COURSE PROJECT MTH517 TIME SERIES ANALYSIS. Study of Long Run relationships and Interdependence between stock markets of US, UK, Japan and India

Transcription

1 COURSE PROJECT MTH517 TIME SERIES ANALYSIS Study of Long Run relationships and Interdependence between stock markets of US, UK, Japan and India Submitted to: AMIT MITRA, PhD Ankit Palliwal, Y6075 Lokesh Pathairia, Y6229 Mandala Rohan, Y6239 Prabhat Kumar, Y6323 Shrey Sahay, Y6411

2 Introduction In the recent decade, markets, businesses, regions, and continents have become more interdependent upon one another. The process of globalization and Economic integration has entailed large cross-border capital flows and resulted in stronger real economic linkages between emerging and developed market economies. Greater real and financial integration may imply higher synchronization between developed and emerging stock markets. The growing international integration of financial markets has prompted several recent empirical studies to examine the mechanism through which stock market movements are transmitted around the world. These studies evaluate how stock returns in one national stock market influence those of another stock market. This academic exercise is aimed at studying whether there exists a long run relationship between the stock market indexes of the developed and emerging market economies. The leading Stock Market Index of India is studied for the existence of long run relationship with Stock Market Indexes of US (DJIA), UK (FTSE 100) and Japan (Nikkei 225). Also, this exercise shall examine the extent to which the movements in the Indian stock market can be explained by the shocks in the three major stock markets. Review of Literature The process of financial liberalization and integration in developing countries has gained much momentum after the works of McKinnon (1973) and Shaw (1973). The authors have been advocating that Financial Liberalization is essential for both financial development and economic development. According to Das (1993), the process of financial liberalization has brought the capital markets closer by linking them more closely. Further, the developments in field of information technology has played an important role in bringing radical changes in the financial services sector by eliminating the geographical barriers to trading and by increasing the speed of fund transfer. The globalization of the financial markets around the world has greatly stimulated the demand and supply of cross-border capital flows. Hawawini (1994) points out that there is an increasing flow of funds from developing countries towards emerging markets like the Latin American markets, the Asian markets, the African markets etc. and, therefore, these markets are becoming increasingly important in terms of portfolio management for institutional investors. Karolyi (1995) focuses on the volatility of returns on the US and Canadian stock indices. He compares the impulse response functions using four specifications of the volatility process. A lot of work has been done on market linkages, market integration, influences and spillovers of one market on another (Yochanan Shachmurove, 2001; Hsin, Chin-Wen, 2004; Balazs Egert & Evzen Kocenda, 2005; Andrew Worthington & Helen Higgs, 2001). Data and data Sources The study uses daily stock returns data. The daily returns data is represented by the log return of the index. The index values for the common working days for the indexes studied have been considered. The data is collected from the official websites of the respective Stock Indexes.

3 Daily returns data is able to capture all the possible interactions. Using weekly or monthly data may block out interactions that last for only a few days. The data are in the respective local currency terms. For the purpose of our analysis, we use the closing times as a point of reference. Refer Table 1(a) for more details. Methodology The cointegration analysis of the indexes considered in the study is conducted so as to analyze whether there exists a long run relationship between them. The residual based approach proposed by Engle and Granger (1987) is used for cointegration analysis. If a pair of time series is cointegrated, there has to be causation in at least one direction (Granger, 1988). To know the direction of causation amongst the cointegrated variables, the causality analysis is conducted using the Granger Test (1969). Techniques used in analysis Unit Root Testing Time series analysis is about the identification, estimation and diagnostic checking of stationary time series. Definition: The sequence {y t } is said to be covariance stationary if for all t and t-s 1. E(y t ) = E(y s )= µ 2. E(y t - µ) 2 = E(y t-s - µ) 2 =σ 2 3. E(y t - µ)(y t-s - µ) = E(y t-j - µ)(y t-j-s - µ) = s That is, the mean, variance and covariance are invariant to the time origin. The Augmented Dickey-Fuller Test for Unit Roots Dickey and Fuller (1979, 1981) devised a procedure to formally test for the presence of a unit root. The Augmented Dickey-Fuller test simply includes AR(p) terms of the X t term in the three alternative models. Therefore we have: The difference between the three regressions concerns the presence of the deterministic elements a 0 and a 2 t. The parameter of interest in all the regression equations is ; if = 0, the series contains a unit root. Tests for Cointegration Consider two time series x t and y t, these two time series can be said to be cointegrated if: (a) both time series (x t and y t ) are I(1) that is become stationary after first differencing, and (b) there is some linear combination of x t and y t that is I(0), that is stationary. When conditions (a) and (b) above hold, it is said that the time series x t and y t are cointegrated. Thus, cointegration is the

