Jurnal Akuntansi-B.Inggris

The Asean Stock Market Integration:
The Effect of the 2007 Financial Crisis on the Asean Stock Indices’
Movements
Adwin Surja Atmadja
Faculty of Economics, Petra Christian University
E-mail: aplin@peter.petra.ac.id
ABSTRACT
This study attempts to examine the existence of cointegration relationship and the short
run dynamic interaction among the five ASEAN stock market indices in the period of before
and during the 2007 financial crisis. The multivariate time series analysis frameworks are
employed to the series in both sub-sample periods in order to answer the hypotheses.The
study finds two cointegrating vectors in the series before the financial crisis period, however
it fails to detect any cointegrating vector in the period of financial crisis. Granger causality
tests applied to the series reveal that number of significant causal linkages between two
variables increase during the crisis period. Moreover, the accounting innovation analysis
shows an increase in the explanatory power of an endogenous variable to another within the
system during the crisis period, indicating that the contagious effect of the 2007-US financial
crisis has entered into the ASEAN capital market, and significantly influenced the regional
indices’ movements.
Keywords: ASEAN, stock market integration, the 2007 financial crisis, regional indices’
movements.
INTRODUCTION
Liberalization of the five ASEAN (Indonesia,
Malaysia, the Philippines, Singapore, and
Thailand) financial markets in 1980s resulted in
enormous capital inflows to this region. By opening
their national borders for foreign investors, the
countries’ financial markets were overwhelmed by
foreign capital in both foreign direct and portfolio
investments giving significant support to their
rapid domestic economic development, as well as
enjoyed rapid financial markets expansion in the
beginning of 1990s. Capital inflows have been
crucial to the rapid – sustained growth in ASEAN
countries (Sachs and Larrain, 1993:577) at that
time, since domestic saving, as commonly in
developing countries, had little role as development
funding.
Triggered by the sharp depreciation of the
Thai baht in the midst of 1997, the disastrous
effects of the 1997 financial crisis were broadly
spread out to the countries’ financial markets
which were dominated by bank loan and portfolio
investment, not by foreign direct investment
(DFAT, 1999:29). The crisis then extensively
affected the world financial markets through its
contagion effects. Market capitalization of the
countries’ stock market was largely contracted due
to a deep depreciation in their stock prices causing
their stock indices then sharply plunged.
However, the downturn in the five ASEAN
rebounded in 1999. After the sharp output
contraction in 1998, growth returned in that year
as depreciated currencies spurred higher exports
(Krugman and Obstfeld, 2003:693). Following the
appreciation of regional currencies in the second
semester of the year, the regional capital and
financial markets started to recover. The regional
stock market indices increased around 42.46% on
average compared to those from two years before
(calculated from IFS 2004). This might indicate
that investors’ confidence started to recover and
they began to invest in the five ASEAN.
During ten years after, the ASEAN’s
economies steadily grew to their new equilibrium.
As a market indicator, the ASEAN capital market
indices apparently fluctuated in a relatively narrow
range dominantly due to small internal shocks in
the short run, but stably moved with positive
trends in the long run. These all mirror that the
ASEAN markets were relatively stable during the
time periods, and their economies were just on the
right tracks.
However, in the second semester of 2007 the
countries experienced significant shocks in their
capital markets due to a contagious effect of the US
financial market turmoil. At the time, the US
financial market deeply suffered from the most
1
JURNAL AKUNTANSI DAN KEUANGAN, VOL. 11, NO. 1, MEI 2009: 1-12
2
significant economic shocks initiated by the subprime
mortgage crisis leading to the downturn in
housing market, and then worsened by the spike in
commodity prices (Yellen 2008:1). The devastating
effects of the 2007 financial crisis in the US then
widely spread throughout the world.
From the facts above, the 2007 financial crisis
may have significant consequences on the variation
of the countries’ stock indices that probably
different with those in non crisis era. The financial
crisis could possibly cause the regional indices
deviate from their long run equilibrium, and the
behaviour of the indices’ movements may be
different with those before. All possibilities may
happen in the regional market depended on how
significant the impact of the financial crisis hit the
market. Therefore, this study will empirically
examine how the 2007 financial crisis has taken
into effect on the five ASEAN stock indices’
movements. To be more specific, this study
attempts to observe the existing of cointegrating
relationships among the five ASEAN stock indices
in the periods of before (pre) and during the 2007
financial crisis in order to portray the long run
interrelations among the indices in the both
periods. The aim is also to answer how and to what
extent the stock indices dynamically interact with
each other in the short run during the given
periods.
CONCEPT OF FINANCIAL MARKET OR
STOCK MARKET INTEGRATION
The basic theoretical concept of financial
market or stock market integration is adopted from
the law of one price. In integrated financial
markets, the assets with the same risk in different
markets will result in the same yield when
measured in a common currency (Stulz 1981:924-
5). However, if the yields are different across the
markets, the arbitrage process will play an
important role in eliminating the differences.
Operationally capital markets integration refers to
the extent that markets’ participants are enabled
and obligated to take notice of events occurring in
other markets by using all available information
and opportunities, while financial market
integration is defined in terms of price
interdependence between markets (Kenen 1976:9).
Moreover, stock market integration is affected by
some factors (Roca 2000:14), such as:
1. Economic integration, which means that the
more integrated the economies of countries, the
more integrated their equity markets (Eun and
Shim 1989: 256).
2. Multiple listing of stocks. This implies that a
shock in a particular stock market can be
transmitted to other stock market through
shares listed in both markets.
3. Regulatory and information barriers. The
higher the barriers, the lower the degree of
stock market integration.
4. Institutionalisation and securitisation. As
institutions are more willing to transfer funds
overseas to increase their diversification
opportunities, the integration will be promoted.
5. Market contagion. The prices between stock
markets can move together due to a contagion
effect (King and Wadwhani 1990:5), and this
contagion effect determines significantly the
dynamic relationships between international
stock markets (Climent and Meneu, 2003:111).
However, in emerging stock markets, this effect
might be smaller than what is widely perceived
(Pretorius 2002:103).
Much research has been done, mainly by using
a cointegration analytical framework, to find and
analyse the existence of integration in stock
market across countries. The results are different
depending on where, when, and how the research
has being conducted. The cointegration analytical
framework has been widely applied to examine the
integration of stock markets across countries. Once
a cointegration vector is found among two or more
stock markets, it indicates the existence of a long
run relationship among them. Thus, stock price
movements in one equity market will affect
another in other markets.
A research conducted by Chung and Liu
(1994:55) found two cointegration vectors between
the U.S and larger Asia Pacific stock markets.
Palac-McMiken (1997:299) also reveals the
existence of cointegration in ASEAN markets
(Malaysia, Singapore, Thailand, and the
Philippines), except Indonesia, during 1987 to
1995. Both results were confirmed by Masih and
Masih (1999:275) who report that some of ASEAN
countries (Thailand, Malaysia, and Singapore)
have a high degree of interdependence with other
Asian (Hong Kong and Japan) and developed (the
U.S. and the U.K.) stock markets. Furthermore,
they also find one cointegration vector among
several major Asian stock markets (Hong Kong,
Korea, Singapore, and Taiwan) and major
developed markets (Masih and Masih 2001: 580-1).
