Stock market efficiency can be defined as the price at which buying or selling a stock would give a "fair" return in relation to the associated risk, after transaction costs have been taken into account. For a market to be efficient information has to be readily available to investors to make informed choices.
However, inside information that could result in better returns than the market is usually withheld from other investors and this is one thing that can create market inefficiency. If one investor has found a profitable trading opportunity and shares this with other traders the advantage is quickly wiped out. A new company can also be inefficient because of a lack of information on the new stocks. Securities laws may be in place that does not require companies to share information and this creates market inefficiency in emerging markets. No matter how much information is provided it will never be enough to predict what will happen in the stock market in the long run but it may be used to predict short term happenings.
Technical analysts use many different methods to try and predict the movements in the stock market in an attempt to gain better returns than the market. One news report suggests that astrologers use movements in the heavens to predict the stock market or to find a mood indicator of the stock market (Can academics predict the market?, 2001). Another method is based on the Fibonacci sequence, using different percentage levels to predict when the best time to buy or sell a share will be. Perhaps the most famous investment based on predictions was the Long Term Capital Management hedge fund, which lost 2.3 billion US dollars in 1999 when Russian markets plummeted because of the devaluation of its currency. This incident presents evidence that the stock market is highly efficient and therefore it is incredibly difficult to beat the market using predictions.
In contrast, there are studies that show stock market anomalies and contradict market efficiency. Persistent stock market anomalies could produce future outperformance of the market but there is no guarantee for this. Market anomalies include the weekend effect, holiday effect, January effect, trading month effect and the Friday the 13th effect. It is possible that these calendar anomalies could be exploited to form an investment strategy and this is what will be discussed in this assignment. There are also technical and fundamental anomalies that could bring better returns than the market.
Persistent stock market anomalies and whether it is possible to harness anomalies as part of an investment strategy.
Trading Rules
Technical analysis is one of the ways that investors try to predict future stock prices from stock history performances. Trading rules are one form of technical analysis. Mills (1992) was the first person to come up with the principle of security price trading rules. In recent years, two trading rules normally employed are Moving Average Oscillator and Trading Range Break-out. 'Buy' or 'sell' signals are generated by comparing short-run moving average and long-run average. The formula is given below:
If short run MA is greater than long run, stock can be bought, in contrast, a sell signal is given out if short run MA is less than the long run MA.
Trading Range Break-out sends signals depending on resistance and support levels. The expressions are as follows:
Rest (m) =max (Pt-1…….Pt-m) Supt (m) =min (Pt-1……Pt-m)
If price exceeds the resistance level then investors will sell, if it exceeds the support level then they will buy. The above trading rules appear to offer potentially profitable trading strategies; however, they are just for a short time period and not suitable for the long-term. Therefore, as the quotation said "No one can consistently predict either the direction of the stock market or the relative attractiveness of individual stocks and thus no one can consistently obtain better overall returns than the market".
Weekend Effect
Weekend effect exists where stock returns are significantly lower between Friday's close and Monday's close. Statistical evidence from FT 30 states the weekend effect has existed in the London International Stock Exchange from 1 July 1935 to 31 December 1994, but was not persistent (Arsad & Coutts, 1996). In contrast, the results of Kim (1988) reported there is persistent weekend effect for the same index.
In table 1 (see Appendix), there are twelve five year sub-samples showing that the mean return for Mondays is negative and there is a significant weekend effect for six of the twelve sub-samples, as well as mean returns being positive and largest on Wednesday and Friday. If there is positive information flow then stock returns will be positive, and there will be negative returns in a bad news environment. Arsad and Coutts found very strong evidence for the existence of the weekend effect in a bad news environment. As table 2 indicates (see Appendix) there is a significant lower return on Mondays than other days in a bad news environment, but no difference between other days with good news. Therefore, they suggested the weekend effect no longer exists in a good news environment.
