1
Effect of market anomalies on expected returns on the JSE: A cross-sector analysis
by
Mpho Innocentia Mahlophe
(22477055)
Dissertation submitted in partial fulfilment of the requirements for the degree
MASTER OF COMMERCE (RISK MANAGEMENT) in the
THE SCHOOL OF ECONOMIC SCIENCE
in the
FACULTY OF ECONOMIC SCIENCES AND INFORMATION TECHNOLOGY
at the
NORTH-WEST UNIVERSITY
VAAL TRIANGLE CAMPUS
Supervisor: Dr PF Muzindutsi Co-Supervisor: Mr W Peyper
ii
ACKNOWLEDGEMENTS
I am thankful to God who has assisted me with the strength to carry out this study, without him the completion of this study would not have been possible.
I thank my parents who have always been supportive of my studies. They have both shown the utmost understanding and love throughout this whole process and without their support, encouragement and prayers this would have not been possible. Jeanette and Isaac Mahlophe are the best parents I could have ever asked for. I thank them for being my pillars of strength throughout the year.
To my supervisors, Dr. P.F Muzindutsi and Mr. W.H Peyer, who have been mentors to me throughout the whole process, I thank you for the guidance, support and encouragement throughout the year. Thanks to my cousin, Bertha Moeletsi, who lent an ear during my stressful periods and for the emotional support.
Finally, I would like to thank the North-West University for aiding me with the finance, space and resources to complete my study successfully.
iii
DEDICATION
To my loving parents, Jeanette and Isaac Mahlophe, I love you and I’m truly thankful to have parents like you.
iv
DECLARATION
I declare that this dissertation titled:
Effect of market anomalies on expected returns on the JSE: A cross-sector analysis
is my own work and all the resources used or quoted have been duly acknowledged by means of in-text citations and complete references, and that I have not previously submitted the dissertation for degree purposes at another university.
___________________________ Mpho Innocentia Mahlophe
v
ABSTRACT
The efficient market hypothesis and behavioural finance have been the cause of much debate for decades, with one theory advocating market efficiency and the other opposing it. The efficient market hypothesis (EMH) assumes that investors always act rationally and stock prices adjust rapidly to new information and should reflect all available information. In contrast, behavioural finance suggests that markets are not rational and investors make irrational decisions, which may lead them to over- or under-price stocks. Researchers for years have been empirically testing these assumptions in stock markets. However, there has been no consensus on which asset-pricing models perform better in capturing the effect of market anomalies and what impact these market anomalies have on the expected returns of different stock market’s sectors.
The aim of the study was to test the effect of selected market anomalies on expected return in different sectors of the Johannesburg Stock Exchange (JSE). More specifically, the study aimed to compare the performance of different asset-pricing models and their ability to account for market anomalies in different sectors of the JSE. Additionally, this study tested the applicability of the recent Fama and French five (FF5-factor) model, in estimating the expected return on the JSE.
The study used a quantitative approach with secondary data over a period of 12 years starting from January 2002 to December 2014. The sample used in the study consists of monthly data obtained from McGregor BFA and the South African Reserve Bank. The study examined for the effects of size, value, January and momentum variables across six sectors of the JSE. This was accomplished by the use of various asset-pricing models such as the Capital asset pricing model (CAPM), the Fama and French three-factor model (FF3-factor), the Carhart four-factor model (C4F) and the recent five-factor model of Fama and French (FF5-factor).
The study showed that whenever the asset-pricing models were not restricted, they tend to capture the market anomalies in four out of the six sectors examined. However, no market anomalies were found present in two of the six sectors analysed. In contrast, when the asset-pricing models are restricted, the asset-asset-pricing models only seem to capture the effects of market anomalies in one of the six examined sectors. The findings in this study suggest that market anomalies are sensitive to model specifications, as restricting the models tends to capture the different market anomalies across the sectors of the JSE. The study also found
vi
that market anomalies differ across sectors and that some sectors are more efficient than others.
The study also reveals that the FF5-factor model is able to account for expected returns on the JSE. In addition, the FF5-factor model tends to perform better when the model is restricted. It is also evident from the findings presented in this study, that the value anomaly loses its predictive power when profitability and investment variables are included in the model. Overall, the study illustrated that market anomalies have an effect on returns of the JSE, that the model specifications play an important role in an asset-pricing model and that the FF5-factor model is applicable on the JSE, however, it is not certain whether four or five FF5-factors apply to the South African market.
Key words: efficient market hypothesis, Johannesburg Stock Exchange, market anomalies,
vii TABLE OF CONTENTS ACKNOWLEDGEMENTS ... ii DEDICATION ... iii DECLARATION ... iv ABSTRACT ... v LIST OF TABLES ... x
LIST OF ABBREVIATIONS ... xii
CHAPTER ONE: INTRODUCTION, BACKGROUND AND PROBLEM STATEMENT ... 1
1.1 INTRODUCTION ... 1 1.2 PROBLEM STATEMENT ... 4 1.3 OBJECTIVES ... 5 1.3.1 Primary objective ... 5 13.2 Empirical objectives ... 5 1.4 IMPORTANCE OF STUDY ... 5
1.5 RESEARCH DESIGN AND METHODOLOGY... 6
1.5.1 Literature review ... 6
1.5.2 Empirical study ... 6
1.6 CHAPTER CLASSIFICATION ... 9
1.6.1 Chapter 1: Introduction, problem statement and background of the study ... 9
1.6.2 Chapter 2: Literature review ... 9
1.6.3 Chapter 3: Research design and methodology ... 10
1.6.4 Chapter 4: Empirical findings and discussion ... 10
1.6.5 Chapter 5: Conclusions and recommendations ... 10
CHAPTER TWO: LITERATURE REVIEW ... 11
2.1 INTRODUCTION ... 11
viii
2.2.1 Forms of market efficiency ... 13
2.2.2 Implications of EMH ... 14
2.3 BEHAVIOURAL FINANCE ... 17
2.3.1 Theories of behavioural finance ... 18
2.3.2 Adaptive market hypothesis ... 21
2.4 MARKET ANOMALIES ... 23
2.4.1 Types of market anomalies ... 23
2.5 EMPIRICAL STUDIES ON THE PRESENCE OF THE ANOMALIES IN STOCK MARKET ... 29
2.5.1 Studies conducted in developed stock markets ... 29
2.5.2 Studies conducted in emerging stock market ... 34
2.5.3 Studies conducted on the JSE ... 37
2.6 SUMMARY AND CONCLUDING REMARKS ... 41
CHAPTER THREE: DATA AND METHODOLOGY ... 44
3.1. INTRODUCTION ... 44