Interestingly, Pretorius (2002:103) reports that
the degree of bilateral trade and the industrial
production growth differential significantly
explained the correlation between two equity
markets, and that the stock markets of countries in
the same region are more interdependent than
those in different regions. Consistent with this
finding, Roca (2000:145) finds the existence of
Atmadja: The Asean Stock Market Integration
3
interdependency among all the ASEAN stock
markets in the short run. However, in contrast to
short run interdependency, he indicates that there
was no cointegration among ASEAN countries as a
group during 1988-1995 and that those stock
markets were not significantly related to each
other in the long run.
Chan, Gup and Pan (1992:289) and DeFusco,
Geppert and Tsetsekos (1996:343) also mention
that there is no cointegration between the U.S and
several Asian emerging stock markets (Hong Kong,
Taiwan, Singapore, Korea, Malaysia, Thailand,
and the Philippines) in the 1980s and early 1990s.
However, these findings somewhat contradicts
with those of Chung et al. (1994) and Masih et al.
(1999). This then implies that the interdependence
among stock markets is not stable over time. For
example, Hung and Cheung (1995:286) assert that
there is no cointegration among stock markets in
some Asia-Pacific countries (Malaysia, Hong Kong,
Korea, Singapore, and Taiwan). However, when
they used US dollar denominated stock prices, it
was reported that those stock markets were
cointegrated after, but not before, the 1987 stock
crash.
Arshanapalli and Doukas (1993:206) also
mention the instability of stock market
interdependence when they tested the effect of
inclusion or omission of the data for the 1987 crisis
and revealed that that it affects the results. They
conclude that the stock markets were highly
integrated during the crisis. Furthermore,
Arshanapalli, Doukas and Lang (1995:72) show
that after the 1987 crisis the stock markets in
emerging markets (Malaysia, the Philippines, and
Thailand) and developed markets (Hong Kong,
Singapore, the U.S., and Japan) are more
interdependent as they found cointegration in the
post-crisis period, but not in the pre-crisis period.
Other researchers, Liu, Pan and Shieh (1998: 59)
also confirm that there is an increase in the
interdependence within Asian-Pacific regional
markets and the stock markets in general post-the
1987 crisis. Similarly, Sheng and Tu (2000:245)
document one cointegration vector between the
U.S. and several Asian stock markets (Taiwan,
Malaysia, China, Thailand, Indonesia, South
Korea, the Philippines, Australia, Japan, Hong
Kong, and Singapore) during the crisis, but none in
the year before the crisis, when they observed the
stock markets using daily data.
Finally, a research recently conducted by
Yang, Kolari and Min (2003:478) examined the
long-run relationship and short-run dynamic
causal linkages among the U.S, Japanese, and ten
Asian emerging markets using daily data of 1997-
1998 periods. They confirm that the stock markets
of those countries have been more integrated after
the 1997 Asian financial crisis than before the
crisis. Both long-run cointegration relationship and
short-run causal linkages among those markets
become more significant during the crisis. These
findings also confirm that the degree of integration
among those countries tends to change over time.
Several points that may be drawn form the
literature review. The implication is that
liberalization of the financial sector in many
countries has caused world or regional stock
markets to be more integrated. Empirical evidence
is given by the presence of cointegration vectors
and significant short-run causal linkages. It is
worth noting that the stock markets of countries in
the same region may be more interdependent than
those in different regions.
RESEARCH METHODOLOGY
Basically, a stock market price index or stock
market index is a portfolio of individual stocks. The
index level corresponds to some average of the
price levels of individual shares. Changes in the
index level give rise to market returns. Thus, the
stock market index, which can be viewed simply as
a portfolio of shares, can commonly be use as an
indicator of the market performance. There are
several factors that determine the level of the
index, such as breadth of index, weighting system,
capitalization adjustment, and dividend effect
(Brailsford Heaney and Bilson 2004:68).
The stock market index of a country may also
be an indicator of short-term portfolio investment
movement in the country. An upward trend of a
stock market index means that there is an increase
in demand of the listed shares in the market. This
indicated that investors are attracted to buy shares
and invest their fund in the country. On the other
hand, a downward trend movement of a stock
market index indicates that the investors are
unlikely to continuously hold the listed shares.
Hence, stock market movements may reflect the
attractiveness of a country for investments,
especially for portfolio investments.
In this study, the daily closing stock price
indices of the five ASEAN countries, which are
Jakcomp of Indonesia; KLSE of Malaysia; PSEi of
the Philippines; STI of Singapore; and SET
Composite of Thailand, are employed as
measurement of the countries’ daily stock index
movements in the periods of before and during the
2007 financial crisis.
Some previous research (Arshanapalli et al,
1993, Chung et al, 1994, Arshanapalli et al, 1995,
Liu et al,1998, Masih et al., 1999, Masih et al,
2001) document that stock markets in the Asian
JURNAL AKUNTANSI DAN KEUANGAN, VOL. 11, NO. 1, MEI 2009: 1-12
4
region are interdependent not only among
themselves, but also with some of the developed
market. Furthermore, those stock markets are
even more interdependent during and after the
financial crisis (Sheng et al 2000; Yang et al 2003)
In the case of the ASEAN, Palac-McMiken
(1997:299) reports the existence of cointegration in
the countries’ stock markets, except Indonesia,
before the 1997 crisis. Yang et al (2003:478)
confirm that both long-run cointegration
relationship and short-run causal linkages among
those markets become more significant during the
crisis period. In contrast, Roca (2000:145) finds the
existence of interdependency among the five
ASEAN’s stock markets in the short run, but not
significantly related in the long run before the 1997
crisis.
Based on these findings, it is hypothesized that
the ASEAN stock indices would have long run
cointegration relationship and short run dynamic
interaction, and that the relationship and the
interaction would be more significant during the
2007 financial crisis.
All daily price index data of the five ASEAN
during the observation periods are obtained from
the Thomson Financial. The index data of all
variables then will be transformed into natural
logarithm forms before conducting the analyses.
In order to examine the movements of the
indices in both periods, the data are then separated
into two sub-sample periods, which are the periods
of: 1) Before the 2007 financial crisis (pre crisis),
which cover the period of Jan 2000 – June 2007, 2)
During the 2007 financial crisis, which cover the
period of July 2007 – May 2009, as it is stated in
several publications (http://en.wikipedia.org,www.
globalissues.org,www.atypon-link.com)
The two most appropriate models that one of
which may suitable for this study are VAR and
VECM. In the Vector autoregressive model (VAR)
all of the variables are endogenous, and
symmetrically treated. A VAR could be very large,
however the simplest VAR model, in standard
form, could be written as (Enders, 2004:265):
Yt = a10 + a11Yt-1 + a12 Zt-1 + eYt.
Zt = a20 + a21Yt-1 + a22 Zt-1 + εZt.
The VAR requires that all variables be
stationary and the appropriate lag length is data
driven (Brooks 2002:333). There are several
available tests for testing for a unit root, the most
common is the Augmented Dicky-Fuller (ADF)
test. Non-stationary variables may be made
stationary by differencing or detrending process.
To define the appropriate lag length, some
tests of information criteria that will be applied in
this study include the likelihood ratio test; Akaike
Information Criterion (AIC); and Schwarz
Bayesian Criterion (SC).
The likelihood ratio test is based on asymptotic
theory and is an F-type approximation. This test
actually compares a restricted VAR (less lags) to an
unrestricted VAR (more lags). Thus, the null
hypothesis of this test is that the restricted model
is correct. However, the shortcoming of this test is
that it may not be useful in small samples. In
addition, the likelihood ratio test is only valid when
the restricted model is tested (Enders 2004:283).
