CORPORATE EARNINGS AND STOCK RETURNS IN INDIA: AN EMPIRICAL STUDY
gfte/314
Samitted to
GOA UNIVERSITY
For the Award of the Degree of Doctor of Philosophy
in
Commerce
L W. Seigefict
lbulen the Sup vivb ion
o Oft. `y. V. sedd g
DEPARTMENT OF COMMERCE GOA UNIVERSITY
GOA - 403 206
2009 4
/ //
preparation of this thesis.
DECLARATION
I, Rocky W. Rebello, hereby declare that the thesis titled "CORPORATE EARNINGS AND STOCK RETURNS IN INDIA: AN EMPIRICAL STUDY"
submitted to the Goa University, Goa, for the award of the Degree of Doctor of Philosophy is the outcome of original and independent research work undertaken by me during the period 2006 - 09. This study is carried out under the supervision and guidance of Dr. Y. V. Reddy, Reader, Department of Commerce, Goa University. It has not been previously formed the basis for the award of any degree, diploma or certificate of this or any other Universities. I have duly acknowledged all the sources used by me in the
—10 11,
1°
1>
41.1
/
Date : 27/04/2009
Place : Goa R. W. Rebello
r. 1X. V. Reddy Supervisor
Dr. Y. V. Reddy
Department of CommerceReader Goa University, Goa
CERTIFICATE
This is to certify that the thesis titled "CORPORATE EARNINGS AND STOCK RETURNS IN INDIA: AN . EMPIRICAL STUDY" for the award of Ph. D. Degr .ee in Commerce, is the bonafide record of the original work done by Shri Rocky W. Rebello, during the period of study under my supervision. This thesis has not formed the basis for the award of any degree, diploma, certificate, associateship, fellowship or similar title to the candidate of this University or any other Universities.
Date : 27/04/2009 Place : Goa
E GEMENTS
With a deep sense of gratitude, I acknowledge the guidance and wholehearted support of my guide and mentor Dr. Y. V. Reddy, Reader, Department of Commerce, Goa University. It was he who motivated me to work on the Artificial Neural Network, a challenging field of study in my subject.
Discussions with him, for which he ever willingly gave of his time, were always a stimulating experience.
Without doubt, the present study has been the culmination of the encouragement given by several individuals and organizations. First, I would like to place on record my gratitude to Prof. B. Ramesh, Head, Department of Commerce, Goa University, for his permission to register with Goa University for my Ph. D. study. He was kind enough to interact with me on the subject on several occasions and offer helpful guidance. I am also thankful to Dr. Anjana Raju, member of the FRC for her valuable suggestions and Dr. K. B. Subhash for his assistance. A word of appreciation is also due to the administrative staff for their willing co-operation.
I am indebted to my friend, colleague and mentor Dr. K. G. Rajan, Reader, Department of Statistics, for his unstinting support throughout the course of my study. He opened up new vistas for me and showed me the light whenever I faltered. Later, he painstakingly reading my drafts and suggested effective changes.
4
•
A special thanks to my colleagues, Yougesh Karunakar, Department of Information Technology, for assisting me in using the ANN Software; Heena Shukla, Department of English, for reading the manuscript and giving vital inputs; Dr. Sanjeevini Gharge, Department of Mathematics, for her readiness-to help; and other staff members of my college who have directly or indirectly been a source of motivation in my endeavour.
I am also grateful to the Principal and the Management of Sheth L. U. J. & Sir M. V.
College of Arts, Science, and Commerce, Andheri (East), Mumbai 400 069 for their support.
Mention must be made of NITIE Mumbai, IIT Mumbai, Department of Commerce, University of Mumbai, and the ICFAI Business School, Hyderabad, who have all given me opportunities to share my views during the seminars organized by them. I acknowledge the assistance provided by the Librarian, NMIMS University, Vile Parle, Mumbai 400 056, in allowing me to use its database facilities.
Last but not the least, this work would never have been possible without the love and support of my family and friends. Thank you for believing in me.
Date: 27/04/2009 R. W. Rebello
CONTENTS
Page a. Declaration by Candidate
b. Certificate by Supervisor ii
c. Acknowledgements iii
d. Contents
e. List of Tables vii
f. List of Figures xi
g. Abbreviations Used xiii
CHAPTER 1
1 INTRODUCTION 1
1.1 Preliminaries 2
1.2 Literature Review 11
1.3 Research Problem 19
1.4 Significance of the Problem 20
1.5 Objectives 20
1.6 Hypotheses 20
1.7 Methodology 21
1.8 Limitations 23
1.9 Chapter Scheme 23
1.10 References 25
CHAPTER 2
2 MULTIVARIATE REGRESSION ANALYSIS 29
2.1 Introduction 30
2.2 Sources of Data 35
2.3 Sampling Design 35
2.4 Data Analysis 42
2.5 Results and Interpretations 54
1
CHAPTER 3
3 PROBABILISTIC GROWTH MODEL 62
3.1 Introduction 63
3.2 Sources of Data 69
3.3 Sampling Design 71
3.4 Data Analysis 75
3.5 Results and Interpretations 125
3.6 References 135
CHAPTER 4
4 ARTIFICIAL NEURAL NETWORK MODEL 138
4.1 Introduction 139
4.2 Sources of Data 176
4.3 Sampling Design 178
4.4 Data Analysis 180
4.5 Results and Interpretations 207
4.6 References 216
CHAPTER 5
5 SUMMARY AND CONCLUSIONS 224
5.1 Summary 225
5.2 Conclusions 239
5.3 Recommendations 240
5.4 Scope for Further Research 243
BIBLIOGRAPHY 244
APPENDICES 271
• • • • • • •
LIST OF TABLES
2.1 2.2
2.3
Pilot Study Results
Average, Maximum and Minimum Variation for Closing Price - Category' Stocks)
Variation of Forecast Stock Price
Page 43
44
•
452.4 Variation in Closing Price (> 3%) based on Stock Prices 46 2.5 Industry-wise Variation CA - Category' Stocks) 47 2.6 Average, Highest, Lowest and Standard Deviation with Lower
and Upper limits CA - Category' Stocks) 52
2.7 R2 and p-Value of Different Firms CA - Category' Stocks) 54 3.1 Average Variation of Sensex Stocks (All Periods) 75 3.2 Maximum and Minimum Variation of Sensex Stocks
(First Period) 76
3.3 Variation for Sensex Stocks (First Period) 77 3.4 Variation in Prices (> 3%) based on Stock Prices (First Period) 79 3.5 Maximum and Minimum Variation of Sensex Stocks
(Second Period) 79
3.6 Variation for Sensex Stocks (Second Period) 81 4
3.7 Variation in Prices (> 3%) based on Stock Prices (Second Period) 82
3.8 Maximum and Minimum Variation of Sensex Stocks
(Third Period) 83
3.9 Variation for Sensex Stocks (Third Period) 84 3.10 Variation in Prices (> 3%) based on Stock Prices (Third Period) 85 3.11 Maximum and Minimum Variation of Sensex Stocks
(Fourth Period) 86
3.12 Variation for Sensex Stocks (Fourth Period) 87
•
3.13 Variation in Prices (> 3%) based on Stock Prices(Fourth Period) 89
3.14 Average Variation of Nifty Stocks 90
3.15 Maximum and Minimum Variation of Nifty Stocks (First period) 91 3.16 Variation for Nifty Stocks (First Period) 92 3.17 Variation in Prices (> 3%) based on Nifty Stock Prices