3.2 SAMPLE SELECTION AND DATA COLLECTION AND ANALYSIS ... 44
3.3 MODEL SPECIFICATION ... 46
3.3.1 Capital asset pricing model ... 47
3.3.2 The Fama and French three-factor model ... 49
3.3.3 The Carthart four-factor model ... 51
3.3.4 Carhart four-factor model augmented with the January effect ... 54
3.3.5 The five-factor model ... 54
3.4 REGRESSION STATISTICS ... 56
3.5 DESCRIPTION OF THE VARIABLES IN THE MODELS ... 57
3.5.1 Calculation of size factor ... 58
ix
3.5.3. Calculation of momentum factor ... 59
3.5.4 Calculation of the profitability factor ... 60
3.5.5 Calculation of investment factor ... 60
3.5.6 Calendar anomalies: The January effect ... 60
3.6 SUMMARY OF THE CHAPTER ... 61
CHAPTER FOUR: RESULTS AND DISCUSSION ... 62
4.1 INTRODUCTION ... 62
4.2 CROSS-SECTOR ANALYSIS RESULTS ... 63
4.2.1 Descriptive statistics ... 63
4.2.2 Analysis of market anomalies in different sectors ... 68
4.3 DISCUSSION OF THE FINDINGS OF THE CROSS-SECTOR ANALYSIS ... 91
4.4 APPLICATION OF FAMA AND FRENCH FIVE-FACTOR MODEL ... 97
4.4.1 Descriptive statistics of variable used in FF 5-factor model ... 98
4.4.2 Results of the Fama and French five-factor models on the JSE ... 99
4.4.3 Discussion of the Fama and French five-factor model results ... 102
4.5 SUMMARY ... 102
CHAPTER FIVE: SUMMARY AND CONCLUSIONS ... 104
5.1 INTRODUCTION ... 104
5.2 SUMMARY OF THE STUDY ... 104
5.2.1 Summary of literature review ... 104
5.2.2 Summary of methodology and main findings ... 105
5.3 CONCLUSIONS AND RECOMMENDATIONS ... 107
5.4 LIMITATIONS OF THE STUDY AND AREAS FOR FURTHER RESEARCH 108 REFERENCES ... 109
x
LIST OF TABLES
Table 2.1: Summary of the three forms of market efficiency and its implications ... 14
Table 2.2: Summary of AMH ... 22
Table 3.1: Summary of the variables ... 58
Table 4.1: Descriptive statistics of all sectors (January 2002 to December 2014) ... 65
Table 4.2: Cross-sector results of the Standard CAPM ... 69
Table 4.3: Cross-sector analysis of the Fama and French three-factor model ... 72
Table 4.4: Cross-sector analysis of the Fama and French three-factor model with CAPM assumption ... 74
Table 4.5: Cross-sector analysis of the Carhart four-factor model (6-month momentum variable) ... 75
Table 4.6: Cross-sector analysis of the Carhart four-factor model with the CAPM assumption (6-month momentum variable) ... 78
Table 4.7: Cross-sector analysis of the Carhart four-factor model with 12-month momentum variable ... 79
Table 4.8: Cross-sector analysis of the restricted Carhart four-factor model (12-month momentum variable) ... 81
Table 4.9: Cross-sector analysis of the Carhart four-factor augmented with January effect (6-month momentum) ... 83
Table 4.10: Cross-sector analysis of the Carhart four-factor model augmented with January effect with CAPM assumption (6-month momentum) ... 86
Table 4.11: Cross-sector analysis of the Carhart four-factor model augmented with January effect (12-Month momentum variable) ... 89
Table 4.12: Cross-sector analysis of the Carhart four-factor model augmented with January effect with CAPM assumption (12 Month momentum variable) ... 91
Table 4.13: Summary of market anomalies present in various sectors... 97
xi
Table 4.15: Application of the Fama and French five-factor model and Wald coefficient restriction test ... 100 Table 4.16: Cross-sector analysis of the Fama and French five-factor model with CAPM assumption and Wald coefficient restriction test ... 101
xii
LIST OF ABBREVIATIONS
AMEX: American Stock Exchange APT: Arbitrage pricing theory C4F: Carhart four-factor model
C4FJ: Carhart four-factor model augmented with the January effect CAPM: Capital asset pricing model
EMH: Efficient market hypothesis
FF3-factor: Fama and French three-factor model FF5-factor: Fama and French five-factor model GLS Generalised least squares
HML High minus low
JSE: Johannesburg Stock Exchange MOM12: Previous 12 month’s returns MOM6: Previous 6 month’s returns
NASDAQ: National association of securities dealers NYSE: New York Stock Exchange
OLS: Ordinary least squares SMB: Small minus big VMG: Value minus growth
xiii
1
CHAPTER ONE: INTRODUCTION, BACKGROUND AND PROBLEM STATEMENT
1.1 INTRODUCTION
Investors are concerned with stock returns and how they fluctuate because of market risk. Market risk can be defined as changes in financial market prices and rates, which reduce the value of a security or a portfolio (Crouhy et al., 2014:25). Investment theory states that there is a positive relationship between risk and return; this is supported by the modern portfolio theory, which suggests that market risk should be the only risk that increases expected return, thus a positive relationship (Elton & Gruber, 1997). However, the modern portfolio theory fails to account for other factors that might have an effect on expected returns other than the market risk. There are two main theories in this regard that are on opposite ends. On the one hand, one is the agreement with market risk being the sole risk factor and this theory is known as the efficient market hypothesis (EMH). On the other hand, a theory known as behavioural finance, believes that other risk factors, along with market risk, account for expected returns. This creates two controversial views in the investigation of the expected return (Malkiel, 2003).
The first view, the efficient market hypothesis, is defined by Fama (1965:3-4) as a market that has large numbers of buyers and sellers who value and analyse securities with the aim of making profit. By doing so, they are able to forecast future market prices for individual stocks, since all the relevant information is freely available to all investors and is taken into account by the market. According to the above definition, in an efficient market, investors should not make above-average returns, as all relevant information about the security is freely available to all investors. Furthermore, prices of securities are supposed to adjust rapidly to new information and all current stock prices should reflect all available information (Reilly & Brown, 2012:140-141). Hence, expected return is exposed to market risk only, according to this EMH approach.
However, over the years it has been discovered that stock markets contradict the assumptions of the EMH. This led to the development of a theory called behavioural finance, which is mostly known to be on opposite ends with the EMH. Behavioural finance refers to a situation where a stock or a group of stocks’ performance deviates from the assumptions of the EMH because of different forces. This situation is best known as market anomalies (Silver, 2011);
2
defined as a distortion of return on a financial market that contradicts the EMH (Latif et al., 2011:1). Market anomalies have been studied for years as researchers look for answers as to why these anomalies affect expected returns. Furthermore, investors seek answers as to whether they should be concerned about the effect these market anomalies have on expected returns.
