Description
The purpose of the present study is to directly examine the relationship between bilateral
exchange rate and stock market index in a bivariate framework during the period of the floating
exchange rate regime in Thailand
Journal of Financial Economic Policy
Linkages between Thai stock and foreign exchange markets under the floating regime
Komain J iranyakul
Article information:
To cite this document:
Komain J iranyakul, (2012),"Linkages between Thai stock and foreign exchange markets under the floating
regime", J ournal of Financial Economic Policy, Vol. 4 Iss 4 pp. 305 - 319
Permanent link to this document:http://dx.doi.org/10.1108/17576381211279280
Downloaded on: 24 January 2016, At: 21:45 (PT)
References: this document contains references to 38 other documents.
To copy this document: [email protected]
The fulltext of this document has been downloaded 478 times since 2012*
Users who downloaded this article also downloaded:
Abdullah M. Noman, Sarkar Humayun Kabir, Omar K.M.R. Bashar, (2012),"Causality between stock and
foreign exchange markets in Bangladesh", Studies in Economics and Finance, Vol. 29 Iss 3 pp. 174-186http://dx.doi.org/10.1108/10867371211246849
Manish Kumar, (2013),"Returns and volatility spillover between stock prices and exchange rates: Empirical
evidence from IBSA countries", International J ournal of Emerging Markets, Vol. 8 Iss 2 pp. 108-128 http://
dx.doi.org/10.1108/17468801311306984
Tian Yong Fu, Mark J . Holmes, Daniel F.S. Choi, (2011),"Volatility transmission and asymmetric linkages
between the stock and foreign exchange markets: A sectoral analysis", Studies in Economics and Finance,
Vol. 28 Iss 1 pp. 36-50http://dx.doi.org/10.1108/10867371111110543
Access to this document was granted through an Emerald subscription provided by emerald-srm:115632 []
For Authors
If you would like to write for this, or any other Emerald publication, then please use our Emerald for
Authors service information about how to choose which publication to write for and submission guidelines
are available for all. Please visit www.emeraldinsight.com/authors for more information.
About Emerald www.emeraldinsight.com
Emerald is a global publisher linking research and practice to the benefit of society. The company
manages a portfolio of more than 290 journals and over 2,350 books and book series volumes, as well as
providing an extensive range of online products and additional customer resources and services.
Emerald is both COUNTER 4 and TRANSFER compliant. The organization is a partner of the Committee
on Publication Ethics (COPE) and also works with Portico and the LOCKSS initiative for digital archive
preservation.
*Related content and download information correct at time of download.
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
5
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Linkages between Thai stock
and foreign exchange markets
under the ?oating regime
Komain Jiranyakul
School of Development Economics,
National Institute of Development Administration, Bangkok, Thailand
Abstract
Purpose – The purpose of the present study is to directly examine the relationship between bilateral
exchange rate and stock market index in a bivariate framework during the period of the ?oating
exchange rate regime in Thailand.
Design/methodology/approach – The monthly data used in this study are the stock market index
or stock prices from the Stock Exchange of Thailand, and the nominal bilateral exchange rate in terms
of baht per US dollar from the Bank of Thailand. The period covers July 1997 to June 2010 with 156
observations. This is the period that the country switched from ?xed to ?oating exchange rate regime.
The stock market return is calculated by the percentage change of stock market index (or stock prices)
while the exchange rate return is the percentage change of the nominal bilateral exchange rate. Three
estimation methods are used to capture the interaction between stock and foreign exchange markets:
bounds testing for cointegration, non-causality test, and the two-step approach with a bivariate
GARCH model and Granger causality test.
Findings – The results of the present study show that bounds testing for cointegration does not
detect the long-run relationship between stock prices and exchange rate. In addition, the non-causality
test fails the diagnostic test for multivariate normality in the residuals of the estimated VAR model.
However, the two-step approach adequately detects the linkages between the stock and foreign
exchange markets. It is found that there exists positive unidirectional causality running from stock
market return to exchange rate return. The exchange rate risk causes stock return to fall as expected.
Moreover, there are bidirectional causal relations between stock market risk and exchange rate risk,
but in different directions.
Research limitations/implications – Since a rising trend in the risk in the foreign exchange
market causes stock return to fall, both domestic and foreign investors should be aware of the risk or
uncertainty in the foreign exchange market because it can cause their portfolio return to fall. For
policymakers, reducing exchange rate risk cannot be done without the associated costs from a rising
risk in the stock market.
Originality/value – This study provides an evidence of volatility (or risk) spillovers in stock and
foreign exchange markets. In addition, the risk in foreign exchange market that adversely affects
return in the stock market is an expected phenomenon under the ?oating exchange rate regime.
Keywords Stock prices, Exchange rates, Bivariate GARCH, Causality, Volatility spillover, Thailand
Paper type Research paper
1. Introduction
Theoretically, the traditional approach or the ?ow-oriented theory states that a
depreciation of domestic currency can have a crucial impact on stock prices (SP) by
increasing ?rms’ competitiveness, while in turn raising their pro?tability. When ?rms
The current issue and full text archive of this journal is available at
www.emeraldinsight.com/1757-6385.htm
The author is grateful to two anonymous referees for their comments and suggestions that are
very helpful in improving the quality of this paper.
Thai stock and
foreign exchange
markets
305
Journal of Financial Economic Policy
Vol. 4 No. 4, 2012
pp. 305-319
qEmerald Group Publishing Limited
1757-6385
DOI 10.1108/17576381211279280
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
5
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
are able to pay more dividends to stockholders, SP will increase. Thus, there should be
a positive relationship between exchange rates and SP. In this case, exchange rates lead
SP. On the contrary, the portfolio balance approachindicates that SPlead exchange rates
on the ground that a rising trend in SP induces foreign investors to invest more in
domestic stocks. This will cause more capital in?ows, which in turn cause domestic
currency appreciation. In addition, a rise in domestic SP causes wealth to increase, and
thus induces investors to increase their demand for money, which results in a rise in
domestic interest rates. Higher interest rates induce capital in?ows, and thus cause an
appreciation in domestic currency. According to this approach, SP lead exchange rates
with a negative relationship. Another approach called “monetary approach” or “asset
market approach” indicates no linkage between SP and exchange rates due to different
factors in?uencing the two variables. The details of the ?ow-oriented model are in
Dornbusch and Fisher (1980), the portfolio balance approach in Branson and Henderson
(1985). A comprehensive review of the monetary approach is in MacDonald and
Taylor (1992).
Empirical studies suggest that there is no long-run relationship between SP and
exchange rates in most countries, speci?cally fromthe results of bivariate cointegration
tests. Furthermore, the direction of causality seems to depend on speci?c characteristics
of each country being analyzed. Many studies employ multivariate cointegration tests to
investigate the long-run relationship between SP and exchange rates, along with other
macroeconomic variables. For examples, Kown and Shin (1999) ?nd no long-run
relationship between SP and exchange rate in a bivariate framework, but the long-run
relationship between SP and other variables, including exchange rate, is detected. Their
?nding shows that SP are not a leading indicator of the exchange rate and other
macroeconomic variables in Korea. Kim (2003) also employs the multivariate
cointegration test to detect the long-run relationship between SP and the dollar
exchange rate in the USA and ?nds negative relationship between SP and the
exchange rate.
Gradual abolishment of barriers to capital ?ows and foreign exchange restrictions,
along with the adoption of more ?exible exchange rates in emerging markets are the
main causes of the volatility of exchange rates and risk in the portfolio diversi?cation
process in the past decades. Therefore, understanding the relationship between
exchange rates and SP will provide guidance for investors to diversify their portfolios.
The linkages between stock markets and foreign exchange markets are explored by
many researchers who employ modern time series econometrics. The mutual relations
between the two markets are examined by using the techniques of cointegration and
causality between SP and exchange rates. The study by Frank and Young (1972) is
known as the inaugurate attempt to link SPto exchange rates. However, the results show
no relationship between SPand exchange rates measured in terms of the prices of the US
dollar. Aggarwal (1981) uses the US monthly data to investigate the relationship
betweenthe two variables and?nds that the relationshipis stronger inthe short runthan
in the long run during the 1974-1978, which is the ?rst stage of the post-Brettonwood
systemthat more volatile exchange rates are observed. Soenen and Hennigar (1988) use
monthly data of the US dollar effective exchange rate and stock market index during the
1980-1986 period to investigate this relationship and ?nd a strong negative relationship.
Ma and Kao (1990) examine SP reactions to exchange rate changes under the ?oating
regime and ?nd that stock returns are minimally or not affected by exchange
JFEP
4,4
306
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
5
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
rate ?uctuations. Bahmani-Oskooee and Sohabian (1992) provide an evidence of
bidirectional causality of effective exchange rate and SP (the S&P 500 index) in the short
run, but not in the long run. Ajayi and Mougoue (1996) indicate the causation running
fromSP to exchange rates in eight industrialized economies. Nieh and Lee (2001) ?nd no
long-run relationship between SP and exchange rates in G-7 countries. Stavarek (2005)
uses monthly data to examine the linkages between SP and effective exchange rates in
EU-member countries and the USA. The results show much stronger causations in
countries with developed capital and exchange rates markets than the less-developed
ones. Causations seemto be unidirectional running fromSPto exchange rates. For Asian
economies, Abballa and Murinde (1997) ?nd the causation running fromexchange rates
to SP in India and the Philippines, but no causations in Pakistan and Korea. Yu (1997)
uses daily data on Hong Kong, Japan, and Singapore stock markets during the 1983-1994
period to examine the causality between exchange rates and SP. The results are mixed,
i.e. bidirectional causality in Japan, unidirectional causation running from exchange
rates to SP in Hong Kong, and no causation in Singapore. Mansor (2000) ?nds short-run
causation running from SP to exchange rates in Malaysia using an analysis in a
bivariate framework. Granger et al. (2000) examine the SP-exchange rates nexus in East
Asian countries using Asian ?u daily data from January 3, 1986 to June 16, 1998. The
so-called “Asian ?u” is the ?nancial turmoil that continued to exert its devastating force
from the third quarter of 1997 to the ?rst quarter of 1998. This was the period that SP
and exchange rates plunged intandeminAsianeconomies. Granger et al. (2000) ?nd that
exchange rates lead SP in South Korea with a positive correlation while SP lead
exchange rates in the Philippines with a negative correlation. The strong feedback
relations (or bidirectional causations) are observed in Hong Kong, Malaysia, Singapore,
Thailand, and Taiwan. Phylaktis and Ravazzolo (2005) apply a multivariate framework
to test for the direction of Granger causality in a group of Paci?c Basin countries, namely
Hong Kong, Malaysia, Singapore, Thailand, and the Philippines using monthly data
from 1980 to 1988. Their results show that stock and foreign exchange markets are
positively related and that the US stock market acts as a conduit for these linkages.
Pan et al. (2007) examine the dynamic linkage between exchange rates and SP for seven
East Asian economies during January 1988 to October 1998, and ?nd that the linkage
varies across economies with respect to exchange rate regimes, trade size, the degree of
capital control, and the size of equity market. Nabiha and Mounira (2009) examine the
dynamic relation between stock and exchange crisis in Mexico, Malaysia, Thailand,
Brazil, and Argentina. Using daily data of stock indexes and the US dollar exchange rate
during 1994 and 2003, they ?nd unidirectional causation running from SP to exchange
rate in Mexico, Argentina, and Brazil, but the causation running from exchange rate to
SP in Malaysia and Thailand.
The interactions between the two ?nancial markets are also investigated using a
generalized autoregressive conditional heteroskedastic (GARCH) model. Giovannini and
Jorion (1987) ?nd that stock return movements impact the foreign exchange markets by
changing stock return forecast that can cause changes in economic activities. Booth and
Rotenberg(1990) indicate that exchange rate movements canimpose the adverse effect on
the level of competitiveness of ?rms that are exposed to exchange rate risk. Choi and
Prasad (1995) ?nd that exchange rate movements affect the value of the US multinational
?rms. Atindehou and Gueyie (2001) estimate the three factor pricing model and ?nd that
Canadian bank stock returns are sensitive to exchange rate risk, speci?callythe risk from
Thai stock and
foreign exchange
markets
307
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
5
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
the US dollar rather than that fromthe Canadian dollar. Kanas (2000) ?nds that spillover
from stock markets to foreign exchange markets is signi?cant for ?ve out of six
developed countries, but spillover fromforeign exchange markets to stock markets is not
signi?cant for all countries. Chen et al. (2004) ?nd a lagged stock market reaction to
exchange rate ?uctuation in New Zealand. Furthermore, they ?nd that the 1997 Asian
?nancial crisis plays an important part in the volatility transmission between the two
?nancial markets by changing the nature of spillover from bidirectional spillover before
the crisis to unidirectional one after the crisis. Choi and Fang (2009) con?rmthe results of
Chen et al. (2004). Narayan (2009) investigates the impact of the Indian rupee depreciation
on the Indian stock market returns using daily data and ?nds that a depreciation of the
Indian rupee raises stock return volatility. However, an appreciation causes higher stock
return and lower return volatility.
