(Luận văn) determinants of foreign exchange rate, case of vietnamese dong and japanese yen

t to ng hi ep UNIVERSITY OF ECONOMICS HO CHI MINH CITY VIETNAM INSTITUTE OF SOCIAL STUDIES THE HAGUE THE NETHERLANDS w n lo VIETNAM - NETHERLANDS PROGRAMME FOR M.A IN DEVELOPMENT ECONOMICS ad ju y th yi pl n ua al n va ll fu DETERMINANTS OF FOREIGN EXCHANGE RATE: CASE OF VIETNAMESE DONG AND JAPANESE YEN oi m at nh z z om l.c gm Mr TRAN VUONG TU k jm ht vb BY n va y te re HO CHI MINH CITY, MAY 2013 n a Lu MASTER OF ARTS IN DEVELOPMENT ECONOMICS th t to ng hi ep UNIVERSITY OF ECONOMICS HO CHI MINH CITY VIETNAM INSTITUTE OF SOCIAL STUDIES THE HAGUE THE NETHERLANDS w n lo ad VIETNAM - NETHERLANDS PROGRAMME FOR M.A IN DEVELOPMENT ECONOMICS ju y th yi pl ua al n DETERMINANTS OF FOREIGN EXCHANGE RATE: CASE OF VIETNAMESE DONG AND JAPANESE YEN n va ll fu oi m at nh z z A thesis submitted in partial fulfilment of the requirements for the degree of vb k jm ht MASTER OF ARTS IN DEVELOPMENT ECONOMICS om n a Lu Mr TRAN VUONG TU l.c gm By n va th HO CHI MINH CITY, MAY 2013 y PhD NGUYEN HOANG BAO te re Academic Supervisor: t to ng hi ACKNOWLEDGEMENT ep This thesis was written at the University of Economics Ho Chi Minh City In addition, w n it was completed in October 2013 During the process of writing, the paper has gained lo ad a lot of experience in writing a thesis and in the area of foreign exchange rate analysis ju y th During the three months of writing this thesis, several persons have contributed in the yi different ways to the quality of this thesis and the paper would like to take this pl opportunity to thank them al n ua Firstly, the paper would like to thank our supervisor PhD Nguyen Hoang Bao for all va the help, guidance, and support The paper would also like to express gratitude to all n professors of the Vietnam-Netherlands Program for the Master in Development fu ll Economics and the classmates who offer to us some useful suggestions Finally, we m oi express special thank to our families and partners for their love and support at nh z z k jm ht vb om l.c gm n a Lu n va y te re th t to ng ABSTRACT hi ep Exchange rate not only plays a very important role in the economic policy of the government of Vietnam in the process of integration into the world economy, but also w n effects many exporters, importers, foreign investors, and commercial banks in the lo ad international transaction y th Japanese economy plays as important as having mainly economic relations with ju yi Vietnamese economy in the export-import trade, foreign direct investment (FDI) capital, pl official development assistance (ODA), etc However, Vietnam government applies the al n ua floating exchange rate policy between Vietnamese Dong and the Japanese Yen va Therefore, the fluctuations of Vietnamese-Japanese exchange rate might great impact n on the trade and investment The exporters and importers of two countries, Japanese fu ll investors, the commercial bankers that having international settlement with Japanese m oi Yen, are in need of defending the exchange rate risk volatility of the exchange rate pairs nh at Our study enhance on analyzing and predicting the fluctuations of Vietnamese-Japanese z z exchange rate The main research question identifies (1) Which Vietnamese and vb ht Japanese macroeconomics variables determine the VND/JPY exchange rate; (2) What k jm the role of the Japanese Yen plays in the economic relationship between Vietnam and gm Japan and (3) Which performance of the multiple regression model and the autoregressive integrated moving average model are in predicting the VND/JPY exchange l.c om rate Methodology focuses on the multiple regression model to define the determinants a Lu Moreover, our study test the reliability in the prediction between multiple