A Study on Impact of WPI, IIP and M3 on the Performance of Selected Sectoral Indices of BSE

A Study on Impact of WPI, IIP and M3 on the Performance of Selected Sectoral Indices of BSE J. Gayathiri 1 and Dr. L. Ganesamoorthy 2 1 (Research Scholar, Department of Commerce, Annamalai University, Tamilnadu, India) 2 (Assistant Professor, Department of Commerce, Government Arts & Science College, Tamilnadu, India) Abstract: The paper aims at identifying the effects on selected sectoral indices of BSE caused by selected macro-economic variables. The macro-economic variables selected for the study are Wholesale Price Index, Index of Industrial Production and Money Supply (M3). The sectoral indices selected are BSE Auto, BSE Oil & Gas and BSE Power. Twelve years data were collected from April 2006 to March 2018. The methodology used in this study is time series econometric techniques i.e. the unit root test, co-integration test and Granger causality test. The findings show that there is co-integration between BSE Auto and money supply, BSE Oil and Gas and Money Supply and BSE Power with all the three selected macro-economic variables but on the other hand there seems to be no Granger cause between WPI, IIP and M3 indices and the selected sectoral indices of BSE. Keywords: Wholesale Price Index, Index of Industrial Production, Money Supply, Sectoral indices I. INTRODUCTION Stock Market is one of the prominent investment avenues. It is risky and high yield investment Avenue. Investors can earn more if they forecast the market. They earn on fluctuations in the market. Many factors causes fluctuations in the market such as macro-economic factors, industry related factors and company related factors. Among them macro-economic factors affect share prices of all companies in the market. But quantum of impact varies from company to company based on nature of factor; similarly the quantum of impact varies sector to sector. Index of a stock market reveals direction of market at a whole. Study of index will exhibit the impact of whole market. Sensex of BSE and nifty of NSE are popular indices in India. Both the stock markets in India have sectoral indices, of which BSE has 19 sectoral indices as on 30 th August, 2018. Hence the researcher has made an attempt to study the impact of three macro-economic factors namely Wholesale Price Index (WPI), Index of Industrial Production (IIP) and Total Money Supply (M3) on the performance three sectoral indices maintained by BSE such as BSE Auto, BSE Oil & Gas and BSE Power. II. REVIEW OF LITERATURE Several Studies have been already conducted to know the impact of macroeconomic variables on the sectoral indices. Some of which are summarized below. Shanmugasundram and Benedict (2013) have attempted to provide an empirical support to identify the risk factors in sectoral indices and CNX Nifty index and also to see the risk relationship at different time intervals. The results showed that there was a significant difference in the mean scores of various time intervals. The results exhibit important implications to individual investors and portfolio managers in terms of reducing portfolio risk and enhancing their returns. Amarasinghe (2016) has attempted to investigate the relationship between IPI and sector market index in Sri Lanka. Result of correlation test showed 84% higher correlation between variables and regression result showed a significant positive relationship among variables. Together, the study concluded that Industrial Production Index will positively impact on Beverage, Food and Tobacco sector Index in Sri Lanka. Izedonmi and Abdullahi (2011) have indicated how macro-economic variables affect the equity return in both developed and emerging stock markets. Ordinary Least Square (OLS) was used and observed that there were no significant effects of those macroeconomic variables on the stocks return in Nigeria. Mundlak et al (1990) have extensively studied the effect of macroeconomic policies on Sectoral prices. The results suggested that economy wide policies had substantial negative effects on both the real exchange rate and the incentives to agricultural exports. Objective of the study To study the impact of selected macroeconomic variables namely, WPI, IIP and M3 on the performance of selected sectoral indices of BSE. III. METHODOLOGY The macro-economic variables selected for the study are Wholesale Price Index (WPI), Index of Industrial Production (IIP) and Total Money Supply (M3). These macro-economic variables are available with http://indusedu.org Page 15

