Testing Investor Sentiment Against Spurious Anomaly Return Predictors

Summary

The document summarizes research on whether investor sentiment genuinely helps explain cross-sectional stock return anomalies or appears predictive because it is persistent. Persistent predictors can spuriously forecast returns when expected returns also vary over time, so the article compares sentiment’s results with simulated autoregressive series. The empirical setting covers eleven anomalies and three related hypotheses: sentiment’s relation to long-short returns, short-side returns, and the lack of a corresponding relation to long-side returns.

Across 200 million simulated regressions, no random series matched sentiment’s joint support for all three hypotheses. The reported evidence is especially consistent for short-side returns, which the source links to limits on short selling. These simulations make a chance explanation less plausible within the chosen setup, but do not prove a causal effect or rule out other model choices. The document is a translated summary of a study, and its conclusions depend on the anomaly sample, simulation design, and assumptions about persistence and expected returns.

Key ideas

• Persistent predictors can appear to forecast returns spuriously when expected returns vary over time.
• The study compares sentiment’s regression patterns across eleven stock anomalies with results from simulated autoregressive predictors.
• The reported evidence finds that random predictors rarely match sentiment’s joint pattern, particularly for short-side returns.
• The simulation results support, but do not prove, sentiment-based mispricing as an explanation for anomalies.

Tags

This summary was written by Stratmill's research agent from the original; it is not a copy of the source.