Print Reference This Reddit This

Literature Review About The Efficiency Market Hypothesis

3.1 Introduction

Many previous researchers have studied in the field of seasonal regularities, particularly for the day of the week effect, January effect, the quarter effect and holiday effect. Some of them considered the seasonal effect is the divergence from the theories of asset pricing behaviour and it could be evidences of market inefficiency (see Lakonishok et al. 1995). Another researcher such as Fama and French (1993) thought that seasonal effect was just anomalies for temporarily exist in the market and the market is still efficient according to the market efficient hypothesis. Other researchers focus on explanation of earing abnormal profit from arbitrage opportunities caused by seasonal anomalies.

In this chapter, we will first discuss the concepts and arguments of the Efficient Market Hypothesis, and followed by theories of Behaviour Finance. Afterwards, irrational exuberance and anomalies which include firm, event, seasonal and accounting anomalies will be introduced. We will focus on previous research review of seasonal anomalies and try to find out how the theories of Efficient Market Hypothesis and Behaviour Finance explained anomalies and what the difference between those two financial theories.

3.2 Efficiency Market Hypothesis

3.2.1 Efficient Market

It was Fama (1970) who first introduced the concept of an efficient market to the literature of financial theories. He defined the efficient market as a market in which all prices of securities quickly and fully reflect all available and relevant information about the assets. The Efficient Market Hypothesis tries to explain that all relevant public and private information are available in the market and will reflect the stock prices. Even professional investors can take advantages from neither technical analysis nor fundamental analysis for investment analysis. All investors act in a “uniform” and “rational” way and they believe that all the prices are fair in efficient market.

In an efficient market, investors are able to use all available information to assist decisions making of trading stocks, also both buyers and sellers have the same set of information. However, some researchers are trying to challenge the Efficient Market Hypothesis, such as Shiryaev and Albert (1999) argue that the concept of efficient market is based on assumption that today’s prices and all available information are considered completely, prices change only when this information is updated. Some researchers believed that past stock price can be used to predict future stock price and fundamental analysis is useful when judge the trend of stock price. Additionally, some researchers found out some irrational stock events like ‘internal bubbles’ or ‘irrational exuberance’, which they thought this will provide arbitrage opportunities from trading stocks.

3.2.2 The Random Walk Theory

New information must be unpredictable; if it is predicted, it should be part of today’s information. Thus if stock prices change with unpredictable information also must move unpredictably. This brings about the argument that stock prices should follow a random walk. The random walk theory was introduced in 1900 by Louis Bachelier, he stated that “stock prices respond in various degrees of efficiency to information and influences as they become known” (Bachelier, 1990, as cited by David and Daniel, 2006, p.27). In 1953 Maurice Kendall published a study that he found stock prices move with no certain pattern, prices’ movement are irrespective of their past performance (Palan, 2004). These empirical observations show that the random walk theory seems to be contrary to the anomalous price behaviour which was examined by Working (1934). Then Alexander found the major evidence of trading rules, he concludes that if return data were based on end-of-period prices, returns would move randomly due to autocorrelation problem (Alexander, 1961). After that some researchers found that traditional tests of random walk theory had errors due to imitative autocorrelation caused by non-synchronous trading, such as (Pope, 1989).Consequently, Fama (1965) divided random walk theory into two hypotheses: continuous price changes are independent and probability distribution decides price changes.

A random walk could be the result of prices that always reflect all current knowledge. In the random walk theory, stock price movement cannot be predicated from previous prices because there is no relationship between two sets of prices, and technical analysis is a waste of time in that prices cannot be predicated from historical data (Faerber, 2007). Predictable stock price movements would bring about damning evidence to stock market inefficiency, because the ability to predict prices indicate that all available information were not ready to reflect stock prices. Therefore, Summers (1986) said that tests of random walk theory are not strong enough to against alternative hypotheses of market efficiency. New tests would reject the random walk model in the US market, as Mackinlay (1988, p.61) finds that random walk model is not consistent with the stochastic behaviour of weekly returns, especially for the smaller capitalization stocks.