only monthly data, hence for uniformity these variables are taken for the study. On the other hand, the sectoral indices selected for the study are BSE Auto, BSE Power and BSE Oil and Gas. The study period taken was 12 years from April 2006 to March 2018. Data for Wholesale Price Index (WPI) were collected from the website of the Office of the economic advisor, Index of Industrial Production (IIP) were collected from the website of Ministry of Statistics and Programme Implementation and the Total Money Supply (M3) were collected from the website of India Macro Advisors (IMA). Sectoral indices data were collected from the BSE website. The twelve years data for all the variables were taken and are checked for stationarity using KPSS Unit root test. The Econometric techniques of Co-integration and Granger causality tests are used further to check the long run equilibrium relationship and the impact of each variable on the other. Unit Root Test The test of stationarity of the selected factors (macro-economic variables and the sectoral indices of BSE) was applied. Kwiatkowski Phillips Schmidt Shin (KPSS) test is applied to investigate whether the time series data used in the study are stationary or not. Generally this is the primary test to conduct before making the analysis for co-integration test, since the test can be conducted only on non-stationary time series data. Johansen Co-integration Test Johansen method of multivariate co-integration is employed to examine the co-integration relationship between macro-economic variables and sectoral indices of BSE. The co-integration analysis is possible for time series that are not stationary. If variables are co-integrated, therefore variables move together in a long run relationship and it implies the existence of error correction model. The model used to calculate Johansen co-integration test between the Macro-economic variables and the sectoral indices of BSE is as follows. Where, r is the number of separate series, T is the number of usable observations and. Granger Causality Test The impact of an independent variable on the dependent variable can be found using the regression analysis. It shows only one side impact of the independent variable on the dependent variable and never shows the otherway impact. But in stock market this method doesn t hold good. Thus Granger causality test is used to test the impact on both the ways that is whether variable X influences variable Y or variable Y may influence variable X. The models used to calculate Granger causality test is presented below. Where, ME is the Macro-economic Variables, SIB is the Sectoral Indices of BSE, are error term assumed and assumed that they are uncorrelated, n is the maximum number of lagged observations and. IV. RESULTS AND DISCUSSIONS Table 1 presents the descriptive statistical properties namely, mean, median, standard deviation, minimum, maximum values, Skewness and Kurtosis which reveals a brief background about the movement of the selected sectoral indices of BSE and other selected macro-economic variables provided with monthly data during the study period. Table1: Descriptive Statistics on Sectoral Indices and Macro Economic Variables Factors BSE Auto BSE Oil & Gas BSE Power WPI IIP Money Supply Mean 11692.46 9638.66 2286.03 148.79 159.00 77014.25 Median 9952.07 9446.94 2125.62 150.80 164.50 74582.80 Maximum 26751.20 16552.40 4548.85 185.90 205.30 140144.80 Minimum 2330.56 5049.78 1386.60 107.80 108.80 27634.90 Std. Dev. 6750.25 2387.19 587.76 27.17 21.97 32960.84 Skewness 0.58 0.74 1.34-0.06-0.39 0.17 Kurtosis 2.07 3.85 5.08 1.39 2.29 1.76 http://indusedu.org Page 16

Jarque-Bera 13.17 17.36 68.92 15.66 6.73 9.90 Probability 0.00 0.00 0.00 0.00 0.03 0.01 Observations 144 144 144 144 144 144 It could be known from table 1 that mean values of the selected indices BSE Auto, BSE Oil and Gas and BSE Power were 11,692.46, 9638.66 and 2286.03 respectively. Their respective standard deviations showed moderate level of deviations from the mean value. On the other hand mean values of macro-economic variables, WPI, IIP and money supply were 148.79, 159 and 77014.25 respectively. Their respective SD showed that there was low level of deviation in WPI and IIP, but a moderate deviation was found for money supply. The Skewness values of the