regression n model and auto-regressive integrated moving average model to examine the VND/JPY th average (ARIMA), Vietnamese Dong, Japanese Yen, Vietnam, Japan y Key Words: VND/JPY exchange rate, multiple regression model, auto-regressive integrated moving te re better forecasting performance n va exchange rate data Hence, auto-regressive integrated moving average model plays t to TABLE OF CONTENTS ng Table of contents hi ep List of tables List of figures w n List of abbreviations lo ad Chapter one: Introduction ju y th 1.1 Background of study .5 yi 1.2 Research question pl 1.3 Research objective al ua 1.4 The outline of paper n Chapter two: Literature review .9 va n 2.1 Theoretical framework fu ll 2.2 Empirical Studies 14 m oi Chapter three: Methodology 17 nh 3.1 Data .17 at z 3.2 The fundamental regression model .18 z vb ht 3.3 Box-Jenkins’ auto-regressive integrated moving average model (ARIMA) 19 k jm Chapter four: The impact of the Japanese Yen in the economic relationship gm between Vietnam and Japan 23 4.1 Overview of the Vietnamese foreign exchange policy 23 l.c om 4.2 Overview of the Japanese foreign exchange policy 25 a Lu 4.3 The impact of the Japanese Yen in the trade, investment, and finance between n Japan and Vietnam 27 Chapter six: Conclusions .41 th 5.3 Forecasting performance .38 y 5.2 The results and summary of findings 34 te re 5.1 Descriptive statistics .32 n va Chapter five: Results: Descriptive data, multiple regression and ARIMA 32 t to 6.1 Summary of study 41 ng 6.2 Policy implication 42 hi ep 6.3 Limitation of our study and suggestion for further research 42 w References 44 n lo Appendix A 50 ad Appendix B 52 y th ju Appendix C 60 yi Appendix D .70 pl n ua al n va ll fu oi m at nh z z k jm ht vb om l.c gm n a Lu n va y te re th t to LIST OF TABLES ng hi Tables 2.1 Description of economic indicators 12 ep Tables 2.2 Empirical Studies 14 w Tables 3.1 Variable sources 17 n lo Tables 3.2 List of variables 18 ad Tables 3.3 The autocorrelation function (ACF) and the partial autocorrelation function y th (PACF) patterns summary 20 ju yi Tables 4.1 Global foreign exchange reserves .26 pl al Tables 4.2 History of the Japan's interventions in the foreign exchange rate 26 n ua Tables 5.1 Description of the variables .32 va Tables 5.2 Correlation test and Anova F-test 33 n Tables 5.3 The result regression model 34 fu ll Tables 5.4 Wald Test 35 oi m nh Tables 5.5 Unit Root Test 36 at Tables 5.6 ARIMA statistical results 37 z z Tables 5.7 The VND/JPY forecasting performance 38 vb jm ht Tables 5.8 Testing of forecasting ARIMA 39 k Tables 5.9 The advantages and disadvantages in the multiple regression and Auto- om l.c LIST OF FIGURES gm regressive integrated moving average model (ARIMA) .40 a Lu Figure 3.1 Box Jenkins Methodology for ARIMA modeling .19 n Figure 4.1 Value of trade balance Vietnam-Japan 28 n va Figure 5.1 The plot of the monthly VND/JPY exchange rate 36 y te re th t to ng hi ep LIST OF ABBREVIATIONS w n lo ad ju y th yi pl ua al Description Autocorrelation function Auto-regressive integrated moving average Bank of Japan Chicago board of trade Central institute for economic management Fundamental analysis Foreign direct investment Gross domestic product General statistics office of Vietnam International Fisher effect industrial production index Japanese Yen Japan-Vietnam economic partnership agreements Light crude oil (West Texas intermediate oil) Ministry of Foreign Affairs of Japan Ministry of Financial of Japan Ministry of Planning and Investment of Vietnam Multiple regression National Assembly's economic Committee Official development assistance Partial autocorrelation function Royal bank of Australia State bank of Vietnam Technical analysis Trade balance U.S Dollar Vietnam general department of customs