3.2.3 Efficient Market Hypothesis

Evensky (1997) states that the Efficient Market Hypothesis considers the relationship of stock prices and actions of buyers and sellers. The assumption of EMH is that securities are always fairy priced, it assumes that the market is made up by rational investors and agents (Focardi and Jonas, 1997, p.166). Thus the implies that new information would expose a firm in order to cause share price move rapidly and rationally, even leading to the direction of the share price movement and size of the movement.

Cuthbertson and Nitzsche (2001) pointed out that one key aspect of the Efficient Market Hypothesis is that traders absorb any relevant information quickly to make decisions for the assets price or returns. The second key element is that rational investors move a market price to a fair value is determined by present value and further dividends. In Efficient Market Hypothesis theory, there is no abnormal profit from trading. The concepts behind the EMH can be used for all speculative asset types, such as stock, bonds and derivatives. Terry (1984, p2) stated that the Efficient Market Hypothesis "provides a theory which helps to fit the facts to the history of stock price levels and price changes."

The validation of the Efficient Market Hypothesis is an empirical problem. Fama (1970) suggested that Efficient Market Hypothesis could be more easily applied if its validity were divided into three sub-themes. Also he said that three levels of efficiency were used to achieve excess returns. Three different level of efficiency are shown below:

Models of share price behaviour (Rutterford & Davison, 2007, P.87)Perfect Markets

Efficient Markets Hypothesis



Fair Game


Random Walk

3.2.4 Three Forms of Efficient Market Hypothesis

Under efficient market hypothesis, there are three types of market efficiencies: the weak form, the semi-strong form and the strong form. Hirschey (2001) defines the weak form of efficiency has been reflected by current prices. This form focuses on historical data, especially historical rates of return which means investors could not obtain profit by analysing the trends of past share prices movement. Roll (1968) first proposes that the weak form could be based on “fair game” policy instead of traditional random walk. Also Levy (1967) attempts to refute the weak form efficiency in the market via positive relationships between share price movements.

Hirschey (2001) also defines that in the semi-strong form efficiency, public information has been already reflected by current prices. In another words, if investors access public and available sources, they would expect it to be reflected in stock prices. Scholes (1972) finds that in the semi-strong form efficiency the share price that has a reaction before the merger announcement has been published, this must be based on the existence of insiders’ information. However, Shleifer (2000) argues that as information becomes public, it will affect prices, so a semi-strong form efficient market is obviously weak form efficient as well. In this market situation, investors would not able to make excess money by accessing only publicly available information.

Finally, strong form efficiency is defined as current share prices reflect all public information and nonpublic information relevant to the firm Hirschey (2001). In this type of market, no abnormal profit can be made by using any information. Argument about strong form efficiency such as corporate officers could access to pertinent information before information release and enable investors earn profit from trading of this information. Like Kaen (2003) states that serious problems for governance and for regulation of financial markets will appear in strong form efficiency from managers and insiders who have information before the information becomes public. And “insiders” are not only involved company directors and employees, this term covers anyone with sensitive information (Arnold, 2005).

Implications for investors

McLaney (2006) pointed out, capital market efficiency implies that seeking to obtain abnormal profit by observing historical data or information of security price movement and analysis of new economic information is a waste of time. Irrational investment behavior could happen when: ignorance exists of the evidence on efficiency; charts of past security price movement show pattern repeating itself; investors do not believe or follow most investment advice for a substantial period and investors know cases of people who have been extremely successful in capital market investment.

3.2.6 Market Anomalies and Efficient Market Hypothesis

The weak form of the Efficient Market hypothesis indicates that investors cannot earn abnormal returns consistently from the prediction of price change which is based on the correlation of past prices and future stock prices. In other words, there is no certain seasonal pattern exists in the market and stock prices move randomly. These price change predictions or seasonal patterns are called anomalies in the stock market (Bodla and Jindal, 2006). Lawrence et al. (2007) says although the Efficient Market hypothesis model is successful when analysing stock prices by fundamental analysis, some anomalies such as high trading volume, high volatility and stock market bubbles remain unexplained. The reason of this unexplained phenomenon is that the Efficient Market hypothesis assumes investors are rational, but some irrational investors would against the predication which leads to anomalies.

