International Research Journal of Finance and Economics ISSN 1450-2887 Issue 16 (2008) EuroJournals Publishing, Inc. 2008 http://www.eurojournals.com/finance.

htm

Long-Run Relationships between an Emerging Equity Market and Equity Markets of the Developed World an Empirical Analysis of Karachi Stock Exchange
Arshad Hasan Muhammad Ali Jinnah University Islamabad H.M.N. Saleem Assistant professor, Islamia University Bahawalpur M. Shoaib Abdullah Assistant professor, International Islamic University, Islamabad

Abstract
This paper examines the long term relationship between Karachi stock exchange and equity markets of developed world for the period 2000 to 2006 by using multivariate Cointegration analysis. Johansen and Juselius multivariate Cointegration analysis indicates that markets are integrated and there exist a long term relationship between these markets. However pair wise Cointegration analysis shows that Karachi stock market is not cointegrated with equity market of US, UK, Germany, Canada, Italy and Australia. Therefore funds managers of US, UK, Germany, Canada, Italy and Australia can get the benefits of portfolio diversification by investing in the Karachi stock market. However, Karachi stock exchange is found to be integrated with France and Japan. Impulse response analysis and variance decomposition analysis indicate that the Karachi stock market is in general independent as most of its shock is explained by its own innovations whereas US and UK markets are exerting some impact on Karachi. Field: Finance

1. Introduction
Globalization is leading the world in a new direction. Financial world is reshaping itself. New market structures and practices are need of time due to financial liberalization and elimination of traditional regulatory barriers and advancement of technology. We are marching towards a globally integrated financial world. Emerging equity markets are attracting the attention of global fund managers because these offer opportunity for portfolio diversification. The benefits and costs of international portfolio diversification need to be considered by anyone holding a financial Portfolio. Similarly, the firm that is considering raising new resource needs to address the requirements of the global marketplace. During last two decades, globalization is most visible feature in financial markets. In globally integrated financial markets, investors and policy makers are interested in taking advantage of efficiency enhancing aspects of market interaction. An efficient monitoring and controlling of this process of market interaction is necessary to avoid the undesirable destabilizing effects.

International Research Journal of Finance and Economics - Issue 16 (2008)

53

Interdependence of markets, businesses, regions, and continents has increased the interest of academician exploring international market linkages. The globalization of economic activity, the increased world wealth, and the reduction in transaction costs associated with the information revolution all direct investors to consider the newly emerging financial markets. Interaction of financial markets is one of the most extensively discussed topics of financial literature. Various factors contributed in this dimension. These include cross border movement of funds, the technological innovations in communications, scientific trading and settlement systems, and the introduction of innovative financial products. Globalizations also played a pivotal role in increase the interest in the study of dynamic inter-linkages among financial markets. So this study is aimed to understand the dynamic inter-linkages between Karachi stock exchange and the equity markets of developed world These countries include Pakistan, USA,UK, France Germany, Japan Canada Italy and Australia. If these markets are not integrated then benefits of portfolio diversification can be reaped by investors of these countries. Moreover, policy makers and regulators in these countries should not worry about any contagious effects. The paper is organized into four sections. Section II provides a brief overview of empirical literature on interlink ages and interaction of equity markets. Section III describes data and methodology adopted in the study. Section IV discusses the empirical results. Section V concludes the results.