Behavioural finance suggests that market anomalies are the result of cognitive limitations, which are responsible for investors making irrational investment decisions, therefore, resulting to inefficiencies in the market caused by prices not being able to deviate from their fair value for long periods (Yalcin, 2010:35). The above statement implies that human behaviour affects investment decisions of investors regarding stocks. These decisions are driven by human behaviour, which causes stock prices to fluctuate, consequently affecting the expected returns of such stocks.
In both EMH and behavioural finance, different models estimate stock returns. In EMH, the Capital asset pricing model (CAPM) of Sharpe (1964), Linter (1965) and Black (1972) has made many waves in the study of finance since its development. The main theme of the model focuses on the mean-variance efficient market portfolio, which encompasses that the market portfolio be constructed in a way that expected returns are maximised for a given level of risk (Fama & French, 1996:74). The CAPM has been favoured for years and many professionals still prefer it today. The model assumes that all investors have similar expectations with regards to risk, that investors have similar net returns and that markets are efficient (Reilly & Brown, 2012:196). An efficient market is one where stock prices rapidly adjust to the arrival of new information, thus current prices reflect all available information about the stocks (Reilly & Brown, 2012:139).
Furthermore, CAPM is used to determine the required rate of return for a particular asset, given the current level of risk and assumes that the only driver of risk is systematic risk, also known as market risk (Reilly & Brown, 2012:196). The CAPM makes use of one risk factor known as market beta and views systematic risk as a key determinant of how a security should be priced in the market, given the level of risk associated with it (Cagnetti, 2002:2). However, there are inconsistencies when it comes to the empirical work of the CAPM, where some of the limitations of the model come into play (Fama & French, 2004). Market beta, as the sole risk factor, is rendered inadequate, as it does not take into account other factors such as firm size, the price-to-earnings ratio and book value/market value, which may influence
3
the price of the asset (Bhatnagar & Ramlogan, 2012:4). All these factors are known as market anomalies. Consequently, since the CAPM was unable to capture these market anomalies, models like the arbitrage pricing theory of Ross (1976) and the three-factor model of Fama and French (1993) were developed to account for the anomalies when estimating expected return. These models that capture the presence of market anomalies are in favour of the behavioural finance view point, that stock returns are affected by factors (anomalies) other than the market risk captured by the CAPM.
Researchers like Fama and French have spent years analysing stock markets and their relationship with expected returns and how they are affected by various factors. Fama and French tested various factors using empirical tools such as the Fama and Macbeth regression (1973), where they use two stages to determine how an asset is priced. Fama and French also made use of the CAPM, which they used as a basis for the development of the Fama and French three-factor model (1993), which includes the size and value anomalies. All these are tools used to aid researchers and financial institutions in estimating risk and return and to help investors make the right investment decision.
Size, leverage, liquidity, momentum, value and January effect are some market anomalies, which go against the assumptions of the efficient market hypothesis (Frankfurter & McGoun, 2002). These market anomalies are able to account for expected return beside market beta and are discussed below. The most studied market anomaly is the size effect, and according to Banz (1981:3-4), the size effect has been in existence for more than four decades. This anomaly is biased towards the size of a firm, where small firms tend to be undervalued, while large firms may be overvalued. This means that the relationship between expected return and size is not proportional to the market value for small firms. Furthermore, since small firms tend to be undervalued they are more risky, thus they provided higher expected returns. In contrast, large firms are overvalued and less risky hence they provided expected returns lower than their given level of risk (Banz, 1981:16).
Given the amount of market anomalies discovered over the years, a number of models have been developed to capture these anomalies. There have been studies that focused on the weaknesses of the EMH (Ross, 1976; Bhandari, 1988; Fama & French, 1992, 1993 & 1996; Jegadeesh & Titman, 1993; Gustafsson & Lundqvist, 2010; Bhatnagar & Ramlogan, 2012). Others exploited the market anomalies left out in the CAPM (Banz, 1981; Fama and French, 1992 & 1993; Van Rensburg, 2001; Acharya & Pedersen, 2005; Mazal, 2009; Auret & Cline,
4
2011; Strugnell et al., 2011; Muller & Ward, 2013). Most of these studies demonstrate that different methods are used for estimation of the expected stock return and hence produced conflicting results, implying that there is no consensus on the effect of anomalies on stock returns. Therefore, conducting a further study on the methods of estimating expected return and the effects market anomalies may shed more light.
1.2 PROBLEM STATEMENT
There has been no consensus on which model captures the effect of market anomalies and what impact these market anomalies have on the expected returns. This is because market anomalies change with the economic climate, stock markets, selected sample, time periods and differ from sector to sector. Furthermore, it has been argued that after market anomalies have been analysed and documented in academic literature, they often disappear, reverse or weaken (Latif et al., 2011:1). This may occur because investors tend to take advantage of these market anomalies and in turn, they lose the effect they have on expected returns. As a result, these regular patterns in returns have been established, hence they lose their predictive power. Thus, there is no definitive conclusion on the causes of market anomalies and their effect on expected returns.
However, behavioural finance states that market anomalies are inconsistencies of asset-pricing theory and are suggestive of market inefficiency or the insufficiencies of asset-asset-pricing models. These inefficiencies are supposed to be captured by the models, leaving no opportunity for investors to be in a position to outperform the market (Schwert, 2003:939). Previous studies (Acharya & Pedersen, 2005; Archana et al., 2014; Banz, 1981; Bhandari, 1988; Borges, 2009; Fama & French, 1992&1993; Muller & Ward, 2013; Strugnell et al., 2011:14; Van Rensburg, 2001) found that anomalies such as calendar effects, size, value, momentum and January effect have a significant effect on stock returns. This suggests that the EMH does not hold, as the market return is affected by different market anomalies. Overall, there is extensive research on market anomalies worldwide but there is no clear approach as to which method is the most appropriate for testing these market anomalies and their effect on expected stock returns, especially when different sectors of the stock market are considered. Thus, a further study on the effect of market anomalies on the expected return across all the JSE sectors and, which asset-pricing models better capture these effects, will shed more light on this topic.
5
1.3 OBJECTIVES
The following objectives have been formulated for the study:
1.3.1 Primary objective
The aim of the study is to test the effect of selected market anomalies on expected return in different sectors of the JSE.
13.2 Empirical objectives
In accordance with the primary objective of the study, the following empirical or secondary objectives were formulated:
to test the effect of selected market anomalies on expected returns on different sectors of the JSE,
to compare the performance of different asset-pricing models and their ability to account for market anomalies in different sectors of the JSE, and
to test the applicability of the recent Fama and French five-factor model (FF5-factor) in estimating the expected return on the JSE.