The main objective of the present study is to directly examine the relationship between
bilateral exchange rate and stock market index in a bivariate framework using monthly
data fromJuly 1997 to June 2010, which is the period of the ?oating exchange rate regime
in Thailand. The afore-mentioned theories are tested. The more recent techniques are
used: the bounds testing for cointegration proposed by Pesaran et al. (2001) and
the non-causality tests proposed byToda andYamamoto (1995). Both procedures provide
simplistic methods to test the nexus between SP and exchange rates. The results of the
present study show that bounds testing for cointegration does not detect the long-run
relationship between SPand exchange rate inThailand. Inaddition, the non-causalitytest
fails the diagnostic test for multivariate normality in the residuals of the estimated vector
autoregressive (VAR) model. To adequately examine the linkages inthe stock and foreign
exchange markets, the results from the two-step approach is also employed. This
approach involves the estimate of a bivariate CCC-GARCH model of Bollerslev (1990)
and the Granger (1969) causality test. The results reveal that there exists positive
unidirectional causality running from stock market return to exchange rate return.
Furthermore, the exchange rate risk causes stock return to fall in the ?oating exchange
rate regime. There are bidirectional causal relations between stock market risk and
exchange rate risk, but in different directions. The paper is organized as follows: the next
section describes the analytical framework of this study. Section 3 presents empirical
results. The last section gives concluding remarks.
2. Methodology
The relationship between SP and exchange rates can be examined by cointegration and
causality tests. If the variables in the model are integrated of order one, I(1) and
cointegrated, the causation is obtained in the vector error correction model, which is
a VAR model in ?rst differences of the series with the inclusion of cointegrating
residuals, i.e. a restricted VAR model. An alternative test in the absence of
cointegration is the standard Granger causality test (Granger, 1969). In case
the variables are not I(1) series, for example they are mixed between I(0) and I(1), the
bounds testing for cointegration can be applied. If cointegration does not exist, the
causal relationship can be tested using the non-causality test in level of the series.
2.1 Cointegration test
The long-run relationship between SP and nominal bilateral exchange rate is
speci?ed as:
JFEP
4,4
308
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
5
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
LSP
t
¼ a þ bLEX
t
þ 1
t
ð1Þ
where LSP is the log of stock market index (or SP) in the Stock Exchange of Thailand,
and LEX is the log of nominal exchange rate.
Pesaran et al. (2001) proposed a new method for testing cointegration called a
conditional autoregressive distributed lag (ARDL) and error correction mechanism.
This is known as “ARDL bounds testing procedure.”
The ARDL model for equation (1) is speci?ed as:
DLSP
t
¼ a
0
þ
X
p
i¼1
b
i
DðLSPÞ
t2i
þ
X
q
j¼0
g
i
DLEX
t2i
þ e
t
ð2Þ
where p and q are the optimal number of lagged differences of log of SP and log of
exchange rate, respectively. By adding the lagged level variables into equation (2), the
computed F-statistic is obtained as shown in equation (3):
DLSP
t
¼ a
0
þa
1
LSP
t21
þa
2
LEX
t21
þ
X
p
i¼1
b
i
DLSP
t2i
þ
X
q
j¼0
g
i
DLEX
t2i
þ e
t
ð3Þ
If cointegration exists, replacing the lagged level variables with the one-period lagged
residuals from the estimate of equation (1) will give the coef?cient of the error
correction term. Unlike other techniques of cointegration test, re-parameterizing the
model into the equivalent vector error correction model is not required.
2.2 Non-causality test
The bivariate causality test with two equations can be used to test for Granger
causality provided that the ?rst difference series are stationary. However, Engle and
Granger (1987) and Granger (1988) show that a VAR model in ?rst differences with
cointegrated variables can be misspeci?ed and lead to inferences with errors. Toda and
Yamamoto (1995) develop the test for causal relationship among variables as an
alternative to the standard Granger causality test. This procedure is conducted in the
sense of non-causalitytest. Inthe case of bivariate VARmodel havingk lags, the variables
in the model appear in their levels. The advantage of this method is that one does not need
to knowa priori whether the variables are cointegrated as long as the order of integration
of series does not exceedthe laglengthof the speci?edVARmodel. To use this method, the
optimal lag length is determined using Akaike information criterion (AIC) or Schwartz
information criterion (SIC), then a VAR of order k
*
¼ k þ d
max
is estimated, where d
max
is the maximumanticipatedorder of integration. AccordingtoRambaldi andDoran(1996),
the Wald tests for linear or non-linear restrictions are valid whether the series is I(0), I(1)
or I(2). Since many researchers may ?nd the biases when using the standard unit root and
cointegration tests, this procedure can prevent such biases. The test of whether the
variables in the model Granger cause each other is a test of the joint restriction where
all coef?cients are zero.
Following Toda and Yamamoto (1995) non-causality test, the VAR model is
speci?ed in the following equations:
Thai stock and
foreign exchange
markets
309
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
5
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
LSP
t
¼ a
0
þ
X
k
i¼1
a
i
LSP
t2i
þ
X
kþd
max
j¼kþ1
a
j
LSP
t2j
þ
X
k
i¼1
b
i
LEX
t2i
þ
X
kþd
max
j¼kþ1
b
j
LEX
tþj
þ u
1t
ð4Þ
and:
LEX
t
¼ a
1
þ
X
k
i¼1
g
i
LEX
t2i
þ
X
kþd
max
j¼kþ1
g
j
LEX
t2j
þ
X
k
i¼1
d
i
LSP
t2i
þ
X
kþd
max
j¼kþ1
d
j
LSP
tþj
þ u
2t
ð5Þ
The error terms in the VAR model are assumed to be white noise. Since the extra
lagged variables are included in the model, the causality test is conducted by testing
for zero restrictions of the coef?cients of all lag variables.
2.3 Two-step approach
Cointegration and causality or non-causality tests might not adequately capture the
interactions between stock and foreign exchange markets due to the associated risk in
these two markets. It is possible that exchange rate and stock returns are positively
related, and that exchange rate risk might cause stock return to fall. Furthermore, there
might be causal relationship between volatilities of the two markets.
The two-step approach is used to explain these relationships. The ?rst step is to
estimate a system of equations to model the stock return and exchange rate return
processes, along with their conditional variances. These estimates can generate stock
return volatility and exchange rate return volatility series. In the second step, these
series, along with the stock return and exchange rate return series are employed in
multivariate Granger causality test. Because the relationship in means and in variances
is thought to take several periods, the two-step approach provides room for the ability
to establish causality even though this approach has been criticized for the generated
volatility or uncertainty series. Speci?cally, in the ?rst step, a bivariate generalized
autoregressive heteroskedastic model with constant conditional correlation
(CCC-GARCH) model of Bollerslev (1990) consists of the following ?ve equations:
r
1;t
¼ a
1;0
þ
X
p
i¼1
a
1;i
r
1;t2i
þ
X
p
i¼1
b
1;i
r
2;t2i
þ 1
1;t
ð6Þ
h
1;t
¼ m
1
þ a
1;1
1
2
1;t21
þ b
1;1
h
1;t21
ð7Þ
r
2;t
¼ a
2;0
þ
X
p
i¼1
a
2;i
r
2;t2i
þ
X
p
i¼1
b
2;i
r
2;t2i
þ 1
2;t
ð8Þ
h
2;t
¼ m
2
þ a
2;1
1
2
2;t21
þ b
2;1
h
2;t21
ð9Þ
h
12;t
¼ r
12
ðh
1;t
Þ
1=2
ðh
2;t
Þ
1=2
ð10Þ
where r
1
is the stock return, and r
2
is the return on exchange rate. The conditional
variances of stock return and return on exchange rate are given by h
1
and h
2
,
JFEP
4,4
310
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
5
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
respectively. The conditional covariance between r
1
and r
2
is given by h
12
. The model
assumes that the coef?cients for the ARCH terms (a
1,1
, and a
2,1
) and the constants,
a
1,0
and a
2,0
, are non-negative. The coef?cients associated with the GARCH terms
(b
1,1
and b
2,1
) are assumed to be greater than or equal to zero. The covariance between
the error terms for the stock return and exchange rate return processes is given by h
12
.
It is assumed that normality of the error terms in equations (6) and (8) is valid.
Additionally, it is assumed that a constant conditional correlation coef?cient r
12
, in
equation (10) can take on values from 21 to 1.
The model is estimated simultaneously in a system of equations to yield estimates
of the conditional means, variances and covariance of stock return and exchange
rate return processes. These estimates are used to derive the two volatility series.
The causal relationships between stock return, stock return volatility, return on
exchange rate and exchange rate return volatility can then be obtained from standard
multivariate Granger causality test. It should be noted that the bivariate GARCH
model with ?ve equations can be used to control for the possibility of correlation
between the error terms in the stock return and exchange rate return processes.
This speci?cation is a simple bivariate GARCH model used by Fountas et al. (2006).
3. Empirical results
The monthly data used in this study are the stock market index or SP from the Stock
Exchange of Thailand, and the nominal bilateral exchange rate (EX) from the Bank of
Thailand. The variable EX is in terms of domestic currency (Thai baht) per US dollar.
The period covers July 1997 to June 2010 with 156 observations. This is the period that
the country switched from ?xed to ?oating exchange rate regime. The stock market
return (r
1
) is calculated by the percentage change of stock market index (or SP) while
the exchange rate return (r
2
) is the percentage change of the nominal bilateral exchange
rate. Expected depreciation of domestic currency can discourage foreign investors to
invest in domestic stocks because loss from exchange rate will be materialized when
they sell those stocks for capital gain. Expected appreciation of domestic currency will
do the opposite. Therefore, exchange rate movements can generate the negative or
positive exchange rate return for foreign investors. The US dollar exchange rate series
is used because it can be a good representation of the price in the foreign exchange
market. The nominal effective exchange rate, which is the weighted average index
of various bilateral exchange rates might not be relevant to many foreign investors
who do not want to imitate the effective exchange rate for their currency portfolios.
Most investors only hold a single or few major currencies. In fact, the US dollar
exchange rate dominates all other major currencies. The large depreciation of Thai
bath against the US dollar at the beginning of the 1997 ?nancial crisis was the main
concern of importers, investors as well as policymakers.
In the ?rst step of the analysis, the data are transformed into logarithmic series.
These series are tested for unit roots using the Augmented Dickey-Fuller (ADF) and
Phillips and Perron (PP) tests with a constant only and a constant and a linear trend.
The optimal number of lags is chosen by AIC for the ADF tests while the optimal
bandwidth for the PP tests are chosen by Newey-West using Bartlett kernel spectral
estimation method. The results are reported in Table I.
For the log of nominal bilateral exchange rate series (LEX), all tests show that this
series is I(0), except for the PP test that includes a linear trend. The test results for the
Thai stock and
foreign exchange
markets
311
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
5
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
V
a
r
i
a
b
l
e
A
D
F
t
e
s
t
(
c
o
n
s
t
a
n
t
)
A
D
F
t
e
s
t
(
c
o
n
s
t
a
n
t
þ
t
r
e
n
d
)
P
P
t
e
s
t
(
c
o
n
s
t
a
n
t
)
P
P
t
e
s
t
(
c
o
n
s
t
a
n
t
þ
t
r
e
n
d
)
L
E
X
2
0
.
9
8
4
[
9
]
(
0
.
7
5
8
)
2
2
.
0
7
8
[
9
]
(
0
.
5
5
3
)
2
2
.
3
7
9
[
3
]
(
0
.
1
5
0
)
2
4
.
0
9
6
[
3
]
(
0
.
0
0
8
)
*
*
*
D
L
E
X
2
1
0
.
6
1
1
[
0
]
(
0
.
0
0
0
)
*
*
*
2
1
0
.
6
4
6
[
0
]
(
0
.
0
0
0
)
*
*
*
2
1
1
.
2
3
7
[
2
2
]
(
0
.
0
0
0
)
*
*
*
2
1
1
.
6
5
5
[
2
3
]
(
0
.
0
0
0
)
*
*
*
L
S
P
2
1
.
5
1
3
[
0
]
(
0
.
5
2
5
)
2
3
.
2
5
5
[
4
]
(
0
.
0
7
8
)
*
2
1
.
6
6
8
[
2
]
(
0
.
4
4
5
)
2
3
.
4
3
6
[
3
]
(
0
.
0
5
0
)
*
*
D
L
S
P
2
1
1
.
6
2
6
[
0
]
(
0
.
0
0
0
)
*
*
*
2
1
1
.
6
1
1
[
0
]
(
0
.
0
0
0
)
*
*
*
2
1
1
.
6
1
2
[
4
]
(
0
.
0
0
0
)
*
*
*
2
1
1
.
6
1
2
[
0
]
(
0
.
