This article was downloaded by: [91.180.109.10]On: 02 May 2014, At: 09:39Publisher: RoutledgeInforma Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House,37-41 Mortimer Street, London W1T 3JH, UK

Applied Financial EconomicsPublication details, including instructions for authors and subscription information:http://www.tandfonline.com/loi/rafe20

Stock market integration and financial crises: the caseof AsiaJian Yang a , James W. Kolari b & Insik Min ca Department of Accounting, Finance & Information Systems, Prairie View A&M University,Texas 77446, USAb Department of Finance, Texas A&M University, College Station, Texas 77843, USAc Department of Economics, Texas A&M University, College Station, Texas 77843, USAPublished online: 07 Oct 2010.

To cite this article: Jian Yang , James W. Kolari & Insik Min (2003) Stock market integration and financial crises: the case ofAsia, Applied Financial Economics, 13:7, 477-486, DOI: 10.1080/09603100210161965

To link to this article: http://dx.doi.org/10.1080/09603100210161965

PLEASE SCROLL DOWN FOR ARTICLE

Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) containedin the publications on our platform. However, Taylor & Francis, our agents, and our licensors make norepresentations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose ofthe Content. Any opinions and views expressed in this publication are the opinions and views of the authors,and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be reliedupon and should be independently verified with primary sources of information. Taylor and Francis shall not beliable for any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilitieswhatsoever or howsoever caused arising directly or indirectly in connection with, in relation to or arising out ofthe use of the Content.

This article may be used for research, teaching, and private study purposes. Any substantial or systematicreproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in anyform to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http://www.tandfonline.com/page/terms-and-conditions

Stock market integration and

financial crises: the case of Asia

JIAN YANG*, JAMES W. KOLARIz and INSIK MIN}

Department of Accounting, Finance & Information Systems, Prairie View A&MUniversity, Texas 77446, USA, zDepartment of Finance and }Department ofEconomics, Texas A&M University, College Station, Texas 77843, USA

This study examines long-run relationships and short-run dynamic causal linkagesamong the US, Japanese, and ten Asian emerging stock markets, with the particularattention to the 1997–1998 Asian financial crisis. Extending related empirical studies,comparative analyses of pre-crisis, crisis, and post-crisis periods are conducted tocomprehensively evaluate how stock market integration is affected by financialcrises. In general, the results for the case of Asia show that both long-run cointegra-tion relationships and short-run causal linkages among these markets were strength-ened during the crisis and that these markets have generally been more integratedafter the crisis than before the crisis. Detailed country-by-country analyses areprovided, which yield a variety of new results concerning the roles of individualcountries in international stock market integration. An important implication ofour findings is that the degree of integration among countries tends to changeover time, especially around periods marked by financial crises.

I . INTRODUCTION

A number of empirical studies have examined long-run

relationships and short-run dynamic causal linkages

among the emerging Asian financial markets and major

developed markets.1 Cointegration studies that investi-

gated long-run relationships focus on the extent to which

nascent stock markets in Asian countries are inter-

nationally integrated and, in turn, have important implica-

tions to diversification potential in Asian stock markets

(Chan et al. 1992; Hung and Cheung, 1995; DeFusco et

al. 1996; Masih and Masih, 2001). Studies that have esti-

mated short-run dynamic causal linkages seek to better

understand the propagation mechanism driving stock mar-

ket fluctuations in different countries, especially with

respect to market crashes (Masih and Masih, 1997a,

1999; Ghosh et al. 1999; Sheng and Tu, 2000). These and

other related studies (Chung and Liu, 1994; Arshanapalli et

al. 1995; Cheung, 1997; Janakiramanan and Lamba, 1998;

Dekker et al. 2001) employ vector autoregression (VAR)

techniques, including cointegration, Granger causality,

impulse response analysis, and forecast error variance

decomposition. In general, the empirical evidence pre-

sented in these studies is mixed with respect to both long-

run relationships and short-run dynamic causal linkages.

This study re-examines Asian stock market integration

around the time of the 1997–1998 Asian financial crisis.

The extant literature is contributed to in several ways.

First, the impact of the Asian financial crisis on those coun-

tries’ stock market integration for three different stages of

the crisis is comparatively documented – namely, pre-crisis,

crisis, and post-crisis periods. It would appear that no

previous work has addressed the issue of how the crisis

altered market integration among Asian countries over

time. Previous studies of market integration invariably

assume that markets are either perfectly integrated, per-

Applied Financial Economics ISSN 0960–3107 print/ISSN 1466–4305 online # 2003 Taylor & Francis Ltdhttp://www.tandf.co.uk/journalsDOI: 10.1080/09603100210161965

Applied Financial Economics, 2003, 13, 477–486

477

*Corresponding author. Email: jian_yang@pvamu.edu1 See Bessler and Yang (2003) for a summary of such analysis conducted in major international stock markets. Also, some researchersfocus on stock return volatility spillover in Asian stock markets, which is not pursued in this study. See Ng (2000) for an excellentexample in this line of research.

Dow

nloa

ded

by [

91.1

80.1

09.1

0] a

t 09:

39 0

2 M

ay 2

014

fectly segmented, or partially integrated but the extent ofintegration is constant over time (Bekaert and Harvey,1995). By focusing on how the Asian financial crisisaffected market integration over time, it is sought to deter-mine if this common assumption holds. Second, recentlydeveloped vector autoregression (VAR) technique of gen-eralized impulse response analysis (Koop et al. 1996;Pesaran and Shin, 1998) is employed to estimate short-run dynamic causal linkages across stock markets. Thisempirical approach is motivated by the existence of strongcontemporaneous correlations among Asian stock marketinnovations as documented in the literature (Cheung, 1997;Janakiramanan and Lamba, 1998). In such instances it iswell known that traditional orthogonalized impulseresponse analysis (or forecast error variance decomposi-tion) based on the widely-used Choleski factorization ofVAR innovations is sensitive to the ordering of the vari-ables. By contrast, generalized impulse response analysis isinvariant to the ordering of the variables. Third and last,while previous studies conduct analyses of major Asianmarkets, more comprehensive information is provided byincluding the developed markets of the USA and Japan aswell as ten Asian emerging stock markets. Several smallemerging stock markets included in this study, such asIndia, Indonesia, Pakistan, and the Philippines, havereceived scant attention in prior research, with the excep-tion of Ghosh et al. (1999). In this regard, 12 markets aremodelled in one system rather than using several smallersubsystems, which is important in making accurate infer-ences concerning cointegration and causality (Hasapis et al.1999).In general, the results for the case of Asia show that both