macro-economic variables WPI and IIP were negative, it implied that these factors had a heavier tail of larger values and there were higher probabilities of negative returns than positive returns. On the other hand, the Skewness values of the macro-economic variable Money Supply and all the three selected sectoral indices were positive and therefore there were higher possibilities for positive returns than negative returns during the study period. The calculated values of Kurtosis for the factors BSE OIL & GAS and BSE POWER were more than 3 which shows that the unconditional distributions of returns are not normal. The calculated values of Jarque-Bera are significant for all the three sectoral indices and all the three macroeconomic variables it shows that the selected variables were normally distributed during the study period. The study has been made to find the impact of selected macro-economic variables on the three sectoral indices. Co-integration test is applied to know the co-integration relationship between the variables. But the test can be applied only when the selected time series data are non-stationary (having unit root). Hence KPSS unit root test has been applied. Table 2 presents the results of KPSS Unit Root Test and for this purpose the following null hypothesis is framed. Ho: The selected sectoral indices and the macro-economic variables do not have unit root. Table 2: KPSS Unit Root Test KPSS Test Factors Critical Values @ 10% Significance Ho Statistics BSE AUTO 1.3131* Significant Rejected BSE OIL & GAS 0.7251* Significant Rejected BSE POWER 0.4323* 1% Level: 0.7390 Significant Rejected Wholesale Price Index 0.8510* 5% Level: 0.4630 Significant Rejected Index of Industrial 0.6549* 10% Level: 0.3470 Significant Rejected Production Money Supply 1.4131* Significant Rejected Notes: (a) * indicates significant at 10% level. Table 2 reveals that the calculated value of KPSS statistics for all the factors namely, BSE AUTO, BSE OIL & GAS, BSE POWER, Index of Industrial Production (IIP), Money Supply (MS2) and Wholesale Price Index (WPI) were significant, hence the null hypothesis was rejected and therefore these factors were nonstationary during the study period, in other words these factors had unit root. Therefore, as per KPSS test all the factors were not stationary and had a unit root, thus proceeding to the next stage. Johansen Co-integration Test Having concluded that each of the variables is not stationary, the study proceeds to examine whether there exists a long run relationship between the selected sectoral indices and the selected macroeconomic variables. For this purpose the Johansen co-integration test has been applied to know the existence of cointegration relationship between the macro-economic variables with selected sectoral indices during the study period. Table 3 presents the results of Johansen Co-integration test between the BSE Auto index and the selected sectoral indices and for this purpose the following null hypothesis is framed. Ho: There is no co-integration relationship between the movements of BSE Auto index and the selected Macroeconomic variables. Table 3: Co-integration Test between BSE AUTO and the selected Macro-economic variables Unrestricted Co-integration Rank Test Variables Pair Trace Statistics Maximum Eigen Value None At Most 1 None At Most 1 BSE AUTO - WPI Eigen Value 0.0302 0.0094 0.0302 0.0094 Trace / Maximum Eigen Value 5.5804 1.3184 4.2621 1.3184 http://indusedu.org Page 17

BSE AUTO - IIP Eigen Value 0.0530 0.0000 0.0530 0.0000 Trace / Maximum Eigen Value 7.5685 0.0015 7.5670 0.0015 BSE AUTO - Money Supply Eigen Value 0.1079 0.0334 0.1079 0.0334 Trace / Maximum Eigen Value 20.5872 4.7230 15.8643 4.7230 Table 3 shows that the calculated values of trace statistics and maximum eigen value of BSE Automobile and Wholesale Price Index (WPI) are 5.5804 and 4.2621 respectively, and the values for BSE Automobile and Index of Industrial Production (IIP) are 7.5685 and 7.5670 respectively, they are less than the critical values, so they are not statistically significant. Hence the null hypothesis is accepted and therefore the sectoral indices in terms of BSE Automobile and the macroeconomic variable in terms of WPI and IIP are not co-integrated, in other words these variables are not moving together. Coming to the next set of variables namely BSE Automobile and Money supply, the calculated values of trace statistics and maximum