Since Fama (1970) introduced the theories of the Efficient Market hypothesis, a great deal of research focused to investigate the randomness of stock price movements. Afterwards, seasonal anomalies in stock market have been documented in the financial literature. The most common seasonal anomalies show that returns follow a seasonal pattern under an assumption of weak market efficiency. Haugen and Jorion (1996) suggest as investors learn experience from past, the seasonal effects will not exist long. And Malkiel (2003) explains that investors attempt to benefit from inefficiency lead to anomaly disappeared. Actually, transactions costs are the key fact to make anomalies disappeared.

3.3 Market anomalies

Stock market anomalies are a long-standing phenomenon. Levy (2002, p.476) defines that “a market anomaly is any event that can be exploited to produce abnormal profits.” However Marcus, Kane and Bodie (2008, p.361) define it as patterns of returns that seem to contradict the efficient market hypothesis. Previous research on market anomalies both in mature stock markets has already led to doubts of the efficient market hypothesis. Some of researchers state that market anomalies indicate market inefficiency. That is because anomalies are not real anomalies during the market efficiency tests, especially when anomalies exist for a long time.

Furthermore, Schlichting (2008) mention that stock market anomalies are caused by individual anomalies which relate to investors’ behaviour during the investment process. He gives a table that shows an overview of the most important types of market anomalies, such as overreaction/underreaction, excess volatility, announcement effect, herding, momentum hypothesis, mean reversion, winner-loser effect and price-book-ratio effect. Some literatures demonstrate the existence of anomalies challenge to choice theories (see Carmerer, 1995 and Starmer, 2000). Research on anomalies makes the market to become more efficient. There are four types of anomalies: firm anomalies, seasonal anomaly, event anomalies and accounting anomalies. This paper concentrates on analysis of seasonal anomalies which following a January effect and the day of the week effect in a separate sections.

3.3.1 Seasonal Anomalies

Seasonal anomalies or calendar effects have divergence from the week form of market efficiency due to stock returns are not random, but they are predictable. Seasonal anomaly is anomaly is based on time, it has been discussed by many professionals in the stock market. In the past 30 years, seasonal regularities have been studied by relative effects such as holiday effect (Lakonishok and Smidt, 1988), January effect (Rozeff and kinney, 1976), the time of the month effect (Ariel, 1987) and the turn of the month (Cadsby and Ratner, 1992). These anomalies are used to explain by asset-pricing models and challenge the efficient market hypothesis. Rozeff and kinney (1976) were the first to document the evidence of anomalies from 1904 to 974 in US stocks, they found highest returns exist in January rather than other months. After that, researchers did research on different types of seasonal anomalies in mature stock markets in succession. For instance, the day of the week effect was tested in the US by French (1980), Gibbons and Hess (1981) and Harris (1986); in Australian, British, Japanese and Canadian markets by Westerfiled, (1985); in Asia-Pacific markets by Ho (1990) and in French, Italian stock markets by Kato (1990) and Barrone (1990).

The best example of seasonal anomaly is the Monday effect which stock returns is significantly negative on Monday (Mehdian and Perry, 2001). If the Monday’s return can be predicted, then investors could generate abnormal returns from establishing a trading rule to observe the seasonal pattern. The Monday effect was first documented in the US stock market by Fields (1931). Afterwards, numerous studies have been done for the Monday effect or with other daily anomalies (see Lakonishok and Levi 1982 and Rogalski, 1984). Most recently, Wang, Li and Erickson (1997) find Monday effect exists in small-cap stock returns due to frequently trading activities and Monday effect for larger stocks has declined because of the dominance of institutional trading. Then, Monday effect has been examined in international stock markets.

3.3.2 Firm anomalies

Firm anomalies come from firm’s characteristics, the most important one is the small firm effect. Studies of Banz (1981) and Keim (1983) find that small firms earn more than large firms at same level of risk, where firm size is defined in terms of market value of equity and risk is market beta. In their research, smaller firm portfolios are more risky, but even the returns are measured by capital asset pricing model, there is still a premium for its portfolios. Later research such as Reinganum (1983) and Blume, Keim and Stambaugh (1984) find that the small firm effect is really significant in January due to higher beta values for the small firms, especially in the first two weeks of January. Therefore, they called this “small-firm-in-January” effect. Aby and Vaughn (1995, p.283) confirm that this result consistent with the capital asset pricing model that suggests that high betas are expected to lead to above high returns.

