Insurance premiums go up and then they go up some more. The amounts can be substantial. This is particularly true for businesses that offer insurance to employees or that insure more types of risks.

And many business owners note that while they pay substantial insurance premiums, the insurance companies often do not have high payouts. This is because there be very few or even no claims submitted.

This is where captive insurance comes in. It is an arrangement where by a business or businesses get into the insurance business for their own risks. To oversimplify, they basically form entities and operate their own insurance companies.

This can make business sense. It can also result in a large tax deduction. That is where the IRS comes in. The IRS has a history of challenging captive insurance arrangements. In these cases the fundamental question is often whether the arrangement truly involves the essential insurance characteristics of risk-shifting and risk-distribution or is it just a tax play?

Given the size of the tax deductions at issue, the court decisions in the tax cases for captives have defined the industry. This brings us to the Swift v. Commissioner, No. 24-60270 (5th Cir. July 2025), case, which gets into whether a captive insurance arrangement has adequate risk distribution.

Facts & Procedural History

The taxpayer was the founder and sole proprietor of an urgent care center. It had 18 locations. He also owned two smaller medical entities that focused on sports rehabilitation and dermatology.

In 2004, the taxpayer explored creating captive insurance companies. He worked with a tax lawyer who specialized in forming and maintaining such entities.

The issue in this case involved the tax years 2012 through 2015. During this period, the taxpayer operated two captives incorporated in the Federation of Saint Christopher and Nevis. Each captive was owned by a trust benefiting one of the taxpayer’s children. The taxpayer and spouse served as trustees. During these four years, the medical practice paid $5.98 million in premiums to the captives. The taxpayer claimed these payments as business expense deductions.

The captives issued two main types of coverage. First, they provided medical malpractice “tail” policies. These policies covered claims related to professional services rendered before the policy period but reported afterward. The policies covered the practice’s physicians back to their start dates. Physicians acknowledged coverage annually and bore responsibility for deductibles and losses exceeding policy limits. Second, the captives issued various nonmedical coverage policies for administrative actions, business income, employment practices, litigation expenses, terrorism, and political violence.

The taxpayer’s attorney advised that the captives needed risk distribution. To achieve this, the captives participated in reinsurance pools. These pools consisted of approximately 100 captive insurance companies. The pools were designed to ensure that at least 30% of each captive’s premiums came from unrelated business through quota-share reinsurance arrangements.

As with most of the articles on our site, the problem started with an IRS audit. The IRS issued notices of deficiency that proposed to disallow the premium payment deductions and imposed 20% accuracy-related penalties. The deficiencies totaled over $2.4 million across the four tax years.

The taxpayer petitioned the U.S. Tax Court. The court sustained both the deficiencies and penalties. The case then went up on appeal, which is the subject of the court opinion we are covering here.

What Constitutes Insurance for Tax Purposes?

The starting point for considering this issue is, what exactly is insurance for tax purposes? It sounds simple, but it is not.

The tax code does not define “insurance.” So, when there is a question, the courts have to determine when premium payments are for “insurance” and qualify for business expense deductions under Section 162(a).

The U.S. Supreme Court established that insurance involves two fundamental elements. These elements are risk-shifting and risk-distribution. Risk-shifting occurs when the insured transfers the financial consequences of potential losses to the insurer. Risk-shifting analysis focuses on whether the insured has genuinely transferred the economic burden of potential losses to another party. This element is usually satisfied in captive insurance arrangements. The captive assumes contractual responsibility for covered claims.

The more challenging requirement typically involves risk-distribution. Risk-distribution spreads those transferred risks across a sufficiently large pool of independent risks. With this requirement, the IRS has consistently emphasized that these requirements must be met in substance–not merely in form. This is the nature of the IRS’s position when litigating these cases.

Why Risk Distribution Matters in Insurance

Risk distribution is the foundation of insurance economics. It distinguishes true insurance from mere self-insurance or family arrangements.

This concept relies on the statistical principle known as the law of large numbers. The law demonstrates something important. When a sufficiently large number of independent risks each have an annual loss probability of X percent, there’s an extraordinarily small likelihood that the actual loss percentage will deviate significantly from X percent.

I am a lawyer, not a mathematician, but I happened across an article by a mathematician explains this concept to laymen using a simple coin-flipping example. It goes like this. If you flip a coin ten times, you might get seven heads and three tails. This represents a significant deviation from the expected 50-50 outcome. However, if you flip that same coin one million times, the percentage of heads will almost certainly approximate 50 percent. Insurance operates on this same principle.

When an insurer covers thousands of independent risks, it can accurately predict total losses for the group. Individual losses remain unpredictable. This predictability allows the insurer to set appropriate premiums. The insurer can maintain adequate reserves and operate profitably while providing meaningful coverage to policyholders.

So back to risk distribution. Without sufficient risk distribution, an insurer faces a problem. A single catastrophic claim could exceed all collected premiums and reserves. This is because the law of large numbers only functions effectively when the underlying risks are truly independent. Risks that are correlated or concentrated in related entities create problems. A single event could trigger multiple claims simultaneously. This defeats the statistical predictability that makes insurance economically viable.

How Many Risks Are Enough for Distribution?

So that is the economics of it. But what does that mean from a practical standpoint? How many risks are enough?

