六個東協國家國際收支的決定因素研究: 以Panel Data分析法 - 政大學術集成

全文

(1)國立政治大學應用經濟與社會發展英語碩士學位學程 International Master’s Program of Applied Economics and Social Development College of Social Sciences National Chengchi University 碩士論文 Master’s Thesis. 六個東協國家國際收支的決定因素研究: 以 Panel Data 分析法 The Determinants of Balance of Payments in Six ASEAN Countries: A Panel Data Analysis. Student: 夏英杰 / Rizal Syaifudin Advisor: 莊奕琦 / Yih-Chyi Chuang. 中華民國 106 年 7 月 July 2017 i .

(2) 六個東協國家國際收支的決定因素研究: 以 Panel Data 分析法 The Determinants of Balance of Payments in Six ASEAN Countries: Using a Panel Data Analysis. 研究生：夏英杰 Student: Rizal Syaifudin 指導教授：莊奕琦 Advisor: Yih-Chyi Chuang. 國立政治大學應用經濟與社會發展英語碩士學位學程碩士論文. A Thesis Submitted to International Master’s Program of Applied Economics and Social Development National Chengchi Univers 中華民國 106 年 7 月 July 2017 ii .

(3) ACKNOWLEDGMENTS. During the writing of this thesis, some people have supported me. I wish to express my gratitude and appreciation to these people. I am indebted to my beloved parents. To my mother, Pratinah, for her endless love, prayer, and support and to my father, Mr. Basuki, for all the text messages and phone calls, which were made every week to keep me going and never give up. I would also like to thank my advisor, Professor Yih-Chi Chuang, for his advice and guidance as well as patience in supervising me from the early stages of this research until the end. He has given me quite an experience in shaping me as a researcher through the past semesters. My appreciation also goes to Professor Fu-Sheng Hung and Professor Ming-Cheng Wang for being part of my defence committee and for giving me advice in improving my thesis better. I am very grateful to have close friends who always supported me. To Broto Wardoyo who always teaches me how to write an academic paper in English and how to improve my English skill. He is also a good listener towards every problem that I faced, especially during my writing process. My appreciation also goes to my Indonesian mates: Jevon Antomi Candra, Ardila Putri, Indira Aryani, Liliek Soelistyo, Safril Mubah, and Sarah Anabarja, for their encouragement. I would also like to thank those individuals who played an important role in the writing of this thesis and to whom I cannot mention one by one, whose names is long enough to be put in this acknowledgment. This thesis is far from perfect, but I am expecting that it will be useful for the readers. Finally, I would not have been able to finish this master programme without the financial support from the Ministry of Education, Republic of China through the Ministry of Education Scholarship Program.. v .

(4) ABSTRACT This research investigates the determinant factors of balance of payments in six ASEAN countries using Keynesian and Monetary approaches during the period of 2002 to 2015. The regression model is conducted with panel data. This thesis first examines the influence of relevant economic variables to balance of payments, such as: exchange rate, gross domestic product (GDP), domestic credit, interest rate, and price level. Then, it tests the robustness of these economic variables with the two control variables: political stability and government’s consumption expenditure. The Hausman test result shows that Fixed Effect Model is a better model than Random Effect Model. According to this model, some economic variables, such as: exchange rate, GDP, and interest rate, are consistent to the Keynesian approach. Meanwhile, variable of price level is consistent to the Monetary approach and variable of domestic credit is inconsistent with, both Keynesian and Monetary approaches. According to the robustness test result, variable of exchange rate is still consistent with Keynesian approach, while variable of GDP and interest rate are partly consistent with Keynesian approach and partly are insignificant to balance of payment. Variable of domestic credit is inconsistent with both Keynesian and Monetary approaches and variable of price level is consistent with the Monetary approach. Variable of political stability shows a negative relationship with balance of payment, whereas variable government’s consumption expenditure is insignificant to balance of payment. This research concludes that Keynesian approach is more appropriate in examining the case of six ASEAN countries. The main reason is because the financial sector in ASEAN countries’ economies are inflexible and susceptible to pressure. In addition, there is a strict capital control in some of those six ASEAN countries. Hence, this research suggests that the government and monetary authorities in those six ASEAN countries should adopt a combination of expansionary fiscal and contractionary monetary policy. The amalgamation of expansionary fiscal and contractionary monetary policy will be efficient in generating adequate capital mobility in order to maintain the stability of balance of payments in the long run. Keywords: balance of payments, Keynesian approach, Monetary approach, panel data.. vi .

(5) TABLE OF CONTENTS Acknowledgments................................................................................................. v Abstract ................................................................................................................. vi Table of contents .................................................................................................. vii 1. Introduction ..................................................................................................... 1 1.1 Problem Statement .................................................................................... 1 1.2 Literature Review...................................................................................... 4 1.3 Organization of thesis ............................................................................... 8 2. Theoretical Framework ................................................................................... 9 2.1 Balance of Payment and Its Components ................................................. 9 2.2 Keynesian Approach and Monetary Approach to Balance of Payment........................................................................................................... 10 2.2.1 Elasticity Approach .......................................................................... 10 2.2.2 Absorption Approach ....................................................................... 11 2.2.3 Monetary Approach ......................................................................... 12 2.3 The Differences between Keynesian Approach and Monetary Approach to Balance of Payment ................................................................... 14 2.4 The Determinant of Balance of Payment .................................................. 15 2.5 Other Control Variables Used in Literature for Robustness Test .............. 17 3. Methodology ................................................................................................... 19 3.1 Data and Description of Variables ............................................................ 19 3.2 Analyses Technique................................................................................... 21 3.3 Estimation Model ...................................................................................... 22 4. The Background of Research Variables .......................................................... 24 4.1 Balance of Payment in Six ASEAN Countries ......................................... 24 4.2 Exchange Rate in Six ASEAN Countries ................................................. 25 4.3 GDP in Six ASEAN Countries.................................................................. 26 4.4 Domestic Credit in Six ASEAN Countries ............................................... 27 4.5 Interest Rate in Six ASEAN Countries ..................................................... 28 4.6 Price in Six ASEAN Countries ................................................................. 29 5. Result of Estimation........................................................................................ 31 5.1 Findings..................................................................................................... 31 6. Conclusion and Policy Implication ................................................................. 36 Bibliography ................................................................................................... 37 Appendix ......................................................................................................... 39 Data ................................................................................................................. 45. vii .

