Volatility Spillovers between Gold Prices, Oil Prices, and the Islamic Stock Market in Pakistan: Evidence from DCC-GARCH and Diebold–Yilmaz Connectedness Analysis

Abstract

Over the past several years, financial markets and commodity markets have become increasingly interconnected, with shocks in one sector affecting the dynamics of others. As these relationships continue to evolve, understanding how volatility moves from one asset class to another has become increasingly important for making investment decisions, managing risk, and maintaining financial stability. This study investigates the volatility spillovers between real assets and the Islamic stock market in Pakistan by examining the dynamic relationships among KMI-30 Index returns, gold returns, and crude oil returns using weekly data from 2016 to 2025. The EGARCH model is applied to capture volatility persistence and asymmetric market behavior, while the DCC-GARCH model is used to estimate the time-varying correlations among the selected markets. Moreover, the Diebold–Yilmaz spillover index is employed to assess the direction and magnitude of volatility spillovers. The empirical results show that oil returns have a significant association with KMI-30 conditional volatility, whereas gold returns have no statistically significant direct effect in Pakistan. The estimated DCC parameters indicate that the relationships among the KMI-30 Index, gold and oil returns are dynamic and highly persistent over the sample period. However, the spillover analysis reveals that gold acts as a net transmitter of volatility within the connectedness network, while the KMI-30 Index and crude oil act as net receivers. The total connectedness index indicates a relatively low-to-moderate level of market interconnectedness, with volatility spillovers present across the selected markets, and scope for diversification benefits. These results provide valuable guidance to investors, portfolio managers, and policymakers in developing risk management and portfolio diversification strategies.