Patent Semi-Comparables

Document Type


Publication Date



Over the last decade, courts have applied increasingly stringent standards to the evidence used to determine patent damages. While this has reduced the risk of awards untethered to the facts, the current focus on strictly comparable licenses covering technology similar to the patent that resemble one-way, royalty-bearing “hypothetical licenses” specified by law has created its own problems, particularly in the valuation of component patents, many of which are incorporated into a single product. The rejection of what we refer to as “semi-comparables”—licenses that deviate in some significant way from the terms of the hypothetical license—has led to distorted incentives, unpredictability, and the exclusion of many of the ex ante transactions that best reflect the incremental value of the invention. We believe that the wholesale exclusion of such licenses is wrong—the problem is not that certain licenses are insufficiently comparable but that courts need better approaches to properly and flexibly interpret evidence of comparables and semi-comparable licenses. Framed in this way, the solution is not to exclude licenses but instead to apply an inclusive but disciplined approach to reasonable royalty determinations that prioritizes objective evidence of a patent’s incremental value even in the form of traditionally excluded “semi-comparable” transactions like technology (as opposed to “patent”) licenses and sales. Though courts have been reluctant to use semi-comparables because of a lack of objective information about their formation, we begin to address this void, drawing upon the collective wisdom of licensing lawyers we interviewed, the nearly two-decade-long career of one of us as a licensing lawyer, and studies of thousands of actual licenses. When a reasonable, evidence-based estimate or upper bound cannot be derived, we consider the expanded use of tailored injunctions, assuming the other eBay elements are met.

This document is currently not available here.