Lessons Learned from the "Avrahami" Case

Gavel and tax blocks

Jeremy Colombik , Richard M. Colombik | August 30, 2017 |

Gavel and tax blocks

There have been many articles and much information on Avrahami v. Comm'r and Feedback Ins. Co., Ltd. v. Comm'r, 149 T.C. 7 (2017). Within the lengthy court opinion, there are many lessons and discussions. So, what can we learn from this case?

First and most important is that the Avrahamis' captive was not operating like an insurance company. As such, it should be no surprise that a company that is not acting as an insurance company would not receive nor be entitled to insurance company benefits. Most people in the captive industry were not surprised by this decision, and, in my view, the court acted reasonably, based on the facts at issue. This article will discuss the main flaws of why this captive was not acting like an insurance company and offer suggestions for how a proper insurance company should and would act.

The areas that the court appears to have focused on are risk distribution, policy language, actuarial standards, and claims. Let's begin our discussion by addressing risk distribution. This, in my view, is the most important lesson you can learn from this case, as risk distribution correlates with policy language, actuarial computations, and payment and processing of claims. That is why I believe it is the main issue with the analysis.

Risk Distribution

Proper policy language, actuarial standards, and payment and processing of claims are needed in order to operate as an insurance company and are needed in an insurance or risk pool if that is the chosen solution for achieving risk distribution. According to the court, there was an insufficient number of brother-sister companies for risk distribution; hence, there was not proper risk distribution. The Avrahamis lost their argument, in part due to their captive having insufficient risk distribution because the risk pool they participated in, Pan American Reinsurance Company, Ltd. ("Pan American"), was found to not be a bona fide insurance company. As such, no risk distribution occurred. That, of course, led to an analysis of what constitutes an insurance company.

In Helvering v. LeGierse, 312 U.S. 531, 61 S. Ct. 646 (1941), to be considered insurance, an arrangement must do the following:

  • involve risk shifting,
  • involve risk distribution,
  • involve insurance risk, and
  • meet commonly accepted notions of insurance.

In Avrahami, the plaintiffs failed the test of what constitutes insurance, based on risk distribution not being present. This case did fail multiple tests based on LeGierse. Not only did they not have risk shifting, but they also had insufficient risk distribution to establish the existence of insurance. If you fail either test, you fail to be an insurance company.

Risk distribution, according to Avrahami, "occurs when the insurer pools a large enough collection of unrelated risks." The Avrahamis argued that they had proper risk distribution through Pan American since they did not have enough brother-sister companies to accomplish risk distribution without Pan American.

The Avrahamis' captive, which was called Feedback, entered into a cross-reinsurance program pool that pooled only 30 percent of their risk. The Tax Court has previously ruled that only 29.6 percent of unrelated risk, at a minimum, needs to be pooled. Therefore, the Avrahamis' counsel argued that 30 percent would have qualified if Pan American was found to be a bona fide insurance company. Pan American was acting unusually. For example, Pan American did not charge a ceding fee, which is standard for an insurance company pooling risk. Even though 100 percent of the risk was reinsured, an insurance company should be getting compensated for services it is providing. Again, Pan American is a business and should be run as a business. Moreover, Pan American, according to the court, was collecting "more than $20 million in premiums in 2009 and in turn agreed to insure up to an aggregate of more than $308 million in losses." Further, Pan American was only holding 2.5 percent of the premiums collected until the end of the policy period. This brings into question how Pan American could have sufficient capital and surplus to pay its claims.

Generally, based on previous claims history, a captive manager would hold onto approximately 25 percent of premiums collected until the end of the policy period, depending on the policies' aggregate limits. If the aggregate limits are higher, which seems to be the case with Pan American, then the holdback on the remaining premiums should correspondingly have been much higher. Further, the Pan American policy pricing was found to be excessive.

When your actuary, who is hired as an expert in this area, has one client in the captive industry, your firm, and testifies in court that "he did not know of any event in history that would have met these requirements" to file a claim for your policies, you are likely at risk. The testimony makes one wonder whether the actuary was merely pricing the policies at the amount the captive manager in this case requested. Again, this is an insurance company, and an insurance company is a business.

Any insurance company, big or small, will expect to have claims made and use its actuary and underwriters to properly price its policies. If you have a policy for which there is virtually no way to have a claim filed, then it does not appear to be within the common concept of insurance. Further, if it is almost impossible to file a claim, why would a business pay hundreds of thousands of dollars for such a policy? One lesson learned is that if you are in a captive, make sure to ask the following: Who is the actuarial firm? Is it reputable? Is it experienced? Does it have other clients? This way, you know that the firm has experience in the industry, has a track record working with other businesses and captive managers, and will price any such policy according to the risk covered and the market rate for such coverage. Further, ask the captive manager about the insurance pool you are in, and ask questions such as the following: Where is it domiciled? How many unrelated entities are in this pool? What is the claims history? How much of the premium you are paying is pooled? What types of business are in this pool? These questions assist in the initial vetting of the insurance pool and determining whether the actuarial firm and captive manager pass the smell test (no, not in the Internal Revenue Code (IRC) but through common sense).