Because of the limitations of the likelihood
ratio test, multivariate generalization of AIC and
SC may be the most suitable alternatives. The
minimum values of AIC and/or SC may validly
indicate the appropriate lags length, as long as the
model’s residual has no serial correlation problem.
Otherwise, the lag length may be too short. Thus, it
is necessary to re-estimate the model using lag
length that yield serially uncorrelated (Enders
2004:338).
In VAR, a block causality test will be used to
examine whether the lags of one variable enter into
the equation for another variable (Enders
2004:283). A variable (y1) is said to be a grangercause
of another (y2) if the present value of y2 can
be predicted with greater accuracy by using past
values of y1, all other information being identical
(Thomas 1997:461). If y1 granger-causes y2, then
the parameters of lags of y1, βi’s, should not equal
zero in the equation of y2. However, it is worth
noting that granger-causality basically means a
correlation between the current value of one
variable and the past (lags) value of others. It does
not mean that movements of one variable
physically cause movements of another (Brooks,
2002:240). Granger causality simply implies a
chronological ordering of movements of the series.
Therefore, it could validly be stated that changes or
movements in one variable (y2) appear to lag those
of another (y1).
The alternative model that probably suitable
to be used is the vector error correction model
(VECM) or cointegration framework analysis,
which is basically is a VAR augmented by the error
correction term (êt-1). The simplest VECM, in
general, takes the form as (Enders 2004:329):
ΔYt = α10 + αY êt-1+ Σ α11(i) ΔYt-i + Σ α12(i) ΔZt-i + εYt.
ΔZt = α20 + αZ êt-1+ Σ α21(i) ΔYt-i + Σ α22(i) ΔZt-i + εYt.
where
êt-1 = (Yt-1 – β1Z1t-1)
Thus, if the parameters of error correction
term (ECT), called speed of adjustments (αY and αZ)
in VECM, are zero, then VECM reverts to a VAR
in first differences (Enders 2004:329).
ΔYt = α10 + Σ α11(i) ΔYt-i + Σ α12(i) ΔZt-i + εYt.
ΔZt = α20 + Σ α21(i) ΔYt-i + Σ α22(i) ΔZt-i + εYt.
However, if the speed of adjustments are not
zero, the larger the speed of adjustments, the
Atmadja: The Asean Stock Market Integration
5
greater the response to previous periods’ deviation
from the long run equilibrium. Thus, a
cointegration relationship is a long term or
equilibrium phenomenon, since it is possible that
cointegrating variables may deviate from their
relationship in the short run, but their association
would return in the long run. A principal feature of
cointegrated variable is that their time paths are
influenced by the extent of any deviation from long run
equilibrium. After all, if the system is to return to long
run equilibrium, the movements of at least some of the
variables must respond to the magnitude of the
disequilibrium. (Enders 2004:328). The VECM result is
also sensitive to its lags length. Thus, it is essential
to use appropriate lag length to get the appropriate
outcomes by conducting the lag order selection
criteria (LR, AIC, or SC) tests.
Unlike VAR, cointegration refers to a linear
combination of non-stationary variables. Thus, it is
necessary to test the existence of unit roots in
observed variables using the ADF test as it is used
in VAR. Cointegration also requires that all
variables in a model be integrated of the same
order. Thus, in order to test the existence of
cointegrated variable, one may use the Engle-
Granger (EG) test, which is a residuals-based
approach, or the Johansen Cointegration test. In
the case of a cointegration relationship does not
exist, a VAR analysis in first difference will then be
the correct specification to conduct the estimation
(Enders, 2004:287).
After estimating the VECM equations, the
VEC Pairwise Granger Causality / Block Exogenity
Wald Tests will be applied to reveal whether
changes in one variable cause changes in another.
If so, then lags of variable should be significant in
the equation for the other variable. If this is the
case, it can be said that the variable grangercauses
another.
A direct interpretation of the cointegration
relations may be difficult or misleading (Lutkepohl
and Reimers 1992:53, Runkle 1987:442). As in a
traditional VAR analysis, innovation accounting,
consist of Impulse Response and Variance
Decomposition Analyses, can provide a solution to
the interpretation problem, and might be the most
appropriate method to explain the short run
dynamic structure of market linkages (Yang et al
2003:479). The analysis would give to answers
whether changes in the value of a given variable
have positive or negative effect on other variables
in the system, or how long it would take for the
effect of that variable to work through the system
(Brooks 2002:341).
A shock to the i-th variable not only directly
affects the i-th variable but is also transmitted to
all of the other endogenous variables through the
dynamic (lag) structure of the VAR. An impulse
response function traces the effect of a one-time
shock to one of the innovations on current and
future values of the endogenous variables. In other
words, impulse response analysis will trace out the
responsiveness of the dependent variables in VAR
to shocks on individual error terms. In this paper,
the generalized type of impulse responses analysis
is employed as orthogonalized impulse responses is
sensitive to the ordering of the variable in the
system. The Generalized Impulses as described by
Pesaran and Shin (1998) constructs an orthogonal
set of innovations that does not depend on the VAR
ordering. The generalized impulse responses from
an innovation to the j-th variable are derived by
applying a variable specific Cholesky factor
computed with the j-th variable at the top of the
Cholesky ordering. Dekker, Sen and Young
(2001:31) found that the generalized approach
provided more accurate results than the traditional
orthogonalized approach for both impulse response
and forecast error variance decomposition analysis
Forecast error variance decomposition,
meanwhile, refers to the proportion of the
movements in a sequence due to its own shock
versus shocks to the other variables (Enders
2004:280). This analysis separates the variation in
an endogenous variable into the component shocks
to the system. Thus, the variance decomposition
provides information about the relative importance
of each random innovation in affecting the
variables in the system. It determines how much of
the s-step ahead forecast error variance of a given
variable is explained by innovations to each
explanatory variable. A shock to the i-th variable
will not only affect that variable, but also can be
transmitted to all of the other variables in the
system. To some extent, impulse responses and
variance decompositions offer very similar
information.
EMPIRICAL RESULTS
The Period of before the 2007 Financial
Crisis
The ADF test applied to all variables at level
within the sub-sample period results in acceptence
(fail to reject) of the null hypothesis that the serries
contain unit root. The existence of a unit root in
Asian stock markets, including the ASEAN is well
established in the literature (Masih et al 1999,
2001). The examination then continues to select
the appropriate lag order. The lag orders suggested
by the three lag order selection criteria result in
serially correlated residual. Therefore, as
mentioned by Enders (2004:338), it is necessary to
re-estimate the model using all possible lag length
JURNAL AKUNTANSI DAN KEUANGAN, VOL. 11, NO. 1, MEI 2009: 1-12
6
until the residual is found serially uncorrelated.
After examination of all possible lag length, the
appropriate lag length is found to be six.
The Johansen Cointegration test then reveals
that there are conflicting results between max
and trace statistic as it is stated in Table 1.
However, as it is suggested by some
econometricians (Johansen and Juselius 1990;
Kasa, 1992; and Serletis and King 1997) that the
trace tends to have more power than the max
because trace takes into account all degrees of
freedom (n-r) of the smallest eigenvalues, then the
number of cointegration vectors suggested by the
trace statistic would be employed. Thus, it may
be concluded that there are two cointegrating
vectors found in the series of the sub-sample period
at 5% level of significance, meaning that the
ASEAN indices are highly interdependent and
significantly related to each other in the long run
during the pre crisis period.
