After a decades long streak of losing most every Tax Court case of consequence involving captive insurance companies, the IRS has managed over the last year to win two, though they are both of questionable relevance to most of the captive insurance industry.
The most recent case is called Reserve Mechanical. Like Avrahami, it involved facts that are atypical to those of most of our clients, facts that made the court’s job in ruling for the IRS easier. Rather than recite those facts at length, it’s more efficient to discuss the court’s conclusions and then mention any relevant details as we go along.
In a nutshell, the court concluded that the captive insurance company in question was not entitled to the tax benefits of Code Section 831(b) for two reasons: (1). There was insufficient risk distribution due to deficient risk pooling and (2) the insurance company was not operating in such a way that it could be viewed as offering insurance “in the commonly accepted sense.” Today we will discuss the first basis for the court’s ruling. We’ll discuss the second basis in next week’s installment of Captivating Thinking.
RISK DISTRIBUTION POOLS
Risk distribution, which is the ability of the insurance company to spread its risks among multiple insureds and risk exposures, is an inherent part of any true insurance arrangement. Many captive insurance companies attempt to achieve risk distribution via participation in risk distribution pools, or simply “risk pools”. In this context, risk pools are arrangements where one captive insurance company assumes some of the risks of multiple other (usually unrelated) captive insurance companies.
Risk pooling can be accomplished in a few different ways. One common way is for the captive insurance company (CIC) in question to sell insurance policies directly to its primary insured. That CIC then “cedes” (that is, transfers) a portion of that risk (usually at least 30% but often as much as 50% or more) to an unrelated reinsurance company. Many other unrelated CICs do the same thing with the same unrelated reinsurance company, so the unrelated reinsurance company therefore serves as the “risk pool”—it assumes various risks from multiple unrelated CICs who in turn insure unrelated insureds of their primary insureds. The unrelated reinsurance company, or the “risk pool”, then “retrocedes” a blended portion of all of the risks that it has assumed back to each of the ceding CICs. By blended risk, we mean that each CIC participating in the arrangements assumes a proportionate share (commonly called a “quota share”) of all the risks assumed by the pool—that is, risks of completely unrelated companies.
The risk pooling mechanism employed in the Reserve Mechanical case was a little different than that described above, but the distinction is irrelevant. The important thing to note is that there are generally multiple separate transfers of risk in any risk pooling arrangement. For instance, in the one described in the last paragraph above, there were three separate risk transfers—the transfer from the primary insured to its related CIC (via the directly written policies), the ceding of some portion of that same risk from the related CIC to the unrelated reinsurance company, and then the retrocession of the reinsurance company’s now “blended” risks back to the CIC.
EACH TRANSFER OF RISK MUST BE ACCURATELY PRICED
For any captive insurance arrangement to be honored as a real insurance arrangement, it’s critical that each transfer of risk be accurately priced using sound actuarial principles. Without sound actuarial pricing, we don’t have a valid “insurance” arrangement.
However, in many risk pool arrangements, including the one employed by Reserve Mechanical, only the directly written policies (those issued by the related CIC directly to its primary insured) were actuarially priced. The other or subsequent transfers (such as between the risk pool reinsurance company and the Reserve Mechanical CIC), had no actuarial pricing support and seemingly no other objective justification. Quite simply, they were not “arm’s length” transfers of risk and associated premiums.
Furthermore, in the Reserve Mechanical case, the premium paid to the risk pool exactly equaled (in dollar terms) the amount of premium subsequently paid by the pool to the CIC in exchange for the latter‘s assumption of the blended risk of the pool. This resulted in a “circular flow of funds”—dollars in equaled dollars out, and essentially to the penny.
This circular flow of funds combined with a lack of actuarial support for the transfer of risk to the pool and its subsequent transfer of blended risk back to the CIC was enough to draw the suspicion of the court. In the absence of an economic justification for the circular flow of funds, the arrangement looked to the court to be deficient so as to not qualify as insurance. Since premiums paid were seemingly unrelated to the amount of risk transferred, there was no true risk distribution, at least not as that term is understood in a traditional insurance context.
