Stocks
This paper shows that a simple, model-free valuation approach using all available accounting data can identify stock mispricing. A strategy that buys underpriced and sells overpriced firms earns **4–10% alpha per year**, outperforming many known anomalies—even without complex valuation models.
Short sellers face unique risks, including loan recalls and rising borrow fees. This paper shows that higher short-selling risk predicts **lower future returns**, **more mispricing**, and **less efficient prices**. These effects are stronger for trades with long expected horizons.
This paper compares machine learning methods to predict stock returns and measure equity risk premia. The best-performing model (a 3-layer neural net) earns out-of-sample Sharpe ratios as high as 2.45.
This paper introduces a new class of asset pricing models using autoencoders. By embedding firm characteristics into a neural network architecture, the authors construct a nonlinear conditional factor model that enforces no-arbitrage. It dramatically outperforms linear benchmarks.
Machine learning (ML) signals look great in backtests—but they often break when exposed to economic restrictions. This paper shows that ML-based strategies lose 50–80% of their alpha when excluding microcaps, distressed stocks, or accounting for transaction costs.
This paper develops the Policy Change Index (PCI) for China, a machine learning-based, language-free signal that predicts upcoming policy shifts by analyzing the page placement of articles in the People’s Daily. The PCI successfully anticipates major Chinese policy events.
This paper uses deep learning to estimate the stochastic discount factor (SDF) that prices all U.S. stocks. The authors combine a feedforward network, LSTM, and a generative adversarial network (GAN) to build a non-linear, no-arbitrage-compliant model.
This paper shows that classic market microstructure measures like VPIN, Amihud, and Roll still have predictive power—even in modern, high-frequency, machine-traded markets.
This paper runs over 2 million trading strategies using CRSP and Compustat data to expose how widespread p-hacking is in finance. After accounting for multiple testing and requiring economic significance, fewer than 20 strategies remain—and none have theoretical justification.
Multinational firms' returns are predictably linked to foreign industry news from their sales regions. This paper shows that investors underreact to economically relevant information abroad.
This paper explains why return predictability—and momentum—concentrates in bad times. Investors rely on different forecasting models, which causes disagreement to spike in recessions. These disagreement shocks create time-series momentum, especially when uncertainty is high and beliefs polarize.
Unusual negative news—news that combines rare language with negative sentiment—predicts sharp increases in market volatility. This paper shows that both firm-specific and market-wide volatility rise significantly after such news, with effects that persist for months.