Research

I study whether disclosure mandates alter the equilibrium of cyberattacks by unintentionally informing cybercriminals. The California Consumer Privacy Act (CCPA) requires companies to disclose their personal information collection practices to consumers, inadvertently informing cybercriminals about the potential benefits of breaching each firm. Using a difference-in-differences design, I find that firms disclosing the collection of valuable personal data face an increased probability of data breaches. These firms also strengthen their cyberdefenses, both in terms of software and employee cybersecurity expertise. Firms trade off cybersecurity costs against the risk of data breaches, with the increase in breach probabilities more pronounced among firms that invest less in cybersecurity. Finally, I find that firms voluntarily disclose more about their cyberdefense and adjust their data-collection policies as additional defense strategies. Overall, this study highlights the trade-off between transparency and cybersecurity risks in today's economy.

We study the economic consequences of managers' vocal delivery quality during earnings conference calls. We introduce a novel measure, vocal delivery quality, that captures the acoustic comprehensibility of audio information for an average listener. Our measure relies on a deep-learning algorithm applied to a large sample of earnings call audio files. Consistent with predictions from the psychology and accounting literatures, we find evidence that the quality of managers' vocal delivery deteriorates when they deliver negative news, such as a decrease in earnings or negative narrative information, and positive but transitory earnings news. We show that the stock market reacts in real time to managers' vocal delivery quality. We also document that the vocal delivery quality has an effect on information intermediaries such as analysts and the media. Overall, our findings underscore the role of vocal dimensions in corporate oral disclosures.

We show a novel mechanism by which mutual fund managers strategically alter their portfolios to take advantage of investors' reliance on Morningstar star ratings. Specifically, funds achieve higher ratings by changing their holdings to induce Morningstar to reclassify them into size/value style boxes with lower average performance, thereby enabling more favorable peer comparison. This practice, which we term  `box jumping', attracts fund flows and higher fees, despite sacrificing return performance and the ratings upgrades reversing within three years on average. These patterns emerge after 2002 when Morningstar ratings began to be based on relative performance within style boxes, and are predictably absent beforehand. We also show that pervasive box jumping creates negative spillover effects on other funds. Together, our findings highlight portfolio recomposition as a novel and strategic lever that funds use to manipulate Morningstar ratings, and that funds box jump despite compromising returns due to investors' fixation on ratings when allocating capital. 

Using two large-scale experiments, we explore how generative AI shapes both overall performance and disparities in investment tasks. When investors are given AI summaries aligned with their sophistication, they become better at processing financial information and making investment decisions. Conversely, misaligned summaries generally have an adverse effect, suggesting AI's ability to benefit investors hinges on personalization of information. We also show AI's benefits accrue disproportionately to individuals with higher financial expertise, which stems from an inherent tradeoff between accessibility for less sophisticated investors and technical precision used by more sophisticated investors. Together, our findings suggest AI improves performance on investment tasks, on average, but also underscore the potential for these tools to widen rather than limit existing performance gaps. 

The literature and anecdotal evidence offer conflicting predictions regarding the relationship between ESG investing and securities lending. Despite the increasing incidence of securities lending by passive funds generally, we find that this is not the case for passive ESG funds. Using high-dimensional fixed effects, we show that passive ESG funds lend 30% less than non-ESG funds at both the fund and stock levels, even when they are holding the same stocks. The forgone revenue is equivalent to 20% of their expenses. We evaluate several explanations and find that our results are consistent with the idea of social norms in investing, where the negative connotation of short-selling labels lending as undesirable behavior for investors pursuing ESG. We add to the understanding of ESG investors' nonpecuniary motivations by quantifying the financial benefits that ESG investors forgo in the securities lending market.