An Investigation of the Interrelations among Macroeconomic Variables in Thailand under Inflation-Targeting for the Post-Financial Crisis Period.

• Author: Hossain, Akhand Akhtar

1. Introduction

   In many developed countries, inflation-targeting has become the preferred monetary policy strategy for maintaining price stability under a flexible exchange rate system. Price stability does not necessarily mean zero inflation. Rather, it means a low rate of inflation that fluctuates within a range of 1 to 3 per cent per annum. In fact, the inflation targeting approach to monetary policy has become popular even in developing countries, although many of them do not meet the requirements of implementing inflation-targeting to achieve price stability in the strict sense. To be more precise, for successful implementation of inflation-targeting for price stability, the country in question must have: first, an independent central bank that has the authority to conduct an independent monetary policy within the framework of a flexible exchange rate system; second, developed money and capital markets where interest rates are market-determined and remain closely related; and third, well-identified monetary policy transmission mechanisms that include interest rate, exchange rate and asset price channels. In the absence of these requirements, it is questionable whether inflation-targeting remains appropriate for developing countries that do not have well-developed money and capital markets and that impose administrative controls on interest rates and foreign exchange.

   In the inflation-targeting monetary policy, a major issue is whether short-term policy interest rates remain linked to real output and prices, especially in the low-inflation environment that has prevailed in many countries since the early 2000s. Contemporary monetary literature argues that, under a floating exchange rate system, monetary aggregates or interest rates (but not both) can be used as tools for monetary policy. Accepting the proposition that long-term inflation is a monetary phenomenon, any choice between policy interest rates and monetary aggregates has become a tactical issue. This choice mainly depends on whether there is a stable and identifiable monetary policy transmission mechanism among real output, prices, interest rates, exchange rates and money. As the relationship between monetary growth and inflation is said to have been weakened, although short-term interest rates are more suitable as a monetary policy instrument under inflation-targeting, the vast majority of evidence suggests that there are co-integral or long-run equilibrium relations among money, real output, and prices, interest rates, and exchange rates across countries and over time. This co-integral relations can be used to derive the dynamic relationship among the four variables, although the short-run relations may not be stable enough for monetary policy analysis. This shows that, although the lack of a stable money-demand relation can be used to justify the use of interest rates as a monetary policy instrument, it does not negate a long-run relation, irrespective of whether the money stock is exogenous or endogenous under a fixed or managed-float exchange rate system. The implication is that, when interest rates become ineffective in a low-inflation environment, monetary aggregates can be used as a monetary policy instrument under inflation-targeting. This is no longer just a matter of theoretical curiosity. As a matter of fact, since Japan, the European Union and the United States have all encountered the zero-bound interest rate problem, (1) this has become a major policy issue. Thailand seems to have a similar problem, as the country's inflation rate has been declining since 2019. This raises the question of whether there is a stable and identifiable monetary transmission mechanism in Thailand.

   The main objective of this paper is to empirically study whether interest rates or monetary aggregates can be effectively used as a monetary policy instrument. A survey of Thailand was conducted using quarterly data from 2000Q2 to 2019Q4. Over the last few years, Thailand has become a major open market economy in Southeast Asia. The following is a brief discussion on institutional arrangements to highlight some of the underlying monetary and exchange rate policy issues.

   Thailand has made major economic progress since the devastating Asian Financial Crisis of 1997-98. Following the crisis, the role of monetary policy in price stability has grown considerably in the country. By implementing the IMF's stabilization programme on 2 July 1997, Thailand stopped fixing its currency, the baht, to the US dollar and established a monetary policy regime combining a "managed float" exchange rate regime and a rule-based monetary policy under episodic control of capital flows and exchange rates. (2) Compared to the first episode of limited or no autonomy in monetary policy, this institutional arrangement made monetary policy at least quasi-independent to achieve and maintain price and output stability in a flexible form of inflation-targeting.