Table 1. The Johansen Cointegration Test For the
sub-sample period of before the 2007
financial Crisis
Trend assumption: Linear deterministic trend
Unrestricted Cointegration Rank Test
Hypothesized
No. of CE(s)
Eigenvalue Trace
Statistic
5 Percent
Critical
Value
1 Percent
Critical
Value
None ** 0.017748 83.02299 68.52 76.07
At most 1 * 0.010864 48.13862 47.21 54.46
At most 2 0.008270 26.85945 29.68 35.65
At most 3 0.004107 10.68322 15.41 20.04
At most 4 0.001368 2.667069 3.76 6.65
Trace test indicates 2 cointegrating equation(s) at the 5% level
Trace test indicates 1 cointegrating equation(s) at the 1% level
Hypothesized
No. of CE(s)
Eigenvalue Max-
Eigen
Statistic
5 Percent
Critical
Value
1 Percent
Critical
Value
None * 0.017748 34.88437 33.46 38.77
At most 1 0.010864 21.27917 27.07 32.24
At most 2 0.008270 16.17623 20.97 25.52
At most 3 0.004107 8.016154 14.07 18.63
At most 4 0.001368 2.667069 3.76 6.65
Max-eigenvalue test indicates 1 cointegrating equation(s) at the
5% level
Max-eigenvalue test indicates no cointegration at the 1% level
The existence of cointegrating vectors resulted
from this study is somewhat consistent with
previous research conducted by Palac-McMiken
(1997:299) and Liu et al (1998:59), but contradicts
with that of Sheng et al (2000:245), in different
period of time. Thus, it can be argued that VECM
is possible to be carried out to estimate the stock
indices of the five ASEAN.
The results of the VECM estimation can be
shown in the two consecutive tables. Table 2
(APPENDIX) shows the estimated cointegrating
vectors, whereas Table 3 report the coefficient of
speed of adjustment.
Table 2. Estimated Cointegrating Vectors
Cointegrating Eq: CointEq1 CointEq2
JAKCOMP 1.000000 0.000000
KLSE 0.000000 1.000000
PSE -2.101383 1.203789
(0.32567) (0.31491)
[-6.45239] [ 3.82264]
SET -0.420384 -0.438546
(0.09796) (0.09472)
[-4.29149] [-4.62993]
STI 1.353018 -2.018991
(0.37949) (0.36695)
[ 3.56532] [-5.50208]
C 1.355101 2.913097
Note: cointegration with unrestricted intercepts and no trends.
Included observations: 1948 after adjusting endpoints
Standard errors in ( ) & t-statistics in [ ]
Table 3. Speed of Adjustment Parameter of the
Error Correction Term (ECT)
Error
Correction:
JAKCOMP
KLSE PSE SET STI
ect1 (α1) -0.004776 5.52E-05 0.009661 0.001528 0.004420
(0.00244) (0.00164) (0.00238) (0.00263) (0.00211)
[-1.95436] [ 0.03365] [ 4.05722] [ 0.58073] [ 2.09935]
ect2 (α2) -0.005994 -0.003282 -0.004991 0.001391 0.005767
(0.00303) (0.00203) (0.00295) (0.00326) (0.00261)
[-1.97914] [-1.61487] [-1.69136] [ 0.42634] [ 2.20980]
Note : cointegration with unrestricted intercepts and no trends
Included observations: 1948 after adjusting endpoints
Standard errors in ( ) & t-statistics in [ ]
As a common practice, Table 2 (APPENDIX)
shows that the first cointegrating vector is
normalized by JAKCOMP, while KLSE is
restricted to zero. Meanwhile, in the second one,
KLSE is used to normalize, while JAKCOMP is
restricted to zero. Based on t-statistic at the 5%
level of significance, JAKCOMP, PSE, SET, and
STI are found significant in the first cointegration
vector, while KLSE, PSE, SET, and STI are
significant in the second one. This means that all of
the significant indices (variables) significantly
contribute to the ASEAN indices’ long run
equilibrium.
With the same critical value of 5%, the speed of
adjustment coefficient for the first and second
cointegrating vector, for KLSE and SET are
statistically zero. This implies that both vectors
have no contribution to the convergence of these
indices to their long run paths, although SET does
have significant influence on any of the
cointegrating relationship, and KLSE affects only
the second one.
Atmadja: The Asean Stock Market Integration
7
In contrast, the speed of adjustment of
JAKCOMP, PSE, and STI are statistically
significant in both vectors. JAKCOMP has
negative influences in both cointegrating
relationship indicating a downward long run
adjustment. Conversely, STI affects the vectors
positively implying an upward long run
adjustment. In the second cointegrating vector,
JAKCOMP will react to a disequilibrium among
KLSE, PSE, SET, and STI. Thus, the vector would
contribute to the convergence of JAKCOMP to its
long run path, even though the index does not have
any significant contribution to the others return to
the long run equilibrium. PSE interestingly has
positive and negative significant impact on the first
and the second cointegration vectors, respectively.
The implication is that PSE would react positively
in the first vector, and negatively in the second one.
The existence of the cointegrating relationship
in the region during the time period could be
caused by some reasons. First, the degree of
economic integration in the ASEAN countries has
risen after the 1997 financial crisis. The
information barriers have also significantly decline
as a result of technological advance in IT
(information technology) and in the markets’
trading operating systems. The other reason is that
the degree of institutionalization and securitization
have increased in the regional market promoting
intra-regional fund transfers to increase
diversification opportunities.
After the VECM estimation is determined, the
next step is to search the existence of granger
causality among variables of each model. The
results of VEC Pairwise Granger Causality Tests
for each country are presented Table 4. Using a 5%
level of significance, the table shows only four
significant causality linkages found among the
variables in the pre crisis period. It also reveals
that none of the other ASEAN indices is
significantly granger caused JAKCOMP during the
period, vice versa. Thus, it may be concluded that
movements of the index during the period
apparently become isolated from the influence of
the others. STI experienced almost the same
condition as JAKCOMP when all other ASEAN
indices do not granger cause the index. However,
somewhat different with JAKCOMP, STI, as well
as SET, granger cause (in uni-directional form)
KLSE meaning that movements in KLSE
appeared to lag those of STI and SET. Moreover,
SET also appears to have bi-directional causality
with PSE.
As a part of the Accounting Innovation
Analysis, the impulse response analysis traces out
the responsiveness of the dependent variable in the
system to shocks to each of the variables (Brooks,
2002:341). The generalized type of the impulse
response analysis will be applied in this study to
observe short run dynamic interactions among the
variables, since orthogonalized impulse responses
is sensitive to the ordering of the variable in the
system. The complete result of the analysis is
presented in Table 5.
Table 4. VEC Pairwise Granger Causality/Block
Exogeneity Wald Tests
Dependent
variable Exclude Chi-sq Prob.
