Determinants of foreign exchange rate, case of vietnamese dong and japanese yen

UNIVERSITY OF ECONOMICS
HO CHI MINH CITY
VIETNAM

INSTITUTE OF SOCIAL STUDIES
THE HAGUE
THE NETHERLANDS

VIETNAM - NETHERLANDS
PROGRAMME FOR M.A IN DEVELOPMENT ECONOMICS

DETERMINANTS OF FOREIGN EXCHANGE
RATE: CASE OF VIETNAMESE DONG AND
JAPANESE YEN

BY

Mr. TRAN VUONG TU

MASTER OF ARTS IN DEVELOPMENT ECONOMICS
HO CHI MINH CITY, MAY 2013


UNIVERSITY OF ECONOMICS
HO CHI MINH CITY
VIETNAM

INSTITUTE OF SOCIAL STUDIES
THE HAGUE
THE NETHERLANDS

VIETNAM - NETHERLANDS
PROGRAMME FOR M.A IN DEVELOPMENT ECONOMICS

DETERMINANTS OF FOREIGN EXCHANGE
RATE: CASE OF VIETNAMESE DONG AND
JAPANESE YEN

A thesis submitted in partial fulfilment of the requirements for the degree of

MASTER OF ARTS IN DEVELOPMENT ECONOMICS

By

Mr. TRAN VUONG TU

Academic Supervisor:
PhD. NGUYEN HOANG BAO

HO CHI MINH CITY, MAY 2013


ACKNOWLEDGEMENT
This thesis was written at the University of Economics Ho Chi Minh City. In
addition, it was completed in October 2013. During the process of writing, the
paper has gained a lot of experience in writing a thesis and in the area of
foreign exchange rate analysis. During the three months of writing this thesis,
several persons have contributed in the different ways to the quality of this
thesis and the paper would like to take this opportunity to thank them.
Firstly, the paper would like to thank our supervisor PhD. Nguyen Hoang Bao for all
the help, guidance, and support. The paper would also like to express gratitude to all

professors of the Vietnam-Netherlands Program for the Master in Development
Economics and the classmates who offer to us some useful suggestions. Finally, we
express special thank to our families and partners for their love and support.


ABSTRACT
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 effects many exporters, importers, foreign investors,
and commercial banks in the international transaction.
Japanese economy plays as important as having mainly economic relations with Vietnamese
economy in the export-import trade, foreign direct investment (FDI) capital, official development
assistance (ODA), etc. However, Vietnam government applies the floating exchange rate policy
between Vietnamese Dong and the Japanese Yen. Therefore, the fluctuations of VietnameseJapanese exchange rate might great impact on the trade and investment. The exporters and
importers of two countries, Japanese investors, the commercial bankers that having
international settlement with Japanese Yen, are in need of defending the exchange rate risk
volatility of the exchange rate pairs.

Our study enhance on analyzing and predicting the fluctuations of VietnameseJapanese exchange rate. The main research question identifies (1) Which
Vietnamese and Japanese macroeconomics variables determine the VND/JPY
exchange rate; (2) What the role of the Japanese Yen plays in the economic
relationship between Vietnam and Japan and (3) Which performance of the multiple
regression model and the auto-regressive integrated moving average model are in
predicting the VND/JPY exchange rate. Methodology focuses on the multiple
regression model to define the determinants. Moreover, our study test the reliability
in the prediction between multiple regression model and auto-regressive integrated
moving average model to examine the VND/JPY exchange rate data. Hence, autoregressive integrated moving average model plays better forecasting performance.
Key Words: VND/JPY exchange rate, multiple regression model, auto-regressive
integrated moving average (ARIMA), Vietnamese Dong, Japanese Yen, Vietnam, Japan

TABLE OF CONTENTS
Table of contents............................................................................................................................ 1
List of tables...................................................................................................................................... 3
List of figures.................................................................................................................................... 3
List of abbreviations.................................................................................................................... 4
1. Chapter one: Introduction.................................................................................................. 5
1.1 Background of study........................................................................................................... 5
1.2 Research question................................................................................................................ 7
1.3 Research objective............................................................................................................... 8
1.4 The outline of paper............................................................................................................. 8
2. Chapter two: Literature review....................................................................................... 9
2.1 Theoretical framework....................................................................................................... 9
2.2 Empirical Studies................................................................................................................... 14
3. Chapter three: Methodology............................................................................................. 17
3.1 Data.................................................................................................................................................. 17
3.2 The fundamental regression model......................................................................... 18
3.3 Box-Jenkins’ auto-regressive integrated moving average model (ARIMA)
..................................................................................................................................................................... 19

