Sentiment Indexes and Next-Month Stock Market Return

Do sentiment indexes usefully predict U.S. stock market returns? In his May 2018 doctoral thesis entitled “Forecasting Market Direction with Sentiment Indices”, flagged by a subscriber, David Mascio tests whether the following five sentiment indexes predict next-month S&P 500 Index performance:

1. Investor Sentiment – the Baker-Wurgler Index, which combines six sentiment proxies.
2. Improved Investor Sentiment – a modification of the Baker-Wurgler Index that suppresses noise among input sentiment proxies.
3. Current Business Conditions – the ADS Index of the Philadelphia Federal Reserve Bank, which combines six economic variables measured quarterly, monthly and weekly to develop an outlook for the overall economy.
4. Credit Spread – an index based on the difference in price between between U.S. corporate bonds and U.S. Treasury instruments with matched cash flows. (See “Credit Spread as an Asset Return Predictor” for a simplified approach.)
5. Financial Uncertainty – an index that combines forecasting errors for large sets of economic and financial variables to assess overall economic/financial uncertainty.

He also tests two combinations of these indexes, a multivariate regression including all sentiment indexes and a LASSO approach. He each month for each index/combination predicts next-month S&P 500 Index return based on a rolling historical regression of 120 months. He tests predictive power by holding (shorting) the S&P 500 Index when the prediction is for the market to go up (down). In his assessment, he considers: frequency of correctly predicting up and down movements; effectiveness in predicting market crashes; and, significance of predictions. Using monthly data for the five sentiment indexes and S&P 500 Index returns during January 1973 through April 2014, he finds that:

• Of 373 months in the test period starting January 1983:
  • The S&P 500 Index is up (down) for 233 (140) months. Accuracies of sentiment indexes in predicting direction range from 60% for Financial Uncertainty to 63% for Investor Sentiment.
  • Crash protection provides a different perspective. For example, Investor Sentiment is wrong about all eight down months during 2008, but Current Business Conditions and Financial Uncertainty are wrong about only two of these months.
  • Gross annualized return for the S&P 500 Index is 7.1%, with standard deviation 15.5%. Gross annualized returns for sentiment index portfolios range from 7.7% for Financial Uncertainty to 12.1% for Improved Investor Sentiment, with standard deviations ranging from 15.0% to 15.2%. Four of five sentiment index portfolios beat a dividend-adjusted benchmark.
  • Gross annualized Sharpe ratio (assuming zero risk-free rate) for the S&P 500 Index is 0.46. Those for sentiment index portfolios range from 0.51 for Financial Uncertainty to 0.81 for Improved Investor Sentiment.
  • Maximum drawdown for the S&P 500 Index is -53%. Those for the sentiment index portfolios range from 30% for Credit Spread to -53% for Investor Sentiment.
• Multivariate regression and, especially, LASSO combinations of sentiment indexes also outperform the S&P 500 Index.

In summary, evidence indicates that the selected sentiment indexes have potential to beat buying-and-holding the S&P 500 Index on a gross basis.

Cautions regarding findings include:

• The sample is somewhat stale, ending with 2014. Also, monthly updates for some sentiment indexes may not be available in as timely a manner as assumed in the paper.
• As noted in the paper, Investor Sentiment, Improved Investor Sentiment and Current Business Conditions indexes are recalibrated monthly with new releases to have zero mean. Such recalibrations change the historical series, introducing look-ahead bias, to the extent that an index value for a particular month may have different signs for different vintages. This recalibration effect disrupts use of these indexes for backtesting.
• All results are gross, not net. Costs of switching between long and short S&P 500 Index positions, and shorting costs while short, would reduce returns. S&P 500 Index futures offer a less costly way to switch, but futures returns likely differ from spot returns.
• Since the test period includes times of high inflation and interest rates, using a zero risk-free rate to calculate Sharpe ratios is problematic.
• Data collection and processing as described in the paper are beyond the reach of most investors and may be costly. Investors delegating these tasks to an investment advisor or fund would bear these costs plus management fees.
• Testing multiple indicators on the same sample of returns introduces data snooping bias, such that the best results overstate expectations. There may also be snooping bias derived from original discovery of the selected sentiment indexes.