Kadau v. Commissioner — Captive Insurance that Wasn’t

The recent memorandum decision in Kadau v. Commissioner (T.C. Memo. 2025-81) provides a clear example of how the Tax Court currently analyzes micro-captive arrangements. Importantly, the opinion does not create new law; rather, it applies the well-established captive-insurance framework developed in Avrahami, Reserve Mechanical, Caylor, and Syzygy. We have previously distinguished how our clients do captive insurance from those prior cases, and we will do the same now with respect to Kadau.

The following analysis outlines the relevant tests and factors applied by the court, noting whether each test is a balancing test or a binary threshold, the relevant factors involved, and how the facts in Kadau map across those factors.

I. The Four-Part Federal Definition of “Insurance”

Insurance premiums are only deductible under federal income tax law if the arrangement qualifies as genuine insurance for federal income tax purposes. Courts ask four questions when determining whether an arrangement constitutes “insurance” for this purpose:

1. Does it involve insurable risks?

2. Does it really shift that risk from the insured to the insurer?

3. Does the insurer distribute that risk among a sufficient number of risk-exposure units?

4. Does the arrangement constitute insurance in the commonly accepted sense?

In micro-captive cases, the disputes almost always center on elements (3) and (4). That is precisely what occurred in Kadau.

II. Element 3: Risk Distribution — A Binary Threshold Test

Risk distribution is not a balancing test. It is a binary determination: either the insurer distributes risk sufficiently among a sufficient number of insureds and/or exposure units to resemble traditional insurance pools, or it does not.

In Kadau, the taxpayer conceded that the captive did not directly insure a sufficient number of exposure units to achieve risk distribution. The question was therefore whether the taxpayer indirectly achieved sufficient risk distribution via its participation in a risk-distribution pool. The Court evaluated the pool using the criteria articulated in Reserve Mechanical:

• whether the pool was a functioning insurance company;

• whether unrelated risks were genuinely pooled;

• whether premiums were actuarially determined;

• whether transactions were conducted at arm’s length;

• whether real risk transfer and distribution occurred.

The Court held that every one of these material factors failed.

1. Whether the Pool Was a Functioning Insurance Company

The Court held that the risk pool was not a functioning insurance company on several grounds.

First, it found that there was a “circular flow of funds” with respect to some of the premiums. In the words of the Court, the premiums “left [the insured], went through the whole apparatus of [the pool], and came back to [the captive], interrupted only by promoter fees and minimal claims.” The Court viewed this as strong evidence that meaningful risk never exited the economic control of the captive owner.

The risk-pool facts were pretty damning. The pool only handled excess-layer coverages for participating captives, including Kadau’s. Money for this excess layer flowed from the insured to the pool and then to the captive. The actual claims submitted to the pool under the excess-layer policies across the entire risk pool’s book of business totaled only $17,611 over the years at issue, resulting in a loss ratio of just 0.035% against approximately $50 million in total premiums received by the pool. No claims were paid under the excess layer for Kadau’s captive, as all payouts occurred under the primary-layer policies which were directly written by the captive.

By contrast, our clients today typically pay all premiums to an unrelated fronting carrier (the pool), which then reinsures a portion of its risk back to the client’s captive insurance company. But unlike in Kadau, the risk pool assumes 51% of the risk on every claim from first dollar to last. Consequently, that pool experiences hundreds of unrelated claims per year totaling millions of dollars, a far cry from the “minuscule” claims (by number and dollar amount) incurred by the pool in Kadau. Our clients’ premiums leave the insured, go through the whole independent apparatus of the risk pool, and come back to the captive interrupted by—not “minuscule” claims—but hundreds of unrelated third-party claims totaling millions of dollars, as well as fees. There is simply no reasonable and honest way to characterize that arrangement as a “circular flow of funds.”