4. Chapter four: The impact of the Japanese Yen in the economic relationship

between Vietnam and Japan.................................................................................................. 23
4.1 Overview of the Vietnamese foreign exchange policy................................ 23
4.2 Overview of the Japanese foreign exchange policy..................................... 25
4.3 The impact of the Japanese Yen in the trade, investment, and finance between

Japan and Vietnam....................................................................................................................... 27
5. Chapter five: Results: Descriptive data, multiple regression and ARIMA
..................................................................................................................................................................... 32

5.1 Descriptive statistics........................................................................................................... 32
5.2 The results and summary of findings..................................................................... 34
5.3 Forecasting performance................................................................................................. 38
6. Chapter six: Conclusions................................................................................................... 41
5


6.1 Summary of study................................................................................................................. 41
6.2 Policy implication.................................................................................................................. 42
6.3 Limitation of our study and suggestion for further research................42
References.......................................................................................................................................... 44
Appendix A......................................................................................................................................... 50
Appendix B......................................................................................................................................... 52
Appendix C......................................................................................................................................... 60
Appendix D......................................................................................................................................... 70

6


LIST OF TABLES
Tables 2.1 Description of economic indicators........................................................ 12
Tables 2.2 Empirical Studies.................................................................................................. 14
Tables 3.1 Variable sources.................................................................................................... 17
Tables 3.2 List of variables...................................................................................................... 18
Tables 3.3 The autocorrelation function (ACF) and the partial autocorrelation function

(PACF) patterns summary........................................................................................................ 20
Tables 4.1 Global foreign exchange reserves........................................................... 26
Tables 4.2 History of the Japan's interventions in the foreign exchange rate 26

Tables 5.1 Description of the variables.......................................................................... 32
Tables 5.2 Correlation test and Anova F-test............................................................. 33
Tables 5.3 The result regression model......................................................................... 34
Tables 5.4 Wald Test..................................................................................................................... 35
Tables 5.5 Unit Root Test.......................................................................................................... 36
Tables 5.6 ARIMA statistical results................................................................................. 37
Tables 5.7 The VND/JPY forecasting performance................................................ 38
Tables 5.8 Testing of forecasting ARIMA...................................................................... 39
Tables 5.9 The advantages and disadvantages in the multiple regression and Auto-

regressive integrated moving average model (ARIMA)...................................... 40
LIST OF FIGURES
Figure 3.1 Box Jenkins Methodology for ARIMA modeling.............................19
Figure 4.1 Value of trade balance Vietnam-Japan.................................................. 28
Figure 5.1 The plot of the monthly VND/JPY exchange rate...........................36

7


LIST OF ABBREVIATIONS
S.No
1
2
3
4
5
6
7
8

9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31

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

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

8


CHAPTER ONE: INTRODUCTION
This paper presents the research projects including studies of users, target research
to identify the factors that influence the exchange rate between the Vietnamese Dong
and Japanese Yen. Accordingly, fluctuations in foreign exchange rate have great
impact on Vietnamese government, the exporters, importers, commercial banks, and
Japanese investors. In addition, they have a need for the research prediction on
foreign exchange rate. Therefore, our research focuses on analyzing and predicting
the fluctuations in VND/JPY exchange rate by multiple regression and auto-regressive
integrated moving average (ARIMA). Our finding is what determinants of VND/JPY
exchange rate in Vietnam.

1.1 Background of study
Vietnam economy increasingly integrated into the world economy in term of trade
and investment. Therefore, exchange rate plays a very important role in the
economic policy of the Vietnamese government. Moreover, many exporters,
importers, foreign investors, and commercial banks, that make the international
transactions, are impacted by the fluctuations in foreign exchange rate. Indeed,
many countries fall into economic hardship due to unstable exchange rate, such
as trade deficit, high inflation, increasingly foreign debts, etc. Therefore, the
exchange rate has attracted special 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 over the world. Many researches have
done in order to predict and analyze the fluctuations in foreign exchange rate.
An opening economy is towards the integration with the world economy as well as
Vietnamese economy. According to the report on April 2013, Vietnamese Ministry of
Planning and Investment, issued "Comprehensive evaluation of Vietnam’s socioeconomic performance five years after the accession to the World trade organization"

9


1

report . This report has identified the economic policy focused on foreign trade of
Vietnam is following the trend of multilateral development cooperation. The
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
exchange rate policy between Vietnamese Dong and U.S Dollar, and keep floating
alongside most of other exchange rates. It creates some difficult, risky factors for
exporters,

importers,

commercial

banks,

and

foreign

investors

during

the

international payment process for non-dollar currencies such as the Japanese Yen,