JAKCOMP KLSE 6.158533 0.4057
PSE 10.79299 0.0950
SET 9.533360 0.1457
STI 7.470766 0.2795
KLSE JAKCOMP 4.013962 0.6748
PSE 11.10882 0.0851
SET 12.83167 0.0458
STI 18.49538 0.0051
PSE JAKCOMP 12.17061 0.0583
KLSE 4.755074 0.5756
SET 40.46320 0.0000
STI 10.91111 0.0912
SET JAKCOMP 4.591306 0.5972
KLSE 10.01266 0.1241
PSE 13.22751 0.0396
STI 4.343344 0.6303
STI JAKCOMP 12.14867 0.0587
KLSE 9.910047 0.1285
PSE 8.841288 0.1827
SET 9.852119 0.1310
Table 5. The Impulse Response to Generalized One
S.D. Innovations
Response
of
Period JAKCOMP
KLSE PSE SET STI
JAKCOMP 1 0.012863 0.003140 0.002791 0.003167 0.004256
2 0.013992 0.003609 0.003377 0.003920 0.005196
3 0.013608 0.004138 0.003762 0.004475 0.005381
4 0.013862 0.004212 0.004669 0.005340 0.006202
5 0.014158 0.004193 0.005338 0.005551 0.006843
6 0.014500 0.004412 0.005403 0.006451 0.007608
7 0.014227 0.004330 0.005254 0.006773 0.007709
KLSE 1 0.002107 0.008631 0.001743 0.002438 0.003392
2 0.002839 0.010238 0.001909 0.003186 0.004508
3 0.002793 0.010509 0.001585 0.003610 0.004519
4 0.003137 0.010794 0.001569 0.003572 0.005168
5 0.003293 0.010699 0.001611 0.003504 0.005511
6 0.003476 0.010717 0.001286 0.003923 0.005946
7 0.003373 0.010630 0.001105 0.003980 0.005859
PSE 1 0.002720 0.002532 0.012534 0.002354 0.002792
2 0.004037 0.003496 0.013744 0.004353 0.004535
3 0.003873 0.003554 0.013203 0.004698 0.004649
4 0.004664 0.003473 0.012587 0.005219 0.005467
5 0.004522 0.003449 0.012576 0.005347 0.005576
6 0.004933 0.003293 0.011914 0.006350 0.005723
7 0.005139 0.003108 0.011751 0.006768 0.005631
SET 1 0.003410 0.003914 0.002602 0.013853 0.005060
2 0.003756 0.004555 0.003256 0.013892 0.005480
3 0.003845 0.004985 0.003782 0.014790 0.006311
4 0.004091 0.005459 0.004515 0.014805 0.006699
5 0.003881 0.004818 0.004692 0.014721 0.006660
6 0.003922 0.005431 0.004932 0.015535 0.007066
7 0.003364 0.004790 0.004150 0.014909 0.007037
STI 1 0.003667 0.004355 0.002469 0.004048 0.011083
2 0.003284 0.003814 0.002652 0.004329 0.011242
3 0.002991 0.003888 0.002978 0.004751 0.011037
4 0.003094 0.003998 0.002944 0.004987 0.011528
5 0.002847 0.004416 0.003423 0.005319 0.012042
6 0.002970 0.004458 0.003290 0.005579 0.012248
7 0.002747 0.004464 0.003485 0.005614 0.011606
JURNAL AKUNTANSI DAN KEUANGAN, VOL. 11, NO. 1, MEI 2009: 1-12
8
As can be seen in Table 5, a generalised
impulse response analysis indicates that one
standard error shock to JAKCOMP would result in
a positive response by changes in STI of 0.0037,
one step ahead. Afterward, the responses have
become smaller ever since. A shock to STI,
commonly believed as the most prominent stock
index in ASEAN, results in second greatest
changes in the other indices in the short run
period. Meanwhile, the greatest contemporaneous
reaction of an index generally due to its own
shocks. This indicates that internal/domestic
shocks in a particular index may have greatest
significant impacts on its movements, and STI
become the most influential stock index in the
region at the time period.
While impulse response functions trace the
effects of a shock to one endogenous variable on to
the other variables in the system, variance
decomposition separates the variation in an
endogenous variable into the component shocks to
the system. As it is mentioned by Enders
(2004:280) the forecast error variance
decomposition tells the proportion of the
movements in a sequence due to its own shock
versus shock to the other variable. A shock to the ith
variable will not only affect that variable, but
can also be transmitted to all of the other variables
in the system.
Table 6 presents the result of the forecast error
variance decomposition of the serries in the period
of before financial crisis. As can be seen from the
table, in general, the proportion movements of the
indices are dominantly due to their own shocks.
Surprisingly, only around 70% of the error variance
of STI was attributable to own shocks in the steps
ahead, while JAKCOMP contributed maximum of
11% to STI’s error variance.
The Period of the 2007 Financial Crisis
The ADF test conducted to the serries at level
reveals the presence of unit root in the serries. The
lags order test then shows three lags length as the
appropriate lag order since the residual is not
serially correlated. However, the Johansen
Cointegration test fails to find the existence of
cointegration vector in the serries. This concludes
that the serries has no cointegrating relationship.
In other words, the indices have no long run
equilibrium during the 2007 financial crisis. The
finding somewhat contradicts with the ones given
by some other researchers (Arshanapalli et al 1993;
Sheng et al 2000, and Yang et al 2003), but
confirms that of Roca (2000:145).
The absence of cointegrating vector in the
series indicates that the cointegration analysis
framework is not possible to be carried out. Hence,
the VAR analysis framework would be applied to
estimate the relationship of the indices, as well as
to reveal the short run dynamic interactions among
the indices.
Table 6. Variance Decomposition
Variance Decomposition of JAKCOMP:
Period S.E. JAKCOMP KLSE PSE SET STI
1 0.012863 100.0000 0.000000 0.000000 0.000000 0.000000
2 0.019020 99.85659 0.010939 0.028547 0.049753 0.054174
3 0.023434 99.50386 0.136144 0.109426 0.186402 0.064169
4 0.027354 98.70463 0.197381 0.416767 0.512016 0.169204
5 0.030979 97.84212 0.214017 0.842533 0.727739 0.373589
6 0.034447 96.85652 0.241250 1.088327 1.165416 0.648487
7 0.037525 95.98986 0.258696 1.232197 1.641083 0.878166
Variance Decomposition of KLSE:
Period S.E. JAKCOMP KLSE PSE SET STI
1 0.008631 5.960468 94.03953 0.000000 0.000000 0.000000
2 0.013407 6.954068 92.87426 0.028773 0.040597 0.102303
3 0.017064 6.971806 92.59963 0.141095 0.195879 0.091595
4 0.020238 7.359674 91.94159 0.229436 0.209507 0.259789
5 0.022955 7.778958 91.22487 0.270537 0.201897 0.523735
6 0.025444 8.198359 90.24934 0.397181 0.289759 0.865363
7 0.027681 8.411056 89.59566 0.531422 0.381012 1.080849
Variance Decomposition of PSE:
Period S.E. JAKCOMP KLSE PSE SET STI
1 0.012534 4.709320 2.360945 92.92974 0.000000 0.000000
2 0.018719 6.761489 2.970640 89.41751 0.670515 0.179843
3 0.023044 7.286824 3.322941 87.93643 1.180083 0.273723
4 0.026505 8.605063 3.336361 85.70276 1.768701 0.587110
5 0.029573 9.250590 3.348796 84.34696 2.224120 0.829537
6 0.032241 10.12412 3.263569 82.40609 3.231500 0.974721
7 0.034710 10.92704 3.118762 80.65164 4.273855 1.028701
Variance Decomposition of SET:
Period S.E. JAKCOMP KLSE PSE SET STI
1 0.013853 6.060589 5.260581 1.058338 87.62049 0.000000
2 0.019639 6.673187 6.267257 1.496936 85.55756 0.005057
3 0.024625 6.681866 6.857899 1.896486 84.48702 0.076732
4 0.028819 6.893286 7.554222 2.489728 82.92112 0.141642
5 0.032441 6.871721 7.475509 3.036180 82.40051 0.216083
6 0.036054 6.746606 7.689571 3.399525 81.89916 0.265139
7 0.039080 6.483311 7.641728 3.442428 82.03551 0.397027
Variance Decomposition of STI:
Period S.E. JAKCOMP KLSE PSE SET STI
1 0.011083 10.94870 10.36302 1.108366 4.525942 73.05397
2 0.015812 9.692145 8.950799 1.464531 5.434735 74.45779
3 0.019334 8.876028 8.824737 1.942618 6.518288 73.83833
4 0.022557 8.401438 8.680124 2.081363 7.188879 73.64820
5 0.025643 7.733370 8.956077 2.414139 7.690111 73.20630
6 0.028485 7.354068 9.084298 2.515628 8.205909 72.84010
7 0.030842 7.066422 9.358043 2.738126 8.729031 72.10838
The VAR analysis, however, requires that the
series must be stationary. Hence, the non
stationary series may be made stationary by
differencing or detrending process. After
transforming the serries into first difference form,
the ADF test is re-employed to ensure that the
series are now stationary. The lag order test then
indicated that the appropriate lag length would be
three. After estimating the series using the VAR in
first difference analysis, the estimated models can
be shown in Table 7.
