浏览： 日期：2020-06-10
One of the most renowned concepts of modern portfolio theory is the Efficient Market Hypothesis. According to the EMH,“capital markets are efficient in an informational sense. A capital market is considered to be efficient if, through trading activities, investors set the price of any particular security in a manner that impounds new information about that security in a direct way”(Financial Advisor Magazine, 2004).
This implies that a market is informationally efficient when current prices reflect all past and currently available information. Hence, the market price of a security can only be altered if new information, which by definition is not fully foreseeable, appears. Most importantly, the EMH presupposes that investors are rational and therefore value securities rationally. When new information comes out, investors evaluate its effect on current and future cash flows, and adjust the price of the stock to reflect the new information.
However, there are some challenges to the EMH which have been extensively documented. The two most broadly known and extensively studied are contrarian and momentum trading strategies, which presume that future returns are predictable on the basis of past returns. Momentum strategies buy stocks which have performed well in the past (winners) and sell stocks which have performed poorly in the past (losers), profiting from return continuation.
Contrarian trading strategies support the opposite method of buying past winners and selling past losers, thus generating profits from return reversals. Both these strategies, when properly carried out have been shown to generate exceptional returns. Contrarian strategies are found to be profitable at very short time horizons up to approximately one month and at long horizons of several years, whereas momentum strategies are found to be profitable at intermediate horizons of up to one year.
A momentum strategy is to buy past 1 to 12 months’ winners and to sell losers in the next 1 to 12 months. It is a straightforward trading strategy, according to which the portfolios are constructed based on cumulative return criterion over certain horizons. A “contrarian” strategy is accomplished in the same way, but winners are sold and losers are bought.
The amount of academic studies documenting the momentum strategy has been remarkable. The momentum strategy gained great importance in 1993 when a study carried out by Sheridan Titman at the University of Texas and Narasimha Jegadeesh, who was then at the University of Illinois was published.
According to Titman and Jegadeesh (Journal of Finance, 1993) “maintaining a long position in past strong performers and shorting past weak performers could earn investors abnormally large returns over a six-to 12-month horizon”. The momentum effect indicates that stocks with high returns over the past 3 to 12 months tend to outperform in the future. Historically, momentum strategies earned profits of about 1% per month over the following 12 months in the stock market (Titman and Jegadeesh, 1993) and in the foreign currency spot market (Okunev and White, 2003).
Several studies in the field of momentum strategies like the ones conducted by Jegadeesh (1990), Jegadeesh and Titman (1993, 2001, 2002), Chan, Jegadeesh and Lakonishok (1996), Rouwenhorst (1998), Grundy and Martin (2001) and Lewellen (2002) have provide evidence that by taking a long position in stocks with the strongest past performance and taking a short position in stocks with the poorest past performance generates excess returns over intermediate investment horizons (3-12 months).
These studies indicate that prices underreact to information and are mainly based on intermediate horizons. In their study, Jegadeesh and Titman show that stocks with higher returns over a particular intermediate investment horizon continue to perform better than firms with lower past returns in the same period. Using data from US stock exchanges, Conrad and Kaul (1998), and Hong et al. (1999, 2000) provide evidence supporting momentum strategies. Furthermore, Liu et al. (1999) have found comparable results in the UK.
Hon and Tonks (2003) expand the sample of Liu et al. to a total of 6600 stocks over the period January 1955 to December 1996. The results of Hon and Tonks are consistent with those of Liu et al. finding significantly positive returns, with the most profitable strategy being the one that holds the momentum portfolio for 6 months after a 12 month ranking period, yielding an annualised return of 16.2%. There is also evidence of abnormal excess return generated by momentum strategies across industries (Moskowitz and Grinblatt 1999).
Rouwenhorst (1998), Chui, Titman and Wei (2000), Griffin, Ji and Martin (2003) state that even after eliminating the effect of individual stock momentum, industry momentum strategies yield significant abnormal returns, across countries and Chen and DeBondt (2004) have found evidence of excess returns by momentum strategies across equity styles. Okunev and White (2003) find that it also exists in foreign currency spot market and Miffre and Rallis (2005) show the evidence of it in commodity futures market. Rouwenhorst (1998) investigates the possibility that momentum is limited to the US, examining momentum strategies in 12 European countries. Rouwenhorst’s results are similar to those of Jegadeesh and Titman, adding extra significance to the importance of the US results.
