Efficiency, Determinants and Performance of Service Sector Stocks Evidence from India

Transcription

1 Conference Proceedings of 3 rd International Conference on Booming Service Sector: From Achievements to Growth Prospects, Pages: 23-36, organised by SGGSCC, DU on Feb. 4-5, ISBN: Efficiency, Determinants and Performance of Service Sector Stocks Evidence from India Vanita Tripathi 1, Arnav Kumar 2 With its unconventional and unprecedented growth on a sustained basis, the contribution of Service sector to India s economy in terms of GDP, income & employment generation, international trade and investment flows, improvement in per capita income and standards of living of general Indians has been both tremendous and amazing. In this light, the behaviour of Service Sector in Indian Stock market is also gaining attention. In a one of its kind study, this paper undertakes a comprehensive examination of Indian Service Sector Stocks from various dimensions, viz., Weak Form Efficiency, Determinants and Performance. We study service sector through nine sectoral indices of BSE (Bank, Energy, Finance, Health Care, IT, Power, Realty, TECK and Telecom) and one broad thematic Service Sector Index of NSE using monthly data from September 2005 to October We measure their adherence to Random Walk Hypothesis (Weak Form Efficiency) through a battery of Unit Root Tests [Augmented Dickey Fuller (ADF), Phillips Perron (PP) & Kwiatkowski-Phillips-Schmidt-Shin (KPSS)] and Variance Ratio Tests [Lo-MacKinlay, Wright s Ranks & Wright s Signs). We further employ proxies for Stock Market [BSE Sensex for Indian Stock Market & MSCI All Country World Index (MSCI ACWI) for Global Equity Markets] and prominent Macroeconomic Variables [Index of Industrial Production (IIP), Inflation (CPI), Interest Rate, Exchange Rate] as Regressors through Multiple Regression Analysis to find out the major determinants of Indian Service Sector Stock Returns. Finally, we also evaluate the performance of these Service sector Indices through three popular Performance measure ratios, i.e., Sharpe Ratio, Treynor Ratio and Jensen s Alpha. Overall, the Efficiency tests reveal that all Indian Service Sector Indices are indeed following the Random Walk Hypothesis and are I(1). All the Regression models were highly significant with very high explanatory powers and no problem of multicollinearity or autocorrelation. Across the board, Exchange Rate from amongst the Macroeconomic Variables and Sensex from Stock Market Variables came out to be the significant determinants of Indian Service Sector Stock Returns. Performance evaluation measures indicated that almost half of the service sector indices lead by Healthcare could significantly outperform the broader market on risk adjusted basis. These results are relevant and have pertinent implications for a broad range of stakeholders related to the field including policy makers, regulators, investors, academicians and researchers. Keywords: Indian Service Sector Stocks, Weak Form Efficiency, Service Stock Return Determinants, Stock Performance Evaluation, Multiple Regression Analysis, Unit Root & Variance Ratio Tests. JEL Classification: E44, G12, G14, G15, G18. 1 Associate Professor, Department of Commerce, Delhi School of Economics, University of Delhi, India. Address: vanitatripathi1@yahoo.co.in 2 Research Scholar, Department of Commerce, Delhi School of Economics, University of Delhi, India. Address: arnavkumardse@gmail.com

