“…In particular, we introduce two new patent-based measures to capture the theoretical distinction between 'exploration' and 'exploitation' in corporate innovation search strategy (March, 1991;Manso, 2011). Our results expand on the recent findings that investors tend to undervalue firms that: (a) invest heavily in R&D and have exhibited historical success in converting R&D investment into sales (Cohen et al, 2013); are efficient in translating R&D expenditure into future patents (Hirshleifer et al, 2013) and (c) produce patents that are relatively more "original" compared to all other patented innovation, proxied by the diversity of technology classes cited (Hirshleifer, Hsu & Li, 2018). While these papers illustrate investors' difficulty in comparing distinct firms across various innovation metrics, such as historical capacity or the originality of their patents, our findings highlight the added difficulty that investors face when interpreting the incremental economic significance of a firm's current innovative output relative to the past innovative output of the same firm.…”