Update On Recent Captive Tax Court Case

In an opinion by Judge Jones, who previously worked in IRS Chief Counsel’s Office before becoming a tax court judge, the United States Tax Court yesterday issued a ruling that was adverse to one of our (CIC Services, LLC’s) clients. Though we are disappointed with the opinion and believe it contains multiple errors that will hopefully be addressed on appeal, we thank the court for its hard work on this very technical case. Judges have hard jobs, especially when the facts and law are as complex as they were in this case.

Even so, we believe that the opinion contains erroneous findings of fact, avoids of much evidence offered at trial, contains conclusions that don’t follow from the facts (non sequitur) and in some cases also misapplies or selectively applies the law. We chronicle many of these errors below and expect them to be resolved on appeal.

However, even should this ruling stand, it’s essential to understand why and how the judge ruled the way she did since it’s ultimately this rationale that will guide our and the industry’s future conduct. The good news is that the opinion hinged on facts and circumstances that were largely unique to the case at hand. To anyone with very similar facts, the ruling, if it stands, is therefore illuminating. To everyone else though, it may not be very relevant at all.

Though we expect this opinion to be reversed on appeal, we at CIC Services, LLC are taking it very seriously. Despite that the fact-specific nature of the judge’s determinations makes extrapolating broader guidance difficult, we are scouring the ruling to discern ways that we or our clients can improve captive insurance arrangements so that those who are interested in claiming them may benefit from any tax benefits available under the Internal Revenue Code. We are pleased to say that the case at hand involved captive insurance transactions from the 2013 to 2016 time frame, and some of the shortfalls identified by the court (even if they are sustained on appeal) were already addressed years ago, making them mostly irrelevant to any transactions in more recent years.

The Court’s Big Picture Errors

Though the details will be reviewed more further below, it’s helpful to begin by noting that the court made five broad categories of errors in reaching its conclusion.

First, in several places throughout the opinion, some of which will be detailed below, the court misconstrued the facts or completely ignored contrary facts without explaining why. In some places she said that there was “no evidence” for things that were actually supported by tremendous amounts of evidence that was either misconstrued or overlooked. We believe this to be reversible error.

Second, the court erred by ignoring essential evidence benefiting the taxpayers from tax years other than those in dispute while at the same time regularly considering evidence offered by the IRS related to the years outside of those in dispute. Justification for this disparity in evidentiary treatment was generally not provided or was suspect. We believe this to be reversible error as well.

Third, the court also consistently compared “excess and surplus” types of insurance policies issued by the captives to “commercial” insurance policies without accounting for differences in the lines of insurance (excess and surplus versus commercial) or the scopes of coverage. This is, with respect, an inapt “apples-to-oranges” comparison that should carry no evidentiary weight. We believe this also to be reversible error.

Fourth, the court regularly criticized Dr. Patel and the captives for failing to do many things that were in fact done on the captives’s behalf by their paid, professional agents or advisors. The court provided no explanation for why the efforts of its paid professionals should not, as a matter of ordinary agency law, be attributed to Dr. Patel and/or the captives. Had the court considered the activities of these paid advisors, it would have been clear that many things the court claimed weren’t done actually were done. We believe this to be another reversible error.

And finally and most surprisingly, the court dutifully mentioned in several places the relevant legal factors that must be weighed in order for it to reach a determination on an given legal issue, and then it either completely ignored many of those factors or instead considered other irrelevant ones. We believe this too to be reversible error.

The Court’s Specific Errors

Risk Distribution via Pooling

 The court notes that, for an insurance company to qualify as offering “insurance” under the Internal Revenue Code, it must (among other things) achieve sufficient risk distribution—that is, it must spread its risk among a sufficient number of independent risk exposures.  However it should be emphasized that risk distribution isn’t merely a requirement of *tax* law, it’s a necessary component of most any valid insurance arrangement under state law and insurance/actuarial theory.  Thus, achieving risk distribution isn’t something  done just for *tax* avoidance reasons but also, and even primarily, for ordinary *insurance* reasons—that is, for business reasons.   Anyone seeking to operate a legitimate insurance company will strive to achieve meaningful risk distribution.  

 Because captive insurance companies are, almost by definition, much smaller than commercial carriers, they sometimes don’t have sufficient independent risk exposures to achieve risk distribution on their directly written policies.  And even when meaningful risk distribution can be achieved on their own, some captives will nonetheless seek to more broadly diversify their risk exposures via other means. 

Consequently, most captive insurance companies in existence achieve risk distribution or more broadly diversify their risk exposures by participating in “risk pools”, arrangements where unrelated and independent captive insurance companies come together to insure a portion of each other’s risks.  Risk pooling ensures that valid claims filed against Company A will be at least in part paid by independent and unrelated Companies B, C, D, etc., and vice versa.

