Lovely Professional University,. Phagwara, Punjab, India. **Assistant Professor,. School of Business,. Lovely Professional University,. Phagwara, Punjab, India.
ISSN: 2249-7137
Vol. 3, Issue. 10, October 2013
ACADEMICIA
P ublis he d b y: S out h A s ia n A c ade m ic R es e arc h J our nals
ACADEMICIA: An International Multidisciplinary Research Journal ( A D o u b le B l i n d R e fe r r e d & R e v ie we d I nt e r na t io na l J o ur na l)
STOCK MARKET CO-INTEGRATION: AN INVESTIGATION OF SOUTH ASIAN COUNTRIES Amitesh Kapoor*; Harendra Singh** *Assistant Professor, School of Business, Lovely Professional University, Phagwara, Punjab, India. **Assistant Professor, School of Business, Lovely Professional University, Phagwara, Punjab, India. ABSTRACT This Report investigated the integration of Asian stock exchanges and indicated the diversification opportunities for potential investors they provide in the long-term. This study examines the stock market co-integration between India and south Asian countries, whether they are co integrated to each other or not which will eventually be helpful for the individual investors along with corporate investors in selecting their investment area and portfolio diversification. If the stock markets have any cointegration than diversification may not be profitable. This Report empirically analyzes the phenomenon of co-integration amongst selected South Asian stock markets. Augmented Dickey Fuller (ADF), Co-integration and Granger Causality tests are applied on the data. KEYWORDS: ADF, Granger Causality, Co-integration, Unit Root, Diversification. ______________________________________________________________________________ INTRODUCTION Portfolio diversification literature talks exceedingly of the stylized fact that diversification across the countries is more effective than diversification across the industries. This has major benefits involved for the investors in terms of risk management. No wonder then, the modern finance has been showing its concerns over the interdependence or linkages of world stock markets since the The South Asian Academic Research Journals http://www.saarj.com 19
ISSN: 2249-7137
Vol. 3, Issue. 10, October 2013
ACADEMICIA
financial crisis in 1987. Globalization, increased strategic economic relations among nations on regional basis and flow of information among stock markets has forced the international investors to rethink their decisions on diversified portfolio investment. If global financial markets exhibit long-term similar behavior then the true benefits of portfolio diversification would not be achieved. Therefore, Co-integration has become a vital area for research in order to understand price adjustment mechanism in stock markets. It provides a long-term measure of diversification opportunities between international financial markets as well as short-term deviations amongst within an error correction model. (Chari and Henry, 2004; Chaudhry, 1996; Cooray, 2004; Darrat and Benkato, 2003). According to Dellas and Hess, 2002; Flannery and Protopapdakis, 2003, strong economic ties and openness in trade among countries also influence the movement of stock prices in their respective markets. If stock prices across countries are moving together (or they are co- integrated) for the long period, it implies that each country‟s firm‟s exposure to the systematic risk is same (Chari and Henry, 2004). Stock markets of South Asian countries i.e. India, Pakistan, Srilanka, and Nepal have been selected for the sake of this study. The monthly index values of top stock markets of selected countries from August 20072012 are taken as data to extract the results. All the indices are denominated in local currencies. Nevertheless, the study in the emerging markets also gaining popularity among the researchers especially after the incident of the Asian turbulent time of Financial Crisis in 1997 by Masih and Masih (1999), Jang and Sul (2002). Other studies have examined the stock market integration by combining both developed and emerging countries, just name a few, from Moon (2001) and Elysian (1998). REVIEW OF LITERATURE Nath et al (2003) studied common Stochastic Trend and Co-Integration in the Emerging Markets A Case Study of India, Singapore and Taiwan.Iit was found that no co-integration between the stock market indices for the entire period and hence no long run equilibrium. They also found mild causality for some years in the study though most of the time these markets have not been interlinked. Narayan et al (2004) examined the dynamic linkages between the stock markets of Bangladesh, India, Pakistan and Sri