گشودگی،عبور ناکامل از طریق نرخ ارز و سیاست پولی

This paper analyses the implications of imperfect exchange rate pass-through for optimal monetary policy in a linearised open-economy dynamic general equilibrium model calibrated to euro area data. Imperfect exchange rate pass through is modelled by assuming sticky import price behaviour. The degree of domestic and import price stickiness is estimated by reproducing the empirical identified impulse response of a monetary policy and exchange rate shock conditional on the response of output, net trade and the exchange rate. It is shown that a central bank that wants to minimise the resource costs of staggered price setting will aim at minimising a weighted average of domestic and import price inflation.

Over the last 6 years a large literature (the so-called “New Open Economy Macroeconomics, NOEM”) has developed examining the optimal conduct of monetary policy in a class of open-economy dynamic general equilibrium models that feature imperfect competition and nominal rigidities.1 One of the models that has recently attracted a lot of attention is the one of Gali and Monacelli (2000). This model combines the open-economy features of the NOEM, with the elegance of the benchmark New-Keynesian closed economy model as, for example, analysed in Woodford (1999). One of the striking findings in Gali and Monacelli (2000) is that the welfare results obtained in the basic New-Keynesian model carry over to its open-economy counterpart. Welfare optimising monetary policy results in a complete stabilisation of the domestic price level. In particular, there is no trade-off between output gap stabilisation and domestic price stability and there is no need for an explicit consideration of the exchange rate.2 This result has proven to be relatively robust with respect to certain extensions of the model. For example, in a two-country set-up Benigno and Benigno (2001) have shown that a policy pursuing domestic price stability can be considered as the optimal outcome in a Nash game between the monetary authorities in two countries. Similarly, Obstfeld and Rogoff (2002) have rejected the necessity of a new international compact on the basis of the argument that policies geared at domestic price stability deliver outcomes that are close to the first best. In another extension, Benigno (2001) shows that achieving domestic price stability continues to characterise the optimal monetary policy when international financial markets are incomplete. One feature that characterises all the models discussed above is the assumption of perfect exchange rate pass-through. There is, however, a lot of empirical evidence that changes in nominal exchange rates affect import prices only gradually. Recently, Campa and Goldberg (2001) estimated pass-through equations for 25 OECD countries over the period 1975–1999. They find that they can reject the hypothesis of complete short-run pass-through in 22 of the 25 countries. In contrast, long-run elasiticities are generally closer to one; Campa and Goldberg (2001) reject long-run pass-through equal to one in only 9 of the 25 countries.3 Based on an empirical analysis of international prices for two magazines, Ghosh and Wolf (2001) argue that sticky prices or menu costs are a better explanation for imperfect pass-through than strategic pricing or international product differentiation. Consistently with the findings of Campa and Goldberg (2001), they find complete long-run pass-through, which typically holds in theories based on sticky prices, but does not hold in theories of international product differentiation. In this paper, we explore the implications of sticky import prices and imperfect exchange rate pass-through for optimal monetary policy. This is done in three steps. In the first step, we develop a completely micro-founded model for an open economy with sticky domestic and import prices, which takes the international interest rate, prices and output as given. This model differs from the benchmark model in Gali and Monacelli (2000) in two important ways. First, as in Monacelli (1999), we introduce a monopolistically competitive import goods sector with sticky prices. Firms in this sector import a homogenous foreign good at a given world price and produce a differentiated import good for the domestic market. Following Calvo (1983) and capturing the presence of menu costs, import firms are only allowed to change their price when they receive a random price signal. In line with the empirical evidence discussed above, the assumption of sticky import prices implies a gradual adjustment of import prices to the level implied by the law of one price. In addition, following the suggestion by McCallum and Nelson (2001), we allow imported goods to be used both in consumption and production. Second, following Ghironi (2000b), we introduce Blanchard–Yaari-type overlapping generations into the Gali and Monacelli (2000) model.4 This allows us to derive a well-defined stationary steady state for consumption, the terms of trade and net foreign assets, around which the model can be linearised. It also allows for a potentially important role of the current account and net foreign assets in the dynamics of the economy, which we do not further explore in this paper.5 In the second step, we calibrate a linearised version of the model using euro area data. As our analysis focuses on the implications of imperfect pass-through for optimal monetary policy, our calibration exercise concentrates on estimating the degree of price stickiness in the domestic and imported goods sectors. In order to do so, we use a new estimation methodology. Using a VAR on euro area data, we estimate the effects of a monetary policy shock on domestic and import prices and on the three variables that drive those prices: output, net exports and the exchange rate. Conditional on the response of the three driving variables and on other structural parameters of the model, we can then estimate the degree of price stickiness in the domestic and imported goods sector by minimising a measure of the distance between the empirical and the model-based impulse responses of domestic and import price inflation to the monetary policy shock. The results of this exercise suggest two conclusions. First, there is a considerable degree of price stickiness in euro area import prices, consistent with the findings mentioned above. Second, the degree of stickiness in import prices is not significantly different from that in domestic prices. In the third and final step, we then analyse the implications of sticky import prices for optimal monetary policy in the calibrated model. We assume that the central bank's mandate is to minimise the distortions that arise from staggered price setting in the domestic and imported goods sector. Following Woodford (1999), we show that the output cost of these distortions is proportional to the relative price variability in the respective sectors, which in turn is proportional to the variance of price inflation in that sector. The resulting loss function can therefore be written as a weighted average of the variance of domestic and import price inflation, where the relative weight depends on the degree of openness of the economy (or the relative importance of both sectors in consumption and production) and the relative degree of price stickiness. As import price inflation will depend on the gap between the sticky import price and the foreign price denominated in local currency, one important implication of this analysis is the introduction of an explicit reason for the stabilisation of the nominal exchange rate in response to other shocks than those that affect foreign prices. The reason is that such movements in the nominal exchange rate create relative price distortions in the imported goods sector. Another important implication is that the combination of sticky domestic and import prices makes the achievement of the flexible price outcome no longer feasible, even if the central bank only cares about domestic inflation stabilisation. The reason is that imperfect exchange rate pass-through makes the exchange rate channel less effective. As a result more of the adjustment needs to be born by the domestic interest rate channel which primarily affects domestic demand. These findings echo the analysis in Erceg et al. (2000), who come to similar conclusions focusing on the trade-off between the stabilisation of sticky price and wage inflation in a closed economy. We discuss the optimal policy response to a domestic productivity shock, a world demand shock and an exchange rate shock. Overall, the results show that an exclusive focus on the stabilisation of domestic prices is no longer optimal, when import prices are sticky and the exchange rate pass-through is gradual. A number of papers have analysed monetary policy behaviour in the presence of imperfect exchange rate pass-through. For example, Devereux and Engel (2001) examine the implications of local currency pricing in the context of the Obstfeld–Rogoff model and argue that in contrast to the findings of Obstfeld and Rogoff (2002), in this case optimal monetary policy in response to real shocks is fully consistent with fixed exchange rates. Other papers are Monacelli (1999), Batini et al. (2000), Devereux (2000) and Adolfson (2001). Those papers analyse the performance of simple monetary policy rules in the presence of imperfect exchange rate pass-through. However, they do not consider the costs of imperfect pass-through and as such ignore the explicit role for exchange rate stabilisation that it implies. This partly explains why the conclusions are sometimes different. For example, Devereux (2000) finds that a rule that stabilises non-traded goods price inflation performs the best, in particular when pass-through is limited. However, the welfare judgement is based on an ad hoc examination of the volatility of output, consumption and inflation. As we show in this paper, in the presence of sticky prices in both the domestic and the imported goods sector, the response of output and consumption will indeed be less than in the flexible price outcome when a productivity shock hits. However, this response is sub-optimal. Another example is Adolfson (2001), who analyses the impact of incomplete exchange rate pass-through when the central bank minimises a standard loss function in inflation, the output gap and interest rate changes. Adolfson (2001) finds that lower pass-through lead to higher exchange rate volatility. However, this result also depends on the fact that exchange rate stabilisation does not explicitly enter the loss function. Our results are most similar to those obtained by Corsetti and Pesenti (2000). In a model with predetermined domestic and foreign prices based on Corsetti and Pesenti (2001), they show that the optimal policy is to minimise the expected value of a CPI-weighted average of mark-ups charged in the domestic market by domestic and foreign producers. The reasons for doing so are different from those in our model. In Corsetti and Pesenti (2000), risk-averse producers respond to the variability of profits from a specific market by increasing the ex-ante price charged in that market. Policy makers can defend domestic consumers’ welfare by committing to stabilise producers’ profits around their equilibrium flex-price level. Corsetti and Pesenti (2000) also find that a low degree of pass-through severely constrains the ability of monetary policy to move the economy towards the flexible price allocation.6 The remainder of the paper is organised as follows. In Section 2 we develop the theoretical model, derive its steady state and a log-linearised version. In Section 3, we derive and discuss the loss function of the central bank, which is based on a minimisation of the resource cost of inefficient relative price variability in the domestic and imported goods sector. Section 4 presents the calibration of the model. We first estimate a VAR using synthetic euro area data over the period 1977–1999. This VAR is used to derive the empirical impulse-response function of a monetary policy shock and an exchange rate shock on the euro area economy (Section 4.1). In Section 4.2 the structural parameters of the price setting processes are estimated. Section 5 analyses the optimal monetary policy response under discretion to productivity, world demand and exchange rate shock. Finally, we make some concluding remarks in Section 6.

