A More Predictive Index of Market Sentiment

Recently, finance literature has turned to non-economic factors such as investor sentiment as possible determinants of asset prices. Using mutual fund data, I calculate a new sentiment measure, a perceived loss index. The advantage of the loss index is that it can determine perceived risk for different categories of equities, including market capitalization, style and sector. Results provide evidence that the perceived loss index outperforms all other sentiment and systematic risk measures in predicting future medium run returns, especially for one- and two-year horizons. This evidence pertains not just to broad market returns but also to capitalization-style and sector specific indice returns as well. In addition, I provide evidence that the loss index can be used as a quantitative measure to detect bubbles and financial crises in financial markets.