This presentation introduces machine learning techniques to investigate whether popular macroeconomic or sentiment factors are better at predicting stock market returns. We find that although either macroeconomic or sentiment variables alone fail to improve the Sharpe ratio of the stock market, combining the factors improves the Sharpe ratio from 0.48 to 0.62 and reduces the investment drawdowns by roughly 30% from 53 percentage points to 36 percentage points. This improvement is significant in both economic and statistical terms. We further evaluate the performance of strategies across business cycle and find that macroeconomic variables tend to outperform sentiment variables during market expansions and underperform during recessions. The combined performance of the macroeconomic and sentiment variables is particularly strong during the late stage of recessions when the stock market is close to its bottom. Our finding is robust to the choice of machine learning technique and indicates that sentiment and macroeconomic information is complementary and, therefore, should be considered jointly by investors.
David Mascio, PhD, is the Founder & Chief Executive Officer of Della Parola Capital Management. He is also an Endowed Chair and the Roland and Sarah George Professor of Applied Finance at Stetson University. Over the past 20 years, Dr. Mascio has served as a university professor, a hedge fund manager, keynote speaker on economic forecasting and the chief investment officer of a billion-dollar trust bank. He has also been published in top academic journals in the area of machine learning in economic and investment forecasting.
Dr. Mascio earned a B.A. in Economics and Business Management from the University of New Mexico, he also earned an MBA from the University of Liverpool (United Kingdom) and a PhD in Finance at EDHEC Business School (Nice, France). He is an accredited Asset Management Specialist (AAMS), and a member of the CFA Institute and CFA society of Orlando.