2 Efficiency, Determinants and Performance of Service Sector Stocks Evidence from India 1. Introduction Service sector, with a contribution of around 52% to nation s GDP in has a special place in Indian economy. With its unconventional and unprecedented growth on a sustained basis, the contribution of Service sector to India s economy not only in terms of GDP, but also in income & employment generation, international trade and investment flows, improvement in per capita income and standards of living of general Indians has been both tremendous and amazing. Indian Service sector had a CAGR of 9 per cent, faster than the overall GDP CAGR of 6.2 per cent in the past four years. It also attracted the largest FDI equity inflows in India amounting to US$ billion which is about 16.8 per cent of the total foreign inflows during the period April May 2015, according to the Department of Industrial Policy and Promotion (DIPP). In this light, the behaviour of Service Sector in Indian Stock market is also gaining attention. We analyse Indian service sector stocks from three dimensions - Weak Form Efficiency, Determinants and Performance. Stock market efficiency in general and its weak form in particular has been extensively researched for the past half century [Fama, 1965, 1970; Vieito et al. (2013)]. In weak-form efficiency, future security prices cannot be predicted by analysing prices from the past as they fully reflect all available or information, i.e., asset prices have no memory and they follow Random Walk. So, it is impossible to consistently outperform the market by earning excess or abnormal returns using any publicly known information. We test whether Indian service sector indices are Weak Form Efficient. When it comes to factors that can predict sectoral returns, the most likely candidates are the broad market stock indices and prominent macroeconomic variables such as Industrial Production, Inflation, Interest rate and Exchange Rate. As sectoral indices are expected to move in tandem with the broader market, a positive relation is expected between them. Increase in economic activity indicated by higher IIP, also should positively impact service sector stock returns as it will result in more cash flows and businesses for service sector firms. But, on the other hand, Inflation and Interest rate are hypothesised to have negative effect on service sector stock returns as they increase the financial and operational costs of business, reduce demand and cash flows while diverting investors to fixed income securities which become more appealing and erodes purchasing power of money. Exchange rate depreciation can have positive impact for export oriented sectors as it will result in more export revenues and should hurt sectors which are net importers due to higher input costs. So, its net effect is debatable and can vary from sector to sector. Third dimension, Performance is equally important to investigate as it will reveal whether the rapid growth in service sector in real economy is matched with corresponding higher returns on service sector indices in stock markets. It will also be interesting to explore whether any of the service sectors has outperformed the broader market on a risk adjusted basis. Rest of the work is sequenced in following manner: Section 2 provides the review of almost non existent literature on the topic, followed by Research Objectives and Hypotheses in Section 3. A detailed Research Methodology is presented in Section 4, trailed by Empirical Analysis and Results in Section 5. Conclusion and Policy Implications are discussed in Section 6 followed by References in the end. 2. Review of Literature As state earlier, the literature on entire sectoral level analysis is scant to rare. Despite our herculean effort, we could not find even a single paper which specifically covers the three dimension of efficiency, determinants and performance for Service sector stocks in any country. As such, we had to limit our survey to few of those who come closest to our theme.

3 Conference Proceedings of 3 rd International Conference on Booming Service Sector: From Achievements to Growth Prospects, Pages: 23-36, organised by SGGSCC, DU on Feb. 4-5, Fama (1965, 1970) reported a strong relationship is present between stock returns and macroeconomic variables, notably, inflation, national output and industrial production. Stock returns are determined by forecasts of more relevant real variables and negative stock returnsinflation relations are induced by negative relationships between inflation and real activity. Tripathi and Seth (2014) conveyed a significant correlation among stock market indicators and macroeconomic factors and identified Inflation, Interest rate and Exchange rate as three principal factors through Factor analysis. They also reported presence of five co-integrating relationships between stock market and macro-economic variables. Tripathi and Kumar (2015 a & b) used granger causality and panel cointegration on BRICS market to conclude that while inflation rate may be significantly related to stock returns in the short run, they do not seem to move together in the long run. Tripathi and Kumar (2015 c) used ARDL model and reported that Stock returns generally lead rather than follow GDP and Inflation. Also, they find significant negative relationship of stock returns with Interest Rate, Exchange Rate and Oil Prices and a positive relationship with money supply. Vieito et al. (2013) reported improved efficiency in G-20 countries especially after the global financial crisis. Tripathi & Kumar (2014) investigated the impact of global financial crisis on Indian stock market at sectoral level using various unit root and variance ratio tests. They reported an increase in inefficiency in Bank, Metal, PSU Bank and Realty sectors in the postcrisis period. Tripathi and Bhandari (2015a) contributed to the related literature by analysing the performance of socially responsible stocks portfolio and general stocks portfolio in Indian stock market using various risk-adjusted measures over the period They find that socially responsible stocks portfolio generated significantly higher returns as compared to other portfolios especially during crisis period. Tripathi and Bhandari (2015b) have compared the performance of ethical mutual funds with their conventional peers and reported that despite having higher risk, ethical funds outperformed on the basis of various risk-adjusted measures and net selectivity returns. 3. Research Objectives and Hypotheses 3.1 Research Objectives Our objectives in this paper are threefold, one each related to the three main dimensions of study of Indian Service sector stocks, viz., Weak Form Efficiency, Determinants & Performance. They are as follows: 1. To investigate whether the broad Indian Service sector Index along with its sub sectoral indices such as Bank, Energy, Finance, Healthcare, Information Technology (IT), Power, Realty, Technology (TECK) and Telecom are Weak Form Efficient. 2. To identify the broad stock market and macroeconomic determinants of Indian Service sector stock returns and comprehend the nature and magnitude of such relationship (if any). 3. To measure and compare the performance of the broad Indian Service sector Index along with its sub sectoral indices such as Bank, Energy, Finance, Healthcare, Information Technology (IT), Power, Realty, Technology (TECK) and Telecom with the Indian and Global benchmark broad market equity indices.