(6) Chapter 1 Introduction. 1.1. Problem Statement Balance of payments (BOP) theory is a part of international economics theories which explains that a nation should have trade relations with other nations in order to collect foreign exchange reserves as the capital for development (Bird, 1981). The number of foreign exchange reserves depends on various factors that influence balance of payments. We can analyse those factors through Keynesian or Monetary approach. Keynesian approach of balance of payment is a one that focuses on the short-term analysis. Keynesian approach assumes that balance of payment in a country automatically would not be in an equilibrium, hence it requires government intervention. Another assumption of Keynesian approach is that wage rate and price level are rigid. Therefore, government should formulate some policies to set the wage rate and price level in order to make new equilibrium. In addition, Duasa (2004) wrote that Keynesian theory analyses balance of payment through the components of trade balance (BOT). Therefore, the disequilibrium in trade balance will lead to disequilibrium in balance of payment. Monetary approach of balance of payment is an explanation of the overall balance of payments by looking at the dynamics of foreign exchange reserves in the long-term adjustment from economic disturbances through money market mechanism. Monetary approach also emphasises that disequilibrium in money market leads to disequilibrium in balance of payment. If the demand of money is greater than the supply of money from the central bank, it will lead to excess demand for money. This condition can be met only by money inflows from abroad. On the other hand, if the money supply from central bank is greater than the demand, the excess supply for money can be eliminated by outflowing the money to other countries. Therefore, we can see the dynamics of money market mechanism leading to a change in the volatility of international reserves and then leads to a change in the volatility of balance of payment, which sometimes can be imbalanced and at other times remains balanced. The equilibrium of balance of payment has a close relationship with economic stability and market behaviour. For example, the global economic crisis in 2008 had some impacts on balance of payment’s behaviour in several ASEAN countries. A 1 .

(7) decrease in export share to several European countries and the US, an increase in Chinese products in ASEAN competitive market, and a high number of import in some ASEAN countries have led to a change in the stability of trade balance. According to the theory, it would lead to a deficit in trade balance. However, trade balance in several ASEAN countries showed a different trend with surpluses during 2008-2009. Singapore had the largest trade balance surplus with US$ 45.03 billion in 2009. The reason is because their income from export was higher than their cost of import. Figure 1.1 shows the current account in six ASEAN countries. According to it, in 2008, with the exception of Malaysia and Brunei Darussalam, trade balance in four ASEAN countries were decreasing. Singapore suffered from the largest decrease in trade balance, reaching US$ 4.0 billion. In 2009, trade balance in those countries were increasing. In Indonesia, the increase was US$ 1.13 billion from the previous year while in Singapore it reached USD 5.08 billion. In Thailand, the increase in trade balance was at US$ 2.25 billion and in the Philippines, it remained at US$ 7.71 billion. In 2010, the trade balance in those countries suffered from a decrease.. TRADE BALANCE Singapore. Thailand. Malaysia. Indonesia. Philippine. Brunei. 200000 150000 100000 50000 0 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015. ‐50000. Figure 1.1. Trade balance in six ASEAN countries (in million unit) Data source: IMF website (2017). Another component which is also determining the equilibrium of balance of payment is capital and financial account. During the period of global economic crisis, there was an increase in capital traffic from developed countries, which were affected due to the crisis, to some developing countries (Dulien, et.al, 2010). A higher investment risk in these crisis countries led to investors’ decision to secure their capital 2 .

(8) by moving it to another country to gain higher returns. On the other hand, these countries also need liquid capital to fund and recover their condition and hence increased their interest rate to attract and provide incentives to investors. Therefore, during this global economic crisis period there were a lot of capital moving back from developing countries to some crisis-ridden countries. However, during the period of 2008-2009, in some ASEAN countries, there were surpluses in capital and financial accounts as well. Figure 1.2 shows capital and financial account in six ASEAN countries. The characteristics of capital and financial account movement in these countries, however, were different. In Indonesia, capital and financial account was decreasing as much as US$ 9.70 billion although the number was still surplus in 2008. In Thailand, the number of capital and financial account was decreasing to US$ 9.9 billion in 2008 although the condition was also still surplus. Surplus of capital and financial account was happening too in Malaysia and the Philippines. In Malaysia, capital and financial account was increasing towards US$ 6.24 billion, whereas in the Philippines it was decreasing to US$ 6.85 billion. A different situation took place in Brunei Darussalam, where the deficit of capital account was taking place as to US$ 1.34 billion on 2008.. CAPITAL AND FINANCIAL ACCOUNT Singapore. Thailand. Malaysia. Indonesia. Philippine. Brunei. 60000 40000 20000 0 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015. ‐20000 ‐40000 ‐60000. Figure 1.2: Capital and financial account in six ASEAN countries (in million unit) Data source: IMF website (2017). 3 .

(9) Other variables, such as: exchange rate, gross domestic product (GDP), domestic credit, interest rate, and price level could also influence balance of payment through mechanism of both trade balance and capital and financial account. According to Keynesian approach, exchange rate, and interest rate have a positive relationship to balance of payment whereas GDP, domestic credit, and price level have a negative relationship to balance of payment. On the other hand, Monetary approach explains that exchange rate and domestic credit have negative relationship to balance of payment, interest rate has positive and negative relationship to balance of payment while GDP and price have positive relationship to balance of payment. Using those two approaches, this research examines the influence of those variables, as listed above, to balance of payment in six ASEAN countries during the period of 2002 to 2015 using panel data analysis. The motivation in doing this research is to provide information on the balance of payment in Thailand, Singapore, Malaysia, Indonesia, the Philippines, and Brunei Darussalam to monetary authorities and respective governments of these countries as one of the recommendations in the making of their economic policies. Moreover, this research provides literature for scholars working on this field in the future since literature on balance of payment in ASEAN countries using panel data estimation is quite limited.. 1.2. Literature Review Previous researchers have examined several variables that influence balance of payment in which they have different results. Duasa (2004) investigated the Malaysian balance of payment uses Keynesian and Monetary approaches from the 1st quarter of 1974 to the 4th quarter of 1995. She used the dynamic model of time series regression to analyse the influence of exchange rate, GDP, price level, money supply, and interest rate. Her finding showed that Keynesian approach is more appropriate in the Malaysian case. Fleermuys (2005) employed Monetary approach to balance of payment in examining the case of Namibia during the period of 1993 to 2003 using Error Correction Model (ECM). He used some independent variables such as GDP, inflation, interest rate, and domestic credit. His research showed that monetary variables are not playing an overwhelming role to determine the balance of payment in Namibia. He also. 4 .

(10) concluded that the balance of payment phenomenon in Namibia is not a monetary phenomenon. Adamu (2007) conducted a study to test the influence of GDP, inflation, interest rate, and domestic credit to balance of payment using Panel Generalised Method of Moments (GMM) estimation in the West African Monetary Zone (WAMZ) using Monetary approach during the period of 1974 to 2008. The research showed that according to both the within-country and cross-country effects suggests that the monetary approach is applicable in the West African Monetary Zone (WAMZ). Ali, et.al (2008) analyzed the influence of political stability on balance of payment and the relative importance of political stability and economic freedom in balance of payment and exchange rate stability. His study focused on ten selected Asian countries of various income levels and tested the factor of political stability and economic freedom in the stability of balance of payment using a simple econometric model with various techniques. His findings showed that stable political stability with visionary leadership leads to improve balance of payment. Umer, et.al (2010) used Error Correction Model (ECM) to investigate the influence of GDP, inflation, interest rate and domestic credit to balance of payment in Pakistan during the period of 1980 to 2008 through Monetary approach. He found that Pakistan’s balance of Payment is not really a Monetary phenomenon. Tijani (2014) conducted a study of empirical analysis of balance of payment adjustment mechanism in Nigeria through monetary approach during the period of 1970 to 2010. He used linear regression estimation to analyse the influence of exchange rate, inflation, balance of trade, domestic credit, and GDP to balance of payment. His findings showed that Monetary approach is more appropriate in the case of Nigeria. Ismalia (2015) also examined the monetary phenomenon to balance of payment in Nigeria during the period of 1986 to 2013. He used Error Correction Model (ECM) estimation to analyse the influence of bank credit to private sector, exchange rate, inflation rate, interest rate, money supply and transparency of the economy to balance of payment. His result also showed that monetary approach is applicable in the case of Nigeria. Brown and Bidemi (2015) investigated fiscal policy measures and balance of payment in Nigeria during the period of 1980 to 2012. He used some independent variables such as government expenditure, government tax revenue, and government debt. The data was analysed using Error Correction Model (ECM). The major findings 5 .

