I still don’t carry dental insurance.  I have not found a plan where the numbers make sense.  The benefits are so limited that by the time I would need them I will have paid more in premiums than the benefits are worth.  The game is not worth the candle.  So I take the money I would otherwise spend on premiums and save it up.  In effect I self-insure.

Businesses face the same choice in managing their various risks.  For most risks they can pay a third party, an insurance company.  But for some risks businesses cannot find a third-party insurance company willing to cover the risks at an acceptable price.  The game is not worth the candle.  It could be because the risks are too specific or esoteric, like covering the risk of bird flu for your chicken farm.  Lesson From The Tax Court: Not Every Decision Comes With An Opinion, TaxProf Blog (November 15, 2021)(discussing unpublished order in Puglisi et al. v. Commissioner). Or the risks might be so massive and the pool so shallow that no third-party insurance company will write coverage.  Here’s a blog post on that.  For those types of risks, businesses must self-insure.

Businesses that need to self-insure can do so by creating a captive insurance company to manage risk.  It’s a “captive” company when the business seeking risk management basically owns it.  And if it’s small enough—e.g. “micro”—it can claim some pretty sweet tax benefits under §831, benefits that help small businesses deal with unusual risk situations, like that bird flu.

But to get those tax benefits, the captive entity must be in the business of insurance.  That’s our lesson for today.  In Jones, et al. v. Commissioner, T.C. Memo. 2025-25 (March 25, 2025), Judge Nega writes a really informative (and loooong) opinion about what that means. 

For the short lesson, see below the fold.

Insurance is the oil in the engine of capitalism, allowing entrepreneurs to take bigger leaps by pooling resources and spreading risk.  As Judge Nega very nicely explains, there are basically two ways that businesses insure various risks.   First, they might purchase insurance from a third-party insurance company.  The business pays the insurance company a periodic fee in exchange for the insurance company agreeing to compensate the business if any of the covered risks materialize.  From the insured viewpoint, this shifts the risks to the insurance company.  Second, a company might self-insure, like I do for my dental risks.

In order to operate profitably, an insurance company needs to spread the risks it is assuming.  Judge Nega is careful to distinguish the insurer’s need to spread risk from the insured’s need to shift risk.  In my baby tax class I explain the risk spreading concept when we discuss annuities.  I’m 64.  This Social Security table says that, on average, a 64 year-old male will live another 17 years at least as of 2021.  (When I google on AI, however, it tells me the average is 19 more years.  Suddenly, I like AI!)  But those are just averages, or “law of large numbers” if you will.  So if I buy a “life-time” annuity, the insurance company will cut me a deal based on the law of large numbers.  They don’t care when I, individually, die.  For every annuitant who outlives the table there will be another who dies early for any one of a number of reasons.  As Paul Simon might have written, there are more than 50 ways to leave your lover by death.  The risks of death are quite diverse.  Thus insurance companies use the law of large numbers to distribute risk and price their product to make their profit.  (I’m thinking here that I would want to buy from a company that expects me to live only 17 more years, right?)

When a business buys insurance from a regular insurance company, the premiums paid are deductible per §162.  They are ordinary and necessary for carrying on the business.  And those payments also constitute gross income to the insurance company (well, most of ‘em; life insurance companies are subject to the taxation regime set out in §§801-811).

In contrast, if a business chooses to self-insure, like I do for my dental risks, they just put money aside.  No deduction for doing that!  But …

…what if a business could create its very own insurance company and shift the business risks to that separate company?  In some sense that what serial LLCs do.   That would be just like self-insuring only now money set aside over time would be deductible premiums.  Even better, Congress has created special tax rules for small (or “micro”) captive insurance companies.  Section 831(b)(2)(A) provides that “every insurance company” that has net written premiums of less than $2.4 million and meets the qualifying requirements set out in subsection (b)(2)(B) does not have to count premiums received as income but need only pay taxes on investment income.  Yes, you will have noticed some emphatic foreshadowing there.

Sweet!  So you just go create a micro-captive company, put money into it and, boom-bada-bing!, you have tax benefits.  Heck, you could then have that captive company “loan” you money.  And don’t worry about all that annoying taxish paperwork.  There are friendly companies, such as Captive Planning Associates, who can help you do all this. 

But….the captive company you create must still be an “insurance company.”  Ain’t no statutory definition of what that means.  It’s the Courts that tell us.  So let’s learn our lesson.

Facts
The tax years at issue 2015 and 2016.  The corporate taxpayer was a Sub S corp and had several individual shareholders, but the Court’s opinion focuses on two: Richard Shor and Sherry Maxon.  So will I.

Mr. Shor started the company, called Sani-Tech West, Inc. (STW), in 1992.  Ms. Maxon joined and worked her way up to eventually become COO and VP for operations.  As its name implies, STW was in the business of providing technical sanitation services, mostly for pharmaceutical and biotech companies.  The business evolved from distributing high-purity reusable components to manufacturing and distributing sterile single-use components such as tubing, filters, clamps, etc.  STW contracted with another company to sterilize what it made.  That involve transporting the finished products to that other company, re-transporting them back to STW facilities (where STW would quality-check the sterility) and then transporting to customers.  STW was well-managed and really successful.  It had big contracts with Big Pharma.

