7 IMPACT OF MACROECONOMIC VARIABLE ON STOCK MARKET RETURN AND ITS VOLATILITY 7.1 Introduction: In the recent past, worldwide there have been certain changes in the economic policies of a no. of countries. The concept of globalization and liberalization has integrated the world into a single market. The era of information technology has brought revolution in the field of research and innovation by providing the data base on various aspects of the economy. Above all the use of various software applications in the field of capital markets helps its users to reach at conclusion with high speed. The macro economic variables changes promptly whenever there is some favorable or unfavorable news in the market, the news like war, terrorist attacks, strikes, unstable governments etc. gives a negative impact on the decision of investors and hence the slow economic growth. On the other hand the news of declaring liberal economic policies, ensuring safety and security, robust infrastructure by govt. leads to the positive impact over investment decisions and hence the economic growth and prosperity of the nation. 154
The capital market has become very efficient with the advent and use of various software applications for data analysis. The Efficient Market Hypothesis (EMH) theory states that an efficient market is one in which stock prices changes rapidly as and when some new information is received. Several studies have found correlation between changes in world economy and macro economic variables, these studies also suggested that the stock market indices are very sensitive towards change in the macro economic variables like inflation rate, Bank rate, FII s, Interest rate etc. The developing countries are generally noticed with a common characteristic i.e. their saving attitude, it leads to availability of funds for further investment and hence high production which robusts the economy year by year, thereby making a continuous rise in GDP. The understanding of behavior of macroeconomic variables, which affects the stock market indices, is very useful for policy makers, investors, traders and all other stake holders. 7.2 Data: This study is restricted to India only because of non availability of the sufficient data in case other selected developing countries Brazil, Russia, China, Mexico.This study establishes the relationship between Indian stock market and macroeconomic variable. BSE Sensex is used in this study as a proxy for the Indian stock market and macroeconomic variable consist Exchange rate, Balance of Trade, average call money market rate, inflation rate, industrial production, 3 months treasury bill yield to maturity, money supply, Gold rate, MSCI, 3 months treasury bill rate of US market, Volume of BSE, Volatility of BSE, Foreign Institutional Investment and Mutual Fund. These macro variables played a major role in Indian economy. Data of 14 macroeconomic variables is collected from the website of Reserve Bank of India, indiastats.com, msci.com, website of Federal Reserve Bank, website of Bombay stock exchange and moneycontrol.com. Stock prices are collected from the website of BSE. The monthly data is used in this study from January 2000 to May 2010. 155
7.3 Method of Analysis: To accomplish this objective unit root test, co-integration test, multiple regression, VAR Granger causality test, Variance decomposition test, Impulse response, have been applied. Detail explanation of these tests is given in the chapter of research methodology. 7.4 Empirical Findings: Descriptive Statistics: The summary statistics for BSE sensex and others macroeconomic variable are given in Table-1. All returns are calculated as percentage change in the monthly closing prices. Then mean of the BSE sensex is 0.0129. The mean is highest in case of foreign institutional investment whereas it is lowest in case of mutual fund. 3 months Treasury bill yield to maturity in India, balance of trade and mutual fund are having negative value of mean. The standard deviation indicates that mutual fund, foreign institutional investment,balance of trade, average call money market rate and 3 months treasury bill rate of US market are relatively more volatile compared to BSE sensex,3 months treasury bill yield to maturity in India, gold rate, industrial production,msci,bse trading volume, money supply, exchange rate, inflation rate. The kurtosis for all the variable is more than 3,it means the frequency distribution assign a higher probability to returns around zero as well as very high positive and negative returns. The Jarque-Bera statistic for all the variable shows that distributions is not normal. 156
