A Preliminary Analysis of Monetary and Inflation Targeting Frameworks for Thailand *

Transcription

1 1 A Preliminary Analysis of Monetary and Inflation Targeting Frameworks for Thailand * Rungsun Hataiseree Monetary Policy Group Bank of Thailand Abstract This paper analyzes main characteristics of monetary policy management under inflation targeting and monetary targeting. It also examines the strengths and weaknesses of the two policy strategies in the application to the context of developing countries like Thailand. Although the results from the empirical analysis in this study remain inconclusive to indicate which sort of policy strategies, inflation targeting or monetary targeting, is more suitable for the actual application in Thailand s case, the findings suggest some specific conditions and important factors that the central bank and concerned authorities have to take into consideration when opting for an appropriate monetary policy framework in the future. In choosing a monetary policy strategy appropriate to a changing structure of Thailand s economic and financial environment, the monetary authorities need to examine further the issues related to the ability to forecast inflation reasonably well over the medium term, the nature of transmission mechanism of monetary and exchange rate policies on inflation, the role of financial innovation and the stability of money demand, a quantitative nature on linkage between instruments and targets of monetary policy, a degree of independence of monetary policy, the nature of the trade-off between the attainment of inflation and other macroeconomic objectives, the nature of shocks affecting the Thai economy in the near term. Keywords: Thailand, Monetary Targeting, Inflation Targeting, Sensible Monetary Strategies 1. Background and Problems Many central banks and monetary authorities are at present contemplating the adoption of an appropriate monetary policy framework suitable for a rapidly changing atmosphere at home and abroad in the next decade when different economic and monetary structures tend to become tightly linked and probably more sensitive to sudden domestic and international shocks. In response to such a changing economic and financial environment, central banks in a number of countries have already chosen monetary targeting as the main strategy for the implementation and the con- * Background paper to a lecture given to central banks delegates at a session of Study Visit on Option for Exchange Rate Policy, organized by ESCAP and the Bank of Thailand, on 22 nd, November, An earlier version of this paper was also presented at the Tenth Convention of the Economic Society of Thailand, November 4, 1999, Bangkok. The author wishes to thank seminar participants in particular Chalongphob Sussangkarn, Sataporn Jinachitra for helpful discussion. Special thanks are also due to Atchana Waiquamdee and Suchada Kirakul for their useful comments and suggestions on the previous version of the paper. The views expressed are those of the author and not necessarily those of the Bank of Thailand.

2 2 duct of monetary policy. This strategy was quite popular in industrial countries during the 1980s, especially in Germany and Switzerland. In the 1990s, however, inflation targeting has become increasingly more important among the central banks in industrialized countries as an alternative framework for the implementation and conduct of monetary policy, in addition to the exchange rate regime and monetary targeting regime which had been used extensively in the 1970s and 1980s, respectively. As one can see, central banks in a number of industrial countries, e.g. Canada, New Zealand, the United Kingdom, Sweden, Finland, Australia, and Spain, had prescribed a policy to maintain a moderate rate of inflation as well as to keep their inflation target in the public eyes. Nevertheless, it is widely agreed that the actual adoption of an appropriate framework for monetary policy is primarily dependent on the foundation and structure of economic and financial systems of an individual country. The successful experiences of a monetary policy management in the history of one country, such as inflation targeting in the case of New Zealand and Canada, or monetary targeting in the case of Germany and Switzerland may not be applied directly to other countries that have different economic and financial structures, per se. Developing countries, as it has been claimed, in particular, seem to have economic and financial structures which are in sharp differences from those of developed countries (Debelle 1997). As Thailand s exchange rate system was changed from the basket-pegging regime to the managed floating regime on July 2, 1997, and as international capital flows have become increasingly volatile over the past many years (see, e.g., IMF, 1999), the Thai monetary authorities find it increasingly important to search for an alternative policy option that can be used as an effective framework for monetary policy management under these new conditions. Under the new economic and financial environment that tends to become increasingly complicated in the future, especially those associated with the external factors, the monetary authorities seem to have little choice but are forced to choose whether to adjust its monetary policy instruments to the changing economic and financial conditions or to use the rule-based framework. Under a rule-based framework, the central bank usually has to announce its ultimate economic targets in advance so that the monetary authorities can easily assess their performance. Among the early attempts to investigate the feasibilities of monetary policies under inflation and monetary targeting were those by Hataiseree (1998a, 1998b, 1998c) and Hataiseree and Rattanalungkarn (1998). However, there appeared to be no empirical analysis to compare the strengths and weaknesses of the two policy strategies in those studies. Even though the monetary targeting framework has yet to be adopted officially by the Bank of Thailand, the concerned officials have closely followed the movements of several monetary indices such as monetary aggregates (both M1 and M2), monetary base, as well as credit aggregates. The information contained in this set of financial variables has been over the years used as supplementary data when the Bank conducts its periodical review and assessment of the ongoing and likely economic and financial conditions of Thailand.

