اثرات شوک سیاست پولی در نرخ ارز: یک روش مدل اصلاح خطای برداری ساختاری

This paper investigates the effects of shocks to US monetary policy on the dollar–yen exchange rate, using structural Vector error correction model (VECM) methods with long-run restrictions. We compare our estimates of the impulse responses with those based on levels Vector autoregression (VAR) with standard recursive order restrictions. The empirical results based on the long-run restrictions are found to be more consistent with standard models of exchange rate determination than the results based on the recursive order restrictions. J. Japanese Int. Economies18 (1) (2004) 99–114.

This paper examines the effects of shocks to US monetary policy on the dollar–yen exchange rate, using structural Vector error correction model (VECM) methods with long-run restrictions. We compare our estimates of the effects with those of Eichenbaum and Evans (1995) based on levels Vector autoregression (VAR) with standard recursive order restrictions. The standard exchange rate model (see, e.g., Dornbusch, 1976) predicts that a contractionary shock to US monetary policy leads to appreciation in US nominal and real exchange rates. However, empirical evidence for two important building blocks of the model is mixed at best. These two building blocks are Uncovered interest parity (UIP)1 and long-run Purchasing power parity (PPP).2 Therefore, it is not obvious whether or not this prediction of the model holds true in the data. Eichenbaum and Evans (1995) directly investigate this prediction by estimating impulse responses to US monetary shocks and find evidence in favor of the prediction, even though their results do not support some aspects of the standard exchange rate model. In order to investigate the impulse responses to a monetary policy shock, it is necessary to identify the shock by imposing economic restrictions on an econometric model. When economic restrictions are imposed, the econometric model is called a structural model. Both the choice of the econometric model and the choice of the set of restrictions can affect the point estimates and standard errors of impulse responses. For this reason, it is important to study how these choices affect the results. Most variables used to study exchange rate models are persistent, and usually modeled as series with stochastic trends and cointegration. In such a case, both levels VAR and the VECM can be used to estimate impulse responses. Levels VAR is more robust than the VECM because it can be used even when the system does not have stochastic trends and cointegration. Perhaps for this reason, it is used in most studies of impulse responses and by Eichenbaum and Evans (1995). However, the structural VECM has some important advantages in systems with stochastic trends and cointegration. First, other things being equal, estimators of impulse responses from the structural VECM are more precise. For example, levels VAR can lead to exploding impulse response estimates even when the true impulse response is not exploding. This possibility is practically eliminated with the structural VECM. Second, it is possible to impose long-run restrictions as well as standard recursive order restrictions to identify shocks. Long-run restrictions are attractive for the purpose of this paper because they are more directly related to macroeconomic models. For example, one of the recursive order assumptions used by Eichenbaum and Evans is that the exchange rate is not included in the Fed's information set when it makes monetary policy decisions. Such an assumption is not found in standard exchange rate models. The assumption is also not very realistic because exchange rate data are available to the Fed up to the minute as pointed out by Faust and Rogers (2000).3 In contrast, long-run restrictions used in this paper such that a monetary policy shock does not affect the real exchange rate and real aggregate output in the long run, are commonly found in the standard exchange rate models. A method of imposing long-run restrictions on the VECM is developed in King et al. (1991) (KPSW for short). This paper employs Jang's (2001a) method rather than the KPSW method.4 Compared to the KPSW method, Jang's method has an advantage in that it requires neither identification nor estimation of each cointegrating vector. This method greatly facilitates the impulse response analysis because identification assumptions for cointegrating vectors can be complicated, and may be inconsistent with some long-run restrictions a researcher wishes to impose to identify shocks. Another feature of Jang's method is that it applies block recursive assumptions to the structural VECM with long-run restrictions. The block recursive system has been well developed in structural VAR and the structural VECM with short-run restrictions, yet has not been studied in the structural VECM with long-run restrictions.

This paper examined the effects of shocks to US monetary policy on the dollar–yen exchange rate and other economic variables, using structural Vector error correction model (VECM) methods with long-run restrictions. We compared our estimates of the impulse responses with those based on levels VAR and recursive order restrictions. An advantage of the VECM compared with the levels VAR is that long-run restrictions can be imposed to identify shocks in the VECM. One of the recursive order restrictions used in Eichenbaum and Evans (1995) is that the exchange rate is not in the Fed's information set. As pointed out by Faust and Rogers (2000), the recursive ordering is not attractive for the purpose of studying effects of monetary policy shocks on exchange rates. It is unlikely that the exchange rate is not included in the Fed's information set given that the data are available up to the minute. It is also unlikely that the exchange rate is predetermined with respect to the monetary policy shock because data for changes in the federal funds rate are available to many exchange rate market participants. Thus, if the monetary policy and exchange rates are simultaneously determined contemporaneously, then the recursive ordering assumptions are not satisfied. In contrast, identification with long-run restrictions does not suffer from this simultaneity problem because no zero restrictions are imposed on . The empirical results based on the long-run restrictions are much more consistent with the standard exchange rate model than those based on the recursive assumption. The VECM with long-run restrictions yields an earlier peak of overshooting than the levels VAR, the VECM results do not yield the price puzzle as the levels VAR do, and the monetary policy shock attributes more to the exchange rate fluctuation in the VECM forecast-error variance decomposition than that it does in the levels VAR. These results are consistent with the view that the simultaneity problem for the exchange rate and the Fed's policy is important.