EL PROBLEMA, OBJETIVOS, HIPÓTESIS Y VARIABLES
2.1. PLANTEAMIENTO DEL PROBLEMA
2.1.2. ANTECEDENTES TEÓRICOS
An extensive of literature have documented seasonal patterns in risky asset returns. The existence of these anomalies has strong implications against the efficient market hypoth- esis. Examples of the patterns include: the Sell in May principle, the Monday effect, the January effect, the Weekend effect, the SAD effect and the Turn-of-the-month effect.
Starting with the Monday effect, it refers a tendency of stock market returns to be rela- tively lower, and often negative on average, on Mondays rather than the other days of the week. This theory also states that lower returns on Monday are driven by the closing per- formance of stocks on the previous Friday. This effect has been observed in international financial markets. Market practitioners identified the Monday effect as early as the 1920s. Kelly (1930) investigated a three-year US stock market data and identified Mondays are the worst days to buy stocks. He ascribed weekend decisions of individual investors to be the cause of lower returns on Mondays. Cross (1973) studied 844 sets of Fridays and the following Mondays price changes in the Standard & Poor’s Composite Stock Index from January 2, 1953 to December 21, 1970 and reported Fridays outperformed Mondays in
terms of both mean percentage change and the percentage of times the Index advanced during the whole eighteen years period. Jaffe and Westerfield (1985) used six distinct samples, including American, Canadian, British, Japanese and Australian equity markets and reported the existence of the Monday effect in international stock markets. Evidence from some emerging markets also supported the Monday effect. Basher and Sadorsky (2006) considered daily closing prices on 21 emerging stock markets and the Morgan Stanley Capital International (MSCI) World index across the period 31 December 1992 and 31 October 2003. Analysing a total number of of 2827 observations and used five different models to examine the day-of-the-week effect. Their estimation results indicate strong day-of-the-week effects in some emerging stock markets even after controlling conditional market risk.
The May principle or the popular expression ‘Sell in May and Go Away’ is another well- known seasonal pattern. It refers to the trading strategy of investors to sell their stock holdings in May and return to the stock market in November to avoid the typically volatile May-October period, also known as the seasonal decline in stock markets. Bouman and Jacobsen (2002) documented that the famous Sell in May effect in stock return was pre- sented in 36 of the 37 countries in their sample, and this effect goes as far back as 1964 in the UK stock market. However, they found no explanations for the seasonal pattern. Maberly et al. (2004) argued that the average negative return patterns in stock markets can be explained by two outliers, namely the 1987 Black Monday crash in world financial markets and the bailout of the Long Term Capital Management hedge fund in 1998.
The Turn-of-the-Month (TOM) effect is a tendency of asset returns to increase during the last two days and the first three days of each month. This tendency is supported by exten- sive academic research, including Barone (1990) who discovered the turn of the month
effect in Italy, Ziemba (1991) in Japan, Cadsby and Ratner (1992) in Australia, Canada, Germany and Switzerland, Hensel and Ziemba (1994) in the U.K. and Martikainen et al. (1995) in Finland. In addition, Kunkel et al. (2003) studied 2153 months of daily stock closing prices from 19 countries from 1988 to 2000, and employed parametric and non- parametric tests to search for evidence of the TOM effect. Their result showed that the TOM effect was presented throughout the 1990s in 16 of 19 countries. Their models supported a significant TOM pattern in stock markets that is independent of any other calendar-related patterns, which can not be explained by outlier observations.
The January effect refers to an increase in stock prices in January along with a drop in stock price at the end of December. The January effect was first brought to the attention of modern finance by Rozeff and Kinney (1976), who calculated the average return on NYSE index from 1904 to 1974 , they found this to be more than eight times higher in January than in any other months. Dyl (1977) and Branch (1977) also confirmed the existence of the January effect in US stock markets. Gultekin and Gultekin (1983) found the January effect also influenced international stock markets. Subsequent research by Reinganum (1983) and Keim (1983) indicated that the January effect in more pronounced on small capitalisation stocks. Chu and TUNG LIU (2004) provided additional evidence of the January effect is size determined , they employed the Markov-switching model to explore the monthly stock returns from 1926 to 1992 and found a significant January effect in small capitalisation stocks only. Branch (1977) introduced the Tax-loss selling hypothesis to explain the January effect, which indicates investors tend to sell stock holdings before the end of tax year to reduce tax payments, and resume stock holdings at the beginning of January.
decades. Recent behaviour finance studies reported a seasonal pattern in equity returns named the SAD effect. The SAD effect also known as Winter Blues, refers to the prin- ciple that stock returns are significantly related to the amount of daylight through the fall and winter. In clinical studies, SAD is a medical symptom where sufferers feel depres- sive when the days become shorter after the autumn equinox, and experimental psychol- ogists have documented that depression led to higher risk aversion, hence stock return are lower when investors suffer from SAD. Kamstra et al. (2003) documented the SAD effect by providing international evidence that stock market returns vary seasonally, ac- cording to the length of the daytime in the fall and winter. Their result indicated that the SAD effect persists in the seasonal cycle of stock returns, even after controlling for some well-known market anomalies and weather factors. They ascribed the cause of the SAD effect to the change of daylight, which has a more pronounced effect on people’s mood than weather factors and mood is related to investor risk perception. Dowling and Lucey (2008) extended the SAD hypothesis research in international stock markets. They grouped countries into those close to the equator against those distance from it, amd by employing the GARCH model, they found a large proportion of SAD coefficients were significant and with the expected sign to support the SAD hypothesis. Moreover, Dowl- ing and Lucey (2008) further provided two key pieces of evidence for the SAD effect. Firstly, they documented a more pronounced SAD effect for those countries further away from the equator, just as SAD is caused by the reduction of daytime in the fall and winter, it is expected that more investors will suffer from SAD in those countries which are far away from the equator. Secondly, they also found a more significant SAD effect in small capitalisation equities, held mostly by individuals, which is an indication of stronger SAD effect on individual investors, given that individual investors are more likely to be affected by the change of mood (Yuan et al., 2006). Kamstra et al. (2012) adopted more statisti-
cal methods to support the existence of the SAD effect in international stock markets. The methods include: ordinary least square (OLS) and the seeming unrelated regression (SUR) with MacKinnon and White (1985) standard errors, and a system of equations gen- eral method of moments (GMM) with heteroskedasticity and autocorrelation consistent (HAC)standard errors.