4 Efficiency, Determinants and Performance of Service Sector Stocks Evidence from India 3.2 Research Hypotheses Based on the above objectives, we have the following three sets of Null and Alternate hypotheses: 1. Null Hypothesis (H 0 ): The Indian Service sector indices are not Weak Form Efficient. Alternate Hypothesis (H 1 ): The Indian Service sector indices are Weak Form Efficient. 2. Null Hypothesis (H 0 ): The broad macroeconomic and stock market variables have no significant relationship with Indian Service sector stock returns. Alternate Hypothesis (H 1 ): The broad macroeconomic and stock market variables have a significant relationship with Indian Service sector stock returns. 3. Null Hypothesis (H 0 ): Their is no significant difference between the performance of Indian Service sector indices with the broad benchmark Indian and Global equity indices. Alternate Hypothesis (H 1 ): Their is a significant difference between the performance of Indian Service sector indices with the broad benchmark Indian and Global equity indices. 4. Research Methodology 4.1 Data To meet the above Research Objectives, the following two sets of variables have been used: A. Stock Market Indices: We have used S&P BSE SENSEX as the broad market Indian equity index and the Morgan Stanley Corporation International s (MSCI) All Country World Equity Index as the global equity benchmark. Further, we have considered the National Stock Exchange s (NSE) Nifty Services Sector Index as proxy for broad Indian service sector equity index. Then, we also incorporate the nine major service sector indices of Bombay Stock Exchange (BSE) to get a more in depth analysis of major constituents of Indian Service sector: S&P BSE BANKEX, S&P BSE Energy, S&P BSE Finance, S&P BSE Healthcare, S&P BSE Information Technology, S&P BSE Power, S&P BSE Realty, S&P BSE TECK and S&P BSE Telecom. Following is a brief description of each of these indices. 1. MSCI All Country World Equity Index: The MSCI ACWI Index offers a single, modern and fully integrated view with no gaps or overlaps, across all sources of equity returns in 23 developed [Australia, Austria, Belgium, Canada, Denmark, Finland, France, Germany, Hong Kong, Ireland, Israel, Italy, Japan, Netherlands, New Zealand, Norway, Portugal, Singapore, Spain, Sweden, Switzerland, UK and United States] and 23 emerging [Brazil, Chile, China, Colombia, Czech Republic, Egypt, Greece, Hungary, India, Indonesia, Korea, Malaysia, Mexico, Peru, Philippines, Poland, Qatar, Russia, South Africa, Taiwan, Thailand, Turkey and UAE] markets ( 2. S&P BSE SENSEX: S&P BSE SENSEX, first compiled in 1986, is calculated on a free-float market capitalization methodology of 30 component stocks representing large, well-established and financially sound companies across key sectors ( 3. Nifty Services Sector Index: The Nifty Services Sector Index is 30 stocks index and includes companies belonging to services sector like Computers Software, IT Education and Training, Banks, Telecommunication services, Financial Institutions, Power, Media, Courier, Shipping etc. (

5 Conference Proceedings of 3 rd International Conference on Booming Service Sector: From Achievements to Growth Prospects, Pages: 23-36, organised by SGGSCC, DU on Feb. 4-5, S&P BSE BANKEX: The S&P BSE Bankex index comprises constituents of the S&P BSE 500 that are classified as members of the banks sector as defined by the BSE industry classification system ( 5. S&P BSE Energy: The S&P BSE Energy is designed to provide investors with a benchmark reflecting companies included in the S&P BSE All Cap that are classified as members of the energy sector ( 6. S&P BSE Finance: The S&P BSE Finance is designed to provide investors with a benchmark reflecting companies included in the S&P BSE All Cap that are classified as members of finance sector ( 7. S&P BSE Healthcare: The S&P BSE Healthcare is designed to provide investors with a benchmark reflecting companies included in the S&P BSE All Cap that are classified as members of the healthcare sector ( 8. S&P BSE Information Technology: The S&P BSE Information Technology is designed to provide investors with a benchmark reflecting companies included in the S&P BSE All Cap that are classified as members of the information technology sector ( 9. S&P BSE POWER: The S&P BSE Power index comprises constituents of the S&P BSE 500 that are classified as members of the heavy electrical equipment & electric utilities sector as defined by the BSE industry classification system ( 10. S&P BSE REALTY: The S&P BSE Realty index comprises constituents of the S&P BSE 500 that are classified as members of the real estate sector as defined by the BSE industry classification system ( 11. S&P BSE TECK: The S&P BSE TECK index comprises constituents of the S&P BSE 500 that are classified as members of the media & publishing, information technology & telecommunications sectors as defined by the BSE industry classification system ( 12. S&P BSE Telecom: The S&P BSE Telecom is designed to provide investors with a benchmark reflecting companies included in the S&P BSE All Cap that are classified as members of the telecom sector ( B. Macroeconomic Variables: The second set of variables are the major Indian Macroeconomic Variables which can be significant determinants of Indian Service sector stock returns. We have considered: the Index of Industrial Production (IIP) which is an indicator of gross industrial economic activity in India; Monthly Inflation measured by Consumer Price Index (CPI); Short term Interest rates (Weighted Average Call Money Rates) and USD INR Exchange rate (1 USD in Indian Rupees). A detailed description of all the Stock Market Indices & Indian Macroeconomic Variables used in the study along with their Time period, Frequency, Source and Symbol for further use is provided in Table 1. Table 1: Description of All Stock Market Indices & Macroeconomic Variables S.No. Variables Time Period Frequency Source Symbol MSCI All Country 1. Monthly Investing.com ACWI World Equity Index 2. S&P BSE SENSEX Monthly BSE Website SENSEX Nifty Services Sector 3. Monthly NSE SERVICE Index 4. S&P BSE BANKEX Monthly BSE Website BANK