2. Literature Review
Kasa (1992) examines the equity markets of the Japan, Germany Canada U.S. and U.K. to identify the common stochastic trends in monthly and quarterly time series for the period 1974 - 1990 by employing Johansen cointegration technique. Results provide evidence about the existence on one cointergrating vector that drives these equity markets. Therefore, there exists a long run relationship between the equity markets of these countries. Roca (1999) investigates the interlinkages among equity markets of Japan, Korea, U.S., U.K., Singapore, Taiwan, Australia and Hong Kong by employing Johansen cointegration technique. He uses weekly equity prices to determine the long run relationship among equity markets. Granger causality test has also been used to find the direction of causality. His results reveal that no cointegration exists between Australia and other markets. However, Australian market is found significantly influenced by U.S. and U.K. markets. Lamba (2005) performs a comprehensive large sample analysis to investigate the presence of long run relationships among South Asian equity markets and the developed equity markets for the period 7/1997 12/2003 by using multivariate cointegration framework. Results reveal that Indian markets is influenced by developed equity market of US, UK and Japan. However Pakistani and Sri Lankan equity markets found relatively independent from the influence of equity markets of developed markets during the entire sample period. He also argues that the three South Asian equity markets are becoming more integrated with each other but at a relatively slow pace. Suchismita (2005) examines the dynamic interaction among Asian equity markets and equity market of US with a special emphasis on Indian equity market. Results reveal that Indian equity market is integrated with Asian and the US markets. Asian markets are in general influenced by Japanese and U.S. equity markets and Indian market is not exception. Further, Indian equity market is also found influencing stock returns in some important Asian markets. Though Indian market is characterized by restrictions on capital flows but still it is segmented with the group of Asian markets. However, degree of integration between the Indian equity and equity markets of the Asian region is not very high so possibility of portfolio diversification still exists. Aggarwal (2003) examine the integration of the three participating equity markets before and after the 1993 passage of NAFTA based on daily, weekly, and monthly data for seven years before and after the passage of NAFTA (1988-2001), unit root tests for the overall period 1988-2001 and the two sub-periods, 1988-1993 (pre-NAFTA) and 1994-2001 (post-NAFTA), indicate that stock prices are

54

International Research Journal of Finance and Economics - Issue 16 (2008)

non-stationary but stock returns are generally stationary for all three markets for all three periods. However, daily, weekly, and monthly equity prices in the three NAFTA countries are cointegrated only for the post-NAFTA period. Similarly, US stock prices are more integrated with both Canadian and Mexican stock prices after the passage of NAFTA. Narayan, Smyth and Nandha (2004) examine the dynamic linkages between the stock markets of Bangladesh, India, Pakistan and Sri Lanka using a temporal Granger causality approach by binding the relationship among the stock price indices within a multivariate cointegration framework. He also examines the impulse response functions In the long run, stock prices in Bangladesh, India and Sri Lanka Granger-cause stock prices in Pakistan. In the short run there is unidirectional Granger causality running from stock prices in Pakistan to India, stock prices in Sri Lanka to India and from stock prices in Pakistan to Sri Lanka. According to them Bangladesh is the most exogenous of the four markets, reflecting its small size and modest market capitalization. Naeem (2000) examines the interlinkages among South Asian equity markets and equity markets of United States and United Kingdom for the period 1/94 to 12/99. Monthly stock market indices of Pakistan, India, Sri Lanka, Bangladesh, United States and United Kingdom has been investigated by using bivariate and multivariate cointegration analysis. Results reveal that no long term relationship exists among these markets in full sample period. However, in pre-nuclear test period cointegration is observed. It is worth mentioning that south Asian markets are not cointegrated with equity markets of the United States and United Kingdom. So potential for diversification of portfolio exists by structuring a portfolio that comprises of investment in equity markets of U.S. or U.K. and any one of the South Asian equity market Glezakos, Merika and Kaligosfiris(2007) examines the short and long-run relationships of between Greek stock exchange and major world financial markets by using Cointegration analysis and granger causality test. Monthly data for 10 countries is used.. The results reveal dominance of the USA financial market and the strong influence of DAX and FTSE on all other markets of the sample. The influence of Germany and the DJ index is especially noticeable on the Athens stock exchange. Numerous numbers of studies has tested the relationships among international stock markets, however, most of these studies focused on highly developed stock markets or East Asian markets. In the current changing economic scenario of Pakistan, a comprehensive study to explore the relationship of Karachi stock exchange with equity markets of developed world is required to explore the opportunity of portfolio diversification by international funds managers.

