There are two age-old adages in the legal profession – one learned in law school and the other through experience. The first is “bad facts make bad law”. The second is that “taxpayers should avoid the Tax Court at all costs”. Both adages are on display in the Tax Court case, Jones, et al v. Commissioner, decided March 25, 2025.

Another Captive Insurance Company Challenged by The IRS

This case involved a so-called “microcaptive” insurance arrangement. We use the term “so-called” because “microcaptive” is not a real word. When typed in a Word document, spell check highlights it as a misspelling. However, this has not stopped the IRS and the Tax Court from using the term.

The IRS coined the term for the first time in Notice 2016-66. It does not appear in the Internal Revenue Code. It is a pejorative term completely made-up by the IRS to influence the Tax Court. And the Tax Court has taken the bait.

The Jones case is the ninth, “microcaptive” captive case that the IRS has won in the Tax Court. (Contrast that to the one and only “microcaptive” case to be tried in a United States District Court before a jury, rather than the Tax Court, which RMC won in April, 2024.) The Judge in the Jones case, Joseph W. Nega, was like the mythical lemming following his predecessor judges off the cliff. His opinion, while well-written, contains nothing to indicate that he did anything other than parrot the opinions of the cases that came before the Jones case. That is unfortunate, because it reinforces the perception that taxpayers do not get a fair hearing in the Tax Court.

What is a Microcaptive

Before we discuss the case, let’s unpack the term “microcaptive”. The Court, in a footnote in the Jones opinion, defines the term as follows:

A captive insurance company is a corporation whose stock is owned by a small number of shareholders, and which handles all or a part of the insurance needs of its shareholders or their affiliated entities . . . A “microcaptive” insurer is a captive insurance company that elects the alternative tax structure provided for under section 831(b).  

This definition is okay as far as it goes. However, the Tax Court, prompted by the IRS, conflates two concepts into one. The first is the concept of a captive insurance company, which the Court accurately describes. The second is the concept of a small insurance company, which is defined by Congress in section 831(b), and covers any non-life insurance company with annual premiums below a certain threshold.

A captive insurance company that makes an election under section 831(b) is a small insurance company. The fact that it is a captive insurance company is immaterial under the express language of section 831(b). As stated above, the term “microcaptive” does not appear in the statute. Importantly, neither does the term “captive”. Yet, the IRS has been treating small captive insurance companies differently than other small insurance companies. And the Tax Court has been happy to go along. However, there is no justification for this disparate treatment in the Internal Revenue Code or any other indication of Congressional intent. The jury in the RMC case understood this. Apparently, Tax Court judges do not.

The Facts of the Jones Case

In many respects, the facts of the Jones case are typical of any captive insurance company. Among the Petitioners were the owners of an s-corporation, which started life as a distributor of products manufactured by others and evolved into manufacturing the products itself. The business was covered by a number of commercial insurance policies; the premiums for which increased year-over-year, even though there were few, if any, insurance claims. When the Petitioners became concerned about the potential for a particular type of claim, its broker contacted a couple of commercial insurance companies, which declined to provide coverage. As a result, their advisors suggested a captive.

The Petitioners were referred to a captive manager who followed a typical path. The captive manager performed a feasibility study to determine whether the business was a good candidate for a captive. An outside actuary was engaged to provide an actuarial study, which identified the risks faced by the business and recommended limits of liability and premiums. The captive was formed, and policies were issued.

The Petitioners’ captive participated in a “reinsurance pool” operated by its captive manager. The Court’s description of the “reinsurance pool” is somewhat confusing. It would appear that the insurance policies were issued by the Petitioners’ captive and a portion of the risk ceded to a reinsurance company operated by the captive manager. The Petitioners’ captive paid reinsurance premiums to the reinsurance company equal to approximately 51% of the premiums paid by its insured. The reinsurance company then ceded pooled risks back to the Petitioners’ captive and paid reinsurance premiums to the Petitioners’ captive in an amount approximately equal to the premiums that had been paid to it by the Petitioners’ captive.