(First Period) 93
*
3.18 Maximum and Minimum Variation of Nifty Stocks(Second period) 94
3.19 Variation for Nifty Stocks (Second Period) 96 3.20 Variation in Prices (> 3%) based on Nifty Stock Prices
(Second Period) 97
3.21 Maximum and Minimum Variation of Nifty Stocks (Third period) 98 3.22 Variation for Nifty Stocks (Third Period) 99 3.23 Variation in Prices (> 3%) based on Nifty Stock Prices (Third Period) 100
3.24 Maximum and Minimum Variation of Nifty Stocks
(Fourth period) 101
3.25 Variation for Nifty Stocks (Fourth Period) 103 3.26 Variation in Prices (> 3%) based on Nifty Stock Prices
(Fourth Period) 104
3.27 Average, Maximum and Minimum Variation for all Prices
- Category' Stocks) 105
3.28 Variation for 'A - Category' Stocks 107
3.29 Variation (> 3%) based on Stock Prices CA - Category' Stocks) 108 3.30 Industry-wise Variations CA - Category' Stocks) - Closing Price 110 3.31 Industry-wise Variations CA - Category' Stocks) - High Price 112 3.32 Industry-wise Variations (A - Category' Stocks) - Low Price 114 3.33 Average, Highest, Lowest and Standard Deviation with Lower
and Upper limits (Closing Price) 117
3.34 Average, Highest, Lowest and Standard Deviation with
Lower and Upper limits (High Price) 120
3.35 Average, Highest, Lowest and Standard Deviation with
Lower and Upper limits (Low Price) 122
3.36 p - Value of Firms 124
4.1 Average Variation of Sensex Stocks (Specific Probes) 180 4.2 Average Variation of Sensex Stocks (All Probes) 181 4.3 Variation for Sensex Stocks (Specific Probes) 182
4.4 Variation in Prices (> 3%) based on Sensex Stocks Prices
(Specific Probes) 183
4.5 Variation for Sensex Stocks (All Probes) 184 4.6 Variation in Prices (> 3%) based on Stock Prices (All Probes) 186 4.7 Averages, Maximum and Minimum Variation of 'A - Category'
Stocks 187
4.8 Variation for 'A - Category' Stocks 188
4 4.9 Variation in Prices (> 3%) based on Stock Prices CA - Category'
Stock) 190
4.10 Industry-wise Variations CA - Category' Stocks) - Closing Price 192 4.11 Industry-wise Variations CA - Category' Stocks) - High Price 194 4.12 Industry-wise Variations CA - Category' Stocks) - Low Price 196 4.13 Average, Highest, Lowest and Standard Deviation with Lower
and Upper limits CA - Category' Stocks) - Closing Price 200 4.14 Average, Highest, Lowest and Standard Deviation with Lower
and Upper limits (A - Category' Stocks) - High Price 202 4.15 Average, Highest, Lowest and Standard Deviation with Lower
and Upper limits (A - Category' Stocks) - Low Price 204 4.16 R2 Value of Different Firms CA - Category' Stocks) 206 5.1 Comparative Averages, Maximum and Minimum Variation
- Category' Stocks) 236
5.2 Comparative Variations for 'A - Category' Stocks 237 ff%
4
LIST OF FIGURES
Page 2.1 'A - Category", Eligible and Sample Firms 38 2.2 'A - Category', Eligible and Sample Industries 38
2.3 Firms Representing Industries 39
2.4 Firms Representing Industries 40
2.5 Industry Names and Number of Firms (Sample) 41
3.1 Normal Distribution 63
3.2 Cumulative Distribution Function 65
4.1 History of Artificial Neural Network 142
4.2 Neuron Forming a Chemical Synapse 144
4.3 Neuron Forming a Chemical Synapse 145
4.4 Artificial Neuron 150
4.5 Artificial Neuron 152
4.6 Feed-forward Network 153
4.7 Symmetrical Curve 156
4.8 Practical Application of ANN 159
4.9 Transfer Functions 165
4.10 MSE with Training Epoch 167
-4.11 Desired Output and Actual Network Output 168 4.12 Artificial Neuron Using Backpropagation Learning 170
4.13 Generalized Feed Forward Network 175
5.1 Variation of Forecast Stock Price (MRA) 227 1
5.2 Variation of Forecast Stock Price (PGM) 229 5.3 Variation of Forecast Stock Price (ANN) 233 5.4 Comparative Averages, Maximum and Minimum Variation (%) 236 5.5 Comparative Variations for 'A - Category' Stocks 238
4
ABBREVIATIONS USED
ANN Artificial Neural Network
APT Arbitrage Pricing Theory
ARCH Autoregressive Conditional
Heteroskedasticity Automobiles - 2/3 Automobiles 2/3 wheeler Automobiles - 4 Automobiles 4 wheeler
BPN Backpropagation Networks
BSE Bombay Stock Exchange
CAPM Capital Asset Pricing Model
Chemicals - I Chemicals Inorganic
Chemicals - 0 Chemicals Organic
Chemicals - S Chemicals Synthetic
Computer - H Computer Hardware
Computer - S Computer Software
DAX Deutscher Aktien IndeX (Germany Stock
Index)
DARPA Defense Advanced Research Projects Agency
E/P Earning to Market Price Ratio
EDIFAR Electronic Data Information Filing and
Retrieval System Electrical Equip Electrical Equipments
ESOP Employee Stock Options
FTSE-100 Financial Times and the London Stock Exchange Index
GA Genetic Algorithms
GANN Genetic Algorithm Neural Networks
GARCH Generalized Autoregressive Conditional
Heteroskedasticity,
GFN Generalized Feedforward Networks
GMM Generalized Method of Moment
HMM Hidden Markov Model
M/B Market to Book Value Ratio
MA Moving Average
MACD Moving Average Convergence/ Divergence
MAD Mean Absolute Deviation
MAPE Mean Absolute Percentage Error
MFI Money Flow Index
MIT Massachusetts Institute of Technology
MLP Multi Layer Perceptrons
MRA Multivariate Regression Analysis
MSE Mean Squared Error
MVA Multivariate Analysis
NASDAQ National Association of Securities
Dealers Automated Quotations
NBER National Bureau of Economic Research
NIC National Informatics Centre
NIFTY NSE 50 Index
Nikkei 225 Nihon Keizai Shimbun (Nikkei), Tokyo Stock Exchange Index
NSE National Stock Exchange
P/E Market Price to Earning ratio
PE Processing Elements
PGM Probabilistic Growth Model
PSP Post Synaptic Potential
RBF Radial Basis Function
RBI Reserve Bank of India
RSI Relative Strength Index
SP 500 Standard & Poor's Index (New York
Stock Exchange)
S & P CNX Standard & Poor's CRISIL NSE Index
Sensex Sensitivity Index (BSE Mumbai)
SETAR Self-Exciting Threshold Autoregression
Models
SO Stochastic Oscillator
SPSS Statistical Package for Social Sciences
SSRN Social Science Research Network
Tanh Hyperbolic Tangent
Textiles - C Textiles Cotton
Textiles - S Textiles Synthetic
VAR Vector Autoregressive Model
i
AODLCir k
Chapter Plan
• Preliminaries
• Literature Review
• Research Problem
• Significance of the Problem Objectives
• Hypotheses
• Methodology
• Limitations
• Chapter Scheme
• References
C ED7
1. INTRODUCTION
1.1 PRELIMINARIES
0 One of the earliest and most enduring questions of modern theory of finance is whether financial asset prices can be predicted. Perhaps because of the obvious analogy between financial investments and games of chance, mathematical models of asset prices have an unusually rich history that predates virtually every other aspect of economic analysis. That many prominent mathematicians and scientists have applied their considerable skills for forecasting financial securities' prices testifies to the fascination for and the challenges of this problem.