4 statistical equivalent of the existence of a long-run economic relationship between I(1) variables. The meaning is that of existence of long-run equilibrium relationship. Testing for Causal Relationship Amongst the Cointegrated Indexes Granger-Causality Test Granger (1988) pointed out that if a pair of time series is cointegrated, then there must be causation in at least one direction. According to the Granger causality (Granger, 1969) approach a variable Y is caused by X, if Y can be predicted better from past values of Y and X, than from past values of Y alone. Four patterns of causality can be distinguished: (a) unidirectional causality from X to Y; (b) unidirectional causality from Y to X; (3) feedback or bi-directional causality; and (d) no causality. For a simple bivariate model, the pattern of causality can be identified by estimating regression of Y and X on all the relevant variables including the current and past values of X and Y respectively and by testing the appropriate hypothesis. Apart from the conventional econometric techniques including co-integration, causality tests and univariate GARCH models, the asymmetric extensions of GARCH models viz., EGARCH and TGARCH shall be employed. VAR Models with Impulse Responses and Variance Decompositions We will analyze the market dynamics, transmission and propagation mechanism driving the stock markets. For this purpose, we will estimate a vector auto regression model (VAR). The VAR model estimates a dynamic simultaneous equation system, free of a priori restrictions on the structure of relationships. Since no restrictions are imposed on the structural relationships between variables, the VAR system can be a flexible approximation to the reduced form of the correctly specified but unknown model of the actual economic structure. Any shock to the stock market is typically characterized by an explosive effect to all financial markets. It is difficult to isolate the effect of a shock to any particular market in such a scenario. where, Z(t) is a 4 x 1 column vector of rates of return of four stock markets, C is the deterministic component comprised of a constant, A(s) are respectively 4x1 and 4x4 matrices of coefficients, m is the lag length, e(t) is the 4 x 1 innovation vector. The advantage of using VAR is that it not only gives us estimates of dependence between the a system of stock markets, but also allows us to shock a particular market and analyze how the shock perpetuates itself throughout the system using the impulse response (IR) analysis. VAR models help to capture the pure effects of artificial shocks introduced by the researcher in a similar manner to dynamic simulations. Another feature of the VAR models pertains to the variance decompositions. The decomposition of variance of the forecast errors of the returns of a given market indicates the relative importance of the various markets in causing the fluctuations in returns of that market. The decomposition allocates the variance of forecast error into

5 percentages that are accounted for by innovations in all other markets including the market s own innovations. Results Results for Co-integration and granger Causality test The results in Table 1(b) indicate that except for the case in which BSE Sensex was regressed on DJIA, in all other cases, the calculated ADF statistic of the residual term was smaller than the critical statistic. In other words, the error generated out of the regression of two non stationary time series is itself non stationary. Hence it could be said that only DJIA and BSE Sensex are cointegrated. Further, the causality test was run for the cointegrated variables so as to determine the direction of causality between the variables. Table 1(c) indicates a unidirectional causality from DJIA to BSE Sensex as per the Bivariate Granger Causality Test. Results for Univariate GARCH Analysis and Asymmetric Extensions Table 4 report the parameter estimates from fitting an AR(p) model for India, UK, USA & Japan for the period Jan to Mar Table 5 reports the results of fitting GARCH (1, 1), EGARCH and GJR models. For the case of India, the coefficients of the α 1, β 1 and γ 1 are all significant. The γ 1 coefficient in the EGARCH model is negative. This implies that a positive returns shock to the Indian stock market produces lower volatility than a negative returns shock. The γ 1 coefficient in the GJR model is 0.22 which is greater than 0 and is significant. This means that leverage effect exists in the Indian stock market and news impact is asymmetric. The magnitude of the differential impact on conditional variance can be ascertained from the α and γ values. Good news to the Indian stock market has an impact of 0.02 while the impact of bad news on the conditional variance is given by 0.24 (α +γ). Based on minimum AIC/SIC values and maximum log likelihood values, it can be concluded that the EGARCH model best captures volatility dynamics of the Indian stock market. As for UK, the coefficients of the α 1, β 1 and γ 1 are all significant except for α 1 in the GJR model. The γ 1 coefficient in the EGARCH model is and is negative and significant. The γ 1 coefficient in the GJR model is 0.11 which is greater than 0 and is significant. Leverage effect is supported as in the case of India. Good news to the UK stock market has an impact of 0.01 while the impact of bad news on the conditional variance is given by 0.12 (α +γ). EGARCH best governs the dynamics of UK stock market volatility. In the case of US, the coefficients of the α 1, β 1 and γ 1 are all significant except for α 1 in the GJR model. This is similar to UK. The γ 1 coefficient in the EGARCH model is and is negative