It is not clear from the Court’s opinion how the “reinsurance pool” actually worked. Even though the Petitioners’ captive issued the policies and ceded a portion of the risk to the reinsurance company, it does not appear that the reinsurance arrangement between the Petitioners’ captive and the reinsurance company was a quota-share reinsurance arrangement, through which the reinsurance company shared the risk from the first dollar. Instead, it appears that the Petitioners’ captive insured a primary layer, and the reinsurance company insured an excess layer, even though no excess policy was issued. In addition, the limits of liability of each layer are not included the opinion, so it is unclear how much of the underlying risk was assumed by the Petitioners’ captive and how much was ceded to the reinsurance company. Further, it is not clear from the opinion what percentage of the total risk assumed by the Petitioners’ captive came from the reinsurance pool as opposed to its related business.

The problem with this arrangement is that there were a number of atypical and bad facts:

1. The actuary never spoke directly to the owners of the business before performing his actuarial study.
2. The policies issued by the captive did not include the risk, which was the alleged reason for forming a captive.
3. The captive was terminated after one year.
4. The investments of the Petitioners’ captive and the reinsurance company were illiquid.
5. The captive made a “loan” to its shareholders for a real estate investment.

While each fact was bad and played a role in the Court’s decision, the worst fact was the “loan” from the captive to its shareholders. The term “loan” is in parenthesis because the Court found that the distribution was not a loan but a constructive dividend. The reasons are that the note provided a below-market interest rate; the shareholders did not pay the note as required by its terms; the due date was allegedly extended without proper documentation; and it was not paid until the Petitioners had been audited, their tax deductions denied, and this lawsuit had been filed.

The Arrangement Was Not Insurance

The Court’s opinion follows the template established by the other “microcaptive” cases that came before it. In fact, it could have been written by any other Tax Court judge in any of the other cases. After stating its findings of fact, the Court discusses the law and then applies the law to the facts. However, it is not clear from the opinion whether the Court is applying the law to the facts of this particular case or to its preconceived notion, established through years of IRS challenges, of the facts of a typical “microcaptive” arrangement. Suffice it to say that the Court found that the payments to the Petitioners’ captive were not insurance premiums and not deductible. It reached that conclusion by holding that the Petitioners’ captive did not distribute risk and was not insurance in the commonly-accepted sense.

There is no value in discussing the Court’s “legal reasoning”. It is the same as every other “microcaptive” case. What would be more valuable is highlighting a few of the inaccuracies, inconsistencies, and hypocrisies in the Court’s opinion that illustrate the Tax Court’s irrational bias against small, captive insurance companies.

The Court correctly states the following:

Neither the Code nor the Treasury regulations define insurance; we are, therefore, guided chiefly by caselaw in determining whether an arrangement constitutes insurance for federal income tax purposes.

This is standard language in all of these cases. However, the courts just skip over the implication of this sentence. The definition of insurance has significant tax consequences for insurance companies and insureds alike. Congress is the entity tasked with making laws. And Congress has made the tax law by enacting the Internal Revenue Code – the current version in 1986 with subsequent amendments. If any entity should define the term “insurance”, it is Congress and not the courts or the IRS. This should be troubling to anybody who respects our system of government and the separation of powers.

The Court also correctly states that, through caselaw, the courts have developed a four-part test to determine whether an arrangement constitutes insurance. It also correctly states that one element of that test is “risk distribution”. However, the Court does not stop at risk distribution. It adds language, which we do not believe appears in any other case, and changes the meaning of this element. The Court defines this prong of the test as follows:

            The insurer distributes the risk of loss among its policyholders; (emphasis added)

As far as we know, the IRS has never required that risk distribution be “among the insurer’s policyholders”. And this requirement essentially eliminates a small, captive insurance company from the definition of insurance. As the Court stated in its footnote cited above, a captive insurance company is formed by a small number of shareholders to insure the risks of its shareholders or their affiliated businesses. By definition, a captive insurance company cannot have a significant number of policyholders. Otherwise, it would not be a captive and could not get a captive license from any jurisdiction. This change in the caselaw-developed definition of insurance shows that Judge Nega decided this case before he heard any evidence or considered the legal arguments of the parties.