Table 8 shows the results of a block causality
test implemented on the series. The table reveals
that, using a 5 % level of significance, more
variables significantly granger cause another in the
crisis period compared to those in pre crisis period.
It means that there are more variables that their
current values have correlation with the past (lags)
value of another implying that the present value of
an index can be predicted with greater accuracy by
using past value of another. This then indicates
that there is an increase in causal linkages among
Atmadja: The Asean Stock Market Integration
9
those indices in the region during the crisis period.
The results are in fact different with those before
the crisis period showing a changing behaviour in
the indices’ movements. For instance, the lags of
SET and STI now significantly enter into the
equation for JAKCOMP, while in the pre crisis
period does not.
Table 7. Vector Autoregression Estimates
JAKCOMP KLSE PSE SET STI
JAKCOMP(-1) 0.072296 0.134318 0.132358 0.101597 0.029438
(0.06530) (0.03643) (0.05607) (0.05610) (0.06375)
[ 1.10716] [ 3.68735] [ 2.36041] [ 1.81101] [ 0.46179]
JAKCOMP(-2) 0.120791 0.083802 0.143389 0.198738 0.096179
(0.06664) (0.03718) (0.05723) (0.05726) (0.06506)
[ 1.81247] [ 2.25413] [ 2.50552] [ 3.47108] [ 1.47829]
JAKCOMP(-3) -0.058141 -0.010836 -0.017686 -0.026591 0.005682
(0.06641) (0.03705) (0.05703) (0.05706) (0.06484)
[-0.87543] [-0.29248] [-0.31011] [-0.46604] [ 0.08764]
KLSE(-1) -0.208902 -0.185895 -0.070152 -0.186469 -0.231196
(0.10815) (0.06033) (0.09287) (0.09291) (0.10558)
[-1.93162] [-3.08130] [-0.75537] [-2.00693] [-2.18980]
KLSE(-2) -0.210401 -0.180766 -0.102958 -0.375966 -0.275983
(0.10879) (0.06069) (0.09342) (0.09346) (0.10620)
[-1.93407] [-2.97869] [-1.10212] [-4.02270] [-2.59867]
KLSE(-3) 0.062985 0.124513 0.038155 0.027387 0.117089
(0.10840) (0.06047) (0.09308) (0.09313) (0.10582)
[ 0.58106] [ 2.05914] [ 0.40990] [ 0.29409] [ 1.10649]
PSE(-1) 0.017456 0.029010 -0.043004 0.048177 0.012092
(0.06255) (0.03489) (0.05371) (0.05374) (0.06106)
[ 0.27907] [ 0.83140] [-0.80062] [ 0.89653] [ 0.19802]
PSE(-2) 0.051068 0.048802 0.014394 0.077747 0.058784
(0.06224) (0.03472) (0.05345) (0.05347) (0.06076)
[ 0.82050] [ 1.40558] [ 0.26931] [ 1.45401] [ 0.96747]
PSE(-3) -0.020801 -0.019708 -0.040438 0.010962 -0.049437
(0.06019) (0.03358) (0.05168) (0.05171) (0.05876)
[-0.34561] [-0.58698] [-0.78239] [ 0.21200] [-0.84138]
SET(-1) 0.135694 -0.014998 0.029879 -0.069210 -0.078839
(0.07317) (0.04082) (0.06283) (0.06286) (0.07143)
[ 1.85453] [-0.36744] [ 0.47554] [-1.10100] [-1.10371]
SET(-2) -0.043142 0.019411 -0.125749 0.016529 0.040208
(0.07254) (0.04047) (0.06229) (0.06232) (0.07082)
[-0.59473] [ 0.47969] [-2.01868] [ 0.26523] [ 0.56778]
SET(-3) -0.149188 -0.042758 -0.077355 -0.060061 -0.147843
(0.07241) (0.04039) (0.06218) (0.06221) (0.07069)
[-2.06042] [-1.05858] [-1.24410] [-0.96552] [-2.09154]
STI(-1) 0.151469 0.077991 0.230158 0.060714 0.124110
(0.07280) (0.04061) (0.06252) (0.06255) (0.07107)
[ 2.08051] [ 1.92034] [ 3.68142] [ 0.97068] [ 1.74621]
STI(-2) 0.048648 0.026292 0.062191 0.044742 0.028764
(0.07342) (0.04096) (0.06305) (0.06308) (0.07168)
[ 0.66258] [ 0.64191] [ 0.98637] [ 0.70931] [ 0.40130]
STI(-3) 0.155315 -0.012852 0.046721 0.128301 0.021717
(0.07118) (0.03971) (0.06113) (0.06116) (0.06949)
[ 2.18190] [-0.32366] [ 0.76432] [ 2.09796] [ 0.31251]
C -0.000179 -0.000638 -0.000853 -0.000801 -0.001001
(0.00093) (0.00052) (0.00080) (0.00080) (0.00091)
[-0.19335] [-1.23409] [-1.07116] [-1.00633] [-1.10644]
Note: Standard errors in ( ) & t-statistics in [ ]
5 % level of significant
Table 8. VAR Pairwise Granger Causality/Block
Exogeneity Wald Tests
Dependent
variable Exclude Chi-sq Prob.
JAKCOMP KLSE 7.633367 0.0542
PSE 0.892896 0.8271
SET 7.966430 0.0467
STI 9.699188 0.0213
KLSE JAKCOMP 18.50092 0.0003
PSE 3.030814 0.3869
SET 1.542379 0.6725
STI 4.229433 0.2377
PSE JAKCOMP 11.80429 0.0081
KLSE 1.972551 0.5781
SET 5.832183 0.1201
STI 15.30169 0.0016
SET JAKCOMP 15.51245 0.0014
KLSE 19.47387 0.0002
PSE 2.766382 0.4291
STI 5.971454 0.1130
STI JAKCOMP 2.368227 0.4996
KLSE 12.88606 0.0049
PSE 1.790673 0.6170
SET 6.127043 0.1056
In order to capture the short run dynamic
interaction among the variables during the
financial crisis period, the generalized impulse
response and the forecast error variance
decomposition, would also be employed. The results
of the generalized impulse response analysis of the
series are presented in Table 9. As it is shown in
the table, during the financial crisis, the
generalised impulse response analysis indicates
that all variables gave greater immediate reactions
to a shock of one variable compared to those in the
pre-crisis era. This implies that the short run
interaction between two indices became more
intense during the 2007 financial crisis period. In
other words, the findings strongly indicate that the
ASEAN indices become more interdependent
during the financial crisis, although they had no
long run equilibrium.