3. Data Description and Methodology

This empirical study is based on weekly closing values of the stock market indices of 09 important equity markets of the developed world. Weekly closing prices of KSE-100 index (Pakistan), S&P500 Index (USA), FTSE 100 (UK) index, CAC 40 index (France), DAX index (Germany), Nikkei 225 index (Japan), TSX Index(Canada), MIB30 index (Italy) and AORD index (Australia) for the period Jan 1,2000 to Dec 30, 2006 has been taken from Yahoo Finance. The continuously compounded rate of return is calculated by using the following formula Rt = ln (Pt/Pt-1) Where: Rt = Return on day t; Pt = Index closing value on week t Pt-1 =Index closing value on week t-1 ln= Natural log. There are several methods for testing the flow of information and co-movement of prices in stock markets across the countries. In this study the emphasis is given to test the inter-market relationship among the stock market in Pakistan with that of equity markets of developed world, via; (i). Descriptive statistics; (ii). Correlation matrix, (iii) Co integration tests, and (iv) Granger causality test

International Research Journal of Finance and Economics - Issue 16 (2008)

55

Cointrgration analysis requires that time series should be integrated of same order. Stationarity of time series has been examined by using unit root tests. Augmented Dickey-Fuller Test and PhillipsPerron Test have been employed for said purpose. The Augumented Dickey Fuller test examines the presence of unit root in an autoregressive model. A simple AR(1) model is yt = yt 1 + ut, where yt is the variable of interest, t is the time index, is a coefficient, and ut is the disturbance term. The regression model can be written as yt = (1)yt1 + ut = yt 1 + ut, where is the first difference operator. This model can be estimated and testing for a unit root is equivalent to testing. = 0. The Dickey-Fuller tests assume that the error terms are statistically independent and have a constant variance which is rather strict assumption so an alternative test, the Phillip-Perron test, is also used to test the stationarity of data. It may be noted that Phillip-Perron test allows the error disturbances to be weakly dependent and heterogeneously distributed. yt= o + 1 yt-1 + t {t- T/2} + ut Test statistics for the regression coefficients under the null hypothesis that the data are generated by yt = yt-1 + ut, where E(ut) = 0. A financial time series is said to be integrated of one order i.e, I (1), if it becomes stationary after differencing once. If two series are integrated of order one, there may have a linear combination that may be stationary without differencing. If said condition fulfills then these are called cointegrated. Cointegration analysis can be conducted by using residual based Engle-Granger (1987) test, or maximum likelihood based Johansen (1988; 1991) and Johansen-Juselius (1990) tests. The Johansen (1988) and Johansen and Juselius (1990) procedure tests the presence of long run relationship between the variables. Johansen and Juselius propose two likelihood ratio tests for the determination of the number of cointegrated vectors. One is the maximal eigenvalue test which evaluates the null hypothesis that there are at most r cointegrating vectors against the alternative of r + 1 cointegrating vectors. The maximum eigen value statistic is given by, max = - T ln (1 - r+1) where r+1,,n are the n-r smallest squared canonical correlations and T = the number of observations. The second test is based on the trace statistic which tests the null hypothesis of r cointegrating vectors against the alternative of r or more cointegrating vectors. This statistic is given by trace = -T ln (1 - i) In order to apply the Johansen procedure, a lag length must be selected for the VAR. A lag length of is selected on the basis of the Akaike Information Criterion (AIC). If Johnston and Juselius test of co-integration confirms that both the price series are cointegrated in the long run, then the system of equations should be modified by inserting an Error Correction Term to account for the short-run divergence of prices from their respective equilibrium values. Granger representation theorem provides that if two variables are cointegrated then Grangercausality must exist in at least one direction, which is a consequence of the relationships described by the ECM. Error Correction Model enables us to capture both the short-run dynamics and long-run relationships between the indices. Hence, the temporal Granger-causality between the variables can be investigated by applying a joint F-test to the coefficients of each explanatory variable in the VECM. Granger causality test captures lead lag relationship within the sample period so Variance decomposition analysis is also used to for out-of-sample causality tests. It partitions the variance of the forecast error of a certain variable into proportions attributable to shocks in each variable in the system. In other words Variance decomposition analysis literally provides a breakdown of the changes in the value of the variable in a given period arising from changes in the same variable in addition to other variables in previous periods. The impulse response analysis investigates the influence of random shock on the markets Impulse responses of returns in various markets to a shock in their own and other market innovations are also examined. Impulse responses show the effect of shocks for different days separately. To see

