Why the Kyocera Case Matters for Tax Teams
The Kyocera AVX Components Corporation R&D court case is one of the most important cautionary tales in recent memory for companies pursuing the R&D tax credit. It showed that even a large, legitimate company with substantial research activity can lose an R&D credit claim if the process behind it does not meet today’s IRS expectations.
In this post, we will break down the key mistakes in the Kyocera case and offer practical steps your tax team can take to create a defensible R&D credit process.
What Went Wrong in Kyocera
Kyocera originally filed a modest R&D credit claim, then filed an amended return increasing the credit to $1.7 million after engaging PwC to perform a multi-year study.
But when the IRS audited the amended claim, several weaknesses became apparent:
- The study relied almost entirely on SME interviews, with no time tracking or contemporaneous project documentation.
- PwC did not retain original survey responses or interview questions.
- The company did not preserve supply or contract research invoices.
- Contemporaneous technical documentation was not produced during discovery.
- Business components were not well defined or substantiated.
- Upper-level management and other high-risk job titles were included without sufficient proof of their qualifying activities.
Ultimately, the court granted summary judgment for the government, and the credit was lost.
How to Avoid These Mistakes
The Kyocera case illustrates several process failures that any company can learn from. Here are six key areas to focus on.
1. Build a Documentation-First Process
One of the core problems in Kyocera was overreliance on SME interviews without supporting documentation. This is no longer sufficient to sustain an R&D credit.
Instead:
- Start by building your process around the documentation your teams already generate in the course of research.
- Collect project artifacts such as design documents, technical drawings, test reports, and iteration records.
- Preserve relevant communications and project tracking data.
SME interviews should be used to complement this documentation, not replace it.
For strategies on improving your documentation process, see Optimizing Your R&D Tax Credit Process Using Today’s Technology.
2. Select and Prepare SMEs Carefully
Not every SME is equally well positioned to contribute to your R&D credit study.
Best practices include:
- Select SMEs who can speak from first-hand knowledge of qualified activities.
- Avoid asking SMEs to represent activities outside their sphere of influence.
- Provide education on what the IRS expects and how to describe activities clearly.
- Encourage SMEs to reference or point to contemporaneous project documentation where possible.
See How to Train SMEs for R&D Tax Documentation for guidance on building a strong SME process.
3. Go Beyond Percentages
In Kyocera, one of the IRS’s core criticisms was that employee time was represented as percentages, with little or no activity-level proof to support those estimates.
So, the IRS wants to see:
- What specific work was performed.
- How that work involved a process of experimentation.
- How it aligns with the four-part test.
Encourage SMEs and project teams to document their qualified activities with as much specificity as possible. Project artifacts should support and validate SME input.
4. Preserve All Contemporaneous Technical Documentation
A major weakness in Kyocera was the failure to retain contemporaneous technical documentation, which severely undermined the taxpayer’s ability to defend its claim.
Your company should:
- Establish policies that ensure key technical documentation is preserved for at least the required retention period.
- Coordinate between tax, engineering, and IT teams to identify what systems store this information.
- Avoid unnecessary purging of project records that may be relevant to R&D credit claims.
The IRS is increasingly requesting this type of documentation during exams, and lacking it can be fatal to your credit.
5. Define Business Components Clearly
In Kyocera, broad and loosely defined business components made it difficult to demonstrate how each met the four-part test.
Your process should:
- Define business components at an appropriate level of granularity.
- Show what uncertainty was involved in each component.
- Document the process of experimentation performed for each component.
- Demonstrate that activities contributing to each component meet qualified criteria.
This requires thoughtful collaboration between tax and engineering during the study process.
6. Address High-Risk Job Titles Proactively
Certain job titles and departments are likely to face IRS scrutiny. In particular:
- Upper-level management is typically presumed to be engaged in direct supervision, not research.
- Support functions need clear evidence of how their activities contribute to qualified research.
If time is included in these categories, prepare robust documentation to demonstrate how the individuals performed qualified activities.
Why This Matters Now
The Kyocera case is part of a broader shift in IRS expectations.
Across recent cases and audit trends, the message is clear: The IRS expects documented proof, not just SME recollections.
Which means taxpayers need to demonstrate activity-level detail, link documentation to business components, and preserve project records that support their claims. Companies that continue to rely on legacy study models built around SME interviews and broad estimates are increasingly at risk.
Bottom Line
The lessons of Kyocera are clear. Building a strong R&D credit process requires a thoughtful blend of documentation, SME input, contemporaneous technical documentation, and cross-functional collaboration.
The good news is that improving your process is achievable. Companies that take the time to modernize their approach can significantly reduce risk while ensuring they claim the full value of the credits they are entitled to.
If your team would like to review your current R&D credit process or explore ways to strengthen it, we are here to help. Contact us for a consultation.