Second, the Court found that the risk pool in Kadau had minimal capitalization and no meaningful insurance operations beyond mere paperwork. It based this conclusion on the following:

(a) Lack of Substantial Assets or Independent Financial Substance

RMC Property retained little to no premiums it received to pay claims, instead apparently sending them quickly to the relatively few captives (10 to 20) as reinsurance premiums, less a 2% to 2.5% fee. Consequently, it held no meaningful reserves to pay potential claims. Its role was essentially pass-through. This created a structure where the pool did not accumulate or manage significant capital to cover pooled risks, as would be expected in a bona fide insurance pool.

By contrast, our clients’ risk-pool arrangement covers nearly 200 captives, or more than ten times the number covered in Kadau. More importantly, it retains tens of millions of dollars on its balance sheet as a contingency reserve to pay claims, segregated in a separate trust to prevent any possibility of malfeasance, passing funds on to the reinsuring captives only after its potential claims liabilities are better known.

(b) Non-Arm’s-Length Arrangement

The minuscule risk-pool capitalization in Kadau was viewed as insufficient to support true risk-bearing, suggesting the setup was not at arm’s length but rather designed to minimize actual financial exposure. The Court noted that the reinsurance agreements capped the captives’ obligations at the amount of reinsurance premiums paid, effectively limiting any real capital commitment from the pool itself.

Once again, our clients’ risk pool retains tens of millions of dollars to cover potential claims, and each reinsuring captive is on the hook for its quota-share of any pooled losses, regardless of how much it received in reinsurance premiums.

(c) Limited to Paperwork and Fee Collection

Operations were confined to issuing “reimbursement policies” (prepared by affiliated RMC Consultants) and collecting/transferring funds. There was no evidence of independent underwriting, actuarial analysis, risk assessment, or ongoing management of pooled risks—functions typically performed by a genuine insurer.

By comparison, our clients undergo a thorough, written risk assessment before participating in the pool. Each risk is separately underwritten, and each policy is supported by actuarial analysis and an actuarial opinion. The pool’s risks are actively managed, with some clients with poor claims histories being expelled by the pool in subsequent years.

(d) Perfunctory Structure

The pool involved only a small number of similar captives (typically 10–20 per year), with quotas not based on sound actuarial principles but rather arbitrary allocations. Premiums were not adjusted for actual experience or losses, and the setup functioned more as a fee-based administrative conduit than an operational insurer. The Court described this as “not substantial insurance operations,” emphasizing the paperwork-heavy nature without real-world application.

We think the Court got some of this analysis wrong. But regardless, our clients’ arrangement is quite different. Their pool consists of hundreds of captives. Each policy is actuarially priced and adjusted each year based on claims experience, evolving insurance-market conditions, and other relevant factors. The pool obtains an actuarial opinion confirming that the premiums it is receiving correspond actuarially to the risks it is assuming. Claims are actively adjusted and regularly paid. In short, it engages in “substantial insurance operations” as any fair insurance professional or executive would define that term.

These conclusions aligned with precedents like Avrahami and Syzygy, where similar micro-captive pools were invalidated for failing to achieve meaningful risk distribution. By contrast, CIC Services’ pool structure is a fronted program with a 51% pool / 49% member captive risk-sharing arrangement. It is also “first dollar” and not a layered structure.

2. Whether Unrelated Risks Were Genuinely Pooled

Because the pool was not an insurance company for the reasons noted above, and because it lacked economic substance, the Court concluded that the supposed sharing of risks among unrelated insureds was illusory. The pool in Kadau retained no assets to back claims, did not, in the Court’s view, retain or meaningfully manage risks, and therefore could not be said to aggregate and distribute those risks. The “unrelated risk” required was defeated not by a lack of unrelated parties per se but because the structure prevented true pooling of those risks.

We believe the Court got some of this analysis quite wrong, but regardless, our clients’ structure is not meaningfully similar. As previously noted, our clients’ fronting carrier (risk pool) retains tens of millions in assets to back claims, pays hundreds of claims each year totaling millions of dollars, shares risk with the reinsuring captives from first dollar to last dollar (rather than the layered approach in Kadau that ensured the pool would ever only have minuscule claims), and independent actuaries validate both the premium calculation and the premium allocation between the pool and reinsuring captives each year.