56

International Research Journal of Finance and Economics - Issue 16 (2008)

the cumulative effect of shocks the variance decomposition is a better tool. The variance decomposition of the stock indices is based on the analysis of responses of the variables to shocks. When there is a shock through the error term we study the influence of this shock to the other variables of the system and thus get information about the time horizon and percentage of the error variance.

4. Empirical Results
Descriptive statistics for the stock indices returns are given in Table 1. These include the distribution of mean, standard deviation, skewness and kurtosis etc. A careful examination reveals that Karachi stock exchange offers the highest return at a reasonable risk level. Approximately at the same level of risk German market is offering negative returns. Stock markets of USA, UK, Germany, Japan and Italy are exhibiting average negative returns for the period under study. Italian market has the worst performance. Australian markets appears to be less risky whereas standard deviation of other markets is not significantly different All of the markets are negatively skewed.
Table 1: Descriptive Statistics
RtnKSE 0.0054 0.0093 0.0367 0.0013 2.6562 -0.6341 0.3041 -0.1762 0.1280 RtnS&P -.00005 0.0013 0.0239 0.0006 3.5896 -0.6788 0.1920 -0.1171 0.0749 RtnCAC .0000 0.0018 0.0275 0.0008 1.6672 -0.3533 0.2121 -0.1213 0.0908 RtnDAX -.00008 0.0034 0.0328 0.0011 1.3892 -0.3463 0.2612 -0.1392 0.1220 RtnFTSE -0.0001 0.0021 0.0209 0.0004 1.7906 -0.4946 0.1582 -0.0886 0.0695 RtnNikkei -0.0002 0.0024 0.0280 0.0008 0.5064 -0.3075 0.2076 -0.1129 0.0947 RtnTSX 0.0012 0.0037 0.0229 0.0005 3.1976 -0.5552 0.2038 -0.1107 0.0931 RtnMIB30 -0.0007 0.0024 0.0281 0.0008 22.998 -2.8795 0.3254 -0.2643 0.0611 RtnAORD 0.0018 0.0026 0.0148 0.0002 1.3286 -0.4704 0.1019 -0.0583 0.0436

Mean Median Std Deviation Variance Kurtosis Skewness Range Minimum Maximum

Table 2 below presents the result of correlation analysis and indicates that Karachi stock exchange is not correlated with the equity markets of the developed world. S& P 500 appears to be strongly correlated with the indices of European countries like UK, France and Germany. Nikkei 225 is weakly correlated the western markets. On the other hand, strong correlation exists among European markets. It may be due to free flow of fund within EU. It appear to a result of financial and economic integration of Europe
Table 2: Correlation Matrix
RtnKSE 1.0000 0.1446 0.1021 0.1289 0.0700 0.0643 0.1554 0.0945 0.1085 RtnS&P 1.0000 0.7539 0.7505 0.7247 0.2946 0.6909 0.6047 0.5313 RtnCAC 1.0000 0.8870 0.8301 0.3967 0.5810 0.7554 0.5501 RtnDAX RtnFTSE RtnNikkei RtnTSX RtnMIB30 RtnAORD

