with Andrew Lo and Stefan Zeume, May 2025. Updated draft!
Abstract: Using tax filings from 374,351 U.S. nonprofit organizations from 2008 to 2020, we provide the first large-scale analysis of endowment prevalence, function, asset allocation, and returns. Endowment use varies systematically across sectors and revenue models. Organizations with endowments scale mission-related spending more effectively and hedge revenue risk through asset allocation. Yet most endowments underperform passive benchmarks, with the weakest performance concentrated among smaller, self-managed funds. Advisory fees are positively correlated with gross returns but negatively with net returns, suggesting overpayment for investment services. Stronger governance, lower discretionary spending, and the use of outsourced CIOs are associated with better performance.
with Dirk Jenter and Lukas Roth, July 2023. R&R at Management Science.
Abstract: This paper analyzes changes in firm value, performance, and behavior caused by CEO deaths. Many deaths trigger large stock price changes—shareholders believe that some CEOs add shareholder value, but others are seen as not optimally matched or overpaid. These value changes are correlated with CEO and firm characteristics (e.g., deaths of old CEOs tend to add value), but much of the variation remains unexplained. The stock price reactions predict changes in operating performance, indicating that shareholders know which CEOs improve performance and which do not. The evidence suggests that CEOs are important, but also that many reduce shareholder value.
October 2017
Abstract: This paper measures the degree of sorting by skill in the U.S. labor market for senior managers. I utilize both wage data and a comprehensive dataset that contains education information on 10,517 executives. I present evidence of a high degree of within-firm positive sorting: better managers are matched with better coworkers. Sorting significantly contributes to variation in productivity across firms and explains a large portion of observed differences in TFP. In addition, sorting explains 20% of variation in compensation. I consider several explanations and suggest that the findings are most easily explained by complementarities in production within executive teams.
December 2016
Abstract: This paper empirically analyzes the labor market for corporate directors and the role compensation plays in this market. Since salaries are not individually negotiated with directors, preferences of directors and firms, rather than wages, determine market outcomes. I estimate preferences from a two-sided matching model and use them in counterfactual experiments. Increasing director pay by $150,000 a year per director allows less desirable firms to attract a better pool of directors. This result suggests that firms can effectively use compensation to build better boards. The framework developed here is used to analyze other important market segments, such as the market for CEO directors.
with Andrew Lo, Zachery Halem, and Sarah Quraishi, Journal of Portfolio Management, November 2022.
Abstract: We examine the long-run impact and implications of an endowment’s spending policy and asset allocation decisions. Using a dynamic model, we explore how different endowment spending rules influence the dynamics of an endowment’s size and future spending. We find that different parameters within each spending rule have significant long-term impact on wealth accumulation and spending capacity. Using Merton's (1993) endowment model and compiled asset allocation data, we estimate the intertemporal preferences and risk aversion of several major endowments. We find significant variation across endowments in their propensity to increase portfolio risk in response to increased spending needs.
with Alexei Zhdanov, Review of Corporate Finance Studies, May 2022.
Abstract: We develop a model of optimal capital structure in imperfectly competitive markets by focusing on a duopoly. The model endogenizes both the financing and investment decisions of firms. We show that in equilibrium the industry leader uses debt conservatively, while the follower uses debt more aggressively and, as the result, defaults first. The model generates novel predictions about the leverage choices of the leader and the follower, their default likelihood, and the degree of leverage dispersion between competing firms. These predictions are strongly supported by the data.
with Evgeny Lyandres and Alexei Zhdanov, Review of Finance, November 2020.
Abstract: This paper shows that the stock market misprices firms’ investment options. We build a real options model of optimal investment under uncertainty to estimate the value of firms’ investment options. We show that firms with valuable investment options have a higher likelihood of being mispriced. Importantly, this mispricing is not one-sided, as such firms are equally likely to be undervalued or overvalued. Our paper adds to the debate on whether public equity markets are myopic and systematically undervalue innovative firms. We show that this is not necessarily the case.