Sentiment makes an impact on the kind of investment choices people make, and therefore, investor greed and fear can be potential predictors of market performance. The challenge of investor sentiment not being observable is addressed by the authors’ finding the appropriate proxies and formulating an index of investor sentiment. Such indexes tend to predict markets better than conventional indicators.
The inverse relationship between investor sentiment and future stock market returns has strong economic effects. But a lack of direct observability of investment sentiment has been an obstacle to putting this phenomenon to use. The authors use established proxies and new statistical tools to construct an aligned investor sentiment index. This index is demonstrated to be a robust predictor of market performance, outperforming macro variables, as well as a predictor of cross-sectional returns by industry, size, value, and momentum. Its predictive power appears to be driven by investors’ biased expectations of future cash flows.
How Is This Research Useful to Practitioners?
Baker and Wurgler (Journal of Finance 2006 and Journal of Economic Perspectives 2007) addressed the issue of the inability to observe investor sentiment by creating an investor sentiment index using six sentiment proxies. The authors extend the work of Baker and Wurgler and many others by further refining the econometric processes used to create the aligned investor sentiment index.
Empirical results show that the aligned index has better forecasting capability for both aggregate and cross-sectional stock market returns than previous investor sentiment indexes. It achieves the monthly in- and out-of-sample R2 values of 1.70% and 1.23%, respectively, compared with 0.30% and 0.15% for the index formulated by Baker and Wurgler, and it is statistically and economically significant.
It also performs better than such major economic predictors as the short-term interest rate, the ratio of dividend yield earnings to price, term spreads, the book-to-market ratio, stock volatility, inflation, corporate issuing activity, the consumption wealth and surplus ratios, and the output gap. It is second only to Kelly and Pruitt’s (Journal of Finance 2013 and Journal of Econometrics 2014) predictor based on the book-to-market ratio. The authors’ investor sentiment index performs better than other behavior predictors monthly and converges with them in the long run.
The authors also consider the driving force linking investor sentiment and asset prices and find that higher prices are driven more by investors’ biased expectation of future cash flows than by a lower discount rate.
Forecasting market returns, asset pricing, and the risk premium are some of the central problems of finance. Insights into investor sentiment contribute to a better estimation of these, and therefore, these findings have broad applications in financial research and practice.
How Did the Authors Conduct This Research?
The authors statistically link the stock market returns and the proxies for investor sentiment to construct the aligned investor sentiment index.
The aggregate stock market return is computed as the continuously compounded log return on the S&P 500 Index (including dividends) minus the risk-free rate.
The proxies used are the same as those used by Baker and Wurgler: (1) closed-end fund discount rate (difference between the net asset values of closed-end stock mutual funds and their market prices), (2) share turnover (ratio of share turnover volume to average shares listed), (3) number of initial public offerings (IPOs), (4) first-day returns of IPOs, (5) dividend premium (log difference of the value-weighted average price-to-book ratios of dividend payers and nonpayers), and (6) share of equity issues in all new issues. The data on these measures, covering July 1965 through December 2010 (546 months), are procured from Wurgler’s website.
The authors construct a more efficient investor sentiment index using the partial least-squares method of Kelly and Pruitt and thereby eliminate the error/noise term that was present in Baker and Wurgler’s index using the first principal component. Such an “aligned” investor sentiment index efficiently incorporates all of the relevant forecasting information from the proxies.
The aligned investor sentiment index addresses the core issue of forecasting asset prices over time and may have a number of applications across various streams of theory and practice of finance. It would be even more useful in emerging markets, which are characterized by less depth, information asymmetry, and restrictions on arbitrage.