نقش سیاست های پولی در رفتار نرخ ارز

We address the role of monetary policy shocks in exchange rate behavior using an inference procedure that allows us to relax dubious identifying assumptions. We find: (i) The peak exchange rate response may be delayed or nearly immediate; (ii) In every otherwise reasonable identification, monetary policy shocks lead to large uncovered interest rate parity (UIP) deviations; (iii) Monetary policy shocks may account for a smaller portion of exchange rate variance than found earlier. While (i) is consistent with overshooting, (ii) implies that the overshooting cannot be driven by Dornbusch's mechanism, and (iii) casts doubt on monetary policy shocks being the main source of exchange rate volatility.

Exchange rate changes are volatile and difficult to explain. Economists have long suspected that monetary policy shocks might play an important role in accounting for this behavior, and a great deal of theoretical and empirical work has been directed at confirming this suspicion. This paper combines recent developments in international finance and econometrics to assess what firm conclusions can be drawn about the role of monetary policy shocks in exchange rate behavior. An obvious starting point for our study is Dornbusch's (1976) overshooting model. Having received over 800 citations,1 this work remains at the core of international finance. Dornbusch's prediction that the exchange rate should initially overshoot its long-run level in adjusting to a monetary shock owes much of its huge appeal to two factors. First, it provides hope of explaining the empirical regularity that exchange rates in the post-Bretton Woods era are more volatile than macroeconomic fundamentals such as the money supply, output, and interest rates. Second, the overshooting conclusion follows directly from three familiar components: the liquidity effect of monetary policy shocks on nominal interest rates, uncovered interest rate parity (UIP), and long-run purchasing power parity (PPP). While overshooting is a dominant theory in international finance, its reliance on uncovered interest rate parity means that when confronted with data, the theory will be enmeshed in a dominant empirical puzzle in international finance—the tendency of the exchange rate to change in the direction opposite to that predicted by UIP. Labelled the forward premium anomaly, this tendency has been extensively documented ( Fama, 1984; Hodrick, 1987; Engel, 1996). Nevertheless, if monetary policy shocks have small effects on currency risk premia, as many models predict, then the Dornbusch story could hold conditionally in response to monetary policy shocks, even if it fails unconditionally. Motivated by these facts, we focus primarily on whether or not the exchange rate overshoots. More specifically, at what lag horizon does the exchange rate peak after a U.S. monetary policy shock? We also consider two related questions: Is the dynamic response of the exchange rate roughly consistent with uncovered interest rate parity? Can U.S. monetary policy explain a large share of exchange rate variance under any reasonable theory? The questions regarding peak timing and conditional UIP shed light on whether the data are consistent with Dornbusch overshooting. The final question addresses whether monetary policy shocks can account for a large share of exchange rate variance under overshooting or any other theory of international finance. Several papers examining the first question (Clarida and Gali, 1994; Eichenbaum and Evans, 1995; Grilli and Roubini, 1996) find that the exchange rate overshoots its long-run value in response to monetary policy shocks, but that the peak occurs after one to three years as opposed to happening immediately as predicted by Dornbusch.2 A typical “delayed overshooting” result is shown in Fig. 1, which gives the estimated dynamic response of the U.S. dollar/U.K. pound and dollar/German mark exchange rates to a stimulative U.S. monetary policy shock in our replication of work by Eichenbaum and Evans (1995). Based on such evidence, a consensus seems to be emerging that the exchange rate shows delayed overshooting and theorists are attempting to rationalize this fact (e.g., Gourinchas and Tornell, 1996).While few papers directly address the conditional UIP question, most find large deviations from UIP. Eichenbaum and Evans (1995), Cushman and Zha (1997) and Kim and Roubini (2000) report that policy shocks generate deviations from UIP that are several times larger than the generated interest rate differential.3 As for the share of exchange rate variability due to monetary policy shocks, papers report a wide range of estimates, between a few percent to over one-half, with little apparent consistency across countries or identification strategies (see Clarida and Gali, 1994; Eichenbaum and Evans, 1995; Rogers, 1999; Kim and Roubini, 2000).4 In all of this work, a highly contentious step is identifying which exchange rate movements are due to monetary policy shocks. All of the literature cited above identifies the policy shocks using the identified vector autoregression approach, currently the dominant approach to identification in the literature. The reason identification is so contentious is that there are few highly credible identifying assumptions. Each of the papers above relies on some assumptions that are dubious at best—a fact that the authors acknowledge. Some papers assume that foreign interest rates do not respond to Fed policy moves until a month after they are made (Eichenbaum and Evans, 1995; Kim and Roubini, 2000). This is inconsistent with the striking movements in foreign rates that regularly follow within minutes of Fed policy announcements. Others further assume that the Federal Reserve ignores any surprising movements in exchange rates and short-term U.S. Treasury Bill rates that have occurred during the month in which they are making their decisions.5 If literally true, this would be quite disappointing to those of us who inform the Federal Reserve Board on a real-time intraday basis regarding surprising movements in financial markets. This paper applies an approach developed by Faust (1998) to more thoroughly assess the robustness of identified VAR conclusions to changes in dubious assumptions. This approach allows one to suspend any dubious assumptions, while maintaining any highly credible assumptions. After suspending the dubious assumptions, one can examine all possible alternative identifications to see if the conclusion is an artifact of the dubious assumptions. It is important to emphasize that this is not a new approach to identifying a policy shock. Rather it is an approach to more thoroughly and completely doing what VAR papers have always done: checking whether important results change when dubious assumptions are suspended. We apply this technique to a standard 7-variable model and a new 14-variable model for both the US–UK and US–German bilateral exchange rates. We find the following: 1. The delayed overshooting result is quite sensitive to dubious assumptions. It is straightforward to find U.S. monetary policy shocks that are reasonable by the standards usually applied in the literature, but in which the peak exchange rate effects are very early (say, within a month after the shock) or very late. The primary reason these have not been found before is that they seem to rely on allowing some simultaneity among interest rates and exchange rates. 2. The conclusion that monetary policy shocks generate large UIP deviations is quite robust. Even when one suspends dubious assumptions and searches all possible ways to complete identification of the U.S. monetary policy shock, it is impossible to find a shock that generates small UIP deviations. Thus, if exchange rates do peak early in response to policy shocks, this overshooting is apparently not UIP-driven as in Dornbusch (1976). 3. The share of exchange rate variance due to the U.S. monetary policy shock is not sharply identified; reasonable estimates go from about zero to over half. This result is fully consistent with earlier work that has reported a similarly large range. We do find, however, that the upper bound in our 14-variable models is somewhat smaller (about 30 percent). Faust (1998) and Leeper et al. (1996) emphasize that including more variables often changes the conclusions from identified VAR work. Thus, an important advantage of the approach used here is that it can be applied in larger models. Our third result suggests that model size may matter. Our results are developed in 5 sections. In 1 and 2, we discuss relevant international finance theory and then an example of our approach to identification. Section 3 lays out the full approach, Section 4 has results, and Section 5 presents conclusions.

