Berezko /
15 April 2019Accounting & tax analysis

Risk pools: guesswork not allowed

“Let’s work the problem, people. Let’s not makes things worse by guessing”—Gene Kranz

The case of Syzygy Insurance Co v Commissioner was decided on April 10 by the Tax Court and adds clarity to the muddy 831(b) risk pool waters.

This latest victory for the Internal Revenue Service addresses a cellular captive reinsuring a protected cell company providing a front on a 100 percent reinsured basis. The cell’s premiums were allocated to two layers of risk: 49 percent of the premium to layer 1, and 51 percent of the premium to layer 2 to reinsure the risks of the captive.

Three-and-a-half months after the policy period, almost all the layer 2 premiums were distributed back to the captive. The end result of this transaction resulted in a situation where 51 percent of the premium attributable to the captive came from a retrocession from the fronting carrier’s risk pool—presumably to satisfy the third party risk rule from the Harper Group case.

A clear road map

In contrast with Avrahami and Reserve Mechanical, the Tax Court provides a relatively clear road map of how it evaluated whether the captive’s premiums were valid deductions for insurance. First, the court evaluated whether the fronting carrier was a bona fide insurance company.

Second, the court assessed whether the captive’s insurance constituted insurance in the commonly accepted sense.

With respect to the fronting carrier, the court took issue with the following:

  • The circular flow of funds between layer 2 and the captive;
  • The fact that the premium rates were about five times as expensive in the captive as on the commercial market;
  • The peculiar clauses in the insurance policies; and
  • The lack of a qualified actuary calculating the insurance premiums;

With respect to the cell, the court identified the following deficiencies as particularly troublesome:

  • The largest source of premium—deductible reimbursements—was not well-managed and there were countless claims that could have been reimbursed that were overlooked by the captive owners;
  • One of the captive’s investments was a split-dollar life insurance policy;
  • Habitually late issuance of captive insurance policies;
  • The lack of quality actuarial studies; and
  • The general lack of claims in the captive.

The court declined to opine as to the sufficiency of the risk pool. The pool apparently had 857 policies in it. Nothing in the opinion discussed whether this constituted a sufficient number of points of risk for risk distribution.

Perhaps the court declined to directly address this issue as it already had two alternative grounds on which to reverse the premium deductions. Or, perhaps worrying about sufficient risk distribution is less of an issue in a post Rent-A-Center/Securitas world.

The court spent time evaluating the deductible reimbursement policies. Given that the parent company had commercial insurance with 11 other carriers, there were numerous opportunities to file reimbursement claims through the captive. Yet, the owners declined to do so.

According to the petitioner, it was because accounting for the deductibles was too much effort. This was a poor defence and shows that the captive manager did a poor job explaining how captive insurance works. The premiums paid to finance the deductible reimbursements should (theoretically speaking) be greater than the claims coming out of the captive.

Financing the deductibles and paying them out of the captive should still result in an overall net saving to the taxpayer. The fact that the captive manager failed to enforce this provision was an enormous oversight.

This issue—the lack of claims—is a touchy one. On the one hand, it’s reasonable to have zero claims in a year. On the other hand, too few incidents and the purported insurable risk starts to look like a non-issue. It is probable that had the captive been operated properly the IRS would not have had any ability to attack the premiums. Further, the reimbursement of the claims by the captive would have kept the parent in an overall better tax position. Again, this fact is hard to ignore as it reveals slipshod management.


There are several general takeaways from the case. First, fronting carriers are evaluated as viable insuring entities on basically the same basis as captives. Second, captive owners need to report their claims and incidents. Third, actuaries are key to a viable captive programme. Fourth, risk distribution is not a set-in-stone issue. There is tremendous leeway in the risk distribution arena.

Above all, the opinion paints the picture of a captive manager running a loose operation with poorly drafted policies, little risk management, and poor policy binding/issuance practices. In addition, the reverse-engineering of premiums to meet an IRS safe harbour shows that this captive, and likely all others in the risk pool, served a tax function more than a risk management function.

Guesswork will get you audited. You cannot guess at the captive insurance company’s premiums. Nor can you guess at the policy language. Nor can you guess at what will qualify as legitimate investment interests for an insurance company.

The strange part is that the captive did not have to be a sham transaction. The insured interests in the captive are standard insurable interests. Had claims been paid and policies issued in a timely manner, then the tight operational practices may have overcome the shortcomings with respect to the lack of a formal actuarial. The court shows us that captives which appear to be constructed to generate tax deductions will not withstand scrutiny.

From Syzygy we see that the Tax Court has several means at its disposal to assess the viability of a captive insurance company. Although the judge declined to issue a bright line ruling, we do see that “fishy” transactions will be investigated.

Ultimately, risk pool validity is an “all of the above” test, weighing various factors. This means that captive managers need to operate a tight ship in their risk pools. Guesswork is fatal.