Patents That Work for a Living

Patents sit at an intermediate stage of corporate innovation. They secure legal claims over technologies, while the value of those claims depends on firms’ ability to incorporate the underlying inventions into new products. As patenting has expanded and strategic patent accumulation has become more common, the size of a patent portfolio has become a noisy proxy for innovation value. A portfolio may contain technologies that support product development, assets held for defensive or bargaining purposes, or inventions that never acquire a commercial role.

Our paper addresses this gap by shifting attention to patent utilization. Specifically, the authors develop a measure of corporate patent utilization to identify how much of a firm’s recent patent stock is reflected in its new product launches. We then examine whether greater utilization predicts subsequent product development, market-share growth, profitability, and valuation. 

Bridging the Language Gap

Constructing the measure is not straightforward. Patent filings and product announcements are written in different registers, one technical and legalistic, the other aimed at investors. A simple keyword match would miss most meaningful connections. We use FastText, a machine-learning language model, to compare patent texts from PatentsView with new product launch descriptions from Capital IQ Key Developments, covering U.S. public firms between 2002 and 2022. A patent-product pair counts as a utilization match when its similarity score falls above the 80th percentile of the full sample. Dividing utilized patents by the firm's total five-year patent stock gives the utilization rate.

Figure 1. Patent-Product Pair Matching Process Illustration

Utilized patents are more closely tied to a firm's core technology fields, receive more internal citations from later patents, and are less likely to be sold. Product launches backed by more utilized patents generate stronger abnormal returns around announcement dates and score higher on a text-based measure of breakthrough quality. The measure is picking up patents that are genuinely embedded in the firm's innovation pipeline, not incidental vocabulary matches.

What Utilization Predicts

A one-standard-deviation increase in patent utilization is associated with roughly 30% more new products and 22% more breakthrough products in the following year, alongside higher cumulative abnormal announcement returns. This is not simply a reflection of firms that launch more products doing better overall, we control for new product intensity throughout, and utilization retains its predictive power independently.

The effects carry through into market position and financial performance. Firms with higher utilization see sales growth and market-share gains of around one percentage point over three years, and improvements in gross profit margin, return on assets, and market-to-book ratio in the subsequent year. The results hold under an instrumental-variable design that uses firms' pre-existing technological exposure to isolate the causal effect.

One finding deserves particular attention. When patents are separated by their estimated economic value at the time of grant, the performance gains are driven almost entirely by the utilization of high-value patents. Commercializing low-value patents contributes little, and in some specifications is associated with weaker outcomes. What matters is not the volume of patent-linked product activity but whether the underlying inventions were worth something to begin with. Firms that put their most valuable patents to work outperform, firms that build product announcements around peripheral filings do not.

The effect is also larger in more competitive product markets, where the pressure to differentiate and defend market share is greater. In those environments, the ability to translate patented technologies into products is a more meaningful source of advantage.

Beyond the Patent Count

For investors, the implication is direct. Patent counts tell you what a firm invented, utilization tells you what it did with the invention. A large portfolio with low utilization may be a weaker asset than it appears. A smaller but actively utilized base, concentrated in high-value patents, may be the more reliable signal of future performance.

For managers, the paper findings sharpen a question that aggregate portfolio metrics tend to obscure: which patents in the firm's stock are actually supporting the product pipeline, and which are not? The data suggest the distinction has real consequences for market position and profitability.

The policy implication runs deeper. Systems that reward patenting without corresponding commercialisation may be encouraging accumulation that produces limited economic return. A portfolio may contain technologies that support product development, assets held for defensive or bargaining purposes, or inventions that never acquire a commercial role. Utilization is a more honest measure of whether intellectual property is generating the returns — in product markets, in firm growth, in broader economic activity — that the system is meant to produce.

______________

Jarrad Harford is the Paul Pigott - PACCAR Professor of Finance and Chair of the Finance and Business Economics Department at the University of Washington’s Foster School of Business, and an ECGI Research Member.

Qiyang He is a Lecturer (Assistant Professor) in Finance at Deakin Business School. 

Buhui Qiu is a Professor of Business and Finance at the University of Sydney Business School. 

This blog is based on a paper presented at the 4th HKU Summer Finance Conference. Visit the event page to explore more conference-related blogs.

The ECGI does not, consistent with its constitutional purpose, have a view or opinion. If you wish to respond to this article, you can submit a blog article or 'letter to the editor' by clicking here.