3 3 An actual adoption of inflation targeting or monetary targeting in the case of Thailand is of particularly important issue. A number of previous studies, such as Hataiseree (1993 and 1994), Hataiseree and Phipps (1996b), and Tseng and Corker (1992), pointed out that the demand for money in Thailand has a stable relationship with a set of macroeconomic variables (such as economic growth and inflation). In other words, the quantity of money and macroeconomic variables tend to move in the same direction, and this movement seems to be stable over the period under study. Nevertheless, the aforementioned studies only investigated the money demand function of Thailand during the periods before Thailand switched its exchange rate system from the basket-pegging regime to the managed floating exchange rate regime on July 2 nd, / Apart from this, those studies were carried out in the period prior to the widespread usage of financial innovations by local financial institutions, and before the Thai financial system became more related to the world financial system. The findings and conclusions reported in those studies may be of limited use for the central bank when it has to issue a monetary policy relevant to a changing economic and financial environment in the future. This paper aims to analyze the major difference between monetary policies under inflation targeting and monetary targeting. It also examines the main characteristics, as well as strengths and weaknesses of the two policy frameworks. In addition, the paper considers the possibility that Thailand would base its monetary policy in the future on either inflation targeting or monetary targeting when the necessary pre-conditions for a successful implementation of these two policy frameworks in the case of Thailand have been put in place. Moreover, this paper attempts to compare and contrast the main characteristics of inflation and monetary targeting frameworks. It also discusses how the central bank can employ these two policy strategies to yield an optimal economic outcome. Most importantly, this paper places more emphasis on the analysis of the appropriate framework for monetary policy for Thailand rather than the discussion of the development of monetary policy management. The rest of this paper contains 4 sections. Section 2 summarizes theoretical background used in defining the framework for monetary policy under inflation targeting and monetary targeting. Section 3 provides a brief discussion of the experiences from several central banks that use these two policy strategies as the main framework for monetary policy management. Section 4 studies and compares necessary conditions that affect the selection of an appropriate strategy for monetary policy in Thailand. Section 5 concludes the paper and provides some suggestive comments for the potential application of monetary targeting or inflation targeting frameworks for Thailand in the future. 2. Main Characteristics of Inflation and Monetary Targeting Frameworks Theoretically, the central bank can control the price level through: 1/ More in-depth analysis of the monetary management under the basket-pegging regime can be found in Hataiseree (1995a, 1997a). For detailed description of the salient features and the likely implications of the managed-floating regime on the Thai economy, see, inter alia. Hataiseree (1997c).

4 4 (i) Exchange rate pegging; (ii) Discretional approach or Just-doit approach. Under this approach, the central bank adjusts its monetary policy according to changing economic and financial conditions; (iii) Monetary Targeting; and (iv) Inflation Targeting. In this paper, however, we will only discuss inflation targeting and monetary targeting because these two policy strategies have been practiced in many countries. 2/ It is beyond the scope of this paper to discuss the exchange rate pegging as it is not what Thailand bases it monetary policy on at present, although this sort of exchange rate regime had been once used extensively as the major part of monetary policy framework in the period before July 2, For some representative views of the main issues associated with the exchange rate pegging, and their implications for monetary policy, the reader is referred to Hataiseree (1995b, 1997a). As for details on the managed float regime, the readers can read, for example, Hataiseree (1997c). 2.1Monetary Policy Strategy under Monetary Targeting Monetary targeting policy is based on the foundation that when the central bank can keep money supply (intermediate target) at the targeted level, it can also keep inflation (ultimate target) low. One of the basic assumptions for monetary targeting is that the money supply must have a stable long-run relationship with GDP or inflation rate. In other words, change in the money supply moves in the same direction with change in GDP. Or change in inflation moves in the same direction with change in money supply. These relationships must be stable over time. According to the economic theory, monetary variables chosen as intermediate targets for monetary policy must have at least the following three properties (Figure 1). First, the chosen monetary variables must have a stable long-run relationship with the ultimate target variables. For instance, changes in the money supply M2 must be closely related to changes in domestic economic expansion (as measured by GDP) and the two variables must move in the same direction over the long run. In other words, the relationship between monetary targets and the ultimate targets of monetary policy should not change too drastically when disturbed by exogenous shocks. Second, the monetary authorities must be able to control the chosen intermediate targets. Being a small, open economy with the relatively fixed exchange rate, particularly before July 1997, the Thai monetary authorities find it quite difficult to control the money supply, the combination of domestic assets and foreign assets, particularly in the period of large capital inflows. In view of this, it is often claimed that monitoring the domestic credits variable alone may prove to be a more effective monetary control than that of M1 and M2 in some particular cases. Third, the chosen intermediate targets must be measurable quickly and accurately. Based on this sort of property, interest rate 2/ The set of countries explicitly practicing inflation targeting includes some small- to medium-sized advanced countries as well as some high-middle income developing countries.

