Research

There is significant heterogeneity in ESG funds’ incentives to engage with portfolio firms. If funds view ESG as a value driver, then these incentives will affect funds’ behavior and thus their impact on firms. We compare ESG funds with similar levels of ESG investments but different incentives to engage. We find that funds with higher incentives to engage, i.e., committed ESG funds, conduct more ESG-related information acquisition, pursue longer-term investment strategies, engage more intensely on ESG issues, and have greater real impacts. Moreover, committed ESG funds have outperformed other ESG funds within subportfolios with higher and more effective ESG engagement.

We study the effect of generative AI, specifically ChatGPT, on academic writing quality and subsequent citations. Leveraging a novel dataset of over 120,000 papers from financial economics, management, and political science, we construct detailed writing quality measures using both traditional linguistic metrics and ChatGPT-based evaluations. We find that higher writing quality strongly predicts future citations. Following the release of ChatGPT in late 2022, writing quality improved significantly, particularly for authors from non-English-speaking institutions, those with non-English names, and those with fewer historical citations. These improvements translate into a narrowing of the citation count gap between these groups and their counterparts, and consequently to a more equal distribution of academic impact.

This paper examines how securities lending affects liquidity management and performance in corporate bond mutual funds. We show that funds engaging in lending hold significantly less cash than non-lenders, suggesting a more aggressive portfolio strategy. While lending funds outperform in normal times due to additional income from lending fees and collateral reinvestment, they underperform during periods of large outflows. Using detailed fund-level and position-level data, we document that lending funds are more likely to recall and sell previously lent bonds when facing redemptions, especially bonds they hold in large ownership, creating price pressure and dragging down performance. Importantly, we find that this mechanism is unique to bond funds and not present in equity funds, reflecting the greater illiquidity of bond markets. Our findings highlight a tradeoff between return enhancement and fragility, and offer new insight into how a common fund practice like securities lending can amplify risks during stress.

Using new SEC data, we study fund derivative use and its impact on performance. Despite small portfolio weights, derivatives contribute largely to fund returns. Contrary to prior research, we find most employ derivatives to amplify, not hedge, equity returns. Using machine learning to classify funds’ derivative strategies reveals high specializations linked to information-related trading, liquidity management, gaining exposure, or hedging motives. Long index derivative users drive the amplification. During the COVID-19 crisis, they significantly increased derivative use more than others and shifted strategies, but initially lost on existing positions and then on newly opened short positions when markets unexpectedly rebounded.

Using mutual funds’ security lending patterns, we propose a data-driven method using association rules to differentiate shorting demand for general hedging and for exploiting firm-specific negative information, which enables us to quantify whether a fund engages in value lending (low quantity and high fee) or volume lending (high quantity and low fee). Contrary to the common prior that security lending is conducted at the fund family or lending agent level, we find rich heterogeneity in funds’ lending strategies within family and lending agent. Despite a lower lending revenue split with lending agents, volume lenders generate higher lending revenue and can better offset fund expense ratio than value lenders. Volume lenders also enhance their yield on cash collateral by lending cash through repurchase agreements to borrowing agents with security lending relations.

This paper investigates portfolio pumping at the fund family level, where non-star fund managers strategically purchase stocks held by star funds in the family to inflate their quarter-end performance. Star funds that engage in such activities show inflated performance after 2002 when the Securities and Exchange Commission increased regulation on portfolio pumping. Stocks pumped by the strategy show strong reversals at the quarter end. Moreover, despite a minor underperformance stemming from portfolio misallocation, non-star fund managers pumping for star funds receive abnormally high subsequent flows, suggesting a pattern of family subsidization.

This paper examines investors’ retirement savings allocation using a hand-collected dataset on 401(k) plans. We find that 83% of investors in our sample hold only 39% of total assets and follow a return-chasing strategy. In contrast, the remaining 17% of wealthy investors with relatively higher financial literacy follow CAPM alpha. This difference between the two investor groups explains why fund flows respond to returns at the plan level but to CAPM alpha at the aggregated fund level. Return-chasing by unwealthy investors is not optimal, as it significantly underperforms a strategy that passively invests in the existing funds in their plans.

This paper empirically shows that information acquisition affects stock returns by reducing firm-level information asymmetry. When firms disclose material information known by insiders, information acquisition reduces asymmetric information and lowers stock returns. The effect is stronger for both unexpectedly good and bad news relative to anticipated news and when investors’ cost of information processing is lower. Using the Northeast Blackout of 2003 as a natural experiment, I explore an exogenous shock in information acquisition and show causal evidence that information acquisition reduces information asymmetry.

We study momentum and its predictability within equities listed at the London Stock Exchange (1820-1930). At the time, this was the largest and most liquid stock market and it was thinly regulated, making for a good laboratory to perform out-of-sample tests. Cross-sectionally, we find that the size and market factors are highly profitable, while long-term reversals are not. Momentum is the most profitable and volatile factor. Its returns resemble an echo: they are high in long-term formation portfolios, and vanish in short-term ones. We uncover momentum in dividends as well. When controlling for dividend momentum, price momentum loses significance and profitability. In the time-series, despite the presence of a few momentum crashes, dynamically hedged portfolios do not improve the performance of static momentum. We conclude that momentum returns are not predictable in our sample, which casts some doubt on the success of dynamic hedging strategies.