1.4 IMPORTANCE OF STUDY
The topic of market anomalies tends to be controversial because the presence of market anomalies tends to vary from sample to sample, implying that it is difficult to generalise the effect of market anomalies on stock returns. Additionally, it has been shown that after market anomalies are analysed and documented in academic literature they often disappear, reverse or weaken. It is, therefore, important to conduct a further study on this topic. This will give investors a broader view of different methods, which can be used to estimate expected returns, as no one model has been said to be accurate. Conducting a sector analysis will be indicative as to which market anomalies impact expected return of individual sectors and this would give South African investors a clear picture of which sectors are affected more by the specific market anomalies. Overall, this study will add to the existing body of knowledge on the effect of market anomalies and stock returns in different sectors.
6
1.5 RESEARCH DESIGN AND METHODOLOGY
1.5.1 Literature review
The literature review constitutes of only secondary information, textbooks, journal articles and other relevant sources were used to collect and review the theory and information. The literature component of this study discussed the theoretical aspect of the efficient market hypothesis, the concepts of the behavioural approach of the stock return, the theoretical aspects of the capital asset-pricing model, the theoretical aspects of the Fama and French three-factor (FF3-factor) and five-factor (FF5-factor) models. It further reviewed the empirical studies that investigated the effect of market anomalies on the stock return in the developed and emerging markets.
1.5.2 Empirical study
1.5.2.1 Data collection and sampling
The research study made use of secondary data collected from McGregor BFA, the JSE and the South African Reserve Bank (SARB). The sample period of study runs from January 2002 to December 2014 in order to cover a full investment cycle. The study made use of monthly and annual observations for the purposes of asset-pricing tests, and to determine the effect market anomalies have on stock returns. Variables used included the expected returns of each sector of the JSE, firm size, value, momentum, the January effect, profitability and investment variables.
1.5.2.2 Data analysis
To achieve the empirical objectives, the study estimated the CAPM, FF3-factor model and the Carhart four-factor model. The study started first by estimating the CAPM, as it is the standard or benchmark asset-pricing model in order to compare the performance of different asset-pricing models; then the FF3-factor model were estimated to examine size and value of anomalies, given by equation (1.1). The study went on to estimate the Carhart four-factor model given by equation (1.2), which extends the FF3-factor model by including an additional variable of momentum. The study extended the four-factor model by including the January effect given by equation (1.3) and finally estimated the FF5-factor model given by equation (1.4).
7 The FF3-factor model is as follows:
Rit−Rft= αi+ βmi(Rmt− Rft) + βsiSMBt+ βviHMLt + et (1.1)
Where:
𝑅𝑖𝑡−𝑅𝑓𝑡 is the excess return of a stock;
𝛼𝑖 is Jensen’s alpha and is the intercept of the regression; 𝛽𝑚𝑖 is market beta;
(𝑅𝑚𝑡− 𝑅𝑓𝑡) is the market return minus the risk free interest rate;
𝑆𝑀𝐵𝑡 is small minus big market capitalisation (proxy for firm size); 𝐻𝑀𝐿𝑡 is high minus low book-to-market ratios (proxy for BE/ME);
𝛽𝑠𝑖 & 𝛽𝑣𝑖 are the factor loadings other than market beta and the beta for momentum; 𝑒𝑖𝑡 is the error term.
The four-factor model of Carhart is as follows:
Rit−Rft= αi+ βmi(Rmt− Rft) + βsiSMBt+ βviHMLt+βmomiMOMt + eit (1.2) Where:
𝑀𝑂𝑀𝑡 is the momentum factor.
The four-factor model of Carhart with January effect is as follows:
Rit−Rft= αi+ βmi(Rmt− Rft) + βsiSMBt+ βviHMLt+βmomiMOMt + βjani Jant+ eit (1.3) Where:
8 FF5-factor model1 is as follows:
Rit−Rft= αi+ βmi(Rmt− Rft) + βsiSMBt+ βviHMLt+ βrmwiRMWt+ βcmaiCMAt+ εit (1.4)
Where:
RMWit is diversified portfolios of stocks with robust and weak profitability; and
𝐶𝑀𝐴it is the difference between the returns on diversified portfolios of the stocks of low and high investment firms.
Hypothesis tested
The following hypothesis were formulated and subsequently tested: Hypothesis test 1
Null hypothesis (Ho): βsi = βvi = 0
Alternative hypothesis (Ha): βsi ≠ 0 and βvi ≠ 0
The null hypothesis states that size and value are jointly equal to zero, whilst the alternative hypothesis states that size and value joint effect is different from zero for both the restricted and unrestricted FF3-factor model in equation 1.1.
Hypothesis test 2 Ho: βsi = βvi = βmomi = 0
Ha: βsi ≠ 0 and βvi ≠ 0 and βmomi ≠ 0
The null hypothesis states that size, value and 6-month momentum effects are jointly equal to zero, whilst the alternative hypothesis states that size, value and 6- and 12-month momentum effects are different from zero for both restricted and unrestricted Carhart four-factor model in equation 1.2.
1 Please note that the FF5-factor model does not include the momentum variable, as the model is derived from the FF3-factor model.
9 Hypothesis test 3
Ho: βsi = βvi = βmomi = βjani = 0
Ha: βsi ≠ 0 and βvi ≠ 0 and βmomi ≠ 0 and βjani ≠ 0
The null hypothesis states that size, value, 6 and 12-month momentum and January effect are jointly equal to zero, whilst the alternative hypothesis states that size, value, 6-month momentum and January effect are different from zero both restricted and unrestricted equation 1.3.
Hypothesis test 4 Ho: βrmwi = βcmai = 0
Ha: βrmwi ≠ 0 and βcmai ≠ 0
The null hypothesis states that investment and profitability variables, in equation 1.4, are jointly equal to zero, whilst the alternative hypothesis states that investment and profitability variables are different from zero.
1.6 CHAPTER CLASSIFICATION
This study comprises of the following chapters:
1.6.1 Chapter 1: Introduction, problem statement and background of the study
The first chapter focuses on the background and the aim of the study. The problem statement, research objectives, as well as the research method are discussed.
1.6.2 Chapter 2: Literature review
Chapter 2 provides a review on the efficient market hypothesis and behavioural finance. Furthermore, the chapter reviews all the relevant literature on market anomalies that impact stock returns, both internationally and in South Africa. This includes the size effect, value effect, momentum effect, January effect, profitability effect and investment effect. Finally, the chapter reviews the empirical studies of both developed and emerging markets.
10
1.6.3 Chapter 3: Research design and methodology
Chapter 3 describes the data used and methodology employed for performing analysis. This is done by describing the sample period, data collection, different sectors and the different variables, which were used as inputs in the different models. The chapter continues by discussing the different models employed in the study.
