My research interests are empirical asset pricing, sustainable finance, artificial intelligence and machine learning, and cryptocurrency markets. My current work studies risk, return predictability, and factor exposures in newer financial markets, with a particular focus on ESG mutual funds and cryptocurrencies.
Status: Available upon request.
This paper examines whether machine-learning methods can predict future return and risk in ESG mutual funds. I compare the predictive content of ESG variables, fund-level characteristics, and stock-level variables constructed from fund holdings. The analysis uses quarterly data from 2002 to 2024 and combines CRSP mutual fund data, Refinitiv ESG scores, ESG fund classifications, fund holdings, and stock-level characteristics.
The main results show that future fund returns are difficult to predict directly, while future realized volatility is substantially more predictable. Importantly, volatility forecasts also contain information about subsequent fund returns. Portfolios sorted on predicted volatility generate economically meaningful return differences, suggesting that predicted risk helps identify variation in expected fund returns. Overall, the paper shows that machine learning is especially useful for understanding risk predictability in ESG mutual funds, while ESG scores alone provide limited incremental predictive power relative to fund characteristics and holdings-based information.Â
Presented at: AI in Finance Conference, 2026 scheduled
This paper studies whether equity-market return and volatility risks are priced in the cross-section of cryptocurrencies. Using a large cryptocurrency universe, we estimate time-varying exposures to stock-market returns and market volatility and test whether these exposures explain differences in expected cryptocurrency returns. The results suggest that cryptocurrencies are not fully detached from traditional financial markets: part of their expected return variation reflects exposure to equity-market risk.
Presented at: SWFA Annual Conference, 2025; FMA PhD Poster Session, 2025; WFC Annual Conference, 2026 scheduled
This paper examines whether cryptocurrency returns are exposed to traditional risk factors from equity, currency, and commodity markets. The results show that cryptocurrencies are connected to several traditional asset classes, with the strongest and most consistent evidence coming from equity-market factors. The paper provides broader evidence that cryptocurrency markets are not fully isolated from traditional financial-market risks.
Presented at: FMA Annual Meeting, 2024