Vietnamese Dong Joint stock commercial bank for foreign trade of Vietnam Vietnam food association World trade organization n va ll fu oi m at nh z z k jm ht vb om l.c gm n a Lu y te re th n va abbreviation AF ARIMA BOJ CBT CIEM FA FDI GDP GSO IFE IPI JPY JVEPA LCO (WTI oil) MFAJ MFJ MPIV MR NAEC ODA PAF RBA SBV TA TB USD VGDC VND Vietcombank VFA WTO n S.No 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 t to CHAPTER ONE: INTRODUCTION ng This paper presents the research projects including studies of users, target research to hi ep identify the factors that influence the exchange rate between the Vietnamese Dong and Japanese Yen Accordingly, fluctuations in foreign exchange rate have great w n impact on Vietnamese government, the exporters, importers, commercial banks, and lo ad Japanese investors In addition, they have a need for the research prediction on foreign ju y th exchange rate Therefore, our research focuses on analyzing and predicting the yi fluctuations in VND/JPY exchange rate by multiple regression and auto-regressive pl integrated moving average (ARIMA) Our finding is what determinants of VND/JPY al n va 1.1 Background of study n ua exchange rate in Vietnam ll fu Vietnam economy increasingly integrated into the world economy in term of trade oi m and investment Therefore, exchange rate plays a very important role in the economic nh policy of the Vietnamese government Moreover, many exporters, importers, foreign at investors, and commercial banks, that make the international transactions, are z z impacted by the fluctuations in foreign exchange rate Indeed, many countries fall into vb ht economic hardship due to unstable exchange rate, such as trade deficit, high inflation, k jm increasingly foreign debts, etc Therefore, the exchange rate has attracted special gm attention to the economists, politicians for the study and research In addition, the exchange rate has become an important topic, which is discussed and analyzed on l.c a Lu fluctuations in foreign exchange rate om over the world Many researches have done in order to predict and analyze the n An opening economy is towards the integration with the world economy as well as th y economic performance five years after the accession to the World trade organization" te re Planning and Investment, issued "Comprehensive evaluation of Vietnam’s socio- n va Vietnamese economy According to the report on April 2013, Vietnamese Ministry of t to report1 This report has identified the economic policy focused on foreign trade of ng Vietnam is following the trend of multilateral development cooperation The hi ep International trade of Vietnam and the rest of world become very exciting However, the Vietnam’s exchange rate policy used to concentrate on implementing the pegged w exchange rate policy between Vietnamese Dong and U.S Dollar, and keep floating n lo ad alongside most of other exchange rates It creates some difficult, risky factors for y th exporters, importers, commercial banks, and foreign investors during the international ju payment process for non-dollar currencies such as the Japanese Yen, Euro, and yi pl Australian Dollar etc Indeed, the General statistics office of Vietnam on the ua al international trade in Vietnam in 2012 said that Vietnam has many international n economies Beside the United State of America as Vietnam’s largest trade partner va n with two-way trade turnover reached US$ 27.6 billion, Vietnam still has many key ll fu trading partners such as Japan (approximately U.S Dollar 24.6 billion, 11.1% of at nh Southeast Asian Nations (ASEAN) oi m Vietnamese total trade)2, South China (U.S Dollar 19.5 billion) and the Association of z Japan, in particular, is a country having highly economic relations with Vietnam in z vb many fields including