long-run cointegration relationships and short-run causallinkages among these markets strengthened during thefinancial crisis and that these markets have generally beenmore integrated after the crisis than before the crisis. Morespecifically, the results reveal that, while the US exertedsubstantial influence on most Asian stock markets in allthree sample periods, Japan had little or no impact onmost markets in the region except during the crisis period.Further results show that, during the non-crisis periods,Japan, Taiwan, and the Philippines were fairly isolatedmarkets, but Thailand and Indonesia were fairly interactiverather than fairly isolated as reported previously in theliterature. Korea, India and Pakistan were fairly endogen-ous, which has not yet been reported in the literature.Singapore instead of Hong Kong appeared to be a regionalleader. An important implication of the findings is that,consistent with Bekaert and Harvey, the degree of integra-tion among countries tends to change over time, especiallyaround periods marked by financial crises.The next section briefly reviews the literature. Section III

describes the empirical framework. Section IV presents theempirical results. The last section gives the summary andconclusions.

II . LITERATURE REVIEW

As mentioned above, the empirical findings of previousstudies on the integration of Asian stock markets aremixed. Chan et al. (1992) and DeFusco et al. (1996)reported no cointegration between US and many Asianemerging stock markets (i.e., Hong Kong, Korea,Singapore, Taiwan, Malaysia, Thailand, and thePhilippines) in the 1980s and early 1990s. By contrast,Arshanapalli et al. (1995) and Masih and Masih (1997a,1999, 2001) reported only one cointegrating vector amongseveral major Asian emerging markets (i.e., Hong Kong,Korea, Singapore, and Taiwan) and major developedmarkets. Chung and Liu (1994) documented two cointe-grating vectors between the US and these larger Asian-Pacific stock markets. Ghosh et al. (1999) documentedpairwise cointegration between USA (and Japan) andsome Asia-Pacific stock markets during the 1997–1998Asian financial crisis. Sheng and Tu (2000) reported nocointegration in the year before the Asian financial crisisbut one cointegrating vector during the crisis between theUS and many Asian stock markets.Most of the aforementioned studies are conducted in

local currency terms. Masih and Masih (1999, 2001) didutilize US dollars as the base currency, and only Hungand Cheung (1995) employ data collected in terms ofboth local currency and US dollars. Using US dollardenominated stock prices, the authors found that fivemajor Asia-Pacific stock markets (i.e., Hong Kong,Korea, Malaysia, Singapore, and Taiwan) were cointe-grated after but not before the 1987 stock crash.Interestingly, they found no cointegration throughout thewhole sample period using stock prices in local currencyterms.While currency numeraire is one explanation for differ-

ences in conclusions reached by different studies on thequestion of Asian stock market integration, other reasonscould be frequency of data, model specification, andsample time period. Frequency of data (i.e., daily, weekly,and monthly) likely has only limited effects on the cointe-gration analysis, as Hakkio and Rush (1991) have shownthat, given a fixed sample period, frequency of data did notaffect cointegration results. With regard to model specifica-tion, many previous studies typically used several smallersubsystems (e.g., bivariate systems) to model relationshipsamong a large number of markets. This practice mayignore potential indirect channels of stock market linkageand generate spurious patterns of stock market linkages(Janakiramanan and Lamba, 1998). Recently, Hasapis etal. (1999) have demonstrated that cointegration and caus-ality inferences are strongly affected by the omission of animportant causing variable in the system. Finally, differentsample time periods could account for the different find-ings. Asian stock markets may (or may not) have beenmore integrated with each other and with the world for

478 J. Yang et al.

Dow

nloa

ded

by [

91.1

80.1

09.1

0] a

t 09:

39 0

2 M

ay 2

014

various reasons, including extensive stock market liberal-ization, increased economic integration within the regionand with the world, technological advances in communica-tion, and stock market crashes. Stock market crashes suchas the one in 1987 have been widely argued to strengthenmajor international as well as Asian stock market linkages.This study contributes to the literature by comprehensivelyexamining Asian stock market integration using a 12country VAR system with different currency numerairesand different sample periods surrounding the Asian finan-cial crisis.

III . EMPIRICAL FRAMEWORK

Long-run and short-run relationships are modelledbetween emerging Asian and the US and Japanese stockmarkets using a cointegrated vector autoregression (VAR)framework. Let Xt denote a vector which includes m non-stationary variables (m ¼ 12 countries). If m price series inXt are cointegrated, they can be expressed using a reducedform error correction model (ECM) with a lag of k� 1(i.e., a lag of k for a levels VAR):

�Xt ¼ �1�Xt�1 þ � � � þ �k�1�Xt�kþ1 þ �Xt�1 þ �þ "t

ð1Þwhere � ¼ �� 0, the matrix � contains short-run adjust-ment parameters to the long-run relationships reflected inthe matrix �, and the rank of � determines the r number ofcointegrating vectors.To comprehensively investigate the impact of the Asian

financial crisis on long-run relationships among samplestock markets, numbers of cointegrating vectors amongthese markets in three periods are compared: pre-crisis (2January 1995 – 31December 1996), crisis (1 July 1997 – 30June 1998), and post-crisis (1 July 1998 – 15 May 2001). Assuggested by Kamin (1999, p. 506) and Corsetti et al. (1999,pp. 370–1), the first year after the crisis was marked bystabilization of exchange rates but continued instability inthe macroeconomy as a whole. As such, it can be consid-ered as a second phase in the financial crisis. For this rea-son the post-crisis period is further divided into twosubperiods: 1 July 1998 – 30 June 1999 (transition period)and July 1, 1999 – 15 May 2001 (post-crisis period).Before testing whether the price series are cointegrated, it

should be established that each univariate series is nonsta-tionary, or Ið1Þ. The existence of a unit root in Asian stockmarket prices is well established in the literature. In par-ticular, Masih and Masih (1999, 2001) have conductedextensive tests which verify the existence of a unit rootfor all Asian stock market index prices. Following Yangand Leatham (2001) and Bessler and Yang (2003), arelatively new procedure is applied for testing stationarityof data series (see also Hansen and Juselius (1995)). Unlike

