a risk pool. Therefore, the court may have to ultimately decide whether 12 entities is sufficient. The major caveat here is that the court will also likely analyze, as it did in Avrahami and other non-831(b) captive insurance cases, whether there is a sufficient number of diverse independent exposure units.
Since risk distribution wasn’t satisfied through affiliated entities, the court had to look at whether risk was distributed to unrelated parties through the Pan American program. As stated, the court did not like Pan American’s product or its operation. The court looked at Pan American and held that it was not a bona fide insurance company, because of the seemingly circular flow of funds, unreasonable premiums and lack of arm’s-length dealings.
The court pointed to facts such as “utterly unreasonable” premiums, the unlikelihood of claims and the thin capitalization of Pan American. A key takeaway is that the court did not condemn all risk pools, but simply said that based on the facts, Pan American was not a bona fide insurance company.
What remains unclear, is how the court would treat risk pools with better facts — i.e. risk pools that operate more as insurance companies and provide a better product than Pan American. We do not expect to get any guidance from the court in Caylor or Wilson, as Caylor does not involve a risk pool and the risk pool used in Wilson appears to be the same one as in Avrahami.
While omitting insurance risks from its analysis, the court touched on them in its facts section, and it is worth mentioning what it said. A common argument made by the IRS in captive insurance audits is that coverages provided by captives are not covering insurance risks, but instead are covering financial or business risks, which would not fall under the definition of insurance. In the facts section of the opinion, the court stated that the taxpayer and IRS agreed that administrative actions and employee fidelity coverage are insurance risks.
The court’s language around those policies seemed friendly. But the language turned a bit chilly with respect to tax indemnity coverage, with the court referring to it as the “most peculiar policy” covered by the captive. As the definition of insurance risk is such a major area of contention between taxpayers and the IRS, we are hoping the court grants some clarity on the issue in either Caylor or Wilson.
Lastly, we do not believe the issue of required capitalization is clear after Avrahami. The court seemingly grudgingly held that Feedback, the captive, was adequately capitalized because it met the minimum capitalization required by its jurisdiction. However, the court also made the argument that Pan American was thinly capitalized because of the circular nature of the premiums (virtually none remained in Pan American) and because it would not be able to cover the nearly $308 million in potential losses under the terrorism coverage. This was because its management did not seem competent to the court to chase payment from all the captives reinsuring those losses. So while the court recognized that Pan American met the jurisdiction’s capital requirements, it found no reason to give them credence where, as here, it found that it was unlikely to be able to pay claims.
So at the end of the day, it is a bit disappointing that Avrahami does not provide more guidance, but it is not surprising, given the fact-intensive nature of the opinion. Caylor and Wilson may help answer some of those questions, but we don’t think that even combined, that this trio of cases will provide sufficient guidance to taxpayers on a go-forward basis to allow good captives to operate without fear of IRS intervention. That is a pity. There is a continuing need for IRS guidance in this area.
Steven T. Miller served as former IRS Acting Commissioner in 2012, but prior to that he served for several years as the Deputy Commissioner for Services and Enforcement, leading all IRS enforcement and service activity. Steven also served as the Commissioner of the Large and Mid-Size Business Division, overseeing IRS audits of large taxpayers and the IRS programs relating to offshore tax compliance and international tax law enforcement. As the Commissioner of the Tax Exempt and Government Entities Division, he supervised the IRS oversight of governments, tax exempt entities and retirement programs.
In a career devoted to government service, Steven has spent the last 25 years with the IRS, serving the agency in a number of diverse and increasingly important roles.
Would you like additional information or do you have questions?