1.6.4 Chapter 4: Empirical findings and discussion
Chapter 4 is based on the theoretical foundation of the preceding chapters; the chapter presents the empirical analysis of the cross-sector and the FF5-factor model. Moreover, the chapter goes on to discuss the findings of the empirical analysis.
1.6.5 Chapter 5: Conclusions and recommendations
Finally, Chapter 5 summarises the study, provides conclusions of the findings, provides recommendations and provides suggestions for future research.
11
CHAPTER TWO: LITERATURE REVIEW
2.1 INTRODUCTION
For many years, efficient market hypothesis (EMH) has been the cause of much debate in financial stock markets. This area of research has attracted many scholars (Bachelier, 1900; Fama, 1965; Fama, 1970; Gibson, 1889; Granger & Morgenstern, 1970; Roberts 1967; Kendall, 1953; Malkiel, 1973 & 1992) for various reasons. First, it was discovered that stock prices moved in a random fashion. Secondly, the EMH argues that new information is independent from other news and arrives in a random fashion. In the third place, the EMH assumes that investors always act rationally and stock prices adjust rapidly to new information and should reflect all available information (Shleifer, 2002:2). Researchers and financial practitioners have been in debate about the EMH and behavioural finance for some of the reasons mentioned above. One of the views is that the EMH strongly believes that investors make rational decisions in order to maximise expected utility (Latif et al., 2011:1). In contrast, behavioural finance suggests that markets are not rational and investors make irrational decisions, which may lead them to over- or under-price stocks (De Bondt & Thaler, 1994). Even though the EMH assumes that markets are efficient only when all relevant information is reflected in the price of a security, behavioural finance considers that markets are not as efficient as the EMH suggests, and to some extent, security prices are predictable. It has been established in past studies (Banz, 1981; Basu 1977; Bhandari 1988; Fama, 1992 &1993; Lakonishok et al., 1994 & Reinganum, 1981) that markets are inefficient as expected returns were found to be higher than market returns.
This chapter aims to put forth the case of EMH and behavioural finance, but with more emphasis placed on behavioural finance as it is an advocate for the presence of market anomalies, which are tested by this study. Therefore, this chapter proceeds with a discussion of the EMH and the random walk theory briefly. It then discusses the three forms of market efficiency and its implications. The chapter continues by discussing behavioural finance and its market anomalies and reviewing empirical studies on stock market anomalies in both developed, emerging stock markets and of the Johannesburg Stock Exchange (JSE). Finally, this chapter ends with the summary and concluding remarks.
12
2.2 EFFICIENT MARKET HYPOTHESIS AND RANDOM WALK THEORY
The theory on stock market efficiency is relatively vast and dates back to the 16th century. A researcher named Kendall (1953) examined the randomness of stock prices. It was assumed stock prices reflected the recurrent patterns of peaks and troughs in economic performance, which should be indicated in those prices. Kendall’s aim was to determine whether normal price cycles occurred in the market, but he established that price movements were random and not recurrent as assumed (Kendall, 1953:11). The results of Kendall’s research were not received positively at first by financial economists as they were of the impression that the stock market was influenced by unreliable market psychology and the irrationality of the market (Bodie et al., 2010:228). However, on further investigation the financial economists were in agreement with Kendall’s findings after reversing their interpretation. Kendall and other financial economists came to the consensus that random price movements signalled an efficient market rather than an irrational one (Shafi, 2014:9). As a result, a new era of research was unleashed by Kendall’s discovery, as researchers sought what influences stock price movements (Kendall, 1953:13).
After Kendall’s (1953) discovery, one frequently asked question by researchers and financial analysts was to what extent past prices could be used to make meaningful predictions about future security prices (Fama, 1965a:34). Various answers to this question have been given, one of which is from technical analysts who try to predict the direction of prices through the study of past market data, mainly prices and volume. Technical analysts argue that history tends to repeat itself and that stock prices are predictable and trends continue to persist until something occurs that changes that pattern or trend (Bodie et al., 2010:233). On the other hand, statisticians can debate that the past is useless in predicting the future and it cannot be used to make meaningful predictions. However, it is evident thatresearchers have looked into the past in order to try to predict what will happen in the future. The view of the statisticians is based on the fact that stock prices follow a random walk, meaning that stock prices are unpredictable and one cannot use yesterday’s stock prices to predict tomorrow’s stock prices (Fama, 1965a:34).
The concept of random walk was established because it was believed that stock markets are efficient. Fama (1970:383) has defined an efficient market to be one where stock prices fully reflect all available information. Furthermore, in an efficient market, investors have no competitive advantage to information, and as a result, they are unable to exploit the use of the
13
information in the hopes of gaining abnormal returns. Therefore, the investment decisions are influenced solely by market efficiency. Additionally, in efficient markets, investors earn merely normal returns when analysing historical, public and inside information (Degutis & Novickyte, 2014:8). This does not generate abnormal returns because the prices move in a random fashion, as a result, the current price of the security is an estimate of the true value or intrinsic value of that security (Yalcin, 2010:24). The intrinsic value of a security is said to be the true value of a security’s potential earnings (Yalcin, 2010:24). The market price of a security is influenced by the actions of competing investors as the actual price of a security wanders randomly about its intrinsic value. Therefore, a market where successive price changes in individual securities are independent is, by definition, a random walk market.
2.2.1 Forms of market efficiency
The EMH is known for having three forms of market efficiency, identified as weak form, semi-strong form and strong form (Latif et al., 2011:2). These forms differ according to the degree of information reflected in security prices, as their classification is based merely on the information reflected in the stock prices (Brealey et al., 2011:317). Each form of the EMH is discussed subsequently in detail.
Weak form of market efficiency
The first form is the weak form of efficiency, which assumes that current prices reflect all security market information. This form of market efficiency suggests that current prices reflect all past information, such as past stock returns and short interest or trading volumes (Mobarek & Keasey, 2002:4). In the weak form efficiency, the use of technical analysis will be rendered useless as the theory of random walk strictly states that stock prices are unpredictable and one cannot use previous market prices in the attempt to forecast future prices (Brealey et al., 2011:317). However, it is possible for an investor to obtain above average returns in this weak form by using public and private information (i.e. fundamental analysis).
Semi-strong form of market efficiency
The semi-strong form assumes that prices rapidly adjust to the release of all public information. In this form of market efficiency, current stock prices reflect all past and public
14
information about the stock (Degutis & Novickyte, 2014:8). Therefore, if markets are in this form of market efficiency, it is expected that prices should immediately adjust to the release of new information such as dividend announcements, acquisition and mergers and to the issue of new shares (Fama, 1970). Thus, in this form it becomes difficult to obtain above average returns, as technical and fundamental analyses are inadequate. However, in semi-strong form of market efficiency an investor can earn above average returns by using information that is not publically available (private information).