Euro, and Australian Dollar etc. Indeed, the General statistics office of Vietnam on the
international trade in Vietnam in 2012 said that Vietnam has many international
economies. Beside the United State of America as Vietnam’s largest trade partner
with two-way trade turnover reached US$ 27.6 billion, Vietnam still has many key
trading partners such as Japan (approximately U.S Dollar 24.6 billion, 11.1% of
2

Vietnamese total trade) , South China (U.S Dollar 19.5 billion) and the Association of
Southeast Asian Nations (ASEAN).
Japan, in particular, is a country having highly economic relations with Vietnam in
many fields including export-import, foreign direct investment capital, Official
development assistance, and etc. According to the Department of Foreign Affairs
under the Ministry of Planning and Investment, in the years 2012-2013, Japan
continues to be the largest donor of official development assistance (ODA) for
3

Vietnam with 40% of total official development assistance (ODA) commitments to
Vietnam. According to the Foreign investment agency under the Ministry of Planning
and Investment, in 2012, Japan was the biggest foreign direct investment investor in
4

Vietnam, accounted for 34.2% of total investment in Vietnam. Thereby, indicating

1

‘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
2
General statistics office of Vietnam 2013, Vietnam Export-Import report 2011-2012, Vietnam

3
Department of foreign affairs 2013, Official development assistance report 2011-2012,
Ministry of Planning and Investment in Vietnam
4
Foreign investment agency 2013, Foreign direct investment report 2011-2012, Ministry
of Planning and Investment in Vietnam

10


the exchange rate fluctuations between the Vietnamese Dong and the
Japanese Yen have great impact on Vietnam's economy in general and
the exporters, importers, banks and Japanese investors in particular.
In studies related to the exchange rate between Vietnamese Dong and foreign
currencies, normally only focused on fixed exchange rate policy between
Vietnamese Dong and the U.S Dollar. There is virtually no study related to the
pairs of floating exchange rate between Vietnamese Dong and the Japanese Yen
or Euro or Pound. Considering the trade, economic relations, and investment
cooperation, the researches on the fluctuations in VND/JPY exchange rate plays
an important role in the trade policy between Vietnam and Japan, in the planning
for investment reception between Vietnam and Japan (official development
assistance and foreign direct investment). It would be clearly explain in chapter
five. In addition, the exporters, importers, Japanese investors, and commercial
banks have the international settlement with Japanese Yen. They are in need of
defending the exchange rate risk volatility of the exchange rate.
Most of all, our study focuses on the analysis of determinants of foreign exchange
rate. Moreover, our study enhances for using the fundamental analysis to define the
determinants of the VND/JPY exchange rate by the multiple regression model and
using the Auto-regressive integrated moving average to examine the VND/JPY
exchange rate. Absolutely, our study investigates to find out what determinants of

VND/JPY exchange rate in Vietnam. Besides that, our study further clarifies the role of
the Japanese Yen in the economic relationship between Vietnam and Japan.

1.2 Research question
Hence, our study can suggest three-research questions as follows: (1) Which
Vietnamese and Japanese macroeconomics variables determine the VND/JPY
exchange rate; (2) What the role of the Japanese Yen plays in the economic
relationship between Vietnam and Japan and (3) Which performance of the multiple

11


regression model and the auto-regressive integrated moving average
model are in predicting the VND/JPY exchange rate.
Our study enhances the multiple regression model and the auto-regressive integrated
moving average model to econometric test the VND/JPY exchange rate.

1.3 Research objective
Our study identifies three objectives as follows: (1) to find out which
macroeconomics variables determine the VND/JPY exchange rate; (2) to identify
the role of the Japanese Yen in the economic relationship between Japan-Vietnam
and (3) To test forecasting performance of the multiple regression model and the
auto-regressive integrated moving average model in the VND/JPY exchange rate.
The exporters, importers, Japanese investors, and commercial banks, that have the
international settlement with Japanese Yen, are our research objective. They always
face to the exchange rate risk volatility of the exchange rate. Therefore, they are
interested in which macroeconomics variables determine the VND/JPY exchange rate
and how to use the econometric model in predicting the VND/JPY exchange rate.