export-import, foreign direct investment capital, Official jm ht development assistance, and etc According to the Department of Foreign Affairs k under the Ministry of Planning and Investment, in the years 2012-2013, Japan gm continues to be the largest donor of official development assistance (ODA) for om l.c Vietnam with 40%3 of total official development assistance (ODA) commitments to Vietnam According to the Foreign investment agency under the Ministry of Planning a Lu and Investment, in 2012, Japan was the biggest foreign direct investment investor in n Vietnam, accounted for 34.2%4 of total investment in Vietnam Thereby, indicating n va th 10 y ‘Comprehensive evaluation of Vietnam’s socio-economic performance five years after the accession to the World Trade Organization’ 2013, report 2013, Central institute for economic management, Vietnamese Ministry of Planning and Investment General statistics office of Vietnam 2013, Vietnam Export-Import report 2011-2012, Vietnam Department of foreign affairs 2013, Official development assistance report 2011-2012, Ministry of Planning and Investment in Vietnam Foreign investment agency 2013, Foreign direct investment report 2011-2012, Ministry of Planning and Investment in Vietnam te re t to XRATE = C(1) + C(2)*IIPVN + C(3)*IRVN + C(4)*CPIVN + C(5)*TRADEVN + C(6)*RICEVN + C(7)*IIPJP + C(8)*IRJP + C(9)*CPIJP + C(10)*TRADEJP + C(11)*OIL + C(12)*XRATE(-1) + [MA(10)=C(13),BACKCAST=2006M02,ESTSMPL="2006M02 2012M12"] ng hi ep Substituted Coefficients: ========================= XRATE = -30.135374111 - 0.0331329298633*IIPVN - 0.187394574032*IRVN + 0.102689340839*CPIVN 0.0397720022944*TRADEVN + 10.2760932463*RICEVN + 0.00867569847378*IIPJP - 4.13190031879*IRJP + 0.282297314597*CPIJP + 0.0627461578307*TRADEJP - 15.0574006791*OIL + 0.987537618531*XRATE(-1) + [MA(10)=0.342905262593,BACKCAST=2006M02,ESTSMPL="2006M02 2012M12"] w n lo ad Restricted Model included XRATE(-1) and MA(10) ju y th yi pl n ua al n va ll fu oi m at nh z z k jm ht vb om l.c gm Estimation Command: ========================= LS(DERIV=AA) XRATE C OIL XRATE(-1) MA(10) a Lu n Estimation Equation: ========================= XRATE = C(1) + C(2)*OIL + C(3)*XRATE(-1) + [MA(10)=C(4),BACKCAST=2006M02,ESTSMPL="2006M02 2012M12"] n va + y te re Substituted Coefficients: ========================= XRATE = 2.99412900742 12.083799246*OIL + 0.991451908091*XRATE(-1) [MA(10)=0.265782214267,BACKCAST=2006M02,ESTSMPL="2006M02 2012M12"] th 61 t to Breusch-Godfrey Serial Correlation LM Test: ng hi ep w n lo ad ju y th yi pl n ua al n va ll fu oi m at nh z z k jm ht vb om l.c gm n a Lu n va y te re th 62 t to ng hi ep w n lo ad ju y th yi pl n ua al n va ll fu oi m at nh z z k jm ht vb Wald Test: om l.c gm n a Lu n va y te re th 63 t to APPENDIX C ng In an ARIMA model, the paper does not have a priori for forecasting model hi ep before model identification takes place ARIMA helps us to choose a “right model” to best fit the time series Put it in a flow chart: w n lo ad ju y th yi Demonstration to find "right ARIMA model (p, d, q)" to fit the time series pl through trial and error The resulting graphs are: n ua al n va ll fu oi m at nh z z k jm ht vb om l.c gm n a Lu n va y te re th 64 t to ng hi ep w n lo ad ju y th yi pl n ua al n va The PACFs are used to determine the order of the AR component, and ACFs to decide the order of the MA component ll fu oi m ARIMA(p,d,q) – p is the order of the AR component; d is the level# of difference; q is the order of the MA component at nh z z k jm ht vb gm From the above graphs, you can see that the time series is likely to have random l.c om walk pattern, which random walk up and down in the line graph Also, in correlogram, the ACFs are suffered from linear decline and there is only one significant spike for PACFs The a Lu graph