0
0
0
)
*
*
*
N
o
t
e
s
:
S
i
g
n
i
?
c
a
n
t
a
t
:
*
1
0
,
*
*
5
a
n
d
*
*
*
1
p
e
r
c
e
n
t
l
e
v
e
l
s
,
r
e
s
p
e
c
t
i
v
e
l
y
;
L
E
X
s
t
a
n
d
s
f
o
r
l
o
g
o
f
e
x
c
h
a
n
g
e
r
a
t
e
w
h
i
l
e
L
S
P
s
t
a
n
d
s
f
o
r
l
o
g
o
f
s
t
o
c
k
p
r
i
c
e
s
;
t
h
e
n
u
m
b
e
r
i
n
b
r
a
c
k
e
t
s
i
s
t
h
e
o
p
t
i
m
a
l
l
a
g
l
e
n
g
t
h
d
e
t
e
r
m
i
n
e
d
b
y
A
I
C
f
o
r
t
h
e
A
D
F
t
e
s
t
s
a
n
d
t
h
e
o
p
t
i
m
a
l
b
a
n
d
w
i
d
t
h
d
e
t
e
r
m
i
n
e
d
b
y
t
h
e
B
a
r
t
l
e
t
t
k
e
r
n
e
l
f
o
r
t
h
e
P
P
t
e
s
t
s
;
t
h
e
n
u
m
b
e
r
i
n
p
a
r
e
n
t
h
e
s
e
s
i
s
t
h
e
p
-
v
a
l
u
e
p
r
o
v
i
d
e
d
b
y
M
a
c
K
i
n
n
o
n
(
1
9
9
6
)
Table I.
Unit root tests
JFEP
4,4
312
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
5
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
log of the stock price series (LSP) show that the series is I(0) by the PP test that include
a linear trend while other tests show that it is I(1). Therefore, it is not conclusive that
the series of dependent and independent variables are I(0) or I(1) because of the mixed
results of the unit root tests. However, both series are stationary in ?rst differences.
The second step is to test for cointegration between the SP and exchange rate series.
Based upon the results of unit root tests shown in Table I, the ARDL bounds testing
procedure speci?ed in equation (3) seems to be suitable for the analysis.
The grid search method is used to select p and q in equation (1). Starting from the
most parsimonious ARDL(1,1) and if the model does not show serial correlation at the
5 percent level using Lagrange multiplier (LM) serial correlation test, then the model is
suitable for testing for cointegration. However, if the serial correlation is present, the
number of lagged ?rst differences will increase. The search continues for all
combinations of p and q until a model that is free of serial correlation is detected.
The suitable ARDL( p, q) model with the lag order of p ¼ 1 for the dependent variable
(LSP) and the lag order of q ¼ 1 for the independent variable (LEX) is called the
“ARDL(1,1)” model. The ARDL(1,1) model is used to test for cointegrating relationship
between LSP and LEX because it is free of serial correlation with the x
2
(2) ¼ 2.609
and its p-value is 0.271. By adding lagged level of the pair of variables, the computed
F-statistic is 1.248. The critical values at the 5 percent level from Table CI(iii) case III in
Pesaran et al. (2001) is 5.73 for the upper bound and 4.94 for the lower bound. Since the
computed F-statistic is below the lower bound critical value, the null hypothesis of
no cointegration can be accepted. In other words, there is no long-run relationship
between SP and exchange rate. This result is consistent with the ?nding of Kown and
Shin (1999) who ?nd that there is no cointegration between stock market index and
foreign exchange rate in a bivariate framework.
The third step is to apply the non-causality test in a VAR model using level of each
pair of the series. The empirical results of the Toda-Yamamoto causality test between
nominal US dollar exchange rate and stock market index are reported in Table II.
Based upon the results of unit root test in Table I, the anticipated maximum order of
integration (d
max
) is one. Using AIC to determine the optimal lag length in the VAR
model, the lag length appears to be two. Therefore, the (k þ d
max
) order VAR is three.
The results in Table II showthat the null hypothesis that LEX does not cause LSP is
accepted, but the null hypothesis that LSPdoes not cause LEXis rejected at the 1 percent
Null hypothesis Modi?ed Wald statistic p-value
LEX does not cause LSP 4.355 (þ) 0.226
LSP does not cause LEX 14.118
*
(2) 0.003
Misspeci?cation tests for the VAR model
Test statistic p-value
JB 15.823 0.000
LM 1.555 0.817
WH 143.083 0.000
Notes: signi?cant at:
*
1 percent; LEX stands for log of exchange rate while LSP stands for log of
stock prices; (þ) indicates the positive sum of the coef?cients of lagged variables, which is positive
causation; JB is the Jarque-Bera statistic for testing the null hypothesis that the residuals are
multivariate normal; LM is the Lagrange multiplier test for serial correlation up to third order in the
residuals, and WH is the White heteroskedasticity test of the residuals
Table II.
Results of non-causality
test between
LEX and LSP
Thai stock and
foreign exchange
markets
313
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
5
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
level of signi?cance. The results indicate that a change of nominal exchange rate does
not cause SP to change even though the sum of the estimated coef?cients of all lagged
independent variables is positive. In the other direction, a rise in SP causes the nominal
exchange rate to fall. Therefore, there seems to be a positive unidirectional causality
between SP and nominal exchange rate.
Further tests are conductedto examine the misspeci?cation of the augmentedVAR(3)
model used in the analysis. In Table II, the Lagrangian multiplier (LM) test statistic
indicates the acceptance of the null hypothesis that there is no serial correlation in the
residuals up to the third order of lags. Additionally, the White heteroskedasticity test
shows that the null hypothesis of the presence of ARCH effect can be rejected at the
1 percent level of signi?cance. However, the Jarque-Bera ( JB) statistic shows that the
residuals are not multivariate normal. Thus, it cannot be concluded that the speci?ed
VAR(k þ d
max
) model is suitable for conducting the non-causality tests. In other words,
the results in Table II are not reliable.
The summary statistics show that both stock and exchange rate returns are
negatively skewed. Excess kurtosis is observed in the exchange rate return series. The
JB statistics reject the null hypothesis of normal distribution in both return series,
which indicates the suitability of a GARCH speci?cation. The causal relationships
between stock and exchange rate returns and their conditional variances, and
especially the impact of exchange rate risk on stock market return under the ?oating
regime are tested using a bivariate CCC-GARCH(1,1) model and the standard Granger
causality test. The CCC-bivariate GARCH model gives the results of stock return
volatility as well as exchange rate return volatility while the standard causality test
gives the causal relationships between the variables of interest.
The results of the estimate of the bivariate AR(p)-CCC-GARCH(1,1) model are
reported in Table III. Assuming constant conditional correlation, the model performs
quite well in the dataset. In addition, the threshold GARCH model proposed by Zakoian
(1994) is also estimated, but the asymmetry is not found. The estimated conditional
correlation is 20.531 which is signi?cant at the 1 percent level. This implies that the
two return processes are not independent. The estimated ARCH and GARCH
parameters are non-negative. In addition, the sum of the coef?cients of the two terms is
0.959 for the r
1
-equation, and 0.815 with insigni?cant GARCH term in the r
2
-equation.
This indicates the stationary conditional variance series. The system is also free of
serial correlation. The multivariate Granger causality test is thus performed on
stationary series, and the results are reported in Table IV.
The results show some crucial ?ndings. First, an increase in stock return seems to
cause exchange rate return to fall, but the result is not signi?cant. Therefore, there is no
evidence of the impact of stockmarket returnonexchange rate. Second, a rise inexchange
rate return signi?cantly causes stock return to rise. Thus, there exists unidirectional
causalityinthe meanequations. Third, volatility causations betweenthe two markets are
observed. An increase in stock return volatility signi?cantly causes exchange rate return
volatility to rise while an increase in exchange rate return volatility causes stock return
volatility to fall. In causality sense, there is bidirectional causality in conditional
variances, but in different directions. In other words, an increase in the risk in the foreign
exchange market causes the risk in the stock market to decrease. However, an increase in
the risk in the stock market causes the risk in the foreign exchange market to rise.
The interesting ?nding here is that the rising risk in the foreign exchange market causes
JFEP
4,4
314
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
5
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
P
a
n
e
l
A
:
s
t
o
c
k
r
e
t
u
r
n
e
q
u
a
t
i
o
n
r
1
,
t
¼
2
2
.
4
9
9
*
*
*
þ
0
.
1
7
0
*
r
1
,
t
2
1
þ
0
.
0
5
8
r
1
,
t
2
2
þ
0
.
1
3
7
r
1
,
t
2
3
þ
0
.
0
1
1
r
1
,
t
2
4
2
0
.
5
7
0
*
*
r
2
,
t
2
1
þ
0
.
3
8
3
r
2
,
t
2
2
(
2
2
.
5
1
7
)
(
1
.
8
0
3
)
(
0
.
6
5
9
)
(
1
.
4
3
0
)
(
0
.
1
2
4
)
(
2
2
.
5
1
0
)
(
1
.
2
7
1
)
(
t
-
s
t
a
t
i
s
t
i
c
i
n
p
a
r
e
n
t
h
e
s
e
s
)
h
1
,
t
¼
2
.
7
9
7
þ
0
.
1
8
0
*
*
*
1
21
,
t
2
1
þ
0
.
7
7
9
*
*
*
h
1
,
t
2
1
(
1
.
2
2
2
)
(
2
.
5
9
5
)
(
1
1
.
3
4
6
)
(
t
-
s
t
a
t
i
s
t
i
c
i
n
p
a
r
e
n
t
h
e
s
e
s
)
P
a
n
e
l
B
:
e
x
c
h
a
n
g
e
r
a
t
e
c
h
a
n
g
e
e
q
u
a
t
i
o
n
r
2
,
t
¼
2
0
.
3
1
6
þ
0
.
1
1
0
r
2
,
t
2
1
þ
0
.
0
3
1
r
2
,
t
2
2
2
0
.
0
4
8
*
*
*
r
1
,
t
2
1
þ
0
.
0
2
4
r
1
,
t
2
2
þ
0
.
0
0
9
r
1
,
t
2
3
þ
0
.
0
3
1
r
1
,
t
2
4
(
2
1
.
6
1
6
)
(
0
.
5
6
7
)
(
0
.
3
5
2
)
(
2
2
.
8
2
0
)
(
1
.
3
7
7
)
(
0
.
5
6
7
)
(
0
.
3
5
2
)
(
t
-
s
t
a
t
i
s
t
i
c
i
n
p
a
r
e
n
t
h
e
s
e
s
)
h
2
,
t
¼
1
.
4
3
8
*
*
*
þ
0
.
7
4
1
*
*
*
1
22
,
t
2
1
þ
0
.
4
7
3
h
2
,
t
2
1
(
3
.
9
4
1
)
(
3
.
2
2
7
)
(
0
.
0
7
4
)
(
t
-
s
t
a
t
i
s
t
i
c
i
n
p
a
r
e
n
t
h
e
s
e
s
)
P
a
n
e
l
C
:
c
o
r
r
e
l
a
t
i
o
n
c
o
e
f
?
c
i
e
n
t
h
1
2
,
t
¼
2
0
.
5
3
1
*
*
*
(
h
1
,
t
)
1
/
2
(
h
2
,
t
)
1
/
2
(
2
3
.
8
1
6
)
(
t
-
s
t
a
t
i
s
t
i
c
i
n
p
a
r
e
n
t
h
e
s
e
s
)
P
a
n
e
l
D
:
s
y
s
t
e
m
d
i
a
g
n
o
s
t
i
c
t
e
s
t
Q
(
4
)
¼
6
.
1
7
5
,
Q
(
8
)
¼
1
8
.
3
7
5
5
(
0
.
9
8
6
)
(
0
.
9
7
4
)
(
p
-
v
a
l
u
e
i
n
p
a
r
e
n
t
h
e
s
e
s
)
N
o
t
e
s
:
S
i
g
n
i
?
c
a
n
t
a
t
:
*
1
0
,
*
*
5
a
n
d
*
*
*
1
p
e
r
c
e
n
t
l
e
v
e
l
,
r
e
s
p
e
c
t
i
v
e
l
y
;
r
1
i
s
t
h
e
m
o
n
t
h
l
y
s
t
o
c
k
r
e
t
u
r
n
c
a
l
c
u
l
a
t
e
d
f
r
o
m
t
h
e
r
a
t
e
o
f
c
h
a
n
g
e
i
n
t
h
e
s
t
o
c
k
m
a
r
k
e
t
i
n
d
e
x
a
n
d
r
2
i
s
t
h
e
m
o
n
t
h
l
y
e
x
c
h
a
n
g
e
r
a
t
e
r
e
t
u
r
n
c
a
l
c
u
l
a
t
e
d
f
r
o
m
t
h
e
r
a
t
e
o
f
c
h
a
n
g
e
i
n
t
h
e
n
o
m
i
n
a
l
b
i
l
a
t
e
r
a
l
e
x
c
h
a
n
g
e
r
a
t
e
;
h
1
i
s
t
h
e
c
o
n
d
i
t
i
o
n
a
l
v
a
r
i
a
n
c
e
o
f
s
t
o
c
k
r
e
t
u
r
n
,
h
2
i
s
t
h
e
c
o
n
d
i
t
i
o
n
v
a
r
i
a
n
c
e
o
f
e
x
c
h
a
n
g
e
r
a
t
e
r
e
t
u
r
n
,
a
n
d
h
1
2
i
s
t
h
e
c
o
n
d
i
t
i
o
n
a
l
c
o
v
a
r
i
a
n
c
e
o
f
s
t
o
c
k
r
e
t
u
r
n
a
n
d
e
x
c
h
a
n
g
e
r
a
t
e
r
e
t
u
r
n
;
t
h
e
l
a
g
s
o
f
r
1
a
n
d
r
2
a
r
e
c
h
o
s
e
n
i
n
s
u
c
h
a
w
a
y
t
h
a
t
t
h
e
s
y
s
t
e
m
d
i
a
g
n
o
s
t
i
c
t
e
s
t
s
h
o
w
n
o
a
u
t
o
c
o
r
r
e
l
a
t
i
o
n
i
n
t
h
e
r
e
s
i
d
u
a
l
s
i
n
t
h
e
m
e
a
n
e
q
u
a
t
i
o
n
s
;
Q
(
k
)
i
s
t
h
e
B
o
x
-
P
i
e
r
c
e
s
t
a
t
i
s
t
i
c
f
o
r
t
h
e
r
e
s
i
d
u
a
l
s
Table III.