Except for the US and the UK, the span of evidence on the international presence of stock momentum is examined by Chui et al. (2000) who document momentum profits in Asian markets with the exceptions of Japan and Korea. Doukas and McKnight (2005) find significant momentum returns in 8 out of 13 European markets, whilst Forner and Marhuenda (2003) find significant momentum returns in the Spanish market.
The studies carried out by DeBondt and Thaler (1985, 1987), Chan (1988), Chopra, Lakonishok, and Ritter (1992), and Richards (1997) show that the contrarian strategy of ranking portfolios according to past returns, buying the stocks with the worst performance and selling those with the best performance, generates positive excess returns over a long holding period (usually three to five years).
The overreaction hypothesis stands in contrast to the underreaction hypothesis supported by proponents of the momentum strategy. The success of the contrarian strategy has been credited to the overreaction hypothesis of DeBondt and Thaler (1985, 1987), DeLong et al. (1990, 1991), and Hong and Stein (1999). In 1996, Chan, Jagadeesh and Lakonishok argued that “Spelling out the links between momentum strategies and contrarian strategies remains an important area of research.”
Market overreaction, first observed by DeBondt and Thaler as mentioned earlier, has introduced a new area of research in finance and their ideas have greatly influenced securities selection for practitioners. Chopra et al. (1992) find that loser portfolios, formed on the basis of prior five-year returns, outperform winners by 5 to 10% per year during the subsequent five years.
Using national stock market indexes, Richards, stresses that during the first six months after portfolio formation, winners continue to outperform losers (momentum), but over the 3 and 4 years horizon, losers start to outperform winners (contrarian). DeBondt and Thaler (1985, 1987) in their study found that the stocks that performed poorly over the previous 3 to 5 years will generate higher returns over the next 3 to 5 years than the stocks that performed well during the same period. They also found that a contrarian strategy of selling past winners and holding past losers generates approximately 8 percent per year from years 3 to 5.
This study shows that contrarian strategies are more profitable over long holding period of 3 to 5 years. Trading strategies of taking long positions on losers and short positions on winners have been employed by Lehman (1990) and Park (1997). Jegadeesh (1990), Martell and Trevino (1990), Lehman (1990), Jegadeesh and Titman (1995), Antoniou, Galariotis, Spyrou (2003), Wang, Yu (2004), and have proved that reversals also exist over shorter horizons (of one week to three months).
For instance, Lehman finds that portfolios of securities that had positive returns in one week typically had negative returns in the following week, while portfolios with negative returns in one week typically had positive returns in the next week. Jegadeesh (1990) finds the same result using the monthly data from 1934 to 1987. Both of them show contrarian strategy yield abnormal returns in the very short forming period. Chopra et al. (1992) find some support for the conclusions of DeBondt and Thaler, but report that only 2.5% of the 14% return on their contrarian strategy is attributable to overreaction.
Nevertheless, evidence of overreaction has continued to emerge from different markets such as Japan (Bremer et al. (1999)) and China (Kang et al. (2002)). UK evidence on overreaction has been largely consistent with the early US study of DeBondt and Thaler (1985). Power et al. (1991) and Macdonald and Power (1991) find evidence of long-term overreaction, although only for a small sample. Another study by Clare and Thomas (1995) provides support for this evidence, but is found to be largely attributable to the size effect.
Dissanaike (1997) finds more than 50 supporting evidence for overreaction in the UK over the period 1993-2000 using FTSE 500 data. Along with these studies of long-term overreaction, very short-term overreaction has been documented by Jegadeesh (1990) and Lehmann (1990), but is now generally considered to be the product of short-term microstructure biases as argued by Kaul and Nimalendran (1990) and Lo and MacKinlay (1990).