6 Efficiency, Determinants and Performance of Service Sector Stocks Evidence from India 5. S&P BSE Energy 6. S&P BSE Finance 7. S&P BSE Healthcare 8. S&P BSE Information Technology 9. S&P BSE POWER 10. S&P BSE REALTY 11. S&P BSE TECK 12. S&P BSE Telecom Index of Industrial Production (IIP) Consumer Price Index, Base 2010 Weighted Average Call Money Rates 1 USD in Indian Rupees Jan August, 2015 Sept Sept Sept Monthly BSE Website ENERGY Monthly BSE Website FINANCE Monthly BSE Website HEALTH Monthly BSE Website IT Monthly BSE Website POWER Monthly BSE Website REALTY Monthly BSE Website TECK Monthly BSE Website TELECOM Monthly Ministry of Statistics IIP Monthly OECD CPI Monthly RBI IR Monthly RBI ER 4.2. Methods First, we start with Descriptive statistics to identify the general statistical properties of the data. Then, in line with our Objectives, we discuss the methods for three main dimensions of this study, viz., Weak Form Efficiency; Major determinants of Indian service sector stock returns and Measuring and comparing performance of Indian service sector indices with Indian and Global benchmark indices. But before that we compute the continuous returns using the log difference: R t = Log I t - Log I (t-1) (1) Where I t = Closing Adjusted Stock Index Value in time t; I (t-1) = Closing Adjusted Stock Index Value in time (t-1) Descriptive Statistics Descriptive statistics quantitatively summarize the patterns and general trends of a dataset in a single value. This can help us identify the average return generated by different indices (Mean); their absolute risk (Standard Deviation); relative risk (Coefficient of Variation); Skewness and Kurtosis values during the study period.

7 Conference Proceedings of 3 rd International Conference on Booming Service Sector: From Achievements to Growth Prospects, Pages: 23-36, organised by SGGSCC, DU on Feb. 4-5, Weak Form Efficiency Tests If mean, variance and auto-covariance (at various lags) are time invariant, then time series data is stationary. Such a series will be mean reverting and will have a constant variance. Unit root and Variance ratio tests are widely used to test stationarity property of time series data. If the Indian service sector indices are Weak Form Efficient or follow Random Walk, then they should have Unit Root or be Non-Stationary or Martingale at level. Thus, to determine whether the Indian service sector stock indices are indeed Weak form Efficient or follow Random Walk or not, we apply a battery of powerful and commonly used Unit root and Variance ratio tests Unit Root Tests We use the Augmented Dickey Fuller (ADF); Philips Perron (PP) and Kwiatkowski Phillips Schmidt Shin (KPSS) Unit Root Tests. They are briefly explained below. 1. Augmented Dickey Fuller (ADF) Unit Root Test IHS (2013): The Augmented Dickey-Fuller (ADF) test constructs a parametric correction for higher-order correlation by assuming that the y series follows an AR (p) process and adding p lagged difference terms of the dependent variable y to the right-hand side of the test p regression: y t = αy t 1 + i=1 β i y t i + x t δ + υ t (2) This augmented specification is then used to test the null and alternative hypothesis - H 0 α = 0 and H 1 α < 0 (p ). 2. Philips Perron (PP) Unit Root Test IHS (2013): Phillips and Perron (1988) propose an alternative (nonparametric) method of controlling for serial correlation when testing for a unit root. The PP method estimates the non-augmented DF test equation: y t = αy t 1 + x t δ + ε t (3) and modifies the t-ratio of the α coefficient so that serial correlation does not affect the asymptotic distribution of the test statistic. The PP test is based on the statistic: t α = t α ( γ 0 f 0 ) 1 2 T(f 0 γ 0 )(se(α )) 2 f s (4) Where α is the estimate and t α the t-ratio of α, se(α ) is coefficient standard error, and s is the standard error of the test regression (p. 478). 3. Kwiatkowski Phillips Schmidt Shin (KPSS) Test IHS (2013): The KPSS (1992) test differs from the other unit root tests described here in that the series y t is assumed to be (trend-) stationary under the null. The KPSS statistic is based on the residuals from the OLS regression of y t on the exogenous variables- x t : y t = x t δ + u t (5) The LM statistic is defined as: LM = S(t) 2 (T 2 f 0 ) t (6)

