THE MONETARY MODEL OF EXCHANGE RATE DETERMINATION - THE CASES OF SINGAPORE AND THAILAND, 2005-2016

The author test the long-run monetary model with determinants of exchange rate for the cases of Singapore and Thailand in recent years. First, theoretical base of the monetary model is presented in details to provide the approach that we use in this report. Estimating results for exchange rate model of the 2 countries are in the next part. Finally, the discussion and implications of the results can be helpful for the government and policy-makers. First of all, we assume both foreign and domestic countries have stable money demand functions...

FOREIGN TRADE UNIVERSITY FACULTY OF INTERNATIONAL ECONOMICS

SUBJECT: ENGLISH FOR SPECIAL PURPOSES 3

TOPIC: THE MONETARY MODEL OF EXCHANGE RATE DETERMINATION: THE CASES OF SINGAPORE AND THAILAND, 2005-2016

Hanoi, 2017

Table of Contents

1 Abstract 1

2 Introduction 1

3 Data 6

4 Results 6

Variales of model 6

Test results 6

5 Conclusion 8

References 9

1 Abstract

The author test the long-run monetary model with determinants of exchange rate for the cases

of Singapore and Thailand in recent years First, theoretical base of the monetary model is presented in details to provide the approach that we use in this report Estimating results for exchange rate model of the 2 countries are in the next part Finally, the discussion and implications of the results can be helpful for the government and policy-makers

2 Introduction

The past three decades have seen enormous growth in the exchange rate Given the importance attached to the exchange rate in the success or failure of an open economy, it is not surprising that evaluating the influences of exchange rate on economic variables is one of the most heavily and interesting research areas in the discipline Exchange rate movements are perhaps the most important factors affecting sales and profit forecasts, capital budgeting plans and the value of international investments In this respect, changes in exchange rate have a significant impact on the world’s political and economic stability and the welfare of individual countries It comes to the question what affects exchange rates and how to measure their influences The simple monetary model of exchange rate determination posits a strong link between the nominal exchange rate and a simple set of monetary fundamentals

This paper sets out to test whether a simple form of the exchange rate model for Singapore and Thailand based on the relationship among nominal exchange rates, money supply and income holds for a decade, from 2005 to 2016 The intuition underlying the simple monetary model is quite attractive analytically: a country’s price level is determined by the supply and demand for foreign currency in that country, and the price level in different countries should

be the same when expressed in a common currency This model serves as a benchmark evaluation tool of the long-run nominal exchange rates between two or more currencies According to the theory of linking between the money market and the foreign exchange market in the monetary model, money supply and income are two main factors that influence

on nominal exchange rate The prior is determined by central bank, foreign investment and exporting cash flow It is said that an increase of country’s money supply causes it’s currency

to depreciate and a decrease of its money supply can lead its domestic currency to appreciate

On the other hand, the later is regarded as money demand in which greater income implies that more goods and services can be bought, so that more money is needed to conduct transactions (Fisher, 1911), as a result, the exchange rate will be upward (Mundell – Fleming model)

The importance of fully understanding the relationship between monetary model and nominal exchange rate is especially in need, in case AEC (ASEAN Economy Community) is going to made an agreement within this year Joining an economy community brings developing countries in South Asia a number of FDI and FPI, a wide range of labor and potential and diverse markets; however, it also challenges those countries to improve their financial infrastructure to face milestone change in the money and foreign exchange market This report could be a good reference to Vietnam in the plan of making investment in Singapore and Thailand in the future of free trade and investment of AEC

Literature Review

There are a number of economists that have rummaged this domain of foreign exchange market We listed below some remarkable works (all are mentioned in Nataliya Boyko (2002)):

Frankel (1982) in his paper “The Mystery of the Multiplying Marks: A Modification of the Monetary Model” proposes to modify the monetary model He adds real financial wealth, which is a stock, (in addition to income, which is flow), as transactions variable in the money demand function, into both flexible-price and sticky-price versions Moreover, the author puts difference between domestic and foreign interest rates equal to the difference between levels

of expected inflation Expected inflation Frankel approximates with logarithmic change of CPI over preceding 12 months He tests both flexible-price and sticky-price models for the German mark – U.S dollar exchange rate for the period 1974-1980 His results support the hypothesis that real financial wealth should be included in the model while real income should be excluded His tests provide some support for the sticky-price The author claims that while monetary model with real financial wealth fits well, the model without wealth fails

in this case

Smith and Wickens (1986) in “An Empirical Investigation into the Causes of Failure of the Monetary Model of the Exchange Rate” analyze possible reasons why the monetary model fails and test a random walk hypothesis for the exchange rate For the test they employ bilateral sterling - U.S dollar and the German mark – U.S dollar exchange rates for the period 3/1973 – 3/1982 Their results show that the breakdown of PPP assumption and misspecification of money demand function are the main causes of the failure of the monetary model If the sources of misspecification are included into the model, it substantially improves explanatory power of the monetary model