(11) showed that in Nigeria, the success of fiscal policy in order to promote balance of payment depended over the level of public revenue available, the direction of public expenditure and its implementation. This research focuses on six variables which are exchange rate, GDP, domestic credit, interest rate, and price which can be categorised as economic variables, and political stability and government’s consumption expenditure as control variables. Previous researches have suggested that those variables have the ability to contribute in influencing balance of payment. Table 1 shows a summary of the findings.. 6 .

(12) Table 1.1. Summary of previous researchers’ findings. Significant. GDP. Inflation. Interest rate. Domestic Credit. Fleermuys (2005) in Fleermuys (2005) long term in short and long term. +. Political Stability Ali et.al . (2008). Adamu (2007) in whithin-country result and in crosscountry effect Umer et.al. (2010) in short term Tijani (2014) Duasa (2004). Fleermuys (2005) in Adamu (2007) in long term whithin-country result and in crosscountry effect Umer et.al. (2010) Adamu (2007) in in short term whithin-country result and in crosscountry effect. -. Duasa (2004). Umer et.al . (2010) in short term. Adamu (2007) in Fleermuys (2005) in Tijani (2014) whithin-country short and long result and in cross- term country effect Insignificant. Umer et.al. (2010) in short term Tijani (2014) Ismalia (2015) Ismalia (2015). Significant. +. Exchange Rate. Trade Balance. Transparency. Ismalia (2005) in short and long term. Money Supply Ismalia (2005) in short term. Bank Credit to Private Sector Ismalia (2005) in short and long term. Duasa (2004) Tijani (2014). Tijani (2014). Ismalia (2015). Duasa (2004). Insignificant. Significant. Government Expenditure. +. -. Brown and Bidemi (2015). Government Tax Government Debt Revenue Brown and Bidemi (2015). Price Level. Brown and Bidemi Duasa (2004) (2015). Insignificant. 7 .

(13) 1.3. Organisation of Thesis This thesis is structured as follows. Chapter 2 is the theoretical framework of both Keynesian approach and Monetary approach, including the relationship of each independent variable to dependent variable. Chapter 3 describes the methodology, including data analysis and model specification. Chapter 4 explores the background of research variables. Chapter 5 explains the research of estimation and the result of theoretical examination. Finally, Chapter 6 is conclusion and policy implication.. 8 .

(14) Chapter 2 Theoretical Framework. 2.1. Balance of Payment and Its Components Balance of payment is a systematic recording of international economic transactions between the citizen of a country and the citizen of another country within a certain period of time. The aim of balance of payment is to provide some information to government over the position of financial debt in economic relations with other countries as well as to assist governments in the shaping of monetary, fiscal, trade and international payment policies (Nopirin, 1999). Fleermuys (2005, p.2) notes that the most important components of balance of payment are current accounts and capital and financial accounts. Current account is one of the components in the balance of payment that records export and import of goods and services, investment income, repayments and principal of foreign debt, as well as shipment balance and money transfer from and to other countries. Surplus in current account indicates that a nation is a net lender to the rest of the world, in contrast to a current account deficit, which indicates that it is a net borrower. Current account surplus will increase a nation’s net assets by the amount of the surplus. While, capital account is a component on the balance of payment that records the value of capital inflow and outflow of a country such as foreign direct investment especially investments made by multinational companies, portfolio investments and other short-term investments, foreign loans provided by national private banks, grants from other country governments and from multilateral donor institutions such as IMF and world bank. Debit transaction will record any capital transaction that cause an increase in a foreign country’s wealth is called as capital outflow. Then, any transaction leading to an increase in capital inflow will be recorded in credit transaction. A surplus on the capital account means that there are more investment funds flowing into the country than out of the country, or capital inflow greater than capital outflow. Balance of payment also can be explained through the nation’s balance sheet of the Central Bank. According to Table 2.1, the central bank’s balance sheet is divided over two parts; asset and liabilities. High-powered money are central bank’s liabilities. High-powered money consists of currency and bank reserves, while, net foreign asset and domestic credit are central bank’s asset. Net foreign asset is the total of foreign 9 .

(15) exchange reserves, gold, and claims on the government or central bank, whereas domestic credit consists to the central bank’s holding of claims, government debt and loans bank of private sector (Dornbush, 1995, p.614).. Table 2.1. Nation’s balance sheet of central bank Assets. Liabilities. Net foreign asset (NFA). High-powered money (H). Domestic Credit (DC) Source: Dornbush (1995). Then, according to Table 2.1, the equation of a change in net foreign asset can be written as: ∆ Where, ∆ and ∆. ∆. ∆ ........................................................................................................ (2.1). is a change in net foreign asset, ∆. is a change in high-power money,. is a change in domestic credit. In words, equation (2.1) can be written that a. change in net foreign asset is equal to a change in high-power money and a change in domestic credit. According to equation (2.1), Dornbush (1995, p.64) noted that ∆ is equal to the balance of payment.. 2.2. Keynesian Approach and Monetary Approach to Balance of Payment The idea of Keynesian approach is based on the macroeconomics theory of John Maynard Keynes (1883-1946). Keynesian theory does not believe that market mechanism automatically can work towards equilibrium point. Keynes argues that equilibrium point can be achieved by government intervention and he also argues that wage rate and price level are rigid and state is always beset with unemployment issues. Keynesian approach is divided into several approaches, such as: elasticity approach and absorption approach. Both elasticity and absorption approach have some weaknesses, such as: those approaches can only be viewed as balance of payment theory in a world without capital flows.. 2.2.1. Elasticity approach Elasticity approach is developed by Robinson (1937) and emphasises changes in the price of goods and services as the main determinant of nation’s balance of payment (Daniels, 2005, p.274). A currency depreciation or appreciation will lead to 10 .

(16) change the domestic currency price paid for import and the price received for exports, then leads to change in the import demanded quantity and export supplied quantity. The amount of a change in the quantity of import demanded and the quantity of export supplied is determined by elasticity of export supply and elasticity of import demand. Furthermore, depreciation will lead to improve trade balance if the absolute sum export supply elasticity and import demand elasticity is greater than unity. The impact of depreciation on domestic currency to improve trade balance takes a long time. It means that the effect of depreciation on domestic currency is not instantly followed by an increase in trade balance. The effect can be seen by J curve, in which J curve shows the differences of depreciation effect on the trade balance in the short run to the long run. The depreciation effect will show worse to trade balance in the short term, but will increase and show better to trade balance in the long term.. 2.2.2. Absorption Approach Absorption approach is developed by Alexander (1952) that emphasises the role of a nation’s expenditures or absorption and income. This approach assumes that price is constant and, therefore, economists view the absorption approach as a short-run approach to balance of payment (Daniels, 2005, p.292). Daniels (2005, p.282) explained that absorption is a national’s total expenditures on the finals good and services. The nation’s absorption can be written as: ≡ Where,. ........................................................................................................ (2.2). is absorption,. expenditure,. is real consumption expenditure, is real investment. is real government expenditure, and. is the real expenditures of a. nation on imported goods and service. On the other hand, real income of a nation is equivalent to the real expenditures on its output of final goods and services. Therefore, real income is equal to real consumption expenditure, real investment expenditure, real government expenditure, and real export in which can be written as; ≡ Where,. ........................................................................................................... (2.3). is real income and is real export. Furthermore, in absorption approach,. current account balance is only representing the difference between foreign real expenditure on export and domestic real expenditure on import, or can be called as trade balance. The formula of trade balance can be written as: ...................................................................................................................... (2.4) . 11 .