Table 7.1 : Descriptive Statistics of Macroeconomic Variable For the Period Jnuary 2000 to May 2010 Std. Variables Mean Dev. Skewness Kurtosis Jarque-Bera Probability BSE sensex 0.013 0.078-0.205 3.865 4.769 0.092 3 months treasury bill rate of Federal bank 0.013 0.368 6.254 56.491 15,465.790 00 3 months treasury bill rate of India - 03 09-0.214 7.594 107.342 00 Average Call Money Market Rate 0.055 0.715 9.907 105.604 55,510.750 00 Balance of trade - 07 2.345-7.818 81.504 32,036.770 00 Foreign institutional investment 0.155 7.373 5.054 47.854 10,834.230 00 Gold rate 0.011 0.037 09 3.589 1.783 0.410 Industrial production 08 0.053-0.422 3.810 6.846 0.033 Mutual fund - 2.195 18.897-5.824 52.744 13,159.190 00 Morgan Stanley Composite Index 00 07-0.791 4.610 26.337 00 BSE Trading volume 0.017 0.183 0.797 4.819 30.477 00 Money supply 0.014 0.011 1.554 6.231 102.198 00 Exchange rate 00 0.014-1.186 8.016 155.219 00 volatility of BSE 0.070 05 0.423 2.992 3.733 0.155 WPI 04 07 0.111 5.256 25.914 00 UNIT ROOT TEST: The study here employed the ADF and PP unit root test to check the stationary of the concerned time series. It is seen from the table the entire variable are stationary at level as their ADF and PP test statistics are more than critical values. So the null hypothesis is rejected and data is found to be stationary. Therefore now VAR Granger causality test, multiple regression and GARCH model can be applied which requires the data to be stationary in order to avoid getting spurious results. 157
Table 7. 2 : Result 0f ADF and PP unit root test ADF PP Variable With Intercept With Trend & intercept With Intercept With Trend & intercept BSE sensex -9.705-9.760-9.821-9.856 3 months treasury bill rate of Federal bank -9.244-9.308-8.917-9.409 3 months treasury bill rate of India -13.738-13.680-13.604-13.549 Average call money market rate -11.984-11.975-12.149-12.190 Balance of trade -10.730-10.712-11.031-119 Foreign institutional investment -116-11.077-117 -11.077 Gold rate -10.554-10.724-10.552-10.720 Industrial production -12.835-12.779-28.247-29.347 Mutual fund -10.912-117 -10.913-11.168 MSCI -8.450-8.449-8.597-8.585 BSE Trading volume -12.607-12.566-12.607-12.578 Money supply -23-1.765-11.257-12.628 Exchange rate -98-9.019-9.071-9.035 volatility of BSE -4.285-4.364-4.285-4.364 WPI -6.957-6.928-6.957-6.928 Critical Values 1% level of significance -3.484-4.034-3.485-4.035 5% level of significance -2.885-3.446-2.885-3.447 10% level of significance -2.579-3.148-2.579-3.149 Note: * Rejection of null hypothesis at 5 per cent level of significance. Johansen s co-integration test: The Result of Johansen s co-integration test are shown in table 4 which explain whether there is any long term relationship between dependent variable and independent variable. The table reveals that variables average call money market rate,foreign institutional investment, mutual fund, MSCI,trading volume of BSE, money supply, exchange rate, volatility of BSE contribute to the co integration system as their calculated values are more than their critical values. So both the Eigen value and Trace statistics shows that there is long term relationship between BSE stock market and the abovementioned macroeconomic variable. But the same is not true with 3 months treasury bill of US, 3 months treasury bill of india, balance of trade, gold rate, industrial production and inflation rate. 158
