1. Intro: The January effect can be explained as a financial anomalywhere the returns of stocks particularly small stocks in January tend to bemuch higher than any other month.
Rozeff and Kinney (1976) was the first tostart off this research on seasonal anomalies, which created an epidemic forother researchers to go on and study this effect in more detail. Marketanomalies are events that can be exploited to earn abnormal returns, implymarket inefficiency. The January effect is a market anomaly, and it is viewedas controversial because till this date, none has a clear explanation of whatit is. Research countries will be the UK and Germany, with FTSE 100 and FTSEsmall caps being the UK indexes and DAX 30 and SDAX being Germany’s indexes.With these index’s, I will examine whether there is a clear indication ofJanuary performing better than all other months and also see whether there is adifference in the small indices compared to the larger index’s.
I will also beoutlining other calendar anomalies related to time like the day-of-the weekeffect and the holiday effect. The datawill be taken from Bloomberg from which the results will be able to tell me whetherthe behaviour of an investor changes once they know they’ll gain more. So themain question I am answering will be: “is the January effect real or is it downto coincidence? Empirical research andliterature reviews will help me prominently to answer this question.
I hope tofind a difference between January and all other months because it will thencause further research to be done as it will be so contradicting to the famousefficient market hypothesis which is believed by so many today. 2. Literature review:2.1 Efficient markethypothesis:This anomaly is proposes that it is impossible to beat themarket because stock market efficient causes existing share prices to alwaysincorporate and reflect all relevant information. Essentially EMH believes thatyou cannot get higher returns than what the market returns are.
Most academicshave supported EMH while most practitioners have not. This is changing though. The term efficient market was first introducedto us by Fama who then went on to extend and refine the concept of efficientmarket hypothesis by defining three forms of market efficiency (Fama 1970); theweak form, the semi-strong form and lastly, the strong form. The weak form iswhere stock prices reflect all information regarding past prices. Semi-strongform efficiency is when stock prices embrace publicly information as well aspast prices. Strong form is where stocks prices incorporate all of thosecharacteristics so; private information, public information and past prices.
Kendall (1953) built on the efficient market hypothesis to comeup with the random walk theory which was heavily built on Regnault (1863)model. Kendall (1953) observed the random walks that stock prices tend to have.However, if this random walks theory is supposedly true, how can we explain themarket getting beaten by Warren Buffett? EMH argument can be refuted by therelevant contradictory studies which we can find from Basu (1972) who foundevidence to suggest a price/earnings ratio study which shows EMH not to workcorrectly.
2.2 The Januaryeffect:The January effect is a seasonal effect where there areincreases in stock prices during the month of January. Studies have shown thatreturns in January are significantly higher than returns in any other month ofthe year.
Sidney Wachtel (1942) was thefirst to notice this effect and he was sure to mention that Harvard Committeeon Economic research 1919 and Richard Owens and Charles Hardy 1920s studieswhich was done before him on seasonality in stock prices showed no evidence ofseasonal tendencies (Wachtel, 1942). Other studies like Rozeff and Kinney(1976) who studied the New York stock exchange to find average return to bemore than 50% higher than other months and Gultekin et al (1983) who sought tofind the January effect present in 15 different countries. Gultekin et al studyshould be enough to prove the January effect as he done it in several differentcountries outlining that it may a global phenomenon. These studies are able backup the possible explanation behind the January effect which is a patternexhibited by stocks where there is a drop in price in December because investorsengage in tax-loss then the stocks recover quickly in the first week of tradingin January and enables investors to gain higher profits (Teall, 2012). However, it has been acclaimed that the January effect is nolonger prominent coming in to the 20th century. Personally, I agreewith this because I believe that it will no longer down to the tax hypothesisbut instead it will be due to behavioural finance. Even before the 21stcentury, researchers like Roll (1983) who did not find evidence regarding thetax-loss selling hypothesis. Also, Pearce and Wilson (1987) believe that theJanuary effect existed before the introduction of a tax system so thatexplanation is void.
2.3 Day of the week:Stocks normally exhibiting larger returns on Fridays incomparison to Mondays: Cross (1973) was the first to observe that, on average,Monday’s stock returns are negatives and Fridays stock returns are positive.This was then backed up by Harris (1986) who has found that the first tradinghour on Monday is characterized by negative returns, while returns are positivefor the same time period on other days. Others go on to expand the day of theweek by linking it in with the January (Rogalski 1984) report that the Mondayand weekend effect are different in January than over the other months. Theyfind that Monday’s returns are, on average, positive in January and negativefor the rest of the year. However, just like all other seasonality anomalies,there have been other studies which have studied this anomaly and rejected it.