eigen values are 20.5872 and 15.8643 respectively, where both these values are greater than the critical values. So the results are statistically significant and hence the null hypothesis is rejected and the alternative hypothesis is accepted. Therefore the sectoral indices in terms of BSE Automobile and the macroeconomic variable in terms of Money supply are co-integrated. On the whole, the two set of variables namely, BSE Automobile WPI and BSE Automobile IIP are not co-integrated which clearly shows that that these variables are not moving together and no similarities are found in their trend. But in the set of variable BSE Automobile and Money supply there seems to be co-integration which states that these variables move together and some similarities are found in their trend. Table 4 reveals the results of Johansen Co-integration test between the BSE Oil Gas and index and the selected sectoral indices and for this purpose the following null hypothesis is framed. Ho: There is no co-integration relationship between the movements of BSE Oil and Gas index and the selected Macroeconomic variables. Table4: Co-integration Rank Test between BSE OIL & GAS and the selected Macro-economic variables Variables Pair BSE OIL & GAS - WPI Unrestricted Co-integration Rank Test Trace Statistics Maximum Eigen Value None At Most 1 None At Most 1 Eigen Value 0.0663 0.0129 0.0663 0.0129 Trace / Maximum Eigen Value 11.3498 1.8097 9.5401 1.8097 BSE OIL & GAS - IIP Eigen Value 0.0764 0.0185 0.0764 0.0185 Trace / Maximum Eigen Value 13.6471 2.6023 11.0448 2.6023 BSE OIL & GAS - Money Supply Eigen Value 0.0909 0.0534 0.0909 0.0534 Trace / Maximum Eigen Value 20.8741 7.6314 13.2427 7.6314 Table 4 indicates the calculated values of trace statistics and maximum eigen value of BSE Oil and WPI are 11.3498 and 9.5401 respectively. Likewise, the calculated values of trace statistics and maximum eigen value of BSE Oil and IIP are 13.6471 and 11.0448 respectively. In both these set of variables, the calculated values are less than the critical values, which states that these variables are not statistically significant. Hence, the null hypothesis is accepted and there is no co-integration among the above mentioned two set of variables. In contrast, the next set of variables namely, BSE Oil and Money Supply the calculated values of trace statistics http://indusedu.org Page 18

and maximum eigen value are found to be 20.8741 and 13.2427 respectively in which trace statistics is more than the critical value but maximum eigen value is less than the critical value. In this scenario, we take into consideration the value of trace statistics into effect. Therefore as per trace statistics they are statistically significant which results in rejection of the null hypothesis and acceptance of the alternative hypothesis. On the whole, while considering the sectoral indices in terms of BSE Oil and the macro-economic variables in terms of Wholesale Price Index and Index of Industrial Production there is no co-integration among the variables which states that these variables move differently in the long run. On the other hand the variables BSE Oil and Money Supply are co-integrated, in other terms they move together in the long run. Table 5 shows the results of Johansen Co-integration test between the BSE Power index and the selected sectoral indices and for this purpose the following null hypothesis is framed. Ho: There is no co-integration relationship between the movements of BSE Power index and the selected Macroeconomic variables. Table 5: Co-integration Rank Test between BSE POWER and the selected Macro-economic variables Variables Pair BSE POWER - WPI Unrestricted Co-integration Rank Test Trace Statistics Maximum Eigen Value None At Most 1 None At Most 1 Eigen Value 0.1038 0.0176 0.1038 0.0176 Trace / Maximum Eigen Value 17.6906 2.4652 15.2255 2.4652 BSE POWER - IIP Eigen Value 0.1024 0.0384 0.1024 0.0384 Trace / Maximum Eigen Value 20.4544 5.4438 15.0107 5.4438 BSE POWER - Money Supply Eigen Value 0.0923 0.0721 0.0923 0.0721 Trace / Maximum Eigen Value 23.8555 10.3954 13.4602 10.3954 Table 5 reports that the calculated values of trace statistics and maximum eigen values of BSE Power and Wholesale Price Index (WI) are 17.6906 and 15.2255 respectively. Likewise the calculated values of trace statistics and maximum eigen value of BSE Power and Index of Industrial Production (IIP) are 20.4544 and 15.0107 respectively. In