The courts have struggled to establish an exact numerical threshold for adequate risk distribution. Instead, courts analyze each case based on its particular facts and circumstances. However, examination of successful captive insurance cases reveals patterns. These patterns show the scale necessary for meaningful risk distribution and the captive insurance industry has picked up on this.

For example, the U.S. Tax Court found adequate risk distribution in the Rent-A-Center, Inc. v. Commissioner case. In that case, the captive provided workers’ compensation, automobile, and general liability insurance for 14,000 to 19,000 employees. The captive also covered 7,000 to 8,000 vehicles and 2,000 to 3,000 stores. Similarly, in Securitas Holdings, Inc. v. Commissioner, the court accepted risk distribution where the captive covered 25 to 45 entities across more than 20 countries. That captive insured more than 200,000 employees and 2,000 vehicles.

These cases show that courts typically require exposure units numbering in the thousands or tens of thousands, not hundreds. The vast scale reflects the statistical reality. Meaningful risk distribution requires substantial numbers of independent risks. This achieves the predictive accuracy that characterizes genuine insurance.

Can Reinsurance Pools Create Risk Distribution?

The question then becomes, what constitutes the appropriate “exposure unit”? If you can define the unit narrowly, then maybe you can get higher numbers and satisfy the risk distribution requirement. Different measurement approaches can yield dramatically different risk counts.

So businesses may not have sufficient direct business to achieve risk distribution. When this happens, they may remedy this through participation in reinsurance pools. These arrangements allow multiple businesses to transfer portions of their risks to a common pool. The business formats its captive and the captives simultaneously assume quota-share responsibility for the pool’s blended liability.

Conceptually, reinsurance can transform a captive’s limited, related risks. It can create participation in a much larger, diversified risk pool. If properly structured, a captive that insures only its parent company’s risks might achieve meaningful risk distribution. This happens by trading those concentrated exposures for a proportional share of the pool’s diverse, unrelated risks.

However, the success of this depends entirely on something specific. The reinsurance transactions must constitute genuine insurance arrangements. They cannot be circular movements of funds designed primarily to create favorable tax characterization. These are the cases that the IRS pushes to litigation. And the courts then examine the structures with particular scrutiny as the cases they see, now, are usually only the ones that are closer calls. The courts analyze whether the structures involve real risk transfer and arm’s-length transactions.

This brings us to the Harper Group v. Commissioner case. In that case, the court established important precedent that the captive insurance industry uses. It found adequate risk distribution where 29 to 33 percent of the captive’s business involved insuring unrelated entities. This created an informal “30 percent rule.” Many practitioners have adopted this as a target threshold for risk distribution for captive insurance arrangements.

What Made This Reinsurance Pool Arrangement Fail?

This brings us back to this case. In this case, the appeals court analyzed the reinsurance pools and concluded that they did not achieve meaningful risk distribution. The court examined multiple factors in determining whether the pools constituted genuine insurance arrangements or merely paper transactions designed to create favorable tax treatment.

The court focused on the circular flow of funds between the captives and the reinsurance pools. The court noted that the captives paid premiums to the pools for reinsurance coverage, but the captives simultaneously received nearly identical amounts back as premiums for their quota-share participation in the pools’ blended risks. The amounts received ranged from 94.98% to 99.59% of the amounts paid across the four tax years. The court noted this as circular arrangement.

The court also considered the pools’ capitalization. They did not have the financial ability to function as genuine insurers as they were underfunded. The court noted that the pools appeared “thinly capitalized.” The court concluded that the pools would struggle to pay meaningful claims. This led the court to question whether any reasonable business would enter such contracts absent tax motivations.

The court also questioned the premium-setting methodology. The parties did not use actuarial analysis to determine appropriate pricing based on covered risks. Instead, the evidence showed that the advisor and actuary “were simply manipulating numbers to design a system where 30% of total premiums would be allocated to reinsurance before being retroceded back.” The pools charged uniform percentages to all participating captives regardless of their individual risk profiles. The pools also allowed captives to choose their own reinsurance percentages for certain coverage types to achieve desired overall allocation targets.

The court noted that the arrangements also included various features designed to discourage actual use of the reinsurance coverage. These features included requiring captives to pay substantial retained limits before making claims. The pools also had authority to exclude members who submitted excessive claims. The court said that these provisions suggested that the arrangements were not intended to function as genuine insurance.

Ultimately, the appeals court sustained the tax court’s opinion. The result was a loss of the business deduction for the insurance premiums.

The Takeaway

This case shows that captive insurance arrangements have to have to be insurance. Participation in reinsurance pools does not always mean there is risk distribution. This is particularly true when the pools operate as circular fund flows rather than genuine insurance arrangements.

Businesses with or considering captive insurance structures need to consider the scale of operations for achieving adequate risk distribution and assess whether their risk profiles involve sufficient independent exposures to support genuine insurance economics. They also need genuine risk transfer rather than circular transactions, adequately capitalized pools, and they should charge actuarially appropriate premiums and operate with meaningful independence from participants.

This is an area where tax planning is needed to try to avoid the type of result in this case.

Watch Our Free On-Demand Webinar

In 40 minutes, we’ll teach you how to survive an IRS audit.

We’ll explain how the IRS conducts audits and how to manage and close the audit.