Another similar anomaly is called neglected firm effect, Arbel and Strebel (1983) find another impact on size-attention or neglect. They confirmed both small firm effect and neglected firm effect are caused by lack of information and limited interest of institutions. Simply speaking, a fewer analysts pay an attention on a stock, the higher the return of stock due to attention of stock analysts does affect share prices (Levy, 2002). Neglected firms tend to be small firms. Small firms bring about more risk, therefore institutions would ignore them and pick larger-capitalization firms. Both of those two anomalies have the same implication: small firms and neglected firms seem to offer a free lunch with surprising regularity (Strong, 2009). More recent study by Beard and Sias (1997) show that after controlling for capitalization there is no evidence of neglected firm premium. In 1987, a article by Merton has already defined neglected firm premium is not a market inefficiency, but is a type of risk premium.

3.3.3 Event Anomalies

Levy (2002) states that event anomalies are price changes when some easily identified event such as an announcement of company have been published. The event anomalies can be examined by event study that will be discussed later. Apart from post-earnings-announcement drift and momentum, another important event anomaly is recommendations from analysts. The more analysts recommend a stock, the more likely the stock price will fall in the near future (Levy, 2002). It is because when analysts find an undervalued stock and recommend it to investors, after investors to buy the stock the price is driven up. As more and more investors buy this stock, push the price even higher. Price will continue goes up until analysts change their buy recommendations to sell recommendations, and then price falls subsequently. Investors will be considered as irrational.

Fama (1998) points out there are two anomalies still challenge to asset pricing and market efficiency, one is post-earnings-announcement drift, and other one is earnings momentum. Both price-earnings-announcement drift and momentum have been doubts by efficient market hypothesis over past few decades, all studies that examine these anomalies take into account transaction costs. In momentum and price-earning-announcement drift portfolio, winners and losers (momentum) and good news firms and bad news firms (price-earnings-announcement drift), represent the highest abnormal return after they are formed (Moskowitz and Grinblatt, 1999).

3.3.4 Accounting Anomalies

Levy (2002, p.480) defines accounting anomalies are changes in stock prices that occur after the release of accounting information. The most typical anomalies are book-to-market ratios (B/M ratio) and P/E ratio. The B/M ratio effect was first documented by Stattman (1980, as cited by Hernandez, 2006), he discovered that average return has positive relation with B/M ratio in U.S. market. Fama and French (1992) show a powerful relationship between average return and book-to-market value and argue that book-to-market still has a stronger role in average return, after controlling size and book-to-market effects, beta has no relation with average return. In another words, firms with high ratios of book value to market value have higher average returns than firms with low book-to-market ratio, book-to-market ratio has capability to predict future returns. This result has been examined by Rosenberg, Reid and Lanstein (1984), Lakonishok et al. (1994). Past valuation errors could exist in return of book-to-market strategy, Lakonishok, Shleifer and Vishny (1994) argue that investors are rely much on past performance trends, overly optimistic and pessimistic expectations by market actual earnings news.

Another accounting anomaly is the P/E ratio anomaly. Nicholson (1960, as cited by Dimson, 1988) published first study about the relation between P/E and total returns, showing that low P/E stocks provide higher return than average stock. This result has been examined by Basu (1977, as cited by Das, 1993) in NYSE stocks. Is the P/E ratio effect another anomaly? Das (1993) says that most low P/E firms are small firms and we can treat P/E effect as a proxy for the size effect. Banz and Breen (1986, as cited by Das, 1993) support the opposite views.

3.4 Behavioral Finance

3.4.1 Introduction of Behavioral Finance

Many of the behavioural concepts and central role were found by Kahneman, Slovic and Tversky in 1979 (as cited by Shefrin, 2002). Behavioral finance first appears and was discussed in Shiller (1989) with an explanation for what observers label excessive stock price volatility (Ineichen, 2007). The key hypothesis of behavioural finance is that “additional variables capturing investor psychology are important to explain movements in the stock market” (Külpmann, 2004).Olsen (1998, p.87) wrotes: "Behavioral finance is focused on the application of psychological and economic principles for the improvement of financial decision making." Shleifer (2000, p.23, as cited by Schlichting, 2008) defines that “at the most general level, behavioural finance is the study of human fallibility in competitive markets.”