6 and significant. The γ 1 coefficient in the GJR model is 0.13 which is greater than 0 and is significant. Leverage effect is supported. Good news to the US stock market has an impact of while the impact of bad news on the conditional variance is given by 0.13 (α +γ). For US, the EGARCH model captures volatility dynamics well based on the lowest AIC/SIC values and highest maximum log likelihood values. Lastly for Japan, the coefficients of the α 1, β 1 and γ 1 are all significant. The γ 1 coefficient in the EGARCH model is and is negative and significant. The γ 1 coefficient in the GJR model is 0.10 which is greater than 0 and is significant. This indicates the presence of leverage effect. Good news to the Japanese stock market has an impact of 0.03 while the impact of bad news on the conditional variance is given by 0.13 (α+γ). As in the case of India, UK and US, the EGARCH model captures volatility dynamics of the Japan market well. On comparing the results and its implications for the Indian stock market, we can see that India exhibits comparatively higher α values than the other three economies in all three models. This implies that the effects of shocks in earlier periods tend to linger around for a longer period than it does in other stock markets. This may imply that the Indian stock market shows less market efficiency than the other markets as the effects of the shocks take a longer time to dissipate. The β 1 parameters capture long term influences on volatility. What is interesting to note that all markets except India exhibit very similar β 1 values. This shows that long term effects have similar influences on market volatility. In the EGARCH and GJR model, the γ 1 values show how good and bad news affects the volatility. India s γ 1 value of (0.24) is comparatively higher than that exhibited by UK(0.12), US (0.13) and Japan(0.13). This means that the leverage effect is higher in the Indian stock market versus that experienced in the other markets. The impact of bad news and shocks has a much greater effect on India. On the whole, it can be seen that the Indian Stock market is very vulnerable to external volatility movements in major markets, especially that of UK, US & Japan. Results for Vector Autoregression and Impulse Response Analyses In the estimation stage, unit root tests are run to analyze time-series properties of the price series. The results from the Augmented Dickey-Fuller test reveal that all the series are stationary in the form of first differences. Table 6, reports the results of the unit root test. The independent variables in the VAR model have to be lagged a certain number of times. The number of lags was chosen based on the Akaike-Schwarz Information Criteria. The Akaike and Schwarz tests suggest a lag order of two. Friedman and Shachmurove (1997) suggest that higher lag orders ensures that all the dynamics in the data is captured in analysis. Eun and Shim (1989) used 15 lags. However, in order to keep the model parsimonious, the Information Criteria governed the lag length in this study. Table 7 shows the parameter estimates obtained from fitting a VAR model. The parameter estimates between India and the markets of US and Japan are found to be significant.