Through the results, these twelve sub-samples not only allow for testing whether there is a consistent weekend effect, they also represent average stock prices caused by news information. However, investors are able to assess risk and return based on news, but it does not mean they have perfect power of prediction, it is because news is random, therefore no one can predict what will happen during the weekend. Indeed, they all have the same expectation because of the free availability of information, thus no one can obtain better overall returns than the market.
From other empirical research in the All Gold Index on the Johannesburg Stock Exchange (JSE), it contrasts to the overwhelming international evidence documented for developed and emerging markets. The result appears to be no weekend effect through 5 January 1987 and 15 May 1997(Arsad & Coutts, 1996). This would support the quotation as the absence of anomalies suggests that the market still obtains better overall returns than any individual.
Table 3 shows the mean Monday return remains negative for two of the three sub-periods. Moreover, there is a significant negative mean return on Tuesday for the second sub-sample between 1990 and 1993. It might be due to particular market behaviour, for example, dividends are generally paid on Mondays in the UK.
To the contrary, Linn and Lockwood suggested there is a weekend effect and it is now accepted by most of financial economists.
Holiday Effect
Holiday effect is considered by many as the largest, oldest and the most important seasonal regularity. It can be classified as pre-holiday, post-holiday, and non-holiday. (Brockman, P and Michayluk, D. 1997)
Ariel (1990) through analysis of US data reports that the trading day before holiday, on average, reveals higher positive returns .The result is supported by Kim and Park (1994) for the US, Japanese and UK markets. Each country has different holidays and arrangements. Therefore, these results can act as international evidence that the holiday effect is a world-wide phenomenon. Table 4 (see Appendix) shows the returns on pre-holiday are higher than non-pre-holiday. In addition, French (1980)'s 'closed-market' assumption states the return following a holiday will be lower than the return for non-holiday. Table 5 and table 6, however, overthrow this assumption, which demonstrates overwhelmingly higher returns before and after holidays (Arsad, 2000).
This evidence influences the emerging market and can be proved in a case, which researches the exits of security price anomalies in the Athens Stock Exchange General Index, over ten years from 14 Oct 1986 to14 Aug 1996. Three major industry indices are considered: Banking, Insurance and Leasing. Holiday effect is the most significant anomaly in the Athens Stock Market. Hence, seasonality anomalies cannot give profitable returns using trading strategies once transaction costs have been considered. This is consistent with the notion of market efficiency; there is no strategy existing that yields abnormal returns (Coutts, 2000) and also supports the quotation.
Mills and Coutts (1995) used UK market data for analysis which suggests that if holiday effect is persistent, the cost of implementing 'trading rules' may be prohibitive. In addition, the round trip transaction costs would render any investment strategy unprofitable. Therefore, it can be concluded that there is no strategy that exists which will consistently yield abnormal returns. This finding also proves the quotation 'no one can consistently obtain better overall returns than the market'.
January Effect
The January effect is present when there is the tendency of stock prices to rise in January relative to December. Economist conducted empirical studies to find out whether the January effect was actually a measure of predicting the market stock value.
The January effect was seen to be present in the international economy based on the research of Reinganum (1983) hypotheses of the 'Tax loss selling' which argues that stock is sold in December to realise losses in order to offset gains accumulated during the year. Investors sell stock that has declined in price during the year to reduce their taxes, thereby causing stock to fall in price. (See Table 7 in Appendix).
Mills and Coutts (1995) reported similar findings and also tried to find ways that this could improve the portfolio performance. They also state that even if the anomalies are persistent, implementing trading rules could be forbidden due to it being expensive.
The January effect is present but as time goes by it become more persistent within the emerging markets sector. Research from Ariel (1990) showed that mean returns were positive in January however this was not high or persistent. These two statements seem to contradict themselves. This is due to the effect that institutional reform has on the January effect.
Mills and Coutts (1995) once again found evidence that supported this but stated that more research was needed and it should be questioned.