Undoubtedly, the body of research on momentum strategies has grown significantly over the past years. Momentum and contrarian trading strategies rely on completely opposite theories and even though such strategies have been practiced to a great extent it is only in the last 10 years that academics have become involved in both quantifying and accounting for their success.
Most academics in their studies have tried to identify the possible reasons for momentum but still have not reached a common explanation for the momentum premium. “In an efficient market trading strategies, based on past prices or publicly available information, are unlikely to succeed as the information that investors seek to exploit is already reflected in prices. Ceteris paribus, a profitable momentum strategy is an anomaly because it is inconsistent with market efficiency”. (Agyei-Ampomah, Sam, 2005)
The momentum premium has been described as a compensation for bearing higher risk (Conrad and Kaul; 1998), a result of data mining (Black, 1993; MacKinlay, 1995), or deceptive and economically insignificant in an attempt to rationalise the anomaly (Lesmond et al, 2004; Hanna and Ready, 2003). However, there are studies that show that the risk-based and data mining claims do not sufficiently explain the momentum premium.
The most likely explanation for the momentum effect seems to be investors’ underreaction/overreaction to information, which is gradually integrated into prices during the next period. Supporters of this view believe that momentum strategies produce higher returns by taking advantage of irrational market behaviour, such as investor underreaction and/or overreaction to information. Daniel et al (1998) suggest that the momentum effect is a product of a delayed market overreaction.
The work of Hong and Stein (1999) and Barberis and Shleifer (2003) also indicate early underreaction and a succeeding overreaction which is consistent with short term price continuations. Doukas and McKnight (2005) find evidence in support of this view, as in Hong and Stein (1999, 2000). Du (2002) argued that investors can be characterised by high or low levels of confidence and thus underreaction arises when investors with low confidence are slow to make decisions.
According to Du (2002) “Delays in acting upon information cause the effects of new information to persist inducing a continuation pattern in returns”. Firm trading level characteristics have also been found to influence momentum profits. Lee and Swaminathan (2000) reported that firms with high trading volume have higher momentum than firms with low trading volume.
Except for the sources of momentum profits, another significant but still puzzling subject has to do with the cost-effective significance of the momentum returns. Although transaction costs are essential in evaluating an investment strategy, past momentum studies fail to appreciate the trading costs associated with the strategy. Jegadeesh and Titman (1993) and Liu et al (1999) both assume a one-off cost of 0.5%.
This is an ambiguous estimation based on average trading costs and ignores the frequency of trades. Lesmond et al (2004) conclude that momentum returns cannot be realised due to considerable transaction costs. In contrast, Korajczyk and Sadka (2004) find that for certain momentum strategies it would require about $6bn for profits generated by momentum strategies to “vanish” due to transaction costs. Although transaction costs could be substantial, they do not completely clarify the existence of the momentum premium, hence further analysis on the economic significance of the momentum strategy is necessary.
For the purposes of this paper, we would like to examine whether there is evidence of abnormal returns in the Greek stock market (Athens Stock Exchange--ASE from now on) and especially in the FTSE 20 index of large capitalization stocks, over a period of 21 years from January 1987 to January 2008 following either a momentum strategy or a contrarian strategy. The motivation of this project is the fact that momentum profitability in emerging stock markets such as the ASE has not been examined to a great extent.
A few years before, Greek stock market was a relatively small and under-examined emerging market. However, innovation and other major reforms that have taken place in the last 20 years caused the market to gain more power. In the last decade an increasing number of new companies were admitted in the ASE in order to raise capital, and a growing number of investors entered the market by investing in corporate stocks. These developments enhanced the local and international investment interest for the ASE, which is now expected to gain the characterization of a more developed market.
The database we will use in this paper consists of the daily spot stock prices of all the stocks that form the FTSE 20 index, which can be sourced from either DataStream or Bloomberg. FTSE 20 index is a large capitalization index which includes the 20 largest companies (blue chips) listed in ASE. We have chosen this period because it covers some interesting periods of stock market behavior and the Greek economy as a whole as indicated by 1) four national elections, 2) the entry of Greece to the “Euro-Zone” (2001) 3) the characterization of the Greek stock market as a developed market since 2001, 4)the “Black Monday” crash in 1987 -- a world-wide phenomenon where all major world markets decreased substantially 5) the crash of the Athens Stock Exchange in 1999 and 6) The ‘free-fall’ of the Greek Stock Market Index from 6500 points (September 1999) to 1800 points (September 2002). All these events make the study of the existence of momentum extra-returns in the ASE to be a rather challenging task and we expect to discover some rather interesting results.