8 Efficiency, Determinants and Performance of Service Sector Stocks Evidence from India Where,f 0 is an estimator of the residual spectrum at frequency zero and where S(t) is a cumulative residual function: S(t) = t r=1 u r and are based on the residuals given by: u t = y t x t δ (0) (p. 479). Note: Three models of estimating ADF and PP tests are available, i.e., trend and intercept; neither trend nor intercept and only intercept. For KPSS, first two variants, i.e., intercept only and trend and intercept are feasible. We have first seen the significance of trend and intercept coefficients for all variables before choosing the model to apply. If for a variable, the trend is found to be significant (Prob. value < 0.05), we have applied the trend and intercept model. If trend is not significant but intercept is significant, we have used the intercept only model. If neither trend nor intercept are significant for a variable, we have used the none (neither trend nor intercept) model Variance Ratio Tests We employ the Lo-MacKinlay (LM), Wright s Rank and Wright s Sign Variance Ratio Tests. 1. Lo-MacKinlay (LM) Test IHS (2013): Lo and MacKinlay (1988) formulate two test statistics for the random walk properties that are applicable under different sets of null hypothesis assumptions about t : First, Lo and MacKinlay make the strong assumption that the t are i.i.d. Gaussian with variance σ 2. Alternately, Lo and MacKinlay outline a heteroskedastic random walk hypothesis where they weaken the i.i.d. assumption and allow for fairly general forms of conditional heteroskedasticity and dependence. (p ). 2. Wright s Rank Test IHS (2013): Wright (2000) proposes modifying the usual variance ratio tests using standardized ranks of the increments, Y t. Letting r ( Y t ) be the rank of the Y t among all T values, we define the standardized rank (r 1t ) and van der Waerden rank scores, (r 2t ):r 1t = (r ( Y t ) T+1 and r 2t = Φ 1 (r (Δ Y t ) (T + 1) ) (7) ) 2 (T 1)(T+1) 12 In cases where there are tied ranks, the denominator in r 1t may be modified slightly to account for the tie handling. The Wright variance ratio test statistics are obtained by computing the Lo and MacKinlay homoskedastic test statistic using the ranks or rank scores in place of the original data (p ). 3 Wright s Sign Test IHS (2013): Wright also proposes a modification of the homoskedastic Lo and MacKinlay statistic in which each Y t is replaced by its sign. This statistic is valid under the martingale difference sequence (m.d.s.) null hypothesis, and under the assumption that μ = 0, the exact sampling distribution may also be approximated using a permutation bootstrap (p. 504). Note: For applying various unit root and variance ratio tests, all indices values have been converted on logarithmic scale by taking their natural log.

9 Conference Proceedings of 3 rd International Conference on Booming Service Sector: From Achievements to Growth Prospects, Pages: 23-36, organised by SGGSCC, DU on Feb. 4-5, Determinants We have used the Multiple Regression Model to identify the broad macroeconomic and stock market variables which are significant in predicting and explaining Indian Service sector stock returns. Our multiple regression models takes the broad service sector index and all nine individual service sector indices [Bank, Energy, Finance, Health Care, IT, Power, Realty, TECK and Telecom] as dependent variables. Each of these dependent variables are then regressed on a set of independent variables comprising the broad Indian and global equity indices [SENSEX and ACWI] and Macroeconomic Factors [IIP, CPI, IR and ER]. So, ten multiple regression models have been run. Thus, we have the following multiple regression equations: Y i = α 1 + β 1 ACWI i + β 2 SENSEX i + β 3 IIP i + β 4 CPI i + β 5 IR i + β 6 ER i + e i (8) Where, Y = Service, Bank, Energy, Finance, Health, IT, Power, Realty, TECK and Telecom. The most probable determinants are the regressors, i.e., broad market stock indices (ACWI & SENSEX +) and prominent macroeconomic variables, viz., Industrial Production (IIP+), Inflation (CPI-), Interest Rate (IR-) and Exchange Rate (ER+/-). We have used stationary series [I(1)] of all the dependent & independent variables. Regressors in the models were also checked for presence of multi-collinearity and no problem was found Performance Evaluation To measure and compare the performance of Indian service sector indices with global and Indian broad market equity indices [ACWI & Sensex], we have used the following three Risk Adjusted Measures for Performance Evaluation: Sharpe Ratio, Treynor Ratio and Jensen s Alpha. 1. Sharpe Ratio This ratio measures the return of the portfolio in excess of risk-free rate, compared to total risk of the portfolio. Since it uses standard deviation as a measure of risk, it does not assume that the portfolio is well diversified. If AR P is the average monthly portfolio return, R F the monthly risk free return and σ P total risk of portfolio, then Sharpe ratio can be calculated as: Sharpe ratio = AR P R F σ P (9) 2. Treynor Ratio This ratio measures the relationship between return of the portfolio, above the risk-free rate, and its systematic risk indicated by portfolio beta (β P ). This ratio is particularly relevant for evaluating the performance of well diversified portfolio, since it considers only the systematic risk. Treynor ratio = AR P R F β P (10) 3. Jensen s Alpha It is used to determine the abnormal return (α) of a security or portfolio of securities over the theoretical expected return. The theoretical return is predicted by the market model (CAPM).