commonly-used augmented Dickey–Fuller (ADF) andPhilips–Perron (PP) tests, in which the null hypothesis isnonstationarity, the null under this new procedure is sta-tionarity. Specifically, in the presence of one cointegratingvector, Johansen (1991) likelihood ratio tests are applied totest whether any 11 (out of 12) price series do not enter thecointegrating vector. This procedure is equivalent to testingthe possibility that one cointegrating vector might arisesimply because the remaining variable is itself stationary.Similarly, in the presence of two cointegrating vectors, like-lihood ratio tests are applied to test whether any ten priceseries are excluded from both cointegrating vectors. Nocointegration among 12 variables implies that all variablesare nonstationary. The likelihood ratio test results (avail-able upon request from the authors) show that stationarityof each price series can be rejected at either 1% or 5%levels for all subperiods and for stock prices in both localcurrency and US dollar terms. Thus, it can be concludedthat each price series is I(1).A priori, it is expected that Asian stock markets are most

cointegrated during the crisis period, which would be cap-tured by the highest number of cointegrating vectors. Asmentioned earlier, the number of cointegrating vectors isdetermined by the rank of � ¼ �� 0. In this respect tracetest statistics of Johansen (1991) can be calculated to testthe null hypothesis that there are at most r cointegratingvectors.Additional testing of the cointegration space spanned by

� can produce further information on long-run marketlinkages. Particular interest is in how many markets areactually excluded in all of the identified long-run relation-ships (if any). This hypothesis can be tested by examiningwhether �i j ¼ 0 for all i ði ¼ 1; . . . ; rÞ cointegration vectorsfor the jth ð j ¼ 1; . . . 12Þ market.Short-run stock market linkages are reflected in the

parameter matrixes � and �i. While the � parametermatrix defines the short-run adjustments to long-runrelationships, the parameter matrixes (�1; . . . ;�p�1) definethe short-run adjustments to changes in the process.However, it is well recognized that, like standard VAR,the individual coefficients of the ECM can be difficult tointerpret (Lutkepohl and Reimers, 1992). In view of thisdrawback, innovation accounting is the most appropriatemethod of summarizing the short-run dynamic structureof market linkages.It is common practice in VAR analyses to rely on a

Choleski factorization to orthogonalize VAR innovationsso that they are uncorrelated contemporaneously.Unfortunately, innovation accounting results based onthe Choleski factorization are sensitive to the ordering ofvariables when the residual covariance matrix is non-diagonal. Here generalized impulse response analysis isapplied as developed in Koop et al. (1996) and Pesaranand Shin (1998), which is invariant to the ordering of thevariables in the VAR model.2

Stock market integration and financial crises 479

Dow

nloa

ded

by [

91.1

80.1

09.1

0] a

t 09:

39 0

2 M

ay 2

014

Under the generalized VAR approach, �Xt is rewrittenas the infinite moving average version of Equation 1, or

�Xt ¼X1

i¼0Ci"t�i t ¼ 1; 2; . . . ;T ð2Þ

The (scaled) generalized impulse response function whichmeasures the effect on �Xtþn of the shock to the jthequation in Equation 1 can be specified as follows:

jðnÞ ¼ ��1=2j j Cn

Xej; n ¼ 0; 1; 2; . . . ; ð3Þ

where �j j is jjth element in the variance–covariance matrix�, and ej is m 1 vector with unity as its jth row and zeroselsewhere.

IV. EMPIRICAL RESULTS

Data

The data consist of daily stock index closing prices of twodeveloped markets and ten Asian emerging markets, i.e.,Hong Kong Hang Seng (HK), India BSE National (ID),Indonesia Jakarta SE composite (IN), Japan Nikkei 225Stock Average (JP), Korea SE Composite (KR),Malaysia Kuala Lumpur Composite (ML), PakistanKarachi SE 100 (PK), Philippines SE Composite (PH),Singapore Strait Times (SG), Thailand Bangkok S.E.T.(TH), Taiwan SE weighted (TW), and US S&P 500 com-posite (US). The sample period is 2 January 1995, to 15May 2001, which includes 1662 daily observations for eachseries. All stock indices are expressed in both local currencyand US dollar terms. The source of data was Datastream.

Cointegration analysis results

The choice of optimal lags for the VAR system isselected based on the Akaike information criterion (AIC).Two lags are chosen for all periods in this study.Diagnostic statistics reveal that the residuals are generallywell behaved and in particular free from autocorrelationproblems. Trace test results in terms of local currencies andUS dollars are reported in Tables 1 and 2, respectively.

As shown in Table 1, using local currencies and exclud-

ing a linear trend term, trace test results indicate no

cointegrating vector exists in the pre-crisis and transition

periods but two cointegrating vectors exist in both the crisis

and post-crisis periods. With a linear trend term, similar

results are found across periods. The finding in the pre-

crisis period is consistent with Chan et al. (1992),

DeFusco et al. (1996), and Sheng and Tu (2000); however,

unlike Sheng and Tu (2000), two rather than one cointe-

grating vector were found during the crisis. These results

suggest that the long-run integration was intensified in the

crisis and post-crisis periods relative to the pre-crisis per-

iod. By implication, it is inferred that the Asian financial

crisis altered the degree of market integration in the region

over time. The results also differ from those of Masih and

Masih (1997b), who concluded that the 1987 stock crash

did not affect the number of common stochastic trends in

major stock markets. Also, it is interesting to note that

long-run integration was weaker during the transitional

period. This finding is interpreted as indirect evidence

that linkages among real economic variables may partly

account for stock market linkage.

Table 2 shows that the results are unchanged using US

dollars. The only difference is that one rather than zero

cointegrating vector is found in the pre-crisis period,

which is consistent with Masih and Masih (1999, 2001).

Thus, allowance for the exchange rate adjustment can

affect the number of cointegrating vectors, in line with

studies by Hung and Cheung (1995) and Bessler and

Yang (2003).

Table 3 gives the results to whether certain stock market

index prices are excluded from the long run relationships

identified by the above trace tests. Based on local curren-

cies, only the Hong Kong stock market is excluded from

the long run relationships during the crisis period, whereas

the Philippines and Taiwan stock markets are ruled out

from the long-run relationships in the post-crisis period.