Indeed, modern financial economics is firmly rooted in early attempts to beat the market - an endeavour that is still of current interest and a matter of hot debate in publications, conferences and cocktail parties.
In general, stock pricing models provide the relationship between the not so well defined variables for a given financial market. There have been several attempts in this direction, but there is no unanimity in identifying the variables, as researchers and investors are constantly bombarded es
with vast quantities of diverse information. This study, then, attempts to
identify the factors, which influence the stock prices more intensely than others do.
Before attempting to identify these factors, it is advisable to take a journey into the history of stocks. Though it is believed that recording of financial transactions came into being as early as 9000 B.C. to 8000 B.C., there is no evidence to prove the existence of such a system. However, from around 2500 B.C. to 1800 B.C., cuneiform — i.e. writing on clay tablets with a reed similar to a stylus - came into use extensively, especially for financial transactions [Edward Chancellor (1999)11. During this period in Mesopotamia, there was a substantial amount of economic activity in agriculture, crafts, ranching, trading, etc. The first bond transactions were documented in cuneiform, where silver had been lent out to a business, and that loan had been transferred to another individual. In addition, the earliest stock or share transactions were also documented in cuneiform, for funding maritime trade expeditions.
Stock exchanges originally existed in the form of 'Euro-Fairs' trading in agricultural and other commodities during the Middle Ages. Credit was commonly given, and therefore supporting documents such as drafts, notes and bills of exchange were created. These were the precursors to modern stock and bond certificates.
During the Roman period, the empire contracted out many of its services to private groups called publicani [Edward Chancellor (1999)11. Shares
in publicani were called `socir (for large co-operatives) and 'particulae', (for over-the-counter shares of small companies). Though the records available of this time are incomplete, Edward Chancellor (1999) 1 states in his book "Devil Take the Hindmost" that there is some evidence that a speculation in these shares became increasingly widespread and that perhaps the first ever speculative bubble in 'stocks' occurred.
During the seventeenth century, certificates of ownership of business came into existence. The first company to issue shares of stock after the Middle Ages was the Dutch East India Company in 1606 [Edward Chancellor (1999)]. The innovation of joint ownership made a great deal of Europe's economic growth possible. The technique of pooling capital to finance the building of ships, for example, made the Netherlands a maritime superpower. Before the adoption of the joint- stock corporation, an expensive venture such as the building of a merchant ship could be undertaken only by governments or by very wealthy individuals or families.
Economic historians found the Dutch stock market of the 1600s particularly interesting: there was clear documentation of the use of stock futures, stock options, short selling, the use of credit to purchase shares, a speculative bubble that crashed in 1695 and changes in trading patterns. Edward Stringham et al (2008)2 also noted that practices such as short selling continued to occur during this time despite the government passing laws against it. This was unusual because it shows individual
parties fulfilling contracts that were not legally enforceable and where the parties involved could incur a loss. Stringham argues that contracts can be created and enforced without state sanction or, in this case, in spite of laws to the contrary.
Since the days of the advancement of the stock market, there has been a relentless effort to unravel the mystery of stock prices and the direction of stock price movements. The few who could predict the direction accurately have benefited from such predictions and created wealth. In pursuit of this goal, several financial economists and market practitioners have attempted to evolve methods and techniques, which would help them to forecast stock prices accurately. However, their efforts were not entirely fruitful and the solution to the mystery continued to elude the players of the stock market.
The famous dramatist Oscar Wilde (1900) 3 once described a cynic as one who "knows the price of everything, but the value of nothing". This description holds good for some analysts and many investors who subscribe to the theory of the 'big fool', which argues that the value of a stock is irrelevant as long as there is a 'bigger fool' around willing to buy the stock from them. While this may provide a basis for some profits, it is a dangerous game to play since there is no guarantee that the latter will still be around, when the time comes to sell.
Equity market professionals use a wide range of analyses to help them make informed trading and investment decisions. Many wish to compare current and historical market situations or review the past performance of an instrument or index. They need tools that draw on real-time and historical stock quotes to enable them to perform these types of analyses.
Technical analysis charts track the historical evolution of stock quotes, trading volumes and other indicators of activity. Technical analysts try to identify buy and sell signals by looking at historical stock market actions.
They pay attention to recurring patterns in historical price movements, to trends and their speed or momentum when making stock trading recommendations.
Relative performance charts, which are also based on historical and real- time stock quotes, enable users to compare the performance of stock quotes against their peers or against sectors or indices over a selected period. Other charts allow users to review how the market moved in the past when certain fundamental levels were reached. An index-earnings growth chart, for example, shows the relationship between earnings growth and stock quotes for the index as a whole. These charts help users identify buying and selling opportunities.
Institutions trading in the equity markets take data-feeds of real-time and historical stock quotes to power their own deSktop applications, analytics
•
and research databases. They use historical end-of-the-day stock quotes for risk management and valuation purposes. Risk groups feed historical end-of-day stock quotes into their systems to run their daily risk reports. Mutual funds use historical end-of-day stock quotes to calculate the valuation of their holdings.
Some foreign equity information products, tailored to the different needs of different users, combine comprehensive news services, real-time market data and powerful analysis tools. They supply real-time and historical end-of-the-day stock quotes in flexible formats to enable institutions to pump market information into their applications and publics of organizations. They provide real-time equity quotes from several exchanges over a decade or two.
Valuating common stock is a complex process, but certainly worth the trouble for both investors and analysts. Over the years, two general
•
approaches have been developed. One method called the discounted cash flow approach estimates the stock's value based on the present value of its future cash flows, such as dividends, operating cash flows or free cash flows, while the other method values a stock based on its current price relative to certain variables such as the company's earnings, revenues or book value.Both the discounted cash flow approach and the relative valuation approach have certain factors in common. To start with, both techniques
are extensively impacted by the investors' required rate of return, because this rate is essentially the discount rate used in many valuation models.