The variance decomposition analysis (Tabel
10) reveals that the proportion of the movements in
an index due to its own shock for all indices
declined during the financial crisis. This means
that in the period of the financial crisis shocks to
other indices have more explanatory power to the
movements of a particular index in the s-steps
ahead. This finding seems reinforce the result of
generalized impulse response analysis that during
the 2007 financial crisis period, the ASEAN’s stock
indices tend to be more interdependent. Thus, it
somewhat confirmed the previous researches done
by Roca (2000:145) and Yang et al (2003:478)
which conclude that interdependency and causal
linkages among the indices become more
significant during crisis.
JURNAL AKUNTANSI DAN KEUANGAN, VOL. 11, NO. 1, MEI 2009: 1-12
10
Table 9. The Impulse Response to Generalized
One S.D. Innovations
Response
of
Period JAKCOMP KLSE PSE SET STI
JAKCOMP 1 0.020461 0.011579 0.008953 0.011621 0.013632
2 0.003635 0.001526 0.002198 0.004214 0.004293
3 0.001716 -0.000584 0.000966 0.000523 0.001221
4 0.000100 0.000434 -0.000181 -0.000541 0.001408
5 -0.000507 -0.000267 -0.000492 -0.000799 -0.000178
6 -0.000571 -0.000327 -0.000373 -0.000433 -0.000329
7 -0.000532 -0.000227 -0.000441 -0.000583 -0.000670
KLSE 1 0.006459 0.011414 0.005095 0.005452 0.006781
2 0.002659 0.000461 0.001297 0.001534 0.002167
3 0.001598 -0.000120 0.001140 0.001414 0.001395
4 -4.01E-05 0.000663 -0.000214 -0.000399 5.15E-05
5 8.40E-05 -0.000142 -0.000101 -0.000276 0.000126
6 -0.000133 -7.84E-05 -8.03E-05 -7.53E-05 8.39E-05
7 -0.000223 -3.72E-05 -0.000149 -0.000202 -0.000250
PSE 1 0.007689 0.007843 0.017570 0.007738 0.007307
2 0.005286 0.003377 0.002215 0.004376 0.005958
3 0.002075 -2.87E-05 0.000618 -4.34E-05 0.001447
4 -0.000102 -0.000320 -0.000544 -0.000358 0.000359
5 -0.000711 -0.000156 -0.000596 -0.000995 -0.000434
6 -0.000525 -0.000297 -0.000408 -0.000535 -0.000410
7 -0.000345 -0.000107 -0.000214 -0.000261 -0.000347
SET 1 0.009984 0.008397 0.007742 0.017578 0.011578
2 0.001362 -0.000435 0.000775 0.000119 0.000884
3 0.003034 -0.000533 0.001747 0.002050 0.002101
4 0.000804 0.001367 0.000655 0.000834 0.002134
5 -0.000160 -0.000455 -0.000346 -0.000645 -0.000140
6 -0.000169 -0.000252 -0.000108 -0.000116 0.000174
7 -0.000474 -0.000120 -0.000405 -0.000544 -0.000503
STI 1 0.013308 0.011867 0.008307 0.013156 0.019974
2 6.65E-05 -0.001392 -0.000281 -0.000578 0.000488
3 0.000879 -0.001056 0.000734 0.000894 0.000598
4 -0.001135 0.000113 -0.001453 -0.002127 -0.000913
5 -0.000329 -0.000385 -0.000354 -0.000535 -0.000465
6 -0.000392 -5.06E-05 -0.000155 -9.08E-05 -8.33E-05
7 -0.000300 -0.000110 -0.000214 -0.000306 -0.000493
Table 10. Variance Decomposition
Variance Decomposition of JAKCOMP:
Period S.E. JAKCOMP KLSE PSE SET STI
1 0.020461 100.0000 0.000000 0.000000 0.000000 0.000000
2 0.021073 97.24338 0.093183 0.174688 1.656038 0.832708
3 0.021264 96.15798 0.878554 0.303776 1.644745 1.014946
4 0.021412 94.83229 0.912108 0.332163 1.758313 2.165129
5 0.021436 94.68301 0.910264 0.353508 1.830443 2.222772
6 0.021445 94.67600 0.909518 0.357529 1.831413 2.225544
7 0.021458 94.61447 0.910093 0.372423 1.851690 2.251322
Variance Decomposition of KLSE:
Period S.E. JAKCOMP KLSE PSE SET STI
1 0.011414 32.02605 67.97395 0.000000 0.000000 0.000000
2 0.011842 34.79170 64.28911 0.182911 0.037178 0.699104
3 0.012084 35.16451 62.80328 0.676161 0.434642 0.921404
4 0.012135 34.86724 62.73966 0.813231 0.661356 0.918514
5 0.012148 34.79750 62.64135 0.817565 0.737312 1.006279
6 0.012152 34.78759 62.60206 0.817460 0.736878 1.056009
7 0.012156 34.79723 62.56663 0.822289 0.743584 1.070267
Variance Decomposition of PSE:
Period S.E. JAKCOMP KLSE PSE SET STI
1 0.017570 19.14909 5.811252 75.03966 0.000000 0.000000
2 0.018664 24.99113 5.212981 66.52072 0.824013 2.451155
3 0.018915 25.53371 5.670186 64.76357 1.220127 2.812400
4 0.018949 25.44521 5.678093 64.59816 1.227054 3.051487
5 0.018986 25.48622 5.680653 64.39330 1.373668 3.066158
6 0.018997 25.53520 5.674576 64.33616 1.390703 3.063358
7 0.019001 25.55543 5.674949 64.30681 1.392098 3.070709
Variance Decomposition of SET:
Period S.E. JAKCOMP KLSE PSE SET STI
1 0.017578 32.25784 3.592081 2.848586 61.30150 0.000000
2 0.017729 32.29985 4.211044 2.919796 60.38029 0.189017
3 0.018269 33.17686 6.197357 3.213120 57.06075 0.351916
4 0.018426 32.80436 6.452377 3.159159 56.10525 1.478857
5 0.018446 32.74148 6.496111 3.163665 56.07662 1.522124
6 0.018455 32.71765 6.500189 3.160603 56.02145 1.600115
7 0.018468 32.73729 6.500537 3.178835 55.97476 1.608586
Variance Decomposition of DLNSTI:
Period S.E. JAKCOMP KLSE PSE SET STI
1 0.019974 44.38852 6.931322 0.493399 7.788191 40.39857
2 0.020116 43.76814 7.577686 0.490308 7.717361 40.44651
3 0.020262 43.32822 8.333029 0.692744 7.747659 39.89834
4 0.020443 42.86986 8.386302 1.122186 8.404362 39.21729
5 0.020451 42.86178 8.393528 1.128666 8.426711 39.18932
6 0.020457 42.87448 8.399131 1.128383 8.425800 39.17221
7 0.020464 42.86514 8.394334 1.130788 8.426286 39.18345
CONCLUSION
The study concludes that two cointegrating
vectors are found in the series before the 2007
financial crisis period indicating the existing of long
run equilibrium in the series during the time
period. However, the study fails to find any
cointegrating vector in the series during the
financial crisis period. The results prove that the
long run relationship of the ASEAN indices has
been removed by the 2007 financial crisis.
