Research and Development tax credits (R&D credits) range in both size and complexity. Similarly, the documentation available to substantiate the credit also varies. Historically, taxpayers could prove that their activities qualified under the Cohan rule (which permitted reasonable estimates) through testimony and estimates. This common-law rule allows the use of estimates when underlying documentation is unavailable. But the rule has limits, as the U.S. Tax Court pointed out in the latest case.

Beyond the limitations of the Cohan rule, the case, George v. Commissioner, T.C. Memo. 2026-10, offers several lessons for both the IRS and taxpayers. The opinion addresses business component classifications, technical uncertainty, statutory exclusions, and the need for documentation supporting the R&D credit and related testimony. In some ways, this case is a win for taxpayers, but it comes with a warning that should not be ignored.

Background

The poultry business is high-volume and low-margin, and the taxpayer, George’s of Missouri, Inc., is one of the largest fully integrated poultry processing companies in the United States. After being referred by their longtime CPA to an R&D consultant, the company claimed the research credit for three years through amended returns. In total, the company claimed more than $4.4M in credits, although much of it would be carried forward to future years.

The basis of the claim was research projects aimed at improving chicken health through feed additives, vaccines, probiotics, phytase, and genetic lines. Some of these projects spanned multiple years, but in all cases, the driving factor for the credit was the cost of feed and medication used to feed the chickens, or the supplies that were expanded during experimentation to create the healthier product. The company kept extensive records of the various feed recipes, dates of the significant events, weekly grower reports, and cost records.

The IRS effectively disallowed the credits, either through examination or by failing to process the refund claims and assessed accuracy-related penalties for two of the years in which the credits were used.

Cohan Rule

The Cohan rule was created by the courts to allow taxpayers to claim deductions even when documentation did not fully support the amount. First established in 1930 in the case of Cohan v. Commissioner, 39 F.2d 540, when Mr. Cohan, a theatrical manager and producer, tried to claim deductions related to his business travels and related entertainment. Unfortunately, Mr. Cohan was unable to produce documentation for these expenses, and the deductions were disallowed upon examination.

However, the Court was convinced that Mr. Cohan incurred business expenses and that it was unlikely he could have properly documented them. As a result, the Court allowed the deductions to be estimated. Thus, the Cohan rule was born, allowing taxpayers to estimate items on their tax returns when documentation is unavailable. This rule has been applied in numerous areas and is frequently discussed in R&D credit cases.

Business Component Classification

A central requirement of the R&D credit is that the research activities must be intended to develop a new or improved business component. A business component can be a product, process, computer software, technique, formula, or invention that is held for sale, lease, or license, or used in the taxpayer’s business.

The IRS has recently prevailed in the case of Grigsby v. United States, 5^th Circ. 2023, by challenging the business component classification: whether the business component is a process rather than a product. In that case, the taxpayer’s misclassification of the business component undermined the taxpayer’s case, preventing the IRS and the Court from properly reviewing how the research activities related to the stated business component.

In this case, the IRS made a similar argument. But the Court was unconvinced. The business components were found to be products, a healthier, better chicken, and not the process of raising the chickens, as the taxpayer initially claimed.

The Court’s analysis focused on whether the research aimed to increase efficiency or reduce costs (a process improvement) or to produce a higher-quality chicken (a product). This analysis led the Court to conclude that the research aimed to produce a chicken that was less susceptible to disease and more uniform in size and other characteristics.

Adaptations

The IRS argued that the business components were merely an adaptation of a previous product. An adaptation is one of the statutory exclusions from the R&D credit. The tax credit is meant to incentivize innovation and the creation of something new or a substantial improvement over an existing product. Adapting an existing product to meet another client’s specific needs does not satisfy the necessary level of innovation.

However, the Court disagreed that the taxpayer’s activities constituted adaptation. It noted that the term “adaptation” must be read in its ordinary sense and determined that it is only minor alterations of a design that are excluded from the credit. The fact that the taxpayers were trying to develop chicken specifications to satisfy customer requirements was not sufficient to exclude the research as an adaptation of an existing business component. First, the Court treated the new chickens as a new business component. Second, the changes used to improve the chicken were not minor modifications and were instead determined to be a “magnitude of change greater than minor alterations.” These two points led the Court to reject the IRS’s argument in favor of the taxpayer.

Documentation

Documentation is one of the cornerstones of the R&D credit. From demonstrating that a new product was developed to showing which activities qualify for the credit to illustrating the experimentation process, documentation supports each part of the tax credit. While the Cohan rule has been used to bridge the gap between real-world documentation created as part of the normal business operations and the ideal, it has limits. This case displays the limitations of the Cohan rule and the extent to which taxpayers should rely only on testimony to support the credit. In the Court’s words: “we do not apply the Cohan rule to estimate expenses paid or incurred if the taxpayer provides ‘no evidence at all that would permit an informed estimate’ of the deduction, basis, or other tax advantage.” And with this, making clear that gone are the days in which an R&D credit can be claimed on testimony alone.

While neither the statute nor the regulations require specific documentation, let alone contemporaneous documentation, and Congress has resisted adding such requirements, H.R. Conf. Rep. No. 106-478, at 132 (1999), the Court here follows a clear line of decisions that discount testimony and require documentation for substantiation. In the 7^th Circuit case of Little Sandy Coal Company, Inc. v. Commissioner, 7^th Circ., 2023, the Court noted that “shortcut estimates of experimentation-related activities will not suffice . . . something more, such as documentation of time spent on such activities, is necessary.” In Betz v. Commissioner, T.C. Memo, 2023-84, the Court did not dismiss the use of testimony but found the presented testimony unconvincing and, without additional documentation, denied the credit. And in Phoenix Design Group v. Commissioner, T.C. Memo, 2024-113, the Court gave preference to the documentation over testimony.

The Court fully disallowed any testimony without documentary evidence, excluded testimony that contradicted the documentation, and allowed only testimony that helped complete the picture it painted. Or stated another way, documentation does not have to provide all of the substantiation for the credit; testimony can assist, but testimony cannot be the only substantiation. In some instances, the taxpayer provided reports showing when specific medications were introduced into the feed at levels above those the company had historically used. Additionally, the company had medical reports of the medical exams (performed post-mortem) on the birds to show an increase in testing and data collection. The Court welcomed testimony bolstering this type of documentation.

Bottom Line

There is room for some celebration about the ruling, but there is much to be concerned about as well. The IRS is not giving up on finding ways to attack business component classifications, exclude activities under statutory restrictions, and limit the use of testimony and the Cohan rule. The taxpayer prevailed in minimizing a full disallowance, sustained some portion of the claimed credit (albeit less than 20%), and overturned the penalties. The table stakes have changed, and the standards of yesteryear are no more. Any R&D credit claimed this year should take into account the Court’s ruling in this case and other recent cases, including Little Sandy Coal Company, Betz v. Commissioner, and Phoenix Design Group v. Commissioner, among others. A study conducted using only testimony and estimates is risky and destined to fail if challenged by the IRS. Creating a proactive plan leveraging business record and building sustainable processes to document eligible activities concurrent with the activities is more necessary now than ever.