Forecasting Market Direction with Sentiment Indices

David Mascio, PhD

Abstract :

Successful market timing strategies depend on superior forecasting ability and the accuracy of market forecasts. We use six predictive models to forecast the S&P 500 Index (SPX) consisting of investor sentiment, current business conditions, economic policy uncertainty, market dislocation information, credit spreads, and financial uncertainty. These indices are combined to create two additional forecast models, a “kitchen sink logistic regression” and a “least absolute square shrinkage and selection operator.” Each model and the combined models are used in a logistic regression analysis to predict the one-month ahead returns of the SPX. In order to determine how successful each strategy is at forecasting the market direction; each prediction is used to adjust the beta of the portfolio. “Beta optimization” refers to a strategy designed to create a portfolio with a beta of 1.0 when the market is expected to go up, and a beta of -1.0 when a bear market is expected. Successful application of this strategy generates returns that are consistent with a call option or an option straddle position; that is, positive returns are generated in both up and down markets. Analysis reveals that the models’ forecasts have discriminatory power in identifying substantial market movements, particularly during the bursting of the tech bubble and the financial crisis. We determine the individual forecast indices and the combined portfolios consistently have higher annual returns and lower monthly drawdowns than the buy-and-hold SPX portfolio, and the benchmark index, Baker and Wurgler (2004) Value-Weighted Dividend Premium (VWDP) model.

Publication date of the thesis

Thesis committee

Supervisor: Frank Fabozzi, EDHEC Business School

External reviewer: Turan G. Bali, Georgetown University

Other committee member: Abraham Lioui, EDHEC Business School