this PDF file THE IMPACT OF POLITICAL EVENT ON EXCHANGE RATE VOLATILITY | Supriyanti | Jurnal Administrasi Bisnis 1 PB
THE IMPACT OF POLITICAL EVENT ON EXCHANGE RATE VOLATILITY
(Study at Bloomberg on January-December 2016)
Ratna Sari Supriyanti
Siti Ragil Handayani
Faculty of Administrative Science
Brawijaya University
Malang
Email: sari.supriyanti@gmail.com
ABSTRAK
Penelitian ini bertujuan untuk menganalisis dampak kejadian politik terhadap volatilitas nilai tukar.
Penelitian ini menggunakan pendekatan kuantitatif dan metodologi event study untuk menganalisis efek dari
British Exit terhadap nilai tukar Poundsterling (GBP) terhadap US Dollar dan US Dollar terhadap Rupiah.
ARCH-GARCH model digunakan dalam penelitian ini untuk mengeksplor volatilitas nilai tukar. Sampel
yang digunakan dalam penelitian ini adalah harga penutupan nilai tukar dan berjumlah 101 data. Hasil
penelitian ini menjelaskan bahwa: 1) Uji Wilcoxon pada Poundsterling (GBP) terhadap US Dollar
menujukkan adanya perbedaan sebelum dan sesudah kejadian British Exit; 2) Uji Wilcoxon pada US Dollar
terhadap Rupiah menunjukkan bahwa ada perbedaan yang terjadi sebelum dan sesudah kejadian British
Exit; 3) Analisis pada nilai tukar Poundsterling (GBP) terhadap US Dollar menggunakan ARCH-GARCH
model menunjukkan adanya volatilitas yang terjadi di sekitar waktu observasi, begitu juga dengan Analisis
pada nilai tukar US Dollar terhadap Rupiah menunjukkan adanya volatilitas yang terjadi di sekitar waktu
observasi; 4) Hasil dari evaluasi ARCH-GARCH model menggunakan ARCH-LM pada Poundsterling
(GBP) terhadap US Dollar dan US Dollar terhadap Rupiah menunjukkan bahwa ARCH-GARCH model
dapat mengatasi unsur heteroskedastisity pada sampel yang diteliti. Hasil penelitian menunjukkan bahwa
efek volatilitas dapat terjadi dikarenakan faktor non ekonomi yang mempengaruhi, terutama faktor politik.
Kata Kunci: Kejadian Politik, Nilai Tukar, Event Study, ARCH-GARCH Model
ABSTRACT
The purpose of this study is to analyze the effect of political event on exchange rate volatility. This study
uses quantitative approach and event study methodology to analyze the effect of British Exit on
PoundSterling (GBP) against US Dollar and US Dollar against Rupiah. ARCH-GARCH model is also used
in this research to explore the volatility of exchange rate. Sample in this study are 101 data from exchange
rate closing price. The results as follows: 1) Wilcoxon test in PoundSterling (GBP) against US Dollar shows
the difference before and after British Exit event; 2) Wilcoxon test on US Dollar against Rupiah shows that
there is a difference before and after British Exit event; 3) Analysis of PoundSterling (GBP) against US
Dollar using ARCH-GARCH model indicates volatility occurs around the observation time, so are Analysis
on US Dollar against Rupiah using ARCH-GARCH model shows volatility around the observation time; 4)
Results from ARCH-GARCH evaluation model using ARCH-LM in PoundSterling (GBP) against US
Dollar and US Dollar against Rupiah indicates that ARCH-GARCH the model could overcome the element
of heteroscedastisity in the data studied. The results show that effects of volatility that could be happened
due to non-economic factors especially political factors.
Keywords: Political Event, Exchange Rate, Event Study, ARCH-GARCH Model
Jurnal Administrasi Bisnis (JAB)|Vol. 55 No. 3 Februari 2018|
administrasibisnis.studentjournal.ub.ac.id
73
INTRODUCTION
The investor and also firms who need to
exchange their currency with other currencies
could do the activity in foreign exchange market.
Foreign exchange market is global decentralized
market for the trading of currencies against one
other currency and it is the largest financial
market in the world in terms of both geographical
dispersion and daily turnover, which is in excess
of $5 trillion (Bank for International Settlements,
2013). More than trillion dollars of currency is
traded each day in the foreign exchange market,
primarily through a network of the largest
international banks.
Exchange rate is also one of macroeconomic
indicators that sensitive on external turmoil
condition. In this case, exchange rate reflects the
economic power as a result of economic global
condition. Among the instruments that are crucial
in economic management and stability of basic
prices is the exchange rate (Were, Geda, Karingi
& Ndung’u, 2001). The more stable exchange rate
value, the more it shows that fundamental strength
of countrie’s economic. The movement of
exchange rate is depends on purchasing power of
exchange rate under the influences of demand and
supply of market. The supply and demand is
affected by several factors. The factors affecting
exchange rate and its volatility are divided into
two groups that are economic and non-economic
factors.
Based on Twarowska and Kąkol (2014),
economic factor consist of rate of economic
growth, inflation rate, interest rate in the country
and abroad, and etc. Non-economic factors consist
of political risk (e.g. risk of armed conflict),
natural disasters, policy approaches and
psychological factors. During the last decades, a
range of studies has emphasized the way in which
the economic and political news and event that
tend to be a political risk affecting the currency
market. There are many studies that examine the
way in which the elective systems, the elections,
the political alliance and the political incertitude
affect both the evolution and the volatility of
capital market and of exchange rate market. The
political events influence the assets price in
different ways. In some cases, these significantly
influence the financial markets, and on the other
hand, the markets sometimes react calmly to
political changes.
In this research is focusing on variable of
political risk and its relation with the exchange
rate movement. The political risk is an important
factor, especially in emergent economies and in
the developing ones. Howell and Chaddick (1994)
define the political risk as being “the possibility of
the political decisions and political events from a
country, including the social ones, to affect the
business environment, so that the investors recall
the investments or the profit leeway reduces”.
Several previous researches did a research in
exchange rate volatility caused by political news
and also events. According to Schwidht (2014) a
coup or riot would give depreciate effect on
Bolivar currencies because the event is resulting
on political instability would create a climate of
fear that would push domestic investors to unload
Bolívares and hold onto Dollars which are more
secure, and would lessen foreign demand for the
Bolívar.
In this research, British Exit event was used
as variable that affect exchange rate. British Exit
event is recent event that happened as a result of
United Kingdom society who wants to exit from
European Union. This event not only shocked at
the level of Europe, but also the worldwide. On
23th June 2016 British was exit from European
Union (EU).
The consequences of British Exit occurs in
many areas like Economic and also political and
social. The Economic consequences for United
Kingdom happened in several areas like Trade,
United Kingdom GDP growth, Inflation, and also
Exchange Rate. The volatility of exchange rate
also reported in Euro Exchange Rate News:
“British Exit has already made a clear impact
upon PoundSterling, with the GBP/EUR exchange
rate falling -6% in the first eight weeks of 2016,
while GBP/USD has dropped -4.5% over the same
period. There is still a long way to fall, according
to Bank of America Merrill Lynch (BaML), which
believes Pound Sterling is still overvalued by
around 3%. ‘Our valuation suggests that Sterling
is overvalued and, in an extreme scenario where
capital inflows temporarily slow in the run-up to
the referendum, it could face large declines.”
Some of impacts made global market is
releasing Poundsterling (GBP) and move their
funds to currencies that they thought is safer than
Poundsterling (GBP) like US Dollar, Yen and
gold. This is because the uncertainty condition
facing by United Kingdom. It is means that US
Dollar is one of stable currencies and become one
of most interested currencies that was used by
investor. Based on middle rate of Bank Indonesia,
close price of rupiah depreciate against US Dollar
to 13.296 and one day before in 13.265 per Dollar.
However several analysts said that the British Exit
was not significant factor affected on economic
Jurnal Administrasi Bisnis (JAB)|Vol. 55 No. 3 Februari 2018|
administrasibisnis.studentjournal.ub.ac.id
74
condition of Indonesia. Governor of Bank
Indonesia, Agus Maragainstojo said that it was
normal for the exchange rate volatility happened
and it was temporary. Although Rupiah is not
liquid as US Dollar, but the effect of US Dollar
appreciate against Poundsterling (GBP) give
impact on Rupiah because the investor tend to
choose US Dollar as safest asset besides gold and
yen, moreover in unstable economic and political
situation
Within this research, the relationship
between effects of political event towards
exchange rate volatility is measured. This research
using two research instruments that are
autoregressive conditional heteroskedasticity
(ARCH)/general
autoregressive
conditional
heteroskedasticity (GARCH) models to explore
the volatility effect of the British exit event and
the paired test to study the level impact of British
Exit. The observation period of this research is
fifty days before and after British Exit which is 5
workdays start from April 14th until September 1th
2016. Based on background above, the title of this
research is “The Effect of Political Event on
Exchange Rate Volatility (Study at Bloomberg
January-December 2016)”.
LITERATURE REVIEW
Exchange Rate
a. Foreign Exchange Rate
Exchange rate indicates the price of
currency when exchanged with other
currencies. Foreign exchange rate is to
measure the value a currency from the
perspective of other currencies (Madura, 2000:
86).
b. Volatility of Exchange Rate
Volatility is a statistic measurement to
measure the price movement at certain period.