7 After estimating a VAR model, we go on to obtain the variance decompositions. The decomposition of variance of the forecast errors of the returns of a given market indicates the relative importance of the various markets in causing the fluctuations in returns of that market. A leading market is one which explains a large percentage of the error variance of other markets while its own forecast error is not explained by innovations in other markets. Table 8 presents the decompositions of the forecast error variance for 1-day, 5-day, 10-day and 15-day horizons. We begin by considering the effect of a shock that originates in the US and then moves to Japan, India and UK. The markets have been ordered according to closing times. Entries show the percentage forecast error variance of the market in the first column explained by the market in the first row. For 5-day, 10-day and 15-day horizons, Table 8 shows that the US market accounts for between 4% and 34% of the forecast error variance of the other markets. The US stock market accounts for approximately 7% of forecast error variances in the Indian market followed by Japan which accounts for 6.5% of error variances of the Indian market. UK has a 0.03% impact on India s error variances. UK is the most endogenous market with almost 36% of its forecast error variance explained by the other markets in the system. This shows how open the UK stock market is and how vulnerable it is to shocks occurring in leading stock markets. Following the analysis on variance decomposition, we investigate the pattern of dynamic impulse response of India to shocks in US, Japan and the UK. The results provide insight on the efficiency of Indian Stock market with respect to the information contained in such shocks. The impulse response coefficients are normalized such that the unit is the standard deviation of the orthogonalized innovation. The initial shock in a variable is set equal to one standard error of innovation at s = 0. Despite different variations of returns across the equity markets considered, the normalized coefficients represent simulated impulse responses of India to each of the four markets to a positive, one-standard deviation shock in the US, Japan and the UK. Figure 3 shows the impulse response of India to the other stock markets. When the lower band crosses the horizontal axis, the response becomes statistically insignificant. The first panel of Figure 3 shows the response of the Indian stock market to a shock in the US market. India s peak response occurs on day one. What is interesting to note is that India s response to a US shock is persistent. The impact of a shock in US lasts for four days. By day four, the impact of the shock becomes insignificant. The second panel shows the response to a shock in the Japanese market. Peak response occurs on the first day of the shock. The impact of the shock dies off by the second day. Panel three shows Indian stock market response to its own domestic shocks. The fourth panel shows India s response to shocks in the UK. Here we can see that India s response to shock from UK is negligible. From all four panels, we can see that the magnitude of India s response to shock from Japan is the highest (approx.0.5) followed by the US (0.4) and the UK (0.1). Hence we can say that India is most sensitive to shocks from the Japan and US market. Conclusions

8 Main Findings This study examines the existence of long run relationships between the Indian stock market and that of US, UK and Japan, the extent to which movements in the Indian stock market can be explained by shocks in the three major markets. Main Results of Co-integration and Granger Causality test It can be concluded that there exists a long run co-integrated relationship between the Stock Markets of US and India. The direction of Causality is from US to India. However, the result cannot be generalized for the Emerging and Developed Market Economies as the results are dependent on the dataset. Main Results of Univariate GARCH Analysis By building a GARCH (1, 1), EGARCH and GJR model for the full sample in order to capture the volatility dynamics of India, Japan, US and UK over a 10 year period, it was found that asymmetry is significant and supported in all four markets. The EGARCH model performs better in capturing volatility dynamics of India, US, UK and Japan. Generally, shocks to the Indian market tend to linger around for a longer period than it does in other stock markets. This may imply that the Indian stock market shows less market efficiency than the other markets as the effects of the shocks take a longer time to dissipate. The Indian stock market shows much higher leverage effect than exhibited by US, Japan and UK. The impact of bad news has a much greater effect on India than it does on the other major markets. Main Results of Vector Autoregression and Impulse Response Analysis Though VAR is not a volatility model, it allows us to look at the extent to which multilateral interaction exists between these markets and the structure of interdependence simultaneously. A VAR analysis is performed on the full sample The variance decomposition results show that the US stock market accounts for about 7% of the forecast error variance of Indian market while the Japan and UK markets account for about 6.5% and 0.03% of forecast error variance respectively. UK is the most endogenous market in the system accounting for about only 63% of its own forecast error variances. Following the analysis on variance decomposition, impulse response technique was used to investigate the pattern of dynamic impulse response of India to shocks in US, Japan and UK. India s response to shocks from Japan and UK lasts for about one day and becomes insignificant. However, India s response to shocks from US is persistent and the impact of shocks from the US market lasts for about four days in the Indian stock market. The UK market was found to have very little influence over the Indian stock market. The Japan and US stock markets are dominant in influencing the Indian stock market. References Das D. (eds.), 1993, International Finance: Contemporary Issues, Routledge Pvt., London.