10 Efficiency, Determinants and Performance of Service Sector Stocks Evidence from India A portfolio with a consistently positive excess return (adjusted for risk) will have a positive alpha and vice-versa. It can be calculated as: R P R F = α + β P (R M R F ) (11) Or, α = R P [R F + (R M R F ) β P ] (12) R P : Average monthly index return; R F : Monthly Risk free return; R M : Average monthly market return; σ P : Total index risk; and β P : Index beta. 5. Empirical Analysis & Results 5.1 Descriptive Statistics Descriptive Statistics results provided in Table 2 shows that for the study period, the highest return is provided by Healthcare sector and lowest by Realty. Absolute risk among stock indices is highest of Realty and lowest of ACWI followed by Healthcare. The relative risk is - also highest for Realty and lowest for Healthcare among the stock indices. Table 2: Descriptive Statistics (All Variables) Variables Mean Std. Dev. Coefficient of Variation Skewness Kurtosis ACWI SENSEX SERVICE BANK ENERGY FINANCE HEALTH IT POWER REALTY TECK TELECOM IIP CPI IR ER Efficiency Results Unit Root Test Results The results of ADF, PP and KPSS unit root tests at level are provided in Table 3 and of those at first difference are provided in Table 4. As per the unit root tests conducted at level, all the service sector stock indices (except Energy) are non-stationary at level, follow random walk and are weak form efficient. However, when their first difference is tested we find that all of them are stationary, i.e., I(1) [integrated of order 1]. They being I(1), can be used in further regression analysis without worrying about emergence of any spurious relationship.

11 Conference Proceedings of 3 rd International Conference on Booming Service Sector: From Achievements to Growth Prospects, Pages: 23-36, organised by SGGSCC, DU on Feb. 4-5, Table 3: Unit Root Tests Results (At Level) ADF Test PP Test KPSS Test Variable Adjusted t-stat. Prob. t-stat. Prob. LM-Stat. t-stat. Prob. ACWI * 0.00 SENSEX * 0.00 SERVICE * 0.00 BANK * 0.00 ENERGY -3.02* * * 0.00 FINANCE * 0.00 HEALTH * 0.00 IT * 0.00 POWER * 0.00 REALTY * 0.00 TECK * 0.00 TELECOM * 0.00 IIP * 0.00 CPI * 0.00 IR -4.29* * * 0.00 ER * 0.00 *Denotes significant at α = Table 4: Unit Root Tests Results (First Difference) ADF Test PP Test KPSS Test Variable Adjusted t-stat. Prob. t-stat. Prob. LM-Stat. t-stat. Prob. ACWI -8.89* * SENSEX -9.94* * SERVICE * * BANK -9.62* * ENERGY * * FINANCE -9.33* * HEALTH * * * 0.01 IT * * POWER -9.48* * REALTY -8.98* * TECK * * TELECOM * * IIP * CPI -6.75* * * 0.00 IR * * ER -9.67* * *Denotes significant at α = Variance Ratio Test Results To further confirm he unit root test results, we conduct LoMackinlay, Wright s Ranks and Wright s Signs Variance Ratio tests. Their results provided in Table 5 affirm that Indian

12 Efficiency, Determinants and Performance of Service Sector Stocks Evidence from India service sector indices are indeed martingale or weak form efficient as the null hypothesis of Random Walk or Martingale is overwhelmingly accepted in all cases. Table 6 summarises the results of all unit root and variance ratio tests. Taken together on a majority basis, these tests affirm that all the Indian service sector indices are Weak Form Efficient. Table 5: Variance Ratio Test Results Lo-MacKinlay Test Wrights Test (Ranks) Wrights Test (Signs) Variable Wald (Chi- Wald (Chi- Max z Stat. Prob. Prob. Square) Square) Prob. ACWI SENSEX * 0.02 SERVICE BANK ENERGY FINANCE * 0.04 HEALTH IT POWER REALTY TECK TELECOM *Denotes significant at α = Stock Indices Table 6: Summary of Unit Root and Variance Ratio Test Results ADF Test Unit Root Tests PP Test KPSS Test Lo- MacKinlay Test Variance Ratio Tests Wright s Rank Test Wright s Sign Test ACWI WE WE WE WE WE WE SENSEX WE WE WE WE WE NWE SERVICE WE WE WE WE WE WE BANK WE WE WE WE WE WE ENERGY NWE NWE WE WE WE WE FINANCE WE WE WE WE WE NWE HEALTH WE WE WE WE WE WE IT WE WE WE WE WE WE POWER WE WE WE WE WE WE REALTY WE WE WE WE WE WE TECK WE WE WE WE WE WE TELECOM WE WE WE WE WE WE WE: Weak Form Efficient, NWE: Not Weak Form Efficient. 5.3 Determinants Results and Model summary of Multiple regression analysis [Equation 8] is provided in Table 7 (Panels A and B). Results indicate that as per our expectations, Sensex has a significant positive relationship with all sectoral indices. Also, ACWI has positive significant relation but only with IT and TECK. As we earlier stated, Exchange rate has significant positive impact on export oriented sectors of Healthcare, IT and TECK. It has significant negative effect on Bank, Finance and Power. Industrial production (IIP) has significant

13 Conference Proceedings of 3 rd International Conference on Booming Service Sector: From Achievements to Growth Prospects, Pages: 23-36, organised by SGGSCC, DU on Feb. 4-5, positive on broad service index and TECK sector. Inflation and Interest rate has no significant relationship with any of the service sector indices. Model summary reveals that all regression models have overall statistical significance (All F- Stats are highly significant even at 1%) and no or very low Autocorrelation (Durbin-Watson Stat. is less than or marginally above 2, except in case of IT). All the regression models also have high explanatory power ranging from 52% in Health to 95% for the overall service sector index. Table 7: Summary Results of Individual Regressions of Service Sector Stock Indices A: Multiple Regression Model Results Regressand Regressors Constant SENSEX ACWI IIP CPI IR ER SERVICE * ** BANK * * ENERGY * FINANCE * * HEALTH * * IT * 0.25* * POWER * * REALTY * TECK * 0.15** 0.15* * TELECOM * Values are Regression Coefficients. * Denotes significant at 5%. ** Denotes significant at 10% level. B: Multiple Regression Models Summary Regression R 2 Adjusted R 2 Durbin- F-Statistic Probability Model Watson Stat. SERVICE * BANK * ENERGY * FINANCE * HEALTH * IT * POWER * REALTY * TECK * TELECOM * * Denotes significant at 5% level. SERVICE

14 Efficiency, Determinants and Performance of Service Sector Stocks Evidence from India BANK ENERGY FINANCE HEALT H

15 Conference Proceedings of 3 rd International Conference on Booming Service Sector: From Achievements to Growth Prospects, Pages: 23-36, organised by SGGSCC, DU on Feb. 4-5, IT POWER REALT Y TECK

16 Efficiency, Determinants and Performance of Service Sector Stocks Evidence from India TELECOM Performance The risk adjusted performance measures [Sharpe Ratio, Treynor Ratio and Jensen s Alpha] along with comparative ranks for all the stock indices is provided in Table 8. As in case of descriptive statistics, the highest return on risk adjusted basis is provided by Healthcare sector and the lowest is provided by realty sector. While, about half of the service sector indices beat the Indian equity market benchmark Sensex, almost all of them (except Reality) outperform the global benchmark ACWI. Table 8: Risk Adjusted Measures of Different Stock Indices Stock Index Sharpe Rank Treynor Rank Jensen s Rank Ratio (%) Ratio (%) Alpha (%) ACWI SENSEX SERVICE BANK ENERGY FINANCE HEALTH IT POWER REALTY TECK TELECOM Conclusion and Policy Implications In this paper, a comprehensive examination of Indian Service Sector Stocks from various dimensions, viz., Weak Form Efficiency, Determinants and Performance was undertaken. We studied service sector through nine sectoral indices of BSE (Bank, Energy, Finance, Health Care, IT, Power, Realty, TECK and Telecom) and one broad thematic Service Sector Index of NSE using monthly data from September 2005 to October 2015.

