Over the past decade, captive insurers and their owners making the 831(b) tax election have come under scrutiny from the IRS. Section 831(b) of the U.S. federal tax code allows an insurance company to pay taxes on their investment income only (as opposed to net income). As a result, quite a few captive insurance companies making the 831(b) tax election have been audited by the IRS for allegedly being set up not to provide insurance, but instead solely to achieve tax benefits.

There have been a few cases in the federal U.S. Tax Court involving insurance companies that made the 831(b) tax election. The IRS argues several reasons why these companies are not legitimate insurance companies. A common argument throughout these cases is that a “circular flow of funds” exists. With one exception, the IRS has prevailed when using this argument—and others. So what differentiates a captive that may fall prey to a circular flow of funds and scrutiny by the IRS? Is the concept of a circular flow of funds an argument under a risk transfer setting? Or is it a fact that’s true for all insurance risk pools? There are some key elements of which captive insurers should be aware.

Circular flow of funds

Before we answer those questions, let’s start with the definition. “Circular flow of funds” describes a financial transaction or a series of financial transactions where 100% of the funds paid from Entity A to Entity B is subsequently paid back to Entity A in a financial transaction or a series of financial transactions. Many captives making the 831(b) election participate in risk pools, where a portion of the premium is paid to a risk pooling mechanism and, in return, the captive receives an equal amount of premium on a basket of risks of other captive insurers that likewise are participants in the risk pool. Under the “substance over form” approach, it is argued that, because this “money in”—with no corresponding true “money out”—transaction lacks any sort of economic substance and has a net result of essentially nothing, one should ignore the intervening steps of the transaction(s) and only consider the two endpoints. When isolating the impact of such transaction(s) in this context, and in the framework of a captive making the 831(b) tax election, the circular flow of funds arguably has no economic purpose other than to provide a tax benefit to the insurer.¹ As we will see later, however, this argument fails to include the other key insurance transactions.

Tax Court decisions

Figure 1 shows a timeline of the major Tax Court decisions involving captives making the 831(b) election. With the exception of Puglisi, the IRS has prevailed in all of the cases, and all captives except for Caylor² participated in a risk pool.

Avrahami v. Commissioner

Benyamin Avrahami and Orna Avrahami, Petitioners v. Commissioner, 149 T.C. No. 7 (2017).

The first major case was brought against Benyamin and Orna Avrahami. The Avrahamis were jewelry store owners who incorporated Feedback Insurance Company (Feedback) in 2007 as a captive electing the 831(b) tax status. Feedback offered a variety of insurance coverages to American Findings Corporation (American Findings), a jewelry store owned by the Avrahamis, and several other entities owned by either the Avrahamis or their children (Avrahami Entities).

Feedback participated in a terrorism risk insurance pool by the name of Pan American Reinsurance Company, Ltd (Pan American). The Avrahami Entities would pay Pan American a premium of $360,000 per year for terrorism coverage. In turn, Pan American would pay $360,000 to Feedback for reinsuring a portion of similar terrorism risks from approximately 100 other pool participants (i.e., captive owners).

As an example, the flow of funds—representing only premiums—for 2010 is outlined in Figure 2. The $810,000 represents premium paid by the Avrahamis to Feedback for their direct insurance policies.

In January 2013, the IRS issued a notice to the Avrahamis bringing many elements of their business arrangement into question, including that Feedback had never paid any claims. Upon conclusion of the trial, the Tax Court ruled in favor of the IRS. Furthermore, Pan American was ruled to have been created only for the tax benefit of its members and found to rely greatly on the circular flow of funds. That is, American Findings would pay a premium to Pan American, only for this same premium to be paid directly to Feedback, with no corresponding claim payments.

Risk pools

The circular flow of funds argument—for premiums only—held true for the other cases involving captives making the 831(b) tax election as well. Except for Caylor, the captive owner paid premiums to a risk pool, only for that risk pool to turn around and pay that same dollar amount of premium (or that same premium minus a small administrative fee) back to the captive insurance companies.

Why did this happen? Because that’s how risk pooling works. For insurance companies participating in a risk pool, the circular flow of funds argument is not an argument, but instead simply a statement of fact. The negative connotation implied by the circular flow of funds argument completely disregards the risk-sharing benefits that captive owners can achieve. Most risk pools don’t retain risk, but instead redistribute risk. For these cases, the risk of each pool participant, i.e., the 831(b) captives, is redistributed to other captive insurance companies. This is done proportionally on the premise that each captive’s risk profile, as measured by its expected losses as a percentage of premium, is the same. Thus, each captive receives premiums exactly equal to what they paid into the risk pool, but with a completely new risk profile. This redistribution of risk is why risk pooling is beneficial to both captive insurance companies and the captive owners. When a claim is reported to the pool, its cost is shared among all pool participants so the company that experienced the loss does not have to pay the full amount. It is risk shifting and risk distribution, using the IRS terminology.

When risk pools are presented with claims, the circular flow of funds argument falls apart. Flow of funds should include all cash flows of the insurer—premiums, claim payments (losses), expenses, and investment income. Here, for simplicity, we have assumed that expenses incurred and investment income earned by the captive are immaterial to whether or not the captive is a true insurer. As such, we are focusing on only premiums and losses.