The measurement shows decreasing and
increasing prices in short period but do not
shows the level of prices except the variance
level form one period to another period. High
volatility reflects the uncommon demand and
supply. In general, volatility measures the
fluctuation average from time series data.
However, the further research develops by
emphasizing the variance value of the data. So
it means that volatility variance as value of
fluctuation data (Sunaryo: 2007 as citied at
Hermayani et al.: 2014).
Time series data, especially in financial
sector data are face the high volatility. High
volatility is indicated by a phase where
fluctuations are relatively high and then
followed by low fluctuation and then move
into high fluctuation again. In other words,
this data has an average and variance that are
not constant (Widarjono, 2009).
c. Factors
Affecting
Exchange
Rate
Movement
Exchange rate deviates from the valuation
basis purchasing power of currencies under
the influences of demand and supply of
currency. The correlation of supply and
demand depends on several factors. Usually
the factors affecting exchange rate and its
volatility are divided into two groups that are
economic
and
non-economic
factors.
Economic factors are distinguished into long
term and short term. Based on Twarowska and
Kąkol (2014), factors that influence the
demand and supply of exchange rate and result
on exchange rate fluctuation are:
1. Economic Factors
a. Short-Term
1) rate of economic growth
2) inflation rate
3) interest rate in the country and
abroad
4) current account balance
5) capital account balance
6) currency speculation
b. Long-Term
1) level of economic development
of the country
2) competitiveness
of
the
economy
3) technical and technological
development
4) size of the foreign debt
5) budget deficit
6) relative domestic and foreign
prices
7) capital flows
2. Non- Economic Factors
a. political risk (e.g. risk of armed
conflict)
b. natural disasters
c. policy approaches
d. psychological factors
d. Exchange Rate Determinant
The spot exchange rate refers to the
current exchange rate. In foreign exchange
currency, there is currency pair means the
quotation of the relative value of a currency
unit against the unit of another currency in the
foreign exchange market. The quotation
Jurnal Administrasi Bisnis (JAB)|Vol. 55 No. 3 Februari 2018|
administrasibisnis.studentjournal.ub.ac.id
75
EUR/USD 1.2500 means that 1 Euro is
exchanged for 1.2500 US Dollar.
e. Political Event
Political event is event that occurs in a
certain palace during a particular interval of
time in term of politic decision that lead on
political event. Every country faces certain
political event which is gives the positive and
negative impact on country condition whether
in economic, politic and social area. Some of
political event can caused the effect which can
be interpreted as risk. Broadly, political risk
refers to the complications businesses and
governments may face as a result of what are
commonly referred to as political decisions or
any political change that alters the expected
outcome and value of a given economic action
by changing the probability of achieving
business objectives.
Conceptual Model and Hypotheses
a. Conceptual Model
Conceptual Model is a model that used
to describe how the theory has the logical
relationship and connect the identified
problems. Conceptual model is also explains
each variables in the research.
Political
event
Exchange
rate
Volatility
Figure 1 Conceptual Model
Source: Theoretical review, June 2017
b. Hypotheses
Based on background and theoretical
review mentioned in this research, the
researcher constructs four hypotheses related
with the problem. There are four hypotheses,
those are:
1) Hypotheses I
H0
: There is no difference on
Poundsterling (GBP) against US Dollar
exchange rate before and after British Exit
H1
: There is difference on
Poundsterling (GBP) against US Dollar
exchange rate before and after British Exit
2) Hypotheses II
H0
: There is no difference on US
Dollar against Rupiah exchange rate
before and after British Exit
H2
: There is difference on US Dollar
against Rupiah exchange rate before and
after British Exit
3) Hypotheses III
H0
: There is no volatility on
Poundsterling (GBP) against US Dollar
exchange rate around British Exit event
H3
:
There
is
volatility
on
Poundsterling (GBP) against US Dollar
exchange rate around British Exit event
4) Hypotheses IV
H0
: There is no volatility on US
Dollar against Rupiah exchange rate
around British Exit event
H4
: There is volatility on US Dollar
against Rupiah exchange rate around
British Exit event
RESEARCH METHOD
Type of Research
The type of research used in this research is
quantitative approach using event study. Event
study refers to measure the impact of economic
and political event (Baldas, Oran: 2014). Besides
the event study methodology used, researcher will
explore the volatility of exchange rate around the
observation period by using ARCH-GARCH
model. By using two instruments research,
researcher will find the effect of political event on
Poundsterling (GBP) against US Dollar, US
Dollar against Rupiah and also explore the
exchange rate volatility.
Location of Research
Research was done by documenting the data
sample
from
websites
those
are
http://www.bloomberg.com/ and another website
that gives the daily, weekly, monthly data about
rate, where these websites as the world trusted
currency authority. This website is chosen because
it provides the complete historical data and
diagram about exchange rate movement.
Research Population and Sample
Population
Research population in this research is the
closing price data of daily foreign exchange rate
movements from January to December 2016.
Sample
Samples of this research are daily exchange
rate of Poundsterling (GBP) against US Dollar
and US Dollar against Rupiah from 5 work days
(Monday to Friday). Purposive sampling was used to
determine the sample. The criteria that can be used in
the determination of the sampling are as follows:
1. US Dollar in the most liquid currencies in the
world (Bank of International settlement, 2013)
Jurnal Administrasi Bisnis (JAB)|Vol. 55 No. 3 Februari 2018|
administrasibisnis.studentjournal.ub.ac.id
76
and also number one in most used currencies
in international trade (Liputan6.com news,
2015).
2. United State is country which most aimed for
goods export of United Kingdom, and it is
means that the exchange of currency with
another currency is more often.
3. Rupiah is chosen because the appreciation of
US Dollar Value will give impacts on Rupiah
currencies, because US Dollar is one of safe
asset intended by investor and also trader
especially in unstable condition.
Data Analysis Technique
This research uses two instrument researches
in order to explore the change of Poundsterling
(GBP) exchange rate movement against US Dollar
and US Dollar against Rupiah. The first
instrument is differences test, where the
instrument used is paired sample test to describe
the exchange rate differences before and after
British exit. The second instrument is ARCHGARCH model analysis to explore the volatility
as instrument research in around the observation
time.
a. Event Study to Measures the Level of
Political Event Effect
1) Normality Test
The basic concept of KolmogorovSmirnov normality test is comparing the
distribution of the data (which will be
tested the normality level) with a standard
normal distribution. The criteria in
determine the data have normal
distribution or not are as follows:
a) If the significant number of
Kolmogorov Smirnov test Sig. < 0.05
so the data have not normal
distribution
b) If the significant number of
Kolmogorov Smirnov test Sig. > 0.05
so the data have normal distribution
2) Paired Test
The purpose of paired test is to
compare the average of two groups
whether having the same significant
average or not. This research uses t-test for
measure the average of closing price of US
Dollar against Poundsterling and US
Dollar against Rupiah as samples during
research period. If the data is normally
distributed, the difference test for the
research would be done by paired sample
T-test. However, if the data is not normally
distributed, the difference test for the
research would be done by wilcoxon
signed rank test.
b. ARCH-GARCH Model Analysis to
Explore the Exchange Rate Volatility
Researcher uses Eviews 8 program as
analysis tool. Steps in Analyzing ARCHGARCH model in this research as follows:
1) Plot of Time Series Data
2) Stationary Test
a. Stationary Test
b. Differentiation level test
3) Identify ARMA-ARIMA model by Using
Correlogram
a. Model
Identification
by
using
Correlogram
b. Selecting the Best of ARIMA Model
4) Detecting the Heteroskedasticity element
by using ARCH-LM Test
5) ARCH-GARCH Mode Estimation by
using Maximum Likelihood
6) Model Evaluation Test
RESULT AND DISCUSSION
A. Descriptive Statistical Analysis
1. Descriptive Statistical Analysis of Poundsterling
(GBP) against US Dollar
Table 1 Descriptive statistical analysis of GBP on
US Dollar
Mean
Median
Poundsterling (GBP) against USD
1.381913
1.41140
Maximum
1.48770
Minimum
1.229800
Std. Deviation
0.068962
N
101
Source: Data processed by researcher, August
2017
Table 1 Shown that the maximum value
is 1.487700 and it shown the highest exchange
rate happened in around the observation
period. The minimum value is about 1.229800
and it is means the lowest value of GBP on US
Dollar exchange rate in observation period.
According to average value on the table above
shows that average of Poundsterling (GBP)
against US Dollar exchange rate is about
1.381913 with the amount of total observation
sample are 101.
2. Descriptive Statistical Analysis of US Dollar
against IDR (Rupiah)
Jurnal Administrasi Bisnis (JAB)|Vol. 55 No. 3 Februari 2018|
administrasibisnis.studentjournal.ub.ac.id
77
Table 2 Descriptive statistical analysis of US
Dollar against IDR (Rupiah)
Mean
Median
Poundsterling (GBP) against USD
1.381913
1.41140
Maximum
1.48770
Minimum
1.229800
Std. Deviation
0.068962
N
101
Source: Data processed by researcher, August
2017
Table 2 Shown that the mean value of
the US Dollar against Rupiah exchange rate is
13261.17 with the amount of observation
sample are 101. This number means that
average to get 1 dollar, people have to change
with 13261.17 in fifty days before and after
British Exit event. The maximum value that
happened in time of observation is about
13705.65 with the minimum value of US
Dollar against Rupiah exchange rate is about
13056.90.