(17) Where,. is trade balance. Then, according to absorption approach trade balance is. determined by the difference between its income and absorption, or it can be written as: ................................................................................................................ (2.5) Or, it can be written as: ......................................................................................................................... (2.6) According to estimation (2.6), the trade balance will surplus if the nation’s income is greater than it absorption, and the trade balance will deficit if the nation’s income is less than it absorption. Then, if the nation’s income is equal to it absorption the trade balance is balanced. Table 2.2 shows the summary of main differences between elasticity approach and absorption approach. Table 2.2. Summary of main differences between elasticity and absorption approaches Balance of Payment No. 1. 2. Elasticity Approach. Absorption Approach. Elasticity approach emphasises changes in the price of goods and service as the main determinant of nation’s balance of payment Elasticity approach assumes that price is not constant. Absorption approach emphasises the role of a nation’s expenditures or absorption and income Absorption approach assumes that price is constant. Source: Daniel (2005). 2.2.3. Monetary Approach Monetary approach of balance of payment defines that balance of payment is the change in international reserve of a country. Monetary approach argues that balance of payment is a monetary phenomenon. It claims that money is the most important role in determining balance of payment. This approach is not refuting that the other important non-monetary factors such as productivity changes, tariff, government spending and taxation on the balance of payment. However, these factors should be linked by money market. Monetary approach also assumes that external disequilibrium is transitory and will revert back to the equilibrium point in the long run. In the monetary approach money market disequilibrium is an important factor causing balance of payment disequilibrium. If demand of money is greater than the supply of money, it will lead to the excess money demand. This condition can be met by inflow of money from overseas. On the other hand, if the money supply is greater than the demand, the excess supply can be eliminated by outflowing the money to the other countries. Therefore, the 12 .

(18) volatility in money market mechanism then will lead to change the number of international reserve and then leads to change the number of balance of payment. The equation of balance of payment according to monetary approach can be derived from money supply, money demand, and money market equilibrium. This model was developed by Johnson (1976) in Tijani (2014) with the following equation as: ·································································· (2.7) , ,. ·································································· (2.8) ··································································· (2.9). Where: . = money supply. = price level. = international reserves. = interest rate. = domestic credit. = equilibrium stock of money. = money demand = level of real domestic income Equation (2.7) shows that money supply is determined by the ability of international reserves and domestic credit level created by monetary reserves of a country. Then, equation (2.8) shows that money demand is influenced by the level of real domestic income, price level, and interest rate. While, equation (2.9) is the equilibrium in the money market. By combining equation (2.7), (2.8) and (2.9), commuting the variables in to percentages, and replacing international reserve variable as dependent variable, we can write the equation of international reserve as: ∆. ∆. , ,. ∆. ····················································· (2.10). Equation (2.10) is the basic monetary approach to balance of payment. Tijani (2014) explains: It postulates that the balance of payments is the outcome of the divergence between the growth of the demand for money and the growth of domestic credit, with the monetary consequences of the balance of payments bringing the money market into equilibrium. An increase in domestic credit brings about an opposite and equivalent change in international reserves, given a stable demand function for money. The coefficient of ∆ is thus known as an offset coefficient. It shows the extent to which changes in domestic credit are. 13 .

(19) offset by changes in international reserves. The monetary approach predicts a value of minus unity for this coefficient in the reserve flow equation (p.71).. 2.3. The Differences between Keynesian Approach and Monetary Approach to Balance of Payment The fundamental philosophical differences between Keynesian and Monetary approaches can be explained by the equation (2.11). According to equation (2.11), money supply or money demand times with the speed of money volatility equal to price time output. The formula can be written as: M.V=P.Y ………………………………………………………………. (2.11) Where: M. = money supply or money demand. V. = the speed of money volatility. P. = price level. Y. = output. According to equation 2.11, Keynesian approach focuses on the real sector, in which this condition is also basically referred as the aggregate demand (P times Y). Meanwhile, monetary approach is focusing on the money supply or money demand (M). In the monetary approach, the price and the volatility of money are fixed and hence M is corresponding with Y, which means that a change in M would be interpreted into a change in Y. Thus, in the equilibrium, the equation will be written as M=Md=Ms and Y=Yd=Ys. Duasa (2004, p.3) wrote the major differences between Keynesian and Monetary approach as shown in Table 2.3 According to Table 2.3, the Keynesian approach analyses the balance of payment phenomenon through the trade balance. Hence, it claims that the trade balance is the most important account on the balance of payment. Some variables such as exchange rate, GDP, domestic credit, interest rate and price will affect trade balance before it extends influence over the balance of payment. Therefore, Keynesian approach argues that disequilibrium in the balance of payment is caused by disequilibrium of trade balance or real forces. On the other hand, Monetary approach analyses the balance of payment phenomenon through both trade balance and capital and financial account. Hence, it claims that international reserve is the most important account in the balance of payment, and therefore, disequilibrium in balance of payment is caused from disequilibrium in international reserve or money forces. 14 .

(20) Table 2.3. The major differences between Keynesian Approach and Monetary Approach to balance of payment.. No. Differences Keynesian Approach. Monetary Approach. 1. Analyze the balance of payment phenomenon only through trade balance. Analyse the balance of payment phenomenon through both trade balance and capital and financial account. 2. Trade balance is the most important account of the balance of payment. International reserve is the most important account of the balance of payment. 3. Disequilibrium in the balance of payment is caused by disequilibrium of real forces. Disequilibrium in the balance of payment is caused by disequilibrium of money forces. Source: Duasa (2004). 2.4. The Determinant of Balance of Payment According to both Keynesian and Monetary approaches, there are some variables that influence balance of payments, such as; exchange rate, GDP, domestic credit, interest rate, and price level. Even though both approaches have similar variables, those variables have different influences on how they can affect balance of payment. Table 2.4 shows the comparison of influences for each coefficient between Keynesian Approach and Monetary Approach to balance of payment.. 15 .

