Thomas R. Varner
Vice-President, Economists Inc., Emeryville, CA

Previous studies of technology license agreements have been limited either to analyzing relatively small sets of agreements or to analyzing a mix of complete agreements and general descriptions of agreements from a variety of sources. This study is unique in that it considers a large number of complete, publicly available agreements from one source, agreements submitted to the U.S. Securities & Exchange Commission (SEC) from 1994 to 2009. The dataset used for this study is also unique in that it contains a large number of royalty-based agreements previously granted Confidential Treatment by the SEC and obtained in unredacted form under the Freedom of Information Act.

I collected a total of 2,963 agreements as part of this study by searching for agreements submitted by high-tech and biotech firms to the SEC—covering over 40 Standard Industrial Classification (SIC) codes—that contained a running, or earned, royalty expressed as a percentage of licensed product sales. This unique dataset is comprised of actual license agreements submitted to the SEC as exhibits rather than simply general descriptions of those agreements, and is based on licenses that meet the SEC’s threshold of materiality. These characteristics, along with the large number of agreements collected, make this dataset well suited to conducting a variety of analyses across multiple industries, as well as analyses of selected subsets of agreements. After presenting my analysis of all agreements in the dataset, I devote the remainder of this paper to findings related to analysis of almost a thousand patent licenses in the dataset. These patent licenses are comprised of Bare Patent licenses (i.e., rights to use patented technology only), Patent plus Know-How licenses (i.e., rights to use patented technology plus rights to additional information such as laboratory data, market analysis, or related studies), and Patent Settlement agreements.