3. Whether Premiums Were Actuarially Determined

In Kadau, the actuarial work was very minimal and formulaic, failing to account for the unique risks of each insured and policy. Marn Rivelle, the primary actuary, conducted the initial actuarial study and feasibility study in July 2012. This was based solely on a questionnaire completed by Curtis K. Kadau and a single phone discussion about the business. Rivelle did not receive or review any additional documents, such as the insured’s existing commercial insurance policies, financial statements, or even its loss history. Rivelle also prepared a 2013 “Split Report” to retroactively justify a change in the premium split from 50–50 (primary layer to the captive, excess layer to the pool) to 40–60. No further actuarial analysis was done for the 2016–2017 shift to a 30–70 split. For the 2017–2018 general liability policy, premiums were simply copied from the insured’s prior commercial carrier (Amerisure) by RMC affiliate Michael Bleiweis, who was not an actuary, without any validation or revaluation.

Premiums were derived from limited input (questionnaire and one call), with no supporting actuarial calculations provided in the study. They remained unchanged over six full tax years (2012–2017) despite evolving business risks, which the Court noted was inconsistent with standard actuarial practices that require periodic reviews and adjustments based on actual experience, loss data, or market changes. The risk splits between the pool and reinsuring captive (e.g., 50–50 initially) were arbitrary and not tied to comprehensive actuarial modeling; later changes were made to encourage more claims in the excess layer without holistic reassessment. Given these obvious deficiencies, the Court concluded that premiums were designed to fit within section 831(b) limits for small insurance companies (capped at $1.2 million initially, later $2.2 million), appearing more like tax-driven targets than risk-based figures.

These actuarial deficiencies, absent in our clients’ risk-pool structure, were a major factor in the Court’s holding that the arrangement did not qualify as insurance for federal tax purposes. By contrast, our clients complete new insurance applications each and every year. Via those applications or otherwise, they inform the underwriter and independent actuaries of their actual loss histories, other coverages, and any material changes to their business model or operations. The actuaries review these documents, and more, to actuarially determine and revise premiums each year, and they support their actuarial determinations with extensive written analysis and opinions.

4. Whether Transactions Were Conducted at Arm’s Length

In examining whether the transactions in question were conducted at arm’s length, and concluding they were not, the Court again noted the circular flow of funds, something it deemed uncommon in arm’s-length insurance arrangements. It also noted that premiums for the excess layer were excessive—2.5–3.5 times higher per dollar of coverage than market rates—and not adjusted over time based on actual experience. Arbitrary splits (e.g., 50–50 initially, later 40–60 or 30–70) were implemented without proper justification, suggesting pricing was driven by tax optimization rather than legitimate risk assessment. Premiums were not actuarially calculated and updated each year.

The excess-layer policies issued by RMC Property included restrictive provisions, such as requiring the primary insurer to admit full liability before any payout, and omitted standard features like a duty-to-defend clause. These nonstandard, limiting terms made the policies unlikely to result in claims and were seen as inconsistent with legitimate arm’s-length insurance products.

For the above reasons and others, the Court concluded that transactions with the risk pool were not conducted at arm’s length. But again, our clients’ arrangements are entirely dissimilar to Kadau and meaningfully more arm’s length. With our clients, there is no circular flow of funds. Premiums are determined by independent actuaries each year with great due diligence and are adjusted each year based on loss histories and other relevant factors. The allocation of premiums between the risk pool and the reinsuring captives is actuarially determined each year and certified in writing. The pool and captives share risk 51/49 from first dollar of claim to last dollar of claim. Hundreds of claims totaling millions of dollars are paid by the pool each year, with each captive bearing its quota-share portion. Policy provisions are more standard. In short, essentially none of the factors noted by the Court in Kadau as indicating a lack of arm’s length apply to our clients.

5. Whether Real Risk Transfer and Distribution Occurred

In Kadau, the risk pool did not operate with normal underwriting, claims administration, or independent oversight. In our clients’ pool arrangement, it does.