The block causality tests employed in both subsample
period reveal that more significant causal
linkages are found in the series during the
financial crisis period compared to those before the
financial crisis. The accounting innovation
analyses conducted to the series also indicate that
the short run dynamic interactions among the
indices tend to be more intense during the financial
crisis period. These all indicate that the indices
become more interdependent during the financial
crisis period since the moment gives rise the
explanatory power of a sequence to the movements
of another.
The general conclusion that may be withdrawn
from this study is that the contagious effect of the
2007-US financial crisis has affected the ASEAN’s
capital market integration, and has changed the
behaviour of the indices’ movements both in the
short run and in the long run.
Thus, the implication policy that can be
suggested is that the diversification of portfolio
within the ASEAN stock markets in the short run
is unlikely to reduce the risk due to the high degree
of financial interdependent of these markets
during the financial crisis.
REFERENCES
Arshanapalli, B. and Doukas, J. 1993,
‘International Stock Market Linkages:
Evidence from the pre- and post-October
1987 Period’, Journal of Banking and
Finance, 17 (1): 193-208.
Arshanapalli, B., Doukas, J. and Lang, L. 1995,
‘Pre and post-October1987 Stock Market
Linkages between U.S and Asian markets’,
Pacific-Bacin Finance Journal, 3 (1): 57-73.
Atmadja: The Asean Stock Market Integration
11
Brailford, T., Heaney, R. and Bilson, C. 2004.
Investment: Concepts and Applications,
Second Edition. Thomson Learning:
Australia.
Brooks, C. 2002. Introductory Econometrics for
Finance. Cambrige University Press:
Cambrige.
Chan, K.C., Gup, B.E. and Pan, M. 1992, ‘An
Empirical Analysis of Stock Prices in Major
Asian Markets and the United States’,
Financial Review, 27 (2): 289-307.
Chung, P. and Liu, D. (1994), ‘Common Stochastic
Trend in Pacific Rim Stock Markets’,
Quarterly Review of Economics and Finance,
34: 241-259.
Climent, F. and Meneu, V. 2003, ‘Has 1997 Asian
Crisis Increased Information Flows between
International Markets’, International Review
of Economics and Finance, 12(1): 111-143.
DeFusco, R.A., Geppert, J.M. and Tsetsekos, G.P.
(1996), ‘Long-run Diversification Potential in
Emerging Stock Markets’, Financial Review,
31(2): 343-363.
Dekker, A., Sen, K. and Young, M. 2001. ‘Equity
Market Linkage in the Asia Pacific Region:
A Comparison of the Orthogonalized and
Generalized VAR Approaches’. Global
Finance Journal, 12: 1-33.
Department of Foreign Affair and Trade of
Australia (DFAT). 1999, Asia’s Financial
Markets: Capitalising on Reform, East Asia
Analytical Unit, Canberra.
Enders, W. (2004). Applied Econometric Time
Series: Second Edition. John Wiley & Sons:
USA.
Eun, C.S. and Shim, S. 1989, ‘International
Transmission of Stock Market Movements’,
Journal of Financial and Quantitative
Analysis, 24 (2): 241-56.
Hung, B. and Cheung, Y. 1995, ‘Interdependence of
Asian Emerging Equity Markets’, Journal of
Business Finance and Accounting, 22 (2):
281-88.
International Monetary Fund 2004. International
Financial Statistic (IFS) Online. Washington
DC.
Johansen, S. and Juselius, K.. 1990, ‘The Full
Information Maximum Likelihood
Procedures for Inference on Cointegration
with Applications to the Demand for Money’,
Oxford Bulletin of Economics and Statistics,
52(2): 169-210.
Kasa, K. 1992, ‘Common Stochastic Trends in
International Stock Markets’, Journal of
Monetary Economics, 29: 95-124.
Kenen, P.B. 1976, ‘Capital Mobility and
Integration: A Survey’, Princeton Studies in
International Finance No. 39, Princeton
University: USA.
King, M and Wadwhani, S. 1990, ‘Transmission of
Volatility between Stock Markets’, Review of
Financial Studies, 3 (1): 5-33.
Krugman, P.R. and Obstfeld, M. 2003. International
Economics: Theory & Policy 6th
Edition. The Addison-Wesley: Boston.
Liu, Y.A., Pan, M.S. and Shieh, J.C.P. 1998,
‘International Transmission of Stock Price
Movements: Evidence from the U.S. and
Five Asian-Pacific Markets’, Journal of
Economics and Finance, 22 (1): 59-69.
Lutkepohl, H. and Reimers, H.E. 1992. ‘Impulse
Response Analysis of Cointegrated System’,
Journal of Economic Dynamics and Control,
16: 53-78.
Masih, A.M.M. and Masih, R. 1999, ‘Are Asian
Stock Markets Fluctuations Due to Mainly
to Intra-regional Contagion Effects?
Evidence Based on Asian Emerging
Markets’, Pacific-Basin Finance Journal, 7
(3): 251-82.
Masih, R. and Masih, A.M.M. 2001, ‘Long and
Short-term Dynamic Causal Transmission
amongst International Stock Markets’,
Journal of International Money and Finance,
20: 563-87.
Palac-McMiken, E.D. 1997, ‘An Examination of
ASEAN Stock Markets: a Cointegration
Approach’. ASEAN Economic Bulletin, 13
(3): 299-311.
Pretorius, E. 2002, ‘Economic Determinants of
Emerging Market Stock Market Interdependence’,
Emerging Markets Review, 3:
84-105.
Roca, E.D. 2000, Price Interdependence among
Equity Markets in the Asia-Pacific Region:
Focus on Australia and ASEAN, Ashgate:
Aldershot.
Runkle, D.E. 1987, ‘ Vector Autoregressions and
Reality’, Journal of Business and Economic
Statistics, 5(4): 437-442.
JURNAL AKUNTANSI DAN KEUANGAN, VOL. 11, NO. 1, MEI 2009: 1-12
12
Sachs, J.D and Larrain, F. B. 1993, Macroeconomics
in the Global Economy, Prentice
Hall: New Jersey.
Serletis, A. and King, M. 1997, ‘Common Stochastic
Trends and Convergence of European Union
Stock Markets’, The Manchester School,
65(1): 44-57.
Sheng, H. and Tu, A. 2000, ‘A Study of
Cointegration and Variance Decomposition
among National Equity Indices before and
during the Period of the Asian Financial
Crisis’, Journal of Multinational Financial
Management, 10 (3): 345-65.
Stulz, R. 1981, ‘On the Effects of Barriers to
International Investment’, Journal of
Finance, 36: 923-934.
Thomas, R.L 1997. Modern Econometrics: an
Introduction. Addison-Wesley: Harlow.
Yang, J., Kolari, J.W. and Min, I. 2003, ‘Stock
Market Integration and Financial Crises:
the Case of Asia’, Applied Financial
Economics, 13: 477-486.
Yellen, Janet L. 2008, ‘Economic Prospects for the
US Economy from a Monetary Policymaker’s
Perspective’, CFA Institute Annual Conference
Vancouver, http://www.cfapubs.org.

Tinggalkan Balasan

Isikan data di bawah atau klik salah satu ikon untuk log in:

Logo WordPress.com

You are commenting using your WordPress.com account. Logout / Ubah )

Gambar Twitter

You are commenting using your Twitter account. Logout / Ubah )

Foto Facebook

You are commenting using your Facebook account. Logout / Ubah )

Foto Google+

You are commenting using your Google+ account. Logout / Ubah )

Connecting to %s


%d blogger menyukai ini: