An Empirical Analysis On The Interaction Effect Between Interest Rate And Exchange Rate Of US Dollars

Interest rate and exchange rate respectively represent the price of money on monetary market and foreign exchange market, and they both are important levers on adjusting economy and leading resource allocation, and also important tools on domestic and foreign monetary policies. There are both high theoretic and practical values of studying on the interaction effect of the two rates. However, at present, studies are still focus on the one-way effect from interest to exchange rate, rarely put emphasizes on the interaction effect of the two rates, and there exists limitation on the indexes, etc.This thesis chooses America which owes highly opened capital market and marketable forming mechanism of both interest and exchange rate as study object, and federal funds rate and nominal effective exchange rate as main indexes. Johansen co-integration test and vector error correction model are adopted to analyze the long-term equilibrium relationship between us dollars interest rate and exchange rate and also the short-term dynamic adjust process. Granger causality test, impulse response functions and variance decomposition are also used to make the study more overall and systemic.The empirical result indicates that there is obvious interaction effect between interest rate and exchange rate in America, and long-term equilibrium and mutual Granger cause relationship exists between them, and this relationship doesn't change with the influence of other economic factors; Although, on one hand, economic factors weaken the interaction effect between us dollars interest rate and exchange rate, on the other hand, they make the effect rapider. Expectation plays a remarkable role in the interaction effect, and it is even more remarkable than the effect which from interest changes to capital movement and asset exchange. And this indicates that people in America pay much attention on economy, at least they are sensitive on interest changes; There exists obvious time-lag on the effect from GDP to us dollars exchange rate, and the time-lag period is about 1 year; Government expenditure may influence the two rates obviously in short-term, American government, consequently, can employ fiscal policy to influence interest rate and exchange rate, furthermore, the author draws a figure of influence route based on the information supplied by the vector error correction model. The innovations of this thesis are as follows: First, innovation on choosing exchange rate index. The author chooses Nominal Effective Exchange Rate (NEER) which more roundly represents the true value of the currency as the study index, and adopts two criterions to choose the main companionate trade countries; Second, innovation on empirical methods. Impulse response functions and variance decomposition are adopted, and the response to each other's innovation and the contribution to each other's forecasting error are consequently studied.