About

I am an Assistant Professor of Finance at Texas A&M University Mays Business School. I obtained my Ph.D. in Financial Economics from Yale School of Management. I conduct research on empirical asset pricing, behavioral finance, machine learning, and natural language processing.

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 Macro Alibi: Subjective Risk Attribution in Analyst Scenarios

with Chen Wang

Abstract: Sell-side analysts over-attribute a stock’s downside risk to macroeconomic forces, near-unconditionally. We document this Macro Bear Bias in scenario-based valuation reports: bear-case narratives overemphasize aggregate macro risk relative to base- or bull-case narratives, even though realized CAPM R-squared is similar across bear, base, and bull outcomes conditional on the same market state. The bear–base macro-attention gap predicts systematic pessimism in subsequent base-case forecasts, and portfolios formed on a nonlinear bias-adjusted signal earn monthly CAPM alphas of up to 1.9%. Mechanism tests favor a cognitive availability-heuristic explanation—analysts anchor downside narratives on salient macro-crisis templates—over a strategic career-concerns explanation. Narrative templates analysts use to rationalize risk can distort analysts’ numerical forecasts and asset prices.

Professional Investors and Media Topics

Abstract: I investigate the impact of media topics 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 news information into 59 distinct topics, and quantify each topic’s time-varying share of news attention and sentiment. I then define a fund as having exposure to a topic if it overweights stocks expected to perform well when the topic grows in importance, and hence attention. I find that the topics that fund managers choose to have high exposure to are high-sentiment topics, but not those with high attention. {This strategy leads to mutual fund underperformance but attracts investor flows.} Topic-oriented strategies account for a large fraction, specifically 37%, of mutual fund tilts, and are a key driver of the underperformance associated with active tilts.

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.