Estimates of the bivariate
AR(p)-CCC-GARCH(1,1)
model of stock market
and exchange rate
returns for the period July
1997 to June 2010
Thai stock and
foreign exchange
markets
315
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
5
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
the stock market return to fall, which is the expected phenomenon under the ?oating
regime. In the events of rising exchange rate uncertainty, foreign investors might
diversifyawayfromthe domestic stocks. However, a rise inthe stockmarket riskdoes not
impose any impact on the exchange rate return. In fact, the Thai stock market is a small
emergingmarket that shouldnot in?uence the foreignexchange market. The dilemmafor
policymakers is that stabilizing the exchange rate will lead to higher level of uncertainty
in the stock market, which will lower the rates of return of domestic stocks. The Bank of
Thailand occasionally intervenes in the foreign exchange market when the exchange
rate is out of line. This practice is used to maintain the position of trade balance of the
countries. Even though the intervention by the Bank of Thailand can stabilizes the
exchange rate, the rising trend in volatility of stock return and the lower stock return
are unavoidable. One of the main ?nding from the present study is from the result of a
positive spillover of risk fromthe stockmarket to the risk in the foreign exchange market.
This evidence gives roomfor the Security Exchange Commission to set proper rules and
regulations for investors so as to mitigate ?uctuations of SPandthus stockmarket return.
As a result, the risk in the foreign exchange market will be reduced. However, these
measures might not be successful if there are excessive speculations in the stock market.
4. Concluding remarks
This study employs the recently developed time series analysis techniques to explore the
causal relationship between SP and exchange rates in an emerging market economy,
namely Thailand. The results from bounds testing for cointegration show that there
is no long-run relationship between SP (stock market index) and exchange rate in a
bivariate framework, which is consistent with the results of other empirical studies.
An alternative approach is to examine the causal links between SP and exchange rate.
The non-causality test is used and the results show that there exists unidirectional
causality between SP and exchange rates. However, the results from non-causality test
should not be reliable because the estimated VAR model does not pass the diagnostic
test of multivariate normality in the residuals.
The results from the two-step approach show that unidirectional causality between
stock and exchange rate return is observed. An increase in the exchange rate return
causes stock return to rise. More importantly, a rising trend in the risk in the foreign
exchange market causes stock return to fall. Both domestic and foreign investors
Hypothesis F-statistic p-value
r
1
does not cause r
2
0.077 (2) 0.630
r
2
does not cause r
1
3.481
* *
(þ) 0.001
h
1
does not cause r
2
1.255 (þ) 0.273
h
2
does not cause r
1
4.081
* *
(2) 0.000
h
1
does not cause h
2
2.238
*
(þ) 0.029
h
2
does not cause h
1
3.784
* *
(2) 0.001
Notes: Signi?cant at:
*
5 and
* *
1 percent level, respectively; r
1
is the monthly stock return calculated
from the rate of change in the stock market index and r
2
is the monthly exchange rate return calculated
from the rate of change in the nominal bilateral exchange rate; h
1
is the conditional variance of stock
return, h
2
is the condition variance of exchange rate return; the optimal lag length determined by AIC
is eight; (þ) denotes positive causation and (2) denotes negative causation
Table IV.
Results from multivariate
Granger causality test
JFEP
4,4
316
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
5
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
should be aware of the risk or uncertainty in the foreign exchange market because it
can cause their portfolio return to fall. For policymakers, reducing exchange rate risk
cannot be done without the associated costs from a rising risk in the stock market.
Instead, some measures that can reduce the stock market risk are bene?cial to the Thai
economy in that the foreign exchange market risk will be reduced. The reservation for
this suggestion is that the measures might not be successful in the presence of
excessive speculations of investors in the stock market.
References
Abballa, I.S.A. and Murinde, V. (1997), “Exchange rate and stock price interactions in emerging
?nancial markets: evidence on India, Korea, Pakistan, and Philippines”, Applied Financial
Economics, Vol. 7 No. 1, pp. 25-35.
Aggarwal, R. (1981), “Exchange rates and stock prices: a study of US capital market under
?oating exchange rates”, Akron Business and Economic Review, Vol. 12 No. 3, pp. 7-12.
Ajayi, R.A. and Mougoue, M. (1996), “On the dynamic relation between stock prices and
exchange rates”, Journal of Financial Research, Vol. 19 No. 2, pp. 193-207.
Atindehou, R.B. and Gueyie, J.-P. (2001), “Canadian chartered bank stock returns and exchange
rate risk”, Management Decision, Vol. 39 No. 4, pp. 285-95.
Bahmani-Oskooee, M. and Sohabian, A. (1992), “Stock prices and the effective exchange rate of
the dollar”, Applied Economics, Vol. 24 No. 4, pp. 459-64.
Bollerslev, T. (1990), “Modelling the coherence in short-run nominal exchange rates:
a multivariate generalized ARCH model”, Review of Economics and Statistics, Vol. 72
No. 3, pp. 498-505.
Booth, A. and Rotenberg, W. (1990), “Assessing foreign exchange exposure: theory and
applications using Canadian ?rms”, Journal of International Financial Management and
Accounting, Vol. 2, pp. 1-22.
Branson, W.H. and Henderson, D.W. (1985), “The speci?cation and in?uence of assets markets”,
in Jones, R.W. and Kenen, P.B. (Eds), Handbook of International Economics, Vol. 2,
Elsevier, Amsterdam.
Chen, J., Naylor, M. and Lu, X. (2004), “Some insights into the foreign exchange pricing puzzle:
evidence from a small open economy”, Paci?c-Basin Finance Journal, Vol. 12 No. 1,
pp. 41-64.
Choi, D.F. and Fang, F. (2009), “Volatility spillover between New Zealand stock market returns
and exchange rate changes before and after the 1997 Asian ?nancial crisis”, Asian Journal
of Finance and Accounting, Vol. 1 No. 2, pp. 106-17.
Choi, J.J. and Prasad, A.M. (1995), “Exchange rate sensitivity and its determinants: a ?rm
and industry analysis of US multinationals”, Financial Management, Vol. 24 No. 3,
pp. 77-88.
Dornbusch, R. and Fisher, S. (1980), “Exchange rates and the current account”, American
Economic Review, Vol. 70 No. 5, pp. 960-71.
Engle, R.F. and Granger, C.W.J. (1987), “Co-integration and error correction: representation,
estimation and testing”, Econometrica, Vol. 55 No. 2, pp. 1057-72.
Fountas, S., Karanasos, M. and Kim, J. (2006), “In?ation uncertainty, output growth uncertainty,
and macroeconomic performance”, Oxford Bulletin of Economics and Statistics, Vol. 68
No. 3, pp. 319-43.
Frank, P. and Young, A. (1972), “Stock price reaction of multinational ?rms to exchange
realignments”, Financial Management, Vol. 1 No. 3, pp. 66-73.
Thai stock and
foreign exchange
markets
317
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
5
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Giovannini, A. and Jorion, P. (1987), “Interest rate and risk premia in the stock market and in the
foreign exchange market”, Journal of International Money and Finance, Vol. 6 No. 1,
pp. 107-23.
Granger, C.W.J. (1969), “Investigating causal relations by econometrics models and cross spectral
methods”, Econometrica, Vol. 37 No. 3, pp. 424-38.
Granger, C.W.J. (1988), “Some recent developments in a concept of causality”, Journal of
Econometrics, Vol. 39 Nos 1/2, pp. 199-211.
Granger, C.W.J., Huang, B.N. and Yang, C.W. (2000), “A bivariate causality between stock prices
and exchange rates: evidence from recent Asian ?u”, Quarterly Review of Economics and
Finance, Vol. 40 No. 3, pp. 337-54.
Kanas, A. (2000), “Volatility spillover from stock returns and exchange rate changes:
international evidence”, Journal of Business Finance and Accounting, Vol. 27 Nos 3/4,
pp. 447-67.
Kim, K. (2003), “Dollar exchange rate and stock price: evidence from multivariate cointegration
and error correction model”, Review of Financial Economics, Vol. 12 No. 3, pp. 301-13.
Kown, C.S. and Shin, T.S. (1999), “Cointegration and causality between macroeconomic variables
and stock market returns”, Global Finance Journal, Vol. 10 No. 1, pp. 78-81.
MacDonald, R. and Taylor, M. (1992), “Exchange rate economics: a survey”, IMF Staff Papers,
Vol. 39 No. 1, pp. 1-57.
MacKinnon, J.G. (1996), “Numerical distribution functions for unit root and cointegration tests”,
Journal of Applied Econometrics, Vol. 11 No. 6, pp. 601-18.
Ma, C.K. and Kao, G.W. (1990), “On exchange rate changes and stock price reactions”, Journal of
Business Finance and Accounting, Vol. 17 No. 3, pp. 441-9.
Mansor, H.I. (2000), “Cointegration and Granger causality tests of stock price and exchange rate
interactions in Malaysia”, ASEAN Economic Bulletin, Vol. 17 No. 1, pp. 36-47.
Nabiha, N. and Mounira, B.A. (2009), “Interaction between stock and exchange crisis in emerging
markets”, in Jay Choi, J. and Papaioannou, M.G. (Eds), Credit, Currency, or Derivatives:
Instruments of Global Financial Stability or Crisis?, International Finance Review, Vol. 10,
Emerald, Bradford, pp. 171-89.
Narayan, P.K. (2009), “On the relationship between stock prices and exchange rate for India”,
Review of Paci?c Basin Financial Markets and Policies, Vol. 12 No. 2, pp. 289-308.
Nieh, C. and Lee, C. (2001), “Dynamic relationship between stock prices and exchange rates for
G-7 countries”, Quarterly Review of Economics and Finance, Vol. 41 No. 4, pp. 477-90.
Pan, M., Fok, R.C. and Liu, Y.A. (2007), “Dynamic linkages between exchange rates and stock
prices: evidence from east Asian markets”, International Review of Economics and
Finance, Vol. 16 No. 4, pp. 503-20.
Pesaran, M.H., Shin, Y. and Smith, R.J. (2001), “Bounds testing approaches to the analysis of level
relationship”, Journal of Applied Econometrics, Vol. 16 No. 3, pp. 289-326.
Phylaktis, K. and Ravazzolo, F. (2005), “Stock prices and exchange rate dynamics”, Journal of
International Money and Finance, Vol. 24 No. 7, pp. 1031-53.
Rambaldi, A.N. and Doran, H.E. (1996), “Testing for Granger non-causality in cointegrated
system made easy”, Working Papers in Econometrics and Applied Statistics, No. 88,
University of New England.
Soenen, L.A. and Hennigar, E.S. (1988), “An analysis of exchange rates and stock prices: the US
experience between 1980 and 1986”, Arkron Business and Economic Review, Vol. 19 No. 4,
pp. 7-16.
JFEP
4,4
318
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
5
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Stavarek, D. (2005), “Stock prices and exchange rates in the EU and the United States: evidence
on their mutual interactions”, Czech Journal of Economics and Finance, Vol. 55 Nos 3/4,
pp. 141-61.
Toda, H.Y. and Yamamoto, T. (1995), “Statistical inferences in vector autoregressions with
possibly integrated process”, Journal of Econometrics, Vol. 66 Nos 1/2, pp. 225-50.
Yu, Q. (1997), “Stock prices and exchange rates: experience in leading East Asian ?nancial
centres: Tokyo, Hong Kong and Singapore”, Singapore Economic Review, Vol. 41 No. 1,
pp. 47-56.
Zakoian, J.M. (1994), “Threshold heteroskedastic models”, Journal of Economic Dynamic and
Control, Vol. 18 No. 5, pp. 931-55.
About the author
Komain Jiranyakul is an Associate Professor at National Institute of Development
Administration, Bangkok, Thailand. He teaches macroeconomics and ?nancial economics.
He recently published papers in Journal of The Asia Paci?c Economy, Economics Bulletin, Journal
of Asian Economics, and Asian Economic Journal. Komain Jiranyakul can be contacted at:
[email protected]
Thai stock and
foreign exchange
markets
319
To purchase reprints of this article please e-mail: [email protected]
Or visit our web site for further details: www.emeraldinsight.com/reprints
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
5
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
This article has been cited by:
1. Arif Billah Dar, Aasif Shah, Niyati Bhanja, Amaresh Samantaraya. 2014. The relationship between stock
prices and exchange rates in Asian markets. South Asian Journal of Global Business Research 3:2, 209-224.