5 5 targets may have the advantage qualification over the money supply because they can be tracked daily whereas changes in the money supply can be tracked only monthly. In Thailand s context, according to the empirical evidences reported in Hataiseree (1993, 1994) and Hataiseree and Phipps (1996b), the narrow money M1 and the broad money M2 were found to have the long-run statistical relationships with the ultimate targets of monetary policy. That is, the changes in M1 and M2 were found to cause changes in GDP in the same direction and in a statistically significant way. Moreover, the relationship between M1 and GDP appears to be more stable than the one between M2 and GDP. In practice, the test to determine which intermediate monetary targets have the most stable relationship with the ultimate targets of monetary policy can be carried out in many ways. One of the most popular methods is to estimate the function of money demand for both M1 and M2. The procedure for conducting the aforementioned testing which will be elaborated in more details in Section 4 involves: (i) estimating the demand for money functions for both M1 and M2; (ii) testing the stability of the demand for money functions using Chow Test (1960); (iii) testing the causal relationship between the quantity of money and income using Granger-causality test (Granger, 1969; Granger, 1988) which is a statistical tool to test if a change in one variable will cause another variable to change in a statistically significant way; and (iv) testing the controllability property of the chosen intermediate targets to determine whether the monetary authorities can have a significant degree of control over such selected targets using Granger-causality test. 2.2Monetary Policy Strategy under Inflation Targeting The monetary authorities operating under inflation targeting normally have to announce that inflation is the ultimate target of monetary policy (Kahn, et al. 1998). When conflicts between inflation targets and other macroeconomic targets (e.g., unemployment, the exchange rate, etc.) arise, the authorities concerned would give their first priority to inflation and issue a monetary policy that aims to stabilize the price level. Under inflation targeting framework, the monetary authorities have to make announcement of inflation targets or a range of inflation target for the future. If the inflation projection for the next 1-2 years tends to fall outside the range of the official target, a series of policy actions needs to be carried out in order to bring the inflation level back into the targeted range. The monetary authorities may have to send a signal to change their policy stance through adjusting the level of short-term interest rates and/or intervening in the foreign exchange market in order to induce the movement of inflation to the targeted range as set originally by the authorities. The formulation and implementation of an inflation targeting strategy, as it has been claimed, resorts in part to the use of Monetary Conditions Index (MCI) as a broad indicator of the overall stance of policy (Freed-

6 6 man, 1995). The MCI index allows the monetary authorities to observe how a central bank monetary policy at any point in time is relatively loose or tight, and to what degree. The index also illustrates the impact that a change in monetary policy has on the economy. The Reserve Bank of New Zealand uses MCI to predict the behaviour and the likely outlook of inflation rates in the future so that they can issue appropriate monetary policies to keep inflation within the targeted range. With the contribution of the MCI index, it is often claimed that monetary policy under inflation targeting is a form of forward-looking policy. In the case of New Zealand, if the value of MCI indicates that inflation tends to rise in the future, the Reserves Bank may have to adjust its monetary policy to curb the inflation rate from rising above the desired level. In this respect, the Reserves Bank of New Zealand would send a signal reflecting the change of its monetary policy stance and keep the market informed of the desired level of MCI and price stability. In doing so, the rate of inflation tends to fall to under the targeted level as set forth in the inflation targeting program. In practice, the daily movements of the exchange rates and interest rates are used to calculate into the value of MCI. Then, the derived figure is compared to the targeted value originally set by the Reserves Bank of New Zealand. 3/ The use of MCI as an indicator for the conduct of monetary policy is primarily based on the premise that...monetary policy affects the economic system and in particular, the inflation rate, through two main transmission mechanisms: (i) the interest rate, which influences the level of expenditure and investment, and (ii) the exchange rate, which influences the price of imports, and ultimately the inflation level.... When the interest rate rises or the domestic currency appreciates, the economy tends to slow down. This would in turn help reduce an inflationary pressure. On the contrary, the fall in the interest rates or the depreciation of domestic currency would help enhance the level of domestic consumption and investment and thus leading to an increase in inflationary pressure in the future (Figure 2). As the interest and exchange rates can affect important components of the economy, many central banks find it increasingly difficult in sending the correct signal about their monetary policy stance to the market. In some cases, interest rates and the exchange rates may have important influence on the economy and inflation rate in the same direction. In other cases, however, they may influence the economy and the inflation in the opposite direction. Consider, for example, a case when the Thai baht appreciates and the interest rates 3/ However, it should be noted that, in February of 1999, the MCI-based structure was abandoned in favour of a more conventional arrangement which targets the overnight cash rate. The reason for this shift lies in part with the shift in thinking about the exchange rate. Recent empirical evidences on the nature of the shocks commonly experienced by the New Zealand economy seem to suggest that shifts in the exchange rate are real in character. As a consequence, they do not require an immediate and offsetting interest rate adjustment to maintain longer-term price stability. The direct price consequences of an exchange rate movement shift in response to a real shock are likely to be transitory and, for that reason, are best ignored. This shift in thinking appears to be consistent with the Reserve bank of New Zealand s dropping of the MCI from its central role in the implementation structure.

