عملکرد تعهد نرخ بهره صفر: قواعد سیاست پولی با وجود حد پایین صفر در نرخ های بهره
|کد مقاله||سال انتشار||مقاله انگلیسی||ترجمه فارسی||تعداد کلمات|
|26300||2008||34 صفحه PDF||سفارش دهید||15123 کلمه|
Publisher : Elsevier - Science Direct (الزویر - ساینس دایرکت)
Journal : Journal of the Japanese and International Economies, Volume 22, Issue 1, March 2008, Pages 34–67
We analyze the monetary policy rules which could be implemented in practice under the zero interest rate constraint. Based on the estimated small structural model for Japanese economy, we investigate which policy rule is superior using stochastic simulations. We modify the estimated Taylor-type rule variously by adding a commitment whereby the zero rate policy will be maintained until the inflation rate rises beyond a specific level. We find that such policy rules can be effective if the commitment is set appropriately. We also find that a nonlinear policy rule incorporating preemptive easing can perform well, mostly without any explicit commitment. J. Japanese Int. Economies22 (1) (2008) 34–67.
Various proposals have been made regarding monetary policy in the presence of the zero lower bound on nominal interest rates. Since the effect of monetary policy declines when facing this constraint, there is a widely shared belief among economists that (1) preemptive monetary easing is important to minimize the likelihood that interest rates will fall to zero, and (2) in cases when interest rates have fallen to zero, “expectations management” which acts on the formation of private-sector expectations toward future monetary policy is important. While macroeconomic theory has played an important role in drawing these general conclusions, there is still room for developing further analyses for policy proposals on the more specific questions of (1) what degree of preemptive monetary easing is appropriate, and (2) how expectations management should be implemented. A number of authors have studied the issue of preemptive monetary easing. For example, in three papers that assumed a purely forward-looking structural model, Adam and Billi (2004a) derived the optimal commitment policy, and Nakov (2004) and Adam and Billi (2004b) derived the optimal discretionary policy. Orphanides and Wieland (2000) and Kato and Nishiyama (2005) derived the optimal discretionary policy under a partially forward-looking model and a purely backward-looking model, respectively. To better depict the actual conditions of the economy, our paper conducts analyses assuming a “hybrid” structural model that accommodates both forward- and backward-looking agents in the IS and AS curves, and seeks optimal policy under a simple policy rule framework, taking into account a possibility of adopting a zero interest rate commitment. Prior research on the importance of expectations management includes Jung et al. (2005) and Eggertsson and Woodford (2003), of which the latter advocates a kind of price level targeting as the optimal targeting rule for a purely forward-looking structural model. Since the structure in the actual economy incorporates various uncertainties, however, there seem to be doubts regarding the feasibility and efficacy of implementing such a specific targeting policy in practice. This leaves the question of what type of options exist for practical expectations management when facing the zero lower bound on interest rates. Analyzing the function of Japan's “policy duration commitment” to maintaining the zero interest rate, which is referred to herein as the “zero interest rate commitment,” is adopted in this paper to examine the practical options.1 Keeping these points in mind, we advance our analysis of practical monetary policy rules in the presence of the zero interest rate constraint using stochastic simulations.2 The remainder of this paper is organized as follows. In Section 2, we explain the simulation framework, including the model and the simulation methods. In Section 3, we conduct a simulation to evaluate whether the zero interest rate commitment, such as one implemented in Japan currently, is effective as a permanent monetary policy rule. In Section 4, we demonstrate that as a permanent policy rule which takes the zero interest rate constraint into account, a nonlinear policy rule incorporating an element of preemptive monetary easing is effective. From Section 5, we assume that the economy is initially subjected to some negative shocks and analyze the conditions where a time-restricted commitment to maintain a zero interest rate is effective. Specifically, in Section 5 we investigate how the optimal setting of such a commitment may differ depending on initial economic conditions. We also analyze the policy effects when the commitment is made based on a price level measure instead of the inflation rate. Then in Section 6, we analyze whether or not the introduction of a zero interest rate commitment is effective in cases where the above-mentioned nonlinear policy rule has already gained credibility from the private sector. Finally, in Section 7 we summarize our findings and consider areas for future research.