Dr. Patel’s captives participated in an independent risk distribution pool in which dozens of independent and unrelated captive insurance companies agreed, via the risk pool mechanism, to collectively cover 51% of each of Dr. Patel’s captives’ valid claims, from first dollar of claim to last dollar of claim, in exchange for the captive ceding 51% of the premiums it received from its insureds to the pool.   Because the pool received exactly 51% of the risk under the policies that Dr. Patel’s captives issued in exchange for exactly 51% of the premium they received for those policies, and because the pool participated in insured losses from first dollar to last dollar of any claim (that is, no arbitrary attachments points), no separate, formal actuarial calculations are needed under the relevant actuarial standards in order to allocate risk and premiums between Dr. Patel’s captives and the risk pool.  With any risk pool structured in this manner, 51% of the risk equals (actuarily speaking) 51% of the premium *by definition*.   The amount of premium needed for the amount of risk assumed by the pool is deemed *self evident* under relevant actuarial authorities.  Multiple witnesses, including actuarial experts, testified to this fact at trial or via their expert opinions.

After assuming 51% of the risk of Dr. Patel’s captives in exchange for 51% of the premiums those captives received, the risk pool then retroceded some of its risk back to Dr. Patel’s captives under a “quota share” arrangement. So while the risk pool reinsured 51% of Dr. Patel’s direct insurance risk in exchange for 51% of the premiums it received on the directly written policies, Dr. Patel’s captive also *reinsured* a “quota share” portion of the risk of the risk pool (excluding his own claims) in exchange for a receiving a “quota share” portion of the premiums received by the pool from all those other insureds.

The end result of this process is that Dr. Patel’s captive and every other captive in the pool had shared liability for each other’s claims, thus distributing their risk among dozens of unrelated insureds with tens or hundreds of thousands of unrelated, independent risk exposures, thereby distributing/diversifying its risks. 

 And, because all other captives in the pool participated in it on essentially the same terms as Dr. Patel’s captives—ensuring risk from first dollar of claim to last dollar of claim also, and using the same actuarial firm for all—the actuarily correct premium paid to Dr. Patel’s captive for assuming its quota share portion of the pool’s risk was *by definition*, approximately equal to the premium that Dr. Patel’s captive paid to the pool in exchange for it assuming 51% of the risk under the policies that his captive issued to his insured entities.   That this is true is purely a function of *math*, accepted actuarial theory and self-evident actuarial standards.  Contrary to the court’s implication, it was not a result of some sort of nefarious or arbitrary “targeting” of premium amounts to ensure a circular flow of funds.  Similar quota share risk pooling arrangements are common both in the captive and “regular” insurance industries and originated with the latter.

Per the court, whether or not a risk pool provides sufficient risk distribution for federal income tax purposes depends on whether or not the pool provides the “functions of an insurance company”.   And, per the court’s own opinion and prior precedent, that depends on the extent to which:

• The risk pool was created for legitimate nontax reasons
• There was a “circular flow of funds”
• The risk pool faced actual and insurable risk
• The policies were “arm’s length contracts”.  
• The entity charged actuarially-determined premiums
• Comparable coverage was more expensive or even available
• The risk pool was subject to to regulatory control and met minimum statutory requirements 
• The risk pool was adequately capitalized
• The risk pool paid claims from a separately maintained account.

Oddly, while emphasizing (rightly) that *all* of the above factors are indeed relevant and essential to determining whether the risk pool provided the functions of an insurance company, the court only analyzes five of them and provides no explanation for omitting the rest.  Importantly, the ones it failed to consider seem to clearly favor the taxpayer.  Failure to consider all legally required factors is, or should be, reversible error.      Let’s review the nine factors noted above one by one.   

Legitimate Nontax Reasons

Much evidence was offered at trial and via the relevant briefs as to why the risk pool was formed.  It was not formed merely to achieve risk distribution for *tax* purposes, though that was important, but also for *insurance* purposes.  As previously explained, an essential element of insurance in general is risk distribution, and any true insurance arrangement will distribute the risks of the insurance company among a large number of independent risk exposures.

Remarkably, the court mentions this factor as relevant to its decision but then never analyzes it nor reaches any conclusion as to whether or not it was satisfied.   Failure to analyze this factor is reversible error, and evidence at trial showed that there were in fact nontax reasons for the risk pool.

Factor 1 above therefore favors the taxpayer, not the IRS.

Circular Flow of Funds

The court next considered whether there was a “circular flow of funds.”  If so, that factor favors the IRS.  If not, then it favors the taxpayer.  

In contending that there was a circular flow, the court makes note of the fact that Dr. Patel’s captives “received payments from [the risk pool] roughly equal to the premiums [the risk pool] was entitled to receive from [the captives].”  While this is true, it’s unclear whether the court applied the appropriate test. 

 The provable fact is that Dr. Patel’s captives did not receive back any of the money they ceded to the risk pool.  Rather, *all* premiums retroceded by the risk pool to Dr. Patel’s captives provably came from *other* participants in the risk pool.  In other words, Dr. Patel’s captives ceded premiums to the pool and then received premiums of approximately the same amount as a retrocession from the pool in exchange for his captives insuring a portion of the pool’s claims that result from losses sustained by *other* insureds in the pool.  

The fact that the the retroceded premiums approximated the ceded premiums in amount was purely a function of math and sound actuarial principles.  It was not, despite any implication to the contrary, a product of manipulation designed to achieve some predetermined outcome or to ensure that Dr. Patel got his ceded premiums back.  

Given that Dr. Patel’s captives never received back any of the premiums they ceded to the risk pool and that all amounts retroceded to his captives came instead from other insureds in the pool, there was no “circular flow of funds” between the risk pool and his captives.  Contending otherwise merely because the amounts ceded and retroceded were similar in total is non sequitur and, we believe, reversible error.  This factor should therefore have been rightfully resolved in favor of the taxpayer.    