Oehler (2000, as cited by Schlichting, 2008) says that in the stock market, behavioural finance indicates psychological and sociological impacts on the stock market in order to explain stock market phenomena-so called anomalies. Similarly, Reilly & Brown (2003, p:196) states that "behavioral finance is concerned with the analysis of various psychological traits of individuals, and how these traits affect how they act as investors, analysts, and portfolio managers." They noted on page 202 that "the major contributions of behavioral finance are both explanations for some of the anomalies discovered by prior academic research, and opportunities to derive abnormal rates of return by acting upon some of the deeply ingrained biases of investors." In contrast to capital market theories, behavioral finance does not assume the market participants to be rational bur irrational (Avramov and Tarun, 2006, as cited by Schlichting, 2008, p: 28).

Behavioral finance tries to understand and explain how the effects of emotion, uncertainty and irrational thinking might influence investors when they make decisions. Such as Malkie, (1973, as cited by Thomas, 2006/2007) proves that people follow their emotions and hearts rather than their reasons and minds when they make decisions. Furthermore, behavioral finance attempts to explain the relationship between stock market efficiency and stock market anomalies. Additionally, behavioral finance tries to find out how to avoid these flaws in decision making (Shiller, 1989).

3.4.2 Arguments of Behavioral Finance

Shefrin (2002) argues that behavioral finance is dependent on the behavior of practitioners. Ritter (2003) stated one of the major problems of behavioral finance regards what bias will be chosen for emphasis, "one can predict either underreaction or overreaction." This problem can be called “model dredging.” Burton Malkiel (2003) offer historical evidence to prove that people follow their emotions and hearts rather than their reason and minds in order to argued that economists’ portraits of people making rational financial decisions based on sufficient market information are simply a non-existent ideal. Thus, arguments of behavioural finance can be divided into two parts. The first part is that irrational investors are not sufficient to cover capital markets inefficient; the second part is that in practice prices do not match intrinsic value. If prices match intrinsic value, then there are no easy profit opportunities. However, the reverse is not necessarily true, if arbitrage activity is limited, then absence of abnormal profit opportunities does not imply that market is efficient.

3.4.3 Behavioral biases

Behavioral biases largely affect investors decision making, some theory papers tried to find out what types could be used to representative behavioral biases. Such as Rayo and Becker (2005) explore types of pressures could produce past payoff and social reference point effects, or Samuelson and Swinkels (2006) find conditions for optimal choice set effects.

To understand various behavioural biases, we should know that heuristics play is an important role to construct investor decisions. “Heuristics are simple, efficient rules of thumb which have been proposed to explain how people make decisions, come to judgments and solve problems, typically when facing complex problems or incomplete information. These rules work well under most circumstances, but in certain cases lead to systematic cognitive biases” (Kahneman, as cited by Parikh, 2009 p.118). Shefrin (2002) summarizes four statements that could be used to define heuristic-driven bias: representativeness, overconfidence, emotion and cognition, conservatism, and aversion to ambiguity. Representativeness means judgments based on stereotypes may lead to misleading. Overconfidence indicates that when people are overconfident, they set their high guess too low and low guess too high. Error of cognitive is that errors stem from the way that people think. The basic idea behind aversion to ambiguity is that people prefer the familiar to the unfamiliar.

Bodie, Kane and Marcus (2009) conclude behavioral biases from four aspects: framing, mental accounting, regret avoidance and prospect theory. Decisions are affected by how choices are framed, and individual investors are risk averse in order to gain. Mental accounting is a specific form of framing to separate certain decisions and help explain momentum in stock prices. Statman (1997) argues that mental accounting is consistent with irrational investment preference. Simply speaking, this means that investors prefer sell stocks with gains rather than losses. Regret avoidance shows that investors will have more regret when they make wrong or bad decisions. But De Bondt and Thaler (1985) supplies that such regret avoidance is consistent with both size and book-to-market effect. High book-to-market firms are able to depressed stock prices. Considering with mental accounting, if investors focus on gains or losses rather than a whole portfolio, they would become more risk averse and concern more about recent poor performance. Prospect theory describes rational risk-averse investors. Higher wealth gives higher satisfaction lead to risk aversion rises, when profit become losses investors should be risk seeking rather than risk averse. Finally, Bodie, Kane and Marcus (2009) point out behavioral biases do not relate to stock prices if rational arbitrageurs could fully expose the mistakes of behavioral investors.