Policy Language

Next, let's address policy language. As I stated earlier in this article, policy language does relate to risk distribution, as valid policies need to be issued in order for an entity to even be considered an insurance company. Policy language in an insurance policy from an insurance company should not have conflicting language, which, in my view, is another lesson learned from this case. The Avrahamis argued that their policies were claims made; however, the policy language contained occurrence policy language as well. It was described by the court as follows: "… between December 2009 and December 2010, terms indicative of both a claims-made policy—the claim must be reported during the policy period—and of an occurrence policy—the claim must occur during the policy period." Again, another lesson learned is to make sure there is clarity in your policy language and the client actually knows what type of policy it is purchasing.

Actuarial Standards

Another essential factor from the instant case is that proper actuarial projections are a key to any insurance company. Proper actuarial computations, pursuant to Actuarial Standards, need to be done by a reputable, nonbiased firm. Not only are actuaries needed to legitimize that you are an insurance company, but as an insurance company you need to price your policies properly within the market, with a profit motive. I mentioned previously that the Avrahamis' actuary was not able to clearly explain his pricing rationale.

According to the court, "… the work of an actuary must be reproducible and explainable to other actuaries." This means that if another actuary from a different firm reviews the work of another actuary, he or she should be able to see how the actuary justified the pricing contained within the actuarial report. The Avrahamis' actuary seemed to be using information from the higher rates that made the policies more expensive than necessary. Also, it is noted that the Avrahamis' actuary used what is called hazard group 2 instead of hazard group 4 when he determined the premiums. This would have reduced the rates, which means decreased premiums. Again, the captive was not acting like an insurance company, with proper actuarial foundation to determine its policy pricing.

The court questioned the actuary about this issue and found that the actuary did not have a coherent explanation in his professional judgment as to how he priced the policies. Again, a reputable actuary can thoroughly explain policy pricing, as actuaries are supposed to have sufficient information that would enable another actuary to review the actuarial report and determine the pricing methodology. This actuary seemed to have only worked on captive attorney Celia Clark's captives and that the attorney established the pricing needed, not the actuary.


The final area I want to address is claims. Claims were incurred but were not filed to be paid. First, there were not any claims filed until 2 months after the Internal Revenue Service (IRS) sent the Avrahamis documents about the audits. Suddenly, the Avrahamis started filing claims. It is unlikely to have no claims, and suddenly after notice of IRS inquiry, claims are filed and paid. Further, based on how these policies were written, just because a claim is filed, it does not necessarily mean that it should or will be paid. This brings us back to the underlying analysis of the court: Was the company operating like an insurance company? Did the company really go through standard insurance and claim procedures? Just because someone filed a claim, it does not mean that all claims are valid and should be paid. Who actually analyzed and examined the claims? Some of the claims were filed well after the policy period, which is not standard with insurance companies. So again, this captive was not operating like an insurance company.


I have reviewed a few topics in the lengthy court opinion and would be remiss in not addressing the captive's investments. According to the court, at around 2010, 65 percent of Feedback, which is the insurance company's assets that the Avrahamis owned, were tied up in long-term illiquid and partial unsecured loans to related parties, which were the Avrahamis' other entities; one of them listed is Belly Button (no jokes about the name, please). This is a very big concern for two reasons. One, it cites back that the loans initially were not approved by the insurance regulator. Also, in my opinion, if this was under a state regulator, most states require annual certified audits and this would have been found and disclosed under an audit. Under a certified audit, there would be required supporting loan documentation, which would have needed regulatory approval.

This much loan activity is an eye-opener for a state or other regulator. An insurance company needs liquidity to pay claims and would generally not have a disproportionate amount of funds illiquid in related party loans. An insurance company must have sufficient capital and liquidity to pay claims. Again, this captive was not acting like or as an insurance company. Thus, it was not eligible for insurance company benefits.


In conclusion, there were many areas in which the Avrahamis' captive was not acting like an insurance company. To receive the benefits of insurance company taxation, a captive must be an insurance company. The court held that the captive at issue was not an insurance company. Hence, it did not qualify for the income tax benefits or taxation that are accorded to an insurance company. This case provides many guidelines for what to be aware of and how a proper insurance company should function. Unfortunately for the parties at issue, the captive failed almost every test and the results were as one would anticipate, a large loss for the taxpayer.

See also "Great Court Ruling That Can Favor the Captive Industry," by Jeremy and Richard Colombik, and Captive.com Thought Leader videos from Jeremy Colombik.

Jeremy Colombik , Richard M. Colombik | August 30, 2017