In addition, all asset valuation techniques are influenced by the estimated growth rate of certain variables, such as dividends, earnings, cash flows or sales. When one of the variables has to be estimated, the result varies because variable inputs are likely to differ from one analyst to another. In other words, when evaluating a stock, prices are likely to be different because investors' required rates of return, as well as estimates of growth rates of earnings like dividends might be different.
A postulate of sound investing is that an investor does not pay more for a stock than its worth. This statement may seem logical and obvious, but it is forgotten and rediscovered at some time in every generation and in every market. There are those who are disingenuous enough to argue that 'value is in the eye of the beholder, and that any price can be justified if there are other investors willing to pay that price, which is
patently absurd. Perceptions may be all that matter when the asset is a painting or a sculpture, but investors do not (and should not) buy most assets for aesthetic or emotional reasons; stocks are acquired for the cash flows expected on them. Consequently, perceptions of value have to be backed by reality, which implies that the price paid for any stock must reflect the cash flows it is expected to generate. The models of valuation described in this study attempt to relate the stock value to the level and the expected growth of cash flows and the risk attached to them.
The proposed study is empirical in nature, aimed at studying the relationship between corporate returns (cash returns) and stock returns (market returns) so as to understand the relationship between earnings (cash flows) and stock prices. Cash earnings are considered in the place of accounting earnings (book profits) in order to avoid accounting bias. The basic aim of this study is to convey to the participants of the market that stock prices largely depend on fundamentals (earnings) rather than on rumours and political or economic events in society. The study also aims at suggesting to the participating firms that if they can release forecast data relating to their earnings for a future period on a continuous basis and disclose deviations thereof on completion of the said period, the stock prices could respond to the changes in earnings rather than to unanticipated elements. This information could make the stock market more transparent and robust, which would put the investors' confidence on a higher plane and hence the market would become more vibrant.
The present work focuses on the Indian Stock Market and studies only those stocks (large cap stocks), which are actively traded on the National Stock Exchange, Mumbai or the Bombay Stock Exchange, Mumbai with reference to the post liberalization period.
In India, it is generally believed that stock prices are not at all rooted in any fundamental factors, but driven by rumours, grapevine, manipulators, speculators, high net worth, institutional investors, etc.
However, in reality, although these factors do play some role in influencing the stock prices, in the long term the fundamentals generally influence the market price. Therefore, the proposed study is an attempt to bring to light the significant factors that influence the stock prices.
Various economists, while trying to understand the fluctuation in stock prices, are confronted with two major variables, viz. expected earnings and expected rate of return (cost of capital). In developed economies, there is a mechanism to evolve projected earnings for corporate sector.
However, in India there is no institutionalized mechanism to project future earnings for corporate firms except their own in-house mechanism. Therefore, such data are not available in the public domain.
If this study can establish the relationship explicitly, then the Regulating Authorities could be convinced to include the projected earnings in the disclosure norms. The second factor is risk free rate of return (cost of capital), which is also critical for valuation along with earnings. However, the cost of capital of a firm does not change as sadistically as the earnings. Therefore, it is assumed that the cost of capital of a firm remains stable during the short term. However, in the long term the cost of capital should be incorporated in the valuation process. Since the cost of capital is subject to the risk premium attached to it, it is impossible to ascertain the cost of capital accurately and maintain it at the same level for the entire period under study. Therefore, it is thought prudent to take the risk free rate of return as the influencing factor instead of the cost of capital.
If this study could establish that, there is a definite relationship between accounting returns (cash basis) and stock returns and that such a relationship could be used to establish future connection, then it would be worthwhile to convey the findings to the regulatory authorities to bring about changes in disclosure norms.
1.2 LITERATURE REVIEW
Eugene Fama (1991)4 in his paper discusses the various hypotheses on efficient markets and their anomalies. The paper also redefines the common definitions of efficient markets and investigates the joint- hypothesis problem, the costs of information and various pricing models.
In this paper the author investigates two problems of market efficiency, the first being information and transaction cost and the second, the joint hypothesis problem. In another paper (1999) 6 the same author states that stock prices fully reflect the most complete and best information available. However, Eugene Fama himself acknowledges that his reading of the market has been a stubborn obstacle for active investors determined to find ways to beat the market.
Darius Palia and Jacob Thomas (1997) 5 write that a common belief among practitioners is that unexpected changes in foreign exchange rates shall affect the market value of certain firms. Given this common belief, the inability to document a strong and systematic
contemporaneous relation between stock returns and exchange rate changes is puzzling.
Paul Krugman (1999) 7 argues that under efficient market hypothesis (EMH), at any given time asset prices fully reflect all available information. That seemingly straightforward proposition is one of the most controversial ideas in all social sciences research, and its implications continue to reverberate through investment practice. The chief corollary to the idea that markets are efficient, that prices fully .reflect all information, is that price movements do not follow any patterns
or trends. This means that past price movements cannot be used to predict future price movements. Rather, prices follow what is known as a 'random walk', an intrinsically unpredictable pattern.
Jing Liu and Jacob Thomas (1999) 8 have, in their paper, attempted to derive and test a relation between current period unexpected returns and unexpected earnings that incorporates revisions in forecasts of future earnings. Their motivation was to emphasize the misspecification in returns/earnings regressions that omits information currently available about future earnings, and to offer a solution.
Pitabas Mohanty (2001) 9 believes that there is now considerable evidence in the US that firm specific characteristics like size, price-to- book value, market risk premium can capture the common variation in stock returns. However, there is no consensus among researchers on
whether an investor can earn risk-adjusted excess returns by investing in small stocks.
Tuomo Vuolteenaho (2001)10 had used a Vector Autoregressive model (VAR) to deconstruct an individual firm's stock return into two components: changes in cash flow (expected cash flow news) and changes in discount rates (expected returns news). By definition', a firm's stock returns are driven by shocks to expected cash flows (cash-
0 flow news) and/or shocks to discount rates (expected-return news). He says that there is a substantial body of research measuring the relative importance of cash flow and expected return news for aggregate portfolio returns, but virtually no evidence is available on the relative importance of these components at the firm level.
Hossein Asgharian and Bjorn Hansson (2002) 11 have investigated the ability of factor-mimicking portfolios to explain expected returns in I multifactor asset pricing models. In particular, the usual manner of constructing factor-mimicking portfolios may result in estimated asset betas (coefficient of the predictor variables) that are quite different from the asset betas against the underlying factors, which may seriously affect the reliability of asset pricing models.
Pastor Lubos and Pietro Veronesi (2002)12 show that uncertainty about a firm's average profitability increases the firm's M/B ratio as well as its idiosyncratic return volatility. They suggest that this uncertainty is
especially large for the newly listed firms, but it declines over time due to learning. Their model therefore predicts that both the M/B and the return volatility of a typical young firm would decline as the firm ages.
