Essays on Finance and Corporate Innovation

Abstract

This thesis consists of two essays on corporate innovation, CEO compensation, firm performance, macroeconomic risk, and stock returns. The first empirical work sheds light on the relationship between failure in innovation, CEO compensation packages, firm performance, and other additional tests. Compensation packages serve as an instrument to motivate CEOs to achieve high performance. On the one hand, CEOs may engage in a reckless shopping spree to build an empire without penalty for poor performance. On the other hand, if the penalty is severe, CEOs will be reluctant to extend the firm's investment in innovation, which is long-term, unpredictable, idiosyncratic, and involves a high probability of failure. To strike a balance between investing in innovation projects and risk aversion, our study suggests that firms might need to design optimal incentive packages that are more tolerant of CEOs who experience failures in innovation. We measure innovation failures by a firm’s non-granted patents (NGPs) per year. The results show that there is a significant reduction post-NGPs year in CEOs’ short-term incentives (jointly and separately). In particular, when we look at the relationship between NGPs and the changes in salary and bonus separately, we find a negative and significant correlation between these factors and NGPs, with more impact on bonus (with a higher Pearson’s r) than salary. This is because bonus schemes are dependent on earnings. In other words, the bonuses of CEOs, particularly those in firms actively engaged in innovation, are associated with the success of these innovations. When we join these two components of CEO compensation, we discover a negative and significant correlation between a CEO’s cash compensation (the sum of both) and NGPs, but with less impact compared to the separate tests. The log we use on total cash compensation may weaken the results. When we examine the relationship between NGPs and the CEO’s long-term incentives, we find a negative and significant correlation between the subsequent CEO’s stock-based compensation and firms’ NGPs. However, we did not observe any significant correlation between NGPs and the changes in other components of the CEO’s long-term compensation, such as option-based compensation. In addition, we computed several compensation variables, including short- and long-term compensation, stock and options, etc., to investigate the same relationship. Our findings show that there is no significant relationship between the changes in these variables and NGPs. The first study also includes a significant section that explores the changes in the firm's performance after the NGP year. We find that the changes in subsequent firm performance, as measured by ROA and ROE, have a significant negative correlation with the NGPs. Furthermore, we set up two supplementary empirical models to investigate the relationship between innovation failure and (1) CEO departure (given that almost 24% of CEOs left their positions post-NGPs) and (2) the changes in R&D investments, as the rest of the CEOs held their positions for a minimum of three years. The results indicate that NGPs have no direct impact on CEO departure following the failure in innovation. Regarding the changes in R&D, we find that R&D is negatively and statistically significantly in response to NGPs. This finding indicates that CEOs tend to sharply reduce the firm’s subsequent investment in R&D, particularly when they face failures in obtaining patents. The second empirical work explores how firms' exposure to macroeconomic risk can help explain the relationship between firms' R&D investment (i.e., innovation input), patents granted (i.e., innovation output), and stock returns. To understand the relationship between these variables, we introduce two critical macro-level risk factors that are missing in the literature: exposure to economic growth, or short-run risk (SRR), and exposure to expected economic growth and uncertainty, or long-run risk (LRR). The latter has two components that capture economic conditions: expansion (high economic growth and/or low economic uncertainty) or recession (low economic growth and/or high economic uncertainty). We first examine the relationship between firms' exposure to SRR and LRR and R&D expenditure. We find that firms with low exposure to SRR engage more in R&D spending, making their future earnings more exposed to LRR. As compensation for LRR dominates, these firms also have higher stock returns. This result shows that firms with higher size-adjusted R&D investment have higher stock returns or risk premiums, providing a risk-based analysis of why R&D investment is considered a risk factor in the literature. We then examine the relationship between patents granted and exposure to SRR and LRR. In contrast to R&D expenditure, we find a positive relationship between patents and SRR and a strong and negative relationship between patents and LRR (i.e., a negative relationship with expected economic growth and/or a positive relationship with future economic uncertainty). This confirms that firms that are frequent innovators have low betas to economic fluctuations, so that patents reduce firms' exposure to future uncertainty. Despite being less exposed to long-term risk, we do not find that their stock returns are lower on average but higher. To further investigate the reason for the latter result, we develop strategies that double-sort firms into the two innovation dimensions, R&D spending and patents granted, in order to compute the corresponding average stock returns and the betas of SRR and LRR for each category. First, we examine the risk exposure of firms with similar levels of R&D spending and show that the decrease in beta to LRR (specifically economic growth) is driven by firms with higher R&D investment. Looking at firms with similar numbers of patents granted per year, we find that the more a firm invests in R&D, the higher its exposure to future macroeconomic conditions. Second, within firms with different levels of R&D spending, we find that there is no relationship between patents granted and stock returns. However, when analyzing the relationship between patent returns and R&D returns, we find that although stock returns increase with the total number of patents, the increase is less pronounced than the increase in R&D expenditure. This shows that the long-term risk of investing in R&D is reduced, but not completely offset, by the granting of patents.