17 Conference Proceedings of 3 rd International Conference on Booming Service Sector: From Achievements to Growth Prospects, Pages: 23-36, organised by SGGSCC, DU on Feb. 4-5, We measure their adherence to Random Walk Hypothesis (Weak Form Efficiency) through a battery of Unit Root Tests [Augmented Dickey Fuller (ADF), Phillips Perron (PP) & Kwiatkowski-Phillips-Schmidt-Shin (KPSS)] and Variance Ratio Tests [Lo-MacKinlay, Wright s Ranks & Wright s Signs). We further employ proxies for Stock Market [BSE Sensex for Indian Stock Market & MSCI All Country World Index (MSCI ACWI) for Global Equity Markets] and prominent Macroeconomic Variables [Index of Industrial Production (IIP), Inflation (CPI), Interest Rate, Exchange Rate] as Regressors through Multiple Regression Model to find out the major determinants of Indian Service Sector Stock Returns. Finally, we also evaluate the performance of these Service sector Indices through three popular Risk Adjusted Performance measure ratios, i.e., Sharpe Ratio, Treynor Ratio and Jensen s Alpha. Overall, the Efficiency tests reveal that all Indian Service Sector Indices are indeed following the Random Walk Hypothesis and are Integrated of order one [I(1)] ruling out the possibility of spurious relationship. All the Regression models were highly significant with very high explanatory powers and no problem of multicollinearity or autocorrelation. Across the board, Exchange Rate from amongst the Macroeconomic Variables and Sensex from Stock Market Variables came out to be the significant determinants of Indian Service Sector Stock Returns. Performance evaluation measures indicated that almost half of the service sector indices lead by Healthcare could significantly outperform the broader market on risk adjusted basis. These results are relevant and have pertinent implications for a broad range of stakeholders related to the field including policy makers, regulators, investors, academicians and researchers. Regulators and Policy makers are worth praising as their decades of efforts in bringing transparency and reforms have resulted in Weak Form Efficiency even at the sectoral level in Indian Stock Market. Exchange rate and Sensex, can provide important clues for future price movements and formulation of investment strategies in case of almost every service sector index. For the investors, Healthcare, IT, Banking and Finance in particular offers promising returns as they comprehensively beat Indian and Global equity benchmarks on all risk adjusted performance measures. This study will also be a valuable contribution to the almost non existent empirical literature on the subject. 7. References Campbell, J; Lo, A. and MacKinlay, A. (1997). The Econometrics of Financial Markets, 2nd Edition. Princeton University Press. Dickey, D. A. & Fuller, A. (1979). Distributions of the estimators for autoregressive time series with a unit root. Journal of American Statistical Association, 74(366), Fama, E. (1965). The behaviour of stock market prices. Journal of Business, 38, Fama, E. (1970). Efficient capital market: A review of theory and empirical work. Journal of Finance, 25(2), IHS Global Inc. (2013). Eviews 8 User s Guide II. Web-link: Lo, A. & MacKinlay, C. (1988). Stock market prices do not follow random walks: evidence from a simple specification test. Review of Financial Studies, 1,

18 Efficiency, Determinants and Performance of Service Sector Stocks Evidence from India Phillips, P. & Perron, P. (1988). Testing for a unit root in time series regression. Biometrica, 75(2), Tripathi, V., & Bhandari, V. (2015a). Socially Responsible Stocks- A Boon for Investors in India. Journal of Advances in Management Research, Emerald publications, Vol. 12, No. 2, Tripathi, V., & Bhandari, V. (2015b). Do Ethical Funds Underperform Conventional Funds? - Empirical Evidence from India. International Journal of Business Ethics in Developing Economies (Forthcoming). Tripathi, V. & Kumar, A. (2014). Sectoral Efficiency of the Indian Stock Market and The Impact of Global Financial Crisis. Journal of Commerce & Accounting Research, 3(4), Tripathi, V., & Kumar, A. (2015a). Short Run Causal Relationship between Inflation and Stock Returns - An Empirical Study of BRICS markets. Asian Journal of Management Applications and Research, Vol.05, No.01. Tripathi, V., & Kumar, A. (2015b). Relationship between Inflation and Stock Returns Evidence from BRICS markets using Panel Cointegration Test. International Journal of Accounting and Financial Reporting, Macro Think Institute, USA, Vol. 4, No. 2, Tripathi, V., & Kumar, A. (2015c). Do Macroeconomic Variables affect stock returns in BRICS Markets? An ARDL Approach. Journal of Commerce & Accounting Research, Macro Think Institute, USA, Vol. 4, Issue 2. Tripathi, V., & Seth, R. (2014). Stock Market Performance and Macroeconomic Factors: The Study of Indian Equity Market. Global Business Review, Vol. 15, No. 2, Vieito, J. P., Bhanumurthy, K. V. & Tripathi, V. (2013). Market efficiency in G-20 countries: the paradox of financial crisis. Annals of Financial Economics, 8(1), Wright, J (2000). Alternative variance-ratio tests using ranks and signs. Journal of Business & Economic Statistics, 18, 1-9.