1.4 The outline of paper

Our study is divided into five chapters as: chapter one briefs on the research
projects including studies of users, target research to identify the factors. They
influence the exchange rate between the Japanese Yen and Vietnamese Dong.
Chapter two presents many theories as the foundation for the study including the
analysis of the basic index of both countries (industrial index, consumer price
index, interest rates, value of balance trade, prices of rice and prices of light
crude oil) and auto-regressive integrated moving average model (ARIMA). Chapter
three presents research methods and research data. Chapter four shows overview
the foreign exchange rate policies of the two countries Vietnam and Japan.
Moreover, it highlights the influence of the Japanese Yen in relation to economic
and trade between Vietnam and Japan. Chapter five explains the results of
finding. Finally, chapter six shows conclusion and further study.
12


CHAPTER TWO: LITERATURE REVIEW
Literature review is essentially an organized collection of theoretical framework and
empirical studies. Chapter two presents theories as the foundation for the study.
There is fundamental and technical analysis. Fundamental analysis shows the
multiple regression model to examine macroeconomic variables of both countries
(industrial index, consumer price index, interest rates, value of balance trade, prices
of rice, and prices of international crude oil). Technical analysis tests the VND/JPY
exchange rate by auto-regressive integrated moving average (ARIMA).

2.1 Theoretical framework
Two main theoretical frameworks will be discussed in this study: (1) Fundamental
analysis is based on a thorough examination of macroeconomic variables that
have an impact on the currency; (2) Technical analysis is the opposite of
fundamental analysis solely focuses on prices, and movements in the past,
ignoring the economic factors and policies. Technical analysis focuses on price

action. It is based on the history of the exchange rate. It is expected that all the
essential information is already included in the price. Our study enhances the
Auto-regressive integrated moving average model in technical analysis.

2.1.1 Fundamental analysis on the multiple regression models
Faust, Rogers, and Wright (2003) emphasize that the exchange rates under a floating
exchange rate depend on the demand and supply of a currency exchange rate. If the
delivery of the currency exceeded demand, the devaluation of this currency and vice
versa. Depending on the asset market model shows "the exchange rate between two
currencies represents the price that balances for shipments and demand for foreign
currency denominated assets." the expectations of the foreign exchange market and
changes in the direction of expectations, changes in demand and supply of money
and the exchange rate. They suggest two important theories: (1) purchasing power
parity and (2) international Fisher effect.

13


The theory of purchasing power parity (PPP) initially formed by Cassel (1918),
employs the long-term equilibrium exchange rate between the two currencies to
balance purchasing power in their home countries. Therefore, the PPP exchange
rate is the rate at which the two currencies are equal by removing the
discrepancies in the price levels between nations. There are two versions of
purchasing power parity theory: (1) the absolute version states that when
converted into a common currency, the price levels in the world should be equal.
In other words, a unit of common currency at home should have the same
purchasing power worldwide; (2) the relative version says that depending on the
changes in the price levels between the two countries the exchange rate between
home currency and foreign currency will be corrected to reflect the changes in the
5

price levels . The relative version can be presented as follows:
et

h
e0 . (1 i

)t

(2.1)

(1 i f )t

where,
et is the spot foreign exchange rate in period t;
e0 is the foreign exchange rate at the beginning of the period;
ih is the inflation rate in the home nation; and,
if is the inflation rate in the foreign nation.
The purchasing power parity is represented by the following approximation

(2.2)

5 Shapiro, A.C. Multinational Financial Management. Seventh edition. Wiley and Sons.

14


Rationally, purchasing power parity is that higher inflation may lead to a
6

currency that should depreciate against a country with low inflation .
Moreover, Zhang and Wu (2011) suggest that the price of good and products
that the country exports or imports are maybe affected the purchasing power.
Hence, purchasing power parity suggests some economic variables that affect
the change of the exchange rate such as the inflation, trade balance, the
industrial, purchasing power and the price of mainly export or import good.
The generalized version of impact international Fisher effect (IFE) said that an
interest rate higher than the share with a low inflation must withstand the
currency with a high annual rate of inflation. Differently, purchasing power
parity means that rates will move; match the changes in the differential rates
of inflation. The combination of these two conditions is the effect of
international Fishermen. International Fisher effect can be specified as:

(2.3)
where
rh is the interest rate in the home country;
rf is the interest rate abroad;
E(et ) is the expected foreign exchange rate at period t; and,
e0 is the foreign exchange rate at the beginning of the period t.
Speculators will receive a positive return when they sell the currency of economy X
and buy the currency of economy Y, if the interest rate in the economy Y is lower than
in economy X. They expect the appreciation of the currency of economy Y. This
purchasing of currency of economy Y speculated or invested in high-yielding assets.
The result is appreciation of the currency by the increasing demand. In the rational,

6 For example, if the rate of inflation in the country X is 15% and the inflation in country Y is 7%, the
currency of country Y maybe appreciate roughly 8% compare of the currency of the country X.