The results using US dollars as numeraire reveal that

three markets are excluded from the long-run relationships

during the crisis and post-crisis periods, four markets

are excluded during the pre-crisis period, and eight

markets are excluded in the transition period. The latter

result confirms our earlier finding that long-run linkage

was weakest among the Asian stock markets in the transi-

tion period.

480 J. Yang et al.

2 In a comparative study of the traditional orthogonalized and generalized VAR analysis, Dekker et al. (2001) found that the generalizedapproach provided more reliable results than the traditional orthogonalized approach based on the Choleski factorization. Also, thegeneralized impulse response analysis instead of the generalized forecast error variance decomposition is used in this study for animportant reason. As pointed out by Masih and Masih (1999, p. 269), the generalized forecast error variance decomposition cannotbe strictly used to isolate responses of a particular market, assuming that all other shocks are not present or not running in conjunctionwith the particular shock in question. In other words, when applying and interpreting generalized forecast error variance decomposition,one should not attribute the shock to the sole variable in the system, as would be appropriate in traditional forecast error variancedecomposition.

Dow

nloa

ded

by [

91.1

80.1

09.1

0] a

t 09:

39 0

2 M

ay 2

014

Generalized impulse response analysis results

Due to the weak stock market linkages found in the transi-

tion period, ECMs only were estimated for the pre-crisis,

crisis, and post-crisis periods. Generalized impulse

response functions estimated from the ECMs provide

insight into how innovations in a particular market in the

system affected other markets through dynamic inter-

actions among markets. Table 4 summarizes these results.

To conserve the space, only the results for day 20 based on

the local currency terms are reported here due to little

difference of dollar-denominated results (results on days 0

to 20 are available from the authors upon request).

Consistent with Bessler and Yang (2003), it is found that

exchange rate adjustments can affect long-run cointegra-

tion relationships but do not significantly affect the short-

run dynamic causal linkage pattern of stock markets.

Following Brocato (1994) and Bessler and Yang (2003),

the analyses employ a forecast horizon of 20 days ahead.3

Similar to Dekker et al. (2001), significant responses are

defined as those that exceed 0.20 unit standard deviations

on day 20. Following Kamin (1999), the Asian markets are

organized into three groups: serious victim markets

(Indonesia, Korea, Malaysia, Philippines, and Thailand),

Stock market integration and financial crises 481

Table 1. Trace tests for cointegration in local currency terms

Panel A: Without linear trend

Number ofcointegrating Pre-crisis Crisis Transition Post-crisis Critical valuevectors period period period period (5%)

r ¼ 0 328.12 380.04 334.69 368.47 338.09r ¼ 1 251.69 292.52 263.73 295.08 289.70r ¼ 2 196.39 220.43 211.20 226.94 244.56r ¼ 3 148.92 174.62 166.42 172.22 203.34r ¼ 4 117.04 137.73 129.36 134.86 165.73r ¼ 5 93.59 104.10 95.92 105.95 132.00r ¼ 6 71.11 75.70 70.04 78.58 101.83r ¼ 7 50.82 54.06 48.48 51.87 75.73r ¼ 8 34.28 33.59 32.04 31.96 53.42r ¼ 9 23.09 20.81 18.60 20.06 34.79r ¼ 10 12.70 10.94 8.86 10.54 19.99r ¼ 11 6.07 3.85 2.69 5.16 9.13

Panel B: With linear trend

Number ofcointegrating Pre-crisis Crisis Transition Post-crisis Critical valuevectors period period period period (5%)

r ¼ 0 305.42 363.32 326.16 361.13 323.93r ¼ 1 231.72 277.21 255.36 288.02 276.36r ¼ 2 178.38 205.68 205.71 219.87 232.60r ¼ 3 131.89 160.07 160.93 165.78 192.30r ¼ 4 103.17 125.36 123.99 130.11 155.74r ¼ 5 79.74 91.86 90.71 101.22 123.03r ¼ 6 58.48 64.60 64.85 73.86 93.91r ¼ 7 41.54 44.13 43.35 47.17 68.86r ¼ 8 28.24 25.43 28.01 29.43 47.20r ¼ 9 17.07 13.76 14.94 18.94 29.37r ¼ 10 6.86 4.51 5.44 9.42 15.34r ¼ 11 0.77 0.24 0.28 4.18 3.84

3 Different from Dekker et al. (2001) but similar to Masih and Masih (2001), this study allows for long-run relationships (if any) whenconducting generalized impulse response analysis. In such a case, the impact of a shock on other markets may not be transitory and dieaway within a few days. Instead, as demonstrated in Masih and Masih (2001), the impact of a shock is most likely to be lasting for a longperiod after it returns to its long-run level. This study is able to explore the possible lasting effect of a shock by employing a longerhorizon of day 20 ahead. Also, Phillips (1998) has recently proved that the imposition of cointegration constraints in the nonstationaryVAR analysis is crucial in yielding consistent results on impulse response analysis and forecast error variance decompositions.

Dow

nloa

ded

by [

91.1

80.1

09.1

0] a

t 09:

39 0

2 M

ay 2

014

less serious victim markets (Hong Kong, Singapore, andTaiwan), and non-victim markets (the US, Japan, India,and Pakistan).Previous work by Janakiramanan and Lamba (1998)

found that the Indonesian market was a relatively isolatedmarket. However, Table 4 shows that quite a few markets,including Hong Kong, India, Malaysia, Philippines,Pakistan, and even the USA, responded significantly toshocks from Indonesia in the pre-crisis period. In thepost-crisis period Hong Kong, Japan, Korea, Malaysia,Philippines, Pakistan, Singapore, and Thailand showed sig-nificant responses to shocks from the Indonesian market.Interestingly, different from what is observed in the case ofother victim markets, the Indonesian market did notincrease its influence on other markets during the periodof the crisis. This evidence agrees with the fact that theIndonesian market did not initiate the crisis. From Table4 it can also be seen that the Indonesian market respondedto shocks from other markets, including the USA and Japan.