Lanka. They used Granger causality test ,Unit root tests, Multivariate Co-integration test. Lamba (2005) observed a detailed, large sample analysis of the short- and long-run relationships between the South Asian markets of India, Pakistan and Sri Lanka and the major developed markets during July 1997 - December 2003. They found that the three South Asian equity markets are becoming more integrated with each other but at a relatively slow pace. . Rahman et al (2009) studied the interactions between stock prices and exchange rates in three emerging countries of South Asia named as Bangladesh, India and Pakistan. Srivastav et al (2010) studied Exchange Rates Movement and Stock Market Volatility to analyzes the relationship between Nifty returns and Indian rupee-US Dollar Exchange Rates and correlation between Nifty returns and Exchange Rates was found to be negative. Khan et al (2011) studied on Co-integration & Causality Analysis among Asian Stock Markets. His findings have strategic implications for the investors interested in diversifying their portfolios across various Asian countries. Iqbal and Khalid and Rafiq (2011) found out the dynamic linkages among the equity market of USA and emerging markets of Pakistan and India covering daily data from January 2003–December 2009. Subhani et al (2011) studied on “Are the Major South Asian Equity Markets Co-Integrated?” Khan et al (2011) studied on Co-integration and Causality The South Asian Academic Research Journals http://www.saarj.com 20
ISSN: 2249-7137
Vol. 3, Issue. 10, October 2013
ACADEMICIA
Analysis of Emerging Equity Markets and Developed Equity Markets to examine the possible integration of Karachi Stock Exchange (KSE 100) with other 28 indices of the world including regional and non-regional developed and underdeveloped countries‟ stock markets. Hooi et al (2011) examined Exchange Rate and Stock Price Interaction in Major Asian Markets: Evidence for Individual Countries and Panels Allowing for Structural Breaks with the purpose to examine the relationship between exchange rates and stock prices in eight Asian countries. Srikanth et al (2011) investigated on Individual Investor Trading and Return Patterns around Earnings Announcements to explore the long-term dynamic relationship exists between the Indian stock market and other selected stock markets in the Asia-Pacific region. Zia et al (2011) examined The Causality between Stock Market and Foreign Exchange Market of Pakistan. Parsva et al (2011) investigated the relation between stock returns and exchange rate for six Middle Eastern countries. Kisaka et al (2012) studied The Causal Relationship between Exchange Rates and Stock Prices in Kenya using Unit Root and Co- integration Tests, Error Correction Model and Granger-causality, Diagnostic Tests, Heteroscedasticity test, the F-test which indicate that exchange rates Granger-causes stock prices in Ken. Saha and Bhunia (2012) investigated Financial Market Integration of South Asian Countries. Harper and Jin and Gleghorn (2012) studied whether the stock markets move together. SCOPE OF STUDY This study examines the stock market Co-integration between India and South Asian countries whether they co-integrated to each other or not which will be insightful for the individual investors and corporate investors in selecting their investment area and portfolio diversification. If the stock markets have any Co-integration than diversification may not be profitable otherwise diversification provides significant profit. This study may use for further research purpose also. HYPOTHESIS CONSTRUCTION 1) Unit root test hypothesis Ho: Variable is stationary or got unit root H1: Variable is not stationary 2) Co integration test: for the BSE and SOUTH ASIAN Nations Ho: There is co integration between the variables H1: There is no Co-integration between the variables 3) Co integration test: for the BSE with individual SOUTH ASIAN Nation Ho: There is Co-integration exist in both the countries H1: There is no Co-integration exist in both the countries
The South Asian Academic Research Journals http://www.saarj.com 21
ISSN: 2249-7137
Vol. 3, Issue. 10, October 2013
ACADEMICIA
