Dynamic Association and Granger Causality between International Commodities Prices, Inflation, Exchange Rate, and Stock Performance: Evidence from Pakistan

Abstract

The purpose of this study is to determine the time varying conditional correlations and Granger causality between four variables: international commodities prices, inflation, exchange rate and stock returns with reference to Pakistan. In this research monthly data of variables is taken from different sources for a period of 23 years from January 2000 to December 2022. In order to determine the time varying conditional correlations DCC GARCH model is applied and to find causality between variables Granger causality method is used. The results of the study are significant which shows that the dynamic conditional correlation exists between all the pairwise variables, which mean that with time the correlations between the pairwise variables will change. Study results also shows that the β coefficients of DCC-GARCH model are significant in all pairwise variables which mean that external shocks can affect the correlations and in long run the correlations between the variables can vary. Secondly we applied Granger causality model in order to determine the causality between variables. Granger causality results are also significant and according to past literature, which shows that unidirectional causality exists from international commodities prices to inflation, inflation to stock returns and from exchange rate to stock returns. Results also show the bi-directional causality exists between inflation and exchange rate. No Granger causality found between international commodity prices and exchange rate and between international commodity prices and stock returns. These results are very important and valuable for the policy makers, investors and businesses. Correlations between asset classes are very important at the time of investment, so the result of DCC-GARCH models helps the investors at the time of portfolio management. This study also put light to understand the behavior of dynamic correlation during financial crises 2007-2009 and during COVID-19 perio