Actual and Insurable Risk

 The next relevant factor announced by the court, and one that it again inexplicably failed to analyze, is whether the risk pool provided the functions of an insurance company—that is, whether or not the risk pool “faced actual and insurable risk.”   It unequivocally *did*. For *every* year in dispute, the risk pool paid millions in claims under policies that it had reinsured, and Dr. Patel’s captives bore their quota share portion of those claims, which amounted to $138,205 dollars over just the few years in dispute and even much more than that in subsequent years (especially the COVID years). Further, the actuaries, state regulators and CPAs all agreed that the reinsurance offered by the risk pool covered actual and insurable risks.

In short, the evidence at trial was overwhelming and undisputed that the risk pool not only *faced* actual and insurable risks but also actually incurred dozens of insured losses related to those risks for *every year* of its existence.   Though the court inexplicably failed to analyzing this factor after emphasizing its importance, it too cuts in favor of the taxpayer.  

Arm’s Length Contracts

 Factor 4 considers whether the reinsurance arrangements between Dr. Patel’s captives and the risk pool were “arm’s length contracts”.  In concluding that it was not, the court ignored the fact that these contracts contained all the standard provisions found in arm’s length insurance contracts and focused instead on the fact that there was supposedly “no evidence of any arm’s-length negotiations in determining the premiums paid to” the risk pool.  Alas, this conclusion is just wrong.

First, whether or not premiums were reasonable is an entirely different factor that gets separately analyzed (see above and below).  Deciding whether or not the contracts were “arm’s length” must therefore involve much more than just analyzing the reasonableness of the premiums.

Second, with very rare exception, nobody “negotiates” their insurance premiums with their carrier.  Actuaries working for the carriers determine the premium and you either pay it and get the insurance or you don’t pay it and…don’t get the insurance.   The idea that *premiums* need to be “negotiated” for an insurance contract to be “arm’s-length” is, with respect, silly.  That cannot be the law.

Third, attorney James Coomes, who ran the risk pool and determined who was in and out and on what conditions, was always familiar with the actuarial pricing work done on all the directly written policies.   Hence, as discussed below, he knew that 51% of those actuarily-determined premiums was in fact the actuarially appropriate premium the pool should demand for assuming 51% of the first-dollar-to-last dollar risk under the policies.   “Negotiation” of those premiums on a case by case basis would not only have made no sense, it would been improper.

Fourth, the court emphasized elsewhere in the opinion that premiums should be “actuarily determined.”   And in this case they *were* (see below for details).  Which is precisely why they were not “negotiated”.  How can one say that premiums should be actuarily *determined* and also that they should be *negotiated*?  That is non sequitur.   

The court tries to bolster its conclusion that the reinsurance arrangement was not “arm’s length” by noting that there was “no actuarial determination of the reasonableness of the 51% of premiums ceded to [the risk pool]”.  But this is clear error.  As explained above, the 51% premium number was *by definition* the proper amount to cede in exchange for the risk pool assuming 51% of any insured loss from first dollar to last dollar.   Actuarily speaking, there could be no other result.  No calculations are needed.  It’s self-evidently true under relevant actuarial standards in the same way that 1 + 1 = 2 is self-evidently true (regardless of whether you “show your math” or not). 

 But if that wasn’t enough, multiple actuaries testified to this fact at trial and in their expert reports, noting that their conclusion is supported by relevant actuarial standards that were likewise introduced into evidence.   Additionally, regulators at the Tennessee Depart of Insurance scrutinized the risk pool arrangement in detail, including pricing, and much evidence of this was offered at trial.   However, the court never acknowledges any of these facts.

 Next, the court criticizes the fact that premiums for the captive insurance policies were markedly higher than premiums for “commercial policies”, and it contends that this also implies that the policies were not arm’s-length contracts.  But the court completely ignores the fact that apples *should* be priced differently than oranges even though both are still fruits.  

 In other words, the pricing of “excess and surplus” captive insurance policies is often radically different from the “commercial” policies the court used for comparison, and actuarily speaking *should* be different, because their *terms* and *scope of coverage* and *rate regulation* are so radically different.   Stated plainly, captive insurance generally offers *much* broader coverage with *far* fewer exclusions, thereby requiring (actuarily speaking) much higher premiums.   In fact, obtaining broader (less restrictive) coverage is a, if not *the*, major reason why captive insurance exists!  This truth was introduced at trial and remains uncontroverted.

Though it defies reason to claim that apples must be priced the same as oranges in order for apples to, like oranges, be considered fruits, that’s effectively what the court concludes.  The court never engages in *any* analysis or comparison of the terms, conditions and exclusions of the relevant captive and commercial policies, something an actuary *must* do in order to accurately price the policies in compliance with actuarial standards. Merely taking note that the excess and surplus captive insurance policies were priced differently than *in*comparable commercial policies says nothing at all about the legitimacy of the pricing of either.

Next, the court criticizes Dr. Patel for failing to do sufficient analysis to determine whether the reinsurance purchased via the risk pool was a reasonable value as compared to commercial insurance.   And it likewise criticized Dr. Patel for doing  no due diligence on each of the dozens of risk pool members before allowing his captive to assume its quota share portion of the pool’s risk each year. And it contended that both of these facts are somehow evidence that the insurance contracts in question are not arm’s length.