4) Co integration between the BSE and Developed countries Ho: There is Co-integration between the BSE & developed Nation SEs H1: There is no Co-integration between the BSE & developed Nations SEs 5) Granger causality Test Hypothesis Ho: Acceptance of null hypothesis that there is no causal relationship among the given variables. H1: Rejected of null hypothesis that there is causal relationship among the given variables RESEARCH METHODOLOGY & DATA Co-integration process is used to investigate any kind of relationship or link between the stock markets in pairs separately. This process will tell about whether a stock market follows some markets more and to some it follows less. The extant of interdependence varies from stock market to stock market. The first body of research focuses on using the Johansen test towards finding Co-integration across the various South Asian stock markets. In this basic model, one regresses the stock market index price of one country against that of the other. If the residuals obtained from the regression are stationary, then a long-run relationship exists between the two countries, or in other words, the stock markets of the countries are co-integrated. Therefore, before I could proceed with the tests of Co-integration, I have to make sure that the Series were non-stationary and hence integrated of order 1. I ran the Augmented Dickey Fuller Tests on the series and the differenced series to confirm that the series were indeed I (1). I used The Schwartz Information Criterion (SIC) (for lag selection as it seems to be the criterion of choice in most studies. Then follows the ADF test to check the stationarity and non stationarity, the Johansen procedure regresses one series against the other then runs an augmented Dickey-Fuller test on the residuals to check if they are stationary. If the residuals are stationary, then it can be assume that the variables are co-integrated. For each time series, the null hypothesis is that there is no cointegration and the alternate hypothesis is that there is one or greater than one co-integrating relationship. If we reject the null, then the next null hypothesis will be that there is one cointegrating relationship and the alternative will be that there are 2 co-integrating relations. After the Johansen test, I moved on to the Gregory-Hansen test, which incorporates Structural breaks into the relationship. Gregory and Hansen explain that it is possible that Co-integration might hold over some (fairly long) period of time, and then shift to a new Long-run relationship. And this test will also use to check Co-integration if Johanen test failed to find out the Co-integration. For the purpose of above-mentioned methodology we have taken the last five years equity data of various stock exchanges. RESULTS AND FINDINGS We have taken the last five-year time series data of Sensex and other south Asian nations like Nepal, Srilanka, and Pakistan. Then apply the correlogram chart to see whether the series have The South Asian Academic Research Journals http://www.saarj.com 22
ISSN: 2249-7137
Vol. 3, Issue. 10, October 2013
ACADEMICIA
any kind of trend of not and after taking the consideration of correlogram, we have applied the ADF test i.e. augmented dickey fuller test to checking the stationary or non stationary for further tests like co integration test, if there is variable is non stationary then only we can apply co integration test. Now in our case when the p < 5% then we reject the null hypothesis and if p > 5% then we have to accept the null hypothesis means that our data is non-stationary. In first case that is intercept only in the T-2 we can see that p value > 5% level of significance and consecutively other T-3 and T-4 P value > than 5% meaning that we have to accept our null hypothesis and we cannot reject the null hypothesis means the data is Non stationary. To check the level of correctness of test we must check the test statistic value and the guideline for this test value is that if the ADF test statistic value is less than Test critical values i.e. statistic value > test critical values. For consideration of values we take only absolute values and ignore negative sign. Make sure that the coefficient of variables must be negative in sign otherwise there is no any kind of value to the test. HERE IS THE SUMMARY TABLE OF CO-INTEGRATION TEST: TRACE TEST
The South Asian Academic Research Journals http://www.saarj.com 23
ISSN: 2249-7137
Vol. 3, Issue. 10, October 2013
ACADEMICIA
INTERPRETATION As we can see in the table for Co-integration between the South Asian nations at None CEs (cointegrated equation) P < 5% i.e. 0.0010 which shows that there is Co-integration between them but when we take the consideration of at most 1 and 2, P > 5% level and TV < CV i.e. 11.95 < 15.49 value, which indicate that there is “No Co-integration between South Asian nations”. Now when we applied for individual countries with the SENSEX the result come out as follows: SENSEX is co integrated with the NEPSE and SENSEX and KSE also have co-integration that is, TV>CV or P