MacDonald and Taylor (1994) in their article “The Monetary Model of the Exchange Rate: Long-run Relationships, Short-run Dynamics and How to Beat a Random Walk” put as the object of the work to show that “at least one of the main exchange rate models – the monetary model – does not behave as badly as is widely thought if it is given better treatment” (MacDonald, Taylor, 1994, p.276) The authors re-examine the flexible-price monetary model for the U.K sterling -U.S dollar exchange rate for the period 1/1976-12/1990 All series were found to be of first order of integration The Johansen cointegration test shows up to three cointegrating vectors These enable the authors to estimate an error correction model They show that the monetary error correction model outperforms random walk forecasting as well

as the basic monetary model MacDonald and Taylor claim that properties of the monetary model can be substantially improved if monetary model is considered as long-run equilibrium condition, which allows short-run dynamics in it They conclude that “the monetary class of exchange rate models, interpreted carefully and with allowance made for complex short-run dynamics, may still be usefully applied, and warrants further research” (MacDonald, Taylor,

1994, p 288)

Diamandis, Georgoutsos, and Kouretas (1996) in their work “Cointegration Tests of the Monetary Exchange Rate Model: the Canadian – U.S Dollar, 1970 – 1994” test the validity

of the sticky price monetary model They consider the Canadian – U.S dollar exchange rate for the period from 1970 to 1994 The model was tested for cointegration and parameter stability

The authors use Augmented Dickey-Fuller (ADF) test to determine the order of integration of variables Since ADF test requires choosing the number of lags to correct for autocorrelation, the authors implement Lagrange Multiplier (LM) test This test helps to choose the number of lags for which no serial correlation was found in the residuals of the regression All variables were found to be of first order of integration The Johansen maximum-likelihood testing cointegration test was used to determine the number of cointegrating vectors One cointegrating vector, that contains all variables of the monetary model, was founded This means that there is a long-run relationship between Canadian-U.S dollar exchange rate, which is described by the monetary model All coefficients, except the U.S output, have predicted signs of the sticky-price monetary model and statistically significant The authors conclude, “the monetary class of models, interpreted carefully, may still be usefully applied” (Diamandis, Georgoutsos, and Kouretas, 1996, p.95)

Rapach and Wohar (2001) in the paper “Testing the Monetary Model of Exchange Rate Determination: New Evidence from a Century of Data” test the long run monetary model They use the basic monetary model and assume that in the steady state domestic and foreign

interest rates are equal Thus, the difference in interest rates is equal to zero and it disappears from the model

The authors use annual data for 14 industrialized countries from the late nineteenth or early twentieth century to the late twentieth century Bilateral exchange rates with U.S dollar were used

Rapach and Wohar implement unit root tests, that estimate cointegration relationship using ordinary least squares (OLS) After that they performed cointegration tests For those countries for which cointegration was found, they, firstly, estimate error-correction models (ECM) and, secondly, compare forecasts of the exchange rate from nạve random walk model and the “monetary fundamentals” Results of estimations show substantial support for the basic long run monetary model for France, Italy, the Netherlands, and Spain; moderate for Belgium, Finland, and Portugal; weak for Switzerland There is no support of the model for Australia, Canada, Denmark, Norway, Sweden, and the United Kingdom There is evidence that “monetary fundamentals” forecast exchange rate for Belgium, Italy, and Switzerland, but there is no evidence for France, Portugal, and Spain

As can be seen from literature review, the class of monetary models of exchange rate determination is still very useful and performs rather well if it is treated properly with some modifications when they are necessary

In this Research Paper, we are using a simple monetary model to show how Singapore money supply, Thailand money supply along with Singapore real GDP and Thailand real GDP influence the SGD/THB exchange rate We analyse the cointergration of these variables with the OLS method

In the section below we present a simple theoretical monetary model and outline our testing steps

The simple monetary model

Our research is mostly based on the simple monetary model which analyses the specific relationship of the nominal exchange rate with the difference between the foreign and domestic money supply and the difference between foreign and domestic real gross domestic product (GDP)

In our analysis, Singapore is denoted as the domestic country and Thailand is denoted as the foreign country

There are some assumptions of the simple monetary model we must follow:

1 The demand for real money balances is a stable function of a small set of real variables

2 Uncovered-interest parity (UIP) holds at all times

3 The supply of money is determined by a stable process

4 Purchasing power parity (PPP) holds over some relevant time horizon

5 Expectations are in some sense rational

First of all, we assume both foreign and domestic countries have stable money demand functions

In which m t is the money supply, pt is the price level, i t is the nominal interest rate, and yt is real output All varibles are stated at time t

The money demand parameters are assumed to be identical in the domestic and foreign countries

The classical model for exchange rate determination is based on the law of one price This law

claims that there can be only one price for a given product at any given time Gold, for example, must cost more or less the same wherever you buy it