Rtn-KSE Rtn-S&P Rtn-CAC Rtn-DAX Rtn-FTSE Rtn-Nikkei Rtn-TSX Rtn-MIB30 Rtn-AORD

1.0000 0.7902 0.3955 0.5745 0.7341 0.5491

1.0000 0.3394 0.5224 0.6477 0.5358

1.0000 0.3558 0.3380 0.4800

1.0000 0.4717 0.5186

1.0000 0.4671

Correlation analysis is weak technique as it does not discuss the cause and effect relationship. In order to take a better picture of the affairs we perform Cointegration analysis that tests the flow of information and co-movement of prices in stock markets across the countries. Cointegration analysis requires time series should be integrated of same order so in first step stationarity of the index series has been tested. Tables 3 reveals that that the time-series are not

International Research Journal of Finance and Economics - Issue 16 (2008)

57

stationary at levels. However first differences of the logarithmic transformations of the series are stationary. So, we say that the series are integrated of order one I(1)
Table 3: Unit Root Analysis
ADF (Level) -0.0757 -1.2993 -1.4081 -1.5210 -1.0044 -1.9268 0.2770 -1.8611 1.9612 -3.4511 -2.8700 -2.5713 ADF (Ist Difference) -8.3037 -9.0714 -8.4555 -8.1186 -8.9080 -7.6430 -7.3218 -9.0830 -8.1926 -3.4511 -2.8700 -2.5713 Phillips-Perron (Level) -0.0074 -1.4809 -1.0037 -1.1346 -1.3259 -1.4036 0.1059 -1.4839 1.8775 -3.4508 -2.8699 -2.5712 Phillips-Perron (Ist Difference) -17.0046 -20.9641 -20.0456 -19.3405 -19.8007 -19.2954 -20.6675 -16.5584 -19.4267 -3.4509 -2.8699 -2.5712

KSE 100 S&P 500 CAC 40 DAX FTSE 100 Nikkei 225 TSX MIB 30 AORD Critical Values 1% 5% 10%

Dickey-Fuller tests assumes that the errors are statistically independent and have a constant variance. This may not be the case with some of the data used here so the less restrictive PhillipsPerron test is used to test the stationarity of a time-series. Tables 3 present the results for the variables in levels and first difference. Phillips-Perron test confirms the results that the series are I(1) so we proceed with Cointegration tests. Maximum likelihood based Johansen (1988; 1991) and Johansen-Juselius (1990) procedure is used to determine the presence of cointegrating vectors in a set of non stationary time series. The first test is based on the trace statistic which tests the null hypothesis of r cointegrating vectors against the alternative of r or more cointegrating vectors. Table 4 present the results of multivariate Cointegration analysis. Trace test indicates 2 cointegrating equations at the 0.05 level. It means markets are integrated
Table 4:
KSE 100 S&P 500 CAC 40 DAX FTSE100 Nikkei 225 TSX MIB30 AORD

Multivariate Cointegration Analysis Trace Statistics

Hypothesis r = 0* r 1* r2 r3 r4 r5 r6 r7 r8 Eigen value 0.1707 0.1534 0.0950 0.0586 0.0540 0.0349 0.0268 0.0244 0.0006 Trace statistics 228.8356 163.7026 105.7349 71.0061 49.9939 30.6595 18.2854 8.8317 0.2215 Critical value 5% 197.3709 159.5297 125.6154 95.7537 69.8189 47.8561 29.7971 15.4947 3.8415 Remarks

Trace test indicates 2 cointegrating equations at the 0.05 level

Then maximal eigenvalue test is used to test the null hypothesis that there are at most r cointegrating vectors against the alternative of r + 1 cointegrating vectors. Table 5 reports the result of Max-eigenvalue test. It also confirms the presence of 2 cointegrating eqn(s) at = 0.05

58
Table 5:
KSE 100 S&P 500 CAC 40 DAX FTSE100 Nikkei 225 TSX MIB30 AORD

International Research Journal of Finance and Economics - Issue 16 (2008)

Multivariate Cointegration Analysis Max-Eigen Value Statistics
Hypothesis r = 0* r 1* r2 r3 r4 r5 r6 r7 r8 Eigen value 0.1707 0.1534 0.0950 0.0586 0.0540 0.0349 0.0268 0.0244 0.0006 Max-Eigen statistics 65.1330 57.9677 34.7289 21.0122 19.3345 12.3740 9.4537 8.6102 0.2215 Critical value 5% 58.4335 52.3626 46.2314 40.0776 33.8769 27.5843 21.1316 14.2646 3.8415 Remarks