B. Event Study Method to Measures the Level
of Political Event Effect
1) The result of Poundsterling (GBP) against
US Dollar (USD) shown that the value of
symp.sig. (2-tailed) is 0.000 less than 0.05
and it is means that H0 accepted and H1
rejected. H0 accepted means that there is
difference on GBP against US Dollar
before and after British exit.
2) The result of US Dollar (USD) against
Rupiah (IDR) showed that asymp.sig. (2tailed) is 0.000 that is less than 0.05 and it
is means that H0 accepted and H1 rejected.
H0 accepted means that there is difference
on US Dollar against Rupiah before and
after British exit.
C. ARCH-GARCH Model Analysis to Explore
the Exchange Rate Volatility
1. Plot of Time Series Data
a. According to plot of time series of
Poundstreling (GBP) against US Dollar
(Figure 1), it can be conclude that the
movement
of
exchange
rate
of
Poundsterling (GBP) against US Dollar
has fluctuates plot and it has decrease
trend.
GBP/USD
1.50
1. Normality Test
Table 3 Result of Normality test of GBP
against USD and USD against IDR
1.45
1.40
1.35
1.30
Variable
P-Value
Meaning
GBP_USD
0.0000
Abnormal
1.25
1.20
18 25
M4
USD_IDR
0.0350
Abnormal
Source: Data processed by researcher, August
2017
The result show that the data of GBP
against US Dollar abnormal because the
significance value is less than α = 5% (0.000 <
0.05) and the data of US Dollar against Rupiah
also has abnormal distribution with the
significance value is less than α = 5% (0.035 <
0.05).
2. Wilcoxon Signed Rank Test
Table 4 Result of Wilcoxon Signed Rank Test of
GBP against USD and USD against IDR
Before_ after
Before_ after
2
9
16 23 30
M5
6
13 20 27
M6
4
11 18 25
M7
1
8
15 22 29
M8
Figure 2 Plot of time series GBP against US
Dollar data
Source: Data Processed by researcher, August
2017
b. The pattern of time series data on US
Dollar against Rupiah showed the
significant increasing movement is
happened in 4th May, 2016 and the lowest
point in 4th July, 2016. Generally, the plot
of US Dollar against Rupiah exchange rate
shows the volatility movement. The data is
fluctuates which the pattern shown the up
movement
and
down
movement
continually.
GBP against USD
Z calc
Asymp. Sig. (2tailed)
-5.251
0.000
USD against IDR
-5.237
0.000
Source: Data Processed by researcher, August
2017
Jurnal Administrasi Bisnis (JAB)|Vol. 55 No. 3 Februari 2018|
administrasibisnis.studentjournal.ub.ac.id
78
USD/IDR
13,800
13,700
13,600
13,500
13,400
13,300
13,200
13,100
13,000
18 25
M4
2
9
16 23 30
M5
6
13 20 27
M6
4
11 18 25
M7
1
8
15 22 29
M8
Figure 3 Plot of time series US Dollar
against Rupiah
Source: Data Processed by Researcher, August
2017
2. Stasionary Test
a. According to unit root test level test of
Poundstreling (GBP) against US Dollar ,
the value of augmented Dickey-Fuller test
statistic is -1.537607 is less than test
critical values which the values are 3.497029, -2.890623, -2.582353, so it is
mean that the data is not stationer. Because
the data is not stationer, so the data will be
tested with unit root test in first difference
level. Based on unit root test in first
difference level, the value of augmented
Dickey-Fuller test statistic is -12.10668
which more than test critical values which
are in level 1% is -3.497727, level 5% is
-2.890926 and in the level of 10% is
-2.582514. It is means the data is stationer.
b. Based on unit root test level test of US
Dollar against Rupiah, the value of
augmented Dickey-Fuller test statistic is 2.973526 is less than test critical values
which the values are -3.497029,-2.890623,
-2.582353. It is mean that the data is not
stationer. Because the data is not stationer
yet, so the differences unit root test has to
be done. Based on unit root test in first
difference level, The value of augmented
Dickey-Fuller test statistic is -10.01595
which more than test critical values which
are in level 1% is -3.498439, level 5% is
-2.891234 and in the level of 10% is
-2.582678. So it is mean that the data is
stationer in first difference level.
3. Identify ARMA-ARIMA model by
Using Correlogram
A. Model Identification by using Correlogram
1) Poundsterling (GBP) against US Dollar
The result of the correlogram test of
Poundstreling (GBP) against US Dollar shows
that the maximum value of AR order (p) is
obtained at lag 1, lag 12 where this lag is cut
off the line and the lag which is used by
researcher is 24. MA order (q) is also obtained
at lag 1 and lag 12 where this lag is cut off the
line.
The first model is ARIMA (1,1,1) and
then ARIMA (1,1,0), ARIMA (0,1,1). The
second mode is ARIMA (12,1,12), ARIMA
(12,1,1), ARIMA (12,1,0). There are 6
tentative models obtained in step of ARMAARIMA analysis. This model will be
regressed by using least square (NS and
ARMA) and will be compared. The tentative
model is measure by using ARMA least
square estimation. The results of regression
are describes below:
Table 5 Result of ARIMA tentative
estimation of GBP against USD
Variable Coefficient PR2
Value
4.33%
ARIMA - 0.000936 0.030
or
(0,1,1)
0.0433
4.19%
ARIMA
-0.205111
0.042
or
(0,1,1)
0.0419
0.914519
0.000 19.61%
ARIMA
or
(1,1,1)
-1.122064
0.000
0.1961
-0.339306
0.004 13.55%
ARIMA
or
(12,1,1)
0.213107
0.000
0.1355
F-stat
0.0375
0.0420
0.0000
0.0020
24.77% 0.0000
or
-0.842085
0.000
0.2477
9.73%
ARIMA
-0.353455
0.003
or
0.0030
(12,1,0)
0.097
Source: Data Processed by Researcher, August
2017
ARIMA
12,1,12
0.189311
model
0.143
2) US Dollar against Rupiah (IDR)
The result of the correlogram test shows
that the maximum value of AR order (p) is
obtained at lag 1 and lag 2 where this lag is cut
off the line. MA order (q) is also obtained at
lag 1, where this lag is cut off the line. The
first model is ARIMA (1,1,1) and then
ARIMA (1,1,0), ARIMA (0,1,1). The second
model is ARIMA (2,1,0), ARIMA (2,1,1).
There are 5 tentative models obtained in
correlogram analysis. This model will be
regressed and will be compared. The results of
regression are describes below:
Jurnal Administrasi Bisnis (JAB)|Vol. 55 No. 3 Februari 2018|
administrasibisnis.studentjournal.ub.ac.id
79
Table 6 Result of ARIMA tentative model by using
least square equation USD against IDR
Variable Coefficient PR2
F-stat
value
8.29%
ARIMA
-0.287225
0.003
0.00385
or
(1,1,0)
0.0829
0.109573
0.109 11.68%
ARIMA
or
0.00257
(1,1,1)
-0.474080 -0.474
0.1168
11.69%
ARIMA
-0.385606
0.000
or
0.00049
(0,1,1)
0.1169
-0.111513
0.296 12.40%
ARIMA
or
0.00185
(2,1,1)
-0.338738
0.001
0.1240
1.99%
ARIMA
-0.140600
0.165
0.16526
or
(2,1,0)
0.0199
Source: Data Processed by Researcher, August 2017
B. Selecting the Best of ARIMA Model
Before determine the ARCH-GARCH
model, the several tentative models is
prepared. The steps of choosing the best
model are same with GBP against USD. The
selected model is determined by using some of
criteria. The First criteria is the biggest R2
value, and then the smallest AIC and SC.
1) Poundsterling (GBP) against US Dollar
Table 7 Result of ARIMA tentative model by
using least square equation GBP against USD
Model
R2
AIC
SC
ARIMA
0.043374
-4.80029
-4.74819
(0,1,1)
ARIMA
0.041937
-4.78918
-4.73676
(1,1,0)
ARIMA
0.196175
-4.94452
-4.86588
(1,1,1)
ARIMA
0.135597
-4.76041
-4.67596
(12,1,1)
ARIMA
0.247728
-4.89935
-4.81490
(12,1,12)
ARIMA
0.097326
-4.73982
-4.68352
(12,1,0)
Source: Data Processed by Researcher, August
2017
Based on table above, the biggest R
squared are showed at ARIMA (1,1,1) and
ARIMA (12,1,12). But, at ARIMA (12,1,12)
there p-value which is not significance with
the p-value of variable AR (12) is 0.143 and it
is more than 0.05. The smallest AIC and SC
is showed at ARIMA (1,1,1). But, at ARIMA
(12,1,12) there p-value which is not
significance with the p-value of AR (12) is
0.143 and it is more than 0.05. So the chosen
model is ARIMA (1,1,1).