In this paper, we have analysed the implications of imperfect exchange rate pass-through for optimal monetary policy in a completely micro-founded open-economy model in which foreign interest rates, prices and output are assumed to be exogenous. The model used may be of interest by itself, as in contrast to many of the existing open-economy models, it has a well-defined steady state and incorporates a non-trivial role for the current account and net foreign asset accumulation. The empirical evidence on gradual exchange rate pass-through into import prices is captured by assuming Calvo-type staggered price setting in the imported goods sector, similar to that in the domestic goods sector. Using euro area data, we show that import prices appear to exhibit the same degree of price stickiness as domestic prices. As discussed in the introduction, a number of papers have recently examined monetary policy behaviour with incomplete exchange rate pass-through and have noted that imperfect pass-through reduces the effectiveness of the exchange rate channel. However, none of these papers have derived the policy implications of the welfare costs that arise because of staggered import price setting. We show that the minimisation of those costs introduces a motive for exchange rate stabilisation in the central bank's loss function. Similar to the analysis in Benigno (1999), the weight on the stabilisation of imported price inflation depends on the degree of openness and the relative degree of price stickiness in the imported goods sector. This cost of exchange rate variability will provide a counterweight to attempts by the central bank to engineer larger exchange rate movements in order to overcome the ineffectiveness of the exchange rate channel. In the light of the central bank's loss function that we derived, it would be interesting to examine how simple policy rules perform in the presence of imperfect exchange rate pass-through. Another interesting issue is to see what the net effect is of imperfect pass-through on exchange rate volatility. For those questions we need to calibrate the processes driving each of the structural shocks. We leave that for future research.