15

the lower interest rate of a currency leads to a lower inflation rate. It
affects the price of currency. Hence, the international Fisher effect said
that interest rates have a significant effect on the exchange rate.
Based on above theories, many economic indicators seem to be important to explain
movements in exchange rates. Historically, Zhang and Wu (2011) suggested many
indicators in examination of AUD/JPY exchange rate. They were the gold price, oil
price, Australian employment data, Australian consumer price index (CPI), Australian
gross domestic product (GDP), Australian trade balance, Royal Bank of Australia
(RBA) rate decisions, Japanese consumer price index (CPI), Japanese trade balance,
Japanese industrial production index, and Bank of Japan (BOJ) interest rate
decisions. Therefore, our study makes decision for applying some economic
indicators to test the fluctuations in VND/JPY exchange rate as follows:

Table 2.1: Description of economic indicators
Economic
7
indicators

The price of
1

rice

The price of
light crude oil
2 (West Texas
Intermediate
oil)

Economic mechanism through which each variable might impact on the
foreign exchange rate
There was a high percentage of total exports of commodity products in
account for Vietnam export, world prices may help explain the price
movements of long-term exchange rate this product as the Vietnamese
Dong.
Sometimes this means that entrepreneurship is Vietnamese Dong just like
trading rice. Vietnam is the second largest producer of rice and the
Vietnamese Dong has generally a positive correlation with the price of rice,
which in the case of an increase in the prices of rice, Vietnamese Dong
tends to grow.
The Japanese economy is heavily dependent on Light crude oil, so import
price fluctuations affect the value of the Japanese Yen. As a rule of
Japanese Yen is only a negative correlation with the price of crude oil. High
oil prices increase the cost of goods and services, as well as slowing the
growth of the Japanese economy. As a rule, if the oil price appreciates, the
Japanese Yen depreciates.

Index of Industrial Production (IPP) is basically the growth in industrial
production of the country. It has a close relation to the growth of Gross
domestic product (GDP) to assess the economic development of an
3 index-industrial
economy. Accordingly, IIP may affect the exchange rate of the domestic
production (IIP)
currency against foreign currencies. In Vietnam, it was published by the
General Statistics Office of Vietnam.
Vietnamese

7 Expected sign of the coefficients that are showed in table 3.2

16


4

Vietnamese
Vietnamese consumer price index (CPI) gauges the average change in retail
consumer price prices for a fixed market basket of goods and services. CPI helps the
index (CPI)
researchers and investors to measure the inflation in Vietnam.

5

The data are important for foreign exchange rate in the long run. Vietnam is
a net importer, meaning that Vietnam imports more goods than it exports.
Changing in the trade balance maybe affects the price of the Vietnamese
Dong.

6

7
8

Vietnamese
trade balance
State bank of
Vietnam (SBV)
interest rate
decisions for
Vietnamese

Dong
Japanese
consumer price
index (CPI)

The State Bank of Vietnam (SBV) likely set the main interest rate which is
based on three important factors: consumer price index, credit growth rate
and money supply. So the exchange rates, therefore, the interest rate policy
emphasis the foreign exchange rate.
Japanese consumer price index (CPI) gauges the average change in retail
prices for a fixed market basket of goods and services. CPI helps the
researchers and investors to measure the inflation in Japan.

Japanese trade Japan plays as well as export country. Therefore, the surplus of trade
balance

balance has impact in the changing of the value of the Japanese Yen.

Index of Industrial Production (IPP) is basically the growth in industrial
Japanese index- production of the country. It has a close relation to the growth of Gross
9 industrial
domestic product (GDP) to assess the economic development of an
production (IIP) economy. Accordingly, IIP may affect the exchange rate of the domestic
currency against foreign currencies.
The Bank of
The Bank of Japan (BOJ) likely set the main interest rate, which is based on
Japan (BOJ)
three important factors: consumer price index, credit growth rate and money
10 interest rate
supply. So the exchange rates, therefore, the interest rate policy emphasis

decision for
the foreign exchange rate.
Japanese Yen

2.1.2 Technical analysis
Technical analysis is a method of analyzing and forecasting the direction of prices
through the study of historical price data, primarily price and volume. A basic
principle of technical analysis is the market price reflects all relevant information, so
their analysis looks at the history of the model of the stock exchange rather than the
external controls such as economic, fundamental and news. So, price action tends to
repeat itself because investors generally tend to be patterns of behavior - such as
technical analysis focuses on trends and conditions can be identified
Auto-regressive integrated moving average (ARIMA), which is one of tools in the
technical analysis, can predict the future price from historical data (time series data
sets). Auto-regressive integrated moving average model is the model for predicting
17


time series data. The model uses no other independent variables but the
prediction will come out from historical exchange rates. Auto-regressive
integrated moving average model requires large run of time series data
and technical expertise on the part of forecaster.
The current values of a time series in terms of past values of itself (the autoregressive component) and past values of the error term (the moving average
terms) were represented by an auto-regressive integrated moving average
model. The number of times a series, which were referred by the integrated
component, to must be differentiated to play stationary. Auto-regressive
integrated moving average model are agnostic in forecasting time series.