both these cases the calculated values of trace statistics and maximum eigen are more than the critical values, which states that these variables are statistically significant. Hence, the null hypothesis is rejected and the alternative hypothesis is accepted. On the other hand, for the next set of variables, BSE Power and Money Supply the calculated values of trace statistics and maximum eigen value are found to be 23.8555 and 13.4602 respectively. Here it is found that the calculated values of trace statistics is more than the critical value but the maximum eigen value seems to be less than the critical value. Hence in this situation, we take in consideration the trace statistics value as the sample selected is quite large. Therefore while taking trace statistics, it is concluded that the variables are statistically significant. Hence, in this case also the null hypothesis is rejected and the alternative hypothesis is accepted. On the whole, it is found that the sectoral indices in terms of BSE Power and the macro-economic variables in terms of WPI, IIP and Money Supply are all co-integrated. It is evidenced from the above table that WPI, IIP and Money supply are moving together with the BSE Power sector. In other words, all these variables have long run equilibrium relationship between each other during the study period. Granger Causality After estimating the long run equilibrium between the selected sectoral indices and the selected macroeconomic variables by using Co-integration test, the next process intends to examine the dynamic interactions between these variables. This is done by means of Granger Causality test. It brings out the results of impact of one variable on another. Table 8 gives the Granger causality test between WPI and selected sectoral indices. http://indusedu.org Page 19

Table 6: Granger Causality between Wholesale Price Index (WPI) and the selected Sectoral indices of BSE Null Hypothesis Observations F-Statistic Prob. Ho BSE AUTO AUTO does not Granger Cause WPI 0.1168 0.8898 Accepted WPI does not Granger Cause AUTO 0.3134 0.7315 Accepted BSE OIL OIL does not Granger Cause WPI 0.1180 0.8888 Accepted WPI does not Granger Cause OIL 0.5058 0.6042 Accepted BSE POWER POWER does not Granger Cause WPI 0.0959 0.9086 Accepted WPI does not Granger Cause POWER 0.2475 0.7811 Accepted Table 6 shows that the Wholesale Price Index (WPI) does not have granger cause on the movement of all the selected sectoral indices namely, BSE Auto, BSE Oil and BSE Power. Since the F-Statistic under Granger causality is not significant for all the selected sectoral indices and hence the null hypotheses in all cases are accepted. Likewise, all the three selected sectoral indices also do not have granger cause on the movement of Wholesale Price Index (WPI). This is because again the F-Statistic under granger causality is not significant between WPI and all the three selected sectoral indices of BSE, therefore the null hypothesis in all circumstances were accepted. It is observed from the table that the fluctuations in Wholesale Price Index (WPI) does not have granger cause on the selected sectoral indices of BSE and in the same way these selected sectoral indices also does not have granger cause on the movements of WPI. Table 7 presents the results of the Granger causality test between IIP and selected sectoral indices. Table 7: Granger Causality between Index of Industrial Production (IIP) and the selected Sectoral indices of BSE Null Hypothesis Observations F-Statistic Prob. Ho BSE AUTO AUTO does not Granger Cause IIP 0.0197 0.9805 Accepted IIP does not Granger Cause AUTO 1.0208 0.3630 Accepted BSE OIL OIL does not Granger Cause IIP 0.0936 0.9107 Accepted IIP does not Granger Cause OIL 0.4901 0.6137 Accepted BSE POWER POWER does not Granger Cause IIP 0.1578 0.8542 Accepted IIP does not Granger Cause POWER 0.4544 0.6358 Accepted It is inferred from table 7 that the F-Statistic values in all combinations between Index of Industrial Production (IIP) and the three sectoral indices of BSE are not significant and hence the null hypothesis are accepted. Therefore it is very clear that there is no granger cause between IIP and the selected sectoral indices of BSE. Similarly, while testing the granger causality between the selected sectoral indices and the IIP also the F- Statistic values are not significant in all combinations and hence the null hypotheses are accepted in this