Evidence from previous analysis shows positive January effect, even though present, may have been faded into non existence and has a number of unanswered questions. On the other hand it suggests anomalies would tend to disappear and have no value once investors become aware of how to predict the market performance. This backs up Malkiel's suggestion that profitable trading opportunities are wiped out quickly.
Other Anomalies
Other anomalies include the trading month and Friday the 13th, which are both calendar anomalies.
The trading month effect states the returns are only positive for the days in the first half of the month and the returns for the days in the second half of the month tend to zero (Ariel, 1987). Ariel's study used data for the US. Data for the UK, Japan and Australia were tested in 1989. From these studies it was shown that no significant monthly effect was present, except from in Australia. Other studies contradict this and show that there is significant trading month effect present in the UK economy. This inconsistency proves that it is difficult to consistently predict the direction of the stock market and different methods of analysis produce different results for the same sets of data.
The Friday the 13th effect is one of the least documented anomalies. This effect is where returns on Fridays which fall on the 13th have lower returns than all other Fridays. This effect is clearly based on superstition about unlucky days and most previous studies have stated that there is limited or no evidence of a Friday the 13th effect.
Fundamental anomalies are another set of anomalies that are based on investing in the value of companies. Investors can overestimate or underestimate the value and growth of companies and this means that returns could be consistently higher than those predicted by the market. A study of the value anomaly showed that even after transaction costs were taken into account, investors could outperform the market, (The Effects of Rebalancing on Size and Book-to-Market Ratio Portfolio Returns, 1995). This contradicts Malkiel's statement which means it may be possible to harness anomalies as part of an investment strategy.
Investment strategy
Investment strategy is a set of rules and process used to invest in the stock market. It mainly focuses on risk-return trade-off for potential investors. Some investors would prefer higher returns by investing in more risky assets than lower rate returns with low risk investment. Therefore it is essential to estimate risk and return on investment before making any investment decisions for meeting long-term goals.
According to the market efficient theory, stock prices reflect all the information that is available to the investors. It means stock return fluctuations depend upon good or bad news environments. In fact, news is random and instantaneous, so it is not easy to predict positive or negative information flow. This makes it difficult to harness anomalies as part of an investment strategy.
The tendency with the weekend effect is for stock returns to be lower on the day before and after the weekend than other weekdays, despite changing environments. Investors should recognise the weekend effect and avoid selling stock on Mondays, and there would be an advantage from buying on Mondays and selling on Fridays. This information is useful when deciding on an investment strategy.
According to the regularity phenomenon, stock returns tend to be higher before and after holidays. Therefore, holiday effect can be used as part of an investment strategy. If a decision is made to invest after a holiday, investors need to collect information that happens during the holiday. If all information gives positive signals then the first trading day post-holiday should make higher returns. However, it is difficult to harness this stock market anomaly as part of an investment strategy because previous research has given mixed results as to the returns before and after holidays.
The January effect has suggested that we can predict what will happen in the stock market in the short term but cannot predict the long term happenings. In an article known as 'data says do not ignore the January effect' it was studied that for the years 1929-2009, 1969-2009 and 1989-2009 the January effect was found to be consistent. Only 34% of returns over 81 years were down. These statistics show that although investors can predict the stock market there are times when events such as a recession or war may cause the stock market to act in a negative way making it not such a good investment strategy for investors unless there is information available that can predict the negative effects that the stock market may have.
Conclusion
Overall, it would be risky to use market anomalies to form an investment strategy. This is firstly because the research tends to fluctuate in the results. For example, the trading month is found to exist only in Australia in one study but other studies show that it is evident in the US and UK markets.
It is also because good and bad news indicators are hard to predict such as war or a poor economy. An investor would have to be quicker than the rest in responding to signals and this is difficult to do.
In addition to this, each investor has his own risk-return trade off so individual attributes would be calculated differently by each investor. Shares which send off negative signals may have to offer higher returns for investors that are more sensitive to risk. Those who are less sensitive to risk may then see this stock as a bargain and automatically gain better returns than their fellow investors.