Conclusion on Element 3:

For all the above reasons, the Court concluded that the risk pool in Kadau did not distribute risk. And because risk distribution is an essential element of any “real” insurance arrangement under the federal tax rules, that alone was sufficient to condemn the transaction and secure the IRS’s victory. However, the Court did proceed to analyze another major factor, which Kadau also failed—whether the arrangement constituted insurance in the “commonly accepted sense.”

III. Element 4: Insurance in the Commonly Accepted Sense — A Balancing Test

Unlike risk distribution, the inquiry into whether an arrangement qualifies as “insurance in the commonly accepted sense” is a holistic balancing test. No single factor is dispositive. The Court examines the totality of the circumstances to determine whether the captive itself operates like a bona fide insurer. The factors typically considered include:

1. the captive’s organization, operation, and regulatory posture;

2. validity and enforceability of policies;

3. reasonableness of premiums;

4. existence of a genuine claims-handling process;

5. adequacy and propriety of reserves;

6. investment strategy and asset composition;

7. consistency with industry practices.

Here is how the Court in Kadau analyzed those factors.

1. Organization, Operation, and Regulation

(a) Organization of the Captive

In considering whether the Kadau captive was organized as a proper insurance company, the Court emphasized that the captive was incorporated in Nevis on October 3, 2012, as a captive insurance company with an initial capitalization of $50,000 (50,000 shares issued to Curtis K. Kadau, the sole shareholder). It had two directors—Kadau and William Jackson Arnold (the latter appointed solely to meet Nevis regulatory requirements and with no material participation)—and officers including Kadau as chairperson, president, and secretary, and Arnold as treasurer. Management was handled by RMC Consultants. The Court noted this minimal structure met basic formalities but lacked independence, as Kadau maintained full control.

We believe the Court’s analysis on this point to be severely flawed and unlikely to be upheld on appeal. The simple truth is that captives in general, and section 831(b) electing ones in particular, were never meant to operate like large commercial insurers, and requiring them to do so makes them economically unviable. For this reason, almost no captive insurance companies have actively working employees. Officers of captives are very typically also officers of the insured. Captive management firms typically independently provide management and oversight as required by state regulators.

The initial actuarial feasibility study by Marn Rivelle (hired by RMC Group), performed when the captive was organized, was based only on a questionnaire from Kadau and one phone call, without reviewing the insured’s financial statements, loss history, or existing commercial policies. This was cited as evidence of superficial organization, not aligned with standard insurance practices. As previously mentioned, our clients organize a captive only after going through a robust risk-assessment process and diligent underwriting and actuarial analysis, all three of which consider the insured’s financial position, loss history, industry profile, risk profile, and existing policies.

(b) Operation of the Captive

The Court assessed operational aspects to determine if the captive functioned like a genuine insurer, focusing on day-to-day activities and economic substance.

Policies Issued and Premiums: The captive issued primary-layer policies starting November 1, 2012, covering risks like excess business interruption, collection risk, and directors and officers liability, all on a claims-made, reimbursement-only basis with deductibles and limits. Premiums were split with companion excess-layer policies from the risk pool (initially 50–50, later adjusted without any documented basis to 40–60 and then 30–70). The Court found policies were nonstandard (e.g., no duty-to-defend clause, limited to corporate officers, no cancellation provisions except for general liability), with premiums excessive, static, and not actuarially sound—evidence of operations prioritizing tax benefits over risk management. None of these things are consistent with standard insurance practices nor with how our clients’ captives are operated.

Reinsurance and Fund Flows: Under a reinsurance agreement, the captive accepted ceded liabilities from the pool, with premiums flowing from the insured to the pool and back to the captive (minus only 2–2.5% fees). This circular flow was highlighted as lacking true operational independence, resembling a pass-through rather than active insurance. Though we believe the Court’s analysis on this point to be flawed, our clients’ structure is quite different, with money being withheld by the pool until claims become better understood and then being reduced not only by fees but by millions of dollars of claims from hundreds of different risk exposures.