This “reasoning” too is fallacious.  First, though neither Dr. Patel nor any W-2 employee of his captives analyzed the “risks, industries [and] ability to fulfill quota share claims” of the pool participants, as the court suggests should be done before allowing his captives to assume their quota share portion of the pool’s risks, other paid professionals engaged by Dr. Patel’s captives *did*.   For example our firm, CIC Services, LLC, the professional, paid “manager” of the captives in question, knew the industry of every insured, knew the relevant risks of each insured business and the pool as a whole, and also performed the bookkeeping for every captive in the pool.  Consequently, the captive manager *did* know the “risks, industries and ability to fulfill quote share claims” of every pool participant, and we had a duty to inform Dr. Patel’s captives and other pool participants if we, in our professional opinion, had concerns about these things.  Evidence for all of this was offered at trial and, remarkably, completely ignored by the judge since these functions were not performed directly by Dr. Patel or by a W-2 employee of the captive.  

 But it wasn’t just us.  The captive’s attorney, who also managed the risk pool, was also familiar with the “risks, industries [and] ability to fulfill quota share claims” of each participant in the risk pool. In fact, evidence was offered at trial showing how Mr. Coomes had previously expelled at least one participant from the risk pool when it proved to be a bad risk for other pool participants, including Dr. Patel’s captives.   And we, as his paid professional captives managers, and also James Coomes, even scrutinized Dr. Patel’s commercial polices and compared them to his captive policies to help ensure that there was little to no overlap or duplicate coverages, something we and Mr. Coomes also did for all others in the pool.  In short, for the court to say that these things were not done is simply not true.  

 The court knew these paid professionals existed.  The court knew that they provided these services on behalf of the captive.  Importantly, the court never concludes that the work done by these professional agents was deficient or unacceptable.  Instead the court just ignores these professionals and their work completely.  They are never mentioned in this part of her opinion.  And only by ignoring them can she conclude (wrongly) that this diligence wasn’t performed on the captive’s behalf.

To deny Dr. Patel and his captives *any* credit for work that was clearly and responsibly done by the captives’s paid professionals, including the organization specifically hired to function as the captives’s “manager”, and to do so without *any* explanation, is reversible error.  

 Once this reversible error is corrected, this factor cuts clearly in favor of the taxpayer.    

Actuarily Determined Premiums

The court next turned to whether or not the premiums between the captives and the risk pool were actuarily-determined.  Despite the fact that the premiums were indisputably determined by an actuary at the time, that two different expert actuaries (one of whom helped write the relevant actuarial standards) reviewed that actuarial work in detail and opined at trial that it was consistent with relevant actuarial standards, that these same two actuaries independently priced the policies in question and arrived at essentially the same results as the captive’s actuary,  that an independent actuary for the State of Tennessee had *contemporaneously* (and *not* in preparation for trial) reviewed the captives’s actuary’s work and deemed it satisfactory for insurance regulatory purposes, the court incredibly concludes that the premiums were not “actuarily determined” at all.

In reaching this conclusion the court ignores all the evidence from the actual actuaries and instead focuses on two minor details, regrettably getting even those wrong.

First,  the court says that the risk pool failed to account for “the different risks of pool members, the types of businesses of pool members, or the geographic location of pool members” when setting its price for reinsurance.  This is not accurate and, even were it accurate, wouldn’t necessarily violate actuarial standards or mean that the premiums were not “actuarily determined”.

As was shown at trial, each participant in the risk pool used the *same* actuarial firm to price its policies each year.   The attorney who managed the risk pool, deciding who gets in and who gets kicked out, had access to the actuaries’ work pricing *all* those policies, as did we as the captive manager.  Further, we and the attorney absolutely knew, as was shown a trial, “the different risks of pool members [and their respective insureds!], the types of businesses of pool members [and their respective insureds], [and] the geographic location of pool members [and their respective insureds].”  And we and the attorney knew that *all* of those factors had, to the extent they were actuarily relevant, been previously considered when the actuaries developed the pricing for the direct written policies.

In short, the court is just wrong that nobody considered the “different risks of pool members, the types of businesses of pool members and the geographic location of pool members” when developing the pricing.  They were in fact considered by all relevant paid professionals of the captive, including the actuary, the attorney and the captive manager.

The court continues its criticism of premium pricing by stating that  there was no evidence that the 51% of direct written premiums ceded to the risk pool was actuarily determined. Again, this is not true.  The 51% figure was true *by definition*—if the risk pool assumes 51% of the first dollar to last dollar risk under the policies directly issued by the captive, and it did, then *by definition* the proper premium to pay it for doing so is 51% of the premiums that were actuarily calculated for those risks.  And this was confirmed at trial by multiple actuaries and by reference to relevant actuarial standards.  The fact that the captives’s actuary didn’t engage in the entirely futile and unnecessary step of writing a formal report explaining that 1 + 1 = 2 and showing his math does *not* mean that the premiums weren’t determined by an actuary applying actuarial standards.   They were, and that was confirmed at trial multiple times by multiple witnesses.