3.5 Irrational exuberance

3.5.1 Background of irrational exuberance

Shiller published one book which is called “Irrational exuberance” in 2000, introduced the background and definition of irrational exuberance. He mentioned that the term of irrational exuberance is derived from one of speeches which entitled as “The challenge of central banking in a democratic society” to describe the dot-com bubble of Alan Greenspan, chairman of the Federal Reserve Board in Washington at December 5, 1996. After this speech, the world stock market has fallen sharply, and then the term irrational exuberance became famous. Actually, the term irrational exuberance is more famous for its impact rather than its definition. Until now, irrational exuberance is often used describe a stronger speculative fervor. Afterward, irrational exuberance is used as a catch phrase for overvalued markets. In this book, he also point out in irrational exuberance, if markets deviate equilibrium, investors will obtain biased information on the level of asset prices in short term, leading to inefficiency.

3.5.2 Explanation of irrational exuberance

Külpmann (2004, p.78) says “Irrational exuberance was a phrase which caught investors”. In his book, he also mentioned that from 2000 “irrational exuberance” came out with stock market boom, at the same time stock market has been overvalued. “ Shiller concludes that society cannot be protected from the effects of waves of irrational exuberance or irrational pessimism, emotional reactions that are part of the human condition.” (Shiller, 2003, as cited by Gansauge, 2004, p.12)

3.5.3 Irrational optimism

Coxe (2006) says when people’s actions are unbalanced or ridiculous and have successful or positive outcomes, it can be irrational optimism. Most of time, we avoid to considering a “Plan B”, it is used to prepare for failure and avoiding irrational optimism. The second problem of irrational optimism is that people usually make negative associations to achieve goals. If you have a solid plan to achieve goals and prepare alternative plans for uncertainty, your goals will not be irrationally optimistic. Furthermore, he point out that we learn some irrational optimism from our parents and friends.

Even the world stock market suffered the worst bear markets in history, many investors remain irrationally optimistic about future long-run stock returns. Those investors believe that the stock value always keep up with inflation over long period, short time investment bring about bubble. In Dimson, Marsh and Staunton’s paper in 2004, they say that companies offer a higher expected return to reward for risks of equities, but many investors overestimate the returns and underestimate the risks of investing in stocks, especially when they held stocks as long as they want.

3.5.4 Irrational pessimism

Irrational pessimism has more potential to research than irrational exuberance. That is because “It is a far easier proposition to convince an investor in a booming market to reduce some risk in exchange for less upside than to convince traumatized investors to go against their instincts and remain in the markets, … But if they do not, the client may miss most of the upside and will likely have a terrible return. “ (Hurley and Fuller, 2001, p.34)

Taylor and Woodford (1999) give one equation for the equity premium in a model with serially uncorrelated consumption growth:


where g is the mean growth rate of consumption, is discount factor close to or even greater than one, is coefficient, is mean riskless interest rate. This equation is evolved from risk free rate puzzle equation, it is investors’ irrational expectation of the equity premium. This model illustrates irrational pessimism between investors ( <) can lower the average risk free rate and increase the equity premium.

3.6 Previous Research Review of Seasonal Effects

There are a load of researches examine the phenomenon of seasonality, especially for its most famous example: the day of the week effect and January effect. In this section, the previous literature in this filed will be reviewed and discussed.