Turning to Korea, only Malaysia, Philippines, andSingapore responded to shocks from the Korean marketin the pre-crisis period, and only Pakistan, Thailand, andTaiwan was responsive in the post-crisis period. Furtherevidence that the Korean market held a passive role inAsian markets is reflected in the fact that only Thailandand Taiwan reacted to news from the Korean market dur-ing the crisis period. This pattern is an exception relative toother victim Asian markets. It is also shown in Table 4 thatthe Korean market responded strongly to shocks fromother markets, particularly after the crisis. Overall, theKorean market appears to be a fairly endogenousmarket, which has not been documented in the previousliterature.With respect to Malaysia, several markets, such as India,

Indonesia, Philippines, Pakistan, and Thailand, had sub-stantial responses to shocks from Malaysia in the pre-crisisperiod. While Singapore showed little response to shocksfrom Malaysia in the pre- and post-crisis periods, Malaysia

482 J. Yang et al.

Table 2. Trace tests for cointegration in US dollar terms

Panel A: Without linear trend

Number ofcointegrating Pre-crisis Crisis Transition Post-crisis Critical valuevectors period period period period (5%)

r ¼ 0 350.36 369.12 339.87 387.76 338.09r ¼ 1 266.72 295.74 269.07 292.39 289.70r ¼ 2 212.22 233.02 215.01 234.54 244.56r ¼ 3 160.43 182.69 169.18 178.64 203.34r ¼ 4 124.22 140.29 129.29 133.16 165.73r ¼ 5 95.51 105.97 92.09 101.24 132.00r ¼ 6 70.57 79.55 64.65 72.69 101.83r ¼ 7 51.96 57.51 46.24 49.40 75.73r ¼ 8 35.15 38.36 31.70 29.90 53.42r ¼ 9 24.69 23.38 17.65 19.50 34.79r ¼ 10 15.07 11.97 8.23 11.29 19.99r ¼ 11 7.04 4.25 2.72 4.45 9.13

Panel B: With linear trend

Number ofcointegrating Pre-crisis Crisis Transition Post-crisis Critical valuevectors period period period period (5%)

r ¼ 0 325.51 351.88 331.45 380.29 323.93r ¼ 1 242.60 278.94 261.64 284.92 276.36r ¼ 2 190.31 216.35 208.90 227.32 232.60r ¼ 3 144.52 166.65 163.09 171.91 192.30r ¼ 4 110.73 125.51 123.55 127.54 155.74r ¼ 5 82.99 91.44 86.37 96.43 123.03r ¼ 6 61.71 66.14 58.97 67.88 93.91r ¼ 7 43.99 44.20 42.10 45.68 68.86r ¼ 8 28.53 27.67 27.66 26.60 47.20r ¼ 9 18.42 15.18 15.14 16.26 29.37r ¼ 10 9.43 4.97 5.72 8.48 15.34r ¼ 11 1.88 0.35 0.32 3.16 3.84

Dow

nloa

ded

by [

91.1

80.1

09.1

0] a

t 09:

39 0

2 M

ay 2

014

responded to shocks from Singapore in these periods. it isinferred that Singapore is more dominant from an infor-mational perspective than Malaysia. This finding is consis-tent with Janakiramanan and Lamba (1998) but contraryto Dekker et al. (2001). The influence of the Malaysianmarket decreased in the post-crisis period, affecting onlythe Indian and Indonesian markets, perhaps due to itsweakened economy after the crisis.India, Indonesia, Pakistan and Thailand were responsive

to shocks from the Philippines market before the crisis and,with the exception of Pakistan, remained responsive afterthe crisis. It is also obvious that the Philippines market hadlittle response to shocks from other markets. Overall, likeDekker et al. (2001), it is found that the Philippines marketwas fairly isolated in non-crisis periods. Also, all otherAsian emerging markets under study (but not the USAand Japan) responded to shocks from the Philippines mar-ket. Such an increased influence on other markets duringthe crisis period is similar to what most other victim mar-kets experienced.Another market identified in previous studies (Ghosh et

al., 1999; Dekker et al., 2001) as relatively isolated isThailand. However, it is found that: Pakistan, India,Indonesia, Korea, Philippines, and Taiwan, significantlyresponded to shocks from Thailand before the crisis;India, Indonesia, Korea, Philippines, and Singaporeresponded to shocks from Thailand after the crisis; andall markets, except Taiwan and the USA, sharplyresponded to the shocks from Thailand during the crisis.These findings imply that Thailand played an importantrole in Asian markets, especially during the crisis.The Hong Kong market is believed to be the most inter-

active and influential market in the Asian region (Masihand Masih, 1999; Dekker et al., 2001). The results reveal

that only Pakistan and Taiwan responded to shocks fromHong Kong in the pre-crisis period. However, the HongKong market does become a more influential market afterthe crisis, with India, Indonesia, Korea, Pakistan, andSingapore all exhibiting strong reactions to shocks fromthis market. As expected, compared to the non-crisis peri-ods, more Asian markets are significantly responsive toinnovations in the Hong Kong market during the crisis(i.e., Indonesia, Japan, Malaysia, Philippines, Pakistan,Singapore, Thailand, and Taiwan, all of which are victimmarkets except Japan and Pakistan).Table 4 shows that the Singapore market is an influ-

ential market in the Asian region. India, Indonesia,Malaysia, Philippines, Pakistan, and Thailand had strongreactions to shocks from the Singapore market beforethe crisis, while Hong Kong, India, Indonesia, Korea,Malaysia, and Thailand responded to innovations fromthis market after the crisis. Like most other victimmarkets, the Singapore market became more influentialmarket during the crisis, with strong reactions fromHong Kong, India, Japan, Malaysia, Philippines,Pakistan, Thailand and Taiwan. Thus, it can be inferredthat the most integrated markets during the crisis arevictim markets.Similar to the case of Korea, few markets, with the

exception of Thailand, responded significantly to a shockfrom Taiwan before the crisis. Such a pattern is alsoobserved after the crisis, with the only exception ofKorea responding to the shock from Taiwan. This suggeststhat Taiwan had little influence on other markets in thenon-crisis periods. Consistent with Ghosh et al. (1999)and Dekker et al. (2001), it is found that Taiwan normallyis an isolated market. However, it is found that most mar-kets, except non-victim countries such as India, Japan and

Stock market integration and financial crises 483

Table 3. Market exclusion tests

Likelihood ratio test statistic (p-value)

Local currency US dollarsExcludedmarket Period 2 Period 3b Period 1 Period 2 Period 3a Period 3b

