Blog - Investing Notes

February 9, 2007 - Aggregate Investor Sentiment and Stock Returns

Is aggregate investor sentiment a useful trading indicator? For what kinds of stocks is sentiment trading most likely to work? In their December 2006 paper entitled "Investor Sentiment in the Stock Market", Malcolm Baker and Jeffrey Wurgler summarize a top down approach to addressing these questions, focusing on the measurement of aggregate sentiment and its relationship to stock returns. They devise a long-run sentiment index based on the six indicators: trading volume as measured by NYSE turnover; the dividend premium; the closed-end fund discount; the number of and first-day returns on Initial Public Offerings; and the equity share in new issues. Using this index and stock return data for 1966-2005, they conclude that:

• Historically, large waves of sentiment map reasonably well to noted stock market trends.
• The correlation between equal-weighted (capitalization-weighted) market returns and changes in the sentiment index is a highly significant 0.43 (0.32).
• When the sentiment index is more than one standard deviation above (below) its historical average, monthly returns average -0.34% (+1.18%) for the value-weighted market and -0.41% (2.75%) percentage points for the equal-weighted market. However, the statistical reliability of these results is only modest. (See first chart below.)
• When sentiment is low (high), the average future returns of volatile stocks exceed (trail) those of bond-like stocks.

The following chart, taken from the paper, shows average monthly equal-weighted and value-weighted market returns (in percent) according to sentiment level in the preceding month. It shows that poor returns tend to follow very high sentiment levels and that equal-weighted returns (emphasizing small-capitalization stocks) are more sensitive to sentiment than value-weighted returns.

The data use to construct the long-run sentiment index are available at Jeffrey Wurgler's NYU web page.

In summary, aggregate investor sentiment offers some value in forecasting stock returns, especially for the most speculative stocks.

For related research, see Blog Synthesis: Sentimental Journey.