2)
US Dollar against Rupiah (IDR)
Table 7 Result of ARIMA tentative model by
using least square equation USD against IDR
Model
R2
AIC
SC
ARIMA
0.116986
11.84183
11.89394
(0,1,1)
ARIMA
0.082913
11.88501
11.93743
(1,1,0)
ARIMA
0.116803
11.86755
11.94619
(1,1,1)
ARIMA
0.124052
11.86546
11.94459
(2,1,1)
ARIMA
0.019959
11.95734
12.01009
(2,1,0)
Source: Data Processed by Researcher, August
2017
Based on table above, the biggest R
squared are showed at ARIMA (0,1,1) and
ARIMA (2,1,1). But, at ARIMA (2,1,1) there
is p-value which not significance with the pvalue of variable AR (2) is 0.2964 and it is
more than 0.05. The smallest AIC and SC is
showed at ARIMA (0,1,1) although the
highest R2 is in ARIMA (2,1,1), but, at
ARIMA (12,1,12) there is p-value which is not
significance with the p-value of AR (12) is
0.2964 and it is more than 0.05. So the chosen
model is ARIMA (0,1,1).
4. Detecting
the
Heteroskedasticity
element by using ARCH-LM Test
1) Poundsterling (GBP) against US
Dollar
Based on ARCH-Lagrange Multiplier
test shows that obs*R2 calculation value is
23.2346, with the probability value is
0.0258. The probability value of chi square
is less than 0.05. It is means that the value
is significance and the data contains of
Heteroskedasticity element. The existence
of Heteroskedasticity element means that
residua variance is not constant and model
used contains of ARCH element. So the
model should be estimated by using
ARCH-GARCH model to overcome the
Heteroskedasticity problem.
2) US Dollar against Rupiah (IDR)
Based on ARCH-Lagrange Multiplier
test shows that obs*R2 calculation value is
11.09382, with the probability value is
0.0009. The probability value of chi square
is less than 0.05. It is means that the value
Jurnal Administrasi Bisnis (JAB)|Vol. 55 No. 3 Februari 2018|
administrasibisnis.studentjournal.ub.ac.id
80
is significance and the data contains of
Heteroskedasticity element. It is means
that residual variance is not constant and
model used contains of ARCH element.
5. ARCH-GARCH Mode Estimation by
using Maximum Likelihood
1) Poundsterling (GBP) against US
Dollar
The ARIMA (1,1,1) model that is
used contains of Heteroskedasticity
element. So Maximum Likelihood model
estimation was needed. The output which
is showed at Maximum Likelihood model
estimation consists of two parts that are
conditional
mean
calculation
and
conditional variance calculation
.
Table 8 Result of ARCH (1) and GARCH(1)
model estimation.
Conditional mean
Probability
Variable
Coefficient
Value
AR (1)
0.0588330
0.0000
MA (1)
-0.913013
0.0000
Conditional Variance
ARCH (1)
1.186651
0.0004
GARCH (1)
0.342789
0.0050
Source: Data Processed by Researcher, August
2017
The result of Maximum Likelihood
regression estimation shows that the
probability value at ARCH (1) and
GARCH (1) are significance. They are
significance whether in conditional mean
and also conditional variance. The
probability value of conditional variance is
significance and it is means that there is
volatility at exchange rate value in
observation period within 50 days before
British Exit event and 50 days after British
Exit event.
The heteroskedasticity element that
was detected in exchange rate value of
Poundsterling (GBP) against Dollar means
that the residual variance of the data is not
constant and fluctuates from one period to
another period. At conditional mean
equation, the probability value for AR (1)
is 0.0000 and MA (1) is 0.0000 and it less
than 0.05, so the value is significance. The
significance of AR (1) and MA(1) means
that the exchange rate value of
Poundsterling (GBP) against US Dollar is
affected by exchange rate value the period
before at lag 1.
2) US Dollar against Rupiah (IDR)
Table 9 Result of ARCH (1) and GARCH
(1) model estimation
Conditional mean
Probability
Variable
Coefficient
Value
MA (1)
-0.326290
0.0219
Conditional Variance
ARCH (1)
1.338500
0.0000
GARCH (1)
0.260482
0.0057
Source: Data Processed by Researcher, August
2017
The result of Maximum Likelihood
regression shows that the probability value
at ARCH (1) and GARCH (1) are
significance. They are significance
whether in conditional mean and also
conditional variance. The probability value
of conditional variance is significance and
it is means that there is volatility at
exchange rate value in observation period
within 50 days before British Exit event
and 50 days after British Exit event. The
residual variance of exchange rate value
(ARCH) is affected by residual at the
period before and lag of residual variance
at the period before (GARCH).
6. Model Evaluation Test
The diagnostic model test should be
done to analyze whether the data formed is
good enough to modeling the data. To
analyze if the data still contains of
heteroskedasticity element or not, ARCHLM test is used. The result of GBP against
US Dollar shown that the value of X2
(Obs*R-squared) is 4.524610 with the
probability value is 0.9720 or 97.2% and it
is more than 0.05. The result of ARCHLM test of US Dollar against IDR revealed
the value of X2 (Obs*R-squared) is
0.291058 with the probability value is
0.5895 or 58.95% and it is more than 0.05.
So it is means that both of sample have not
contain of ARCH element and the model
could overcome Heteroskedasticity .
Jurnal Administrasi Bisnis (JAB)|Vol. 55 No. 3 Februari 2018|
administrasibisnis.studentjournal.ub.ac.id
81
CONCLUSION AND SUGGESTION
Conclusion
1. The result of Wilcoxon signed rank test on
daily closing price of Poundsterling against
US Dollar shows that there is difference at
observation period which are 50 days before
and 50 days after British Exit.
2. The result of daily closing price of US Dollar
against Rupiah 50 days before and 50 days
after British Exit event shows that there is
difference on the exchange rate.
3. The result of ARCH-GARCH model shows
that there is heteroskedasticity element on
Poundsterling against US Dollar and also US
Dollar against Rupiah. It is means that residual
variance is not constant and change over
especially in observation period. Applying the
GARCH model at closing price of
Poundsterling (GBP) against US Dollar shows
that the data have time varying volatility
problem.
4. Model evaluation test for poundsterling (GBP)
against US Dollar and also US Dollar against
Rupiah show that the ARCH-GARCH could
overcome the heteroskedasticity problem. The
residual variance of the data is constant or the
data has not contains of ARCH element.
Suggestion
1. For Further researcher
a. Researcher should find out and improve
other information that can affect exchange
rate market. The short period can be used
in the future research in order to minimize
the bias result so it will be better to use
long term period in future research.
b. For Rupiah, the perceived effect is not felt
directly, like has been mentioned in
background in chapter 1.
2. For Company
Company should anticipate about the event
that can be viral in media. It is worth
mentioning that in the long run the exchange
rate evolution will always reflect the level of
competitiveness, the level of productivity from
countries and the investor sentiment against a
country. And it will affect the productivity of
the company.
REFERENCES
Adiwijaya, Setiawan. 2016. Dampak Brexit, BI:
Ekonomi Indonesia Baik, Bisnis Tempo
Article, accessed at May 17, 2017 from
https://bisnis.tempo.co/read/782799/dampa
k-brexit-bi-kondisi-ekonomi-indonesiabaik.
Baldas Ulkem, Oran Adil. 2014. Event Studies in
Turkey. Accessed at February 3, From
http://www.sciencedirect.com/science/artic
le/pii/S22144501 4000271
.
Bank of International Settlements. 2013. Triennial
Central Bank Survey. Foreign Exchange
Turnover in April 2013: preliminary
global results Monetary and Economic
Department, accessed at February 2, 2017
http://www.bis.org/publ/rpfx13fx.pdf.
Fruen, L. 2016. What is Brexit? Everything you
need to know about the UK leaving the
EU. The sun Article, November 22, 2016
from https://www.thesun.co.uk./.
Hermayani
et
al.
2014.
Mengatasi
Heteroskedastisitas pada Model ARIMA
dengan Menggunakan ARCH-GARCH
(Studi Kasus: Indeks Harga Konsumen
Provinsi Kalimantan Timur Tahun 20052012), Jurnal Eksponensial Volume 5,
Nomor 1, accessed at May 22, 2017 from
https://fmipa.unmul.ac.id/files/docs/[9]%2
0jURNAL%20hERMAYANI%20ok.doc.
Howell, L. & Chaddick, B. 1994. Models of
Political Risk for Foreign Investment and
Trade: An Assessment of Three.
Madura, Jeff. 1995. Manajemen Keuangan
Internasional. Alih Bahasa oleh Emil
Salim. 2000. Edisi Keempat. Jilid 1.
Jakarta: Erlangga.
Madura, jeff. 2008. International finance
management.
Ninth edition. South
western: Thomson.
Twarowska, Katarzyna & Kakol, Magdalena.
2014. Analysis of Factors Affecting
Fluctuation the Exchange Rate of Polish
Zloty
against
Euro.
International
Conference: Slovenia, accessed from,
http://www.toknowpress.net/ISBN/978961-6914-09-3/papers/ML14-652.pdf.
Were, M., Geda, A., Karingi, S. & Ndung’u, N.
2001. Kenya’s Exchange Rate Movement
in a liberalized environmental: An
empirical Analysis. Kenya Institute for
Public Policy Research and Analysis
(KIPPRA), Nairobi.
