Swift decision, slow war: the captive insurance industry at risk

For years, experts in captive insurance warned that the IRS was preparing to renew its offensive against mid-sized to large captive insurers. The Fifth Circuit’s recent holding in Swift v. Commissioner opens the door wider than ever before. The IRS is armed with new weapons posing an existential threat to the captive insurance industry.

A troubling precedent

The case centres on Dr. Bernard Swift Jr., founder of Texas MedClinic, who created two small captives (so-called microcaptives) to finance business risk. The IRS denied deductions for premiums paid to these captives, arguing the arrangements failed to constitute true insurance. Both the Tax Court and the Fifth Circuit sided with the IRS, finding a lack of “risk distribution” and insufficient resemblance to “insurance in the commonly accepted sense”.

The real peril for the industry rests in how the courts interpreted “risk distribution”.

Dr. Swift’s captives financed risk in two ways: directly insuring medical malpractice risks of about 200 doctors and insuring various other risks via participating in risk pools. While the risk pooling structure raised several concerns, the Fifth Circuit’s analysis of directly insured medical malpractice risks handed the IRS new ammunition with which to expand its ongoing war on the captive insurance industry.

The Tax Court held that insuring 199 physicians did not provide enough risk distribution. The court concluded that this pool paled against broader, more diverse ones upheld in earlier cases because a pool of 199 doctors fails to trigger the law of large numbers. Accordingly, the lack of risk distribution invalidated the entire insurance transaction. Unfortunately, this constitutes a misunderstanding of the nature of risk.

The Fifth Circuit confused a means to an end with the end itself. In other words, the Fifth Circuit views the law of large numbers as necessary to assess whether a transaction has proper risk distribution because this phenomenon allows an actuary confidently to forecast the expected loss cost within a policy period. However, risk distribution is not “the law of large numbers” in sheep’s clothing. Rather, risk distribution is intertwined with risk reduction. The whole purpose of distributing risk is to ensure risks are minimised so the solvency of an insurance company is assured. There is no magic in the law of large numbers. Accordingly, there is more than one way to skin a cat.

The Fifth Circuit brushed aside arguments that a pool of 199 doctors results in a 93% reduction in relative risk. One of Swift’s experts pointed out that 199 doctors result in millions of interactions with patients per policy year, which is more than enough risk data points to constitute a distributed risk. The IRS countered that medical malpractice insurance policies are priced by multiplying the actuarial rate by the number of doctors. In other words, the IRS argued that the number of agents within a system is the proper measurement of risk distribution, not the total number of opportunities for loss.

And yet, the IRS’s experts conceded that full-time physicians’ medical malpractice insurance rates are higher than part-time physicians. The key difference between a full-time and part-time employee is the number of hours worked per pay period. The sole reason a full-time physician commands a higher insurance premium is that a full-time physician has more opportunities to make a mistake than a part-time physician.

Medical malpractice insurance pricing hinges on the total opportunities for loss (the total interactions with individual patients) not merely the total number of physicians. Unfortunately, the Fifth Circuit found the IRS’s experts more persuasive.

This case is the first meaningful discussion of risk distribution in the circuit courts in decades. Case law regarding the nature of risk distribution arises out of the case of Helvering v. Le Gierse. The Le Gierse holding, drafted shortly before the USA’s entry into the Second World War, rests on simplistic and outdated concepts. Modern actuarial science leverages advanced predictive models, generalised linear models and statistical metrics such as the coefficient of variation (CV), absolute risk reduction (ARR) and relative risk reduction (RRR) to quantify and price risk.

For example, in a 199-physician pool, the CV can empirically reveal how the variability of outcomes drops. Likewise, ARR and RRR, routine in the field of epidemiology, offer evidence that a well-structured captive functions as a true risk intervention. Yet the courts signalled that modern analytical tools “miss the point”. Risk pooling, under this judicial interpretation, requires not the achievement of statistical predictability, but the satisfaction of headcount-based quotas. The law of large numbers, in this way, becomes a rigid gatekeeper.

IRS leverage and the chilling effect on innovation

This legal rigidity gives the IRS a powerful advantage. By clinging to an outdated understanding of risk distribution, the IRS can challenge any captive that doesn’t look like a big commercial insurer. In Swift, the IRS even criticised the captives’ low loss ratios, an 8% loss ratio against an alleged industry standard of 66%, arguing this shortfall proved the insurance arrangement wasn’t genuine. Under this logic, only large corporations that can emulate the practices of major commercial insurers qualify for captive arrangements.

The market message is unmistakable: unless you look like an admitted carrier, your captive insurance arrangement is suspect. Accordingly, any single-parent captive not participating in a risk pool is at risk. This includes almost every single so-called microcaptive insurance company, as well as a significant chunk of the middle market’s captive insurance programmes. This will chill innovation and discourage smaller businesses from setting up sound, regulated self-insurance solutions.

Rethinking the legal standard

Now, the true fight will focus on single-parent captives and other 831(b) structures that do not participate in risk pools. Under the Swift framework, the IRS can challenge any captive that doesn’t clear this tight view of risk distribution. The industry must answer with strategic recalibration. It’s time to press the courts, forcibly and directly, to reconsider the Le Gierse regime in the light of modern actuarial science.

The message is simple: insurability stems from achieving predictability through rigorous statistical modelling and pricing, not from arbitrary pool sizes or headcounts. Captive insurers can and do achieve true risk distribution through proper exposure management, expert premium setting and modern risk modelling.

The legal doctrine must evolve accordingly. Otherwise, data-driven risk management might lose ground to legal doctrine stuck in the 1940s. That would be a setback not only for the insurance industry but for every business that seeks to manage its risks prudently and innovatively.

Matthew Queen (pictured) is the owner of The Queen Firm LLC, a law firm providing captive insurance consulting services, focused on providing advice and guidance to business owners. He is also an adviser to the 831(b) Institute. He can be contacted at: matthew@thequeenfirm.com

Did you get value from this story?  Sign up to our free daily newsletters and get stories like this sent straight to your inbox.