Calendar anomaly dan kapitalisasi pasar : analisis pada pasar saham indonesia

Stock market anomaly is defined as an abnormality that occurs on the stock trading that causes investor to have the opportunity to obtain abnormal return that should not be possible according to efficient market hypothesis (EMH) (Singh 2014). Calendar anomaly, one of stock market anomalies, is the repeating pattern that only occurs at certain times, such as on a particular month or day (Deyshappriya 2014). Two forms of calendar anomalies that widely studied are month of the year anomaly (MOYA), also known as the January effect, and day of the week anomaly (DOWA) (Dicle and Levendis 2012, Darrat et al. 2011, Das and Rao 2011, Bouman and Jacobsen 2002). Each type of calendar anomaly can be exploited to construct a trading strategy. MOYA, for instance, contain the opportunity to obtain abnormal return by buying stocks on January. Whereas DOWA, indicates that investors could obtain highest return on Friday. Numerous findings of calendar anomaly in many countries indicate that those countries stock markets are still inefficient. Indonesia is one of the countries that also experience calendar anomaly (Anwar and Mulyadi 2012, Iramani and Mahdi 2006, Werastuti 2012, Naughton and Veeraghavan 2005). There is another theory about calendar anomaly which states that the anomaly tends to occur in small cap stocks (Bodie et al. 2008). To clarify the theory, analysis of calendar anomaly needs to be done on different stock capitalization. When different patterns of anomalies are found in different stock capitalization, different strategies should be formulated for each type of capitalization. Another interesting thing that can be analyzed related to calendar anomaly is to analyze whether calendar anomaly also exist on individual stocks, unlike most of the previous studies that only analyzed at market level (indices) (Dicle and Levendis 2012). In this study the indices used are LQ45 and Pefindo25. The LQ45 index is used to describe the "behavior" of large capitalization stocks while the Pefindo25 index is used to represent a group of stocks with medium and small market capitalization. The analysis of the existence of MOYA and DOWA is done both on index level and individual stock level. The number of stocks observed is 58 stocks, 35 shares of the LQ45 index and 23 shares of the Pefindo25 index. Those stocks are the most commonly listed stocks on both indices during the observation period of 2010-2015. To examine the existence of DOWA and MOYA in stock returns dummy variables regression model was used. With daily or monthly return as dependent variable and days or months as independent dummy variables. The regression model was estimated using two estimator options namely Ordinary Least Squares (OLS) method and ARCH / GARCH Model. Estimator applied for each model is depends on wether or not the problem of heterokedastisitas in the return series. DOWA is found both in LQ45 and Pefindo25, with Monday having the lowest return and Wednesday has the highest return. In individual stocks DOWA more occur in LQ45 that is at 74.29 percent of shares than Pefindo which only occurs in 47.38 percent shares. On the individual stocks level DOWA is mostly found in LQ 45 stocks, which is found 74.29 percent of total LQ45 stocks, rather than Pefindo25 which is only found in 47.38 percent of total Pefindo25 stocks. The difference between the returns on Wednesday and the other days on the LQ45 index ranged from 0.15 to 0.41 percent, while in Pefindo25 the difference ranged from 0.15 to 0.32 percent. DOWA pattern on Indonesia stock market is alleged to have shifted from previous DOWA pattern finding that was Monday as the day with lowest return and Thursday as day with highest return. MOYA is found on index LQ45 with pattern that is March as month with highest return and August as month with lowest return, whereas at index Pefindo25 only found August as month with lowest return. MOYA on the LQ45 index only occurs in 45.71 percent of the stocks while in Pefindo25 the number is fewer that only 34.78 percent of stocks. March return of LQ45 is 1.18 to 8.4 percent higher than of other months. The low return of August allegedly due to the influence of the Eid holiday that often fall in August and around August in the observed period. The period of holiday is suspected to cause a decline in stock trading activity that ultimately lowers the return