Variable 3 months treasury bill rate of Federal bank 3 months treasury bill rate of India Average Call Money Market Rate Balance of trade Foreign institutional investment Gold rate Industrial production Mutual fund Morgan Stanley Composite Index BSE Trading volume Money supply Exchange rate volatility of BSE WPI * denotes rejection of the hypothesis at the 0.05 level able 7.3 : Results of Johansen's Co-integration Test Hypothesized No.of Trace 5% critical Prob Max-Eigen 5% critical CE(S) statistics value. Statistic value 0 None * 26.4044 15.4947 08 22.5744 14.2646 0.05 At most 1 3.83 3.8415 03 3.83 3.8415 0 None * 27.3898 15.4947 05 24.6403 14.2646 0.09 At most 1 2.7495 3.8415 73 2.7495 3.8415 0 None * 27.5266 15.4947 05 23.5641 14.2646 0 At most 1 * 3.9625 3.8415 65 3.9625 3.8415 0 None * 22.263 15.4947 41 18.7591 14.2646 0 At most 1 3.5039 3.8415 12 3.5039 3.8415 None * 35.8764 15.4947 0 29.1335 14.2646 At most 1 * 6.7429 3.8415 0 94 6.7429 3.8415 0 None * 20.8385 15.4947 71 19.0905 14.2646 0.18 At most 1 1.748 3.8415 61 1.748 3.8415 0.01 None * 18.5471 15.4947 68 18.3213 14.2646 0.63 At most 1 0.2258 3.8415 46 0.2258 3.8415 0 None * 26.6754 15.4947 07 1727 14.2646 0 At most 1 * 9.6727 3.8415 19 9.6727 3.8415 None * 42.7307 15.4947 0 3896 14.2646 At most 1 * 4.6211 3.8415 0.03 16 4.6211 3.8415 0 None * 24.8519 15.4947 15 20.9455 14.2646 0 At most 1 * 3.9064 3.8415 81 3.9064 3.8415 None * 41.0351 15.4947 0 23.6757 14.2646 At most 1 * 17.3594 3.8415 0 17.3594 3.8415 0 None * 29.4976 15.4947 02 24.9103 14.2646 0.03 At most 1 * 4.5873 3.8415 22 4.5873 3.8415 0 None * 28.6686 15.4947 03 19.7853 14.2646 0 At most 1 * 8.8832 3.8415 29 8.8832 3.8415 0 None * 21.1791 15.4947 62 20.8325 14.2646 0.55 At most 1 0.3466 3.8415 6 0.3466 3.8415 159
VAR Granger Causality test: The table represents the result of VAR Granger causality test between macroeconomic variable and Indian stock market. It shows that there is unidirectional causality between BSE sensex and industrial production, BSE sensex and inflation rate, morgan Stanley composite index of developed market and BSE sensex. This reveals that developed market affects the Indian stock market. It is also seen that BSE sensex can be used as leading indicator for the change in industrial production and inflation rate. It is observed that there is bidirectional relationship between BSE sensex and 3 months treasury bill rate of Federal bank, BSE sensex and 3 months treasury bill rate of India. Table shows that variables average call money market rate, balance of trade, foreign institutional investment, gold rate,mutual fund, BSE trading volume,money supply, exchange rate, volatility of BSE does not cause BSE sensex and not affected by BSE sensex. Therefore variables 3 months treasury bill of Federal bank, 3 months treasury bill of india and morgan Stanley index of 24 developed market can be used as leading indicator for the performance of Indian stock market. 160
Table 7.4: Results of VAR Granger Causality test Null Hypothesis: Wald statistic P-value BSE sensex return does not cause 3 months treasury bill rate of Federal bank 154 05 3 months treasury bill rate of Federal bank does not cause BSE sensex return 17.670 01 BSE sensex return does not cause 3 months treasury Bill rate of India 18.664 09 3 months treasury bill rate of India does not cause BSE sensex return 22.590 02 BSE sensex return does not cause average call money market rate 1.030 0.905 Average call money market rate does not cause BSE sensex return 4.309 0.366 BSE sensex return does not cause Balance of Trade 1.628 0.804 Balance of Trade does not cause BSE sensex return 2.898 0.575 BSE sensex return does not cause Foreign Institutional Investment 1.566 0.815 Foreign Institutional investment does not cause BSE sensex return 1.532 0.821 BSE sensex return does not cause Gold Rate 2.991 0.559 Gold Rate does not cause BSE sensex return 2.820 0.588 BSE sensex return does not cause Industrial Production 8.266 04 Industrial Production does not cause BSE sensex return 02 0.837 BSE sensex return does not cause Mutual Fund 2.366 0.669 Mutual Fund does not cause BSE sensex return 5.809 0.214 BSE sensex return does not cause MSCI 1.073 0.300 MSCI does not cause BSE sensex return 10.417 01 BSE sensex return does not cause BSE trading volume 5.582 0.233 BSE trading volume does not cause BSE sensex return 2.818 0.589 BSE sensex return does not cause money supply 6.683 0.154 Money supply does not cause BSE sensex return 0.618 0.961 BSE sensex return does not cause exchange rate 7.490 0.112 Exchange rate does not cause BSE sensex return 6.364 0.174 BSE sensex return does not cause volatility of BSE 3.425 0.489 Volatility of BSE does not cause BSE sensex return 2.522 0.641 BSE sensex return does not cause WPI 19.308 0.013 WPI does not cause BSE sensex return 10.209 0.251 Variance Decomposition: Variance decomposition explains the percentage of forecast variance due to each innovation in bivariate VAR framework. The table indicates that BSE sensex explains nearly 88% of its own forecast variance while remaining 12% variance of sensex are explained by 3 months treasury bill rate of federal bank and 3 months treasury bill rate of federal bank explain 84% of its own 161