HK 4.70 (0.094) 16.90 (0.000)* 4.55 (0.032)* 6.71 (0.034)* 0.67 (0.411) 33.51 (0.000)*ID 16.63 (0.000)* 9.14 (0.001)* 11.80 (0.000)* 2.01 (0.364) 0.57 (0.446) 22.31 (0.000)*IN 11.27 (0.003)* 9.59 (0.008)* 17.21 (0.000)* 11.43 (0.003)* 3.17 (0.074) 13.16 (0.001)*JP 12.83 (0.001)* 15.63 (0.000)* 6.77 (0.009)* 7.62 (0.022)* 5.28 (0.021)* 38.48 (0.000)*KR 29.12 (0.000)* 13.41 (0.001)* 29.50 (0.000)* 2.76 (0.250) 0.22 (0.635) 19.68 (0.000)*ML 16.62 (0.000)* 10.28 (0.005)* 2.08 (0.148) 13.58 (0.001)* 0.39 (0.528) 14.29 (0.000)*PH 19.04 (0.000)* 3.05 (0.217) 29.38 (0.000)* 1.07 (0.584) 11.43 (0.000)* 14.02 (0.000)*PK 15.51 (0.000)* 15.29 (0.000)* 3.64 (0.056) 6.07 (0.047)* 3.23 (0.072) 8.05 (0.017)*SG 11.03 (0.004)* 12.90 (0.001)* 25.35 (0.000)* 6.01 (0.049)* 1.18 (0.275) 0.37 (0.830)TH 37.99 (0.000)* 8.03 (0.01)* 21.69 (0.000)* 16.23 (0.000)* 0.87 (0.350) 2.58 (0.274)TW 11.29 (0.003)* 4.33 (0.114) 0.05 (0.815) 14.15 (0.000)* 5.19 (0.023)* 1.35 (0.507)US 17.90 (0.000)* 11.30 (0.003)* 0.14 (0.701) 15.17 (0.000)* 12.76 (0.000)* 16.49 (0.000)*

Note: For Period 1 and Period 3a in US dollars terms, the test is performed given one cointegrating vector. For other periods the test isperformed given two cointegrating vectors. An asterisk denotes the rejection of the null hypothesis at the 5% significance level.

Dow

nloa

ded

by [

91.1

80.1

09.1

0] a

t 09:

39 0

2 M

ay 2

014

484 J. Yang et al.

Table 4. Impulse responses by country to one standard deviation innovations in each stock market (day 20)

Responses by country

Sub-period HK ID IN JP KR ML PH PK SG TH TW US

Indonesia (IN)(a) 0.27 0.29 0.63 0.14 0.11 0.30 0.26 0.50 0.19 �0.01 0.07 0.28(b) 0.47 0.20 0.75 0.18 �0.06 0.68 0.23 0.40 0.46 0.32 0.19 0.03(c) 0.31 0.18 0.98 0.27 0.48 0.27 0.24 0.20 0.45 0.33 0.15 0.07

Korea (KR)(a) �0.08 0.18 �0.14 0.19 0.48 �0.20 �0.23 0.05 �0.21 0.06 �0.18 0.07(b) 0.08 0.08 0.10 0.08 0.89 0.10 0.01 0.19 0.07 0.54 0.20 0.06(c) 0.18 0.08 0.14 0.11 0.75 �0.05 0.06 �0.27 �0.01 0.20 0.24 �0.02

Malaysia (ML)(a) 0.01 0.41 0.49 0.05 0.10 0.42 0.49 0.77 0.10 0.29 0.18 0.00(b) 0.27 0.04 0.48 0.07 0.05 0.98 0.37 0.14 0.43 0.52 0.23 0.03(c) �0.07 0.26 0.29 0.01 �0.06 1.02 0.07 0.04 �0.11 0.10 0.10 �0.03

Philippines (PH)(a) 0.11 0.26 0.30 0.08 �0.05 0.16 0.57 0.29 0.00 0.21 �0.18 0.12(b) 0.46 0.30 �0.25 0.15 �0.21 0.24 0.67 0.64 0.50 0.33 0.32 0.04(c) 0.04 0.22 0.25 0.01 0.19 �0.01 1.04 �0.01 0.00 0.23 0.03 0.03

Thailand (TH)(a) �0.07 0.35 0.24 �0.03 0.30 0.08 0.24 0.56 0.11 0.59 �0.25 �0.08(b) 0.61 0.21 0.31 0.22 0.71 0.97 0.49 0.50 0.59 1.04 0.18 0.10(c) 0.17 0.43 0.27 0.15 0.60 0.07 0.33 0.18 0.31 0.94 �0.02 0.11

Hong Kong (HK)(a) 0.18 0.09 0.15 �0.19 �0.00 0.07 0.02 0.41 0.03 �0.07 �0.25 0.07(b) 0.89 0.18 0.22 0.20 0.02 0.52 0.32 0.40 0.57 0.26 0.25 0.02(c) 0.71 0.78 0.40 0.18 0.74 �0.09 0.16 0.25 0.26 0.16 0.01 0.05

Singapore (SG)(a) 0.11 0.56 0.61 0.05 �0.01 0.48 0.56 0.70 0.48 0.66 0.06 �0.01(b) 0.85 0.28 0.13 0.21 �0.01 0.80 0.43 0.70 0.97 0.53 0.36 0.05(c) 0.21 0.66 0.70 0.14 0.46 �0.20 0.06 �0.18 0.67 0.31 0.09 �0.05

Taiwan (TW)(a) 0.04 0.08 �0.04 �0.14 �0.08 0.06 0.04 0.04 �0.04 �0.23 0.52 �0.04(b) 0.38 �0.09 0.41 0.16 0.73 0.61 0.43 0.30 0.42 0.55 1.00 �0.02(c) 0.02 0.20 0.00 0.08 0.29 0.09 0.02 0.11 0.02 0.00 0.99 0.02

United States (US)(a) 0.73 0.36 0.72 0.58 0.00 0.44 0.82 0.15 0.29 0.32 �0.06 0.70(b) 1.02 0.34 0.40 0.58 0.53 0.39 0.60 0.47 0.71 1.19 0.63 0.89(c) 0.84 0.36 0.26 0.56 1.05 0.40 0.45 0.14 0.62 0.54 0.52 0.99