[Abstract] [Full Text] [PDF]
2. Walid M. A. Ahmed. 2014. Dynamic interactions between Egyptian equity and currency markets prior
to and during political unrest. Applied Financial Economics 1-13. [CrossRef]
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
5
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
doc_389299362.pdf
The purpose of the present study is to directly examine the relationship between bilateral
exchange rate and stock market index in a bivariate framework during the period of the floating
exchange rate regime in Thailand
Journal of Financial Economic Policy
Linkages between Thai stock and foreign exchange markets under the floating regime
Komain J iranyakul
Article information:
To cite this document:
Komain J iranyakul, (2012),"Linkages between Thai stock and foreign exchange markets under the floating
regime", J ournal of Financial Economic Policy, Vol. 4 Iss 4 pp. 305 - 319
Permanent link to this document:http://dx.doi.org/10.1108/17576381211279280
Downloaded on: 24 January 2016, At: 21:45 (PT)
References: this document contains references to 38 other documents.
To copy this document: [email protected]
The fulltext of this document has been downloaded 478 times since 2012*
Users who downloaded this article also downloaded:
Abdullah M. Noman, Sarkar Humayun Kabir, Omar K.M.R. Bashar, (2012),"Causality between stock and
foreign exchange markets in Bangladesh", Studies in Economics and Finance, Vol. 29 Iss 3 pp. 174-186http://dx.doi.org/10.1108/10867371211246849
Manish Kumar, (2013),"Returns and volatility spillover between stock prices and exchange rates: Empirical
evidence from IBSA countries", International J ournal of Emerging Markets, Vol. 8 Iss 2 pp. 108-128 http://
dx.doi.org/10.1108/17468801311306984
Tian Yong Fu, Mark J . Holmes, Daniel F.S. Choi, (2011),"Volatility transmission and asymmetric linkages
between the stock and foreign exchange markets: A sectoral analysis", Studies in Economics and Finance,
Vol. 28 Iss 1 pp. 36-50http://dx.doi.org/10.1108/10867371111110543
Access to this document was granted through an Emerald subscription provided by emerald-srm:115632 []
For Authors
If you would like to write for this, or any other Emerald publication, then please use our Emerald for
Authors service information about how to choose which publication to write for and submission guidelines
are available for all. Please visit www.emeraldinsight.com/authors for more information.
About Emerald www.emeraldinsight.com
Emerald is a global publisher linking research and practice to the benefit of society. The company
manages a portfolio of more than 290 journals and over 2,350 books and book series volumes, as well as
providing an extensive range of online products and additional customer resources and services.
Emerald is both COUNTER 4 and TRANSFER compliant. The organization is a partner of the Committee
on Publication Ethics (COPE) and also works with Portico and the LOCKSS initiative for digital archive
preservation.
*Related content and download information correct at time of download.
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
5
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Linkages between Thai stock
and foreign exchange markets
under the ?oating regime
Komain Jiranyakul
School of Development Economics,
National Institute of Development Administration, Bangkok, Thailand
Abstract
Purpose – The purpose of the present study is to directly examine the relationship between bilateral
exchange rate and stock market index in a bivariate framework during the period of the ?oating
exchange rate regime in Thailand.
Design/methodology/approach – The monthly data used in this study are the stock market index
or stock prices from the Stock Exchange of Thailand, and the nominal bilateral exchange rate in terms
of baht per US dollar from the Bank of Thailand. The period covers July 1997 to June 2010 with 156
observations. This is the period that the country switched from ?xed to ?oating exchange rate regime.
The stock market return is calculated by the percentage change of stock market index (or stock prices)
while the exchange rate return is the percentage change of the nominal bilateral exchange rate. Three
estimation methods are used to capture the interaction between stock and foreign exchange markets:
bounds testing for cointegration, non-causality test, and the two-step approach with a bivariate
GARCH model and Granger causality test.
Findings – The results of the present study show that bounds testing for cointegration does not
detect the long-run relationship between stock prices and exchange rate. In addition, the non-causality
test fails the diagnostic test for multivariate normality in the residuals of the estimated VAR model.
However, the two-step approach adequately detects the linkages between the stock and foreign
exchange markets. It is found that there exists positive unidirectional causality running from stock
market return to exchange rate return. The exchange rate risk causes stock return to fall as expected.
Moreover, there are bidirectional causal relations between stock market risk and exchange rate risk,
but in different directions.
Research limitations/implications – Since a rising trend in the risk in the foreign exchange
market causes stock return to fall, both domestic and foreign investors should be aware of the risk or
uncertainty in the foreign exchange market because it can cause their portfolio return to fall. For
policymakers, reducing exchange rate risk cannot be done without the associated costs from a rising
risk in the stock market.
Originality/value – This study provides an evidence of volatility (or risk) spillovers in stock and
foreign exchange markets. In addition, the risk in foreign exchange market that adversely affects
return in the stock market is an expected phenomenon under the ?oating exchange rate regime.
Keywords Stock prices, Exchange rates, Bivariate GARCH, Causality, Volatility spillover, Thailand
Paper type Research paper
1. Introduction
Theoretically, the traditional approach or the ?ow-oriented theory states that a
depreciation of domestic currency can have a crucial impact on stock prices (SP) by
increasing ?rms’ competitiveness, while in turn raising their pro?tability. When ?rms
The current issue and full text archive of this journal is available at
www.emeraldinsight.com/1757-6385.htm
The author is grateful to two anonymous referees for their comments and suggestions that are
very helpful in improving the quality of this paper.
Thai stock and
foreign exchange
markets
305
Journal of Financial Economic Policy
Vol. 4 No. 4, 2012
pp. 305-319
qEmerald Group Publishing Limited
1757-6385
DOI 10.1108/17576381211279280
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
5
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
are able to pay more dividends to stockholders, SP will increase. Thus, there should be
a positive relationship between exchange rates and SP. In this case, exchange rates lead
SP. On the contrary, the portfolio balance approachindicates that SPlead exchange rates
on the ground that a rising trend in SP induces foreign investors to invest more in
domestic stocks. This will cause more capital in?ows, which in turn cause domestic
currency appreciation. In addition, a rise in domestic SP causes wealth to increase, and
thus induces investors to increase their demand for money, which results in a rise in
domestic interest rates. Higher interest rates induce capital in?ows, and thus cause an
appreciation in domestic currency. According to this approach, SP lead exchange rates
with a negative relationship. Another approach called “monetary approach” or “asset
market approach” indicates no linkage between SP and exchange rates due to different
factors in?uencing the two variables. The details of the ?ow-oriented model are in
Dornbusch and Fisher (1980), the portfolio balance approach in Branson and Henderson
(1985). A comprehensive review of the monetary approach is in MacDonald and
Taylor (1992).
Empirical studies suggest that there is no long-run relationship between SP and
exchange rates in most countries, speci?cally fromthe results of bivariate cointegration
tests. Furthermore, the direction of causality seems to depend on speci?c characteristics
of each country being analyzed. Many studies employ multivariate cointegration tests to
investigate the long-run relationship between SP and exchange rates, along with other
macroeconomic variables. For examples, Kown and Shin (1999) ?nd no long-run
relationship between SP and exchange rate in a bivariate framework, but the long-run
relationship between SP and other variables, including exchange rate, is detected. Their
?nding shows that SP are not a leading indicator of the exchange rate and other
macroeconomic variables in Korea. Kim (2003) also employs the multivariate
cointegration test to detect the long-run relationship between SP and the dollar
exchange rate in the USA and ?nds negative relationship between SP and the
exchange rate.
Gradual abolishment of barriers to capital ?ows and foreign exchange restrictions,
along with the adoption of more ?exible exchange rates in emerging markets are the
main causes of the volatility of exchange rates and risk in the portfolio diversi?cation
process in the past decades. Therefore, understanding the relationship between
exchange rates and SP will provide guidance for investors to diversify their portfolios.
The linkages between stock markets and foreign exchange markets are explored by
many researchers who employ modern time series econometrics. The mutual relations
between the two markets are examined by using the techniques of cointegration and
causality between SP and exchange rates. The study by Frank and Young (1972) is
known as the inaugurate attempt to link SPto exchange rates. However, the results show
no relationship between SPand exchange rates measured in terms of the prices of the US
dollar. Aggarwal (1981) uses the US monthly data to investigate the relationship
betweenthe two variables and?nds that the relationshipis stronger inthe short runthan
in the long run during the 1974-1978, which is the ?rst stage of the post-Brettonwood
systemthat more volatile exchange rates are observed. Soenen and Hennigar (1988) use
monthly data of the US dollar effective exchange rate and stock market index during the
1980-1986 period to investigate this relationship and ?nd a strong negative relationship.
Ma and Kao (1990) examine SP reactions to exchange rate changes under the ?oating
regime and ?nd that stock returns are minimally or not affected by exchange
JFEP
4,4
306
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
5
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
rate ?uctuations. Bahmani-Oskooee and Sohabian (1992) provide an evidence of
bidirectional causality of effective exchange rate and SP (the S&P 500 index) in the short
run, but not in the long run. Ajayi and Mougoue (1996) indicate the causation running
fromSP to exchange rates in eight industrialized economies. Nieh and Lee (2001) ?nd no
long-run relationship between SP and exchange rates in G-7 countries. Stavarek (2005)
uses monthly data to examine the linkages between SP and effective exchange rates in
EU-member countries and the USA. The results show much stronger causations in
countries with developed capital and exchange rates markets than the less-developed
ones. Causations seemto be unidirectional running fromSPto exchange rates. For Asian
economies, Abballa and Murinde (1997) ?nd the causation running fromexchange rates
to SP in India and the Philippines, but no causations in Pakistan and Korea. Yu (1997)
uses daily data on Hong Kong, Japan, and Singapore stock markets during the 1983-1994
period to examine the causality between exchange rates and SP. The results are mixed,
i.e. bidirectional causality in Japan, unidirectional causation running from exchange
rates to SP in Hong Kong, and no causation in Singapore. Mansor (2000) ?nds short-run
causation running from SP to exchange rates in Malaysia using an analysis in a
bivariate framework. Granger et al. (2000) examine the SP-exchange rates nexus in East
Asian countries using Asian ?u daily data from January 3, 1986 to June 16, 1998. The
so-called “Asian ?u” is the ?nancial turmoil that continued to exert its devastating force
from the third quarter of 1997 to the ?rst quarter of 1998. This was the period that SP
and exchange rates plunged intandeminAsianeconomies. Granger et al. (2000) ?nd that
exchange rates lead SP in South Korea with a positive correlation while SP lead
exchange rates in the Philippines with a negative correlation. The strong feedback
relations (or bidirectional causations) are observed in Hong Kong, Malaysia, Singapore,
Thailand, and Taiwan. Phylaktis and Ravazzolo (2005) apply a multivariate framework
to test for the direction of Granger causality in a group of Paci?c Basin countries, namely
Hong Kong, Malaysia, Singapore, Thailand, and the Philippines using monthly data
from 1980 to 1988. Their results show that stock and foreign exchange markets are
positively related and that the US stock market acts as a conduit for these linkages.
Pan et al. (2007) examine the dynamic linkage between exchange rates and SP for seven
East Asian economies during January 1988 to October 1998, and ?nd that the linkage
varies across economies with respect to exchange rate regimes, trade size, the degree of
capital control, and the size of equity market. Nabiha and Mounira (2009) examine the
dynamic relation between stock and exchange crisis in Mexico, Malaysia, Thailand,
Brazil, and Argentina. Using daily data of stock indexes and the US dollar exchange rate
during 1994 and 2003, they ?nd unidirectional causation running from SP to exchange
rate in Mexico, Argentina, and Brazil, but the causation running from exchange rate to
SP in Malaysia and Thailand.
The interactions between the two ?nancial markets are also investigated using a
generalized autoregressive conditional heteroskedastic (GARCH) model. Giovannini and
Jorion (1987) ?nd that stock return movements impact the foreign exchange markets by
changing stock return forecast that can cause changes in economic activities. Booth and
Rotenberg(1990) indicate that exchange rate movements canimpose the adverse effect on
the level of competitiveness of ?rms that are exposed to exchange rate risk. Choi and
Prasad (1995) ?nd that exchange rate movements affect the value of the US multinational
?rms. Atindehou and Gueyie (2001) estimate the three factor pricing model and ?nd that
Canadian bank stock returns are sensitive to exchange rate risk, speci?callythe risk from
Thai stock and
foreign exchange
markets
307
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
5
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
the US dollar rather than that fromthe Canadian dollar. Kanas (2000) ?nds that spillover
from stock markets to foreign exchange markets is signi?cant for ?ve out of six
developed countries, but spillover fromforeign exchange markets to stock markets is not
signi?cant for all countries. Chen et al. (2004) ?nd a lagged stock market reaction to
exchange rate ?uctuation in New Zealand. Furthermore, they ?nd that the 1997 Asian
?nancial crisis plays an important part in the volatility transmission between the two
?nancial markets by changing the nature of spillover from bidirectional spillover before
the crisis to unidirectional one after the crisis. Choi and Fang (2009) con?rmthe results of
Chen et al. (2004). Narayan (2009) investigates the impact of the Indian rupee depreciation
on the Indian stock market returns using daily data and ?nds that a depreciation of the
Indian rupee raises stock return volatility. However, an appreciation causes higher stock
return and lower return volatility.
