| Summary: | 碩士 === 國立臺灣科技大學 === 管理研究所企業管理學程 === 88 === Monetary policy, international trade and macroeconomic conditions are important variables for research about stock market. I attempt to find the interactions between 4 group stock indexes and 25 macroeconomic variables both in short and long term.
The sample period for this study consists of 108 months during 1991/01-1999/12. There are four steps in this study. First is to test the stationary of all variables by ADF unit root test. Second is to use the stepwise regression to find the macroeconomic variables which have significant relations with 4 group stock indexes. Third is to adopt the cointegrated analysis to confirm the long-term trends between variables. Final is to use ECM model to find the leading indicators for group stock indexes.
The data fits the stationary after transforming to first difference style. All of the fourteen pairs exist significantly cointegrated relations. The empirical results of ECM model are as following:
1. The average wage in manufacturing industry and M1B are good leading indicators of Plastics & Chemical.
2. M1B is a one-term leading indicator of Paper & Pulp and Banking at the same time, and three-term leading indicator of Electrical.
3. Because import price index, exchange rate (NT/USD), and the average wage of each person in manufacturing industry have two-way interactions with Electrical significantly, they are effective indicators for Electrical in the short term.
4. The average wage in manufacturing industry and M1B have not only one-way but also two-way interactions with the four Group Stock Indexes.
5. There are four pairs of variables which don''t have significant interactions in the short term, and three of them aren''t chosen from stepwise regression.
6. The relations between Electrical and macroeconomic variables are more significant. Paper & Pulp and Banking are only affected by M1B one-term lagged, so they aren''t influenced by macroeconomic variables relatively.
|
|---|
