Using technical rules to enhance the predictability of the standard GARCH model for the volatility of stock indices

Abstract

The main purpose of this study is to explore whether technical analysis can improve the predictability of Generalized Autoregressive Conditional Heteroscedasticity (GARCH(1 1)) for the volatility of stock index Four technical analysis categories filter rule support and resistance channel breakout and moving average were applied creating 1 107 rules in total and the rules were used individually on the realized volatility of stock index to obtain signals We discuss if the GARCH(1 1)-augmented model with technical analysis signals can provide better predictability for stock index volatility than the benchmark model(GARCH(1 1)) Employing the mean absolute error (MAE) and mean squared error (MSE) as performance measure we conduct Hsu Hsu and Yen’s (2014) Step-SPA(k) test to control for the data snooping bias Dow Jones Industrial Average data from 2008 to 2012 and 1993 to 1992 are used as the sample Our results showed that analyzing realized volatility signals with technical rules does not lead to a significant improvement in the predictability of the GARCH(1 1) model for volatility of stock indices

Date of Award	2015 Jul 28
Original language	English
Supervisor	Meng-Feng Yen (Supervisor)