Do weak predictability of technical indicators still enhance forecasting power for the Korean stock market returns?

초록

Recent empirical studies suggest that technical indicators offer substantial predictive power for future equity market risk premiums and can effectively complement forecasting models based on macroeconomic variables. This paper assesses the predictive power of technical indicators along with macroeconomic variables, examining the extent to which technical indicators enhance forecasts of future equity risk premiums in the Korean market. We evaluate the performance of individual variables and groups of macroeconomic variables and/or technical indicators by parsimoniously extracting principal components from these variables. Using monthly data from October 2000 to December 2023, we conduct both in-sample and out-of-sample predictability tests to provide the economic implications of principal component predictive regressions. Contrary to prior findings, our results reveal that technical indicators in Korea exhibit little or weak predictive power at a monthly frequency. Nevertheless, integrating information from technical indicators with macroeconomic variables substantially improves the ability to predict the Korean market’s equity risk premium. The in-sample R^2 statistic for a predictive regression based on principal components extracted from all variables significantly surpasses any other in-sample R^2 statistic obtained from individual variables or principal components extracted solely from either macroeconomic variables or technical indicators. These in-sample results are supported by out-of-sample testing, leading to the conclusion that incorporating information from technical indicators into macroeconomic variables enhances the predictability of the equity market risk premium in Korea.