2.2 Empirical Studies
Historically, there are many researches related on the foreign exchange

rate that showed as follows:
Table 2.2 Empirical Studies
Authors
Research
Dobre, I. and They
applied
the Box-Jenkins
Maria,
A. methodology to estimate unemployment
(2008)
rate during monthly data from 1998 to
2007 period.
Faust,
J., They examine the forecasting performance
Rogers, J.H. of standard macro-economic models of
and Wright, J, foreign exchange rates in real time, using
H., (2003),
the main economic indicators database of
the organization for economic co-operation
and developments (OECD)
.They
calculated out of sample forecasts, as they
would have been made at the time, and
compares them to a random walk
alternative.

Result
The evidence is presented that ARIMA
(2,1,2) plays the most adequate model
for the unemployment rate.

Kilian,
L.(1997)

The evidence is presented that the linear
VEC model framework underlying the
empirical study is likely to be misspecified, and that the methodology for
constructing bootstrap p-values for longhorizon regression tests may be
fundamentally awed.

He illustrated the use of four major
exchange rates by a new bootstrap method.
It was for small-sample inference in long
horizon regressions by analyzing the long
horizon predictability. In addition, the
findings are reconciled with those of an
earlier study.

The resulting of time series testing
indicates that standard macro-economic
models have large effects on exchange
rate predictability.

18


Massarrat,
and Khan,
(2013),

M. He used auto-regressive integrated moving
A., average (ARIMA) as well as building the
model to develop a predicting model for
gold price.

After that checking the accuracy of the
model in predicting performance
Meese, Richard They issue that the study in a number of
A., and Rogoff, time series and structural models, which
Kenneth (1982) was based on out of sample predicting
accuracy.

He have used Box-Jenkins for making
the predicting model.
ARIMA(0,1,1) is the most suitable
model to be used for predicting the gold
price.
As the result, they discovered that one to
twelve month horizons for the
USD/DEM, USD/GBP, USD/JPY and
trade-weighted dollar exchange rates was
estimated by a random-walk model

Meyler, Aidan Auto-regressive integrated moving
and
Kenny, average (ARIMA) time series models was
Geoff
and applied for forecasting Irish inflation.
Quinn, Terry They identified the ARIMA models - the
(1998),

Box Jenkins approach on forecast
performance.

They suggested that the approach focus
on maximizing in-sample ‘goodness of
fit’ and minimizing out of sample
forecast errors. Thus, the approach
followed is unashamedly one of ‘model
mining’ with the aim of optimizing
forecast performance.
The harmonized index of consumer
prices (HICP) and some of its major subcomponents play practical issues in
ARIMA time series forecasting.

Nochai, R. and The objective of the research is to find an
Nochai,
T., appropriate ARIMA model for predicting
(2006),
oil palm price and wholesale price of oil
palm of Thailand in the period from 2000
to 2004.

ARIMA (2,1,0) plays for predicting farm
price of oil palm.
ARIMA (1,0,1) or ARMA(1,1) were
suitable for predicting wholesale price of
oil palm.

Zhang, Y., and Their study focuses the relationship The prediction result of the ARIMA
Wu, H., (2011), between the AUD/JPY exchange rate and model is more accurate than those based

some economic fundamentals by using a on the fundamental regression model.
regression model.
Source: Author’s collection from different sources
Summarily, the study finds that a topic relating to the analysis of exchange rate volatility is not
a new subject in the world. Two main theoretical frameworks will be discussed in this study:
(1) Fundamental analysis is based on a thorough examination of macroeconomic variables
and policies that have an impact on the currency. Our study makes decision for applying
some economic indicators for VND/JPY exchange rate. They are the price of rice, the price of
light crude oil (West Texas Intermediate oil), Vietnamese industrial production index,
Vietnamese consumer price Index (CPI), Vietnamese trade balance, State bank of Vietnam
(SBV) interest rate decisions for Vietnam Dong, Japanese industrial production index,
Japanese consumer price

19


index (CPI), Japanese trade balance, and Bank of Japan (BOJ) interest rate decisions
for Japanese Yen. (2) Technical analysis is Auto-regressive integrated moving
average (ARIMA) model that can predict the future price from historical data (time
series data sets). Recently research, M. Massarrat, M. and Khan, A., (2013), has made
research and analysis related to gold price prediction tools in Auto-regressive
integrated moving average (ARIMA). Results suggest that ARIMA (0,1,1) is the most
suitable model to be used for predicting the gold price. Ying Zhang and Hailun Wu
(2011) use the auto-regressive integrated moving average (ARIMA) model and
multiple regression model to analysis Australia Dollar and Japanese Yen. The
prediction result of the auto-regressive integrated moving average model (ARIMA) is
more accurate than those based on the multiple regression model.