(21) Table 2.4. Relationship between Exchange Rate, GDP, Domestic Credit, Interest Rate, and Price to Balance of payment in Keynesian Approach and Monetary Approach Balance of Payment Independent Variables Exchange Rate. GDP. Domestic Credit. Interest rate. Price Level. Keynesian Approach +. -. -. Explanation Depreciation in domestic curency, output , M , BOT , BOP Domestic Y , M , BOT , BOP Domestic Credit , Ms , i ,I ,Y , M , BOT , BOP. +. Domestic i , I , Agregat Demand , Y ,M , BOT , BOP. -. Domestic P , Demand of export ,BOT , BOP. Monetary Approach -. +. -. Explanation Depreciation in domestic curency, X and demand of domestic curency , R , BOP Domestic Y , demand of domestic money ,R , BOP Domestic Credit , Ms , excess supply in domestic money, capital outflow ,R , BOP Domestic i , Demand foreign exchange , R , BOP. ?. +. Domestic i , R , BOP. , capital inflow. Domestic P , Demand of domestic money , R , BOP. . The Keynesian approach to balance of payment explains that depreciation in domestic currency will lead to an increase of output. An increase in output leads to an increase in import. This condition will lead to a decrease of trade balance and hence, a balance of payment decreases. Then, an increase in domestic income will lead to a negative way to balance of payment. The reason is because an increase in domestic income leads to an increase in imports and therefore the trade balance and the balance of payment decrease. As for domestic credit, an increase in domestic credit will lead to an increase in money supply. An increase in money supply leads to a decrease in interest rate, which will increase investment, then increases income. An increase in income leads to an increase in import. The impact of an increase in import leads to a decrease in trade balance. A decrease in trade balance leads to a decrease in balance of payment. As for interest rate, an increase in domestic interest rate leads to a decrease in investment. A decrease in investment will lead to a decrease in aggregate demand. A decrease in aggregate demand leads to a decrease in domestic income, and then leads to a decrease in import. A decrease in import leads to an increase in trade balance, and hence will improve balance of payment. Price level has negative relationship to balance. 16 .

(22) of payment. The reason is the lower domestic price level will increase demand of export, and hence will lead to increase trade balance and will improve balance of payment. According to monetary approach to balance of payment, exchange rate has a negative relationship to balance of payment. Depreciation in domestic exchange rate will lead to an increase in export and demand for domestic currency, which will improve international reserves and then improve balance of payment. Then, GDP has a negative relationship to balance of payment. An increase in domestic income will lead to an increase in demand of domestic money. Therefore, an increase in demand of domestic money leads to an increase in international reserves then leads to an increase in balance of payment. An increase in domestic credit will lead to a decrease in balance of payment, because a rise in domestic credit leads to an increase in money supply. If there is an excess in money supply, this condition will lead to capital outflow increase. Then, an increase in capital outflow leads to a decrease in international reserves then leads to decrease in balance of payment. The effect of interest rate in the monetary approach is not clear, because interest rate has two ways in influencing balance of payment. An increase in domestic interest rate leads to a decrease in demand foreign exchange, then leads to reduce international reserve and balance of payment decreases. On the other hand, an increase in domestic interest rate also leads to an increase in capital inflow. Then, an increase in capital inflow will increase international reserves, hence, balance of payment increases. An increase in domestic price level will lead to increase on the money demand of a country, and hence international reserves increases and then balance of payment increases.. 2.5. Other Control Variables Used in Literature for Robustness Test Political stability can be defined as the durability and integrity of a current government regime. The index of political stability is determined on the basis of the number of terrorist attacks and violence which takes place in that state. World Bank measures the index of political stability in the range between 2.5 to -2.5. The higher the number of index means that the political stability of country is good. Good political stability has a potential to increase the balance of payment of a country. In the meanwhile, government’s consumption expenditure is defined as any money spent by the government for funding their operations including health, social service, national defense, unemployment packages, and government bailouts to bank but excludes government military expenditures that are parts of government capital 17 .

(23) formation. The number is shown in the percentage of GDP. Higher government’s consumption expenditure will lead to an increase in income and then leads to increase import. An increase in import will lead to a decrease in trade balance, hence, balance of payment of payment decreases.. 18 .

(24) Chapter 3 Methodology. 3.1. Data and Description of Variables This research uses both cross-section and annual data. These data are collected from World Bank website and International Monetary Fund website, and Philippine Central Bank website such as data of net foreign asset, exchange rate, GDP, domestic credit, interest rate, price level, political stability, and government’s consumption expenditure in six ASEAN countries, i.e. Singapore, Thailand, Malaysia, Indonesia, Philippines, and Brunei Darussalam in the period of 2002 to 2015. Table 3.1 shows the summary of operational variables definition. Table 3.1. Definition of operational variables Variables. Balance of payment. Exchange rate. GDP. Domestic Credit. Interest rate. Price level. Definition Balance of payment is measured by net foreign asset. Net foreign asset is used as balance of payment proxy because it calculates the total international reserve and gold reserve of a country. Balance of payment is used as dependent variable. The data are converted in US$. Exchange rate is measured by US$ per domestic exchange rate of a country in the end of period GDP is measured by real GDP of a country at the base year price 2010. Domestic credit is measured by net domestic credit. Net domestic credit is the sum of net claims on the central government and claims on the other sectors of the domestic economy. Data are converted in US$ Interest rate is measured by lending rate of a country. According to world bank, lending rate is the bank rate that usually meets the short- and medium-term financing needs of the private sector. This rate is normally differentiated according to creditworthiness of borrowers and objectives of financing. Price level is measured by consumer price index of a country at the base year price 2010 (2010=100).. 19 . Unit of Measure. Data Source. US$ billion. World Bank website. US$. IMF website, Philippine Central Bank website. US$ billion. World Bank website. US$ billion. Same as above. Percentage. Same as above. Index. Same as above.

(25) Political stability. Government’s consumption expenditure. Political stability is measured by index of political stability of a country. The index of political stability ranges between 2.5 to -2.5. The number of -2.5 means that political stability is very bad and 2.5 means that political stability is very good. Government’s consumption expenditure is measured by general government final consumption expenditure of a share of GDP in a country.. Index. Same as above. Percentage. Same as above. Net foreign asset is used as a dependent variable and as a balance of payment proxy, because it calculates the total international reserve and gold reserve of a country. Net foreign asset is measured in US$ billion units. Exchange rate is measured by US$ per domestic exchange rate of a country. GDP variable is measured by real GDP at the base year price 2010 in US$ billion units, and domestic credit variable is measured by net domestic credit in US$ billion units. Lending rate is used as the proxy of interest rate and is measured in percentage unit. Consumer price index is used as proxy of price level and is measured in index unit. The index of political stability is used as the proxy of political stability and is measured in index unit, whereas general government final consumption expenditure of share of GDP is used as government’s consumption expenditure proxy and is measured in percentage unit. The descriptive statistics of the mean maximum value, minimum value and standard deviations of estimation model are presented in Table 3.2. According to Table 3.2, variable of balance of payment has a mean of US$ 55.6 billion, maximum value of US$ 161 billion, minimum value of US$ 2.18 billion, and the standard deviation of US$ 42.40 billion. Variable of exchange rate has a mean of US$ 0.19, maximum value of US$ 0.53, minimum value of US$ 0.0001, and the standard deviation of US$ 0.21. GDP variable has a mean of US$ 283 billion, maximum value of US$ 988 billion, minimum value of US$ 12.8 billion, and the standard deviation of US$ 228 billion. Then, variable of domestic credit has a mean of US$ 127 billion, maximum value of US$ 351 billion, minimum value of US$ 690 million, and the standard deviation of US$ 96.10 billion. Variable of interest rate has a mean of 7.49 %, maximum value of 18.95 % minimum value of 4.59 %, and the standard deviation of 3.29 %. Then, variable of price has a mean of 95.81 index, maximum value of 132.30 index, minimum value of 54.91 index, and the standard deviation of 14.30 index. Political stability variable has a mean of 0.14 index, maximum value of 1.40 index, minimum value of -2.10 index, and the. 20 .