Though the court is right that the 51% risk/premium split between the captive and the risk pool was chosen in part to comply with tax law, it does not follow therefrom that the premiums weren’t actuarily determined.  So long as 51% of the premium was the actuarily appropriate amount to pay in exchange for the pool assuming 51% of the risk under the policies, and it was, then the premiums were, in fact, “actuarily determined”.  And it seems utterly improper and ill-advised to hold the captives’s and risk pool’s good faith attempts to *comply* with the tax laws (by distributing sufficient risk via the pool) against them as evidence of some improper tax motivation.

In short, the evidence is overwhelming that the premiums were actuarily determined, and the court only avoids that conclusion by misinterpreting the facts and engaging in non-sequitur, both reversible error.  This factor should have rightly been resolved in favor of the taxpayer.  

The Existence of “Comparable Coverage”

The next relevant factor, and *another* one that the court entirely failed to even consider despite previously indicating it to be legally significant to its determination, is whether comparable reinsurance coverage was more expensive or even available.

For quota share risk pools to work efficiently and serve their insurance purpose, it’s usually important that all insureds in the pool use a common actuarial firm, that the risk ceded to the pool by each insured captive actuarily equates to the same percentage of premiums for each insured (in this case, 51%), and that the pool be “policed” by a common person who is looking out for the interest of all pool participants, etc. For these reasons and others, there generally was no “comparable” reinsurance coverage available to Dr. Patel’s captives from other reinsurers or risk pools. While this may not have been verified personally by Dr. Patel each year, it was known by paid professional agents acting on behalf of his captives, and is self-evident to anyone working in this industry.    

 Maybe even more importantly, much evidence was introduced at trial showing that Dr. Patel didn’t trust commercial carriers, and for good reason—namely, he had very bad experiences with them denying coverage for past claims.  And consequently, all else being equal, or sometimes even unequal, he generally preferred purchasing much more reliable there-when-you-need-it captive insurance, such as that offered by the risk pool, to the much less reliable “deny, delay, defend” commercial reinsurance.  Evidence at trial showed that this proved to be a wise decision on his part.  The risk pool covered 51% of his $2.1 million in insured COVID, key person and storm claims, as contractually agreed, while his commercial carriers didn’t pay a dime.  

 Consequently, this factor (whether commercial insurance was more expensive or even available), though inexplicably not analyzed by the court even after it specifically declared its relevance, also cuts in the taxpayer’s favor.  The court’s failure to consider this factor is a reversible error.

Regulatory Control and Statutory Compliance 

The next relevant factor was whether the risk pool was subject to regulatory control and met statutory requirements.  It unquestionably was, and the Tennessee Department of Insurance even specifically concluded that it was a reinsurer qualified to conduct business in Tennessee (after doing *extensive* due diligence on the risk pool that was confirmed at trial by the former Tennessee regulator who performed it).

The court conceded these facts and concluded specifically that the risk pool “was organized and regulated as a reinsurance company under state and international law.” But then, inexplicably, the court concludes that “these insurance-like traits cannot overcome its other failings” without even explaining what “traits” or “other failings” it’s even referencing!

In other words, the relevant factor at hand, per the court’s own determination, was whether the risk pool “was subject to regulatory control and met statutory requirements.”  After conceding that it *was* under regulatory control and without noting a single deficiency in meeting *any* statutory requirement, the court then inexplicably pivots to imply that this factor isn’t satisfied because unnamed “insurance like traits” can’t overcome other unnamed “failings”.  Once again, this is non sequitur and reversible error.  Given that the court found explicitly that the risk pool “was organized and regulated and a reinsurance company under state and international law” and noted no deficiencies in its regulation or in satisfying any statutory requirements, this factor too resolves in favor of the taxpayer.  

Adequate Capitalization 

The next factor, which the court also mysteriously fails to analyze, is whether or not the risk pool was adequately capitalized.  The undisputed testimony at trial was that the risk pool met all regulatory requirements regarding its capitalization.  Given that the Tax Court has consistently deferred to regulatory determinations on this matter, this factor also resolves in the taxpayer’s favor. The court’s failure to consider it and so conclude is reversible error.

Claims Paid from a Separately Maintained Account

The final factor is whether or not the risk pool paid claims from a separately maintained account.  It indisputably did. Every single one and hundreds in total.

Not only did the risk pool pay all claims from a separately maintained account, but the evidence at trial showed that this was in fact an explicit regulatory requirement of the Tennessee Department of Insurance.

This final factor therefore resolves in the taxpayer’s favor, though the court once again mysteriously failed to analyze it even after explicitly mentioning it.   Its failure to consider this factor and award it to the taxpayer is reversible error.

Conclusion

Of the nine separate factors that the court said are relevant to determine whether or not the risk pool served the functions of an insurance company, all favor the taxpayers when each is fully and properly analyzed in the light of the record and rid of non sequiturs.   That being the case, the risk pool provided the “functions of an insurance company”, and the captives in question achieved risk distribution via the risk pool.   Consequently, the rest of the court’s risk distribution analysis is irrelevant and won’t be covered in this write-up.

