Stock Price and Exchange Rate Dynamics in an Emerging Economy: Evidence from Bangladesh

Abstract

This study examines stock price and exchange rate dynamics in Bangladesh by utilizing the daily data of stock price and exchange rate of BDT/USD from June 2003 to December 2016. The study period has been divided into two sample periods; one from June 02, 2003, to January 25, 2013, and the other from January 28, 2013, to December 30, 2016. The sample period one (June 02, 2003, to January 25, 2013) represents the period with more variance for both stock price and exchange rate because within this period different turbulent events have occurred (e.g., world economic meltdown, stock market crash etc.) while sample period two (January 28, 2013, to December 30, 2016) represents less variance for both stock price and exchange rate. In both periods it is observed that stock market has more variability than that in exchange market. This study divides the analysis into three parts; one is individual efficiency analysis in terms of random walk hypothesis through different parametric and non-parametric tests for both stock market and exchange market, the second one is joint efficiency investigation through cointegration and causal relationship (long and short-run relationship) between them and the final one is the volatility dynamics in within and between those two markets. The individual efficiency analyses confirm the no randomness for stock price and exchange rate in both sample periods. Therefore, stock market and foreign exchange (FX) market are individually inefficient in Bangladesh. Joint Efficiency investigation presents no long-run relation between stock market and FX market over the period of greater variance, but a unidirectional long-run relationship has been found running from stock market to FX market over the period of lower variance. On the other hand, bidirectional nonlinear causal relationship (stock price leading the relationship) is found in high variance period; but no linear or nonlinear short-run relationship exist in low variance period. Thus, the evidence of connectivity between stock market and FX market denies joint market efficiency in Bangladesh. Finally, the volatility dynamics within and between stock and FX market has been investigated for both periods. The volatility dynamics within both markets approves significant past news effect and asymmetric news effect and volatility clustering within each market. The volatility dynamics between financial markets indicates volatility spillover (cross-market relationship) between stock market and FX market. It confirms the return spillover from stock market to foreign exchange market where it is negative for high variance period and positive for low variance period. Thus, portfolio balance approach is observed, that is, past stock return has significant explanatory power over conditional mean of FX return. Volatility spillover investigations confirm negative bidirectional volatility spillover phenomena explaining Government intervention and behavior of foreign investors in the exchange market for both periods with less and more variance. The study also provides the evidence of effectiveness of cross market positive news in reducing the volatility in the financial market. These observations affirm the absence of weak form joint efficiency, that is, information of one market can be used to predict another market. In general, the thesis contributes to the existing literature by confirming how the level and direction of connection between stock market and FX market changes with the extent of volatility in the sample period in Bangladesh. It also presents how information contents of the stock market affect FX market in Bangladesh and vice versa. Overall, the study enriches the knowledge of functionality, volatility of and between stock and FX markets and also has implications for investors and regulators.