Purchasing power parity (PPP) is assumed to hold

where et is the nominal exchange rate measured in the number of units of foreign currency per

unit of domestic currency

Solving the fisrt and the second equations for pt* and pt and substituting the resulting expression into the third equation, yields:

𝑒𝑡 = (𝑚𝑡∗− 𝑚𝑡) − 𝛼𝑡(𝑖𝑡∗− 𝑖𝑡) − 𝛼2(𝑦𝑡∗− 𝑦𝑡) And the following equation is a basic form of the monetary model that establishes a long-run relationship between the nominal exchange rate and a simple set of monetary fundamentals

Mark and Sul (2001) impose the additional restriction that α=1, yielding the simple form of the monetary model:

The long-run monetary model requires these three variables: et, (m*t – mt), (y*t – yt) to be

cointegrated, and so we estimate the following cointegrating relationships:

et = β0 + β1 (m t * - m t) + β2 (y* - y ) + ui

3 Data

The authors have access to the International Financial Statistics database of IMF (International Monetary Fund) in order to collect the necessary data for our research paper in the tables below

The data used in this study consist of quarterly observations from 2005Q1 to 2016Q4 for the nominal exchange rate (foreign currency per U.S dollar), the money supply relative to the U.S., and real GDP relative to the U.S for 2 countries: Singapore and Thailand In our report, Singapore is home country and Thailand is foreign country

The authors have handled this data by Excel software, then use the OLS model and test the above theory by the Eviews 8 software

4 Results

Variales of model

Theory: The basic monetary model represents the long-run relationship between the nominal exchange rate and the money supply If money supply in home country increases, the nominal exchange rate will be rised (Other factors are constant)

In our report, Singapore is home country and Thailand is foreign country, we have the model:

e = β 0 + β 1 (m 2 * - m 2 ) + β 2 (y * - y ) + u i

Variables:

 e: Nominal exchange rate

 m2 * : money supply of Thailand

 m2 : money supply of Singapore

 y * : real GDP of Thailand

 y : real GDP of Singapore

 m2 * - m 2 = ∆m2: the difference between money supply of Thailand and money supply

of Singapore

 y * - y = ∆GDP : the difference between real GDP of Thailand and real GDP of Singapore

 ui : Effects of other variables to exchange rate

Test results

Cointegration test results:

We estimate the coefficients of the monetary model using Ordinary Least Square (OLS), and results are presented as follows:

Table 1 Table of regression coefficient

Independent Variable

-8 ]*

(1.83e-9)

-7]*

(4.36e-8)

Note: * Level of significance is 5%

Source: Author’s calculation using Eviews 8

The model of exchange rate is:

e = 0.0434 + 1.24e -8 (m 2 * - m 2 ) + 1.78e -7 (y * - y ) (1)

Where:

 β0 = 0.0434 > 0, implying that if m2 * is equal to m2 and is y * equal to y then the exchange rate will be 0.0434

 β1 = 1.24e-8 (=4.1597*10 -4 ) > 0, implying that if ∆m2 increases 1 unit, then average of

e will increases 4.1597*10-4 units

In the specific case, on 05/14/2017, 1 Baht Thailand (THO) was exchanged for 0.04 Dollar Singapore (SGD) If the Singapore Government increases the money supply by 10 millions USD dollar (that equal 14.091 millions SGD), the exchange rate between Dollar Singapore and Baht Thailand will be rise and reached 0.044159 SGD/THO

 β2 = 1.78e-7 (=1.623*10 -3 ) > 0, implying that if ∆GDP increases 1 unit, then average e

will increases 1.623*10 -3 units

Omitted variable test result

Ramsey RESET Test shows that the model (1) is non - omitted variable

Collinearity test:

Coefficient Corvariane Matrix

The table shows that variables of model (1) is non - collinearity

3.3.3 Normality test

p – value = 0.4834 > 0.05

Nomarlity test shows that the model (1) is normality

3.3.4 Heterokedasticity test

P – value = 0.2097 > 0.05

White Test shows that the variance of error in model (1) is constant

5 Conclusion

We test the monetary model using longitudinal data for over a decade and find support for a simple form of the long-run monetary model of Singapore dollar exhange rate determination for Thailand as in the precedenting studies (Frankel (1982), Smith and Wickens (1986), and MacDonald and Taylor (1994)) However, it would be more insightful and informative to analyse the model with other advanced methods such as DOLS FM-OLS, ARIMA and VEC model

For policy making process, the governments can exploit this connection in which the exchange rate of an economy affects aggrregate demand through its effect on export and import prices

If government of home country keeps low the exchange rate (or keeps the domestic currency undervalued), export will be stimulated because the prices of exported goods or services fall and imported components more expensive In this case, home country will export more and import less That makes the balance of payment surplus and GDP increase

Following our report, the government can hold down exchange rate by manipulating money

supply In the model above, the home country increases its money supply (m2) and ∆m2 reduces That leads to the increase of e According to the principles of macroeconomics, if