Empirical work on the role of monetary policy shocks in explaining exchange rate behavior is impeded by the lack of fully credible identifying assumptions. This paper applies an inference approach that allows us to test the robustness of existing conclusions to relaxing dubious assumptions and to changes in the number of variables in the model. We find that the delayed overshooting result is sensitive to dubious assumptions. This conclusion comes from loosening the standard assumption of recursiveness in money market variables to allow plausible simultaneity. We also find that monetary policy shocks generate large expected root mean square UIP deviations. Even when imposing very little on the behavior of the money shock, we are unable to find policy shocks that generate interest rate and exchange rate responses roughly consistent with UIP. Finally, the results suggest that U.S. monetary policy shocks may explain less of the observed exchange rate variability than previously believed. In our 7-variable model, policy shocks can account for much of the variance of the exchange rate; in the 14-variable model, we find it highly unlikely that U.S. policy shocks account for more than one-third of exchange rate variance. These results have important implications for what stylized facts theorists should be attempting to explain and they present a mixed bag for theorists hoping that relatively conventional theories will do the trick. The results allow for an early peak in the exchange rate, which might give a role for the conventional overshooting model. Unfortunately, the bulk of the variance of the exchange rate after policy shocks is due to large deviations from UIP. This is inconsistent with Dornbusch overshooting, and, indeed, no conventional models we are aware of generate large and variable foreign exchange risk premia in response to policy shocks that have the modest effects on output and interest rates that we find. Perhaps models in which large ex post UIP deviations arise from information problems offer greater hope.