Cointegration and Granger Causality Tests of Stock Price and Exchange Rate Interactions in Malaysia

Abstract

I. Introduction Recent high exchange rate fluctuations in sphere of growing trade and financial liberalization have attracted a great deal of interest from both economists and policy-makers. Of particular interest has been question of domestic economies' exposure to exchange rate risk. It is noted that fluctuations in exchange rate can substantially affect values of firms, through changes in terms of competition, changes in input prices, and changes in values of foreign currency-denominated assets (Bodnar and Gentry 1993). Domestic firms with foreign operations are obviously affected directly and, given these various channels of influences, even firms with no foreign operations may be indirectly affected. Accordingly, firms' stock prices and stock market may react to changes in exchange rates. Conversely, changes in stock prices may influence movements in exchange rate, via firms' portfolio adjustments (Bahmani-Oskooee and Sohrabian 1992) or outflows of capital (Qiao 1996). The latter is likely to be operative if changes in stock prices are sufficiently persistent to generate or destroy confidence in market. Empirically, there are quite a number of studies that attempt to determine impact on stock prices of exchange rate changes. The findings, however, are not uniform across various studies. Some studies documented positive effects of exchange rate changes on stock market (Aggrawal 1981), while others found negative effects (Soenen and Hennigar 1988). Yet other studies concluded that exchange rate changes have no significant impact on stock market (Solnik 1984). In a more recent study, Bahmani-Oskooee and Sohrabian (1992) evaluated interactions between Standard and Poor's Composite Index and effective exchange rate of dollar, using monthly observations from July 1973 to December 1988. Applying standard co-integration and Granger tests, they found bi-directional causality between stock price index and exchange rate. However, there was no long-run relationship between two variables. Using daily data for cases of Japan, Hong Kong and Singapore, Qiao (1996) found stock price-exchange rate causal nexus to be different across countries. Specifically, direction of causation is bi-directional for Japan, is unidirectional from exchange rate to stock returns for Hong Kong, and is non-causal for Singapore. He also noted presence of a strong long-run relationship in these three countries. Recently, Abdalla and Murinde (1997) investigated same issue for emerging markets of India, Korea, Pakistan and Philippines. They noted that understanding of issue is crucial, as these countries attempt to develop their financial markets and, at same time, move towards more flexible exchange rates. Their analysis is based on monthly observations from January 1985 to July 1994, using IFC stock market indices for countries and real effective exchange rates. The results from their studies suggested unidirectional causality from exchange rates to stock prices in all countries, except Philippines. In case of Philippines, Abdalla and Murinde found stock price to Granger caused exchange rate. Additionally, they documented presence of a long-run relationship for cases of India and Pakistan. These results led them to conclude that, for cases of unidirectional causality from exchange rates to stock prices, the respective governments of these emerging markets should well be cautious in their implementation of exchange rate policies. The objective of this study is to extend existing studies on stock price-exchange rate causal relationship by investigating issue for another emerging market, Malaysia. The country is one of open and fast-growing economies in region. The development of its stock market is also Exceptional. …