variance while BSE sensex explain remaining 16% of variances. It means both the variable cause each other changes. BSE sensex explain about 4% of the variance of 3 months treasury bill rate of india whereas 3 months treasury bill rate of india explain 3% of variance of BSE sensex,it means both the variable affect each other. In the case of BSE sensex and industrial production, BSE sensex explain only 1% of the variances of the industrial production and industrial production explain about 6% of the variance of BSE sensex. It means industrial production cause BSE sensex changes. BSE sensex and MSCI shows that sensex explain only 10% of variance of MSCI while MSCI explain about 40% of the variance. It strongly shows that MSCI cause the sensex changes. Sensex explain 3% of the variance of inflation while at the same lag inflation explains 8% of the variance of sensex. It means both the variable cause each. Table 7.5: Results of Variance decomposition of BSE sensex and other macro variables Variance decomposition lags BSE sensex 3 months treasury bill rate of federal bank BSE Sensex 2 99.942 0.058 5 87.590 12.410 10 87.421 12.579 3 months treasury bill rate of federal bank 2 6.213 93.787 5 15.843 84.157 10 15.899 841 BSE sensex 3 months treasury bill yield to maturity BSE sensex 2 99.954 06 5 95.116 4.884 10 94.865 5.135 3 months treasury bill yield to maturity 2 0.119 99.881 5 3.150 96.850 10 3.201 96.799 BSE Sensex Industrial production BSE Sensex 2 99.991 09 5 99.097 0.903 10 99.075 0.925 Industrial production 2 3.549 96.451 5 5.541 94.459 10 5.624 94.376 BSE Sensex MSCI BSE Sensex 2 90.240 9.760 5 90.157 9.843 10 90.232 9.768 MSCI 2 38.972 618 162
5 43.763 56.237 10 44.750 55.250 BSE Sensex WPI BSE Sensex 2 99.458 0.542 5 97.718 2.282 10 96.818 3.182 WPI 2 2.178 97.823 5 6.197 93.803 10 7.924 92.076 Impulse Response: The impulse response traces the responsiveness of the dependent variable in the VAR to shocks to each of the endogenous variables. So, for each variable from each equation of the VAR separately, a unit shock is applied to the error, and the effects upon the VAR system over time are noted. The ordering of the endogenous variables may affect the results of impulse response; hence the generalized impulses are considered for the analysis in order to neutralize the ordering effect. It is observed that BSE sensex respond positively to foreign institutional investment, average call money market rate, balance of trade, BSE rading volume and MSCI particularly up to 3 lags but it respond to others variable in opposite direction. 163
Fig-7.1 Impulse Response to a shock in BSE sensex, Foreign Institutional Investment, Mutual Fund, Exchange Rate, Balance of Trade, ACMMR, WPI, Industrial Production, 3-month Treasury bill, Money Supply, Gold Rate 10g, MSCI23DM, 3-months Treasury bill of Federal bank, Trading volume of BSE, Volatility of BSE. Response to Cholesky One S.D. Innovations ± 2 S.E. Response of SER01 to SER01 Response of SER01 to SER02 Response of SER01 to SER03 Response of SER01 to SER04 - - - - - - - - Response of SER01 to SER05 Response of SER01 to SER06 Response of SER01 to SER07 Response of SER01 to SER08 - - - - - - - - Response of SER01 to SER09 Response of SER01 to SER10 Response of SER01 to SER11 Response of SER01 to SER12 - - - - - - - - Response of SER01 to SER13 Response of SER01 to SER14 Response of SER01 to SER15 - - - - - - Result of Regression Model: Results of regression equation is presented in the table 3 which indicates that only two variables are statistically significant with BSE return at the 5% level of significance. MSCI represents the return of the developed market which has the positive relationship with BSE return. it means whenever there is any positive movement in the return of the developed market then BSE return will also move in the positive direction. Whereas WPI shows inverse relationship with the BSE 164