scenario also. Therefore, the three selected sectoral indices namely, BSE Auto, BSE Oil and BSE Power also do not have granger cause on the movement of the macro-economic variable Index of Industrial Production (IIP). It is found from the above results that IIP does not have granger cause on the movements of selected sectoral indices of BSE. Likewise the three sectoral indices also do not have granger cause on the movements of Index of Industrial Production (IIP). Therefore it is very clear that in either of the way there is no influence caused by the macro-economic variable and the sectoral indices. Table 8 presents the results of the Granger causality test between Money Supply and selected sectoral indices. http://indusedu.org Page 20

Table 8: Granger Causality between Money Supply and the selected Sectoral indices of BSE Null Hypothesis Observations F-Statistic Prob. Ho BSE AUTO AUTO does not Granger Cause MONEY 0.1686 0.8450 Accepted MONEY does not Granger Cause AUTO 0.3973 0.6729 Accepted BSE OIL OIL does not Granger Cause MONEY 0.0892 0.9147 Accepted MONEY does not Granger Cause OIL 2.6257 0.0760 Accepted BSE POWER POWER does not Granger Cause MONEY 0.0369 0.9638 Accepted MONEY does not Granger Cause POWER 1.0591 0.3496 Accepted It is known from table 8 that the calculated values of F-statistic of Money supply on the three sectoral indices movements were not significant level. Hence the null hypotheses were accepted and therefore the Money supply does not have granger cause on any of the three selected sectoral indices namely, BSE Auto, BSE Oil and BSE Power. In the same way, in case of impact of sectoral indices on the money supply also, the calculated values of F-Statistic were not significant. Hence the null hypothesis were accepted in all these cases also which shows that the selected three sectoral indices does not have granger cause on the macro-economic variable Money Supply. It is evidenced from the Granger causality test that Money Supply does not have impact on the movement of the three selected sectoral indices namely, BSE Auto, BSE Oil and BSE Power. Likewise these three sectoral indices also do not have any impact on the macroeconomic variable Money supply. Therefore, in either of the way there seems to be no granger cause among the variables. V. CONCLUSION This paper studies the effect of macro-economic variables namely, Wholesale Price Index, Index of Industrial Production and Money supply on sectoral indices of BSE. Very few efforts were done earlier to find out the effect of these factors on the sector wise return. From that perspective this paper demands special attention for finding the effect of macro-economic variables on the sectoral indices of BSE. The study found unit root on selected macro-economic variables and selected sectoral indices of BSE. The results showed that there seems to be co-integration between the macro-economic variable money supply and all the three selected sectoral indices of BSE. In the same way there is co-integration between BSE Power and the entire three selected macro-economic variable. It shows that these variables move together in the long run. However, there seems no granger cause among any of the variables, which states that none of the variables are accountable for the movement of selected sectoral indices and macro-economic variables in either of the way. The study helps the investors in providing the information regarding the impact of market determinants on the sector wise effects. VI. REFERENCES [1] Amarasinghe, A. (2016). A study on the impact of industrial production index (IPI) to beverage, food and tobacco sector index with special reference to Colombo Stock Exchange. Procedia food science, 6, 275-278. [2] Izedonmi, P. F., & Abdullahi, I. B. (2011). The effects of macroeconomic factors on the Nigerian stock returns: A sectoral approach. Global Journal of Management and Business Research, 11(7), 25-29. [3] Mundlak, Y., Cavallo, D., & Domenech, R. (1990). Effects of macroeconomic policies on sectoral prices. The World Bank Economic Review, 4(1), 55-79. [4] Shanmugasundram, G., & Benedict, D. J. (2013). Volatility of the Indian Sectoral Indices A Study with reference to National Stock Exchange. International Journal of Marketing, Financial Services & Management Research, 2(8), 1-11. [5] www.bseindia.com [6] www.indiamacroadvisors.com [7] www.mospi.gov.in [8] www.eaindustry.nic.in http://indusedu.org Page 21