Strong form of market efficiency
Finally, the strong form assumes stock prices fully reflect all past, public and private information. In this form, technical, fundamental and superior market analysts are theoretically not supposed to outperform the market, as all past, public and private information are already priced-in and is reflected in the current price (Latif et al., 2011:2). If a market is efficient in the strong form then it is evident that it is also efficient in the semi-strong and weak form efficiency; nonetheless, a market can be efficient in the weak form efficiency but not in the semi and strong form efficiency. Table 2.1 below is a summary of the forms of market efficiency and the type of implications each form entails. The implications of EMH are discussed in the sub-section that follows.
Table 2.1: Summary of the three forms of market efficiency and its implications
Market prices reflect
Forms Past data Public data Private data Implications Weak Fundamental analysis Semi-strong Fundamental analysis Strong
Source: Own construct
2.2.2 Implications of EMH
Efficient market hypothesis has implications on the trading strategies used in a stock market, which is said to be efficient. The strategies include trading strategies, such as technical analysis and fundamental analysis and portfolio management strategies, such as active and
15
passive portfolio management. These applications of strategies are dependent on the degree of efficiency in the market and each is discussed details below.
2.2.2.1 Technical analysis
Technical analysis is a methodology used to predict the direction of prices through the study of past market data, mainly prices and trading volumes (Eriotis et al., 2006:75). Technical analysts assume that what happened in the past will repeat itself in the future (Reilly & Brown, 2012:161). Furthermore, technical analysts believe that stock prices are predictable and that trends will continue to persist until something occurs that will change that pattern or trend (Bodie et al., 2010:233). Technical analysts do not attempt to determine intrinsic values like fundamental analysts analyse patterns and indictors on charts to determine future performance of the security (Mpofu et al., 2010:75). The kind of indicators or patterns technical analyst look for are peaks, troughs, booms, trends and they also make use of past market prices, trading volumes and market activity in order to make a decision of buying or selling a specific stock. Thus, technical analysts use past trends and patterns in the hopes of generating abnormal returns. However, generating abnormal returns is impossible in an efficient market as technical analysis is reliant on the informational inefficiency of a market. 2.2.2.2 Fundamental analysis
In contrast to technical analysis, fundamental analysts attempt to determine the intrinsic value of a security, which is said to be the true value of a security and is dependent on the potential earnings of the security (Fama, 1965). Fundamental analyst attempt to determine the intrinsic value by means of conducting macroeconomic factor analysis, industry analysis and company valuation (Mpofu et al., 2013). All this is in hopes of determining whether the security’s price is below or above the intrinsic value and if it tends to move towards the intrinsic value (Fama, 1965b:55). Therefore, conducting fundamental analysis aids in making meaningful forecasts of future security prices. However, the EMH states that all relevant information is reflected already in stock prices, therefore, fundamental analysis maybe adequate in the weak form of market efficiency (Shostak, 1997:28).
2.2.2.3 Portfolio management and the implication of active and passive management Another strategy affected by the market efficient is portfolio management, which is used by investors to select their portfolios. Portfolio management entails selecting a portfolio of stocks rationally by selling over-valued stocks and buying under-valued (Reilly & Brown,
16
2012:164). Portfolio managers do so to help investors and institutions to meet their investment objectives, to allocate their funds efficiently and to balance the investors risk and reward (Bodie et al., 2011:378). Investors can use a passive strategy, which aims at utilising a buy-hold strategy as, in efficient markets; stocks are fairly priced, given all available information is reflected in prices. Therefore, it makes no sense to spend large amounts of money on brokerage fees, which do not increase the expected return of a stock (Bodie et al., 2011:378).
Contrary to a passive strategy, an investor or institution can opt for an active strategy, which aims at buying undervalued stocks and selling overvalued stock frequently in the hopes of purposefully outperforming the market (Reilly & Brown, 2012:165). Active portfolio management goes hand in hand with fundamental analysis as portfolio managers conduct in-depth macroeconomic and industry analysis and in scrutinising a firm’s financial statements in the hopes of finding something other analysts have overlooked (Reilly & Brown, 2012:164). Active portfolio managers are often encouraged to purse medium and small size stocks as information on those stocks are less publicised and, therefore, smaller numbers of analysts follow these stocks. As a result, the intrinsic value of these stocks tends to differ from the prices. This may be due to the fact that small size stocks are less publicised and, therefore, less information is available on small sized firms and these neglected stocks may be less efficient hence, the size anomaly is evident in most stock markets. As a result, active managers have the opportunity to outperform the market by adding small sized firms into their portfolios (Reilly & Brown, 2012:164).
2.2.2.4 Violation of EMH assumptions
There has been much controversy about the EMH and in instances, it has been established that market movements can go against the assumptions of the EMH. This scenario of deviation from the EMH assumptions is known as a market anomaly. A market anomaly can be defined as a regular pattern in an asset’s return, which is unknown (Bodie et al., 2010:240). Since this pattern is regular, it implies that investors are able to take advantage of it, as there is some predictability to it. Examples of market anomalies, which seem to predict superior returns, arise from factors such as a stock’s price-earnings ratio. Basu (1977, 1983) established the phenomenon that portfolios of low price-earnings ratio earn higher returns than portfolios of high price-earnings ratio. Furthermore, the P/E effect seemed to persist even when returns are adjusted for portfolio risk (Basu, 1977). This could be caused by
17
mispricing stocks in the market given their price-earnings ratio or due to incorrect risk-adjusted returns.
Another market anomaly is one, which is commonly analysed and it is known as the size effect, as first documented by Banz (1981). The size effect is indicative that small capitalisation companies on average earn higher expected return than their larger counterparts. Researchers such as Keim (1983), Roll (1983), Blume and Stambaugh (1983) and Rozeff and Kinney (1976) recognise a similar market anomaly which occurs in the first two weeks of January known as the small-firm-in January effect which deals with movements in stock prices in a given period. The main assumption of the January effect is that small firms tend to earn higher returns than large firms in the market do in the first two to three weeks of January (Mazal, 2009:7). It is evident that the January effect may be difficult to merge with EMH as it appears regularly and because it is well publicised.
Additionally, to the small-firm-in-January effect is related closely to the neglected-firm effect, which refers to small firms, being neglected by large institutional investors as information on small firms is not easily accessible, making them more risky and thus they have to reward investors with higher expected returns (Arbel & Strebel, 1983). Similarly, Amihud and Mendelson’s (1986) study on liquidity effects also supports the small-firm-in-January effect as large institutional investors neglect small firms and thus the stocks become less liquid. As a result, these small firms provide abnormal returns in January as they compensate for lack of information and liquidity (Yalcin, 2010:33).