The main objective of the present study is to directly examine the relationship between
bilateral exchange rate and stock market index in a bivariate framework using monthly
data fromJuly 1997 to June 2010, which is the period of the ?oating exchange rate regime
in Thailand. The afore-mentioned theories are tested. The more recent techniques are
used: the bounds testing for cointegration proposed by Pesaran et al. (2001) and
the non-causality tests proposed byToda andYamamoto (1995). Both procedures provide
simplistic methods to test the nexus between SP and exchange rates. The results of the
present study show that bounds testing for cointegration does not detect the long-run
relationship between SPand exchange rate inThailand. Inaddition, the non-causalitytest
fails the diagnostic test for multivariate normality in the residuals of the estimated vector
autoregressive (VAR) model. To adequately examine the linkages inthe stock and foreign
exchange markets, the results from the two-step approach is also employed. This
approach involves the estimate of a bivariate CCC-GARCH model of Bollerslev (1990)
and the Granger (1969) causality test. The results reveal that there exists positive
unidirectional causality running from stock market return to exchange rate return.
Furthermore, the exchange rate risk causes stock return to fall in the ?oating exchange
rate regime. There are bidirectional causal relations between stock market risk and
exchange rate risk, but in different directions. The paper is organized as follows: the next
section describes the analytical framework of this study. Section 3 presents empirical
results. The last section gives concluding remarks.
2. Methodology
The relationship between SP and exchange rates can be examined by cointegration and
causality tests. If the variables in the model are integrated of order one, I(1) and
cointegrated, the causation is obtained in the vector error correction model, which is
a VAR model in ?rst differences of the series with the inclusion of cointegrating
residuals, i.e. a restricted VAR model. An alternative test in the absence of
cointegration is the standard Granger causality test (Granger, 1969). In case
the variables are not I(1) series, for example they are mixed between I(0) and I(1), the
bounds testing for cointegration can be applied. If cointegration does not exist, the
causal relationship can be tested using the non-causality test in level of the series.
2.1 Cointegration test
The long-run relationship between SP and nominal bilateral exchange rate is
speci?ed as:
JFEP
4,4
308
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
5
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
LSP
t
¼ a þ bLEX
t
þ 1
t
ð1Þ
where LSP is the log of stock market index (or SP) in the Stock Exchange of Thailand,
and LEX is the log of nominal exchange rate.
Pesaran et al. (2001) proposed a new method for testing cointegration called a
conditional autoregressive distributed lag (ARDL) and error correction mechanism.
This is known as “ARDL bounds testing procedure.”
The ARDL model for equation (1) is speci?ed as:
DLSP
t
¼ a
0
þ
X
p
i¼1
b
i
DðLSPÞ
t2i
þ
X
q
j¼0
g
i
DLEX
t2i
þ e
t
ð2Þ
where p and q are the optimal number of lagged differences of log of SP and log of
exchange rate, respectively. By adding the lagged level variables into equation (2), the
computed F-statistic is obtained as shown in equation (3):
DLSP
t
¼ a
0
þa
1
LSP
t21
þa
2
LEX
t21
þ
X
p
i¼1
b
i
DLSP
t2i
þ
X
q
j¼0
g
i
DLEX
t2i
þ e
t
ð3Þ
If cointegration exists, replacing the lagged level variables with the one-period lagged
residuals from the estimate of equation (1) will give the coef?cient of the error
correction term. Unlike other techniques of cointegration test, re-parameterizing the
model into the equivalent vector error correction model is not required.
2.2 Non-causality test
The bivariate causality test with two equations can be used to test for Granger
causality provided that the ?rst difference series are stationary. However, Engle and
Granger (1987) and Granger (1988) show that a VAR model in ?rst differences with
cointegrated variables can be misspeci?ed and lead to inferences with errors. Toda and
Yamamoto (1995) develop the test for causal relationship among variables as an
alternative to the standard Granger causality test. This procedure is conducted in the
sense of non-causalitytest. Inthe case of bivariate VARmodel havingk lags, the variables
in the model appear in their levels. The advantage of this method is that one does not need
to knowa priori whether the variables are cointegrated as long as the order of integration
of series does not exceedthe laglengthof the speci?edVARmodel. To use this method, the
optimal lag length is determined using Akaike information criterion (AIC) or Schwartz
information criterion (SIC), then a VAR of order k
*
¼ k þ d
max
is estimated, where d
max
is the maximumanticipatedorder of integration. AccordingtoRambaldi andDoran(1996),
the Wald tests for linear or non-linear restrictions are valid whether the series is I(0), I(1)
or I(2). Since many researchers may ?nd the biases when using the standard unit root and
cointegration tests, this procedure can prevent such biases. The test of whether the
variables in the model Granger cause each other is a test of the joint restriction where
all coef?cients are zero.
Following Toda and Yamamoto (1995) non-causality test, the VAR model is
speci?ed in the following equations:
Thai stock and
foreign exchange
markets
309
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
5
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
LSP
t
¼ a
0
þ
X
k
i¼1
a
i
LSP
t2i
þ
X
kþd
max
j¼kþ1
a
j
LSP
t2j
þ
X
k
i¼1
b
i
LEX
t2i
þ
X
kþd
max
j¼kþ1
b
j
LEX
tþj
þ u
1t
ð4Þ
and:
LEX
t
¼ a
1
þ
X
k
i¼1
g
i
LEX
t2i
þ
X
kþd
max
j¼kþ1
g
j
LEX
t2j
þ
X
k
i¼1
d
i
LSP
t2i
þ
X
kþd
max
j¼kþ1
d
j
LSP
tþj
þ u
2t
ð5Þ
The error terms in the VAR model are assumed to be white noise. Since the extra
lagged variables are included in the model, the causality test is conducted by testing
for zero restrictions of the coef?cients of all lag variables.
2.3 Two-step approach
Cointegration and causality or non-causality tests might not adequately capture the
interactions between stock and foreign exchange markets due to the associated risk in
these two markets. It is possible that exchange rate and stock returns are positively
related, and that exchange rate risk might cause stock return to fall. Furthermore, there
might be causal relationship between volatilities of the two markets.
The two-step approach is used to explain these relationships. The ?rst step is to
estimate a system of equations to model the stock return and exchange rate return
processes, along with their conditional variances. These estimates can generate stock
return volatility and exchange rate return volatility series. In the second step, these
series, along with the stock return and exchange rate return series are employed in
multivariate Granger causality test. Because the relationship in means and in variances
is thought to take several periods, the two-step approach provides room for the ability
to establish causality even though this approach has been criticized for the generated
volatility or uncertainty series. Speci?cally, in the ?rst step, a bivariate generalized
autoregressive heteroskedastic model with constant conditional correlation
(CCC-GARCH) model of Bollerslev (1990) consists of the following ?ve equations:
r
1;t
¼ a
1;0
þ
X
p
i¼1
a
1;i
r
1;t2i
þ
X
p
i¼1
b
1;i
r
2;t2i
þ 1
1;t
ð6Þ
h
1;t
¼ m
1
þ a
1;1
1
2
1;t21
þ b
1;1
h
1;t21
ð7Þ
r
2;t
¼ a
2;0
þ
X
p
i¼1
a
2;i
r
2;t2i
þ
X
p
i¼1
b
2;i
r
2;t2i
þ 1
2;t
ð8Þ
h
2;t
¼ m
2
þ a
2;1
1
2
2;t21
þ b
2;1
h
2;t21
ð9Þ
h
12;t
¼ r
12
ðh
1;t
Þ
1=2
ðh
2;t
Þ
1=2
ð10Þ
where r
1
is the stock return, and r
2
is the return on exchange rate. The conditional
variances of stock return and return on exchange rate are given by h
1
and h
2
,
JFEP
4,4
310
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
5
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
respectively. The conditional covariance between r
1
and r
2
is given by h
12
. The model
assumes that the coef?cients for the ARCH terms (a
1,1
, and a
2,1
) and the constants,
a
1,0
and a
2,0
, are non-negative. The coef?cients associated with the GARCH terms
(b
1,1
and b
2,1
) are assumed to be greater than or equal to zero. The covariance between
the error terms for the stock return and exchange rate return processes is given by h
12
.
It is assumed that normality of the error terms in equations (6) and (8) is valid.
Additionally, it is assumed that a constant conditional correlation coef?cient r
12
, in
equation (10) can take on values from 21 to 1.
The model is estimated simultaneously in a system of equations to yield estimates
of the conditional means, variances and covariance of stock return and exchange
rate return processes. These estimates are used to derive the two volatility series.
The causal relationships between stock return, stock return volatility, return on
exchange rate and exchange rate return volatility can then be obtained from standard
multivariate Granger causality test. It should be noted that the bivariate GARCH
model with ?ve equations can be used to control for the possibility of correlation
between the error terms in the stock return and exchange rate return processes.
This speci?cation is a simple bivariate GARCH model used by Fountas et al. (2006).
3. Empirical results
The monthly data used in this study are the stock market index or SP from the Stock
Exchange of Thailand, and the nominal bilateral exchange rate (EX) from the Bank of
Thailand. The variable EX is in terms of domestic currency (Thai baht) per US dollar.
The period covers July 1997 to June 2010 with 156 observations. This is the period that
the country switched from ?xed to ?oating exchange rate regime. The stock market
return (r
1
) is calculated by the percentage change of stock market index (or SP) while
the exchange rate return (r
2
) is the percentage change of the nominal bilateral exchange
rate. Expected depreciation of domestic currency can discourage foreign investors to
invest in domestic stocks because loss from exchange rate will be materialized when
they sell those stocks for capital gain. Expected appreciation of domestic currency will
do the opposite. Therefore, exchange rate movements can generate the negative or
positive exchange rate return for foreign investors. The US dollar exchange rate series
is used because it can be a good representation of the price in the foreign exchange
market. The nominal effective exchange rate, which is the weighted average index
of various bilateral exchange rates might not be relevant to many foreign investors
who do not want to imitate the effective exchange rate for their currency portfolios.
Most investors only hold a single or few major currencies. In fact, the US dollar
exchange rate dominates all other major currencies. The large depreciation of Thai
bath against the US dollar at the beginning of the 1997 ?nancial crisis was the main
concern of importers, investors as well as policymakers.
In the ?rst step of the analysis, the data are transformed into logarithmic series.
These series are tested for unit roots using the Augmented Dickey-Fuller (ADF) and
Phillips and Perron (PP) tests with a constant only and a constant and a linear trend.
The optimal number of lags is chosen by AIC for the ADF tests while the optimal
bandwidth for the PP tests are chosen by Newey-West using Bartlett kernel spectral
estimation method. The results are reported in Table I.
For the log of nominal bilateral exchange rate series (LEX), all tests show that this
series is I(0), except for the PP test that includes a linear trend. The test results for the
Thai stock and
foreign exchange
markets
311
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
5
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
V
a
r
i
a
b
l
e
A
D
F
t
e
s
t
(
c
o
n
s
t
a
n
t
)
A
D
F
t
e
s
t
(
c
o
n
s
t
a
n
t
þ
t
r
e
n
d
)
P
P
t
e
s
t
(
c
o
n
s
t
a
n
t
)
P
P
t
e
s
t
(
c
o
n
s
t
a
n
t
þ
t
r
e
n
d
)
L
E
X
2
0
.
9
8
4
[
9
]
(
0
.
7
5
8
)
2
2
.
0
7
8
[
9
]
(
0
.
5
5
3
)
2
2
.
3
7
9
[
3
]
(
0
.
1
5
0
)
2
4
.
0
9
6
[
3
]
(
0
.
0
0
8
)
*
*
*
D
L
E
X
2
1
0
.
6
1
1
[
0
]
(
0
.
0
0
0
)
*
*
*
2
1
0
.
6
4
6
[
0
]
(
0
.
0
0
0
)
*
*
*
2
1
1
.
2
3
7
[
2
2
]
(
0
.
0
0
0
)
*
*
*
2
1
1
.
6
5
5
[
2
3
]
(
0
.
0
0
0
)
*
*
*
L
S
P
2
1
.
5
1
3
[
0
]
(
0
.
5
2
5
)
2
3
.
2
5
5
[
4
]
(
0
.
0
7
8
)
*
2
1
.
6
6
8
[
2
]
(
0
.
4
4
5
)
2
3
.
4
3
6
[
3
]
(
0
.
0
5
0
)
*
*
D
L
S
P
2
1
1
.
6
2
6
[
0
]
(
0
.
0
0
0
)
*
*
*
2
1
1
.
6
1
1
[
0
]
(
0
.
0
0
0
)
*
*
*
2
1
1
.
6
1
2
[
4
]
(
0
.
0
0
0
)
*
*
*
2
1
1
.
6
1
2
[
0
]
(
0
.