3.6.1 The Day of the Week Effect

The day-of-the-week patterns in stock returns challenges to the efficient market hypothesis. “For seasonal of calendar effects, the distribution of common stock returns is not identical for all days of the week, such an effect was called ‘day-of-the-week’ or ‘weekend’ or ‘Monday’ effect, depending on the days that are significant” (Tsangarakis, 2007, p.1447). Weekend effect or the day-of-week effect is an anomaly which is much more significant at holiday weekends or any day in the week. As it can be seen that the day-of-week effect is stock returns link to a particular day in the week. While Fama (1965) report that variance of Monday is 20% greater than other days. Cross (1973) was first found negative returns during weekend for most common stocks and other financial assets. Since French (1980) observed that stock returns are high on the last trading day and low in first trading day in the week, the day-of-week effect or weekend effect became popular for financial study. Explanations for this anomaly come from three aspects: market settlement procedures (see Gibbons and Hess, 1981), measurement errors in stock prices (see Keim and Stambaugh, 1984) and impact from information has been published during the week (see Damodaran, 1989).

The day-of-week effect has been gradually examined in US stock markets, and the Monday effect and positive return on Friday are the most common seasonalities. Most of studies tested the Standard & Poor’s Composite Index and the Dow Jones Industrial Index in US, such as Gibbons and Hess (1981) and Keim and Stambaugh (1984). Wang et al. (1997) uses a long time period from 1962 to 1993 to show the Monday effect in the last two weeks of a month. Also for US market, Peiro′ (1994) confirms the Monday’s positive returns. Other studies analyze the effect with volatility of returns (see Ho and Cheung, 1994, and Choudhry, 2000).According to Singal (2006), the weekend effect is defined as the return of Friday minus the following Monday’s return will equal to zero.

After US stock market, there are several studies to examine weekend effects in the stock market outside US, such as Brusa et al. (2003) report that weekend effect exist in stock market of Brazil, France and Japan, a reverse weekend effect pattern appears in US stock market, but no weekend effect in UK, Hong Kong, Australia stock market or Indian stock market (Raj and Kumari, 2006). For less-developed markets, more recent study about Day-of-the-Week effect shows return of Mondays and Tuesdays is low than Wednesdays and Fridays, there is no evidence of this effect in Asia stock market (Hui, 2005). Negative return on Monday in Singapore was examined by Tan and Tat (1998). Negative returns on Tuesday were found by Choudhry (200) for India, South Korea, Taiwan, Thailand for Balaban (1995) and Turkey by Oguzsoy and Guven (2003). Daily return of Thursday is higher than Friday in Chinese stock market (Mookherjee and Yu, 1999a), excluding Chinese stock markets (Cai et al. 2006)

The reason of day-of-week effect is considered as one of the most interesting market anomalies and is continually studied until now, because significant day-of-week effect would very helpful for developing profitable trading strategies (Syed and Sadorsky, 2006). This means investors could buy stocks at abnormally low return days and sell them on abnormally high returns. However, tests by Connolly (1989) point out the results of the day-of-week effect are affected by adjustments for sample size, heteroskedasticity, autocorrelation, and leptokurtosis. He observed that the day-of-week effect has disappeared by 1975 in US stock markets.

Apart from explanations of the day-of-week effect, arguments from Wang et al. (1997) point out the day-of-week effect can be considered as an interaction of other seasonalities such as January effect. And Chen et al (2001) try to use spillover effect to explain day-of-week phenomenon. Furthermore, some analysts argued that weekend effect is result from bad news which has been published after the close of trading on Friday and during the weekend, rational investors will consider the impact of bad news over weekend into price before weekend, this helps to eliminate the weekend effect (Damodaran, 2002). As we consider the day-of-week effect a puzzle and despite different theories, this puzzle has not been satisfactorily solved yet.

Chen et al. (2001) examined the day-of-the-week regularity in Chinese stock market. They collected the data from Shanghai A share Index, B share Index and Shenzhen A share Index and B share Index from different beginning time to 31 December 1997. They test the null hypothesis as returns of each day in the week is equal to zero by a generalized autoregressive conditionally heteroskedasticity (GARCH) model. According to analysis, they found that Tuesday anomaly appears after 1995 and they suggest day-of-the-week anomaly in China is highly depends on the estimation method and sample period. Finally, they point out that if taking transaction costs into account, arbitrage profits from anomalies are small which is consistent with the efficient market hypothesis.

Another paper from Ogunc (2009) investigated the seasonal regularities, particularly the day-of-the-week effect and January effect in the Chinese stock market. They chose the same data types as Chen et al.’s paper from 1990 to 2006, but divide data into limit period and non-limit period to against the entire period. In this research, they found some January and weekend effects exist in this market and Shanghai A index is a high volatility market.

