About

I am a Ph.D. candidate in Finance at Yale School of Management. I conduct research on asset pricing, machine learning, and natural language processing.

I am on the 2024-2025 academic job market.

Job Market Paper

Active Mutual Funds and Media Narratives

Abstract: I investigate the impact of media narratives on the portfolio strategies of active equity mutual funds. Using 1.5 million Wall Street Journal articles from 1984 to 2023, I use ChatGPT to distill media narratives into 59 distinct topics, and quantify each one’s time-varying attention and sentiment. I find that mutual funds increase their exposure to topics with high sentiment, but not necessarily to those with high attention. While this strategy leads to mutual fund underperformance, it also attracts investor flows. Topic-oriented exposures account for a large fraction of the variation in mutual fund tilts, and are a key driver of the negative alpha associated with active tilts.

Publications

The Virtue of Complexity in Return Prediction

with Bryan Kelly and Semyon Malamud
The Journal of Finance 79, no. 1 (2024): 459-503. Link

Abstract: Much of the extant literature predicts market returns with “simple” models that use only a few parameters. Contrary to conventional wisdom, we theoretically prove that simple models severely understate return predictability compared to “complex” models in which the number of parameters exceeds the number of observations. We empirically document the virtue of complexity in US equity market return prediction. Our findings establish the rationale for modeling expected returns through machine learning.

Working Papers

The Virtue of Complexity Everywhere

with Bryan Kelly and Semyon Malamud

Abstract: We investigate the performance of non-linear return prediction models in the high complexity regime, i.e., when the number of model parameters exceeds the number of observations. We document a “virtue of complexity” in all asset classes that we study (US equities, international equities, bonds, commodities, currencies, and interest rates). Specifically, return prediction R-squared and optimal portfolio Sharpe ratio generally increase with model parameterization for every asset class. The virtue of complexity is present even in extremely data-scarce environments, e.g., for predictive models with less than twenty observations and tens of thousands of predictors. The empirical association between model complexity and out-of-sample model performance exhibits a striking consistency with theoretical predictions.

Robust Prediction after Structural Breaks

[Draft Available Upon Request]

Abstract: I propose a new modeling approach for time series prediction after structural breaks. The method incorporates a time trend variable into non-linear predictive models to effectively handle coefficient variations over time. By optimizing the bias-variance tradeoff, this approach significantly improves prediction accuracy and optimal portfolio Sharpe ratio compared to both linear and non-linear standard models. I construct Monte Carlo simulations to examine the finite sample performance of the proposed procedures. Empirically, the paper demonstrates improved prediction performance for U.S. equity market returns. These findings establish the robustness of machine learning predictions in the presence of structural breaks.