0
0
0
)
*
*
*
N
o
t
e
s
:
S
i
g
n
i
?
c
a
n
t
a
t
:
*
1
0
,
*
*
5
a
n
d
*
*
*
1
p
e
r
c
e
n
t
l
e
v
e
l
s
,
r
e
s
p
e
c
t
i
v
e
l
y
;
L
E
X
s
t
a
n
d
s
f
o
r
l
o
g
o
f
e
x
c
h
a
n
g
e
r
a
t
e
w
h
i
l
e
L
S
P
s
t
a
n
d
s
f
o
r
l
o
g
o
f
s
t
o
c
k
p
r
i
c
e
s
;
t
h
e
n
u
m
b
e
r
i
n
b
r
a
c
k
e
t
s
i
s
t
h
e
o
p
t
i
m
a
l
l
a
g
l
e
n
g
t
h
d
e
t
e
r
m
i
n
e
d
b
y
A
I
C
f
o
r
t
h
e
A
D
F
t
e
s
t
s
a
n
d
t
h
e
o
p
t
i
m
a
l
b
a
n
d
w
i
d
t
h
d
e
t
e
r
m
i
n
e
d
b
y
t
h
e
B
a
r
t
l
e
t
t
k
e
r
n
e
l
f
o
r
t
h
e
P
P
t
e
s
t
s
;
t
h
e
n
u
m
b
e
r
i
n
p
a
r
e
n
t
h
e
s
e
s
i
s
t
h
e
p
-
v
a
l
u
e
p
r
o
v
i
d
e
d
b
y
M
a
c
K
i
n
n
o
n
(
1
9
9
6
)
Table I.
Unit root tests
JFEP
4,4
312
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
5
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
log of the stock price series (LSP) show that the series is I(0) by the PP test that include
a linear trend while other tests show that it is I(1). Therefore, it is not conclusive that
the series of dependent and independent variables are I(0) or I(1) because of the mixed
results of the unit root tests. However, both series are stationary in ?rst differences.
The second step is to test for cointegration between the SP and exchange rate series.
Based upon the results of unit root tests shown in Table I, the ARDL bounds testing
procedure speci?ed in equation (3) seems to be suitable for the analysis.
The grid search method is used to select p and q in equation (1). Starting from the
most parsimonious ARDL(1,1) and if the model does not show serial correlation at the
5 percent level using Lagrange multiplier (LM) serial correlation test, then the model is
suitable for testing for cointegration. However, if the serial correlation is present, the
number of lagged ?rst differences will increase. The search continues for all
combinations of p and q until a model that is free of serial correlation is detected.
The suitable ARDL( p, q) model with the lag order of p ¼ 1 for the dependent variable
(LSP) and the lag order of q ¼ 1 for the independent variable (LEX) is called the
“ARDL(1,1)” model. The ARDL(1,1) model is used to test for cointegrating relationship
between LSP and LEX because it is free of serial correlation with the x
2
(2) ¼ 2.609
and its p-value is 0.271. By adding lagged level of the pair of variables, the computed
F-statistic is 1.248. The critical values at the 5 percent level from Table CI(iii) case III in
Pesaran et al. (2001) is 5.73 for the upper bound and 4.94 for the lower bound. Since the
computed F-statistic is below the lower bound critical value, the null hypothesis of
no cointegration can be accepted. In other words, there is no long-run relationship
between SP and exchange rate. This result is consistent with the ?nding of Kown and
Shin (1999) who ?nd that there is no cointegration between stock market index and
foreign exchange rate in a bivariate framework.
The third step is to apply the non-causality test in a VAR model using level of each
pair of the series. The empirical results of the Toda-Yamamoto causality test between
nominal US dollar exchange rate and stock market index are reported in Table II.
Based upon the results of unit root test in Table I, the anticipated maximum order of
integration (d
max
) is one. Using AIC to determine the optimal lag length in the VAR
model, the lag length appears to be two. Therefore, the (k þ d
max
) order VAR is three.
The results in Table II showthat the null hypothesis that LEX does not cause LSP is
accepted, but the null hypothesis that LSPdoes not cause LEXis rejected at the 1 percent
Null hypothesis Modi?ed Wald statistic p-value
LEX does not cause LSP 4.355 (þ) 0.226
LSP does not cause LEX 14.118
*
(2) 0.003
Misspeci?cation tests for the VAR model
Test statistic p-value
JB 15.823 0.000
LM 1.555 0.817
WH 143.083 0.000
Notes: signi?cant at:
*
1 percent; LEX stands for log of exchange rate while LSP stands for log of
stock prices; (þ) indicates the positive sum of the coef?cients of lagged variables, which is positive
causation; JB is the Jarque-Bera statistic for testing the null hypothesis that the residuals are
multivariate normal; LM is the Lagrange multiplier test for serial correlation up to third order in the
residuals, and WH is the White heteroskedasticity test of the residuals
Table II.
Results of non-causality
test between
LEX and LSP
Thai stock and
foreign exchange
markets
313
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
5
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
level of signi?cance. The results indicate that a change of nominal exchange rate does
not cause SP to change even though the sum of the estimated coef?cients of all lagged
independent variables is positive. In the other direction, a rise in SP causes the nominal
exchange rate to fall. Therefore, there seems to be a positive unidirectional causality
between SP and nominal exchange rate.
Further tests are conductedto examine the misspeci?cation of the augmentedVAR(3)
model used in the analysis. In Table II, the Lagrangian multiplier (LM) test statistic
indicates the acceptance of the null hypothesis that there is no serial correlation in the
residuals up to the third order of lags. Additionally, the White heteroskedasticity test
shows that the null hypothesis of the presence of ARCH effect can be rejected at the
1 percent level of signi?cance. However, the Jarque-Bera ( JB) statistic shows that the
residuals are not multivariate normal. Thus, it cannot be concluded that the speci?ed
VAR(k þ d
max
) model is suitable for conducting the non-causality tests. In other words,
the results in Table II are not reliable.
The summary statistics show that both stock and exchange rate returns are
negatively skewed. Excess kurtosis is observed in the exchange rate return series. The
JB statistics reject the null hypothesis of normal distribution in both return series,
which indicates the suitability of a GARCH speci?cation. The causal relationships
between stock and exchange rate returns and their conditional variances, and
especially the impact of exchange rate risk on stock market return under the ?oating
regime are tested using a bivariate CCC-GARCH(1,1) model and the standard Granger
causality test. The CCC-bivariate GARCH model gives the results of stock return
volatility as well as exchange rate return volatility while the standard causality test
gives the causal relationships between the variables of interest.
The results of the estimate of the bivariate AR(p)-CCC-GARCH(1,1) model are
reported in Table III. Assuming constant conditional correlation, the model performs
quite well in the dataset. In addition, the threshold GARCH model proposed by Zakoian
(1994) is also estimated, but the asymmetry is not found. The estimated conditional
correlation is 20.531 which is signi?cant at the 1 percent level. This implies that the
two return processes are not independent. The estimated ARCH and GARCH
parameters are non-negative. In addition, the sum of the coef?cients of the two terms is
0.959 for the r
1
-equation, and 0.815 with insigni?cant GARCH term in the r
2
-equation.
This indicates the stationary conditional variance series. The system is also free of
serial correlation. The multivariate Granger causality test is thus performed on
stationary series, and the results are reported in Table IV.
The results show some crucial ?ndings. First, an increase in stock return seems to
cause exchange rate return to fall, but the result is not signi?cant. Therefore, there is no
evidence of the impact of stockmarket returnonexchange rate. Second, a rise inexchange
rate return signi?cantly causes stock return to rise. Thus, there exists unidirectional
causalityinthe meanequations. Third, volatility causations betweenthe two markets are
observed. An increase in stock return volatility signi?cantly causes exchange rate return
volatility to rise while an increase in exchange rate return volatility causes stock return
volatility to fall. In causality sense, there is bidirectional causality in conditional
variances, but in different directions. In other words, an increase in the risk in the foreign
exchange market causes the risk in the stock market to decrease. However, an increase in
the risk in the stock market causes the risk in the foreign exchange market to rise.
The interesting ?nding here is that the rising risk in the foreign exchange market causes
JFEP
4,4
314
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
5
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
P
a
n
e
l
A
:
s
t
o
c
k
r
e
t
u
r
n
e
q
u
a
t
i
o
n
r
1
,
t
¼
2
2
.
4
9
9
*
*
*
þ
0
.
1
7
0
*
r
1
,
t
2
1
þ
0
.
0
5
8
r
1
,
t
2
2
þ
0
.
1
3
7
r
1
,
t
2
3
þ
0
.
0
1
1
r
1
,
t
2
4
2
0
.
5
7
0
*
*
r
2
,
t
2
1
þ
0
.
3
8
3
r
2
,
t
2
2
(
2
2
.
5
1
7
)
(
1
.
8
0
3
)
(
0
.
6
5
9
)
(
1
.
4
3
0
)
(
0
.
1
2
4
)
(
2
2
.
5
1
0
)
(
1
.
2
7
1
)
(
t
-
s
t
a
t
i
s
t
i
c
i
n
p
a
r
e
n
t
h
e
s
e
s
)
h
1
,
t
¼
2
.
7
9
7
þ
0
.
1
8
0
*
*
*
1
21
,
t
2
1
þ
0
.
7
7
9
*
*
*
h
1
,
t
2
1
(
1
.
2
2
2
)
(
2
.
5
9
5
)
(
1
1
.
3
4
6
)
(
t
-
s
t
a
t
i
s
t
i
c
i
n
p
a
r
e
n
t
h
e
s
e
s
)
P
a
n
e
l
B
:
e
x
c
h
a
n
g
e
r
a
t
e
c
h
a
n
g
e
e
q
u
a
t
i
o
n
r
2
,
t
¼
2
0
.
3
1
6
þ
0
.
1
1
0
r
2
,
t
2
1
þ
0
.
0
3
1
r
2
,
t
2
2
2
0
.
0
4
8
*
*
*
r
1
,
t
2
1
þ
0
.
0
2
4
r
1
,
t
2
2
þ
0
.
0
0
9
r
1
,
t
2
3
þ
0
.
0
3
1
r
1
,
t
2
4
(
2
1
.
6
1
6
)
(
0
.
5
6
7
)
(
0
.
3
5
2
)
(
2
2
.
8
2
0
)
(
1
.
3
7
7
)
(
0
.
5
6
7
)
(
0
.
3
5
2
)
(
t
-
s
t
a
t
i
s
t
i
c
i
n
p
a
r
e
n
t
h
e
s
e
s
)
h
2
,
t
¼
1
.
4
3
8
*
*
*
þ
0
.
7
4
1
*
*
*
1
22
,
t
2
1
þ
0
.
4
7
3
h
2
,
t
2
1
(
3
.
9
4
1
)
(
3
.
2
2
7
)
(
0
.
0
7
4
)
(
t
-
s
t
a
t
i
s
t
i
c
i
n
p
a
r
e
n
t
h
e
s
e
s
)
P
a
n
e
l
C
:
c
o
r
r
e
l
a
t
i
o
n
c
o
e
f
?
c
i
e
n
t
h
1
2
,
t
¼
2
0
.
5
3
1
*
*
*
(
h
1
,
t
)
1
/
2
(
h
2
,
t
)
1
/
2
(
2
3
.
8
1
6
)
(
t
-
s
t
a
t
i
s
t
i
c
i
n
p
a
r
e
n
t
h
e
s
e
s
)
P
a
n
e
l
D
:
s
y
s
t
e
m
d
i
a
g
n
o
s
t
i
c
t
e
s
t
Q
(
4
)
¼
6
.
1
7
5
,
Q
(
8
)
¼
1
8
.
3
7
5
5
(
0
.
9
8
6
)
(
0
.
9
7
4
)
(
p
-
v
a
l
u
e
i
n
p
a
r
e
n
t
h
e
s
e
s
)
N
o
t
e
s
:
S
i
g
n
i
?
c
a
n
t
a
t
:
*
1
0
,
*
*
5
a
n
d
*
*
*
1
p
e
r
c
e
n
t
l
e
v
e
l
,
r
e
s
p
e
c
t
i
v
e
l
y
;
r
1
i
s
t
h
e
m
o
n
t
h
l
y
s
t
o
c
k
r
e
t
u
r
n
c
a
l
c
u
l
a
t
e
d
f
r
o
m
t
h
e
r
a
t
e
o
f
c
h
a
n
g
e
i
n
t
h
e
s
t
o
c
k
m
a
r
k
e
t
i
n
d
e
x
a
n
d
r
2
i
s
t
h
e
m
o
n
t
h
l
y
e
x
c
h
a
n
g
e
r
a
t
e
r
e
t
u
r
n
c
a
l
c
u
l
a
t
e
d
f
r
o
m
t
h
e
r
a
t
e
o
f
c
h
a
n
g
e
i
n
t
h
e
n
o
m
i
n
a
l
b
i
l
a
t
e
r
a
l
e
x
c
h
a
n
g
e
r
a
t
e
;
h
1
i
s
t
h
e
c
o
n
d
i
t
i
o
n
a
l
v
a
r
i
a
n
c
e
o
f
s
t
o
c
k
r
e
t
u
r
n
,
h
2
i
s
t
h
e
c
o
n
d
i
t
i
o
n
v
a
r
i
a
n
c
e
o
f
e
x
c
h
a
n
g
e
r
a
t
e
r
e
t
u
r
n
,
a
n
d
h
1
2
i
s
t
h
e
c
o
n
d
i
t
i
o
n
a
l
c
o
v
a
r
i
a
n
c
e
o
f
s
t
o
c
k
r
e
t
u
r
n
a
n
d
e
x
c
h
a
n
g
e
r
a
t
e
r
e
t
u
r
n
;
t
h
e
l
a
g
s
o
f
r
1
a
n
d
r
2
a
r
e
c
h
o
s
e
n
i
n
s
u
c
h
a
w
a
y
t
h
a
t
t
h
e
s
y
s
t
e
m
d
i
a
g
n
o
s
t
i
c
t
e
s
t
s
h
o
w
n
o
a
u
t
o
c
o
r
r
e
l
a
t
i
o
n
i
n
t
h
e
r
e
s
i
d
u
a
l
s
i
n
t
h
e
m
e
a
n
e
q
u
a
t
i
o
n
s
;
Q
(
k
)
i
s
t
h
e
B
o
x
-
P
i
e
r
c
e
s
t
a
t
i
s
t
i
c
f
o
r
t
h
e
r
e
s
i
d
u
a
l
s
Table III.