Japan (JP)(a) 0.01 0.15 0.21 0.62 0.06 �0.01 0.14 0.25 0.14 0.08 0.12 0.04(b) 0.38 �0.18 0.90 0.88 0.42 0.40 0.50 �0.17 0.41 0.67 0.15 0.00(c) �0.04 0.66 0.26 0.87 0.51 �0.04 �0.01 0.04 0.03 0.01 0.16 0.08

India (ID)(a) �0.22 0.59 �0.18 �0.07 0.11 0.03 0.11 0.02 �0.12 0.04 0.36 �0.12(b) 0.77 1.06 0.72 0.18 0.44 0.88 0.60 0.28 0.67 0.59 0.05 0.12(c) 0.19 1.06 0.00 0.16 0.33 0.11 0.12 0.02 0.14 0.10 0.27 �0.02

Pakistan (PK)(a) �0.14 0.06 �0.10 �0.05 �0.07 �0.13 �0.09 0.44 �0.06 �0.05 �0.15 �0.09(b) 0.22 0.06 0.23 0.01 �0.05 0.18 0.14 1.05 0.22 0.37 0.15 �0.02(c) 0.12 �0.17 0.08 �0.02 �0.26 0.13 0.02 0.77 �0.12 0.09 0.12 �0.03

Note: The three subperiods are denoted as follows: (a) pre-crisis period, (b) crisis period, and (c) post-crisis period. Countries areabbreviated as follows: HK (Hong Kong), ID (India), IN (Indonesia), JP (Japan), KR (Korea), ML (Malaysia), PH (Philippines),PK (Pakistan), SG (Singapore), TH (Thailand), TW (Taiwan), and US (United States).

Dow

nloa

ded

by [

91.1

80.1

09.1

0] a

t 09:

39 0

2 M

ay 2

014

the USA, significantly responded to shocks from Taiwanduring the crisis period.Examining the responses of the victim markets to shocks

from non-victim markets provides further evidence on howthe crisis affected linkages among the Asian markets. Table4 reports the size of responses to a US shock for each of theother countries. In general, all Asian markets significantlyresponded to US shocks throughout all the three periods.Responses to US shocks among most Asian markets wasintensified during the crisis; however, some countries suchas Japan, India, and Malaysia experienced no substantialchange in response to US shocks across the three differentsample periods. Conversely, the US market normally didnot respond to Asian markets in the three sample periods,which is consistent with previous literature (e.g., Masih andMasih, 1999; Sheng and Tu, 2000).The role of Japan as the leader in the Asian region has

been a contentious issue. While Ghosh et al. (1999) as wellas Masih and Masih (2001) argue that Japan is a marketleader, the present results suggest that it did not play apivotal role in the non-crisis periods. Only India andPakistan exhibited an appreciable response to the shocksfrom Japan before the crisis. Also, the Japanese market didnot become much more influential after the crisis, as onlyIndia, Indonesia, and Korea responded to shocks from thismarket. Thus, consistent with Masih and Masih (1997a,1999), Japan did not have much influence on other Asianmarkets in non-crisis periods. Also, Japan showed little orno response to shocks from other markets in such periods.It’s infered that Japan is a relatively isolated market undernormal market conditions, as found in earlier studies byDekker et al. (2001) as well as Bessler and Yang (2003).Once again, during the financial crisis, more Asian marketssignificantly responded to innovations in the Japanese mar-ket, including Hong Kong, Indonesia, Korea, Malaysia,Philippines, Singapore, and Thailand, all of which are vic-tim markets.Both India and Pakistan are small Asian emerging mar-

kets and not identified as victim markets. Only two markets(Hong Kong and Taiwan) before the crisis and another twomarkets (Korea and Taiwan) after the crisis responded toshocks from India. Thus, it is obvious that India has littleinfluence on other Asian markets under normal marketconditions. By contrast, during the crisis, eight victim mar-kets (Hong Kong, Indonesia, Korea, Malaysia,Philippines, Pakistan, Singapore, and Thailand) exhibitedstrong reactions to innovations from India. Lastly, Table 4reports the Asian markets’ responses to shocks from thePakistan market. No other markets showed significantresponses to shocks from this market in the non-crisisperiods, with the exception of Korea after the crisis.However, during the non-crisis periods, Pakistan, as wellas India, responded substantially to shocks from quite afew other markets, including the US, Singapore, HongKong, Thailand, Indonesia, and Philippine. In this sense,

India and Pakistan are not isolated from other markets; onthe contrary, they are quite endogenous as their stockmovements are to a large extent driven by other marketmovements during the non-crisis periods while the reversemay not hold. During the crisis, however, four victimmarkets, including Hong Kong, Indonesia, Singapore,and Thailand, significantly responded to innovationsfrom the Pakistan market.Overall, the crisis caused the victim markets to be more

responsive to external shocks from non-victim markets.Thus, an important new finding is that the crisis not onlyled markets affected by the crisis to be more integrated witheach other but also caused them to be more responsive tothe outside world.

V. SUMMARY AND CONCLUSIONS

This study examines the long-run relationship and short-run dynamics among the US, Japanese, and ten Asianstock markets, with the particular attention to the 1997–1998 Asian financial crisis. Extending related empiricalstudies, comparative analyses of pre-crisis, crisis, andpost-crisis periods are conducted to comprehensivelyevaluate how stock market integration is affected byfinancial crises. An error correction model (ECM) isemployed to estimate long-run relationships betweenmarkets, and generalized impulse response functionsare utilized to provide insights into short-run causaldynamic linkages among Asian and developed stockmarkets.In general, the empirical results reveal that long-run

cointegration relationships among these markets werestrengthened during the crisis and that these marketshave been more integrated after the crisis than before thecrisis. The results for the US and Japanese stock markets’impact on emerging Asian markets agree with the previousstudies (e.g., Masih and Masih, 1999; Bessler and Yang,2003) on the roles of these two markets in the internationalstock markets. The US substantially influenced the Asianmarkets in all three sample periods but was almost unaf-fected by the Asian markets. Conversely, Japan has little orno influence on the Asian markets except during the finan-cial crisis. Further empirical evidence indicates that Japan,Taiwan and Philippines are fairly isolated markets, whichis consistent with Dekker et al. (2001), among others.Unlike prior studies (e.g., Janakiramanan and Lamba,1998; Ghosh et al. 1999; Dekker et al. 2001), it is foundthat Indonesia and Thailand are integrated with severalother Asian markets during non-crisis periods, ratherthan being isolated markets. Korea, India and Pakistanappear to be fairly endogenous markets, which has notbeen documented in the previous literature. Also, HongKong is not as interactive in non-crisis periods as reportedin prior studies and, therefore, is not as influential as pre-

Stock market integration and financial crises 485

Dow

nloa

ded

by [

91.1

80.1

09.1

0] a

t 09:

39 0

2 M

ay 2

014

viously believed (e.g., Masih and Maish, 1999; Dekker etal. 2001). In this regard, Singapore appears to be a marketleader in the Asian region.

