Bayesian analysis of dynamic linkages among gold price, stock prices, exchange rate and interest rate in Pakistan

Understanding of complex relationships among economic variables has much significance for investors, researchers and policy makers alike. This study is conducted to examine dynamic linkages among gold price, stock prices, exchange rate and interest rate nexus using monthly data of Pakistan economy ranging from 2001-1 to 2014-12. Search of the best model motivates the use of Bayesian inference. Hence, the study is also used to compare performance of classical VAR model and Bayesian VAR model under four types of priors. It may be considered another significant contribution in terms of methodological framework. All the four variables are nonstationary at level but stationary at first difference. However, no long-run relationship among the variables is found by employing three cointegration tests, i.e. JJ test without considering structural break, GH test considering one unknown structural break and FLY test allowing two unknown structural breaks. Hence short-run relationship is analyzed in the study. Bayesian VAR model under independent normal inverse Wishart priors is selected as the best model which is then used to conduct impulse response analysis. Inverse bilateral relationships between stock prices and gold price as well as between rupee value and gold price are determined where as positive bilateral relationship between stock prices and rupee value is explored. It implies that stock prices and rupee value move downward during recessionary periods but gold becomes more pretty and vice versa also hold. Hence, gold is not only considered as safe haven but it is also considered as an alternative investment during adverse fluctuations in stock and foreign exchange markets of Pakistan. Changes in stock prices and rupee value are negatively responded by monetary policy makers in Pakistan but changes in gold prices are not considered by monetary authorities. Results also reveal that monetary policy actions significantly and adversely affect the three markets under consideration. Direct relationship between nominal interest rate and exchange rate in Pakistan conform to the International fisher effect theory.