Jurnal Administrasi Bisnis (JAB)|Vol. 55 No. 3 Februari 2018|
administrasibisnis.studentjournal.ub.ac.id
82
(Study at Bloomberg on January-December 2016)
Ratna Sari Supriyanti
Siti Ragil Handayani
Faculty of Administrative Science
Brawijaya University
Malang
Email: sari.supriyanti@gmail.com
ABSTRAK
Penelitian ini bertujuan untuk menganalisis dampak kejadian politik terhadap volatilitas nilai tukar.
Penelitian ini menggunakan pendekatan kuantitatif dan metodologi event study untuk menganalisis efek dari
British Exit terhadap nilai tukar Poundsterling (GBP) terhadap US Dollar dan US Dollar terhadap Rupiah.
ARCH-GARCH model digunakan dalam penelitian ini untuk mengeksplor volatilitas nilai tukar. Sampel
yang digunakan dalam penelitian ini adalah harga penutupan nilai tukar dan berjumlah 101 data. Hasil
penelitian ini menjelaskan bahwa: 1) Uji Wilcoxon pada Poundsterling (GBP) terhadap US Dollar
menujukkan adanya perbedaan sebelum dan sesudah kejadian British Exit; 2) Uji Wilcoxon pada US Dollar
terhadap Rupiah menunjukkan bahwa ada perbedaan yang terjadi sebelum dan sesudah kejadian British
Exit; 3) Analisis pada nilai tukar Poundsterling (GBP) terhadap US Dollar menggunakan ARCH-GARCH
model menunjukkan adanya volatilitas yang terjadi di sekitar waktu observasi, begitu juga dengan Analisis
pada nilai tukar US Dollar terhadap Rupiah menunjukkan adanya volatilitas yang terjadi di sekitar waktu
observasi; 4) Hasil dari evaluasi ARCH-GARCH model menggunakan ARCH-LM pada Poundsterling
(GBP) terhadap US Dollar dan US Dollar terhadap Rupiah menunjukkan bahwa ARCH-GARCH model
dapat mengatasi unsur heteroskedastisity pada sampel yang diteliti. Hasil penelitian menunjukkan bahwa
efek volatilitas dapat terjadi dikarenakan faktor non ekonomi yang mempengaruhi, terutama faktor politik.
Kata Kunci: Kejadian Politik, Nilai Tukar, Event Study, ARCH-GARCH Model
ABSTRACT
The purpose of this study is to analyze the effect of political event on exchange rate volatility. This study
uses quantitative approach and event study methodology to analyze the effect of British Exit on
PoundSterling (GBP) against US Dollar and US Dollar against Rupiah. ARCH-GARCH model is also used
in this research to explore the volatility of exchange rate. Sample in this study are 101 data from exchange
rate closing price. The results as follows: 1) Wilcoxon test in PoundSterling (GBP) against US Dollar shows
the difference before and after British Exit event; 2) Wilcoxon test on US Dollar against Rupiah shows that
there is a difference before and after British Exit event; 3) Analysis of PoundSterling (GBP) against US
Dollar using ARCH-GARCH model indicates volatility occurs around the observation time, so are Analysis
on US Dollar against Rupiah using ARCH-GARCH model shows volatility around the observation time; 4)
Results from ARCH-GARCH evaluation model using ARCH-LM in PoundSterling (GBP) against US
Dollar and US Dollar against Rupiah indicates that ARCH-GARCH the model could overcome the element
of heteroscedastisity in the data studied. The results show that effects of volatility that could be happened
due to non-economic factors especially political factors.
Keywords: Political Event, Exchange Rate, Event Study, ARCH-GARCH Model
Jurnal Administrasi Bisnis (JAB)|Vol. 55 No. 3 Februari 2018|
administrasibisnis.studentjournal.ub.ac.id
73
INTRODUCTION
The investor and also firms who need to
exchange their currency with other currencies
could do the activity in foreign exchange market.
Foreign exchange market is global decentralized
market for the trading of currencies against one
other currency and it is the largest financial
market in the world in terms of both geographical
dispersion and daily turnover, which is in excess
of $5 trillion (Bank for International Settlements,
2013). More than trillion dollars of currency is
traded each day in the foreign exchange market,
primarily through a network of the largest
international banks.
Exchange rate is also one of macroeconomic
indicators that sensitive on external turmoil
condition. In this case, exchange rate reflects the
economic power as a result of economic global
condition. Among the instruments that are crucial
in economic management and stability of basic
prices is the exchange rate (Were, Geda, Karingi
& Ndung’u, 2001). The more stable exchange rate
value, the more it shows that fundamental strength
of countrie’s economic. The movement of
exchange rate is depends on purchasing power of
exchange rate under the influences of demand and
supply of market. The supply and demand is
affected by several factors. The factors affecting
exchange rate and its volatility are divided into
two groups that are economic and non-economic
factors.
Based on Twarowska and Kąkol (2014),
economic factor consist of rate of economic
growth, inflation rate, interest rate in the country
and abroad, and etc. Non-economic factors consist
of political risk (e.g. risk of armed conflict),
natural disasters, policy approaches and
psychological factors. During the last decades, a
range of studies has emphasized the way in which
the economic and political news and event that
tend to be a political risk affecting the currency
market. There are many studies that examine the
way in which the elective systems, the elections,
the political alliance and the political incertitude
affect both the evolution and the volatility of
capital market and of exchange rate market. The
political events influence the assets price in
different ways. In some cases, these significantly
influence the financial markets, and on the other
hand, the markets sometimes react calmly to
political changes.
In this research is focusing on variable of
political risk and its relation with the exchange
rate movement. The political risk is an important
factor, especially in emergent economies and in
the developing ones. Howell and Chaddick (1994)
define the political risk as being “the possibility of
the political decisions and political events from a
country, including the social ones, to affect the
business environment, so that the investors recall
the investments or the profit leeway reduces”.
Several previous researches did a research in
exchange rate volatility caused by political news
and also events. According to Schwidht (2014) a
coup or riot would give depreciate effect on
Bolivar currencies because the event is resulting
on political instability would create a climate of
fear that would push domestic investors to unload
Bolívares and hold onto Dollars which are more
secure, and would lessen foreign demand for the
Bolívar.
In this research, British Exit event was used
as variable that affect exchange rate. British Exit
event is recent event that happened as a result of
United Kingdom society who wants to exit from
European Union. This event not only shocked at
the level of Europe, but also the worldwide. On
23th June 2016 British was exit from European
Union (EU).
The consequences of British Exit occurs in
many areas like Economic and also political and
social. The Economic consequences for United
Kingdom happened in several areas like Trade,
United Kingdom GDP growth, Inflation, and also
Exchange Rate. The volatility of exchange rate
also reported in Euro Exchange Rate News:
“British Exit has already made a clear impact
upon PoundSterling, with the GBP/EUR exchange
rate falling -6% in the first eight weeks of 2016,
while GBP/USD has dropped -4.5% over the same
period. There is still a long way to fall, according
to Bank of America Merrill Lynch (BaML), which
believes Pound Sterling is still overvalued by
around 3%. ‘Our valuation suggests that Sterling
is overvalued and, in an extreme scenario where
capital inflows temporarily slow in the run-up to
the referendum, it could face large declines.”
Some of impacts made global market is
releasing Poundsterling (GBP) and move their
funds to currencies that they thought is safer than
Poundsterling (GBP) like US Dollar, Yen and
gold. This is because the uncertainty condition
facing by United Kingdom. It is means that US
Dollar is one of stable currencies and become one
of most interested currencies that was used by
investor. Based on middle rate of Bank Indonesia,
close price of rupiah depreciate against US Dollar
to 13.296 and one day before in 13.265 per Dollar.
However several analysts said that the British Exit
was not significant factor affected on economic
Jurnal Administrasi Bisnis (JAB)|Vol. 55 No. 3 Februari 2018|
administrasibisnis.studentjournal.ub.ac.id
74
condition of Indonesia. Governor of Bank
Indonesia, Agus Maragainstojo said that it was
normal for the exchange rate volatility happened
and it was temporary. Although Rupiah is not
liquid as US Dollar, but the effect of US Dollar
appreciate against Poundsterling (GBP) give
impact on Rupiah because the investor tend to
choose US Dollar as safest asset besides gold and
yen, moreover in unstable economic and political
situation
Within this research, the relationship
between effects of political event towards
exchange rate volatility is measured. This research
using two research instruments that are
autoregressive conditional heteroskedasticity
(ARCH)/general
autoregressive
conditional
heteroskedasticity (GARCH) models to explore
the volatility effect of the British exit event and
the paired test to study the level impact of British
Exit. The observation period of this research is
fifty days before and after British Exit which is 5
workdays start from April 14th until September 1th
2016. Based on background above, the title of this
research is “The Effect of Political Event on
Exchange Rate Volatility (Study at Bloomberg
January-December 2016)”.
LITERATURE REVIEW
Exchange Rate
a. Foreign Exchange Rate
Exchange rate indicates the price of
currency when exchanged with other
currencies. Foreign exchange rate is to
measure the value a currency from the
perspective of other currencies (Madura, 2000:
86).
b. Volatility of Exchange Rate
Volatility is a statistic measurement to
measure the price movement at certain period.
