Title: Forecasting the daily dynamic hedge ratios in emerging European stock futures markets: evidence from GARCH models
Authors: Taufiq Choudhry; Mohammad Hasan; Yuanyuan Zhang
Addresses: Southampton Business School, University of Southampton, UK ' Kent Business School, University of Kent, UK ' School of Securities and Futures, Southwestern University of Finance and Economics, Sichuan, China
Abstract: This paper empirically estimates and forecasts the hedge ratios of three emerging European and one developed stock futures markets by means of seven different versions of GARCH model. The seven GARCH models applied are bivariate GARCH, GARCH-ECM, BEKK GARCH, GARCH-DCC, GARCH-X, GARCH-GJR and GARCH-JUMP. Daily data during January 2000-July 2014 from Greece, Hungary, Poland and the UK are applied. Forecast errors based on these four stock futures portfolio return forecasts (based on forecasted hedge ratios) are employed to evaluate out-of-sample forecasting ability of the seven GARCH models. The comparison is done by means of model confidence set (MCS) and modified Diebold-Mariano tests. Forecasts are conducted over two non-overlapping out-of-sample periods, a two-year period and a one-year period. MCS results indicate that the GARCH model provides the most accurate forecasts in five cases, while each of the GARCH-ECM, GARCH-X and GARCH-GJR models constitutes model confidence set in four cases at a reasonable confidence level. Models selection based on modified Diebold-Mariano tests further corroborate results of the MCS tests. Differences between the portfolio returns also indicate the high forecasting ability of GARCH-BEKK and GARCH-GJR models.
Keywords: forecasting; hedge ratio; generalised autoregressive conditional heteroscedastic; GARCH; emerging market; volatility.
International Journal of Banking, Accounting and Finance, 2019 Vol.10 No.1, pp.67 - 100
Accepted: 21 Sep 2018
Published online: 16 Apr 2019 *