From 1992 to 2014 STW managed its various business risks through a general insurance contract with the Hartford (apparently paying something like $97k in yearly premiums in 2014.  Op. p. 17).   It also managed risk by using commercial delivery services to transport finished goods to customers.  As Judge Nega observes “Damage claims from customers were not frequent enough nor monetarily significant enough for STW to seek additional insurance coverage solely for those damages.”   Op. at 7.  Customer contracts provided that damaged goods could be returned for replacement but that almost never happened.

STW used its own vehicles to transport products to and from the sterilization company.  Until 2014 STW used flatbed trucks.  But twice in 2014 “a box containing sterilized products flew off an STW truck.” Op at 7.  The loss was, according to Mr. Shor’s testimony, “small potatoes.” Op. at 8.  Still, to manage that risk, STW started using a “fully enclosed box truck.”  Id.  That’s a sensible solution  The opinion is silent on what that cost, but this website suggests prices between $45k and $90k depending on size and condition.  STW also switched insurance carriers in 2017 from Hartford (who said they did not cover that risk) to Sentry Insurance.  Not only did that drop their yearly premiums to something like $67,000, but it gave more coverage: “The insurance purchased through Sentry would have covered any losses incurred were STW’s products to, once again, fly out of its trucks.”  Op. at 8.  Less cost, more coverage.  Good work!

Now comes the weird part.  Curiously enough, in 2015, Mr. Shor and Ms. Maxon decided to engage Captive Planning Associates to form a micro-captive insurance company, to manage “unknown” risks.  If that makes you scratch your head in wonder, you are not alone!  Judge Nega writes “We do not believe that Shor’s and Maxson’s purported concerns about STW’s “unknown” risks were the motivating reasons for pursuing captive insurance.”  Op. at 8.  My oh my, what could possibly be another motivation?

After killing a small forest of electrons, Planning Associates green-lighted the creation of a micro-captive entity, called Clear Sky Insurance Co., Inc. (CSI).  The new company was created under Montana law and owned by Mr. Shor (75%) and Mx. Maxon (25%).  It was capitalized at the Montana minimum of $250,000.

A few gigabytes later, all the folks involved decided that STW would pay this new captive entity a yearly premium of … wait for it … wait for it … $813,000.  Deductible by STW and not taxable to CSI.  And THAT amount was a last-minute reduction after a December 22, 2015 plea from Mr. Shor that he just could not afford the initial proposed premium of $1.16 million (which amount was, just coincidentally, just under then then statutory cap of $1.2 million in §831(b)). Op. at 19.

All the paperwork was completed by the end of 2015 and the coverage period started on December 31, 2015.  All nice and legal.  In August 2016 Mr. Shor took an unsecured personal loan from CSI for $400,000 to help him expand STW business in a curiously circular transaction that I am sure was all nice and legal, but that I do not quite understand.  (Mr. Shor eventually repaid the $400,000 principal to CSI in 2020, after the Tax Court litigation began, but without having made any interest payments).

Then, suddenly, late in 2016 Mr. Shor changed his mind about having a captive insurance company and refused to have STW renew the policies.  So the program lapsed on December 31, 2016.  Still STW took a deduction for the $813k it paid to CSI and CSI did not report that amount as income.  And all of those positions flowed through to the individual shareholders, including Mr. Shor and Ms. Maxson.

The IRS audited and disallowed the deductions asserting deficiencies against the various shareholders, ranging from a low of $1,100 to a high of $265k for CSI and then $258k for Mr. Shor and $82k for Ms. Maxon.

Lesson: If It Doesn’t Walk Like a Duck It’s Probably Not a Duck
As explained above, an entity cannot be a micro-captive insurance company unless it is, actually, an insurance company.  Section 831(c) provides that an insurance company is any “company more than half of the business of which during the taxable year is the issuing of insurance or annuity contracts or the reinsuring of risks underwritten by insurance companies.”

Hokay, but that does not tell us what “insurance” means.  Judge Nega explains that the Courts have developed a 4-factor test to decide whether a particular contractual arrangement constitutes insurance.

1. the arrangement must involve an insurable risk of loss;
2. the arrangement must shift the risk of loss from the insured to the insurer;
3. the insurer must distribute the risk of loss among its policyholders; and
4. the arrangement must be considered insurance in the commonly accepted sense.

These are not a Wobbly Table of Factors.  As Judge Nega puts it “[f]ailure to satisfy all four factors is fatal to characterizing an arrangement as insurance.” Op. at 33.   He cites AMERCO v. Commissioner, 96 T.C. 18 (1991) for that proposition.  There, the Tax Court called these four factors “principles” and explained that:

“[t]hese four principles do not yield a definition of insurance. They do, however, create what we believe is the proper framework to be adopted when addressing a question of the existence of insurance for Federal tax purposes. They are not independent or exclusive. Instead, we read them as informing each other and, to the extent not fully consistent, confining each other’s potential excesses.” 96 T.C. at 38.