Claims Handling: A claims committee (Kadau and Arnold) met three times for a total of 53 minutes, with no independent adjuster. Claims were paid only under the collection-risk policy in 2016 ($112,089 and $73,460 after deductibles) and 2017 ($74,761 out of $162,189), split into multiple invoices without thorough review. No claims occurred in the first four years or under excess layers. The Court saw this minimal, Kadau-dominated process as evidence of perfunctory operations, not genuine claims adjudication. By contrast, our clients’ captives pay hundreds of claims per year via their participation in the risk pool, and all claims are independently reviewed by professional claims adjusters and paid only if the independent adjuster approves.

Investments: Assets included cash, an annuity, and a $6 million whole life insurance policy on Kadau (over half the portfolio by 2017, selected to cover personal mortgages). The policy was canceled post-audit and replaced with preferred stocks. The Court criticized the lack of diversification, due diligence, or risk-hedging alignment, viewing investments as personal rather than operational. Though some of our clients may use life insurance as an investment vehicle within the captive insurance company, those decisions are made for the benefit of the captive (rather than the business owner), after significant due diligence, with a specific eye toward risk hedging, and for the same compelling reasons that community banks often invest large portions of their Tier 1 capital in bank-owned life insurance and Fortune 1000 companies invest large portions of their nonqualified deferred compensation plans in corporate-owned life insurance.

(c) Regulation of the Captive

The Court examined regulatory compliance to assess if the captive was regulated as an insurance company.

Licensing and Capitalization: It held a Nevis captive insurance license, approved for specific coverage lines with anticipated premiums under $1.2 million. Minimum capitalization was met ($50,000 initial, totaling $350,000 including a policy purchase). However, the Court deemed this “marginal” and insufficient for true risk-bearing, noting no employees and reliance on Kadau/Arnold for operations. By contrast, our clients domicile their captives domestically rather than internationally and, consistent with domestic standards, capitalize their captives in a manner consistent with the risks borne by the captive and as approved by actuaries and regulators, most often with $250,000 or more of highly liquid assets.

Ongoing Compliance and Oversight: The Kadau captive had no ongoing due diligence on premiums, reinsurance, or risks; actuarial reviews were limited to the 2012 study without updates. Separate accounts for claims were not maintained, and investments were not linked to the liquidity and other needs of the captive insurance company. The Court found regulatory control limited to Nevis licensing, with operations lacking arm’s-length standards and exhibiting circular fund flows—evidence that regulation was formal but not substantive. As previously noted, none of our clients’ captives suffer from these deficiencies.

These factors collectively led the Court to conclude that the captive was not organized, operated, or regulated as a bona fide insurance company, but rather as a tax-avoidance vehicle, thus failing the first prong of the “insurance in the commonly accepted sense” balancing test.

2. Valid and Binding Policies

The Court evaluated several aspects of the policies’ form, substance, and practical application to assess whether policies issued by the captive were valid and binding:

Policy Structure and Terms: The policies were scrutinized for essential elements of a binding insurance contract, including identification of the insured (Surface Engineering), an effective coverage period (typically November 1 to November 1 annually), specified risks covered (e.g., excess business interruption, collection risk, directors and officers liability), stated premiums, deductibles, and policy limits, and signatures from authorized representatives (e.g., issued by RMC Consultants on behalf of Risk & Asset). However, the Court highlighted deficiencies: the policies were reimbursement-only (requiring the insured to pay losses first and seek repayment), lacked standard provisions like a duty-to-defend clause in the directors and officers liability coverage (leaving the insured vulnerable during litigation), had narrow or undefined terms (e.g., “Officer” and “Director” were not clearly defined, potentially limiting coverage to the corporation rather than individuals), included few exclusions, and omitted cancellation options except for the general liability policy. Some policies also had internal inconsistencies, such as conflicting binding clauses. Though we believe the Court gets part of this analysis wrong, our clients’ structures lack similar defects.