The measurement shows decreasing and
increasing prices in short period but do not
shows the level of prices except the variance
level form one period to another period. High
volatility reflects the uncommon demand and
supply. In general, volatility measures the
fluctuation average from time series data.
However, the further research develops by
emphasizing the variance value of the data. So
it means that volatility variance as value of
fluctuation data (Sunaryo: 2007 as citied at
Hermayani et al.: 2014).
Time series data, especially in financial
sector data are face the high volatility. High
volatility is indicated by a phase where
fluctuations are relatively high and then
followed by low fluctuation and then move
into high fluctuation again. In other words,
this data has an average and variance that are
not constant (Widarjono, 2009).
c. Factors
Affecting
Exchange
Rate
Movement
Exchange rate deviates from the valuation
basis purchasing power of currencies under
the influences of demand and supply of
currency. The correlation of supply and
demand depends on several factors. Usually
the factors affecting exchange rate and its
volatility are divided into two groups that are
economic
and
non-economic
factors.
Economic factors are distinguished into long
term and short term. Based on Twarowska and
Kąkol (2014), factors that influence the
demand and supply of exchange rate and result
on exchange rate fluctuation are:
1. Economic Factors
a. Short-Term
1) rate of economic growth
2) inflation rate
3) interest rate in the country and
abroad
4) current account balance
5) capital account balance
6) currency speculation
b. Long-Term
1) level of economic development
of the country
2) competitiveness
of
the
economy
3) technical and technological
development
4) size of the foreign debt
5) budget deficit
6) relative domestic and foreign
prices
7) capital flows
2. Non- Economic Factors
a. political risk (e.g. risk of armed
conflict)
b. natural disasters
c. policy approaches
d. psychological factors
d. Exchange Rate Determinant
The spot exchange rate refers to the
current exchange rate. In foreign exchange
currency, there is currency pair means the
quotation of the relative value of a currency
unit against the unit of another currency in the
foreign exchange market. The quotation
Jurnal Administrasi Bisnis (JAB)|Vol. 55 No. 3 Februari 2018|
administrasibisnis.studentjournal.ub.ac.id
75
EUR/USD 1.2500 means that 1 Euro is
exchanged for 1.2500 US Dollar.
e. Political Event
Political event is event that occurs in a
certain palace during a particular interval of
time in term of politic decision that lead on
political event. Every country faces certain
political event which is gives the positive and
negative impact on country condition whether
in economic, politic and social area. Some of
political event can caused the effect which can
be interpreted as risk. Broadly, political risk
refers to the complications businesses and
governments may face as a result of what are
commonly referred to as political decisions or
any political change that alters the expected
outcome and value of a given economic action
by changing the probability of achieving
business objectives.
Conceptual Model and Hypotheses
a. Conceptual Model
Conceptual Model is a model that used
to describe how the theory has the logical
relationship and connect the identified
problems. Conceptual model is also explains
each variables in the research.
Political
event
Exchange
rate
Volatility
Figure 1 Conceptual Model
Source: Theoretical review, June 2017
b. Hypotheses
Based on background and theoretical
review mentioned in this research, the
researcher constructs four hypotheses related
with the problem. There are four hypotheses,
those are:
1) Hypotheses I
H0
: There is no difference on
Poundsterling (GBP) against US Dollar
exchange rate before and after British Exit
H1
: There is difference on
Poundsterling (GBP) against US Dollar
exchange rate before and after British Exit
2) Hypotheses II
H0
: There is no difference on US
Dollar against Rupiah exchange rate
before and after British Exit
H2
: There is difference on US Dollar
against Rupiah exchange rate before and
after British Exit
3) Hypotheses III
H0
: There is no volatility on
Poundsterling (GBP) against US Dollar
exchange rate around British Exit event
H3
:
There
is
volatility
on
Poundsterling (GBP) against US Dollar
exchange rate around British Exit event
4) Hypotheses IV
H0
: There is no volatility on US
Dollar against Rupiah exchange rate
around British Exit event
H4
: There is volatility on US Dollar
against Rupiah exchange rate around
British Exit event
RESEARCH METHOD
Type of Research
The type of research used in this research is
quantitative approach using event study. Event
study refers to measure the impact of economic
and political event (Baldas, Oran: 2014). Besides
the event study methodology used, researcher will
explore the volatility of exchange rate around the
observation period by using ARCH-GARCH
model. By using two instruments research,
researcher will find the effect of political event on
Poundsterling (GBP) against US Dollar, US
Dollar against Rupiah and also explore the
exchange rate volatility.
Location of Research
Research was done by documenting the data
sample
from
websites
those
are
http://www.bloomberg.com/ and another website
that gives the daily, weekly, monthly data about
rate, where these websites as the world trusted
currency authority. This website is chosen because
it provides the complete historical data and
diagram about exchange rate movement.
Research Population and Sample
Population
Research population in this research is the
closing price data of daily foreign exchange rate
movements from January to December 2016.
Sample
Samples of this research are daily exchange
rate of Poundsterling (GBP) against US Dollar
and US Dollar against Rupiah from 5 work days
(Monday to Friday). Purposive sampling was used to
determine the sample. The criteria that can be used in
the determination of the sampling are as follows:
1. US Dollar in the most liquid currencies in the
world (Bank of International settlement, 2013)
Jurnal Administrasi Bisnis (JAB)|Vol. 55 No. 3 Februari 2018|
administrasibisnis.studentjournal.ub.ac.id
76
and also number one in most used currencies
in international trade (Liputan6.com news,
2015).
2. United State is country which most aimed for
goods export of United Kingdom, and it is
means that the exchange of currency with
another currency is more often.
3. Rupiah is chosen because the appreciation of
US Dollar Value will give impacts on Rupiah
currencies, because US Dollar is one of safe
asset intended by investor and also trader
especially in unstable condition.
Data Analysis Technique
This research uses two instrument researches
in order to explore the change of Poundsterling
(GBP) exchange rate movement against US Dollar
and US Dollar against Rupiah. The first
instrument is differences test, where the
instrument used is paired sample test to describe
the exchange rate differences before and after
British exit. The second instrument is ARCHGARCH model analysis to explore the volatility
as instrument research in around the observation
time.
a. Event Study to Measures the Level of
Political Event Effect
1) Normality Test
The basic concept of KolmogorovSmirnov normality test is comparing the
distribution of the data (which will be
tested the normality level) with a standard
normal distribution. The criteria in
determine the data have normal
distribution or not are as follows:
a) If the significant number of
Kolmogorov Smirnov test Sig. < 0.05
so the data have not normal
distribution
b) If the significant number of
Kolmogorov Smirnov test Sig. > 0.05
so the data have normal distribution
2) Paired Test
The purpose of paired test is to
compare the average of two groups
whether having the same significant
average or not. This research uses t-test for
measure the average of closing price of US
Dollar against Poundsterling and US
Dollar against Rupiah as samples during
research period. If the data is normally
distributed, the difference test for the
research would be done by paired sample
T-test. However, if the data is not normally
distributed, the difference test for the
research would be done by wilcoxon
signed rank test.
b. ARCH-GARCH Model Analysis to
Explore the Exchange Rate Volatility
Researcher uses Eviews 8 program as
analysis tool. Steps in Analyzing ARCHGARCH model in this research as follows:
1) Plot of Time Series Data
2) Stationary Test
a. Stationary Test
b. Differentiation level test
3) Identify ARMA-ARIMA model by Using
Correlogram
a. Model
Identification
by
using
Correlogram
b. Selecting the Best of ARIMA Model
4) Detecting the Heteroskedasticity element
by using ARCH-LM Test
5) ARCH-GARCH Mode Estimation by
using Maximum Likelihood
6) Model Evaluation Test
RESULT AND DISCUSSION
A. Descriptive Statistical Analysis
1. Descriptive Statistical Analysis of Poundsterling
(GBP) against US Dollar
Table 1 Descriptive statistical analysis of GBP on
US Dollar
Mean
Median
Poundsterling (GBP) against USD
1.381913
1.41140
Maximum
1.48770
Minimum
1.229800
Std. Deviation
0.068962
N
101
Source: Data processed by researcher, August
2017
Table 1 Shown that the maximum value
is 1.487700 and it shown the highest exchange
rate happened in around the observation
period. The minimum value is about 1.229800
and it is means the lowest value of GBP on US
Dollar exchange rate in observation period.
According to average value on the table above
shows that average of Poundsterling (GBP)
against US Dollar exchange rate is about
1.381913 with the amount of total observation
sample are 101.
2. Descriptive Statistical Analysis of US Dollar
against IDR (Rupiah)
Jurnal Administrasi Bisnis (JAB)|Vol. 55 No. 3 Februari 2018|
administrasibisnis.studentjournal.ub.ac.id
77
Table 2 Descriptive statistical analysis of US
Dollar against IDR (Rupiah)
Mean
Median
Poundsterling (GBP) against USD
1.381913
1.41140
Maximum
1.48770
Minimum
1.229800
Std. Deviation
0.068962
N
101
Source: Data processed by researcher, August
2017
Table 2 Shown that the mean value of
the US Dollar against Rupiah exchange rate is
13261.17 with the amount of observation
sample are 101. This number means that
average to get 1 dollar, people have to change
with 13261.17 in fifty days before and after
British Exit event. The maximum value that
happened in time of observation is about
13705.65 with the minimum value of US
Dollar against Rupiah exchange rate is about
13056.90.