This is the first layer: a 4-factor “framework.”  As applied to today’s case, Judge Nega concludes that CSI was not in the business of insurance because “CSI did not distribute the risk of loss among its policyholders, and the CSI Program was not insurance in the commonly accepted sense.”  Op. at 33.

I will spare you Judge Nega’s very well-written and detailed analysis on why CSI did not properly distribute the risk of loss.  It’s great reading but it does involve a Wobbly Table of 9 Factors.  All you need to know is that it’s a WTF test.

More pertinent to today’s lesson are the reasons why Judge Nega finds that CSI’s operation was not insurance in the commonly accepted sense.  This involves a second layer of analysis and here again we run into a WTF test.  But it’s a short one and Judge Nega lists the important factors as follows:

“whether (1) CSI was organized, operated, and regulated as an insurance company; (2) CSI was adequately capitalized; (3) the policies CSI issued were valid and binding; (4) the premiums CSI charged were reasonable and the result of an arm’s-length transaction; and (5) CSI paid claims.”  Op. at 41-42.

Judge Nega’s basic problem that CSI was neither organized nor run like a real insurance business.  It may have looked like a duck on paper but it did not walk like a duck.  It was an epic fail on the first, third and fourth factors because it was just sloppy.

First, to the extent it operated, it walked like a crippled duck.  It lasted only a year and “even during its brief stint as a captive insurer, it did not operate as an insurance company normally would.”  Op. at 42.  Here Judge Nega takes a robustly functionalist approach, emphasizing that “we must look beyond the formalities and consider the realities of the purported insurance transactions.” Id.

The basic problem was the slip-shod manner in which CSI was created and run.  To start with, “no one performed any underwriting analysis when crafting the CSI policies.”  For example, the CSI policy for STW insured against earthquakes and floods but “there is no evidence of…anyone inspecting the site to assess seismic or flooding exposure or even considering flood or seismic maps that might be relevant to a coastal location in Southern California.” Op. at 43-44.

Worse, CSI’s purported “loan” of $400k to Shor was totally out of line for a real business.  Or to quote Judge Nega’s more lawyerly phrasing, “it is not commercially reasonable” for an insurance company to lend over 40% of its investment portfolio as an unsecured loan at way below market rate and without any evaluation of the borrower’s credit.  Just. Sloppy.  And then CSI extended the loan’s term in violation of its formal Financial Operations Manual.  And never received interest.

Second, even the formal policy documents issued by “CSI” (or, most likely, the canned documents used by the organizer) were missing “essential elements.”  Worse, they “had terms that deviated significantly from commercial industry standard.”  Op. at 46.  Just. Sloppy.

Third, the amazingly high premiums were pulled out of someone’s big ass..umptions.  They bore no discernible relation to the purported coverage.  Judge Nega noticed how the supposedly properly calculated premiums were revised just because the boss casually commented that “they seemed too high.”  Op. at 47.  So they dropped from $1.1 million to $813,000.  And even then they were “between 48 and 63 times as expensive as the insurance STW purchased from Hartford.”  Op. at 46.  That may have made sense if proportional to the coverage but “STW submitted no claims to CSI during the coverage period, despite a range of 16 different policy coverages.”  Id.  Sloppy hogs.

Bottom Line:  Don’t be sloppy.  Whether a company is in the business of insurance is a well-developed common law concept.  Because the inquiry is so dependent on facts and circumstances, taxpayers cannot afford to be sloppy.  You can’t just create the duck, you have to walk like a duck.  Oh, and don’t use AI.

Coda 1: How much money did these taxpayers waste on this frolic?  They had to have spent a lot setting everything up.  And then the IRS audit costs.  And then in Tax Court the taxpayers were represented by the firm of Panitz and Kossoff.   All of this had to be expensive.  The IRS put on experts.  The taxpayers put on experts.  I wonder whether the taxpayers had—well—insurance to cover those litigation costs!

Coda 2:  In January 14, 2025, Treasury issued these final regulations regarding when micro-captive insurance arrangements will not be honored.  I highly recommend reading the 2023 proposed regulation’s Preamble to get a sense of what Treasury was attempting to do in the regs.  But notice that the basic question in this case—what constitutes doing business as an insurance company—is not addressed in the regulations.  The preamble explains that the common law test—here used by Judge Nega—for determining what constitutes “insurance” continues to apply.

Bryan Camp is the George H. Mahon Professor of Law at Texas Tech University School of Law.  He invites readers to return on the first Monday of each month (or Tuesday if Monday is a federal holiday) to TaxProf Blog for another Lesson From The Tax Court.  

[Editor’s Note:  If you would like to receive a daily email with links to each Lesson From The Tax Court and other tax posts on TaxProf Blog, email here.]