Introduction
Although market efficiency has been extensively researched by academics throughout the twentieth century its history dates back to Bachelier (1900), it was not popularised however until Eugene Fama (1970) published his paper on efficient capital markets in the Seventies and because of this work, is generally considered to be the father of the Efficient Market Hypothesis (EMH). In more recent years the EMH has been interrogated by many academics and there is an abundance of research and empirical studies on the subject, Dimson and Mussavian (1998:91) wrote 'The concept of efficiency is central to finance' but in more recent years, this theory has been challenged by many academics. The focus of this paper will be to critically analyse the validity of the semi strong form of market efficiency more than forty years after Fama published his original work.
The EMH and a critical analysis of the semi-strong form
It is important that we first define what is meant by market efficiency, market efficiency broadly means that the prices of securities always reflect all available information. If market efficiency is to be believed then no group or individual is able to make an abnormal return on investment because they are acting on information available to all investors and as new information emerges investors will re-evaluate the price of the security, 'Whenever investors find that the expected return is not equal to the required return (Å™ = r), a market price adjustment occurs' (Gitman, 2009:343).
Whilst there are three different forms of market efficiency, we are interested in that of the Semi-strong form and Fama (1970:383) describes this as 'tests, in which the concern is whether prices efficiently adjust to other information that is obviously publicly available', examples of public information could be mergers, profit warnings and other events. If the semi strong form of EMH exists then the price of a security would quickly and accurately change to represent the current market value following the release of any new information to the market. For the purposes of this paper I will be dividing semi-strong form efficiency into three different categories which will be event studies, cross sectional predictability of returns and finally, calendar effects.
Event studies are interested in how a company's stock adjusts to unpredictable events such as stock splits, mergers, takeover bids and other events that can be traced back to a particular moment in time.
An example of the rapid change to daily returns could be demonstrated by looking at the stock price of Rolls Royce and the effect of an unforeseeable event.
On at 03:45 on November 4th 2010 a Quantas aircraft was forced to make an emergency landing in Singapore because of a malfunction to one of its Rolls Royce engines (Quantas, 2010), as the LSE opened Rolls Royce's stock dropped sharply as shown in appendix I. This price adjustment shows the market reacting to an unforeseeable event and the equilibrium price has significantly lowered to reflect the market value demonstrating an efficient market in its most basic form.
Another area of event studies is the changes in a stock's price to a dividend announcement, Pettit studies the affect these announcements have and concludes that his study 'lends support to the proposition that the market is reasonably efficient on both a monthly and daily basis' (1972:1007) although recognises that immediately prior to the announcement there is a slight anticipation effect, most likely caused by insider information. Fama et al (1969) also investigated the effect of stock splits on the value of shares and found that their results supported the efficient market hypothesis stating that 'the information implications of a split are fully reflected in the price of a share at least by the end of the split month but most probably almost immediately after the announcement date' (Fama et al. 1969:20).
Since the publication of Fama's original work there have been a number of papers detailing anomalous evidence in event studies, Ball and Brown (1968) were the pioneers of the concept of post announcement drift which shows the positive or negative drift of a securities price after an announcement of earnings, this research was also supplemented by Ball (1978) when conducting further research into earnings and dividends. The phenomenon of post announcement drift has been demonstrated by a number of academics to date but there is speculation as to whether the drift is caused by market inefficiency, or by a failure in the pricing model.
Another area of proposed market inefficiency is the negative long run on initial public offerings (IPO's) as documented by Ritter (1991) and further investigated by Loughran and Ritter (1995), these studies found that if an investor were to invest in an IPO at the end of the first day of public trading and to hold them for three years it would provide a return to the investor less than the return from investing in a group of similar firms that had not recently floated, in other words Ritter has demonstrated that the IPO's studied generated a substantial underperformance when compared to similar securities to a period of up to three years. Friesen and Swift conduct a similar study but also investigate the abnormal initial returns giving two possible explanations:
Some suggest the large initial returns result from initial undervaluation, in which case the first day return simply reflects an adjustment from the
offer price to the fundamental value. Others suggest that initial returns
reflect investor overreaction to new information on the first day of trading.
(Friesen and Swift, 2009:1285)
The perceived breach of market inefficiency with IPO's again calls into question whether this is but another failure of the pricing model, or in fact the market not acting efficiently, an issue known as the joint hypothesis problem.
Cross Sectional Predictability of Returns looks at the patterns that emerge when focusing on one dimension of a firm's character, for example, the size effect and the book-to-market effect. The size effect is concerned with the apparent higher than average risk adjusted returns from smaller companies, Rolf Banz was the first to investigate the size effect and concludes from his research that 'On average, small NYSE firms have had significantly larger risk adjusted returns than large NYSE firms over a forty year period.' (Banz 1981:16). Similarly, Basu (1977) published work showing that companies with low price to earnings ratios have, on average, higher risk adjusted returns although acknowledges that trading costs and tax liabilities would usually hinder an investor from making abnormal profits. Another series of studies such as Kothari et al (1995) and Rosenberg et al (1985) describe the correlation between a securities book to market ratio and its future returns, the studies show a strong positive correlation suggesting that there is the potential for abnormal returns and a breach of the EMH.