B. Event Study Method to Measures the Level
of Political Event Effect
1) The result of Poundsterling (GBP) against
US Dollar (USD) shown that the value of
symp.sig. (2-tailed) is 0.000 less than 0.05
and it is means that H0 accepted and H1
rejected. H0 accepted means that there is
difference on GBP against US Dollar
before and after British exit.
2) The result of US Dollar (USD) against
Rupiah (IDR) showed that asymp.sig. (2tailed) is 0.000 that is less than 0.05 and it
is means that H0 accepted and H1 rejected.
H0 accepted means that there is difference
on US Dollar against Rupiah before and
after British exit.
C. ARCH-GARCH Model Analysis to Explore
the Exchange Rate Volatility
1. Plot of Time Series Data
a. According to plot of time series of
Poundstreling (GBP) against US Dollar
(Figure 1), it can be conclude that the
movement
of
exchange
rate
of
Poundsterling (GBP) against US Dollar
has fluctuates plot and it has decrease
trend.
GBP/USD
1.50
1. Normality Test
Table 3 Result of Normality test of GBP
against USD and USD against IDR
1.45
1.40
1.35
1.30
Variable
P-Value
Meaning
GBP_USD
0.0000
Abnormal
1.25
1.20
18 25
M4
USD_IDR
0.0350
Abnormal
Source: Data processed by researcher, August
2017
The result show that the data of GBP
against US Dollar abnormal because the
significance value is less than α = 5% (0.000 <
0.05) and the data of US Dollar against Rupiah
also has abnormal distribution with the
significance value is less than α = 5% (0.035 <
0.05).
2. Wilcoxon Signed Rank Test
Table 4 Result of Wilcoxon Signed Rank Test of
GBP against USD and USD against IDR
Before_ after
Before_ after
2
9
16 23 30
M5
6
13 20 27
M6
4
11 18 25
M7
1
8
15 22 29
M8
Figure 2 Plot of time series GBP against US
Dollar data
Source: Data Processed by researcher, August
2017
b. The pattern of time series data on US
Dollar against Rupiah showed the
significant increasing movement is
happened in 4th May, 2016 and the lowest
point in 4th July, 2016. Generally, the plot
of US Dollar against Rupiah exchange rate
shows the volatility movement. The data is
fluctuates which the pattern shown the up
movement
and
down
movement
continually.
GBP against USD
Z calc
Asymp. Sig. (2tailed)
-5.251
0.000
USD against IDR
-5.237
0.000
Source: Data Processed by researcher, August
2017
Jurnal Administrasi Bisnis (JAB)|Vol. 55 No. 3 Februari 2018|
administrasibisnis.studentjournal.ub.ac.id
78
USD/IDR
13,800
13,700
13,600
13,500
13,400
13,300
13,200
13,100
13,000
18 25
M4
2
9
16 23 30
M5
6
13 20 27
M6
4
11 18 25
M7
1
8
15 22 29
M8
Figure 3 Plot of time series US Dollar
against Rupiah
Source: Data Processed by Researcher, August
2017
2. Stasionary Test
a. According to unit root test level test of
Poundstreling (GBP) against US Dollar ,
the value of augmented Dickey-Fuller test
statistic is -1.537607 is less than test
critical values which the values are 3.497029, -2.890623, -2.582353, so it is
mean that the data is not stationer. Because
the data is not stationer, so the data will be
tested with unit root test in first difference
level. Based on unit root test in first
difference level, the value of augmented
Dickey-Fuller test statistic is -12.10668
which more than test critical values which
are in level 1% is -3.497727, level 5% is
-2.890926 and in the level of 10% is
-2.582514. It is means the data is stationer.
b. Based on unit root test level test of US
Dollar against Rupiah, the value of
augmented Dickey-Fuller test statistic is 2.973526 is less than test critical values
which the values are -3.497029,-2.890623,
-2.582353. It is mean that the data is not
stationer. Because the data is not stationer
yet, so the differences unit root test has to
be done. Based on unit root test in first
difference level, The value of augmented
Dickey-Fuller test statistic is -10.01595
which more than test critical values which
are in level 1% is -3.498439, level 5% is
-2.891234 and in the level of 10% is
-2.582678. So it is mean that the data is
stationer in first difference level.
3. Identify ARMA-ARIMA model by
Using Correlogram
A. Model Identification by using Correlogram
1) Poundsterling (GBP) against US Dollar
The result of the correlogram test of
Poundstreling (GBP) against US Dollar shows
that the maximum value of AR order (p) is
obtained at lag 1, lag 12 where this lag is cut
off the line and the lag which is used by
researcher is 24. MA order (q) is also obtained
at lag 1 and lag 12 where this lag is cut off the
line.
The first model is ARIMA (1,1,1) and
then ARIMA (1,1,0), ARIMA (0,1,1). The
second mode is ARIMA (12,1,12), ARIMA
(12,1,1), ARIMA (12,1,0). There are 6
tentative models obtained in step of ARMAARIMA analysis. This model will be
regressed by using least square (NS and
ARMA) and will be compared. The tentative
model is measure by using ARMA least
square estimation. The results of regression
are describes below:
Table 5 Result of ARIMA tentative
estimation of GBP against USD
Variable Coefficient PR2
Value
4.33%
ARIMA - 0.000936 0.030
or
(0,1,1)
0.0433
4.19%
ARIMA
-0.205111
0.042
or
(0,1,1)
0.0419
0.914519
0.000 19.61%
ARIMA
or
(1,1,1)
-1.122064
0.000
0.1961
-0.339306
0.004 13.55%
ARIMA
or
(12,1,1)
0.213107
0.000
0.1355
F-stat
0.0375
0.0420
0.0000
0.0020
24.77% 0.0000
or
-0.842085
0.000
0.2477
9.73%
ARIMA
-0.353455
0.003
or
0.0030
(12,1,0)
0.097
Source: Data Processed by Researcher, August
2017
ARIMA
12,1,12
0.189311
model
0.143
2) US Dollar against Rupiah (IDR)
The result of the correlogram test shows
that the maximum value of AR order (p) is
obtained at lag 1 and lag 2 where this lag is cut
off the line. MA order (q) is also obtained at
lag 1, where this lag is cut off the line. The
first model is ARIMA (1,1,1) and then
ARIMA (1,1,0), ARIMA (0,1,1). The second
model is ARIMA (2,1,0), ARIMA (2,1,1).
There are 5 tentative models obtained in
correlogram analysis. This model will be
regressed and will be compared. The results of
regression are describes below:
Jurnal Administrasi Bisnis (JAB)|Vol. 55 No. 3 Februari 2018|
administrasibisnis.studentjournal.ub.ac.id
79
Table 6 Result of ARIMA tentative model by using
least square equation USD against IDR
Variable Coefficient PR2
F-stat
value
8.29%
ARIMA
-0.287225
0.003
0.00385
or
(1,1,0)
0.0829
0.109573
0.109 11.68%
ARIMA
or
0.00257
(1,1,1)
-0.474080 -0.474
0.1168
11.69%
ARIMA
-0.385606
0.000
or
0.00049
(0,1,1)
0.1169
-0.111513
0.296 12.40%
ARIMA
or
0.00185
(2,1,1)
-0.338738
0.001
0.1240
1.99%
ARIMA
-0.140600
0.165
0.16526
or
(2,1,0)
0.0199
Source: Data Processed by Researcher, August 2017
B. Selecting the Best of ARIMA Model
Before determine the ARCH-GARCH
model, the several tentative models is
prepared. The steps of choosing the best
model are same with GBP against USD. The
selected model is determined by using some of
criteria. The First criteria is the biggest R2
value, and then the smallest AIC and SC.
1) Poundsterling (GBP) against US Dollar
Table 7 Result of ARIMA tentative model by
using least square equation GBP against USD
Model
R2
AIC
SC
ARIMA
0.043374
-4.80029
-4.74819
(0,1,1)
ARIMA
0.041937
-4.78918
-4.73676
(1,1,0)
ARIMA
0.196175
-4.94452
-4.86588
(1,1,1)
ARIMA
0.135597
-4.76041
-4.67596
(12,1,1)
ARIMA
0.247728
-4.89935
-4.81490
(12,1,12)
ARIMA
0.097326
-4.73982
-4.68352
(12,1,0)
Source: Data Processed by Researcher, August
2017
Based on table above, the biggest R
squared are showed at ARIMA (1,1,1) and
ARIMA (12,1,12). But, at ARIMA (12,1,12)
there p-value which is not significance with
the p-value of variable AR (12) is 0.143 and it
is more than 0.05. The smallest AIC and SC
is showed at ARIMA (1,1,1). But, at ARIMA
(12,1,12) there p-value which is not
significance with the p-value of AR (12) is
0.143 and it is more than 0.05. So the chosen
model is ARIMA (1,1,1).