Two cointegrating equations at the 0.05 level

Now we examine whether these markets are pair wise co integrated with each other for the period under study. The Table 6 presents the results of the pair wise cointegration tests for the entire sample period. We find that Karachi stock market is not cointegrated with equity market US, UK, Germany, Canada, Italy And Australia. However markets are found integrated with France and Japan. Therefore funds managers of US, UK, Germany, Canada, Italy And Australia can get the benefits of portfolio diversification by investing in the Karachi stock market. However European markets and us markets indicate pair wise Cointegration.
Table 6: Bivariate Co Integration Analysis
Hypothesis r=0 r1 r=0 r1 r=0 r1 r=0 r1 r=0 r1 r=0 r1 r=0 r1 r=0 r1 Eigen value 0.020816 0.011343 0.028216 0.021736 0.022437 0.019187 0.021658 0.018961 0.033169 0.02013 0.019788 0.013331 0.020436 0.006271 0.01523 0.004595 Trace statistics 11.29069 3.970087 17.65873 7.669574 14.68079 6.761222 14.32265 6.680852 18.86938 7.096901 11.62552 4.670334 9.374737 2.189339 6.943377 1.602683 Critical value 5% 15.49471 3.841466 15.49471 3.841466 15.49471 3.841466 15.49471 3.841466 15.49471 3.841466 15.49471 3.841466 15.49471 3.841466 15.49471 3.841466 Remarks No Cointegration Cointegrated No Cointegration No Cointegration Cointegrated No Cointegration No Cointegration No Cointegration

KSE 100 -S&P 500 KSE 100 -CAC 40 KSE 100 DAX KSE 100 -FTSE 100 KSE 100 - Nikkei225 KSE 100 -TSX KSE 100 MIB30 KSE 100 -AORD

According to representation theorem, if two variables are cointegrated then Granger-causality must exist in at least one direction. Results of Granger causality are reported in Table 7. Rejection of the null hypothesis at 5% indicates that there exist unidirectional granger causality between developed markets and Karachi equity market.

International Research Journal of Finance and Economics - Issue 16 (2008)

Table 7: Granger Causality Test (Lag 2)
F-Statistic 5.28447 2.05946 3.91492 1.47016 3.26551 2.10154 4.31258 0.02584 6.32212 0.37932 7.04085 0.38645 2.03193 0.04082 3.5461 2.58651 Probability 0.00549 0.12909 0.02083 0.23132 0.03936 0.12383 0.01413 0.9745 0.00201 0.68461 0.00101 0.67976 0.13265 0.96 0.02989 0.07674

59

Null Hypothesis: Rtn S&P 500 does not Granger Cause Rtn KSE 100 Rtn KSE 100 does not Granger Cause Rtn S&P 500 Rtn CAC 40 does not Granger Cause Rtn KSE 100 Rtn KSE 100 does not Granger Cause Rtn CAC 40 Rtn DAX does not Granger Cause Rtn KSE 100 Rtn KSE 100 does not Granger Cause Rtn DAX Rtn FTSE100 does not Granger Cause Rtn KSE 100 Rtn KSE 100 does not Granger Cause Rtn FTSE100 Rtn Nikkei225 does not Granger Cause Rtn KSE 100 Rtn KSE 100 does not Granger Cause Rtn Nikkei225 Rtn TSX does not Granger Cause Rtn KSE 100 Rtn KSE 100 does not Granger Cause Rtn TSX Rtn MIB 30 does not Granger Cause Rtn KSE 100 Rtn KSE 100does not Granger Cause Rtn MIB 30 Rtn AORD does not Granger Cause Rtn KSE 100 Rtn KSE 100 does not Granger Cause Rtn AORD