Delivery and Timing: The Court considered whether the policies were delivered in a timely manner to create binding obligations. Many policies were issued on a claims-made basis with retroactive effective dates, but some were not delivered until after the coverage period had begun or even after potential claims arose, which the Court viewed as unusual and indicative of a lack of genuine intent to bind parties in advance. Once again, the Court simply gets the analysis here wrong. Though retroactive effective dates are not common, it is very common for even ordinary commercial insurance policies to be formally delivered after the policy period has already commenced. Regardless, these are not defects common to our clients’ captive arrangements.

Comparison to Commercial Insurance Standards: The policies were compared to arm’s-length commercial insurance products. The Court found them nonstandard and overly restrictive, with terms that made payouts unlikely (e.g., excess-layer policies required the primary insurer to admit full liability first). This suggested the policies were not designed for real risk transfer but for tax benefits. Though it is common for captive insurance policies to have nonstandard terms, it is not common in the insurance industry, or among our clients’ policies, for those terms to be structured so as to make payouts unlikely. If anything, the opposite is generally more true both in the captive insurance industry in general and among our clients.

As for whether the policies in question were valid and binding, the Court deemed the evidence “mixed” but ultimately concluded that it weighed slightly against the petitioners, contributing to the overall ruling that the arrangement did not constitute insurance. Had the Court properly applied the relevant tests, we believe it would have reached the opposite conclusion, but regardless, our clients’ policies do not suffer the same defects.

3. Premium Reasonableness

This one was easy for the Court. It analyzed the issue by assessing whether premiums were actuarially sound, comparable to market rates, and set at arm’s length. As previously noted, they obviously were not. By contrast, our clients’ premiums are determined by independent actuaries each year after significant data gathering and due diligence.

4. Claims Handling

A two-person committee (Kadau and Arnold) met three times for 53 minutes total to approve claims. No independent adjuster was used. The claims process was internal and perfunctory, with quick approvals and minimal scrutiny. Even so, the Court concluded that payment of claims weighed slightly in petitioners’ favor, but the conflicted, superficial process undermined legitimacy. As previously noted, all our clients’ claims are paid only after thorough review and approval by an independent, professional claims adjuster.

5. Investment Portfolio

The Court concluded that the captive’s investments, primarily a cash-value life insurance policy, were chosen with insufficient due diligence and primarily to benefit the personal needs of Mr. Kadau, such as paying off his personal mortgage in the event of his death, rather than the business needs of the insurance company. This factor therefore weighed against the taxpayer.

Some commentators have suggested that the Court’s issue was with life insurance itself. This position is unnuanced and naïve. Corporations like banks and even Fortune 1000 companies commonly, after due diligence, purchase life insurance on their officers as investments for the benefit of the company, using the cash-surrender value and/or death benefit to fund nonqualified deferred compensation plans, store Tier 1 capital, or otherwise. There is nothing innately improper about a corporation, including a captive insurance company, doing the same. In fact, a properly structured life insurance policy can be the ideal investment vehicle for some captives, providing far more mid- and long-term liquidity than many other options available. But it is essential that the insurance be purchased only after sufficient due diligence, for the benefit of the captive itself and not primarily for the benefit of its owner, and that the policy be owned, funded, and structured accordingly. And finally, that all of that be documented.

In short, the Court implicitly makes clear that the presence of life insurance in a captive is not inherently disqualifying in the “commonly accepted sense” analysis. Real commercial insurance companies sometimes hold life policies or structured products too. It was the context and magnitude, combined with the broader operational deficiencies, that transformed the life insurance component into a negative factor. On a different factual matrix—one with real risk distribution, market-based premiums, credible underwriting, legitimate claims activity, and proper investment due diligence—the mere presence of a life insurance asset would not be fatal.

Conclusion on Element 4: After weighing the above factors and facts, which differ markedly from our clients’ facts, the Court concluded that the Kadau captive was not offering insurance in the “commonly accepted sense,” and so premiums paid to it were not deductible for this reason also.

IV. Overall Conclusion

The benefits of a captive insurance company can be extraordinary. But anyone forming one who hopes to deduct the premiums, or claim the benefits of the section 831(b) election, must observe the necessary formalities noted above. With proper professional assistance, that should not be difficult. Alas, too many taxpayers either do not get the proper assistance or ignore it.