9 Hawawini G., 1994, Equity Price Behaviour: Some Evidence from Markets Around the World, Journal of Banking and Finance, 18, pp Shaw E.S., 1973, Financial Deepening in Economic Development, Oxford University Press, New York. McKinnon R.I., 1973, Money and Capital in Economic Development, Brookings Institution, Washington D.C. Yochanan Shachmurove, 2001."Dynamic Co-movements of Stock Indices: The Emerging Middle Eastern and the United States Markets," Penn CARESS Working Papers ddffc4204cf90a8523fb64134, UCLA Department of Economics. Karolyi, G Andrew, "A Multivariate GARCH Model of International Transmissions of Stock Returns and Volatility: The Case of the United States and Canada," Journal of Business & Economic Statistics, American Statistical Association, vol. 13(1), pages 11-25, January. Hsin, Chin-Wen, "A multilateral approach to examining the comovements among major world equity markets," International Review of Financial Analysis, Elsevier, vol. 13(4), pages Balazs Egert & Evzen Kocenda, "Contagion Across and Integration of Central and Eastern European Stock Markets: Evidence from Intraday Data," William Davidson Institute Working Papers Series wp798, William Davidson Institute at the University of Michigan Stephen M. Ross Business School. Andrew Worthington & Helen Higgs, 2001."A multivariate GARCH analysis of equity returns and volatility in Asian equity markets,"school of Economics and Finance Discussion Papers and Working Papers Series 089, School of Economics and Finance, Queensland University of Technology. Watson, Mark W. (1994). Vector Autoregressions and Cointegartions, in Robert F. Engle and Daniel L. McFadden, editors, Handbook of Econometrics. Chapter 47, pp Vol. 4. Amsterdam: Elsevier. Tim bollerslev, Robert E. Engle and Daniel B. Nelson (1994), ARCH Models, in the Handbook of Econometrics Vol. IV (eds. Robert E. Engle and Daniel Mcfadden), Amsterdam: North Holland Press. Chris Brooks, Introductory Econometrics for Finance, 2002, Cambridge University Press.

10 Appendix: Tables and Figures Table 1(a). Data Sources and Software used Country Index Source Developed Market Economies US DJIA Japan Nikkei UK FTSE Emerging Market Economies India SENSEX Alternative data Sources Google Finance finance.google.com Yahoo Finance finance.yahoo.com Boston Finance finance.boston.com Software EViews, R, STATA Table 1(b) Unit Root Test Results of Residuals based on OLS Regression of Sensex on DJIA, FTSE 100 and Nikkei 225 Residuals based on OLS regression of Sensex on ADF test statistic Test critical values(10% level) Stationary DJIA FTSE Nikkei

11 Table 1(c) Results of Bivariate Granger Causality Test Lags Null Hypothesis Obs F-statistic Probability 1 BSE does not Granger Cause DJIA DJIA does not Granger cause BSE E-21 2 BSE does not Granger Cause DJIA DJIA does not Granger cause BSE E-27 3 BSE does not Granger Cause DJIA DJIA does not Granger cause BSE E-27 4 BSE does not Granger Cause DJIA DJIA does not Granger cause BSE E-28 5 BSE does not Granger Cause DJIA DJIA does not Granger cause BSE E-27 Table 2: Market indices and market opening and closing times Country Index Abbreviation GMT India BSE SENSEX SENSEX 04:25-10:00 UK FTSE 100 FTSE :30-16:30 US Dow Jones Industrial Average DJIA 14:30-21:00 Japan Nikkei 225 Nikkei :00-02:00 03:30-06:00 Table 3: Descriptive statistics for daily market returns in local currency terms: Jan Mar India Japan UK USA Mean Median Maximum Minimum Std. Dev Skewness Kurtosis Jarque-Bera

12 Figure 1: Plot of stock price indices for India, UK, USA & Japan SENSEX DJIA FTSE100 Nikkei225

13 Figure 2: Plot of returns for India, UK, USA & Japan 15 Returns_SENSEX 12 Returns_DJIA Returns_FTSE Returns_Nikkei

14 Table 4: Parameter estimates from fitting AR (p) for India, UK, USA & Japan from Jan Mar Country India UK USA Japan Parameter Estimates AR(1) 0.04(0.00)* 0.04(0.00)* -0.04(0.02)* -0.04(0.01) -0.04(0.01)* -0.05(0.00)* -0.05(0.00)* - - AR(2) (0.00) -0.08(0.00)* (0.00)* -0.06(0.00)* AR(3) AR(4) AR(5) (0.01)* (0.00)* (0.03)* (0.01) AIC SIC Log Likelihood Q(10) 17.17(0.04)* 11.77(0.16) 59.16(0.00)* 37.56(0.00)* 8.67(0.27) 12.42(0.19) 7.55(0.47) 11.34(0.25) 5.26(0.72) Q 2 (10) (0.00)* (0.00)* (0.00)* (0.00)* (0.00)* (0.00)* (0.00)* (0.00)* (0.00)* The p-values are indicated in the parentheses. I check for 10% level of significance. P-values greater than 0.1 indicate insignificance while p-values less than 0.1 indicate significance. * denotes statistical significance at the 10% level