Moreover, this effect is stronger for firms that pay no dividends, confirming another prediction of the model. The model is also endorsed by the observation that M/B declines faster for younger firms.
G. P. Samanta and Kaushik Bhattacharya (2002)13 in their paper have discussed the issue of whether the spread between Earning to Market Price (E/P) ratio and interest rate contains useful information about the movement of stock market. The results of their study reveal that though the spread seems to have reasonably strong causal influence on returns, the causal model helps in achieving slightly better forecasts than the random walk model. However, they are not clear as to whether the spread can be used as a profitable business strategy.
Andrew Ang and Jun Liu (2003)14 have developed a model to consistently value cash flows with changing risk-free rates, predictable risk premiums and conditional betas in the context of a conditional Capital Asset Pricing Model (CAPM). Practical valuation is accomplished with an analytic term structure of discount rates, with different discount rates applied to expected cash flows at different horizons.
John Y. Campbell and Motohiro Yogo (2003)15 in their paper argue that tests of the predictability of stock returns may be invalid when the predictor variable is persistent and its innovations are highly correlated with returns. They also suggest two methods to deal with the problem.
The first one is a pretest that determines predictability of stock, when the conventional t-test is misleading and the second, a new test of predictability that always leads to correct inference and is more efficient when compared to existing methods.
Francis A. Longstaff and Monika Piazzesi (2003)16 have attempted to quantify the risk premium attached to the standard asset-pricing theory.
They have emphasized that equilibrium asset values can be expressed as the expected product of a pricing kernel and the cash flows from those assets.
Burton G. Malkiel (2003)17 in his paper presents a defence of passive financial investment (indexing) strategies in all types of investment markets both nationally and internationally. He justifies the case of such strategies by relying on the theory of efficient market hypothesis and suggests that the information generally available about individual stock or about the market as a whole is reflected in market prices immediately.
Lakshmi Narasimhan S. and H. K. Pradhan (2003)18 find that the Indian stock market has witnessed drastic changes during the past decade under the broad stock market liberalization measures. In their study, the
authors have tested the validity of conditional CAPM for Indian stock market and found that the risk premium changes with changing economic conditions. The risk premium varies over time and it is negatively correlated with the index of industrial production. They also argue that the risk premium increases during a recessionary phase rather than during an expansionary phase.
Ajay Pandey (2003) 19 believes that modeling and forecasting the volatility of
•
capital markets are important areas of inquiry and research in financial economics with the recognition of time-varying volatility, volatility clustering and asymmetric response of volatility to market movements. This stream of research has been aided by various conditional volatility (Autoregressive Conditional Heteroskedasticity / Generalized Autoregressive Conditional Heteroskedasticity - ARCH/GARCH type) models proposed to handle these empirical regularities.Jeremy J. Siegel (2003) 20 defines a bubble as "a sharp rise in the price of an asset or a range of assets in a continuous process, with the initial rise generating expectations of further rises and attracting new buyers - this concerns speculators interested in profits from trading in the asset rather than its use or earnings capacity".
Eugene F. Fama and Kenneth R. French (2004) 21 argue that the capital asset pricing model (CAPM) is still widely used in applications, such as a
estimating the cost of capital for firms and evaluating the performance of
managed portfolios. The attraction of the CAPM is that it offers powerful and intuitively pleasing predictions about how to measure risk and the relation between expected return and risk.
Torben G. Andersen, Tim Bollerslev, Francis X. Diebold and Clara Vega (2005) 22 have discussed how markets arrive at prices. There is perhaps no question more central to economics. Their paper focuses on price formation in financial markets where the question looms especially large.
How, if at all, is news about macroeconomic fundamentals incorporated into stock prices, bond prices and foreign exchange rates?
Unfortunately, the process of price discovery in financial markets remains poorly understood.
John Y. Campbell and Samuel B. Thompson (2005) 23 wrote that towards the end of the last century, financial economists came to take the view that aggregate stock returns are predictable. During the 1980s, a number of papers studied valuation ratios such as the dividend-price ratio, earnings price ratio or smoothed earnings-price ratio. Around the same time, several papers pointed out that yields on short-term and long-term treasury and corporate bonds were correlated with subsequent stock returns.
Naiping Liu and Lu Zhang (2005) 24 state that recent studies have used the value spread to predict aggregate stock returns to construct cash- flow betas that appear to explain the size and value anomalies. Their
work shows that two related variables, the book-to-market spread (the book-to-market of value stocks minus that of growth stocks) and the market-to-book spread (the market-to-book of growth stocks minus that of value stocks) predict returns in different directions and exhibit opposite cyclical variations. More importantly, value spread mixes information on the book-to-market and market-to-book spreads and appears less useful in predicting returns.
to, Pandey I. M. (2005) 25 explores the significance of profitability and growth as drivers of shareholders wealth, measured by the market-to-book value (M/B). The author has studied the relationship between profitability (economic profitability) on the one hand and M/B ratio on the other. He has used panel data, employed Generalized Method of Moment (GMM) estimator and found that there is a strong positive relationship between profitability and M/B ratio. Growth on the other hand, is negatively related to M/B ratio.
Narasimhan Jegadeesh and Joshua Livnat (2006) 26 state in their paper that there are significant positive associations between earnings surprises and abnormal returns, around the preliminary earnings announcements as well as in the post-earnings announcement period.
Since earnings is a summary measure of material economic events that affect a firm in a given period, the intense focus on earnings surprises by investors and academics is natural.
Lewellen Jonathan, Stefan Nagel and Jay Shanken (2006) 27 argue that asset pricing tests are highly misleading in the sense that apparently strong explanatory power, in fact provides exceptionally weak support for a model. They offered a number of suggestions for improving empirical tests and evidenced that several proposed models do not work as satisfactorily as originally claimed.
Jacob K. Thomas and Huai Zhang (2006) 28 state that their study is motivated by the apparent gap between predictions regarding the determinants of market price to earning ratios (P/E ratio) and empirical evidence. While P/E ratio should be positively related to expected growth rate and negatively related to risk and the level of interest rates, prior evidence suggests weak relations at the portfolio level.
1.3 RESEARCH PROBLEM
The above studies illustrate that various attempts have been made to ascertain the value of stocks by identifying the unexpected earnings, dividend/price relationship, book value/market value relationship, discounted value of dividends, earning/market value relationship etc.
However, no attempts have been made to relate accounting returns (cash flows) to stock returns and use this relationship as a benchmark to predict the stock prices. This relationship could also be collated with the cost of capital as the latter has undergone a radical change vis-à-vis the integration of Indian economy with the global economy.
1.4 SIGNIFICANCE OF THE PROBLEM
If the research community identifies the relevant variables that influence the stock returns and communicates the same to the investing community in specific, and market participants in general, it will benefit them all in arriving at a fair market value of stocks. It will also enable the market participants to bring about transparency in market operations and help to build confidence in the investing community. This will lead to create stability in the market and make markets less volatile.