If an investor thought he had a substantial strategy to beat the market, they would have to keep the information to themselves because it may stop working when the strategy becomes common knowledge. This is one of the points argued by Malkiel and is one of the reasons why long term investment strategy does not seem possible.
The other problem with keeping an investment strategy secret is that an investor could be accused of using insider information to cheat in order to gain better returns than the market. It is illegal to "become an insider" in order to make larger profits than normal.
There are too many deciding factors in creating an investment strategy to be able to predict a way of consistently gaining better returns than the market. Therefore, harnessing market anomalies as part of an investment strategy is not likely to produce better returns than the market.
References
Paper Resources
Ariel, R.A (1990) High stock returns before holidays: existence and evidence on possible cause, Journal of Finance, 45, 1611-1626
Arsad, Z and Coutts, J. Andrew (1997). Security Price Anomalies in The London International Stock Exchange: a 60 year perspective. Department of Statistics, Herriot Watt University, Applied Financial Economics 1997,7,455-464, Tayler & Franes Ltd.
Arsad, Z and Coutts, J. Andrew (2000). Security Price Anomalies in The London International Stock Exchange: a 60 year perspective. Department of Statistics, Herriot Watt University, Applied Financial Economics 2000,10,561-571, Tayler & Franes Ltd.
Brockman, P and Michayluk,D et al (1997). The Holiday Anomaly: An Investigation of Firm Size versus Share Price Effects. http://www.questia.com/googleScholar.qst;jsessionid=LGwKkyJjFXhvyv1TChtkrLZK20VTDnXm2yF7dW8zSGCdC2QgwSZv!-2131880813!-867442119?docId=5001527379 [Accessed25/02/2010]
Coutts, J Andrew & Kaplanidis, Christos & Roberts, Jennifer, (2000). Security Price Anomalies in an Emerging Market: The Case of the Athens Stock Exchange," Applied Financial Economics. Taylor and Francis Journals, vol. 10(5), pages 561-71, October.
French, K. R.(1980) Stock returns and the weekend effect, Journal of financial Economics, 8,55-69
Keith Redhead (2003) - Introducing Investments: A personal finance approach, pp.204-232. Pearson Education Limited
Kim, C. W and Park, K.(1994) Holiday effects and stock returns: further evidence, Journal of Financial and Quantitative Analysis , 29, 145-157
Kim. S. W. (1988) - Capitalising on the weekend effect. Journal of Portfolio Management, pp. 14, 59-63.
Linn, S. C. and Lockwood, L. J. (1988) - Short term stock price pattens: NYSE, AMEX, OTC. Journal of Portfolio Management, pp. 15, 30-34.
Mills, T. C. and Coutts, J. A.(1995) Calendar effects in the London stock exchange FT-SE indices, European Journal of Finance, 1.79-93
Patrick Dennis, Steven B. Perfect, Karl N. Snow, and Kenneth W. Wiles, "The Effects of Rebalancing on Size and Book-to-Market Ratio Portfolio Returns," Financial Analysts Journal, May-June 1995.
Z. Arsad & J. A. Coutts (1996) - Applied Economics Letters, Volume 3, pp. 797-801. Routledge, Taylor & Francis Group.
Online Resources
Ryan, O. (2001). Can academics predict the market.
.http://news.bbc.co.uk/1/hi/business/1425639.stm
[Accessed09/03/10]
Lounsbury, J (2010). Markets: The January Effect.
http://seekingalpha.com/article/186994-markets-the-january-effect
[Accessed27/02/10]
Moosa, I.(2007). The Vanishing January Effect.
http://www.eurojournals.com/IRJFE%20ISSUE%207%20moosa.pdf
[Accessed27/02/10]
Investorhome(1998). Fundamental Anomalies.
http://www.investorhome.com/anomfun.htm
[Accessed09/03/10]