Insurance in the Commonly Accepted Sense

For a captive insurance company to be respected as an insurance company for federal income tax purposes, the company must, in addition to achieving risk distribution, provide “insurance in the commonly accepted sense”.  In deciding this question, the court indicated that the following questions must be resolved and balanced:

• whether the company was organized, operated, and regulated as an insurance company;
• whether it was adequately capitalized;
• whether the policies were valid and binding;
• whether premiums were reasonable and the result of arm’s-length transactions;
• whether claims were paid;
• whether policies covered typical insurance risks; and
• whether there was a legitimate business reason for acquiring insurance from the captive.

Organized, Regulated and Operated as an Insurance Company

With regard to the first question above, the court concludes that “there is no dispute” that the captives were organized and regulated as insurance companies.  The question therefore is simply whether they were *operated* as one.

In concluding it was not, the court first criticizes (without offering reason or any legal basis)  the captives for not operating through “employees of their own that performed services”.  This is an odd criticism because prior Tax Court cases also involving captives with no employees, and who likewise acted only through hired professionals, found those arrangements to be acceptable and legitimate.   Quite simply, until this opinion, there was no known requirement that captives operate via paid W-2 employees rather than via professional independent contractors. If this indeed is a requirement to “operate” as an insurance company, and it should not be, then even many very large captives will fail this test.

In analyzing whether is captives were operating as insurance companies, the court next cast doubts on Dr. Patel’s motivation for forming them.  The court said that “there is no credible evidence that Dr. Patel’s conversations about forming a captive centered around preventing a future disaster.”

However, to arrive at this conclusion the court had to either entirely ignore or completely discount the sworn testimony of multiple fact witnesses—Dr. Patel himself; members of Dr. Patel’s staff who were involved in forming the captives; myself (Sean King); Bryan Ridgeway (also with CIC Services), Dr. Patel’s financial advisor, Christopher Fay; the captive attorney, James Coomes; and the Tennessee regulator.  Not a single witness at trial who helped Dr. Patel form his captives testified that Dr. Patel was primarily tax motivated.  All indicated that Dr. Patel was clearly motivated to form the captive by his prior and devastating business disaster, a disaster that was not covered by his commercial insurance policies.

Incredibly, without even *mentioning* the sworn testimony of all of these fact witnesses, the court instead focuses on minor, cherry-picked details to reach a conclusion as to Dr. Patel’s real motives in forming the insurance companies.

For example the court mentioned that Dr. Patel had read a book on asset protection before forming his captives, as if that fact harmed rather than bolstered Dr. Patel’s argument that he was motivated by his prior catastrophic business failure and the multi-million dollar losses he sustained as a result of it.   Oddly, the court does not here mention that Dr. Patel also bought and read books on captive insurance in the years following that disaster, though evidence of that was offered at trial.

The court also criticized Dr. Patel for maintaining his commercial coverages rather than replacing them with captive insurance.  But, as was explained at trial, most of the commercial coverages that he retained, such as medical malpractice insurance, were not generally available or suitable (for regulatory or practical reasons) to be insured through captive insurance arrangements in general, and through this one in particular.  In other words, replacing many of his commercial coverages with captive insurance simply wasn’t an option, and the paid professionals acting on behalf of his captives unquestionably knew that.

The court then criticized Dr. Patel for a supposed lack of a formal “feasibility studies” to determine “whether a captive was necessary and, if so, what policies were required.”  In fact such a study *was* prepared.  And in any event, the relevant legal question isn’t whether or not the captives were “necessary” but merely whether they were advisable.  I and many others personally testified at the trial as to why Dr. Patel believed each captive was advisable (a prior catastrophic and uninsured loss that he endured), and our testimony was consistent with Dr. Patel’s own on the subject.  Not a single witness offered contrary testimony.

And finally, the court next criticizes Dr. Patel for not doing “any due diligence with respect to the reinsurance or quota share agreements” with the risk pool.  As previously noted above, while it’s true that Dr. Patel may not have done much due diligence on these matters, he definitely did some.  But more importantly, the paid professionals managing his captive did that diligence extensively, and overwhelming evidence of that was offered at trial. The court simply ignores this under the implied theory that unless this work was done by Dr. Patel or W-2 employees of the captives, it simply doesn’t count.

The court then uses these selectively chosen and often untrue facts to conclude that the captives were not “operated” as insurance companies despite being licensed as such, regulated as such, accepting premiums as such, issuing policies as such, paying more than $2.5 million in direct and indirect claims as such (including claims of unrelated third parties every year of its existence), the auditing CPA deeming them to be such, etc.   Judges are certainly free to weight the evidence however they deem appropriate, but contending that there was “no evidence” that the captives were operated as insurance companies when, in fact, they were and there was an extraordinary amount of evidence for it, must be reversible error.

When these errors are reversed, this factor is properly resolved in favor of the taxpayer.  

Capitalization

The next factor the court considered was capitalization.  After noting that the captives in question met the minimum capitalization requirements of their domiciles, the court concluded that they were adequately capitalized and resolved this factor in favor of the taxpayer. 

Valid and Binding Policies

Next, the court analyzed whether the policies issued by the captives were “valid and binding”.  The court noted that it has previously ruled that policies are “valid and binding” when “each insurance policy identified the insured, contained and effective period for the policy, specified what was covered by the policy, stated the premium amount and was signed by an authorized representative of the company”.  The undisputed evidence showed that all of these factors were present in this case, and the court never concludes otherwise in its opinion.