(26) standard deviation of 1.11 index. Finally, Government’s consumption expenditure variable has a mean of 12.99 %, maximum value of 27.17 unit, minimum value of 7.26 %, and the standard deviation of 4.52. Table 3.2. The mean, maximum value, minimum value and standard deviation. Mean Maximum Minimum Std. Dev.. Balance of Payment 55.60 161.00 2.18 42.40. Exchange Rate 0.19 0.53 0.00010 0.21. GDP. Domestic Credit. Interest Rate. 283 988 13 228. 127 351 0.69 96.10. 7.49 18.95 4.59 3.29. Price Level. Political Stability. 95.81 132.30 54.91 14.30. -0.14 1.40 -2.10 1.11. Government's Consumption Expenditure 12.99 27.17 7.26 4.52. 3.2. Analysis Technique This research uses panel data regression model to estimate the model since it enables to control the individual heterogeneity so that the estimation result is unbiased and, at the same time, it increases the degree of freedom due to an increase in the number of observations. Panel data also can minimise the problem of omitted variable. In panel data analysis, there are two analysis models: Fixed Effect Model and Random Effect Model. Fixed Effect Model explains the relationship between predictor and outcome variable within an entity such as: religion, area, company, country, and person. Each entity has specific characteristics in where at times it can be influencing the independent variables and at other time it cannot. The advantage of Fixed Effect Model is that Fixed Effect Model controls all time variant differences between the individuals, so the estimated coefficients of the Fixed Effect Models would not be biased due to the omitted time variant characteristics, such as: race, gender, religion, culture, etc. Fixed Effect Model assumes that there is a correlation between an entity’s error term and independent variables, and the time variant characteristics are unique to the individual and should not be correlated with other individual characteristics. The equation for Fixed Effect Model can be written as: Yit= β1Xit + αi + μit ...................................................................................... (3.1) Where: α i (i=1) = the unknown intercept for each entity (n entity-specific intercepts). Y it. = the dependent variable, where i = entity and t = time.. X it. = represents one independent variable. β1. = the coefficient of independent variable 21 . .

(27) μtit. = error term. Random Effect Model is used to test whether the differences across entities have some influences on the dependent variable. The advantage of Random Effect Model is that the time variant variables can be included to the model, in which in the Fixed Effect Model these variables are absorbed by intercept. According to Gujarati (2003), this model assumes that individual components are not correlated with each other and are not auto-correlated across both cross-section and time series. Random Effect Model also allows generalisation of the inferences beyond the sample used in the model. The equation Random Effect Model can be written as: Yit= β1Xit + αi + μit+ εi ................................................................................ (3.2) Where: α i (i=1) = the unknown intercept for each entity (n entity-specific intercepts). Y it. = the dependent variable, where i = entity and t = time.. X it. = represents one independent variable. β1. = the coefficient of independent variable. μtit. = the combined time series and entity component. εi. = the entity or individual specific error component. Hausman test is conducted to find the best result between Fixed Effect Model and Random Effect Model. Hausman test can be performed by comparing the P-Value (Prob) and critical value. The Hausman test hypothesis can be written as: H0. : Random Effect Model. H1. : Fixed Effect Model. From those hypotheses, H0 is not rejected if the chi-square statistic < chi-square table or P-value (Prob) > α (5%).. 3.3. Estimation Model This research uses regression model based on Keynesian and Monetary approach in which balance of payment is selected as the dependent variable and exchange rate, GDP, domestic credit, interest rate and price level are the independent economic variables. This model is adopted from Tijani (2014). In his work, Tijani develops a simple OLS regression model at level value and takes into account variables of exchange rate, inflation rate, trade balance, domestic credit, and GDP. This study also uses level value but only takes into account, variables of exchange rate, GDP, interest rate, domestic credit, price level, political stability and government’s 22 .

(28) consumption expenditure, using panel data analysis technique. Furthermore, this research is conducted as follows: First, we examine the influence of relevant economic variables balance of payment. Second, we test the robustness of the economic variables with the two control variables, political stability and the size of government by government’s consumption expenditure. Then, the empirical model can be written as; NFA it= β1 Exch it + β2 GDP it + β3 DC it + β4 i it + β5 CPI it + β6 Polstab it +β7 GSit + μt ·························································································································· (3.3) Where: NFA. = net foreign asset. Exch. = exchange rate. GDP. = gross domestic product. DC. = domestic credit. i. = interest rate. CPI. = consumer price index. Polstab. = political stability. GS. = government’s consumption expenditure. μt. = error term. 23 .

(29) Chapter 4 The Background of Research Variables. 4.1. Balance of Payment in Six ASEAN Countries Balance of payment performance in six ASEAN countries shows an increasing trend in almost all of those countries during 2002 to 2012. According to Figure 4.1, Singapore has the biggest balance of payment position in those periods, followed by Thailand in the second position, Malaysia in the third position, Indonesia in the fourth position, the Philippines in the fifth position and Brunei Darussalam in the sixth position. Overall, the trend of balance of payment shows the resiliency of the domestic economy to external fluctuation, especially the impact of the global crisis economy in 2008. Improvement in the performance of trade balance and capital account is the main factor. High export growth and diversified export destination markets to some Asian countries as well as the characteristics of the export products and the value of the exchange rate in ASEAN region which remains competitive contribute to balance of payment improvement. In addition, the global economic fluctuation directly affects the dynamics of capital and financial transactions, particularly foreign portfolio investments that records surpluses in some countries such as Malaysia and Singapore (see Figure 1.2). However, starting from 2013 to 2014, balance of payments in some countries was decreasing such as in Thailand, Indonesia, the Philippines and Brunei Darussalam. In Indonesia, a decrease in balance of payment is due to an increase in import that leads to trade balance deficit and an increase in capital outflow due to improvement of economic condition in European and American countries after the crisis. In Thailand, a decrease in balance of payments is because the flooding in 2012 disturbed the domestic economics and hence led to a decrease in export. In Malaysia, a decrease in balance of payment is due to an increase in capital outflow as the reaction to uncertainty to the rate of US Federal Reserve bond purchases program. In Brunei Darussalam, a decrease in balance of payments is due to a decrease in the number of oil exports. In 2015, balance of payment in those six countries have increased again because of improvement in domestic economic condition.. 24 .