Avrahami v. Commissioner: Hypothetical

What would have happened if American Findings did in fact have a claim? As an illustrative example, suppose that American Findings had a $5.525 million terrorism claim (policy limit for 2010). To keep things simple, let’s also assume that no other companies in the Pan American risk pool had a claim. According to the structure of the agreement, American Findings would file this claim with Pan American, and Feedback would assume from Pan American $99,284 of the losses related to this claim (or 1.797%, equal to Feedback’s share of the total pooled premiums). The premium flow of funds would be the same as outlined above in Figure 2. The hypothetical losses flow of funds are outlined in Figure 3.

What happens to the flow of funds now? For the Avrahami companies, while premiums paid to the risk pool exactly equal the premiums returned, claim payments of $5.525 million are insured, with only $99,284 of this insured by Feedback. Absent the captive and the risk pool, the Avrahami companies would have absorbed the full $5.525 million loss—a very different flow of funds! Similarly, the other pool participants’ flow of funds is also not circular, as each captive or captive owner, despite having no claims of their own, now must pay a portion of the Avrahami claim.

Why is all this important? American Findings benefits from being able to access the reinsurance market (through Pan American), as one of the many benefits of captive insurance is risk shifting through the risk pool. Losses incurred by American Findings ($5.525 million) do not equal losses assumed by Feedback ($99,284). Similarly, if another company (Company X) in the risk pool had a $5.525 million terrorism claim, and no other companies (including American Findings) had a claim, Feedback would be responsible for $99,284 of the claim. In this second example, another pool participant would have benefitted from the reinsurance afforded by the risk pooling, and American Findings, now acting in its capacity as a reinsurer of other captives, must pay a portion of someone else’s losses (see Figure 4). Again, the key concept here is that when claims are presented, the amounts received by Pan American and paid by Feedback are not equal, meaning the flow of funds—inclusive of claims—would not be circular.

What if there are no losses?

The presence of claims subject to a risk pool will almost always defeat the circular flow of funds argument. However, losses are not required, only the risk of losses. Consider the federal Terrorism Risk Insurance Act (TRIA) program that started in 2003, which functions in a similar fashion to the risk pool described above. Through 2021, premiums paid to insurers for terrorism insurance were approximately $59.7 billion. However, no claims have ever been presented to the federal government and the IRS has never challenged the deductions taken by policyholders for premiums paid for terrorism insurance.

The reason there have been no claims and why this isn’t a problem for the IRS is that this coverage is low-frequency/high-severity. There is a very low chance of a claim (low frequency), but if a claim does occur it most likely will have a very large dollar amount (high severity). Most of the coverages written by these captives and risk pools are similar in nature. The fact that there hasn’t been a claim or that there are very low loss ratios doesn’t mean that there won’t be a claim in the future, or that there wasn’t a risk of a large claim in the past. That is the nature of insurance. Most drivers buy auto insurance, but most drivers do not get into accidents. We buy insurance in case there is an accident.

Call your actuary!

The critical piece in defending against the circular flow of funds argument is that the premiums paid to the captive insurance company are reasonable, or actuarially sound. The IRS often likes to compare premiums paid to captives making the 831(b) election with premiums paid for similar coverages purchased in the traditional commercial insurance market. That argument fails to recognize that captive insurance companies are designed to fill coverage gaps in the traditional insurance market and are providing completely different coverages that often aren’t available in the commercial market. As such, the comparison is not valid. Thus, instead of asking whether these premiums are higher or lower than other coverages purchased in the commercial market, the better questions to ask are:

• Who determined these premiums?
• Are the premiums actuarially sound?
• Were the premiums established using generally accepted actuarial techniques and reasonable assumptions?
• Did the actuaries abide by the applicable actuarial standards of practice?

The best defense against a future circular flow of funds argument is to have an actuary involved in the planning stages of captive formation and pricing of the insurance coverages from the time that a captive is formed. A robust pricing review should be conducted by an actuary on an annual basis.

In summary

Flow of premium funds analyses are, in and of themselves, insufficient in determining whether or not a captive is a bona fide insurance company. Actual claim experience, as well as the probability of future reported claims, needs to be considered. In the hypothetical examples above for the Avrahami case, one large insurance loss eliminated the circular flow of funds argument. In addition, in the first hypothetical example, American Findings benefitted by having the remaining amount of the claim, over $5.4 million, paid by the rest of risk pool. However, risk pools don’t need to have claims to demonstrate that a true risk transfer and insurance program exists.

For many small captive owners, making the 831(b) election may make sense from a tax perspective. The key point is to make sure the captive itself is a legitimate insurance company, and that the formation of the captive is not primarily driven by federal tax benefits. This includes involving an actuary from the very beginning to ensure the annual premiums calculated are reasonable and in accordance with actuarial standards of practice.

¹ Cummings, Jasper L. (November 2, 2011). Circular Cash Flows and the Federal Income Tax. American Bar Association. Retrieved October 4, 2022, from https://www.americanbar.org/groups/taxation/publications/tax_lawyer_home/11spr/ttl-spr11-cummings/.

² The Caylor case is different because the captive didn’t participate in a risk pool.