The court also noted that it has previously found policies to be “valid and binding” when “the insured filed claims for covered losses and the captive insurance company paid them.”   The undisputed evidence showed that the captives in question paid claims under policies for each year in dispute and in all subsequent years, more than $2.5 million in total, and the court never concludes otherwise in its opinion. The court acknowledges that claims were in fact paid.

Finally the court noted that it also considers other things in deciding whether policies are valid and binding, such as whether there are conflicting policy terms.  The court did not make note of *any* such conflicting policy terms in its opinion.

And yet, after laying out these standards for “valid and binding” policies, standards that the captives both passed with flying colors, the court inexplicably proceeds to rely on *other* previously unnamed factors to conclude that these policies were not “valid and binding”.  Specifically the court noted that the captive policies had (1) some “atypical provisions” (such as a somewhat restrictive “claims made” requirement) that it deemed “unfavorable” to the insureds, (2) “excess” policy language and high premiums, and (3) didn’t permit refunds of premium upon cancellation of the policy.

Relying *only* on the above three listed factors, *none* of which even if true are by the court’s own precedent, or as a matter of law or logic, relevant to whether or not the policies were actually “valid and binding” contracts, the court concludes that evidence of the binding and enforceable nature of these policies is “mixed” and therefore that this factor is “neutral” (favors neither the taxpayer nor the government). However, if the court had instead applied the prior precedent mentioned in its own opinion, and if it had considered only factors actually relevant to the enforceability (the “valid and binding” nature) of the policies—this factor cuts very clearly in favor of the Patels.  Deeming it merely “neutral” required the court to introduce new and irrelevant standards and resort again to non sequitur.  This is reversible error.

Reasonableness of Premiums

Next the court turned to the “reasonableness of premiums” factor.

In analyzing this factor, the court accused Dr. Patel of “targeting” the monetary limit of Section 831(b).  But fixing a premium budget that is purposefully intended to stay *under* the Section 831(b) limits, so as to ensure the best possible tax treatment for the captives, is *not* the same as “targeting” the monetary limit of Section 831(b).   Dr. Patel could have paid *more* premiums to his captives and could have insured even more risks, and yet he limited his premiums to remain below the premium Section 831(b), insuring these other risks elsewhere or it other ways or not at all.  As multiple courts have noted, intentionally structuring a transaction to take advantage of *statutorily* available tax benefits does *not* by itself mean that the transaction is somehow improper (see the Summa Holdings case as but one example).

The court then criticizes the work of the actuary, accusing the actuary of using “ill defined” factors to “increase total premiums as close as possible to” the Section 831(b) premium limit.   But to reach that conclusion, the court entirely ignores the fact that two separate very prominent actuaries who served as expert witnesses for Dr. Patel independently did their own premium calculations and arrived at premium amounts *very* close to those of Dr. Patel’s actuary. Further, at least one of those actuaries reviewed the work of the captives’ actuary in detail and concluded that the quality of that work was consistent with all relevant actuarial standards.  This would suggest that, while the factors used by the actuary may have seemed “ill defined” to the layperson judge, they were not “ill defined” at all to the professional actuaries who reviewed the work.  That is what *should* matter.

The court is dismissed the work of these experts actuaries who confirmed that the captives’s  pricing was actuarily sound, accurate and principled simply on the grounds that the expert reports were “prepared for the purpose of litigation” (as *all* expert reports are, including those of the government’s experts, which the court cites favorably at times).  Even were that rationale alone sufficient to entirely disregard their expert analysis (in which case, why was the testimony of the government’s experts not disregarded on the same grounds?), the court inexplicably fails to mention the fact that the State of Tennessee’s Department of Insurance contemporaneously (and *not* in preparation for litigation) reviewed the captive actuarial pricing as part of its regulatory due diligence, and it *too* found the pricing to be reasonable.

In other words, the court dismisses or ignores both the testimony of the unquestionably qualified actuary experts, one of whose work is cited very favorably in prior tax court cases, and *also* the work of actuaries contemporaneously hired by state regulators, and instead relies on dubiously-supported accusations that the premiums were “targeted” at a certain level in order to maximize tax benefits, and, as we will see, dubious comparisons to incomparable commercial policies, to conclude that the premiums paid to the captive were not “reasonable.”

This is reversible error both because aiming for a statutorily defined benefit doesn’t by itself render transaction improper and because reaching this conclusion required the court to ignore incredible amounts of relevant contrary evidence.

The court next compares the premiums that Dr. Patel’s businesses paid to his captives to premiums paid for third party commercial policies, concluding that the captive premiums were, by comparison, too high.   But the court did not mention or consider, as you actuarily must, the fact that the captive policies in question are much broader in scope than typical commercial policies.    It also ignored all the evidence introduced showing that it’s not reasonable or actuarily sound to compare premiums for “surplus and excess” lines of coverage, such as those issued by the captives, to “commercial” coverages.   Once again, the judge seems to fall for the fallacy that if orange (commercial insurance) are fruits and apples (captive excess and surplus insurance) aren’t priced the same as oranges, then apples are not fruit.   This is yet another example of non sequitur.