20

CHAPER THREE: RESEARCH METHODOLOGY
Chapter three explains the number of observations and sources of database
that applied for testing the multiple regression model and Auto-regressive
integrated moving average model. Our study also emphases the Box-Jenkins
methodology in auto-regressive integrated moving average plays the model
identification and model selection, parameter estimation and model checking,
It examine the past value in time series, in order to make forecasts. Moreover,
the paper formulates a multiple regression model, which economic
fundamentals influence the price movement of the VND/JPY exchange rate.

3.1 Source of the data sets
The data in this study is mainly based on the monthly from January 2007
(the 1

st

observation) to December 2012, which are includes 84

observations. Sources of data sets are follows:
Table 3.1: Variable sources
Variable

Name of website

Sources

Light crude oil (West
Chicago board of trade Retrieved April 4, 2013 from
Texas Intermediate oil)

/>Vietnamese export rice Vietnam food association Retrieved April 18, 2013 from
price
/>Vietnamese industrial
production index,
Vietnamese consumer
price index (CPI)

General statistics office Retrieved April 18, 2013 from


Vietnamese trade balance Vietnam general
Retrieved April 18, 2013 from
department of customs
Interest rate of
Vietnamese Dong

The State bank of
Vietnam

Retrieved April 7, 2013 from
/>
VND/JPY exchange rate Joint stock commercial Retrieved April 8, 2013 from
bank for foreign trade of />Vietnam (Vietcombank) tes/

21


The Japanese consumer
price index (CPI),
Japanese trade balance,

Japanese industrial
production index and
interest rate decision of
Japanese Yen

Bank of Japan

Retrieved April 4, 2013 from
/>
Sources: Author’s collection

3.2 The multiple regression model
In our study, the paper formulates a multiple regression model, which is
used to determine how much the economic fundamentals such as the
price of a commodity, interest rates, and so on influence the price
movement of the VND/JPY exchange rate. The model is specified as:
Xratet = α1 + α2CPIjpt + α3CPIvnt + α4IIPjpt + α5IIPvnt + α6IRjpt + α7IRvnt +
α8∆tradejpt + α9∆tradevnt + α10∆ricet + α11∆oilt + ε t
The variables are defined in the below table:
Table 3.2: List of variable
Variable

Description

Xratet

Value of VND/JPY rate over time t.

CPIjpt

The consumer price index (CPI) index is
calculated as a percentage to reflect the
change in relative prices of Japanese
consumer goods over time t.

CPIvnt

The consumer price index (CPI) index is
calculated as a percentage to reflect the
change in relative prices of Vietnam
consumer goods over time t.

Expected
sign of the
coefficients

+

IIPjpt

Index-Industry Products referred to as the
IIP index is an indicator of industrial
activity produced of Japan

IIPvnt

Index-Industry Products referred to as the
IIP index is an indicator of industrial
activity produced by the General Statistics
Office of Vietnam announced by time t

-

The interest rate of Japanese Yen at time t.

+

IRjpt

+

Theory at literature
reviews
International Fisher
Effect

International Fisher
Effect

Purchasing Power
Parity
Purchasing Power
Parity

International Fisher
Effect

22

IRvnt

The interest rate of Vietnam Dong at time
t.

-

International Fisher
Effect

∆tradejpt

The percentage change in the Japanese
trade balance at time t.

+

Purchasing Power
Parity

∆tradevnt This is the percentage change in the
Vietnam trade balance. Since there is an
adverse balance of trade, we use the first
difference divided by the absolute value of
the lagged trade balance at time t.

Purchasing Power
Parity

-

∆ricet

The percentage change in the Vietnam
export rice price at time t.

*

∆oilt

The percentage change in the crude oil
price at time t.

