UC San Diego

As Intellectual Property (IP) increasingly acquires a central role in today's knowledge economy, tools for valuation of IP assets for facilitating new markets for their trade, and for effective generation and management of IP assets is becoming central to firms' strategies. An increase in patenting activity and patent litigations have also led to important policy debates regarding the potentially harmful impact of the patent system and proposed reforms to it. In this thesis, I explore how fundamental questions around valuation of patents, strategic management of patents, and policy issues around the patent system can be addressed via empirical econometric analysis of data. First, I explore the relationship between publicly available patent characteristics and the measure of patent values. Using patent specific value estimates, I find that the number of received citations explains the value measure far more than other indicators such as the citations made, number of claims, countries in which a patent is applied, and prosecution effort. I also find that focusing on specific technology areas is important for a more precise understanding of the indicators of patent values. Second, I estimate whether the value of innovations can be predicted to any degree early in their life cycle. I exploit the establishment of screening processes by organizations to predict of value of innovations before filing them as patents, to save high costs incurred in pursuing potentially non valuable innovations with the patent offices. I find that despite a large effort expended by groups of leading subject matter experts, the prediction power of the value of early innovations remains weak. This finding suggests that innovations are unpredictable by nature, even to those with the best state -of-the-art knowledge in the area of innovation, and explains why we observe a highly skewed distribution of their value in general. Third, I analyze how the patenting behavior of start-up firms is related to patent thickets, i.e., by the number of existing IP rights holders and by the amount of dispersion of those rights, while controlling for observable firm and market characteristics. I investigate the impact of patent thickets on both quality and quantity of patents produced by start-up firms founded between 2001-05 in rapidly evolving semiconductors, digital communications, and telecommunications industries. I find that on average, start-up firms file more and higher quality patents in technology areas with higher number of existing IP rights holders