آیا هدفگذاری نرخ تورم بانک های مرکزی باید به به حرکات نرخ ارز پاسخگو باشد؟

This paper examines whether it is optimal for inflation-targeting central banks to respond to exchange-rate movements. The paper finds that exchange-rate movements can provide a signal on the developments in the economy that the central bank cannot perfectly observe. The results suggest that when the degrees of exchange-rate pass-through and international financial integration are high, it is optimal for the central bank to pay more attentions to exchange-rate movements. These results however depend on two conditions: 1) the ability of the central bank to observe the true exchange-rate process and 2) the number of real frictions in the model economy.

Over the past decade, one interesting trend in global financial markets has emerged. An increasing number of countries have abandoned exchange rate pegs to allow their currencies to float freely and adopted inflation targeting as their monetary policy framework.1 Such a switch has led these countries to uncharted territory of volatile exchange-rate fluctuations. From a theoretical standpoint, exchange-rate volatility may have a negative effect on economic welfare.2 From a policy perspective, excessive exchange-rate fluctuations may render macroeconomic instability.3 This raises a question whether inflation-targeting central banks, most of which rely on short-term nominal interest rates as their instrument, should respond to exchange-rate movements.4 One of the earliest discussions on this topic is offered by Obstfeld and Rogoff (1995) who point out that it may be advantageous for the central bank to ease in response to a substantial appreciation. Ball (1999) and Taylor (1999) provide the first two formal analyses on whether the central bank should respond to exchange-rate fluctuations. These studies compare macroeconomic outcomes generated by Taylor interest-rate rules that respond to exchange-rate fluctuations with those that do not respond to the exchange rate. Subsequent studies are extensions of the Ball–Taylor framework whereby Taylor interest-rate rules are evaluated on a variety of open-economy dynamic stochastic general equilibrium (DSGE) models.5 The results from these studies suggest that including the exchange rate in a Taylor interest-rate rule yields either no gains or fairly small improvements in terms of economic stability. In a recent contribution that is closely related to the present paper, Leitemo and Soderstrom (2005) [henceforth, LS] extend the Ball–Taylor framework by studying the gains from including the exchange rate in Taylor interest-rate rules when there is uncertainty regarding the exchange rate model. For a wide range of model specifications, their results confirm the findings in the literature that the gains from including the exchange rate in Taylor interest-rate rules are small and sometimes a separate response to the exchange rate can even reduce welfare. Furthermore, Taylor rules that respond to the rate of exchange rate depreciation are more sensitive to model uncertainty than the standard Taylor rules. The present paper is an attempt to examine the role of the exchange rate in the conduct of monetary policy. Rather than relying on Taylor interest-rate rules, I pursue a different approach by deriving optimal monetary policy under commitment. Under the goal of maintaining price stability and full employment, optimal monetary policy under commitment, which can be implemented via flexible inflation-targeting, yields the best outcome that the central bank is capable of implementing. While the analysis of optimal policy under commitment has been applied to various issues before, to the best of my knowledge, this is the first attempt at deriving the optimal reaction function to study whether it is optimal for the central bank to respond to exchange-rate fluctuations. 6 This paper also presents two other features that are relatively new in the literature of monetary analysis in open-economy settings and thus distinguish it from LS. First, this paper examines a scenario in which the central bank has an imperfect knowledge on the state of the economy. It has been generally agreed among academics and policymakers that modern-day central banks have to conduct monetary policy amid imperfect information and uncertainty.7 Note however that while LS focus the analysis on the full-information scenario as well as model uncertainty on the exchange-rate determination, this paper focuses on the analysis of partial information on the state of the economy in general. That is, the present paper examines the role of the exchange rate not only when the exchange rate process is uncertain but also when the inflation process is uncertain.8 Second, this paper examines optimal monetary policy in a large-scale open-economy DSGE model estimated by Bayesian techniques. The model features the export/import sectors, capital, money and net foreign assets, with several frictions including incomplete exchange-rate pass-through and imperfect capital mobility. LS, on the other hand, rely on a smaller model with fewer frictions and shocks. The present paper finds that under the optimal equilibrium, the nominal exchange rate fully depends on the state of the economy. Therefore, if the central bank can perfectly observe the state of the economy and can respond to the state of the economy directly, there will be no further gain from responding to the exchange rate. This result lends support to the findings in the literature that are derived under the full-information setting. However, if it becomes difficult for the central bank to perfectly verify the state of the economy, the exchange rate may provide information useful for the monetary-policy making process and in this case, there may be some benefits from responding to the exchange rate. The present paper also finds that the importance of the exchange rate in the policy-making process hinges on several conditions. First, when there is a high degree of uncertainty regarding the determination of the exchange rate, it will be optimal for the central bank to reduce weights given to the exchange rate in the rate-setting process and instead pay more attention to other indicators such as inflation and monetary growth.9 Second, when the exchange-rate pass-through is complete and when the degree of capital mobility is high, the central bank should pay more attention to exchange-rate movements. Finally, the presence of real frictions such as habit persistence and investment adjustment costs, limits the ability of domestically-related indicators, such as inflation, to provide clear signals on the dynamics of the economy and thus increases the importance of the nominal exchange rate in the rate-setting process.10 The remainder of the paper proceeds in the following manner. Section 2 sets up the model used for the analysis. Section 3 presents the policy problem of a central bank who can commit, and solves for optimal policy under commitment. Section 4 considers whether it is optimal for the central bank to respond to exchange-rate fluctuations under the full-information case. In Section 5, the assumption of full information is relaxed. Under the partial-information scenario, the central bank solves an optimal filtering problem, using exchange rates as an indicator. In Section 6, I conduct a sensitivity analysis to find out which assumptions in the paper's model are necessary to obtain the paper's results. Section 7 concludes.

This paper examines whether central banks under the regime of inflation targeting should respond to exchange-rate fluctuations. I find that under the commitment equilibrium, the nominal exchange rate is fully determined by the state of the economy. Hence if the central bank can perfectly observe the state of the economy and has already set the nominal interest rate to respond to the state of the economy, there will be no additional gain from responding to the exchange rate directly. Specifically, in an environment with full information in which the central bank can perfectly observe the state of the economy, all the information necessary for the monetary-policymaking process has already been contained in the state of the economy. Fluctuations in exchange rates do not reveal any extra information. This result is consistent with those derived in several other papers, including Ball, 1999 and Adolfson, 2001 and Batini et al. (2003). These papers examine the role of the exchange rate in the full-information environment and find very small or no improvements from including the exchange rate in policy rules. However, when the central bank cannot perfectly observe the state of the economy, movements in exchange rates can provide a signal on the developments in the state of the economy. In this case, it is optimal for the central bank to take into account exchange-rate movements in its rate-setting process. I also find that when the degrees of exchange-rate pass-through and capital mobility are high, the central bank should play even more attentions to exchange-rate fluctuations. The importance of the nominal exchange rate to the rate-setting process however depends on two critical conditions. First, the degree of uncertainty regarding the exchange-rate process must be small, relative to the uncertainty regarding the inflation process. Second, there must be a sufficient number of domestically-related real frictions included in the model economy.