چه چیزی سیاست پولی را پس از بحران شرق آسیا پیش می برد؟ قوانین سیاست پولی نرخ بهره یا نرخ ارز

This paper estimates a simple small open macroeconomic model to analyse the effectiveness of monetary policy rules (MPRs) where either the nominal interest rate or the nominal exchange rate is the policy instrument. The aim is to ascertain which of those MPRs are best suited for a selection of inflation targeting economies of Asia. Normally, one would associate inflation targeting with interest rate rules but it is thought that, due to fear of floating, exchange rate rules may well be more effective given the openness of these economies. It is found that interest rate rules seem to better reflect the prevailing policy regime than exchange rate rules. It is also found that stronger relationships pertaining to the interest rate rules are found in the case of Korea and Thailand than for Indonesia and the Philippines. Exchange rates appear to be very influential in determining the value of the nominal interest rate but not in a policy sense.

One of the dominant issues surrounding the choice of monetary and exchange rate policy in Asia is that of the role of the exchange rate and what policy the crisis-affected countries actually implemented after the crisis. According to many observers, the exchange rate option for developing countries boils down to one between flexibility, on the one hand, and credible pegging, on the other. Countries have, however, been advised to steer clear of arrangements that lie anywhere between these polar extremes (i.e. those in the “middle”) as they were viewed as inherently unstable.1 Of the crisis-affected countries, we observe that Korea, Thailand, Indonesia and the Philippines opted for flexibility by instituting inflation targeting regimes. Each of these countries has passed legal and institutional legislations supporting their respective inflation targeting arrangements. These legislations so passed provide for many facets of the new monetary policy regime including the appointment of key personnel and their tenure, the independence and autonomy of the monetary authority, the stated objectives of monetary policy and the responsibilities and accountability with respect to the achievement of those objectives.2 Malaysia chose the opposite corner and implemented a rigid fix of the ringgit to the US dollar which it subsequently forsaked in July 2005 (see Table 1).Under conventional wisdom regarding inflation targeting, this is potentially problematic. Inflation targeting, as a normative statement, is usually associated with floating exchange rates and is usually implemented by an interest rate monetary policy rule (MPR) (see Clarida et al., 2001 and Debelle, 2001). Any involvement of the exchange rate as an objective will dilute the effectiveness of inflation as the primary target. However, there is evidence to suggest that the inflation targeting countries still held some desire to fix their respective currencies after the effects of the crisis had subsided (for instance, see Cavoli & Rajan, 2007b).3 Furthermore, any involvement of the exchange rate as an instrument (say, a monetary conditions index, MCI) would compromise the effectiveness of the interest rate. However, some recent literature has examined the possibility of inflation targeting being implemented through an exchange rate MPR. Parrado (2004) and Cavoli and Rajan (2007a) have found that this is a feasible policy system and empirical results seem to suggest that a MPR specified in this way may actually be implemented in Singapore. This paper expands on and contributes to the literature on this topic by investigating the behaviour of the policy variables that matter for policy making. It does this by estimating a series of monetary policy rules (MPRs) for Korea, Thailand, Indonesia and the Philippines separately. We will estimate interest rate MPRs versus exchange rate MPRs for each country. Our interest is whether, despite an interest rate based inflation targeting based policy for many Asian countries, the strength of the relationship between the exchange rate and other important macroeconomic variables is such that the exchange rate may actually be a more appropriate instrument. 4 The key to this study is not to analyse the MPRs strictly as reaction functions. Rather, it is an opportunity to examine them as a way of providing information about ‘elasticities’ that are present in the MPRs. The basic premise is that if the relationship between the instrument and the target variable is strong, even if it is not currently viewed by the monetary authority as an instrument, it serves to suggest that it should possibly be an instrument—it provides a normative statement about the conduct of monetary policy. Put another way, the nature of the feedback loop between the structural model and the policy rule is sufficiently robust to suggest that the instrument to that policy rule may be effective. The remainder of the paper is as follows: The following section presents a simple macroeconomic model and derives an optimal MPR for the interest rate and the exchange rate. Using this model and the optimal MPRs as our organising framework, Section 3 discusses the estimation technique and some data issues. Section 4 presents the results of the empirical investigation and Section 5 concludes.

Ultimately, the question that is being asked in this paper is whether or not the relationships between key macroeconomic variables and the possible policy variables justify the particular monetary policy stance taken in the countries sampled? The short answer is that the results are mixed.16 Based on the above analysis, Korea and Thailand appear justified in selecting an inflation target while Indonesia and the Philippines do not. This is entirely consistent with the stylised facts regarding the take up of inflation targeting in Asia. Korea and Thailand were the first to implement this system while the Philippines and Indonesia adopted it much later. This may be reflected in the data. Unfortunately, there is no evidence of possible exchange rate rules for Indonesia and the Philippines—suggesting another metric should be used to capture the salient characteristics of monetary policy for these countries. Even though the evidence of the existence of inflation targeting through the MPRs is perhaps not as strong as first thought, one must also consider that this analysis ignores the potential improvement in credibility that one associates with inflation targeting through greater transparency and accountability. Furthermore, this analysis ignores the obvious benefits of inflation targeting in providing a nominal anchor for inflationary expectations. From the Granger Causality and VAR coefficient results, there would appear to be some support for the existence of an inflation targeting regime for Thailand and Korea and that, under this regime, the nominal interest rate reacts to inflation and output (gap). The IRFs supported this suggestion particularly for inflation and particularly for Thailand. Hence, there is some justification in the use of inflation targeting for Thailand and Korea. The interest rate results suggest that the relationships are weaker for Indonesia and the Philippines. This is perhaps not surprising as both of these countries did not institute inflation targeting regimes until later. The interest rate reaction to the exchange rate, while very strong and statistically significant, did not appear to reflect policy. This is quite surprising given the importance of the exchange rate in these open economies. One would have expected a policy driven response. Unfortunately, the potential exchange rate rules did not reflect a policy response. These relationships appeared to reflect an association captured through parity conditions.