CMStatistics 2020: Start Registration
View Submission - CFE
Title: Can we forecast better in periods of low uncertainty? The role of technical indicators. Authors:  Sam Pybis - Manchester Metropolitan Universty (United Kingdom) [presenting]
Olan Henry - University of Liverpool (United Kingdom)
Maria Ferrer Fernandez - University of Liverpool (United Kingdom)
Michalis Stamatogiannis - University of Liverpool Management School (United Kingdom)
Abstract: We examine the importance of periods of high versus low financial uncertainty when forecasting stock market returns with technical predictors. Our results suggest that technical predictors perform better in periods of low financial uncertainty and should be avoided due to poor forecasting performance in periods of heightened uncertainty. In-sample, we report disentangled R^2 statistics, and out-of-sample we show these results continue when forecasting the equity risk premium. We show similar results when forecasting the volatility of returns with technical predictors. We measure periods of heightened and low financial uncertainty in a regime switching framework. Overall, our results provide insight into the mechanism that suggests that, when uncertainty rises, investors' opinions polarize leading to a breakdown of predictability based on technical indicators.