1.5 OBJECTIVES
1.5.1 To establish the relationship between accounting returns (cash basis) and risk free rate of return (as independent variables) with market returns (as dependent variable).
1.5.2 To determine the expected stock price based on the relationship established under 1.5.1.
1.6 HYPOTHESES
1.6.1 There is a significant relationship between the earnings and risk- free rate of return of the firm on the one hand and stock price on the other.
1.7 METHODOLOGY
For the purpose of this study, the following three different techniques have been used. A brief description of these techniques is given hereunder. At the same time, a detailed explanation for all the three techniques is given in Chapters 2, 3 and 4.
4IL 1.7.1 Multivariate Regression Model:
The Multivariate Regression Analysis (MRA) technique is an extension of simple regression analysis. The regression that measures the relationship between two variables becomes a multiple regression when it is extended to include more than one independent (predictor) variable such as X1, X2, X3, X4, etc, in trying to explain the dependent variable Y. In the case of simple regression analysis, the R 2 measures the strength of the relationship, but an additional R 2 statistic called the adjusted R2 is computed to counter the basis that will induce the R 2 to keep increasing as more independent variables are added to the regression. Like simple regression, multivariate regression is a powerful tool that allows the examination of the determinants of any response variable.
1.7.2 Probabilistic Growth Model:
This tool is newly developed and is used to forecast the price of stocks. In this model, it is assumed that stock price is a function of growth rate, subject to occurrence of such a growth rate. To capture non-linear behaviour of stocks, it is necessary to ascertain the lognormal growth rate instead of the simple growth rate. Therefore, the lognormal growth rate is derived for all the observations. Another important part of this model is that it lays emphasis on the probability of occurrence of such a growth rate, which is calculated by using the cumulative probability for standard normal distribution.
1.7.3 Artificial Neural Network Model:
The third tool is the Artificial Neural Network. An artificial neural network is an information-processing model that is inspired by the way human nervous systems process information. The key element of this model is the new structure of the information processing system. It comprises of a large number of interconnected processing elements (neurons) working in harmony to decipher a particular problem. An artificial neural network is configured for a specific application, such as pattern recognition or data classification, through a learning process. Neural networks, with their remarkable ability are able to derive meaning from complicated or imprecise data. These can be used to mine
patterns and discover trends that are too complex to be noticed by either humans or other computer software programs.
1.8 LIMITATIONS
The main limitation of the study is timely availability of data. These models cannot be used as a black box but should be used judiciously.
These are user-specific techniques; therefore, the user should have a thorough knowledge of the techniques used in this study.
1.9 CHAPTER SCHEME
The chapter scheme given below has been followed in presenting the details of the study conducted:
Chapter 1: This chapter covers introduction encompassing the preliminary background of the study, a literature review, the research problem and its significance, research objectives, hypotheses, methodology, limitations and the chapter scheme.
Chapter 2: In this chapter, the Multivariate Regression model along with sources of data, type of data used, sampling design, sample size, data analysis, results and interpretations are
discussed. Relevant references made in the chapter are stated at the end.
Chapter 3: This, chapter covers the explanation of the Probabilistic Growth model, including sources of data and type of data used, sampling design, sample size, data analysis and finally results and interpretations. References made in the chapter are stated at the end.
Chapter 4: This part discusses the various aspects of the Artificial Neural Network model, including sources of data, type of data used, sampling design, sample size, data analysis and results and interpretations. References made during the discussion are given at the end of the chapter.
Chapter 5: Finally, in this chapter, all the observations made during the entire study are summarized, conclusions are drawn, recommendations are made and scope for further research is suggested.
1.10 REFERENCES
1 Chancellor, E. (1999), "Devil Take the Hindmost: A history of financial speculation", Penguin Books.
2. Stringham, E. Boettke P. and Clark J. R. (2008), "Are regulations the answer for emerging stock markets", Evidence from the Czech Republic and Poland" Quarterly Review of Economics & Finance, Elsevier, Vol. 48, pp. 541 - 566.
3. Wilde, 0. (1900), "Nothing ... except my genius" Alastair Rolfe (Compiler), Stephen Fry (Introduction) 1997, Penguin Books.
4. Fama, E. (1991), "Efficient capital market II", Journal of Finance, Vol.
46, pp. 1575 - 1617.
5. Palia, D. and Thomas J. (1997), "Exchange rate exposure and firm valuation: New evidence for market efficiency", Harvard (Financial decisions and control workshop) and Stanford (Accounting Summer Camp), USA, http://www.som.yale.edu/Faculty/
jkt7/papers/fx.pdf.
6. Fama, E. (1999), "Think you can beat the market? Eugene Fama still says you can't", Capital Ideas, Vol. 2, pp. 57 - 58.
7 Krugman Paul (1999), "Market Efficiency", http://www.deanlebaron . com/book/ultimate/chapters/mkt eff.html, www.web.mit.edu/krugman/,
8. Liu, J. and Thomas J. (1999), "Stock returns and accounting
• earnings", The Journal of Accounting Research, Vol. 38, pp. 71 - 101.
9. Mohanty, P. (2001), "Efficiency of the market for small stocks"
presented for NSE Research Initiative, Research Paper No. 1, Mumbai, http://www.nseindia.com/content/research/res_papers.htm.
10. Vuolteenaho, T. (2001), "What drives firm level stock returns", Working Paper No. W-8240, NBER, MA, USA.
4
11. Hossein, A. and Bjorn H. (2002), "'A critical investigation of the explanatory role of factor mimicking portfolios in multifactor asset pricing models", EFA Berlin Meetings Discussion Papers, SSRN:
http://ssm.com/abstract=302338.
12. Lubos, P. and Veronesi P. (2002), "Stock valuation and learning about profitability", The Journal of Finance, Vol. 58, pp. 1749 - 1789.
13. Samanta, G. P. and Bhattacharya, K. (2002), "Is the spread between E/P ratio and interest rate informative for future movement of Indian stock market"? NSE Research Initiative, Research Paper No. 7, NSE, Mumbai, India.
14. Ang, A. and Liu J. (2003), "How to discount cash flows with time- varying expected returns", Working Paper No. 10042, NBER, MA, USA.
• 15. Campbell, J. Y. and Motohiro Y. (2003), "Efficient tests of stock return
16. Longstaff, F. A. and Piazzesi M., (2003), "Corporate earning and the equity premium" Working Paper No. 10054, NBER, MA, USA, SSRN: http://ssrn.com/abstract=461375.
17. Malkiel, B. G. (2003), "Passive investment strategies and efficient markets", European Financial Management, Vol. 9, pp. 1 - 10.
18. Narasimhan, L. S. and Pradhan H. K. (2003), "Conditional CAPM and cross sectional returns - A study of Indian securities market", presented for NSE Research Initiative, Research Paper No. 23, Mumbai, http://www. n se i nd i a .com/contentJresearch/res_papers. htm.