The argument against this form of market inefficiency is that there is a higher risk factor involved when investing in small firms that might not be taken into account by the pricing model, in such cases the abnormal return might be considered a proxy for risk. The abnormal returns might also be considered compensation for problems such as higher trading fees, a larger bid/ask spread or perhaps informational asymmetry, all problems associated with trading smaller firms stocks.
Calendar Events investigate a pattern in stock returns that re-occur at observed points in time, for example, the January anomaly and the Weekend effect. The January anomaly was first studied by Rozeff and Kinney (1976) who found that there was a higher mean return during January than in any other month throughout the year, this research brought about a number of different possible explanations to the anomaly including the tax selling hypothesis which suggested investors were more likely to sell underperforming stocks in January to decrease their capital gains tax payment although further research on the subject has suggested that 'tax-loss selling cannot explain the entire January seasonal effect. The small firms least likely to be sold for tax reasons (prior year 'winners') also exhibit large average January returns' (Reinganum, 1983:103). Another proposed explanation for the January anomaly is the Accounting information hypothesis which suggests that most firms have fiscal year ends in December causing uncertainty amongst investors and depressing the stock price, once the company's year end results are disclosed investors are more inclined to invest because of the availability of the accounting information, however, Reinganum and Gangopadhyay (1991) show the continuance of the January effect on smaller firms regardless of the end of their fiscal year.
The day of the week effect, first documented by Kenneth French suggests that it would be possible for investors to make abnormal returns by buying and selling securities in time with regular market movements, he noted that 'the return for Monday was negative and lower than the average return for any other day' (French, 1980:59), if the day of the week effect is observed then it would in theory be more profitable on average to sell stock on a Friday than it would be on a Monday, similarly it would be wise not to buy securities on a Friday, but instead to wait until Monday. Dimitris Kenourgios and Aristeidis Samitas also tested this hypothesis on securities traded in the Athens stock market and found that 'The day of the week effect patterns in return and volatility might enable investors to take advantage of relatively regular market shifts by designing and implementing trading strategies' (Kenourgios and Samitas, 2008:87)
There are a number of other studies that have been conducted on calendar events throughout the year including the Halloween effect as first documented by Bouman and Jacobsen (2002), or Krueger and Kennedy (1990) who suggest that investors can outperform the market in the United States by reacting to the outcome of the yearly Superbowl, many of these studies show a cyclical effect in market prices which should not be possible within an efficient market.
We must also consider the rationality of investors when looking at market efficiency, Fama (1970) recognised that for a market to be efficient, the investors within that market must be rational but there have been instances in the past where investors have acted in an irrational manner and a good example of this type of irrational investing would be the internet bubble. When the internet was popularised at the end of the 20th century there was a vast amount of speculative value added to any company with a '.com' at the end of its name, companies that had no technical affiliation were able to add '.com' to their name in order to achieve increases in their stock value, this speculation caused the stock prices of internet companies to rise dramatically and as prices continued to rise the amount of investors trying to profit from this emerging market increased thus inflating the price further, this type of behaviour associated with market bubbles shows 'the presence of speculative rather than fundamental reasons for investing' (Simon 2003:18) and demonstrates a clear example of market inefficiency.
Conclusion
Throughout this paper I have tried to provide literature that both supports and questions the semi strong form of EMH but it is important to note that the body of literature surrounding this topic is vast and it is difficult to draw any firm conclusions, it is also important that we consider two fundamental problems that arise with any tests of the EMH. Market efficiency must be tested alongside a pricing model which determines the expected market return, this model may not always account for such things as proxies for risk and this fundamental issue is called the Joint Hypothesis problem. Another matter to consider is that if a trading strategy which was able to predict market movements were created then all rational investors would be using it and therefore eliminate any potential abnormal returns, there is also the ex-ante versus ex-post problem with any trading strategy which states that just because an investment strategy may have resulted in abnormal returns in the past does not mean that using the same strategy in the future will definitely result in abnormal returns.
Fama returns to the subject in 1991 with a critique of how well the EMH has stood up after twenty years of scrutiny and says 'The cleanest evidence on market-efficiency comes from event studies … evidence tilts me toward the conclusion that prices adjust efficiently to firm-specific information.'(Fama 1991:1607) With Fama's comments in mind it is impossible to escape the amount of studies that have shown some form of inefficiency, especially with studies of calendar events that have withstood scrutiny over long periods of time.
Appendix
I)
http://www.londonstockexchange.com/exchange/prices-and-markets/stocks/summary/company-summary.html?fourWayKey=GB0032836487GBGBXSET1