نتیجه گیری انگلیسی
This paper has examined, based on stochastic simulation analyses, what kinds of monetary policy rules are effective assuming a small-scale structural model giving consideration to the zero lower bound on nominal interest rates. The main conclusions obtained in this paper can be summarized as follows. • A zero interest rate commitment based solely on the inflation rate does not necessarily show good policy performance in promoting economic stability when recognized as a permanent policy rule. Rather, there is a risk that economic stability may be impaired because, when the commitment is in effect, the conditions of the GDP gap are not reflected in monetary policy. • A nonlinear optimal simple rule, whereby a conventional linear Taylor-type rule is optimally “curved” as the nominal interest rate approaches zero, performs well as a permanent monetary policy rule that takes the zero interest rate constraint into account. This policy rule has the effect of diminishing the “cost” of the zero lower bound via preemptive monetary easing. The desirable shape of the curve can be determined depending on structural parameters such as the target inflation rate and the long-term natural rate of interest. The nonlinear optimal simple rule provides hints regarding (i) how high the interest rate indicated by a normal Taylor-type rule should rise before the transition from zero to positive interest rates, and (ii) the speed at which a normal Taylor-type rule should be reinstated once interest rates have turned positive. • A zero interest rate commitment can be effective when it is interpreted as a “time-restricted” monetary policy rule that will be terminated after the prevailing deflation is overcome. The content of the zero interest rate commitment—that is, the threshold rate of inflation adopted as a prerequisite for exiting the zero interest rate policy—can be optimally selected based on the economic conditions, such as the size and persistence of the demand and supply shocks, when the commitment is introduced. • We can assume a zero interest rate commitment based on a price level measure, instead of the inflation rate, but such a commitment is superior in strictly limited cases with an extremely large demand shock. • On the whole, the policy performance of a Taylor-type rule with an optimally set commitment does not match the performance of the nonlinear optimal simple rule that has gained credibility. This suggests that communications to ensure the formation of appropriate private-sector expectations regarding the central bank's stance toward preemptive monetary easing are important. In cases where the zero lower bound on nominal interest rates is encountered before such a policy stance is sufficiently recognized by the private sector, the action of making a zero interest rate commitment may effectively convey the “message” on the monetary policy stance as the qualitative explanations by the central bank do. • After the nonlinear optimal simple rule has gained complete credibility, except for a few cases with extremely large demand shocks, setting some sort of zero interest rate commitment does not enhance the policy effect. The marginal benefits from adding a zero interest rate commitment are less after the credibility is established than when the credibility is insufficient. • These analytical results are considered to be qualitatively robust. However, in quantitative evaluations, the conclusions may change depending on the economic environment including the level of the long-term natural rate of interest and the targeted rate of inflation. It should also be pointed out again that the relative performance of each policy rule depends on the initial demand and supply shocks. Table 9 shows the ranking, for each initial shock, of all the policy rules analyzed in this paper in terms of the total loss under a short-term simulation. For example, in case (1) with the largest demand shock, the nonlinear optimal rule with an inflation rate based commitment (where the optimal threshold rate is 0.0%) performs best, and the baseline Taylor-type rule with a price level based commitment performs second best. The third best is the baseline rule with an inflation rate based commitment where the optimal threshold rate is high at +1.5%+1.5% because of the need to compensate for the lack of recognition of the preemptive monetary easing. In case (2) with the second largest demand shock, the nonlinear optimal rule with an inflation rate based commitment is also the best policy rule as in case (1), but the optimal threshold rate becomes −0.5%−0.5%, which is lower than in case (1). The ranking of the baseline Taylor-type rule with a commitment based on price level falls to fourth place and the nonlinear optimal rule without any commitment obtains second place. In contrast, in cases (3) and (4) where the initial demand shock is relatively small, the nonlinear optimal rule without any commitment performs best and the one with an inflation rate based commitment is second best. In case (5) where the initial demand shock is very small, any zero interest rate commitment is undesirable regardless of whether the policy stance toward preemptive monetary easing has gained credibility or not. The policy implications from these points are that when approaching the zero lower bound on nominal interest rates the central bank should (i) try to identify the feature of the economic shocks accurately and (ii) judge the extent to which the preemptive monetary easing policy is recognized.This paper also examined the following two possible functions of a zero interest rate commitment23: (1) The message function of transmitting information to the private sector regarding the contents of the nonlinear monetary policy rule nearby zero interest rates. (2) As a policy instrument to realize a better economic path that would be impossible to achieve without such a commitment. The results of our analyses indicate that, when the private sector does not recognize the contents of the nonlinear policy rule, a zero interest rate commitment that is appropriately set in accordance with the economic conditions is highly likely to demonstrate function (2). Under this function the contents of the zero interest rate commitment are directly incorporated into the policy rule, and have the effect of stabilizing the economy. In contrast, under (1) the introduction of the zero interest rate commitment has the potential function of working on the formation of private-sector expectations regarding preemptive monetary easing before facing zero interest rates as well as the policy response just after interest rates return from zero to a positive level. While we do not statistically verify that point in this paper, we believe there is a substantial likelihood that this potential function will become manifest as the understanding of the desirable monetary policy under the zero lower bound permeates. The analyses in this paper, especially the quantitative contents, are in part dependent on the accuracy of the presumed model and the appropriateness of the assumptions made. In this context, we may list two particular points at issue. (1) Our analyses have adopted the policy judgment criteria regarding the relative importance of inflation stability, GDP stability and interest rate stability by assuming that Japan's prior monetary policy has been implemented in an optimal manner. In other words, under this approach we have evaluated policy performance assuming that preference in the conduct of prior monetary policy should be unchanged. The counterargument to this notes that there is possibility that Japanese monetary policy has not been conducted optimally in the past, and that a different method could be used to set the policy judgment criteria. One example of such an alternative approach, as explained in footnote 10, is to calibrate the deep parameters of the economic structure and then derive the loss function from the economic theory. Compared with the settings used in this paper, the conclusions under that method are to give greater emphasis to inflation stability and to place very little emphasis on GDP stability and interest rate stability. In that case, one would expect substantial changes to the analytical findings. One may surmise that the trend would then be toward a high evaluation of the zero interest rate commitment under each case, since qualitatively this would make it possible to ignore the loss of GDP stability that is a side effect of the commitment. In short, when the economic shocks and other settings are given, the conditions for setting the optimal commitment would likely change toward being more accommodative. On the other hand, it is not clear what changes this would bring to the nonlinear optimal simple rule. In general, it is important to discuss further on the best criteria for policy judgment. (2) It would also be desirable to investigate how the optimal monetary policy changes when the leeway for supporting economic conditions with fiscal policy changes as the concern regarding future tax changes depending on the seriousness of fiscal deficits. In such cases, qualitatively, a more accommodative monetary policy might become optimal with more serious fiscal deficits. These are remaining issues for the future.