*

* a considerations was unclear

3.3 Box-Jenkins’ auto-regressive integrated moving average model (ARIMA)

Meyler, Aidan and Kenny, Geoff and Quinn, Terry (1998) proposed the BoxJenkins methodology for forecasting Irish inflation. The Box-Jenkins
methodology plays the model identification and model selection, parameter
estimation and Model checking for examining the past value in time series,
in order to make forecasts. It has predictability in foreign exchange rate is
accepted in many researchers and countries. They illustrates as follows:

Figure 3.1: Box Jenkins methodology for ARIMA modeling

Sources: Meyler, Aidan and Kenny, Geoff and Quinn, Terry (1998)

Step 1 (Data collection and examination): Meyler, Aidan and Kenny, Geoff and

Quinn, Terry (1998) recommended a lengthy time series of data (at least 50
observations) is required for univariate time series forecasting. In study, the
23


VND/JPY monthly database shows 84 observations from January 2007
st

(the 1 observation) to December 2012 (the 84

th

observation).

Step 2 (Determine stationary of time series): Gujarati (2002) defined a
stationary time series is a series with constant mean, variance and autocovariance at various lags. He also introduced a rule of thumb to choice an
appropriate lag length. It is recommended to compute the autocorrelation
function (ACF) up to one-third to one-quarter the length of the time series.
Step 3 (Model identification and estimation): As mentioned above, d is order of nonseasonal differences to make time series stationary. The next task is to determine the
value of p and q the paper use the graphical properties of the autocorrelation function
(ACF) and the partial autocorrelation function (PACF). Gujarati, (2002) proposed
theoretical

patterns

of

the

autocorrelation

function

(ACF)

and

the

partial

autocorrelation function (PACF), which are summarized in follows:

Table 3.3 The autocorrelation function (ACF) and the partial
autocorrelation function (PACF) patterns summary
Type of Model
AR(p)
MA(q)
ARMA(p,q)

Typical pattern of ACF
Decays exponentially or damped
Sine wave pattern or both
Significant spikes through lags q
Exponential decay

Typical pattern of PACF
Significant spikes through lags p
Declines exponentially

Exponential decay

Source: Gujarati, D. N. (2002). Basic Econometrics. 4th edition

Step 4 (Diagnostic checking): In this step, model must be checked for adequacy
by considering the properties of the residuals whether the residuals from an
ARIMA model must has the normal distribution and should be random. An overall
check of model adequacy is provided by the Ljung-Box Q statistic. The test
statistics is to examine whether residual is a “white noise” or not. If it is a “white
noise” then the model is accepted. Otherwise, the procedure will be reset to the
beginning. Q-statistics and Normality test can be used in this step.
Step 5 (Forecasting and forecast evaluation): Forecasts for one period or several periods into
the future with the parameters for a tentative model have been selected. In

24


this research paper, the paper utilizes E-views 6 with “forecast” package
build and evaluate ARIMA model.
Meyler, Aidan and Kenny, Geoff and Quinn, Terry (1998) listed that some main
benefits of using the auto-regressive integrated moving average as follows: First,
they believe that it is useful if the time series have a big variables. Second, they
emphasize that this reduced some error that occurs with multiple repressors.
Third, the multiple models always require the huge and correlated variable in time
series. Forecasting by multiple models is highly dependent on the un-available
variables and source of predicting, therefore, its result face uncertainty.

However, Meyler, Aidan and Kenny, Geoff and Quinn, Terry (1998) synthesize
some limitations of auto-regressive integrated moving average (ARIMA); some
of identification approaches plays as subjective. Moreover, the talent and

experience of the forecasters can be depended by the reliability of the chosen
model (although other modeling approaches are applied by this criticism as
well). It is not embedded within any underlying theoretical model or structural
relationships. It is not clear to the economic significance of the chosen model.
Furthermore, unlike with structural models, it is not possible to run policy
simulations with Auto-regressive integrated moving average models.
Model for non-seasonal series are called autoregressive integrated moving average
model, denoted by ARIMA (p,d,q). Here p indicates the order of the autoregressive
part, d indicates the amount of differencing, and q indicates the order of the moving
average part. If the original series is stationary, d= 0 and the auto-regressive
integrated moving average models reduce to the ARIMA (p,d,q) model.
Summarily, based on theories that are showed in the literature review, methodology
chapter undertook to build the multiple regression model from selected variables in
table 2.1. It is specified as: Xratet = α1 + α2CPIjpt + α3CPIvnt + α4IIPjpt + α5IIPvnt
+ α6IRjpt + α7IRvnt + α8∆tradejpt + α9∆tradevnt + α10∆ricet + α11∆oilt + εt. Moreover, our

study builds the steps to test the ARIMA model by the Box-Jenkins methodology with
five steps: data collection and examination, determine stationary of time series,
25