Too hot to deduct

As a parent of young children, I've learned that bedtime stories are especially good at simplifying complex topics and I often use the same technique in my work. 

If you think about it, executive compensation at a private C corporation is a Goldilocks problem. Pay a shareholder-employee too much, and the IRS can recast the excess compensation as a non-deductible dividend. Pay too little, and the company forfeits a legitimate deduction and leaves taxable income sitting inside the corporation. Ultimately, we want an amount that's "just right" for the services actually performed, along with the ability to prove it. That second part — proof — is where most closely held companies come up short.

Take the case of Clary Hood, Inc. v. Commissioner. Marion "Buddy" Hood built Clary Hood into a successful land excavation and grading company. After the company generated $44 million in revenue in 2015, it paid Hood a bonus of roughly $5 million, followed by another $5 million bonus in 2016 when the company generated $69 million in revenue. There was no employment agreement and no formal board process. Hood and his wife were the company's only two directors, and they set his compensation themselves. It's worth noting they had a clear view of the company's annual profits for the year before they finalized Hood's bonus number.

The IRS argued that Hood's bonus was a dividend in disguise like the fictional grandmother with suspiciously large teeth in the children's story. The outcome was more nuanced than a simple loss. The Tax Court agreed that part of the increase was legitimate since Hood had genuinely been underpaid in prior years and the compensation "catch up" was recognized justification for the outsize bonus. But The Tax Court's decision required the company to prove, rather than simply assert, that the specific amount it paid Hood was reasonable. Ultimately the Court allowed less than the company claimed. Again, the facts supported a raise, but the documentation did not. For a closely held company, that gap is where the deduction is won or lost, and here a meaningful piece of the bonus ran too hot to deduct fully.

The mechanics of failing one test two ways

Treasury Regulation 1.162-7 sets out a two-part standard for deducting compensation. First, the pay must be purely for services, and second, the pay must be "reasonable" based on what comparable businesses pay for similar services under similar circumstances. In practice, courts tend to collapse the first question into the second. If the amount looks reasonable, intent rarely gets separately litigated. If the amount looks inflated, the argument hinges on what is "reasonable."

Cases of recharacterization follow typical patterns such as: (a) no arm's-length bargaining overcompensation; (b) compensation tracking ownership percentage more closely than the job performed; (c) no company history of ever paying a dividend; and (d) profits getting fully absorbed by "salary," thus leaving little taxable income. Clary Hood checked several of these boxes. 

A compensation benchmarking analysis can help fortify the deduction.  

What defeats the recharacterization argument

By benchmarking executive compensation to multiple independent sources and by drawing from landmark court cases, deductibility can be supported and documented. In addition to market data, we have precedented approaches to defining "reasonable."

The Multifactor Test, drawn from Mayson Manufacturing Co. v. Commissioner and refined in Elliotts, Inc. v. Commissioner, documents genuine compensatory intent by weighing the executive's actual duties, qualifications and unique contributions, and tying them to the role rather than to the ownership percentage.

The Independent Investor Test, from Exacto Spring Corp. v. Commissioner, answers the "profits siphoned out as salary" question with a number instead of by an assertion. If the company's return on equity remains at a level that would satisfy an outside investor, even after paying the compensation in question, that's strong evidence that the payment reflects real services rather than a disguised distribution.

It's important to note that a compensation benchmarking analysis is part of corporate governance, not a replacement for it. 

Putting it together

Consider a hypothetical founder-CEO of a $45 million metal fabrication company. Let's assume he had been paid the same $390,000 for many years despite absorbing CFO-level duties, carrying personal guarantees and driving more than half of the company's revenue through direct customer relationships. The board practiced disciplined corporate governance by commissioning an independent compensation benchmarking analysis and by setting compensation policies prior to year-end. That analysis placed the market range for the CEO's role between $1.75 million and $2.25 million, with a Multifactor Test supporting a conclusion near the top of that range. Additionally, the Independent Investor Test was upheld with strong equity returns despite the elevated compensation expense. Adjusting the CEO's salary to $2 million added over $1.6 million in deductible compensation annually, resulting in reduced corporate tax of $330,000 annually.

Hypothetical founder-CEO compensation analysis

Source: Withum, 2026

The lesson runs in both directions. A history of under-compensation is not, by itself, a defense any more than a large bonus is automatically a dividend. What separates a defensible position from an audit adjustment is a contemporaneous, well-supported number that is set before the IRS or a court sets it. Reasonable compensation is not a slogan; it's a conclusion that a company should be prepared to document. Doing the work in advance keeps a company's deduction and footing intact when the question is finally asked.