19. Pandey, A. (2003), "Modeling and forecasting volatility in Indian capital markets", Working Paper No. 2003-08-03, IIMA, Research and Publication Department, Ahmedabad, India.
20. Siegel, J. J. (2003), "What is an asset price bubble", European Financial Management, Vol. 9, pp. 11 - 24.
21. Fama, E. F. and French K. R. (2004), "The capital asset pricing model: Theory and evidence", Journal of Economic Perspectives, Vol. 18, pp. 25 - 46.
22. Andersen, T. G., Bollerslev T., Diebold F. X. and Vega C. (2005),
"Real-time price discovery in stock, bond and foreign exchange markets", Working Paper 11312, NBER, MA, USA, SSRN: http://
ssrn.com/abstract=560642.
wt
23. Campbell, J. Y. and Thompson Samuel B. (2005), "Predicting the equity premium out of sample: Can anything beat the historical average", Working Paper No.11468, NBER, MA, USA, SSRN:
http://ssrn.com/abstract=755704.
24. Liu, N. and Zhang L. (2005), "The value spread as a predictor of returns", Working Paper 11326, NBER, MA, USA, SSRN: http://
ssrn.com/abstract=532703.
25. Pandey, I. M. (2005), "What drives the shareholders value", Working Paper Series, Working Paper No. 2005-09-04, Indian Institute of Management, Ahmedabad, India.
26. Jegadeesh, N. and Livnat J. (2006), "Revenue surprises and stock returns", Journal of Accounting and Economics, Vol. 41, pp. 147 - 171.
27. Jonathan, L. Nagel S. and Shanken J., (2006) "A skeptical appraisal of asset-pricing tests", Working Paper No. 12360, NBER, MA, USA.
28. Thomas, J. K. and Zhang H. (2006), "Another look at P/E ratios", Annual Winter Accounting Workshop, University of Southern California, California, USA.
I ft \L )-.1j tA752810K
Chapter Plan
• Introduction
• Sources of Data
• Sampling Design
• Data Analysis
• Results and Interpretations
• References
HA TER la
2. MULTIVARIATE REGRESSION MODEL
2.1 INTRODUCTION
It is worthwhile to look back at the history and assumptions of multivariate regression analysis before we embark on its use. The history of regression goes back to 18 th Century. The earliest form of regression was the method of least squares, which was published by Adrien Marie Legendre' in 1805 and by Carl Friedrich Gauss 2 in 1809.
The 'least squares' is Legendre's term. However, Gauss claimed that he had known the method since 1795. Legendre and Gauss both applied the method to the problem of determining the orbits of bodies around the Sun. Gauss3 published a further development of the theory of least squares in 1821, including a version of the Gauss-Markov theorem.
The term 'regression' was coined by Francis Galton 4 , a cousin of Charles Darwin 5 , in the nineteenth century to describe a biological phenomenon.
The phenomenon was that the heights of descendants of tall ancestors tend to regress down towards a normal average. For- Galton, regression had only this biological meaning, but his work was later extended by
*
Udny Yule and Karl Pearson et al b to a more general statistical context.
Now, the term 'regression' is often synonymous with 'least squares curve fitting'.
Classical assumptions stated by George Carrie' for regression analysis indude:
a. The sample must be a representative of the population for the inference prediction.
b. The error is assumed as a random variable with a mean of zero, conditional on the explanatory variables.
c. The independent variables (predictors) are error-free. If this is not so, modeling may be done using errors in variables model.
d. The predictors must be linearly independent, i.e. it must not be possible to express any predictor as a linear combination of the others.
e. The errors are uncorrelated, that is, the variance covariance matrix of the errors is a diagonal matrix and each non-zero element is the variance of the error.
f. The variance of the error is constant across observations (homoscedasticity). If not, weighted least squares or other methods might be used.
Univariate analysis consists in describing and explaining the variation in a single variable. Bivariate analysis does the same for two variables taken together (co-variation). According to George H. Dunteman 8
multivariate analysis (MVA) considers the simultaneous effects of many variables taken together. A crucial role is played by the multivariate normal distribution, which allows simplifying assumptions to be made
0
(such as the fact that the interrelations of many variables can be reduced to information on the correlations between each pair), which makes it feasible to develop appropriate models. Multivariate Analysis models are often expressed in algebraic form (as a set of linear equations specifying the way in which the variables combine with each other to affect the dependent variable) and can be thought of geometrically. Thus, the familiar bivariate scatter-plot, according to C. Huang at e1 9 , is in the two dimensions, representing two variables, which can be extended to higher-dimensional spaces, and Multivariate Analysis can be thought of discovering how the points cluster together.
The most familiar and often-used variants of MVA include extensions of regression analysis and analysis of variance, to multiple regression and multivariate analysis of variance respectively, both examine the linear effect of a number of independent variables on a single dependent variable.
A common use of multivariate analysis is to reduce a large number of inter-correlated variables into a much smaller number of variables, preserving as much as possible of the original variation, whilst also having useful statistical properties such as independence. In the case of regression analysis, the R2 measures the strength of the relationship, but an additional R2 statistic called the adjusted R2 is computed to counter the basis that will induce the R 2 to keep increasing as more independent variables are added to the regression. Like regression, multivariate regression is a powerful tool that allows examining the determinants of any variable.
In this part of the study, Multivariate Regression Analysis is used to forecat the price of stocks. While Stock Price is taken as dependent variable (explained variable), Cash Flow and Risk Free Rate of Return are taken as independent variables (explanatory variables). By applying multivariate regression analysis, the estimated values for all the observations are ascertained. Subsequently the quarterly logarithmic growth rates of the estimated values are ascertained for the entire period for every company included in the sample. In order to capture non- linearity of stocks, it is essential to use the logarithmic growth rate instead of simple growth rate. From these quarterly logarithmic growth rates, the geometric mean growth rate is ascertained for the entire period, and that is used for forecasting the stock prices.
The frequency of data for which growth rate ascertained is important. It may be for a day, week, month or quarter. It is also assumed that such growth is valid for future period. Accordingly, weight for the period for which forecasting is planned for, should be incorporated while using this model. Using such estimated growth rate, period of forecast and spot price, forecast price are estimated. The analysis is carried out to verify the validity of the hypothesis set out for study. Software package such as Statistical Package for Social Sciences (SPSS) and Data Analysis Tool of Microsoft excel is used. The models used in the analysis are described below as stated by G. S. Maddala l° and William Greene":
4
•
LINEAR MODEL
Yi = Poi +
13i 04 +
P2 (X2)1 (2.1)LOG LINEAR MODEL
log Y1 = Poi + Pi log (X1)1 + P2 log (X2)1 (2.2) EXPONENTIAL MODEL
In Yl = Poi + 131 In (X1)1+ P2 In ((2)1 (2.3)
Where,
= 1, 2, 3, n, Number of observations,
Explained (dependent) Variable (Market Price), X1 = Explanatory (independent) variable (Cash flow),
X2 = Explanatory (independent) variable (Risk-Free Rate of Interest),
log = Ordinary Log, In = Natural Log, f3 Intercept,
Partial Slope Co-efficient of the first explanatory variable, /32 = Partial Slope Co-efficient of second explanatory variable, u, = Stochastic Disturbances.