2)
US Dollar against Rupiah (IDR)
Table 7 Result of ARIMA tentative model by
using least square equation USD against IDR
Model
R2
AIC
SC
ARIMA
0.116986
11.84183
11.89394
(0,1,1)
ARIMA
0.082913
11.88501
11.93743
(1,1,0)
ARIMA
0.116803
11.86755
11.94619
(1,1,1)
ARIMA
0.124052
11.86546
11.94459
(2,1,1)
ARIMA
0.019959
11.95734
12.01009
(2,1,0)
Source: Data Processed by Researcher, August
2017
Based on table above, the biggest R
squared are showed at ARIMA (0,1,1) and
ARIMA (2,1,1). But, at ARIMA (2,1,1) there
is p-value which not significance with the pvalue of variable AR (2) is 0.2964 and it is
more than 0.05. The smallest AIC and SC is
showed at ARIMA (0,1,1) although the
highest R2 is in ARIMA (2,1,1), but, at
ARIMA (12,1,12) there is p-value which is not
significance with the p-value of AR (12) is
0.2964 and it is more than 0.05. So the chosen
model is ARIMA (0,1,1).
4. Detecting
the
Heteroskedasticity
element by using ARCH-LM Test
1) Poundsterling (GBP) against US
Dollar
Based on ARCH-Lagrange Multiplier
test shows that obs*R2 calculation value is
23.2346, with the probability value is
0.0258. The probability value of chi square
is less than 0.05. It is means that the value
is significance and the data contains of
Heteroskedasticity element. The existence
of Heteroskedasticity element means that
residua variance is not constant and model
used contains of ARCH element. So the
model should be estimated by using
ARCH-GARCH model to overcome the
Heteroskedasticity problem.
2) US Dollar against Rupiah (IDR)
Based on ARCH-Lagrange Multiplier
test shows that obs*R2 calculation value is
11.09382, with the probability value is
0.0009. The probability value of chi square
is less than 0.05. It is means that the value
Jurnal Administrasi Bisnis (JAB)|Vol. 55 No. 3 Februari 2018|
administrasibisnis.studentjournal.ub.ac.id
80
is significance and the data contains of
Heteroskedasticity element. It is means
that residual variance is not constant and
model used contains of ARCH element.
5. ARCH-GARCH Mode Estimation by
using Maximum Likelihood
1) Poundsterling (GBP) against US
Dollar
The ARIMA (1,1,1) model that is
used contains of Heteroskedasticity
element. So Maximum Likelihood model
estimation was needed. The output which
is showed at Maximum Likelihood model
estimation consists of two parts that are
conditional
mean
calculation
and
conditional variance calculation
.
Table 8 Result of ARCH (1) and GARCH(1)
model estimation.
Conditional mean
Probability
Variable
Coefficient
Value
AR (1)
0.0588330
0.0000
MA (1)
-0.913013
0.0000
Conditional Variance
ARCH (1)
1.186651
0.0004
GARCH (1)
0.342789
0.0050
Source: Data Processed by Researcher, August
2017
The result of Maximum Likelihood
regression estimation shows that the
probability value at ARCH (1) and
GARCH (1) are significance. They are
significance whether in conditional mean
and also conditional variance. The
probability value of conditional variance is
significance and it is means that there is
volatility at exchange rate value in
observation period within 50 days before
British Exit event and 50 days after British
Exit event.
The heteroskedasticity element that
was detected in exchange rate value of
Poundsterling (GBP) against Dollar means
that the residual variance of the data is not
constant and fluctuates from one period to
another period. At conditional mean
equation, the probability value for AR (1)
is 0.0000 and MA (1) is 0.0000 and it less
than 0.05, so the value is significance. The
significance of AR (1) and MA(1) means
that the exchange rate value of
Poundsterling (GBP) against US Dollar is
affected by exchange rate value the period
before at lag 1.
2) US Dollar against Rupiah (IDR)
Table 9 Result of ARCH (1) and GARCH
(1) model estimation
Conditional mean
Probability
Variable
Coefficient
Value
MA (1)
-0.326290
0.0219
Conditional Variance
ARCH (1)
1.338500
0.0000
GARCH (1)
0.260482
0.0057
Source: Data Processed by Researcher, August
2017
The result of Maximum Likelihood
regression shows that the probability value
at ARCH (1) and GARCH (1) are
significance. They are significance
whether in conditional mean and also
conditional variance. The probability value
of conditional variance is significance and
it is means that there is volatility at
exchange rate value in observation period
within 50 days before British Exit event
and 50 days after British Exit event. The
residual variance of exchange rate value
(ARCH) is affected by residual at the
period before and lag of residual variance
at the period before (GARCH).
6. Model Evaluation Test
The diagnostic model test should be
done to analyze whether the data formed is
good enough to modeling the data. To
analyze if the data still contains of
heteroskedasticity element or not, ARCHLM test is used. The result of GBP against
US Dollar shown that the value of X2
(Obs*R-squared) is 4.524610 with the
probability value is 0.9720 or 97.2% and it
is more than 0.05. The result of ARCHLM test of US Dollar against IDR revealed
the value of X2 (Obs*R-squared) is
0.291058 with the probability value is
0.5895 or 58.95% and it is more than 0.05.
So it is means that both of sample have not
contain of ARCH element and the model
could overcome Heteroskedasticity .
Jurnal Administrasi Bisnis (JAB)|Vol. 55 No. 3 Februari 2018|
administrasibisnis.studentjournal.ub.ac.id
81
CONCLUSION AND SUGGESTION
Conclusion
1. The result of Wilcoxon signed rank test on
daily closing price of Poundsterling against
US Dollar shows that there is difference at
observation period which are 50 days before
and 50 days after British Exit.
2. The result of daily closing price of US Dollar
against Rupiah 50 days before and 50 days
after British Exit event shows that there is
difference on the exchange rate.
3. The result of ARCH-GARCH model shows
that there is heteroskedasticity element on
Poundsterling against US Dollar and also US
Dollar against Rupiah. It is means that residual
variance is not constant and change over
especially in observation period. Applying the
GARCH model at closing price of
Poundsterling (GBP) against US Dollar shows
that the data have time varying volatility
problem.
4. Model evaluation test for poundsterling (GBP)
against US Dollar and also US Dollar against
Rupiah show that the ARCH-GARCH could
overcome the heteroskedasticity problem. The
residual variance of the data is constant or the
data has not contains of ARCH element.
Suggestion
1. For Further researcher
a. Researcher should find out and improve
other information that can affect exchange
rate market. The short period can be used
in the future research in order to minimize
the bias result so it will be better to use
long term period in future research.
b. For Rupiah, the perceived effect is not felt
directly, like has been mentioned in
background in chapter 1.
2. For Company
Company should anticipate about the event
that can be viral in media. It is worth
mentioning that in the long run the exchange
rate evolution will always reflect the level of
competitiveness, the level of productivity from
countries and the investor sentiment against a
country. And it will affect the productivity of
the company.
REFERENCES
Adiwijaya, Setiawan. 2016. Dampak Brexit, BI:
Ekonomi Indonesia Baik, Bisnis Tempo
Article, accessed at May 17, 2017 from
https://bisnis.tempo.co/read/782799/dampa
k-brexit-bi-kondisi-ekonomi-indonesiabaik.
Baldas Ulkem, Oran Adil. 2014. Event Studies in
Turkey. Accessed at February 3, From
http://www.sciencedirect.com/science/artic
le/pii/S22144501 4000271
.
Bank of International Settlements. 2013. Triennial
Central Bank Survey. Foreign Exchange
Turnover in April 2013: preliminary
global results Monetary and Economic
Department, accessed at February 2, 2017
http://www.bis.org/publ/rpfx13fx.pdf.
Fruen, L. 2016. What is Brexit? Everything you
need to know about the UK leaving the
EU. The sun Article, November 22, 2016
from https://www.thesun.co.uk./.
Hermayani
et
al.
2014.
Mengatasi
Heteroskedastisitas pada Model ARIMA
dengan Menggunakan ARCH-GARCH
(Studi Kasus: Indeks Harga Konsumen
Provinsi Kalimantan Timur Tahun 20052012), Jurnal Eksponensial Volume 5,
Nomor 1, accessed at May 22, 2017 from
https://fmipa.unmul.ac.id/files/docs/[9]%2
0jURNAL%20hERMAYANI%20ok.doc.
Howell, L. & Chaddick, B. 1994. Models of
Political Risk for Foreign Investment and
Trade: An Assessment of Three.
Madura, Jeff. 1995. Manajemen Keuangan
Internasional. Alih Bahasa oleh Emil
Salim. 2000. Edisi Keempat. Jilid 1.
Jakarta: Erlangga.
Madura, jeff. 2008. International finance
management.
Ninth edition. South
western: Thomson.
Twarowska, Katarzyna & Kakol, Magdalena.
2014. Analysis of Factors Affecting
Fluctuation the Exchange Rate of Polish
Zloty
against
Euro.
International
Conference: Slovenia, accessed from,
http://www.toknowpress.net/ISBN/978961-6914-09-3/papers/ML14-652.pdf.
Were, M., Geda, A., Karingi, S. & Ndung’u, N.
2001. Kenya’s Exchange Rate Movement
in a liberalized environmental: An
empirical Analysis. Kenya Institute for
Public Policy Research and Analysis
(KIPPRA), Nairobi.
Jurnal Administrasi Bisnis (JAB)|Vol. 55 No. 3 Februari 2018|
administrasibisnis.studentjournal.ub.ac.id
82