Modelling the joint dynamics of financial assets using MGARCH family models: Insights into hedging and diversification strategies

Abstract

This study analyzes the conditional correlations and thereafter constructs the optimal portfolios for G-12 countries’ stock market returns and national benchmark bonds, crude oil, gold and the volatility index (VIX) returns. We use daily data ranging from January 1, 1994 to May 3, 2021 and compare the DCC, ADCC and GO-GARCH methods to investigate the past shocks and volatility transmissions. The rolling estimation techniques are employed to construct one-step-ahead forecasts of the dynamic conditional correlations and the optimal hedge ratios of G12 markets and other variables. For most of the situations studied, the volatility index (VIX) generates the best effective hedge to stock returns for these markets. The national benchmark bond indices create the second-best hedge. The risk and downside risk measures suggest that a sole stock exhibits the greatest risk and the expected maximum loss compared to a mixed bond-stock, a mixed VIX-stock, or a mixed gold-stock portfolio. The results are robust to alternative modeling specifications, model selection, as well as distributional assumptions.