The abovementioned market anomalies have created much controversy in the field of finance. These market anomalies are contradictory and violate the assumptions of EMH; they are documented regularly and there is significant empirical evidence supporting their existence.It has been documented that these anomalies can be taken advantage of by the use of simple trading strategies, thus resulting in profits. The persistence of these anomalies as suggested might be due to the mispricing of asset-pricing model, which at times fail to adjust for risk.
2.3 BEHAVIOURAL FINANCE
The efficient market hypothesis, since its development, has established two important predictions. First, is that stock prices reflect all available information to investors and secondly, it would be challenging for active traders to outperform passive traders
18
purposefully. As a result, developing measures of the true or intrinsic value of a security and testing if prices match these values, are problematic. Therefore, the main focuses of market efficiency tests are centred on the performance of active traders (Bodie et al., 2010:261). This results in active traders being unable to outperform passive strategies. Examples include the Standard and Poor 500 stock index, which outperforms the market 60 percent to 80 percent of the time, therefore attesting that stock markets are efficient to some extent (Ricciardi & Simon, 2000:2).
However, a new school of thought was developed, which contradicts the theory of EMH. This new school of thought is known as behavioural finance. The main aim of behavioural finance is to attempt to understand the thought process and reasoning of investors when making decisions about purchasing a stock (Ricciardi & Simon, 2000:2). The theory combines theories based on finance and psychology to explain the presence of stock market anomalies.
Behavioural finance argues that, even though EMH believes that markets are rational, investors do not always make their investment choices rationally. Therefore, behavioural finance is centred around the theory, which tries to explain how people think, known as cognitive psychology and limits to arbitrage, which allows investor irrationality to be substantial and have long-lived impact on stock prices (Shefrin & Statman, 2011). The purpose of this sub-section is to discuss the two categories of behavioural finance theory and to define market anomalies, which are examined in this study.
2.3.1 Theories of behavioural finance
Behavioural finance encompasses various theories; however, it is made up of two major schools of thoughts, namely the heuristics decision process and prospects theory. The category of heuristic decision process deals with measures people use during the decision making process, while prospects theory deals with the idea that people do not always behave rationally (Ricciardi & Simon, 2005:5).
2.3.1.1 Heuristics decision process
The word heuristics refers to a rule of thumb used to make the decision making process easier, however, using a rule of thumb does not always work out when circumstances change; therefore, it may result in suboptimal investment decisions (Ritter, 2004:3). Some heuristic
19
biases include representativeness, overconfidence, anchoring, adjustment, and lastly framing. Each of these biases is discussed in details in the sub-sections below.
2.3.1.1.1 Representativeness bias
Representativeness heuristics is a decision-making process concerned with the probability of an event occurring under uncertainty (Kishore, 2004:6). It is said that when a person is confronted with a situation, the brain refers to experiences, which are similar to the current situation to aid in making a decision about a current event or situation (Tversky & Kahneman, 1974).
2.3.1.1.2 Overconfidence bias
Overconfidence is defined by Mahajan (1992:330) as an overestimation of probabilities for a set of events. In simple terms, this definition means that individuals overestimate the probability of an event occurring or overestimate their own abilities. Gervais and Odean (2001) examined this, and they conducted a model in terms of which traders become overconfident as they become successful. This infers that the traders in their model learn from their success and failure and in that way they become better traders and more confident in their abilities (Gervais & Odean, 2001:19). However, some individuals do not necessarily learn from their past mistakes and failures and as a result this adds to the overconfidence problem (Ricciardi & Simon, 2000:4). Other researchers have examined overconfidence in gender. Research done by Barber and Odean (2001:289) revealed that men tend to be more overconfident with their trading abilities and trade more than women. An example of overconfidence is when an investor ‘puts all his eggs in one basket’ referring to the issue of lack of diversifying one’s portfolio. This may be caused by the preference of the investor or because of familiarity of the investor with a particular stock or bond (Ritter, 2003:4).
2.3.1.2 Anchoring and adjustment bias
Anchoring and adjustment refers to a decision-making mechanism where one has to make a decision under uncertainty using a reference point (Kishore, 2004:6). For example, the information persons already have and can adjust until a final decision is reached (Kishore, 2004:6). For example, if you are an investor and you are planning to buy stocks in a particular company, as you have heard that the company is in the process of merging with a huge rival company. It is expected that the purchase will enable the company to own 50 percent of the market share in that industry, resulting in more profits for the company and
20
dividends for investors. As an investor, you will base your decision on past and new information entering the market on the two companies. The more news you hear about how well the merger is going, the more you consider buying the shares. However, if you hear that the merger proceedings are not going well and other shareholders are selling their shares, the more inclined you will be as an investor not to purchase the stocks. Hence, investors anchor and adjust their decisions based on the information they already have and information arriving in the market at the time they decide to purchase stocks.
2.3.1.2.1 Framing bias
Framing refers to how individuals or investors present and perceive a concept (Ritter, 2004:4). Well-known researchers for examining this type of behaviour are Tversky and Kahneman (1981). They conducted a series of experiments in which they asked students from various universities to answer numerous problem sets in order to establish the attitudes of people towards the risks involving gains and losses Tversky & Kahneman, 1981:457-458). They established that participants made decisions based on how the questions were presented, and even if participants were asked the same question in a different manner, they gave different responses all the time.
2.3.1.2.2 The prospects theory
The second category is the prospects theory, based on the assumption that people do not always act rationally and that their choices under uncertainty are influenced by biases motivated by psychological factors (Ricciardi & Simon, 2000:5). It is presumed that probabilities do not always coincide with preferences, which are considered as a function of decision weights (Schwartz, 1988:82). It suggests that these decision weights tend to overweigh small probabilities and under-weigh moderate and high probabilities. The prospects theory includes risk and loss aversion, and regret aversion which are discussed below.
2.3.1.2.3 Risk and loss aversion bias
Risk and loss aversion bias states that investors tend to be more risk averse when faced with gains and more loss averse when faced with the possibility of a loss. For example, consider two investment portfolios, A and B. Portfolio A offers a sure gain of R 10 000 and portfolio B offers a possibility of gaining 80 percent of R 13 000 (13 000*0.8= 10400) with a possibility of a 20 percent chance of gaining nothing. Most people may opt for portfolio A,
21
illustrating that the majority of investors become risk averse when faced with gains (Kahneman & Tversky, 1979). In contrast, investors can become loss averse when faced with the possibility of an 80 percent chance gaining nothing and a sure gain of 20 percent. For example, portfolio B offers much higher gains than portfolio A. This illustrates that when investors are faced with a probability of them facing a loss, they would take the chance instead of the sure gain of 20 percent.