The Court further defines risk distribution in a way that excludes small insurance companies, whether a captive or not, by focusing on a theory known as the “law of large numbers”. The Court says the following:

In essence, by entering into insurance contracts with a large group of similarly situated insureds and “writing policies for enough independent risks, an insurer can predict 'the frequency and amount of loss within this larger group . . . [and] set the premium at a level that will cover the losses, cover the insurer's overhead and expenses, and earn a profit.

The Court also says:

Generally, risk distribution occurs when the insurer pools a sufficiently large collection of independent risks. See Rent-A-Center, 142 T.C. at 24. Risks are independent when “the likelihood of a loss under one policy is independent of the likelihood of a loss under a separate policy”. In other words, independent risks “are generally unaffected [*35] by the same event or circumstance.”

There are two problems with these statements. The first is that the case cited by the Court to support his decision involved a large, captive insurance company established by a multi-national corporation with a number of subsidiaries and thousands of independent risks, not a small, captive insurance company. The second is that a small insurance company – again, whether a captive or not – cannot insure a large number of independent risks because its annual premiums are limited by section 831(b). For the tax year in question in the Jones case, a small insurance company was limited to premiums of $1.2m. How many independent risks can an insurance company assume for that level of premium? The answer is not enough to satisfy either the IRS or the Tax Court. This is another example of the Tax Court’s irrational bias against small, captive insurance companies that has no basis in fact or law.

There is one other problem with the Court’s statements. Judge Nega overlooks the fact that a “reinsurance pool” used by many small, captive insurance companies to achieve risk distribution, by structure, contains many independent risks unlikely to be affected by the same event or circumstances. By including multiple businesses in many different industries, a “reinsurance pool” will almost never have multiple insureds impacted by a single event or circumstance. This is exactly how the “law of large numbers” works as described by Judge Nega. So, even though a small, captive insurance company cannot take advantage of the “law of large numbers” on its own, the “reinsurance pool” in which it participates certainly does. Judge Nega’s failure to recognize this fact affirms his inherent bias against “microcaptives”.

The Court correctly states that an insurance company generally charges actuarially-determined premiums. The Court then recognizes that there is no statutory or regulatory definition of the term “actuarially-determined” when it says the following:

Because neither the Code nor the regulations define “actuarially determined” premiums in the context of captive insurance, we have relied on caselaw for guidance.

This leads us to the same problem we have with the definition of insurance. Congress has not provided a statutory definition of “actuarially-determined”. Neither has the IRS. In fact, the IRS has been asked on numerous occasions to provide a standard for “actuarially-determined” premiums. And it has refused. Why should the IRS provide such guidance to taxpayers when it knows that the Tax Court will do its bidding? This is highlighted by the fact that the premiums in the Jones case were, in fact, determined by an actuary using methodologies generally accepted in the profession. Yet, the Court found the premiums deficient, even though I can guarantee that the Judge is not an actuary.

Moral of the Case

As stated in the first section of this article, this case reaffirms two concepts that we already knew. The first is that bad facts make bad law. This case had significant bad facts from the failure of the actuary to meet with the insured business to the surrender of the captive’s insurance license after only one year to the captive providing financing to its shareholders for a real estate investment unrelated to the captive.

The second is the irrational bias of the Tax Court against small, captive insurance companies. The Tax Court has never seen a small, captive insurance company that it likes, and it may never. So far, the IRS is 9 for 9 in Tax Court cases. Part of the reason is that it can choose the cases with the worst facts. But a bigger part of the reason is that, when it comes to captive insurance companies, the Tax Court is willing to distort the law to give the IRS a win. That is why a taxpayer should do whatever it takes to bring its case in a United States District Court in front of a jury, if it possibly can.

For more information on this case or a fuller discussion, please call us at 239-298-8210.