7 7 fall (Table 1). In a small open economy, like Thailand, both exchange rates and interest rates can affect domestic demand and inflation (e.g., exchange rates through exports and imports, interest rates through domestic investment and consumption). The monetary authorities, therefore, need to take into account the exchange rates and interest rates effects when they evaluate the effectiveness of their monetary policies. As a way to increase an effective assessment, the central bank can use the value of MCI to figure out the net effect of exchange rate and interest rate variables on aggregate demand and inflation. Against the background mentioned above, coupled with continuous changes in the behaviour of interest and the exchange rates, many central banks find it increasingly difficult to evaluate whether the monetary policy stance is tightening or loosening. Where interest rates are high (low) but the domestic currency appreciates (depreciates), inflation tends to decline (rise). In view of this, the central banks need to take into account both interest rates and the exchange rates when they evaluate the liquidity conditions of the monetary system. 2.3 Comparative Aspects of Inflation and Monetary Targeting Frameworks Theoretically, three main features of monetary targeting framework for monetary policy include: A program that aims to maintain the growth of the money supply, with the ultimate goal of controlling inflation and national income. The most important objective under this program is for monetarists to control and manage the prescribed monetary flow on a daily basis. The transmission of monetary policy which assumes that the amount of money supply directly effects the price level, but has very little influence on the economic activity. The philosophy of classical dichotomy asserts that the effects of monetary policy are difficult to determine due to long and variable lags. Moreover, it states that the economy could adjust itself to a new equilibrium following a shock to the economy without resorting to any fine-tuning measures for monetary management. Under this view, inflation is seen as a product of the pressure from price expectation rather than from other exogenous variables. The context of monetary policy which asserts that monetary policy has the major role to play in maintaining the stability of the price level, while other macroeconomic policies are claimed to have very little roles. The view of this kind seems to be in contrast with Keynesian economists who believe that excess demand (as a result of imbalance, excess monetary flow, or other rapid economic changes) would put upward pressure on the production capability of the economy, and hence on inflation rate. According to Svensson (1997), monetary policy framework with focus on inflation targeting has the following key elements :

8 8 The rule of monetary management under this framework would put the rate of inflation as the ultimate objective of monetary policy. However, a clear resolution has not been made about the variables chosen as intermediate targets for a low inflation rate. The transmission of monetary policy seems to place more emphasis on the causal relationship running from domestic production to inflation, while the exchange rate appears to gain a more significant role as the channel of monetary policy transmission. 3. Inflation and Monetary Targeting Frameworks: An International Perspective The framework and strategy for the implementation and the conduct of monetary policy have rapidly changed during the past two decades. In the 1970s, central banks of numerous leading industrialized countries such as Germany, the United States and Japan have utilized monetary targeting as the primary strategy for monetary policy. This has primarily been in response to the escalating inflation during and that resulted from the oil crisis. However, since the early 1980s, the central banks of many industrialized countries have gradually shifted away from monetary targeting. This was because the targeted monetary levels have been made more volatile by the increased pace of financial innovation and the intensified movements of international capital flows. Towards the late 1980s, central banks of many industrialized nations have turned their attention towards an inflation targeting strategy as the focal point of their monetary policy. The list includes the central banks of New Zealand, Canada, England, Sweden, Finland, Australia, Spain and Israel (Green 1996, Svensson 1997, Kahn, et al. 1998). In practice, the central banks of developed countries set ultimate targets for inflation rates that should be beneficial to the economy, economic growth, and the stability of the price level by keeping monetary aggregates from expanding over the set value. These monetary aggregates may be narrowly categorized into M1, M2, and M3. To cite an example, the Federal Reserve Bank of the United States defined its first official monetary target in 1970 under the Federal Open Market Committee: FOMC. Every six weeks the FOMC set figures that M1 and M2 should not exceed. Other central banks in the G-7 countries, including inter alia, Canada, Germany, and the U.K. subsequently followed the U.S. practices by officially announcing the use of monetary targeting as the main framework and strategy for the conduct of monetary policy (Table 2). In the late 1980s, monetary management based on the monetary targeting framework lost its popularity as there have been increasing evidences suggesting that the demand for money function seemed to be unstable due to the increasing pace of financial innovation (Stock and Watson, 1989; Friedman and Kuttner, 1992). The findings of this kind have resulted in the abolition of monetary targeting practices by a number of central banks in the advanced economies, including in particular the Bank of Japan and the Bank of Canada. The findings that monetary targeting had lost its effectiveness makes many cen-

9 9 tral banks in industrial countries (New Zealand, Canada, Finland, Sweden, and Norway) realize that it is indispensable to develop a new monetary strategy to manage their monetary policies, to evaluate their economy, as well as to control price level and inflation. The Reserve Bank of New Zealand began to use inflation targeting in March 1990, followed by the Bank of Canada in February Monetary Policy Management under Monetary Targeting Even though monetary targeting policies have become less important these days, the central bank of Germany (Deutsche Bundesbank) still based its monetary management on monetary targeting. As widely known, the Bundesbank is the only central bank in the world which has resorted to the use of monetary targeting as the main framework for the conduct and implementation of monetary policy. Two main features of monetary targeting in Germany are the intention for transparency and clear communication between the central bank and the public. Examples in this regard include the detailed disclosure of the calculation of inflation targets and monetary targets (M3). Moreover, the central bank of Germany regularly gives information to the public regarding its attempt to move towards the targeted level of money supply growth as initially set by the monetary authorities. The provision of the information of the kind is normally carried out through various means, including in particular annual and monthly publications as well as speeches of the Board members (Mishkin and Posen 1997). As is evident from the Germany s experience, the actual of inflation may occasionally deviate from the predetermined targets. The actual rates of money supply growth may overshoot or undershoot the targets. Strategy for monetary targeting in Germany is, therefore, only a way for monetary officials to inform the public how the policy makers would operate to reach the targeted inflation. Moreover, it has been seen as a tool to increase the responsibility of the central bank. It is important to note that, under the monetary targeting policy, it is a common practice for the Bundesbank to set a targeted range for the inflation in an explicit way. In view of this, it is often claimed that there are no major differences between the monetary targeting strategy and inflation targeting strategy as both strategies seem to place greater emphasis on the use of inflation as the ultimate target of monetary policy. As one can see, the Deutsche Bundesbank sometimes allow M3 to deviate from the targeted value to keep domestic price stability. In other words, Germany would sometimes forego its monetary targets and relax the level of M3 in order to keep inflation low. The management of monetary targeting policy in Germany is by no means greatly different from those operating under the inflation targeting in other countries. It is interesting to note, however, that the management of monetary policy under the framework of monetary targeting primarily targets the inflation level-just as the management of monetary policy under the framework of inflation targeting regime. For monetary targeting, monetary aggregates are used as intermediate targets for the conduct and implementation of monetary policy. In the case of inflation targeting, however, inflation forecasts are used as intermediate targets for the conduct and implementation of monetary policy. Under an inflation targeting regime,

10 10 short-term interest rates tends to be pushed upwards in the event that forecasted inflation shows a tendency of stabilizing at a higher level than the established targeted inflation. Further details on this matter may be found in Svensson (1997), among others. 3.2Monetary Policy Management under Inflation Targeting In response to the empirical finding that the demand for money fails to have a stable relationship with a set of important macroeconomic variables such as inflation and national income, central banks of many industrialized countries have decided to turn away from monetary targeting framework. Several central banks, especially those of industrialized countries, switched to the more popular inflation targeting framework which tends to give more emphasis on the setting of a clear goal for the inflation rate. Table 3 characterizes some salient features of monetary policy under the framework of inflation targeting. From this table, one can observe that the MCI has been used as a meaningful indicator in managing monetary policy by the central banks of many countries, most notably the central banks of New Zealand and Canada that uses MCI as the operating target for the formulation and implementation of monetary policy. The first group of developed countries to use inflation targeting were New Zealand, Canada, and the U.K. From the experiences of these countries, one may conclude that inflation targeting has helped monetary authorities to efficiently control and keep inflation within the targeted range. Following the adoption of inflation targeting in 1990, the inflation rate of New Zealand has constantly declined. Indeed, the country s inflation rate in the period after 1992 has been stable at the targeted level of 0% and 2%. In the case of Canada, inflation targeting was put in place in February Soon afterwards, the central bank of Canada was faced with pressure to mitigate the effect of tax increase from pushing inflation up. Nonetheless, under inflation targeting regime, such an increase in the tax rate would have minimal impact on the price level of commodities for only one short period. It is likely that the tax increase of the kind would have little influence on price increase in a longerterm basis. In light of this, it hardly exerted significant impact on inflation rate. Most importantly, after the initial effect of the tax increase ended, the inflation rate has decreased quickly that it undershoots the targeted rate. Indeed, the actual inflation rate has been constantly under 2% since 1992 and remains at a low level since then. This sort of evidence seems to demonstrate the success of the Central Bank of Canada operating under inflation targeting framework. After a season of drastic changes in the world s financial market in August 1992, the British government had to withdraw itself from the European Exchange Rate Market (ERM). As a result, the country was not able to use the exchange rate as a nominal anchor to managing its monetary policy like it did in the past. The exchange rate has lost its role as a nominal anchor that monetary authorities rely on to maintain inflation. In response to the aforementioned changing atmosphere, the British government has turned to inflation targeting in order to restore the inflation rate towards the targeted level. This sort of monetary strategy has been seen as a useful technique in managing

11 11 monetary policy and thus contributing to the rapid reduction of the inflation rate. Since 1993, inflation rate was held under 2.5%, which is the targeted value. The success of the U.K. inflation targeting regime has convinced other developed nations, including in particular Australia, Finland, Sweden, Spain, and Israel to follow suit. Nevertheless, several economists believe that one should be cautious when analyzing the success of inflation targeting countries. As has been widely claimed, those countries switched to inflation targeting during the period when inflation rates were simultaneously dropping throughout the world. Such the declining trend of global inflation may have caused inflation to fall to the targeted figures. As such, there seems to be no clear evidence in supporting the view that the adoption of inflation targeting regime help lower inflation rate (Debelle, 1997). Under inflation targeting regime, the central banks tend to rely on more economic indicators to manage their monetary policy than those operating under monetary targeting regime. The central banks operating under monetary targeting regime tend to place a greater emphasis on the use of monetary variables to manage their monetary policy. However, inflation targeting seems to provide the central banks more room to manoeuvre their monetary policy to changes in the economic and financial conditions at home and abroad. The central banks operating under inflation targeting are claimed to have several tactics for use to achieve their targeted inflation rate. First, the chosen price level (which is the base in calculating inflation targets) may not incorporate the effect of some sensitive factors that are generally important in the economy. The central banks of many developed countries targeted their inflation rate in conjunction with an indicator of price level called the core inflation. As is widely known, the concept of core inflation does not take into account some sort of sensitive factors such as energy and commodity prices, the effects from changes in the exchange rate, and fluctuations in the Terms of Trade. Second, targeted inflation is normally prescribed as a range rather than a single value. Third, short-term inflation targets tends to be periodically revised to accommodate supply shocks and the effects of the exchange rate. Against the background mentioned above, one can see that the transition from monetary targeting to inflation targeting framework is rather time-consuming and reflects the central banks response to changing economic conditions, development of the ideas of how the economy works, and the limitations of monetary policy. 4. The Application of Monetary Targeting and Inflation Targeting for Thailand Over the past many years, there have been an increasing number of countries, especially those of advanced economies, that have resorted to the rule-like approach for the management of their monetary policy. The move towards this sort of approach is intended to create more transparent policies and make it easy for the public to evaluate the central bank s operation. Many central banks have opted for the use of monetary targeting or inflation targeting to achieve their aims. The application of these two strategies of policy management in developing

12 12 countries, like Thailand, requires a careful examination of the benefits as well as the constraints of the two approaches. Additionally, the necessary preconditions for the successful implementation of these two monetary policy strategies, particularly those related to the institutional infrastructure, need to be carefully evaluated. As is widely agreed, at this stage, Thailand has not based its monetary policy on either inflation targeting or monetary targeting. In view of this, direct evaluation remains impossible. This is clearly different from the situations in developed countries like Canada and New Zealand that have accumulated substantial experiences with regard to the management of monetary policy under monetary targeting or inflation targeting. One possible way to carry out in this study is to analyze certain features of the two strategies within the set of necessary conditions which need to be met in the context of Thailand for a successful monetary policy management. 4.1Preliminary Considerations for Monetary Targeting for Thailand To base their policy management on monetary targeting, the monetary authorities need sufficient statistical evidence indicating the existence of stable long-run relationship between monetary variables (such as the money supply and interest rates) and real economic variables (such as national income and price level). The most important question remains: Can the Thai monetary authorities infer from the statistical evidence which monetary variables are leading economic indicators on which they can base their monetary policies to achieve the ultimate objectives of economic policy? Previous studies (e.g. Hataiseree 1993, 1998a, 1998b) seem to provide some empirical evidences in support of the use of monetary targeting for Thailand. That is, the money demand for M1 was found to be statistically stable over the long run. The empirical evidences obtained in those studies also suggest that changes in the money supply M1 were significantly and closely related to changes in inflation rate. In addition, changes in the monetary base were found to cause changes in M1 in the short-run. More importantly, there were also evidences indicating that the central bank can have reasonable control over the monetary base. In short, the narrow money M1 can be adequately used as a leading indicator for the assessment of the behaviour of the inflation in the future. The empirical evidences obtained from previous studies coupled with the finding in this paper that the narrow money M1 has high degree of controllability, along with the fact that the central bank can better control M1 than M2 suggest that M1 seems to be more stable, predictable, and controllable. In view of these, M1 should be chosen as an useful indicator that the monetary authorities might use as an intermediate target to achieve the ultimate objectives of promoting a low inflation environment with stable economic growth. Nevertheless, one needs to be cautious when making inference and applying statistical results obtained form previous studies in the conduct and implementation of monetary policy. As one can see, those studies were carried out before the exchange rate regime was switched form the basket-pegging regime to the managed float regime on the 2nd of July In this regard, the empiri-

13 13 cal results carried out under these new conditions may be significantly different from the previous results. Although the narrow money M1 may be an useful indicator for monetary policy, it is important that the monetary authorities should resort to the use of some other economic indicators such as the interest rates, the exchange rate, and the changes in world money market in their periodical evaluation and constant monitoring of the stance of monetary policy. In this paper, we tested whether, and to what extent, the results obtained from previous studies may have changed in the face of new economic and financial conditions. The data used in this part cover the period from January 1990 to June 1998 as compared to the years 1981 to 1994 as being the case in previous studies. As is evident from Tables 4 and 5, the demand for M1 were found to have the most long-run stable relationship with the ultimate target of nominal GDP when compared to M2. More importantly, in the short run, changes in M1 were found to cause changes in national income in a more predicable direction. The evidences reported in Table 6 suggest that the money supply M1 seems to have a better predictability property when compared to M2. The evidence mentioned above seems to suggest that the central banks can use M1 to gauge the future directions and movement of national income and the inflation better than M2. The results shown in Table 7 also indicate that M1 has high controllability. This is because changes in the monetary base were found to cause changes in M1 in a more foreseeable direction than that of M2. Additional results from this study enable us to infer that M1 should be used as a leading economic indicator, or an intermediate target, for managing the monetary policy. The information of this kind can be used to enhance the efficiency of the application of the monetary targeting in the context of Thailand in the future. It is important to note, however, that the unstable relationship between monetary variables and the ultimate targets of monetary policy as well as the problem associated with the controllability of the monetary variables have enabled the Germany s monetary authorities to exercise their monetary targeting policy in a flexible manner. That is, the actual figures of intermediate monetary targets were on occasion allowed to deviate from the predetermined figures for some time when the actual values showed signs of overshooting or undershooting the targeted level. The government on several occasions would not intervene in the money market to bring the chosen intermediate monetary targets back to the targeted levels. The unstable relationships between monetary variables and economic variables have often been cited as one of the main reasons behind such an inactive government s intervention. This sort of evidence makes monetary targeting loose its importance as a robust monetary strategy. 4.2Preliminary Considerations for Inflation Targeting for Thailand The Role of Interest Rate and the Exchange Rate on Inflation As mentioned earlier, inflation targeting tends to give more weight to various economic variables (both monetary variables and other real economic variables) in evaluating

14 14 the current behaviour and the future outlook of inflation. On the other hand, monetary targeting tends to give more attention to monetary variables as the main anchors for monetary policies to achieve the targeted inflation. Only few studies on monetary policy management under inflation targeting investigated the conditions in Thailand. One major study on inflation targeting is conducted by Hataiseree and Rattanalungkarn (1998) which established a model to explain the behavior of inflation in Thailand from January 1990 to July This model incorporates some important monetary variables, such as interest rates and the exchange rates, as well as some variables reflecting the developments in the real sector such as oil prices and prices of other relevant commodities. The empirical analysis reported in that study appears to indicate that both interest rate and the exchange have exerted significant impacts on aggregate demand and, thus, on inflation. The degree of relative importance of the interest rate and the exchange rate, the so-called the MCI ratio, was found to be in the value of 3.3:1. The finding of this kind seems to suggest that the interest rate have more influence on inflation than the exchange rate in Thailand s context. The finding of this kind seems to conform with the economic and financial environment during the period under study. As one can see, the previous framework of monetary policy was operating under the basket-pegging regime, particularly in the period before the move to the new exchange rate regime in mid Under such circumstances, it is hardly surprising to observe that variations in the interest rates were far larger than those of the exchange rates. As such, the components of the aggregate demand tended to be influenced by interest rate variable rather than the exchange rate variable. Nonetheless, it is likely that, in the period after the move towards the managed floating rate regime on the 2nd of July 1997 onwards, the exchange rate tends to play a more important role as a mechanism by which monetary policy may affect real economic variables and inflation. Additionally, it has been argued that this new exchange rate regime has given the Thai monetary authorities more room to manoeuvre its exchange rate policy in the sense that an adjustment in the nominal exchange rate under the new regime can be implemented continuously, e.g., through a periodical intervention in the exchange market, in response to developments at home and abroad and without the risk of being politically unacceptable which was the case under the previous exchange rate regime. With such rapidly changing economic and financial environment coupled with the switching towards the new exchange rate regime, the role of the exchange rate has become increasingly important as a channel through which monetary policy may have affected the real sector and, thus, the inflation rate (M.R. Chatu Mongol Sonakul, 1998b and Hataiseree, 1997c) The Role of the MCI for the Conduct of Monetary Policy The empirical results reported in Hataiseree and Rattanalungkarn (1998) suggest that changes in the direction of MCI which can be used as an useful indicator to reflect how tight or loosening monetary policy is have been closely linked to changes in

15 15 the direction of inflation. 4/ This sort of information suggests that the Thai policymakers may use MCI as a monetary indicator for the assessment of the behaviour of inflation in the short run. Notwithstanding, three observationsregarding the role MCI in the application of inflation targeting framework and in the monetary transmission process need to be pointed out below: First, in monitoring the inflation rate, the MCI should not be used as the only indicator for the assessment of the future outlook of the inflation. As indicated earlier, the construction of the MCI are not only included the exchange rates and interest rates variables, but also included some other factors, such as crude oil price and the price of agricultural products, which are deemed to have potential influences on the behaviour of the inflation. It is important to note, however, that there have been an increasing discussion as to whether there is the need to incorporate factors other than the exchange rate and the interest rate, such as the variables reflecting the yield of financial assets in the construction of the MCI (see, e.g., Brash, 1998). However, the inclusion of this kind of variables has never been applied in actual situations or empirical studies. Second, although the expected depreciation of domestic currency may cause inflation rate to rise in the future, it is not necessary to increase the short-term interest rate in order to prevent inflationary pressure. The rationale behind this lines in that there may be some other factors, which are not included in the inflation-forecasting equation, that may affect the inflation in an opposite direction to that of the interest rate. Examples along this line include the lack of confidence on the part of private sector which may lead to a rapid downfall of domestic consumption and investment, the lack of trust in financial institutions, the credit crunch problem. Third, the application of the MCI needs to be administered with caution. Presently, the MCI is just a monetary indicator. Therefore, the management of inflation must also take other indicators into consideration. Most importantly, one must fully understand the behavior of the Thai economy which is not an easy task because the economy may have to take time to adjust itself under the IMF program. One must also be able to differentiate between shocks with exerting a temporary impact and shocks with having a longlasting effect. There is also a 4/ The finding in the study of Hataiseree and Rattanalungkarn (1998) also indicates that exchange rate has assumed a more important role in the transmission of the monetary policy to the real economy and the inflation.

16 16 need to understand clearly whether such a shock stems from the demand side or supply side. Although the empirical results reported in the study of Hataiseree and Rattanalungkarn (1998) suggest that the MCI appears to be an useful indicator for inflation, the actual adoption of an inflation targeting framework as the main strategy for monetary policy management in the case of Thailand may have to take the following aspects into account: If the monetary authorities were to use the MCI as an important indicator for its monetary policy management in the future, they may have to consider different options. First, the central bank could use the MCI as an Operating Target as in case of New Zealand and Canada. In this regard, the central bank needs to set in advance the appropriate value of the MCI deemed to be consistent with the underlying economic and financial conditions. Under this framework, the authorities tend to conduct a periodical intervention in order to guide the actual MCI to the targeted value by adjusting the short-term interest rates (Freedman, 1995). Second, central banks may use the MCI solely as an useful indicator of the policy stance as being the case for those operating in many Scandinavian countries. The Degree of Tolerance for the movements of the MCI needs to be reasonably established. The important message is that in order for the MCI to be effective, the authorities need to be flexible in implementing the policy instruments, thus preventing the use of MCI as a fixed formula. The empirical evaluation of the monetary conditions and its likely outlook, using the MCI in real terms may yield different results from those using the MCI in nominal terms as being the case in the study by Hataiseree and Rattanalungkarn (1998). Empirical results of the two approaches need to be compared to ascertain whether or not there is any statistically significantly different Further Considerations for the Application of Inflation Targeting Theoretically, the formulation and implementation of monetary policy under an inflation targeting framework places a greater emphasis on transparency and accountability. The success of this strategy, however, relies on the credibility of the central bank whether the financial market believes that the central bank can intervene successfully to achieve the targeted inflation. If the financial market looses its confidence in the central bank, it would be quite difficult for the Bank to conduct the monetary policy to guide the underlying inflation towards the predetermined rate of inflation. The reason for this is that future price expectation perceived by various economic agents may have been quite different from the targeted inflation set by the Bank. The monetary authorities also need to choose a suitable period when administer-