of correlogram suggests that ARIMA(1, 0, 0) may be an appropriate model Then, the n n before further finding AR(p) and MA(q) va paper takes the first-difference of "xrate" to see whether the time series becomes stationary 65 th also get a correlogram graph y paperneed to generate it "Dxrate=d(xrate)" Then, you will use it to draw a line graph and te re To see whether first difference can get level-stationary time series or not, the t to ng hi ep w n lo ad ju y th yi pl n ua al n va ll fu oi m at nh z z k jm ht vb om l.c gm n a Lu n va Unit root test on Dxrate y te re th 66 t to ng hi ep w n lo ad ju y th yi pl n ua al n va ll fu oi m at nh z z k jm ht vb gm Result from unit root test show that the Augmented Dickey-Fuller test statistic = 6.96358 is outside the range of the test critical values at 1%, 5% and 10%, hence the paper l.c om reject the null hypothesis that D(xrate) has a unit root or in other words, D(xrate) is n a Lu stationary graph and is white noise as shown no significant patterns in the graph of correlgram And the y support that the ARIMA(0,1,0) is suitable for the time series te re unit root test also confirms the first-difference becomes stationary The strong evident n va Now, the first-difference series "Dxrate" becomes stationary as showing in line th Then, the paper can construct the ARIMA model as following steps: 67 t to ng hi ep w n lo ad ju y th yi pl n ua al n va ll fu oi m at nh z z k jm ht vb om l.c gm n a Lu n va y te re Seems that AR(1) still exist The paper can another test LAGRANGE MULTIPLIER TEST (LM TEST) to confirm: Lags = means the paper test whether both AR(1) and AR(2) =0 th 68 t to ng hi ep w n lo ad ju y th yi pl n ua al n va ll fu oi m at nh z z From the result, P-value = 0.0844 > 0.05 >>> The paper fail to reject the H0 that both AR(1) and AR(2) = So the paper continue test with lags = (test whether AR(1) = 0) k jm ht vb om l.c gm n a Lu n va y te re th 69 t to ng hi ep w n lo ad ju y th yi pl n ua al n va ll fu oi m at nh z The test reject H0 that AR(1) = >>> AR(1) does exist Now, the paper turn to estimation equation again to run regression with both AR(1) and AR(2) included: z k jm ht vb om l.c gm n a Lu n va y te re th 70 t to ng hi ep w n lo ad ju y th yi pl n ua al n va ll fu oi m at nh z z k jm ht vb om l.c gm n a Lu n va y te re th Now, since there is no significant spikes of ACFs and PACFs, and the LM test strongly reject the null hypothesis, it means that the resduals of this selected ARIMA model 71 t to are white noise, so that there is no other significant patterns left in the time series, then the ng paper can stop at here and don't need to further consider another AR(p) and MA(q) hi ep The ARIMA(2,0,0) is a good model with white noise and without AR w n lo ad ju y th yi pl n ua al va The criterions to judge for the best model are as follows: n - Relatively small of BIC (Schwarz criterion which is measured by nLog(SEE)+kLog(n)) ll oi m - Relatively high adjust R2 fu - Relatively small of SEE at nh - Q- statistics and correlogram show that there is no significant pattern left in the ACFs and z PACFs of the residuals, it means the residuals of the selected model are white noise z k jm ht vb om l.c gm n a Lu n va y te re th 72 t to The ARIMA (1, 0, 1) ng hi ep w n lo ad ju y th yi pl n ua al n va ll fu oi m at nh z z k jm ht vb om l.c gm n a Lu n va y te re th 73 t to APPENDIX D ng hi The forecasting performance results of fundamental regression ep w n lo ad ju y th yi pl n ua al n va ll fu oi m at nh The forecasting performance results of ARIMA (2,0,0) model z z k jm ht vb om l.c gm n a Lu n va y te re th 74 t to The forecasting performance results of ARIMA (1,0,1) model ng hi ep w n lo ad ju y th yi pl n ua al n va ll fu oi m at nh z z k jm ht vb om l.c gm n a Lu n va y te re th 75