(30) Figure 4.1. Balance of payment in six ASEAN countries Data source: World Bank website (2017). 4.2. Exchange Rate in Six ASEAN Countries Generally, an increase or decrease in exchange rate in ASEAN countries is influenced by the international trade of each country. In addition, the exchange rate of Brunei Darussalam and Singapore has equality of value. Both Brunei Darussalam and Singapore have signed a Currency Interchangeability Agreement since 1967 in which Brunei dollar is accepted in Singapore as "customary tender" and, likewise, Singapore dollar is accepted for payments in Brunei Darussalam.1 As a result, the exchange rate of Singapore and Brunei increased as much as US$ 0.00691 from 2008 to 2009, while in Malaysia, it increased as much as US$ 0.00750 from 2008 to 2009. In Thailand, the exchange rate increased as much as US$ 0.000691 from 2008 to 2009, and in Indonesia it increased as much as US$ 0.000009 from 2008 to 2009. Then, in the Philippines, the exchange rate increased as much as US$ 0.000513 from 2008 to 2009..                                                              1 Monetary Authority of Singapore. "The Currency Interchangeability Agreement". Retrieved July 8, 2017.. 25 .

(31) EXCHANGE RATE $0.600000 $0.500000 $0.400000 $0.300000 $0.200000 $0.100000 $0.000000 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 Singapore. Thailand. Malaysia. Indonesia. Philippine. Brunei. Figure 4.3. Exchange rate in six ASEAN countries Data source: IMF website, World Bank website (2017). 4.3. GDP in Six ASEAN Countries According to Figure 4.3, the GDP in six ASEAN countries has shown a positive trend during the period of 2002 to 2015, in which Indonesia’s GDP is on the first position and Thailand’s GDP is on the second position. However, in 2009, there were a decrease in GDP of some countries such as Singapore, Thailand, Malaysia, and Brunei Darussalam due to the impact of global economy crisis. In Singapore, the GDP decreased to US$ 1.245 million from 2008 to 2009 and in Thailand it decreased as many as US$ 2.358 million. In Malaysia, the GDP decreased to the point of US$ 6.176 million and in Brunei Darussalam it decreased to US$ 329 thousand. A different situation was undergoing in Indonesia and the Philippines, in which the GDP in these two countries were increasing. In Indonesia, the GDP increased to US$ 31.448 million, and in the Philippines, it increased to US$ 2.105 million. In 2015, the GDP of Brunei Darussalam decreased to the point of US$ 77,000 as the impact of the decrease in oil export. Export is the one of the sectors that produces GDP. In some countries, such as Singapore, Malaysia, Thailand, and Brunei Darussalam, export is the crux of their economy. In Singapore, manufacturing export is the most important factor in determining their GDP, while in Malaysia, the GDP is crutched by manufacturing export and agricultural export. In Thailand, the GDP is crutched by agricultural export, while in Brunei Darussalam oil export is the most important factors to determine GDP. However, the GDP in Indonesia and the Philippines is crutched more by consumption sector. 26 .

(32) Figure 4.3: GDP in six ASEAN countries Data source: World Bank website (2017). 4.4. Domestic Credit in Six ASEAN Countries Figure 4.4 shows that domestic credit shares in almost all countries under studies have positive trend during the period of 2002-2015. The domestic credit of Thailand increased on an average of US$ 21.776 million from 2004 to 2014 and was distributed in service sectors, particularly in professional services industry as working capital credit. In Malaysia, the domestic credit increased on an average of US$ 20.248 million from 2004 to 2014 and was distributed to agricultural sector, particularly in the crude and palm oil industry as working capital credit. In Singapore, the domestic credit increased on an average of US$ 20.290 million from 2004 to 2014 and was distributed to the real estate sector as investment credit. Then, in Indonesia the domestic credit increased on an average of US$ 16.996 million from 2004 to 2014 and was distributed to trade sector, particularly on the small and medium-sized micro industries as working capital credit. In the Philippines, the domestic credit increased on an average of US$ 5.982 million from 2004 to 2014 and was distributed among real estate, manufacturing, information and communication sectors as investment credit. In Brunei Darussalam, the domestic credit increased on average as much as US$ 201.206 thousand.. 27 .

(33) Figure 4.4. Domestic credit in six ASEAN countries Data source: World Bank (2017). 4.5. Interest Rate in Six ASEAN Countries In some ASEAN countries, the lending interest rate is still high. The reason is due to banking inefficiency and a high fund cost, high inflation, high risk of legal uncertainty, legal compliance costs and corporate governance, high concentration in banking industry structure, and high concentration in the deposit market.2 Figure 4.5 shows the interest rate in six ASEAN countries during the period 2002 to 2015. According to Table 4.5, Indonesian interest rate was the highest in comparison with other ASEAN countries during the period 2002 to 2015. From 2002 to 2015, the average of Indonesian interest rate was 14.02 percent, in which on 2002 recorded the highest interest rate as many as 18.95 percent and in 2013 recorded the lowest interest rate as many as 11.66 percent. The interest rate of Philippines is at the second position. The average of the Philippines interest rate from 2002 to 2015 was 7.96 percent, in which in 2005 recorded the highest interest rate as many as 10.18 percent and in 2013 recorded the lowest interest rate as many as 11.8 percent. Thai interest rate is at the third position. The average of Thai interest rate from 2002 to 2015 was 6.55 percent, in which in 2006 recorded the highest interest rate as many as 7.35 percent and in 2004 recorded the lowest interest rate as many as 5.5 percent..                                                              2. Asosiasi Pengusaha Indonesia. “Policy brief apindo: Tingginya suku bunga kredit perbankan Indonesia”. (2016). http://apindo.or.id/id/press/read/policy‐brief‐apindo‐tingginya‐suku‐bunga‐ kredit‐perbankan‐indonesia Retrieved July 10, 2017. . 28 .

(34) Interest Rate 20 18 16 14 12 10 8 6 4 2 0 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 Singapore. Thailand. Malaysia. Indonesia. Philippine. Brunei. Figure 4.5. Interest rate in six ASEAN countries Data source: World Bank website (2017). 4.6. Price Level in Six ASEAN Countries According to Asian Development Bank, an increase in the price level in some ASEAN countries is mainly due to the increase in fuel price and an increase in staple food price.3 According to Figure 4.5, Indonesia had the highest increase in price level during 2013 to 2015. The consumer price index is increasing as many as 15.39 index during the period of 2013 to 2015. The reason was due to the revocation of fuel subsidies by Indonesian government which brought an impact over the increase in staple food price. In the Philippines, the consumer price index was increasing as many as 6.23 index during the period of 2013 to 2015 because of an increase in fuel price. In Singapore, the consumer price index was increasing as to the 0.57 index during the period of 2013 to 2015 because of an increase in price in energy, food, and commodity sectors. Consumer price index of Thailand increased as many as 2.34 index during 2013 to 2014 due to the flood they suffered in 2012 that led to an increase in staple food price. In Malaysia, the consumer price index increased as many as 5.7 index during 2013 to 2015 because of an increase in staple food price..                                                              3 BBC Indonesia. com.” Laju inflasi ancam Asia”. (2008).. http://www.bbc.co.uk/indonesian/news/story/2008/06/printable/080615_asia_inflation.shtml Retrieved July 10, 2017. . 29 .

(35) Price Level 140 120 100 80 60 40 20 0 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 Singapore. Thailand. Malaysia. Indonesia. Philippine. Figure 4.6: Price in six ASEAN countries Data source: World Bank website (2017). 30 . Brunei.