15 Table 5: Parameter estimates of fitting GARCH (1, 1), EGARCH and GJR for India, UK, USA & Japan for Jan 1996 Mar 2009 Model GARCH(1,1) EGARCH GJR/TGARCH Parameter Estimates India AR(1) 0.09(0.00)* 0.09(0.00)* 0.10(0.00)* AR(5) -0.04(0.01)* -0.04(0.00)* -0.04(0.00)* Ω 0.25(0.00)* -0.04(0.00)* 0.31(0.00)* α (0.00)* 0.18(0.00)* 0.02(0.00)* β (0.00)* 0.91(0.00)* 0.77(0.00)* γ (0.00)* 0.22(0.00)* AIC/SIC 3.93/ / /3.91 Log Likelihood UK AR(1) -0.02(0.12) -0.02(0.21) -0.02(0.16) AR(2) -0.04(0.00)* -0.03(0.06)* -0.04(0.02)* AR(5) -0.06(0.00)* -0.05(0.00)* -0.05(0.00)* Ω 0.01(0.00)* -0.09(0.00)* 0.01(0.00)* α (0.00)* 0.12(0.00)* 0.01(0.16) β (0.00)* 0.98(0.00)* 0.91(0.00)* γ (0.00)* 0.11(0.00)* AIC/SIC 2.99/ / /2.98 Log Likelihood US AR(1) -0.01(0.52) -0.01(0.60) (0.69) AR(5) -0.02(0.13) -0.01(0.57) -0.01(0.29) Ω 0.01(0.00)* -0.09(0.00)* 0.02(0.00)* α (0.00)* 0.12(0.00)* 0.003(0.59) β (0.00)* 0.97(0.00)* 0.91(0.00)* γ (0.00)* 0.13(0.00)* AIC/SIC 3.00/ / /2.98 Log Likelihood Japan AR(2) -0.02(0.17) 0.01(0.49) -0.02(0.20) AR(5) -0.02(0.13) -0.02(0.30) -0.02(0.20) Ω 0.04(0.00)* -0.10(0.00)* 0.05(0.00)* α (0.00)* 0.16(0.00)* 0.03(0.00)*

16 β (0.00)* 0.96(0.00)* 0.89(0.00)* γ (0.00)* 0.10(0.00)* AIC/SIC 3.57/ / /3.56 Log Likelihood The p-values are indicated in the parentheses. I check for 10% level of significance. P-values greater than 0.1 indicate insignificance while p-values less than 0.1 indicates significance. * denotes statistical significance at the 10% level Table 6: Results of the unit root test ADF PP Optimal Lag Length (based on AIC) (based on Newey-West bandwidth) India * * 6 5 UK * US * * 19 1 Japan * * 1 11 *Statistically significant at specified optimal lag length at the 1% level

17 Table 7: Vector Autoregression Estimates Jan Mar US JAPAN INDIA UK US(-1) [ ]* [ ]* [ ]* [ ]* US(-2) [ ] [ ]* [ ]* [ ]* JAPAN(-1) [ ]* [ ]* [ ]* [ ]* JAPAN(-2) [ ] [ ]* [ ] [ ]* INDIA(-1) [ ] [ ]* [ ] [ ] INDIA(-2) [ ] [ ] [ ] [ ] UK(-1) [ ] [ ]* [ ] [ ]* UK(-2) [ ]* [ ]* [ ] [ ]* C [ ] [ ] [ ] [ ] t-statistics are indicated in the brackets. The 10% critical for a two-tailed test with large df (>120) is Statistical significance is denoted by *

18 Figure 3: Response of India to shocks in US, Japan & UK Response to Cholesky One S.D. Innovations ± 2 S.E. 2.0 Response of SENSEX to DJIA 2.0 Response of SENSEX to Nikkei Response of SENSEX to SENSEX 2.0 Response of SENSEX to FTSE

19 Table 8: Variance Decompositions Market ordering: US, Japan, India, UK Innovation in market of: Variance Decomposition in markets of: Horizon(days) US JAPAN INDIA UK US JAPAN INDIA UK Cholesky Ordering: US JAPAN INDIA UK