To further investigate the dynamic responses between the variables we also calculated the impulse response of the VAR system and results are given in Table 8. An impulse response function traces the effect of a one time shock to one of the innovations on current and futurer values of the endogenous variables. The following charts show the impulse response of one of the variables on other variables. Here, it appears that Karachi stock market seems to be the exogenous as most of its shock is explained by its own innovations. However, US and UK, German and Japan markets are exerting some impact on Karachi stock exchange.
Impulse Response Analysis
Res pons e of KSE100 One S.D. Innovations
Response of kse to kse
.04 .03 .02 .01 .00 -.01 1 2 3 4 5 6 7 8 9 10 .04 .03 .02 .01 .00 -.01 1 2 3 4 5 6 7 8 9 10

Response of KSE100 to S&P500

.04 .03 .02 .01 .00 -.01 1 2

Response of KSE100 to CAC40

10

Response of KSE100 to DAX

.04 .03 .02 .01 .00 -.01 1 2 3 4 5 6 7 8 9 10 .04 .03 .02 .01 .00 -.01 1

Response of KSE100 to FTSE100

.04 .03 .02 .01 .00 -.01 2 3 4 5 6 7 8 9 10 1

Response of KSE100 to Nikkei225

10

Response of KSE100 to TSX

.04 .03 .02 .01 .00 -.01 1 2 3 4 5 6 7 8 9 10 .04 .03 .02 .01 .00 -.01 1 2

Response of KSE100 to MIB30

.04 .03 .02 .01 .00 -.01 3 4 5 6 7 8 9 10 1 2

Response of KSE100 to AORD

10

60

International Research Journal of Finance and Economics - Issue 16 (2008)

Finally, we conduct variance response analysis to see to what extent shocks to certain markets are explained by other markets. It provides some further evidence on the patterns of linkages amongst stock markets and enhances our insights about reaction of markets to system wide shocks. It also helps to the pattern of responses propagation over time. Table 8 gives the decomposition of forecast error variance for the variables whereas results for other markets are given in appendix.
Table 8:
Period 1 2 3 4 5 6 7 8 9 10

Variance Decomposition Analysis

S. E 0.04 0.04 0.05 0.05 0.06 0.06 0.07 0.07 0.07 0.07 KSE 100 100.00 96.95 94.53 94.44 94.39 94.72 94.82 95.03 95.14 95.26 S&P 500 0.00 0.06 1.61 1.33 1.33 1.39 1.31 1.25 1.25 1.22 CAC 40 0.00 0.44 0.53 0.46 0.40 0.36 0.33 0.29 0.27 0.25 DAX 0.00 0.48 0.51 0.81 0.70 0.63 0.60 0.56 0.54 0.51 FTSE 100 0.00 1.50 1.30 1.35 1.36 1.22 1.22 1.17 1.13 1.10 Nikkei 225 0.00 0.03 0.43 0.49 0.48 0.44 0.39 0.37 0.34 0.31 TSX 0.00 0.02 0.57 0.51 0.73 0.66 0.70 0.71 0.72 0.72 MIB 30 0.00 0.16 0.14 0.16 0.15 0.14 0.15 0.14 0.13 0.13 AORD 0.00 0.36 0.38 0.45 0.45 0.45 0.49 0.48 0.49 0.50

The Karachi stock market seems to be the exogenous as most of its shock is explained by its own innovations. It can be seen that variation in Karachi equity market are greatly explanative by them whereas US and UK markets are exerting some impact on Karachi stock exchange during 2000-2006. Results for other markets are reported in appendix B Indicate the developed markets exert significant impact on each other