2.2 SOURCES OF DATA
Data required for the purpose of study are collected from various sources. Market prices are picked up from published data of BSE12 (Bombay Stock Exchange) - Mumbai, Cash Flows are taken from Electronic Data Information Filing and Retrieval (EDIFAR)13 system (National Informatics Centre) and Risk Free Rate of Return is culled from Reserve Bank of India's (RBI)14 website.
Monthly closing price of BSE was used in this study, assuming that it fairly represents the average stock price, which is adjusted for bonus shares, share splits, right shares, Employees' Stock Option (ESOPs)etc.
and then converted it into average quarterly price. Quarterly Operating Cash Flows are culled from Cash Flow Statement of the Companies.
Risk free rate of return is collected from the data published by the RBI.
These data are collected for all the companies included in the 'A - Category' stocks of BSE Mumbai.
2.3 SAMPLING DESIGN
Before the actual study, a pilot study was carried out on 10 randomly selected 'A-Category' stocks. As the results were extremely satisfactory, it was decided to continue with 'A - Category' stocks for the purpose of this study and the sample was selected by stratified random sampling procedure.
All the companies classified as 'A - Category' stocks by BSE Mumbai were identified. BSE Mumbai classifies a company as 'A - Category' if its turnover is more than Rs. 500 Crores per annum. Thus, the sample encompasses only large companies and actively traded stocks. In general, these companies form a base for market movements and for formulating Sensex and similar BSE and NSE (National Stock Exchange) Indices. These companies broadly represent most of the industries in
0 Indian economy.
There are 206 listed companies classified as 'A - Category' stocks as on 31 st March 2007 (Appendix i). Out of these, 187 companies are traded on Bombay Stock Exchange for a period 3 years and more (Appendix ii).
All the remaining companies are traded on Bombay Stock Exchange for a period less than 3 years; therefore, such companies are excluded from the sample.
As per S & P CNX (Standard & Poor's CRISIL NSE Index) Classification, there are 76 types of industries (Appendix iii). However, all the 206 'A - Category' companies of BSE Mumbai cover only 54 types of industries (Appendix iv). Nevertheless, these 54 types of industries envelop majority of industries included in the S & P CNX Classification. Since the sample covers 187 companies out of 206 companies, these 187 companies represent 53 types of industries (Appendix v), broadly encompassing entire 'A - Category' stocks.
However, adequate data are available for 173 companies only (Appendix vi) out of 187 companies selected as sample, and therefore analysis is carried out on these companies only. In the case of remaining 14 companies, adequate data were not available; therefore, analysis could not be carried out. These 173 companies cover 51 types of Industries (Appendix vii).
Companies listed with the stock exchanges started to file quarterly returns from the year 2001 onwards. Therefore, quarterly data are available for a maximum period of 7 years. Since several companies are listed with the stock exchanges from the year 2003 onwards, the data for these companies are available for a period of 5 years and less.
Since quarterly cash flow data can be culled from quarterly data of listed companies, it is thought prudent to use quarterly data instead of yearly data. At the same time, quarterly stock prices could be determined from the collected data. Similarly, the quarterly risk free rate of return (Treasury bill return) is derived from the collected data.
Figure 2.1 depicts number of firms selected as population (A - Category), eligible (> 3 years) and sample firms. Figure 2.2 depicts number of industries covered under population (A - Category), eligible and sample firms.
206
A - CATEGORY ELIGIBLE
1 2% I
173
SAMPLE
FIRMS
Figure 2.1 'A - Category", Eligible and Sample Firms
INDUSTRIES
53
51
A - CATEGORY ELIGIBLE SAMPLE
Figure 2.2 'A - Category', Eligible and Sample Industries
Figure 2.3 and figure 2.4 depict the number of firms in each industry included in the 'A - Category' firms, number of firms in each industry included in eligible firms and number of firms in each industry included in the sample. The numbers given in the figure indicates the number of firms included in the sample, eligible and 'A - Category' firms for each industry.
0 SAMPLE
■ ELIGIBLE 0 POPULATION _ - -
Aluminium Auto Ancillaries Automobiles - 2/3 Automobiles - 4 Banking Bearing Castings Cement Chemicals - Chemicals - Chemicals - Cigarettes Compressors Computer - H Computer - S Construction Consumer Durables Detergents Diversified Electrical Equip
Engineering Engines Fertilizers Finance Finance - Housing Food Processing Gas
4, 4, 4 5, 5, 5
20, 21, 24
5, 5, 5
J 14,16,16
2,2, 2 7, 7, 7 4, 4, 4 1, 2, 2 3, 4, 5 1, 1, 1 2, 2, 2 3, 3, 3 2, 2, 2
NO. OF FIRMS
❑ SAMPLE ELIGIBLE
❑ POPULATION
Health Care Hotels Infrastructure Media Metals Mining Miscellaneous Oil Exploration Packaging Paints Paper Products Personal Care Pesticides Petrochemicals Pharmaceuticals Power Refineries Shipping Steel Tea & Coffee Telecommunication
Textile Products Textiles - Synthetic Trading Travel & Transport Tyres Watch
3, 3,
2, 2,
0, 0,
2, 2,
1, 1,
2, 2,
1, 1,
2, 2,
2, 2,
4, 5,
0, 2, 4, 4, 3
2
1
3
1
3
1
2
2
5
2
5 20, 21, 22
6, 6, 9
7, 8, 8
2, 2, 2
8, 9, 9 1, 1, 1
5, 3, 2
2, 2, 2
4, 4,4
1, 1, 1
1, 1, 2
1, 1, 1
1, 1, 1
NO. OF FIRMS
Transport,
_T yres, 1 1
Aluminium, 2
Telecommunication, 2
Trading, 1 Textiles -
S, 4 Tea & Coffee, I
Tex iles, 2
Shipping, 2 Refineries, 7 Power, 6
Pharmaceuticals, 20
Petrochemicals, Personal
Care, 4 Paper, 2 Packa Oil Explor ion, 1 Paints, 2 Metal Miscellaneous, 2
Finance - H, Health Care, 3 Media, 2
Finance, 1
Watches, 1
Auto Ancillaries, 3
Automobiles, 3 Bearing, 1
Automob' es, 6 Castings, 1
Cement, 5 Chemicals - I,1 Chemicals- 0, 1 Chemicals S, 1 Cigarettes, 1 Compressors, 1
Computer - H, 5 Construction, 1
Consumer Durables, 1 Detergents, 1 Diversified, 7
Engines, 2 Electrical Equ, 4 Engineering , 1
Fertilizers, 3 Steel, 8
otels, 2 Foo Process
Figure 2.5 depicts the industry wise classification of firms included in the sample for this study (173 Companies).
Figure 2.5 Industry Names and Number of Firms (Sample)