This paper aims to extent previous studies for the ASE by employing the same methodology that was used in one of the most significant studies about momentum strategy by Jegadeesh and Titman (1993, 2001, and 2002). In their study, they tested whether momentum strategy yields abnormal return using the sample of all the stocks traded on the New York Stock Exchange and American Stock Exchange for the testing period from 1965 to 1989. Koutmos et al (1993), Antoniou et al (2005), Spyrou and Mandalis (2003), in their studies have provided evidence of return predictability in ASE using the methodology employed by DeBondt and Thaler (1985).
From the dataset described in section 5, we will compute the returns of each stock that comprise the FTSE 20 index using the formula below:
where is the stock return for period t, is the spot stock price at period t and is the spot stock price at period t-1.
We select the stocks based on their performance over the last 3, 6, 9, 12 months, the formation period. When the returns are calculated, the stocks will be ranked from highest to lowest based on the historical return. The next step will be to engage in a series of strategies to create long positions in the stocks with the highest return and create short positions in the stocks with the lowest return.
The return on the momentum portfolio is then measured as the difference between the returns of the winner and the loser stock. The procedure is moved forward to create new winner, loser and momentum portfolios. For example, in the 3-1 strategy, the winner/loser portfolio is formed at the end of December, based on past 3 months return, holding this portfolio for 1 month, and forming a new portfolio at the end of January.
Once a stock exposure is initiated it is held for a period, at which time the foreign exchange position is re-evaluated. The return of momentum is simply defined as the difference in the returns of the winner and loser stock. We can form 16 distinct strategies; (1-1, 1-3, 1-6, 1-12, 3-1, 3-3, 3-6, 3-12, 6-1, 6-3, 6-6, 6-12, 12-1, 12-3, 12-6, 12-12). A brief description of the 1-1 strategy follows:
1-1 strategy:
Both the ranking and the holding period for this strategy is 1 month. We calculate the returns for all stocks beginning from January 1987 to November 2007 and rank them from the most attractive stock to the least attractive for each month. Once the strongest and weakest stocks are identified, a long/short position is taken by buying the most profitable stock and selling the least profitable one.
For each month, we calculate the profits/losses of our “sub-portfolios” and at the end of the year we calculate the total profits/losses of our portfolio. For example, in this strategy the winner/loser portfolios were formed at the end December , based on the previous month’s returns, holding this portfolio for 1 month, and forming a new portfolio at the end of January. We will follow the same procedure to form the other strategies, by changing the ranking and holding period for each one and we will also carry out the contrarian strategy for the same ranking and holding periods under which, the weakest momentum stock is bought and the strongest momentum stock is sold.
After carrying out the study we expect that the momentum profitability effect will be present in the Greek stock market and these results would be consistent with the results in previous studies about the existence of momentum profits in the European and emerging markets and the studies for the existence of momentum profitability in the ASE.
A study published in the Financial Times on February 18th by Dimson, Marsh and Staunton states that “momentum investing in equity markets delivers striking and remarkably persistent excess returns” They conclude: “Though costly to implement on a stand-alone basis, all investors need to be acutely aware of momentum. Even if they do not set out to exploit it, momentum is likely to be an important determinant of their performance”.
The study of both momentum and contrarian investment strategies is a very interesting and challenging discipline in finance, not only because the effectiveness of such strategies suggest a rejection of the EMH, but because research into the sources of these profits lies at the centre of comprehending the behaviour of market participants and how markets price securities. From the perspective of investors the significance is equally great, indicating the methods of safe and profitable investment, and giving answers to the questions of corporate managers who wish to know how the market will react to announcements about the wellbeing of their companies and their plans for the future.