2.3.1.2.4 Regret aversion bias
Risk aversion bias is related to the theory of regret, which has to do with the emotional reaction a person has when deciding between two objects. This theory of regret assumes rationality as decisions are based on expected returns and expected regret (Ricciardi & Simon, 2000:5). Regret can be described as an emotional reaction caused by the opportunity cost of sacrificing one object for the alternative (Mohr & Fourie, 2009:7).
2.3.2 Adaptive market hypothesis
Adaptive market hypothesis (AMH) is a fairly new field of study, which tries to reconcile EMH and behavioural finance. The theory of AMH is compiled from various literature of researchers such as Barrett et al. (2002), Farmer (2002), Lo (2002, 2004 & 2005), Mangel and Clark (1988), Pinker (1997), Simon, (1982), Trivers, (1985) and Wilson (1975). The AMH is composed of several components such as evolutionary psychology, behavioural ecology, complex systems evolutionary biology and bounded rationality in economics.
The debate around EMH has caused much controversy, which has led to the field of behavioural finance that opposes the major assumptions of EMH. It is difficult at times to reach a common ground with both these financial theories. Therefore, the field of AMH introduces a new approach to financial markets, which is influenced greatly by developments in the discipline of evolutionary psychology (Farmer & Lo, 1999; Farmer, 2002) in terms of which market participants evolve, compete and adapt to changing market conditions.
Wilson’s (1975) application of the principles of reproduction, competition and natural selection to social interactions resulted in fascinating explanations for human behaviour. For example, this refers to how people select their partners, morality, ethics, altruism, kin selection and language and how that could be used in the context of finance and economics. As a result, this enables EMH and behavioural finance to reconcile. The subsection that
22
follows will illustrate how AMH mergers both EMH and behavioural finance. Table 2.2 below is a summary of how AMH combines EMH and behavioural finance.
Table 2.2: Summary of AMH
EMH Behavioural Finance AMH
Rationality
Irrationality, prospects theory, decision heuristics process
Combination of financial and cognitive psychology
theory
Market efficiency Market inefficiency
Market are not fully efficient and efficiency depends on market participants, market
size and economic conditions Once risk factor Multiple risk factors
Source: Own construct
The AMH is a new theory aimed at reconciling EMH and behavioural finance. It is viewed as a new version of EMH, which incorporates psychological biases. EMH is a combination of market conditions, number of participants, market size and the ability of security prices to reflect information instantaneously (Neely et al., 2009). The theory of AMH suggests that market participants are dependent on economic profits for their survival in order for market interactions and financial innovation to be derived readily. This implies that in a large market where resources and prices of securities are easily available, there are a large number of investors competing for those stocks. Therefore, this market tends to be more efficient as investors compete for these stocks, adapt to market conditions and bring prices of stocks back to their intrinsic value (Lo, 2005). In contrast, behavioural finance suggests psychologists apply heuristic to finance before reconciling EMH and behavioural.
As a result, markets undergo profits and losses due to changing market conditions, opportunities that exists and as new participants enter and exit the market and opportunities shift, participants will be affected by such changes. Behaviourists believe that the downfall of rational thinking is caused by greed and fear and that the ability to adapt improves the chances of achieving average returns (Lo, 2005). Therefore, it can be concluded that there is an important link between emotion and rationality when making decisions, as they are not conflicting but rather complementary (Damasio, 1994).
23
2.4 MARKET ANOMALIES
Market anomalies are patterns found in cross-sectional or time series data such as security prices, which cannot be captured by the asset-pricing model and is not predicted by traditional finance theory (Schwert, 2003:939). The documentation of market anomalies signals a transitional phase towards a new standard or paradigm. It has been discovered that market anomalies are due to empirical results which are dependent on a joint null hypothesis stating that markets are informational efficient and security returns behave rationally according to a pre-specified model known as the CAPM (Keim, 2006). However, if the joint null hypothesis is rejected, it cannot be said that it was due to either part of the joint null hypothesis. As a result, market anomalies often are seen as a form of market inefficiency and such a conclusion is presumptuous as the rejection of the joint null hypothesis may be due to an incorrect asset-pricing model (Keim, 2006:1).
It has been documented that some of these anomalies disappear when documented. This is a result of the actions of investors when they try to exploit the return patterns, or if these anomalies were sample specific, while other anomalies continue to persist hence researches stand firm that it is not due to market inefficiency (Latif et al., 2011). This section aims at discussing the types of market anomalies inherent around the globe and factors that may lead to the presence of these market anomalies on the JSE.
2.4.1 Types of market anomalies
There are various types of market anomalies, which have continued to violate the assumptions of EMH. These anomalies include calendar anomalies, fundamental anomalies and technical anomalies and each of them is discussed below.
2.4.1.1 Calendar anomalies
Calendar anomalies are those that relate to a specific period such as day-of-the-week. They include Monday effect, weekend effect, turn-of-the-month, January effect and holiday effect to name a few. These are known as patterns in security returns that relate to the calendar (Hansen & Lunde, 2003:1).
An assumption usually referred to when researchers discuss calendar anomalies is that the past behaviour of a stock’s price patterns contains information about future price patterns.
24
This infers that investors can exploit these calendar anomalies by trying to predict future price patterns. However, these anomalies are contradictory with the main theory of market efficiency (EMH). The weak form of market efficiency stipulates that all past information is already factored into the prices of securities and cannot be used to predict future price or in any way to purposefully outperform the market (Young, 1994; Faweson et al., 1996; Poshakwale, 1996). Some of the different calendar effects are discussed in the sub-section below.
Day-of-the-week or weekend effect
This effect was first documented by Osborne (1962) and analysed by Cross (1973), French (1980), Gibbons and Hess (1981), Keim and Stambaugh (1983). The day-of-the-week effect states that expected returns and normal returns are not the same for each day of the week. Hess (1981) documented that Friday yielded higher returns than any other day and Monday had the lowest returns. Furthermore, for any Monday effect to be a true stock market anomaly, Monday returns must not only be low or negative, but they must also be significantly different from the returns during the rest of the week (Ajayi et al., 2004:59). Thus, investors’ behaviours on different days of the week violate the EMH assumption of rationality.
Holiday effect
Lakonishok and Smidt (1988) and Petengill (1989) documented that the returns tend to be higher before public holidays due to the behaviour of investors. Investors tend to be more optimistic and they are more inclined to trading before a public holiday, as a result there is as an increase in returns and trading volumes, whilst after public holidays returns and trading volumes remain low (Zafar et al., 2012:7262).
January effect
The month of January sets the standard for the calendar year and thus much significance is placed on it. It is believed that small company stocks tend to generate higher returns than any other asset classes in the first two weeks of January (Keim, 1983; Mazal, 2009). This normally occurs between the last trading day in December of the previous year and the fifth trading of the new year in January (Reinganum, 1983). This is due to investors selling off their losing positions at the end of December, as funds have to report end of year figures.