(36) Chapter 5 Result of Estimation. 5.1. Findings In the first step, the Fixed Effect Model and Random Effect Model of the panel data are conducted to analyse the influence of economic variables such as exchange rate, GDP, domestic credit, interest rate, and price level to balance of payment. Table 5.1 shows the result of Fixed Effect Model, Random Effect Model, and Hausman test. Table 5.1. Estimation results of balance of payment: Fixed Effect Model vs. Random Effect Model Balance of payment. Constant Exchange rate GDP Domestic credit Interest rate Price level Observation F statistic (Prob) F statistic Adjusted R-square. Fixed Effect Model. Random Effect Model. (1) -108. (2) 41.10. (0.0006)***. (0.0167)**. 129. 93.30. (0.0153)**. (0.0000)***. -0.10. 0.051. (0.0510)*. (0.0019)***. 0.19. 0.324. (0.0000)***. (0.0000)***. 3.29. -2.97. (0.0750)*. (0.0049)***. 1.24. -0.39. (0.0004)***. (0.0090)***. 84. 84. 99.91. 23.67. (0.0000)***. (0.0000)***. 0.92. 0.58. Hausmant Test Chi-square statistic. 352.866384. Probability. (0.0000)***. Chi-square df Note: * significance at 10% level ** significance at 5% level *** significance at 1% level. 5. According to Table 5.1, column (1) and (2) show the regression estimation of the relevant economic variables on balance of payment. The Hausman test result shows 31 .

(37) that the number of chi-square statistic is 352.86 and the number of chi-square table is 15.09. The chi-square table result is based on the α = 5%, df = k= 5, and the total of observation number = 84. Hence, this research concludes that the number of chi-square statistic > chi-square table. It means that this research should reject H0 or, in other words, the best result is Fixed Effect Model. Hausman test shows that H0 hypothesis of random effect model is rejected. Moreover, the fixed effect model in column (1) has an adjustment R-square value of 0.92 which is greater than 0.58 in the random effect model. Both results show that the use of fixed effect model is more appropriate for the estimation of the six ASEAN countries. From column (1) exchange rate variable has a positive and significant coefficient. It means that a decrease in exchange rate will lead to a decrease in balance of payment on the ceteris paribus assumption. Therefore, the positive effect of exchange rate variable on balance of payment is consistent with Keynesian approach. Variable of GDP has a negative and significant coefficient. It means that an increase in GDP will lead to a decrease in balance of payment on the ceteris paribus assumption. Therefore, variable of GDP has a negative and significant influence on the balance of payment. This is consistent with Keynesian approach. Variable of domestic credit has a positive and significant coefficient. It means that an increase in domestic credit will lead to an increase in balance of payment on the ceteris paribus assumption. This finding, however, is not consistent with both Keynesian approach and Monetary approach. Interest rate variable has a positive and significant coefficient. It means that an increase in interest rate leads to an increase in balance of payment on the ceteris paribus assumption. This finding is consistent with the Keynesian approach. Price variable has a positive and significant coefficient. It means that an increase in price leads to an increase in balance of payment on the ceteris paribus assumption. This finding, however, is consistent with the Monetary approach. Table 5.2 shows the robustness test with other control variables in Fixed Effect Model. Column (3) shows the influence of economics variables and political stability to balance of payment in Fixed Effect Model. Column (4) shows the influence economics variables and government’s consumption expenditure to balance of payment in Fixed Effect Model. Then, column (5) shows the influence economics variables and two control variables to balance of payment in Fixed Effect Model.. 32 .

(38) According to regression result of Table 5.2, each of the exchange rate variable on the column (3), (4) and (5) has a positive coefficient and significant to influence balance of payment variable, and this finding is still consistent with Keynesian approach. This research finding is also similar to the finding of Duasa (2004). Her research suggested that exchange rate has a negative and significant influence on balance of payment. However, this finding is not similar to the finding of Ismalia (2005) who found that exchange rate has a positive and significant influence on balance of payment in both short and long terms, and Tijani (2014) found that exchange rate is insignificant to balance of payment. Each variable of GDP on the column (3), (4) and (5) has a negative coefficient. However, only coefficient of column (4) that deems significant to influence balance of payment and consistent to Keynesian approach, while the coefficient of column (3) and column (5) are not significant. The reason is because in the six ASEAN counties, almost every country has adopted budget deficit. Therefore, they are more focused to use their GDP for paying their state debts and the debts interest. In addition, some countries also use their GDP for paying their civil servant salary and use as the capital for building domestic infrastructures. This finding is similar to the finding of Duasa (2004) who found that GDP has a negative and significant influence on balance of payment. However, this finding is not similar to the findings of Fleermuys (2005), of Adamu (2007) in both within-country result and cross-country effect, of Umer, et.al (2010) in the short term and of Tijani (2014). Their finding suggested that GDP has a positive and significant influence on balance of payment. Each variable of domestic credit on column (3), (4) and (5) has a positive coefficient and significant. This finding is still not consistent with both Keynesian and Monetary approaches, because domestic credit almost in these six ASEAN countries is more distributed to production credit than consumption credit, particularly as investment credit and working capital credit. Hence, an increased domestic credit leads to an increase in capital of production sectors. In the assumption of fixed production cost, an increase in capital leads to an increase in quality of goods and services. Then, an increase in quality of goods and service will lead to an increase in competitive price, then it will lead to an increase in export. An increase in export leads to an increase in trade balance, hence balance of payment increases. This finding is not similar to the findings of Fleermuys (2005) in the long term, of Adamu (2007) in both within-country result and cross-country effect, and of Umer, et.al (2010) in the short term. Their 33 .

(39) findings suggested that domestic credit has a negative and significant influence on balance of payment. Furthermore, this finding is also not similar to the finding of Tijani (2014) who found that domestic credit is insignificant to balance of payment.. Table 5.2. Robustness test with other control variables. Balance of payment (3). (4). -103. -108. (5) -103. (0.0008)***. (0.0025)*** 129. (0.0032)*** 139. (0.0226)** -0.10. (0.0125)** -0.06. (0.0539)* 0.19. (0.3158) 0.16. (0.0000)*** 3.29. (0.0003)*** 2.30. 1.13. (0.0777)* 1.24. (0.2172) 1.13. (0.001)***. (0.0004)***. -10.40. -. (0.001)*** -10.40. Constant Exchange rate. 140 (0.0076)***. GDP. -0.056 (0.3125). Domestic credit. 0.163 (0.0001)***. Interest rate. 2.31 (0.2115). Price Political stability Government’s consumption expenditure. (0.0306)** -. 3.36. (0.0317)** -47.08. -. (0.9974). (0.962). 84. 84. 96.08. 89.58. 84 86.85. (0.0000)***. (0.0000)***. 0.93. 0.92. Observation F statistic (Prob) F statistic Adjusted R-square Note: * significance at 10% level ** significance at 5% level *** significance at 1% level. (0.0000)*** 0.93. Each variable of interest rate on the column (3), (4) and (5) has a positive coefficient. However, only coefficient of column (4) that shows significant to influence balance of payment and consistent to Keynesian approach, while the coefficient of column (3) and column (5) are not significant. The reason is because high interest rate in some countries in those six in ASEAN countries leads to a decrease in real sector. A decrease in real sector leads to a decrease in import as the impact of a decrease in income. A decrease in real sector also will lead to a decrease in export as the impact of an increase in financing in production sector. Therefore, when the two effects of a 34 .