Estimates of the bivariate
AR(p)-CCC-GARCH(1,1)
model of stock market
and exchange rate
returns for the period July
1997 to June 2010
Thai stock and
foreign exchange
markets
315
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
5
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
the stock market return to fall, which is the expected phenomenon under the ?oating
regime. In the events of rising exchange rate uncertainty, foreign investors might
diversifyawayfromthe domestic stocks. However, a rise inthe stockmarket riskdoes not
impose any impact on the exchange rate return. In fact, the Thai stock market is a small
emergingmarket that shouldnot in?uence the foreignexchange market. The dilemmafor
policymakers is that stabilizing the exchange rate will lead to higher level of uncertainty
in the stock market, which will lower the rates of return of domestic stocks. The Bank of
Thailand occasionally intervenes in the foreign exchange market when the exchange
rate is out of line. This practice is used to maintain the position of trade balance of the
countries. Even though the intervention by the Bank of Thailand can stabilizes the
exchange rate, the rising trend in volatility of stock return and the lower stock return
are unavoidable. One of the main ?nding from the present study is from the result of a
positive spillover of risk fromthe stockmarket to the risk in the foreign exchange market.
This evidence gives roomfor the Security Exchange Commission to set proper rules and
regulations for investors so as to mitigate ?uctuations of SPandthus stockmarket return.
As a result, the risk in the foreign exchange market will be reduced. However, these
measures might not be successful if there are excessive speculations in the stock market.
4. Concluding remarks
This study employs the recently developed time series analysis techniques to explore the
causal relationship between SP and exchange rates in an emerging market economy,
namely Thailand. The results from bounds testing for cointegration show that there
is no long-run relationship between SP (stock market index) and exchange rate in a
bivariate framework, which is consistent with the results of other empirical studies.
An alternative approach is to examine the causal links between SP and exchange rate.
The non-causality test is used and the results show that there exists unidirectional
causality between SP and exchange rates. However, the results from non-causality test
should not be reliable because the estimated VAR model does not pass the diagnostic
test of multivariate normality in the residuals.
The results from the two-step approach show that unidirectional causality between
stock and exchange rate return is observed. An increase in the exchange rate return
causes stock return to rise. More importantly, a rising trend in the risk in the foreign
exchange market causes stock return to fall. Both domestic and foreign investors
Hypothesis F-statistic p-value
r
1
does not cause r
2
0.077 (2) 0.630
r
2
does not cause r
1
3.481
* *
(þ) 0.001
h
1
does not cause r
2
1.255 (þ) 0.273
h
2
does not cause r
1
4.081
* *
(2) 0.000
h
1
does not cause h
2
2.238
*
(þ) 0.029
h
2
does not cause h
1
3.784
* *
(2) 0.001
Notes: Signi?cant at:
*
5 and
* *
1 percent level, respectively; r
1
is the monthly stock return calculated
from the rate of change in the stock market index and r
2
is the monthly exchange rate return calculated
from the rate of change in the nominal bilateral exchange rate; h
1
is the conditional variance of stock
return, h
2
is the condition variance of exchange rate return; the optimal lag length determined by AIC
is eight; (þ) denotes positive causation and (2) denotes negative causation
Table IV.
Results from multivariate
Granger causality test
JFEP
4,4
316
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
5
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
should be aware of the risk or uncertainty in the foreign exchange market because it
can cause their portfolio return to fall. For policymakers, reducing exchange rate risk
cannot be done without the associated costs from a rising risk in the stock market.
Instead, some measures that can reduce the stock market risk are bene?cial to the Thai
economy in that the foreign exchange market risk will be reduced. The reservation for
this suggestion is that the measures might not be successful in the presence of
excessive speculations of investors in the stock market.
References
Abballa, I.S.A. and Murinde, V. (1997), “Exchange rate and stock price interactions in emerging
?nancial markets: evidence on India, Korea, Pakistan, and Philippines”, Applied Financial
Economics, Vol. 7 No. 1, pp. 25-35.
Aggarwal, R. (1981), “Exchange rates and stock prices: a study of US capital market under
?oating exchange rates”, Akron Business and Economic Review, Vol. 12 No. 3, pp. 7-12.
Ajayi, R.A. and Mougoue, M. (1996), “On the dynamic relation between stock prices and
exchange rates”, Journal of Financial Research, Vol. 19 No. 2, pp. 193-207.
Atindehou, R.B. and Gueyie, J.-P. (2001), “Canadian chartered bank stock returns and exchange
rate risk”, Management Decision, Vol. 39 No. 4, pp. 285-95.
Bahmani-Oskooee, M. and Sohabian, A. (1992), “Stock prices and the effective exchange rate of
the dollar”, Applied Economics, Vol. 24 No. 4, pp. 459-64.
Bollerslev, T. (1990), “Modelling the coherence in short-run nominal exchange rates:
a multivariate generalized ARCH model”, Review of Economics and Statistics, Vol. 72
No. 3, pp. 498-505.
Booth, A. and Rotenberg, W. (1990), “Assessing foreign exchange exposure: theory and
applications using Canadian ?rms”, Journal of International Financial Management and
Accounting, Vol. 2, pp. 1-22.
Branson, W.H. and Henderson, D.W. (1985), “The speci?cation and in?uence of assets markets”,
in Jones, R.W. and Kenen, P.B. (Eds), Handbook of International Economics, Vol. 2,
Elsevier, Amsterdam.
Chen, J., Naylor, M. and Lu, X. (2004), “Some insights into the foreign exchange pricing puzzle:
evidence from a small open economy”, Paci?c-Basin Finance Journal, Vol. 12 No. 1,
pp. 41-64.
Choi, D.F. and Fang, F. (2009), “Volatility spillover between New Zealand stock market returns
and exchange rate changes before and after the 1997 Asian ?nancial crisis”, Asian Journal
of Finance and Accounting, Vol. 1 No. 2, pp. 106-17.
Choi, J.J. and Prasad, A.M. (1995), “Exchange rate sensitivity and its determinants: a ?rm
and industry analysis of US multinationals”, Financial Management, Vol. 24 No. 3,
pp. 77-88.
Dornbusch, R. and Fisher, S. (1980), “Exchange rates and the current account”, American
Economic Review, Vol. 70 No. 5, pp. 960-71.
Engle, R.F. and Granger, C.W.J. (1987), “Co-integration and error correction: representation,
estimation and testing”, Econometrica, Vol. 55 No. 2, pp. 1057-72.
Fountas, S., Karanasos, M. and Kim, J. (2006), “In?ation uncertainty, output growth uncertainty,
and macroeconomic performance”, Oxford Bulletin of Economics and Statistics, Vol. 68
No. 3, pp. 319-43.
Frank, P. and Young, A. (1972), “Stock price reaction of multinational ?rms to exchange
realignments”, Financial Management, Vol. 1 No. 3, pp. 66-73.
Thai stock and
foreign exchange
markets
317
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
5
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Giovannini, A. and Jorion, P. (1987), “Interest rate and risk premia in the stock market and in the
foreign exchange market”, Journal of International Money and Finance, Vol. 6 No. 1,
pp. 107-23.
Granger, C.W.J. (1969), “Investigating causal relations by econometrics models and cross spectral
methods”, Econometrica, Vol. 37 No. 3, pp. 424-38.
Granger, C.W.J. (1988), “Some recent developments in a concept of causality”, Journal of
Econometrics, Vol. 39 Nos 1/2, pp. 199-211.
Granger, C.W.J., Huang, B.N. and Yang, C.W. (2000), “A bivariate causality between stock prices
and exchange rates: evidence from recent Asian ?u”, Quarterly Review of Economics and
Finance, Vol. 40 No. 3, pp. 337-54.
Kanas, A. (2000), “Volatility spillover from stock returns and exchange rate changes:
international evidence”, Journal of Business Finance and Accounting, Vol. 27 Nos 3/4,
pp. 447-67.
Kim, K. (2003), “Dollar exchange rate and stock price: evidence from multivariate cointegration
and error correction model”, Review of Financial Economics, Vol. 12 No. 3, pp. 301-13.
Kown, C.S. and Shin, T.S. (1999), “Cointegration and causality between macroeconomic variables
and stock market returns”, Global Finance Journal, Vol. 10 No. 1, pp. 78-81.
MacDonald, R. and Taylor, M. (1992), “Exchange rate economics: a survey”, IMF Staff Papers,
Vol. 39 No. 1, pp. 1-57.
MacKinnon, J.G. (1996), “Numerical distribution functions for unit root and cointegration tests”,
Journal of Applied Econometrics, Vol. 11 No. 6, pp. 601-18.
Ma, C.K. and Kao, G.W. (1990), “On exchange rate changes and stock price reactions”, Journal of
Business Finance and Accounting, Vol. 17 No. 3, pp. 441-9.
Mansor, H.I. (2000), “Cointegration and Granger causality tests of stock price and exchange rate
interactions in Malaysia”, ASEAN Economic Bulletin, Vol. 17 No. 1, pp. 36-47.
Nabiha, N. and Mounira, B.A. (2009), “Interaction between stock and exchange crisis in emerging
markets”, in Jay Choi, J. and Papaioannou, M.G. (Eds), Credit, Currency, or Derivatives:
Instruments of Global Financial Stability or Crisis?, International Finance Review, Vol. 10,
Emerald, Bradford, pp. 171-89.
Narayan, P.K. (2009), “On the relationship between stock prices and exchange rate for India”,
Review of Paci?c Basin Financial Markets and Policies, Vol. 12 No. 2, pp. 289-308.
Nieh, C. and Lee, C. (2001), “Dynamic relationship between stock prices and exchange rates for
G-7 countries”, Quarterly Review of Economics and Finance, Vol. 41 No. 4, pp. 477-90.
Pan, M., Fok, R.C. and Liu, Y.A. (2007), “Dynamic linkages between exchange rates and stock
prices: evidence from east Asian markets”, International Review of Economics and
Finance, Vol. 16 No. 4, pp. 503-20.
Pesaran, M.H., Shin, Y. and Smith, R.J. (2001), “Bounds testing approaches to the analysis of level
relationship”, Journal of Applied Econometrics, Vol. 16 No. 3, pp. 289-326.
Phylaktis, K. and Ravazzolo, F. (2005), “Stock prices and exchange rate dynamics”, Journal of
International Money and Finance, Vol. 24 No. 7, pp. 1031-53.
Rambaldi, A.N. and Doran, H.E. (1996), “Testing for Granger non-causality in cointegrated
system made easy”, Working Papers in Econometrics and Applied Statistics, No. 88,
University of New England.
Soenen, L.A. and Hennigar, E.S. (1988), “An analysis of exchange rates and stock prices: the US
experience between 1980 and 1986”, Arkron Business and Economic Review, Vol. 19 No. 4,
pp. 7-16.
JFEP
4,4
318
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
5
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Stavarek, D. (2005), “Stock prices and exchange rates in the EU and the United States: evidence
on their mutual interactions”, Czech Journal of Economics and Finance, Vol. 55 Nos 3/4,
pp. 141-61.
Toda, H.Y. and Yamamoto, T. (1995), “Statistical inferences in vector autoregressions with
possibly integrated process”, Journal of Econometrics, Vol. 66 Nos 1/2, pp. 225-50.
Yu, Q. (1997), “Stock prices and exchange rates: experience in leading East Asian ?nancial
centres: Tokyo, Hong Kong and Singapore”, Singapore Economic Review, Vol. 41 No. 1,
pp. 47-56.
Zakoian, J.M. (1994), “Threshold heteroskedastic models”, Journal of Economic Dynamic and
Control, Vol. 18 No. 5, pp. 931-55.
About the author
Komain Jiranyakul is an Associate Professor at National Institute of Development
Administration, Bangkok, Thailand. He teaches macroeconomics and ?nancial economics.
He recently published papers in Journal of The Asia Paci?c Economy, Economics Bulletin, Journal
of Asian Economics, and Asian Economic Journal. Komain Jiranyakul can be contacted at:
[email protected]
Thai stock and
foreign exchange
markets
319
To purchase reprints of this article please e-mail: [email protected]
Or visit our web site for further details: www.emeraldinsight.com/reprints
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
5
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
This article has been cited by:
1. Arif Billah Dar, Aasif Shah, Niyati Bhanja, Amaresh Samantaraya. 2014. The relationship between stock
prices and exchange rates in Asian markets. South Asian Journal of Global Business Research 3:2, 209-224.
[Abstract] [Full Text] [PDF]
2. Walid M. A. Ahmed. 2014. Dynamic interactions between Egyptian equity and currency markets prior
to and during political unrest. Applied Financial Economics 1-13. [CrossRef]
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
5
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
doc_389299362.pdf