We examine the predictability of expected stock returns across horizons using machine learning. We use neural networks, and gradient boosted regression trees on the U.S. and international equity datasets. We find that predictability of returns using neural networks models decreases with longer...

Based on 58,256 news articles published in the Financial Times during a 15-year period that cover companies in the DJIA, we find that a trading strategy that longs stocks with the most negative news and shorts stocks with the least negative news is not profitable. Consistent with this result, we...

A war-related factor model derived from textual analysis of media news reports explains the cross section of expected asset returns. Using a semi-supervised topic model to extract discourse topics from 7,000,000 New York Times stories spanning 160 years, the war factor predicts the cross section...

Investors' return expectations are pivotal in stock markets, but the reasoning behind these expectations remains a black box for economists. This paper sheds light on economic agents' mental models - their subjective understanding - of the stock market, drawing on surveys with the US general...

We study how investors respond to inflation combining a customized survey experiment with trading data at a time of historically high inflation. Investors' beliefs about the stock return-inflation relation are very heterogeneous in the cross section and on average too optimistic. Moreover, many...

In this paper, we use accounting fundamentals to measure systematic risk of distress. Our main testable prediction—that this risk increases with the probability of recessionary failure, P(R|F)—is based on a stylized model that guides our empirical analyses. We first apply the lasso method to...