Investor attention predicts short-term stock moves ^{CEFR B2}

New research demonstrates that investor attention, measured at the stock level and aggregated across the market, helps predict short-term stock returns. The study, authored by Zhi Da of the University of Notre Dame with coauthors from Baruch College and National Taiwan University, is published in Management Science.

The team measures retail attention with Google’s daily search volume index and institutional attention with Bloomberg’s “Daily Maximum Readership” score. For each stock they compute abnormal attention and average these values across all stocks to form two daily market indexes: Aggregate Retail Attention (ARA) and Aggregate Institutional Attention (AIA). They then run regressions of market returns on ARA and AIA to test predictive power.

The study finds two consistent patterns. First, higher retail attention predicts lower returns over the following week; popular stocks with heavy retail buzz tend to underperform later because individual investors often arrive late and push prices too high. Second, higher institutional attention predicts higher returns, particularly before major announcements, since institutional investors often begin research ahead of news and their interest can signal forthcoming uncertainty and higher required returns. The authors also show that top-down measures, such as searches for broad market terms, perform poorly compared with this bottom-up approach. The findings can help explain market moves and inform investing decisions.