An important implication of the findings is that thedegree of integration among countries tends to changeover time, especially around periods marked by financialcrises. As Bekaert and Harvey (1995) have noted, previousresearch assumes that stock markets are either perfectlyintegrated, perfectly segmented, or partially integratedbut the extent of integration is constant over time. Basedon evidence gathered from regime-switching models, theyshowed that this assumption does not hold. Extending theirproposition to the case of the Asian financial crises, it isalso found that Asian stock market integration can be timevariant.

REFERENCES

Arshanapalli, B., Doukas, J. and Lang, L. (1995) Pre and post-October 1987 stock market linkages between US and Asianmarkets, Pacific-Basin Finance Journal, 3, 57–73.

Bekaert, G. and Harvey, C. R. (1995) Time-varying world marketintegration, Journal of Finance, 50, 403–44.

Bessler, D. A. and Yang, J. (2003) The structure of interdepen-dence in international stock markets, Journal of InternationalMoney and Finance, 22, forthcoming.

Brocato, J. (1994) Evidence on adjustments in major nationalstock market linkages over the 1980s, Journal of BusinessFinance & Accounting, 21, 643–67.

Chan, K. C., Gup, B. E. and Pan, M. (1992) An empirical analysisof stock prices in major Asian markets and the United States,Financial Review, 27, 289–307.

Cheung, D. (1997) Pacific Rim stock market integration underdifferent federal funds rate regimes, Journal of BusinessFinance and Accounting, 24, 1343–51.

Chung, P. and Liu, D. (1994) Common stochastic trends in PacificRim stock markets, Quarterly Review of Economics andFinance, 34, 241–59.

Corsetti, G., Pesenti, P. and Roubini, N. (1999) What caused theAsian currency and financial crisis? Japan and the WorldEconomy, 11, 305–73.

Dekker, A., Sen, K. and Young, M. (2001) Equity market in theAsia Pacific region: A comparison of the orthogonalized andgeneralized VAR approaches, Global Finance Journal, 12, 1–33.

DeFusco, R. A., Geppert, J. M. and Tsetsekos, G. P. (1996)Long-run diversification potential in emerging stock markets,Financial Review, 31, 343–63.

Ghosh, A., Saidi, R. and Johnson, K. (1999) Who moves theAsia-Pacific stock markets–US or Japan? Empirical evidencebased on the theory of cointegration, Financial Review, 34,159–70.

Hansen, H. and Juselius, K., (1995) CATS in RATS:Cointegration Analysis of Time Series, ESTIMA, Evanston,Illinois.

Hakkio, C. S. and Rush, M. (1991) Cointegration: how short isthe long-run? Journal of International Money and Finance, 10,571–81.

Hassapis, C., Pittis, N. and Prodromidis, K. (1999) Unit roots andGranger causality in the EMS interest rates: the GermanDominance Hypothesis revisited, Journal of InternationalMoney and Finance, 18, 47–73.

Hung, B. and Cheung, Y. (1995) Interdependence of Asian emer-ging equity markets. Journal of Business Finance andAccounting, 22, 281–88.

Janakiramanan, S. and Lamba, A. S. (1998) An empirical examin-ation of linkages between Pacific-Basin stock markets,Journal of International Financial Markets, Institutions andMoney, 8, 155–73.

Johansen, S. (1991) Estimation and hypothesis testing of cointe-gration vectors in Gaussian Vector Autoregressive Models,Econometrica, 59, 1551–80.

Kamin, S.B. (1999) The current international financial crisis: howmuch is new? Journal of International Money and Finance, 18,501–14.

Koop, G., Pesaran, M.H. and Potter, S.M. (1996) Impulseresponse analysis in nonlinear multivariate models, Journalof Econometrics, 74, 119–47.

Lutkepohl, H. and Reimers, H. (1992) Impulse response analysisof cointegrated systems. Journal of Economic Dynamics andControl, 16, 53–78.

Masih, A. M. M. and Masih, R. (1997a) A comparative analysisof the propagation of stock market fluctuations in alternativemodels of dynamic causal linkages, Applied FinancialEconomics, 7, 59–74.

Masih, A. M. M. and Masih, R. (1997b) Dynamic linkages andthe propagation mechanism driving major internationalstock markets, Quarterly Review of Economics and Finance,37, 859–85.

Masih, A. M. M. and Masih, R. (1999) Are Asian stock marketfluctuations due mainly to intra-regional contagion effects?Evidence based on Asian emerging stock markets, Pacific-Basin Finance Journal, 7, 251–82.

Masih, R. and Masih, A. M. M. (2001) Long and short termdynamic causal transmission amongst international stockmarkets, Journal of International Money and Finance, 20,563–87.

Ng, A. (2000) Volatility spillover effects from Japan and the US tothe Pacific-Basin, Journal of International Money andFinance, 19, 207–33.

Pesaran, M. H. and Shin, Y. (1998) Generalized impulse responseanalysis in linear multivariate models, Economics Letters, 58,17–29.

Phillips, P. (1998) Impulse response and forecast error varianceasymptotics in nonstationary VARs, Journal ofEconometrics, 83, 21–56.

Sheng, H. and Tu, A. (2000) A study of cointegration and vari-ance decomposition among national equity indices beforeand during the period of the Asian financial crisis, Journalof Multinational Financial Management, 10, 345–65.

Yang, J. and Leatham, D. J. (2001) Currency convertibility andlinkage between Chinese official and swap market exchangerates, Contemporary Economic Policy, 19, 347–59.

486 J. Yang et al.

Dow

nloa

ded

by [

91.1

80.1

09.1

0] a

t 09:

39 0

2 M

ay 2

014