   Thailand operated on an exchange rate targeting system until the late 1990s. Extant literature suggests that the fixed exchange rate regime under freely mobile capital led to the financial crisis, as large-scale capital movements created a boom-bust cycle. Once the fixed exchange rate regime became untenable, it was discarded in favour of monetary-targeting as the exchange rate floated. Operating within a weak monetary-targeting approach to monetary policy from July 1997 to May 2000, the Bank of Thailand (BOT) stabilized prices after the crisis. Later, the BOT changed its monetary policy strategy to adopt flexible inflation-targeting to keep inflation stable without affecting output. The major reason for this shift in monetary-targeting was the presumed instability of the money demand function caused by the crisis, which could have loosened the causal relationship between money growth and inflation. (3) After abandoning the monetary aggregate in May 2000, the BOT began deploying a short-term policy interest rate (4) to manage aggregate demand despite the lingering uncertainty about the transmission mechanisms of monetary policy. This uncertainty originated from Thailand's underdeveloped money and capital markets and its de facto managed float exchange rate system, which aimed to maintain exchange rate stability during shocks of both domestic and foreign origin. Under these circumstances, one of the major results of inflation-targeting was excessive volatility in money growth, which in turn kept inflation volatile. The resultant volatility of real interest rates and real exchange rates affected trade and capital flows, hampering economic growth and volatility (Hossain 2015a, 2015b).

   This paper develops a five-variable monetary policy transmission model within an SVEC modelling framework to examine the relative contributions of the policy interest rate and a monetary aggregate to real output, prices, interest rates and exchange rates. The model is estimated using quarterly data for the 2000Q2-2019Q4 period after imposing both short- and long-term restrictions to make the model theoretically consistent and empirically significant. In order to ensure that the model captures the effects of external shocks on the economy, it is estimated using two external variables--the global oil price and the global interest rate (represented by rate of US federal funds).

   The rest of the paper is organized as follows. The next section adapts classical monetary theory to derive the relationships among money, real output, prices, interest rates and exchange rates. The third section then illustrates an SVEC model, which is used to investigate the interactions among the four variables in the presence of the global oil price and the US federal funds rate as external variables. The subsequent section presents the empirical results, including the impulse responses of the system variables and their predictive error variance decompositions. The fifth section reports some results obtained concerning the robustness tests of the SVEC model. The final section summarizes the main findings, draws policy implications and provides concluding remarks.

2. Relationships among Money, Real Output, Prices, Interest Rates and Exchange Rates

   Based on the principles of classical monetary theory, this section specifies a monetary model of inflation that could study any causal link between inflation and money growth, determined exogenously by a central bank in a flexible exchange rate system. Consider the equilibrium condition of the money market:

   M/P = [m.sup.d] (y, NIR) (1)

   where M is the money stock, y is real income, P is the price level, and NIR is a representative nominal interest rate. This money market equilibrium condition provides a relationship among money, real income and price level, assuming that a long-run, stable money demand relationship [m.sup.d5] exists in a country like Thailand.

   Equation (1) can be expressed as a proportional change or as a first difference, indicating that there is a one-to-one relationship between inflation and money growth, adjusted for economic growth and interest rate variations:

   [pi] = [lambda] - [[omega].sub.1][g.sub.y] + [[omega].sub.2][DELTA]NIR (2)

   where [pi] is the inflation rate, [lambda] is the money growth rate, [g.sub.y] is the economic growth rate, [[omega].sub.1] is the income elasticity of demand for money, [[omega].sub.2] is the semi-interest elasticity of demand for money, and [DELTA] is the first-difference operator. Equation (2) suggests that the inflation rate increases positively and proportionately to the rate of money growth and decreases with the rate of economic growth. (6) At a steady-state, the growth rate of output approximates the technological progress, while the interest rate approximates the long-term real interest rate and the rate of inflation that is consistent with the rate of money growth. This long-term relationship remains consistent with the rational expectations hypothesis because the...