In short, it was both reasonable and legally permissible for Dr. Patel to seek to limit premiums paid to his captive to less than the Section 831(b) limit.  The work of the captives’s actuary was independently and *contemporaneously*  reviewed by state regulators and deemed *reasonable*.   And the pricing of inapt commercial policies is, as a matter of both logic and actuarial science, irrelevant to the pricing of incomparable captive surplus and excess types of policies.  Reversing these errors would award this factor to the taxpayer.

Did the Captive Pay *Any* Claims?

Next, the court considers the factor of “whether [the captives] paid *any* claims” (Emphasis added).  The court notes that no *direct* claims were filed by Dr. Patel’s captive during the years under audit.  However, as was shown at trial, this was simply because no directly insured losses actually occurred during those years, something that all actuaries, including the government’s own, agreed was not unusual for the types of policies issued.

However and very importantly, directly insured losses *did* occur in subsequent years, claims *were* filed for those directly insured losses, and more than $2.1 million in direct claims *were* paid under those same policies in subsequent years. Those claims related to COVID, loss of a key employee and storm losses.

Evidence for these directly insured claims, including undisputed evidence that they were in fact paid, and the detailed process of reviewing and approving the claims prior to payment, was offered at trial. And yet the court’s opinion makes *no* mention of *any* of these direct claims when considering the question of “whether [the captives] paid *any* claims”.

The court then notes that the captives did indeed pay a total of $138,205 in risk pool related claims during the years in dispute.  That alone requires that the factor of “whether the captives pay any claims” be resolved in the taxpayer’s favor.   And if we expand the inquiry, as we should, to other years, such as the COVID years, we see that its risk pool claims increased markedly over time and even much more was paid.

In short, despite the court’s reluctance to acknowledge the significance of this truth, risk pool claims *were* paid in *every* year of the captives’s existence at issue in the case and in *all* subsequent years.   And those claims increased over time.  But more, the captives also paid all directly insured claims (totaling over $2.1 million) and the court never contends or implies otherwise.  It just completely and inexplicably ignores this undisputed fact.

By logic and the court’s own announced standard (whether or not *any* claims were paid), this factor indisputably cuts heavily in favor of the Patels. And yet the court concludes, after ignoring the more than $2.5 million in claims actually paid, that “these relatively small payments…might weigh slightly in favor of the Patels.”   Astounding.

What makes the court’s reluctance to award this factor to the taxpayer even more astounding is that the court was perfectly willing in other contexts to consider evidence outside of the tax years in dispute when doing so supported the government’s case, but it fails to do so here (such as by failing to even acknowledged the COVID claims) and gives absolutely no explanation or justification as to why.  Rightfully analyzed this factor weighs *heavily* in flavor of the Patels.

Typical Insurance Risks

The next factor the court announced as relevant is whether the policies insured against typical insurance risks.  Alas, inexplicably, the court *again* fails to consider this factor at all, even after announcing it as relevant.  This is reversible error.

Evidence at trial showed that many of the coverages available via the captives were in fact available from third party insurers, albeit with far less favorable terms.   Those that weren’t are nonetheless commonly offered by captive insurance companies.  Multiple actuaries at trial testified that the risks insured by the captive were in fact insurable risks.  A regulator from the Tennessee Department of Insurance confirmed this also at trial.

In short, this factor too resolves in favor of the taxpayer. The court’s failure to analyze it is reversible error.  

Whether There Was a Legitimate Business Reason for the Captive

This is yet another factor that the court announced as relevant and then inexplicably otherwise fails to consider.

Evidence that the captives were formed for legitimate business reasons and served legitimate business purposes was overwhelming.   A few years prior to forming the captives, Dr. Patel had suffered an extraordinary business disaster that cost him and many friends and family millions.   Importantly, his commercial policies did not cover any of these losses.  Every witness who testified at trial as to their understanding of Dr. Patel’s purpose in forming the captives, including me (Sean King), agreed that this disaster, and insuring against a wide variety of risks that could similarly devastate him in the future, was top of his mind when he was forming the captives.

And the wisdom of that decision was confirmed by the large COVID loss and other losses he later suffered, losses that were insured by his captive and ultimately 51% paid by unrelated insurers who participated in the risk distribution pool.  In short, there can be no doubt that the captives were formed for and served *a* legitimate business purpose.   Neither the law nor logic require that tax considerations be entirely absent from the decision making process.

This factor too resolves in favor of the taxpayer.  Why did the court ignore it?  This too is reversible error.

Conclusion

Based on the flawed analysis as outlined above, the court concludes that “the Patels have not proven that the payments they seek to deduct as insurance expenses were for insurance in the commonly accepted sense.”  With respect to the court, this is just wrong. 

When properly rid of non sequitur, cherry-picking of the applicable legal factors and selective presentation of evidence, it’s clear that the captives in question offered insurance “in the commonly accepted sense”.  Every single relevant factor noted by the court resolves in favor of the taxpayer.  

Grand Conclusion

When the court’s errors are corrected, it’s clear that Dr. Patel’s captives achieved risk distribution via their participation in the risk distribution pool.  It’s also clear that they provided “insurance in the commonly accepted sense.”  Consequently, they should be treated as valid and legitimate insurance companies under federal income tax law.

We hope that Dr. Patel appeals this decision.  We are confident that an appellate court will recognize the errors in the trial court opinion and rule in favor of Dr. Patel and his captives.

Sean King, JD, CPA | Principal