WHEN PREMIUM EQUALS RISK AND RISK EQUALS PREMIUM
Despite the (misguided or perhaps malicious?) comments by some industry pundits to the contrary (Jay Adkisson, I’m looking at you), there are, in fact, instances in which an underlying economic justification for a circular flow of funds may exist. And there are likewise instances in which separate actuarial pricing for subsequent risk transfers may be unnecessary or, more properly, redundant.
For instance, when a CIC cedes a fixed percentage of its total risks (from first dollar risk to last dollar risk) under each direct written policy to the risk pool, then most any actuary will confirm that no separate pricing study is required for that subsequent risk transfer, at least if we assume that the policies originally issued by the CIC were underwritten and priced correctly. In such cases the actuarial pricing of the directly written policies is sufficient and controls subsequent transfers of that risk because a fixed portion of that exact same risk is subsequently transferred. Since there has been no chance to the nature of the risks and the underlying policies were priced accurately, risk equals premium and premium equals risk in such arrangements. So, transferring some fixed percentage of the overall risk means also transferring that same percentage of the overall premium.
For instance (and subject to exceptions that are irrelevant to today’s discussion), if the CIC’s directly-written policies are accurately underwritten and priced by an actuary and the CIC subsequently cedes 50% of ALL of its risks on each policy (from first dollar to last dollar) to the pool reinsurance company – then – actuarily speaking and assuming the pool reinsurer has confidence in the initial pricing, the reinsurance company is simply due half of the premium received by the CIC.
By definition, ceding half the risk of loss on each policy means ceding half the premium if the policies are correctly priced. This is not a controversial statement. Most any actuary would readily confirm it once given sufficient context. However, in light of the Reserve Mechanical decision, having written actuarial acknowledgment of this logical fact is strongly advised.
RESERVE MECHANICAL’S PROBLEM
But notice that the logic above does NOT work unless (1) the risk transferred remains unchanged so that the original actuarial pricing still holds, and (2) a fixed portion of each and every risk is transferred. In the case of Reserve Mechanical, the structure didn’t transfer a fixed portion of each and every risk and the risks that were subsequently transferred changed.
In the Reserve Mechanical risk pool, the individual CICs retained ALL of the risk of loss up to a certain dollar amount for each claim under each policy and then transferred only the risk of loss above that dollar amount to the risk pool reinsurer. Said another way, only the “stop loss” portion of the risk was actually assumed by the pool. And only a blended portion of that stop loss risk was then ceded to each individual CIC. All losses below the stop loss amount were retained by the individual CIC, and each CIC bore its quota-share of any losses in excess of the stop loss amount.
Because only stop loss coverage was pooled, and because claims below the stop loss amount are almost certainly likely to be more frequent in occurrence, separate actuarial pricing was certainly necessary to establish how much of the actuarial risk of loss was retained by each captive and how much was transferred to the stop loss carrier. Furthermore, if the stop loss amounts were set at different levels for different risks for policies issued by different CICs, then every single policy would essentially need to be repriced to determine the correct premium associated with the actuarial risk transferred.
In Reserve Mechanical, such pricing did not exist. Rather, the amount transferred between each CIC and the risk pool reinsurers was simply calculated as the amount of money required to ensure a circular flow of funds—that is, the amount needed to insure that each CIC got back exactly what was paid to the risk pool. This was enough for the court to question the legitimacy of the arrangement:
The evidence shows that the stop loss pool was divided among the captives so that reinsurance premiums equaled the portion of direct premiums paid by each captive’s affiliated insureds. We conclude that the amounts that [the pool] was to pay [the CIC] under the quota share arrangement were not determined at arm’s length or using objective criteria.
WHAT THE COURT DIDN’T SAY
Some commentators, such as the ever-alarmist-IRS-apologist, Jay Adkisson, contend that the court effectively ruled that ANY circular flows of funds between a captive and a pool “does not provide risk distribution” and therefore that “quota-share risk pools simply don’t work.” This is not in fact what the court ruled at all, nor is such a conclusion even a remotely reasonable inference from the ruling.