3.6.2 January Effect

Wachtel (1942) first documented reference for a January effect in stock returns. 34 years later, January effect is found by Rozeff and Kinney (1976) as stock return of US markets are higher in January than other months during 1904 to 1974. Since 1976, abnormally large returns in January on common stocks became the one of the most intriguing issues. In 1983, Roll, Keim and Reinganum find a positive relation between price increase in January and capital losses in December in small firms. Also Kato and Shallheim (1985, as cited by Elton et al. 2010, p.404) confirm that abnormal return in January and strong relationship between return and size. Associated negative return in December as considered of small company effect which is based on ‘tax-loss selling’ hypothesis, and have been researched over decades (Lo and McKinlay, 1997). Gultekin (1983) find return patterns of 17 countries are higher in January than in non-January months. Obviously, January seasonality is unstable with market efficiency, in the efficient market investors should readjust their portfolios in order to eliminate abnormal returns from January.

Recently, some professionals observed January effect for a wide range of developed stock market as well. Evidence shows that January effect was dramatically reduced during 1987 to 1996 period and it does not exist for other assets, apart from stocks Levy (2002). In the paper of Chen et al. (2010), they examine the possible January effect on Asian stock market price and find that January effect has largely disappeared from Asian markets. Also, there is no evidence of January effect or monthly seasonality in ALL Gold Index on the Johannesburg stock market from 1987 to 1997 (Coutts and Sheikh, 2000), or Hong Kong market (Cheung and Coutts, 1999). Even July effect in Kuwait stock market rather than January effect (Al-Saad and Moosa, 2005).

Girardin and Liu (2004) use a time series modelling to test the seasonal anomalies in Chinese stock markets. They only covered the A share markets in Shanghai and Shenzhen from 1993 to 2003 period with both monthly and quarterly data. According to their findings, they suggest that using unobserved-components models help them to find out there is no evidence to show changes in seasonal patterns, but they do found positive unchanged June effect and a negative December effect since 1993 in Chinese stock market.

Some explanations can be used for January effect. January effect may be caused by transaction costs (Stoll and Whaley, 1983), tax-loss selling effects (Ritter, 1988), risk premium or expected returns (Change and Pinegar, 1989 and Kramer, 1994), even by transactions of cash or liquidity (Ogden, 1990). Moreover, Kohers and Kohli (1992) suggest January effect is caused by business cycle, and Ligon (1997) says that January effects accompany with high trading volume and low interest rates. Tax loss selling hypothesis is the most widely acceptable explanation. This hypothesis suggests that investors taking advantage via sell decreased price stocks at the end of year in order to realize capital losses, this push the stock prices go down. In the following year, stock prices will bounce up as the selling pressure is relieved. This phenomenon is usually appears in small firms due to its volatility are more than large firms. Branche (1977), Dyl (1977) and Roll (1983) supported this hypothesis, whereas Jones et al. (1987) had inconsistent results with this hypothesis. The second important explanation for the January effect is the information hypothesis, which means firms always release their important accounting information at the year-end lead to uncertainly depress on stock prices. After the uncertainty has been solved, the stock prices rise again. According to the information hypothesis, firms which have a non-December fiscal year end should earn high returns in the month following their fiscal year end, but not in January (Kim, 2006). However, Reinganum and Gangopadhyay (1991) do not have the consistent results with this hypothesis. The third explanation is microstructure explanation for January effect. Keim (1983) finds the bid-ask spread has impact on tendency of closing prices. This phenomenon usually exist in small and low-priced stocks, as Keim argues that large returns of small firms in the first several trading days are contribute to the trading pattern somehow.

All previous literatures and expiations have not explained the January effect completely over the years. Because if the trend of January effect increase or decline (Gu and Simon, 2003), even it disappear in a market, then there are should be some new or unidentified factors may exist to cause the change of abnormal return in January. Thus, further research need to reveal any trend of January anomalies interact with other market anomalies.

Need help with your literature review?

Our qualified researchers are here to help. Click on the button below to find out more:

Literature Review Service

Related Content

In addition to the example literature review above we also have a range of free study materials to help you with your own dissertation: