This study explores the role of knowledge interdependencies on the termination of patented inventions. Termination refers to the abandonment of inventive efforts that are no longer deemed promising. We argue that high interdependencies between an inventive effort and the other inventions in the same research program will increase the cognitive burden on managers and decrease the likelihood of termination. Further, in the presence of interdependencies, managers are likely to rely on heuristics for termination decisions. We focus on two such heuristics: interdependencies of an invention with those in other research programs and the level of external competition in the research program. We test our hypotheses with longitudinal data on patent terminations through non‐payment of renewal fees in the pharmaceutical industry.

Effective management of innovation portfolios requires termination of opportunities that are no longer promising. Most current tools on termination assume that opportunities to be evaluated are independent from one another. This assumption may limit their usefulness in increasingly complex research domains, such as pharmaceutical R&D. In this study, we investigate how interdependencies among inventions influence firms’ tendency to terminate those inventions. Our results on patent terminations show that a patent that is more interdependent with other patents in the same research program is less likely to be terminated. This suggests that managers may have difficulty in evaluating the inherent value of interdependent opportunities. This result is stronger when the patent is less interdependent with those in other research programs or in a more competitive area.