The circular flow of funds was only an issue in this case because, as the quote from the court above implies, (1) a fixed percentage of all the overall risk wasn’t transferred from the pool to the CIC, (2) the risk transferred was different from the original risk assumed by each captive due to stop-loss layering, (3) no arm’s length actuarial analysis determined or priced the portion of the risk that was transferred from the pool to the CIC, AND (4) the amount that was transferred from the pool to the CIC was calculated specifically to ensure a circular flow of funds—that is, to ensure that everyone ended up with the same amount of money they began with. The court makes quite clear that these factors were the source of its concern:
The perfect matching of payments under the corresponding stop loss endorsements and quota share policies (from insureds to [the pool], and from [the pool] to captives) indicates that the quota share arrangement was not the product of arm’s-length considerations.
And why is that so? The court tells us why in the very next sentence:
[The primary insured’ s] risks that were insured through [the pool] were different from the risks that [the pool] ceded to [the CIC] under the quota share policies.
And why were they different? Because only the stop loss layer was pooled, and there was no actuarial delineation of the portion of the risk of loss retained by each captive versus the portion absorbed by the pool. The payment of funds from one party to the other was simply designed to ensure a circular flow and not to account for the underlying risk transfer economics of the transaction. Again, the court makes the nature of its concern crystal clear:
The same amount that [the primary insured] and the other insureds were obligated to pay [the pool] for the stop loss coverage was to be paid to [the CIC related to the primary insured] pursuant to the quota share arrangement. [The CIC] has not explained why these amounts were the same. It has not explained how all [pool] clients in the quota share arrangement would be able to transfer a particular set of risks (i.e., those associated with their affiliated insureds) and assume in exchange a blended portion of completely different risks for exactly the same premium price. [emphasis added]
Note that if the taxpayer HAD been able to explain why these amounts were the same, which is easily done with an actuarial approach when a fixed percentage of fixed risks (from first dollar to last) is transferred on a quota-share basis, and if the risks subsequently transferred HAD been the same, then the court would have reached a very different conclusion. Regrettably for Reserve Mechanical, that was not the case.
ZERO POOLED CLAIMS ENSURED PERFECT CIRCULARITY OF CASH FLOW
It’s important to note (because the court emphasized it) that no claims were paid by the pool in Reserve Mechanical’s case because none of the underlying policies experienced claims sufficient to trigger the stop loss protection offered by the pool. Because the pool paid no claims, the circular flow of funds was perfect. Dollars paid in to the pool exactly equaled dollars returned to the original payers (or their CIC). As the court said:
The CIC] never recorded and it does not contend that it had any losses or expenses in connection with its purported quota share liabilities. Accordingly, the end result for each tax year under the quota share arrangement was that [the CIC] would receive payments from [the pool] in exactly the same amount as the payments that [the pool] was entitled to receive from [the primary insured of the CIC] and the other insureds for the stop loss coverage.
And again, all without any actuarial justification!
IS THIS THE END FOR RISK POOLS?
Contrary to Jay Adkisson’s contentions, the court’s decision simply does not address pools in which (1) a fixed percentage of the total risk (from first dollar to last dollar) is transferred, (2) the risk transferred is unchanged from the originally-priced risk, (3) the percentage of premium transferred exactly corresponds to the percentage of risk transferred, (4) the pool regularly and consistently pays claims, and (5) payments from the pool to the CIC are reduced by these pool claims, thereby interrupting the near-perfect circular flow.
Neither I nor any other lawyer or CPA that I know (and I’ve spoken with many) interpret this case to mean that all quota-share risk pools are dead. Quota-share arrangements are not something invented by the captive insurance industry for some nefarious tax purpose but rather are common risk distribution structures employed in innumerable commercial insurance contexts.
The obvious issue in the Reserve Mechanical case wasn’t the existence of a quota-share arrangement or even the circular flow of funds, it was simply that the circular flow of funds had no underlying actuarial or economic justification. When the flow of money is simply designed from the outset to ensure circularity with disregard to the amount of underlying risk transferred, it’s just not real insurance. Fortunately, that’s simply not the case with many other risk pools, and the court’s ruling therefore has little to no bearing on them. If anything, the court’s rationale in Reserve Mechanical explicitly supports the legitimacy of these latter types of pooling arrangements.
MORE ON THIS SUBJECT SOON
Coming soon in Captivating Thinking, we’ll discuss the second basis for the court’s ruling against the Reserve Mechanical CIC, namely that it wasn’t operated in such a way that it would be said to offer “insurance” in the “commonly accepted sense.”