return, if there is any increase in the inflation then BSE return will respond negatively. The entire variables are explaining 50% of the return of the BSE, as value of R 2 IS.505. Value of Durbin- Waston statistics is 2.220 which show the absence of autocorrelation. It means the model is good and fit for the data. Table 7.6: Result of Regression Model t- Variable Coefficient Std. Error Statistic Prob. Constant 0.014 00 0.724 0.471 Federal bank interest rate 0.017 0.016 11 0.291 Indian interest rate -04 0.078-0.309 0.758 average call money market rate -01 09-0.131 0.896 Balance of Trade -03 02-1.213 0.228 Foreign Institutional Investment 00 01-0.223 0.824 Gold rate 0.197 0.168 1.177 0.242 Industrial Production -00 0.115-0.177 0.860 Mutual Fund 00 00-0.781 0.437 MSCI* 1.198 0.127 9.412 00 Volume of BSE -0.016 0.032-0.492 0.624 Money Supply -0.387 0.530-0.729 0.467 Exchange Rate 0.599 0.489 1.226 0.223 Volatility of BSE 0.145 0.227 0.641 0.523 Wholesale Price Index* -1.967 0.852-2.308 03 R-squared 0.505 Mean dependent var 0.014 Adjusted R-squared 0.439 S.D. dependent var 00 S.E. of regression 00 Akaike info criterion -2.688 Sum squared resid 0.372 Schwarz criterion -2.340 Log likelihood 176.290 F-statistic 7.660 Durbin-Watson stat 2.220 Prob(F-statistic) 00 * Significant at 5 percent level of significance GARCH model: It is observed from the table that no one variable is having statistically significant impact on the volatility of BSE return. The. The coefficient of the GARCH term is larger than the ARCH term 165
which indicates that effect of past volatility is higher than the recent past information. The total of the ARCH term and GARCH term is less than 1 which presents that model is perfectly structured. Table 7.7: Result of the GARCH model Variable Coefficient Std. Error z- Statistic Prob. Constant 05 03 1.344 0.179 ARCH -03 0.075-0.837 0.403 GARCH 0.490 0.345 1.419 0.156 Federal Interest rate 01 06 0.162 0.871 Indian Interset rate -0.017 0.015-12 0.279 average call money market rate 00 01-0.099 0.921 Balance of trade 00 01 0.273 0.785 Foreign institutional investment 00 00 1.382 0.167 Gold rate -03 0.034-0.091 0.928 Industrial Production -02 03-0.070 0.944 Mutual Fund 00 00 1.446 0.148 MSCI -0.018 06-0.683 0.495 Volume of BSE 09 06 1.476 0.140 Money supply -09 07-0.644 0.520 Exchange rate -07 0.115-0.586 0.558 Volatility of BSE -0.010 0.034-0.283 0.777 Wholesale price index 0.114 0.150 0.761 0.447 * Significant at 5 percent level of significance 7.5 Conclusion: In today globalization scenario where Indian stock market getting integrated with world markets, it has become important to understand the fundamental macroeconomic variable affecting the market at domestic and global level. This study made an effort to examining the impact of macroeconomic variable on stock market return and stock market volatility. It is found that developed market have a significant bearing on the Indian capital market and work as a leading 166
indicator for the Indian capital market. MSCI explain the 40% of the variance of the BSE sensex return. MSCI shows a positive relationship with BSE sensex. Treasury bill interest rate of federal bank and India also also cause change in BSE return. The impact of the inflation is also statistically significant while explaining the BSE return. Some other variable like FII, exchange rate, money supply, average call money market rate, gold rate, volume and volatility of BSE have no significant impact on the the Indian capital market. In the case of volatility no one variable is found to be statistically significant. The finding that the foreign institutional investment has been a very insignificant factor in moving the stock market is very surprising because it is contrary to the common perception that it is the foreign portfolio investment that moves the stock market. The finding can be considered as a reassurance for domestic market s strength. It becomes a crucial input for our policy makers as well as the regulators. 167