5. Conclusion
Karachi stock exchange offers the highest return i.e 0.54 % per week at a reasonable risk level whereas European and American stock markets generally show an average negative return during the period under study which is indicative of good economic performance of the country. Further, Karachi stock exchange is not correlated with the equity markets of the developed world it means opportunities of diversification exist. It is also evident that USAs stock index S& P 500 is strongly correlated with the indices of European countries like UK, France and Germany. It may be due to free flow of funds within these countries as result of elimination of barriers on capital flow. Moreover, strong correlation exists among European markets which is in line with the concept of financial and economic integration of Europe. Correlation analysis is weak technique as it does not discuss the cause and effect relationship so Cointegration and Granger causality is tested. Before application of Johansen-Juselius maximum likelihood ratio test, stationarity of index series is tested by using Augmented Dickey Fuller test and Phillips-Perron test. Both tests confirms that series are I(1).. Trace test and Max-eigen value test both indicates 2 cointegrating equations at the 0.05 level. Thus according to multivariate Cointegration analysis markets are integrated and there exist a long term relationship between these markets. However pair wise Cointegration analysis shows that Karachi stock market is not cointegrated with equity market US, UK, Germany, Canada, Italy and Australia. However market is found integrated with France and Japan. Therefore funds managers of US, UK, Germany, Canada, Italy and Australia can get the benefits of portfolio diversification by investing in the Karachi stock market. However European markets and US markets indicate pair wise Cointegration which is in line with results of our informal correlation analysis. Impulse response analysis and variance decomposition analysis indicate that the Karachi stock market appears exogenous as most of its shock is explained by its own innovations whereas US and UK markets are exerting some impact on Karachi.

International Research Journal of Finance and Economics - Issue 16 (2008)

61

References
[1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] Kasa (1992), Common stochastic trends in international stock markets. Journal of Monetary Economics, 29: 95 - 124. Lamba(2005), Analysis of the Short- and Long-Run Relationships Between South Asian and Developed Equity Markets, International Journal of Business, 10(4), 2005 ISSN: 10834346 Roca E.D.(1999), Short-term and long-term price linkages between the equity markets of Australia and its major trading partners, Applied Financial Economics. 9: 501 -511. Suchismita B. (2005), Indian, US and Asian Stock Markets Recent Trends in Interlinkages Money And Finance, Jul- Dec 2005 Naeem(2000), Stock Market Linkages: Evidence from South Asia unpublished Paresh Narayan; Russell Smyth; Mohan Nandha(2004),Interdependence and dynamic linkages between the emerging stock markets of South Asia Accounting and Finance, Volume 44, Number 3, November 2004, pp. 419-439(21) Phillips, P., and P. Perron.(1988),Testing for a unit root in time series regression, Biometrica, 75 (June):335-346 Johansen, S. (1991),Estimation and hypothesis testing of cointegrating vectors in Gaussian vector autoregressive models, Econometrica, 59 (November):1551-1580. Johansen, S.(1992),Determination of cointegration rank in the presence of a linear trend, Oxford Bulletin of Economics and Statistics, 54 (3):383-397. Johansen, S., and K. Juselius (1990), Maximum likelihood estimation and inference on cointegration with application to the demand for money, Oxford Bulletin of Economics and Statistics, 52:169-209 Johansen, S. (1988),Statistical analysis of cointegrating vectors, Journal of Economic Dynamic and Control, 12:231-254. Aggarwal R., N A. Kyaw (2003), Equity Market Integration in the NAFTA Region:Evidence from Unit Root and Cointegration Tests, Granger, C.W.J.(1988),Some recent developments in a concept of causality, Journal of Econometrics, 39 (1/2):199-211 Engle, R.F., and C.W.J. Granger.(1987),Co-Integration, error correction: Representation, estimation and testing. Econometrica 55:1251-1276. Gunasinghe (2004), Behaviour of Stock Markets in South Asia An Econometric Investigation Accounting and Finance, Volume 44, Number 3, November 2004, pp. 419-439(21) Glezakos M., Merika A., Kaligosfiris. H (2007), Interdependence of Major World Stock Exchanges: How is the Athens Stock Exchange Affected? International Research Journal of Finance and Economics Issue 7 (2007)

62

International Research Journal of Finance and Economics - Issue 16 (2008)

Appendix