Exchange rates or stock prices, what causes what: A firm level empirical investigation

The study employs cointegration, the standard Granger causality tests and vector error correction modeling technique to investigate the cause-effect association between exchange rates and stock prices for Pakistan. It uses weekly data for 70 individual securities and the trade-weighted exchange rate over the span from January 1, 1999 to March 31, 2004. The results of cointegration tests show that there is no co-movement between the said variables for most of the examined firms. On the issue of causation, the evidences are mixed. In some cases causation runs from stock prices to exchange rate whereas for some firms’ stock prices are affected by the changes in trade-weighted exchange rate. However, the analysis findings are generally supporting the asset market approach to exchange rate determination that reports no link between the said variables.