Gibbons and Hess (1981) suggest that this effect is not the result ofmeasurement errors in recorded prices (Gibbons and Hess 1981)2.4 The holidayeffect: The holiday effect is the tendency for a stock market togain on the final trading day before an exchanged-mandated long weekend orholiday such as Christmas. The holiday effect was first introduced by Fields(1934) which is the earliest proclaimed seasonality effect of this paper. Theholiday effect is very much to do with behavioural finance as it could beproven that the behaviour of investors will influence their buying and sellingof shares.
Studies by Brockhman and Michayluk (1998) suggest that investorswill tend to buy shares before holidays due to the ‘high spirits’. In my opinion, this seasonality effect is themost difficult to test because essentially you’ll have to find qualitative datawhich shows behaviour of investors increase due to the holidays rather than anyother reason which is harder to prove than getting simple quantitative data. Researchdone by Vergin and McGinnis (1999) shows how the holiday effect has disappearedfor the large companies but are still available for the smaller companies whichis similar to the January effect. 3. Methodology:My methodology will include using Bloomberg and excel. Iwill need to find data regarding FTSE 100, FTSE small caps, DAX 30 and SDAX. Thedata period needs to be large enough in order to get a fair representation andreliable results so data is from 1975 to 2016; even though there was afinancial crisis in the 2000s, it will be interesting to this whether this hadan effect on the January effect.
I choose FTSE 100 because it is an index onthe London Stock exchange featuring 100 companies with the highest marketcapitalisation. FTSE small cap index is an index of small marketcapitalisation. DAX 30 is a market index of the top 30 German companies whotrade on the Frankfurt Stock Exchange (FSE). SDAX on the other hand is an indexon the FSE consisting of 50 small to medium sized companies. The necessary datafrom Bloomberg will be extracted to excel and from excel, I will create anexcel regression output. Lastly, I have to bear in mind that Germany end offinancial year is in September.
My data will be an average of the years from 1975 to 2016.There should be a positive difference between the return for January and thereturn for the other 11 months of the year. I will be using the monthly pricereturns formula of the index:Eq (1) Rit =Oi+ ?idit +Eit Rit: Monthly return on the stock market index and t isequal to the timeOi: Overall average of the 11 months?i: Return difference between January and the 11 othermonths dit: Dummy variable; January is i=1Eit: Error termWe hope that the intercept and slope is equal to the meanreturn of January.4. Further research:My research study merely focussed on two indexes from the UKstock market and two index’s for Germany’s stock market.
In my opinion, a fairrepresentation for further research of this topic would be if the researchertook a minimum of three different countries from each continent in the worldand measures that index return for both small cap companies, medium capcompanies and large cap companies. In doing this, it will interesting to see ifJanuary effect is a real phenomenon which happens in all stock marketsregardless of the size of the company or whether it just happens in the UK withsmall cap companies. It would also be fascinating to see whether there are anyother effects which affect the January effect mainly focussing on social or politicaleffects. I did not get to go into enough detail about the behavioural financeof the January effect which would be interesting to see how the cognitive biasassumes people make predictable mistakes when assessing information could heavilyaffect the January effect.5. ConclusionMy expectation of this project is to find whether the Januaryeffect still holds till this day and if not, when did this effect seemed to dieout. Due to Reinganum (1983) study onmarket behaviour of small firms in January which showed results of smallerfirms receiving larger returns and January especially in the first trading daysin January. I myself will be hoping to find effects which have never beenstudied.
If it is true what they say about the January effect being cause by tax-loss,we should find returns in January significantly higher than all other monthsand for Germany, we should find returns to be significantly higher in Octoberdue to their end of tax year being in September. The main hypothesis of the January effect which is thetax-loss hypothesis may not be the only reason the January effect takesplace. There could be a number ofdifferent reasons why the January effect happens and as psychology has gotbigger of the year and the number of researchers have increased there ispotential for psychology researchers to have a go at explaining the psychologicaleffects investors have which may in fact lead to the January effect. However, some may argue that this will takeaway the while meaning of the January effect being down to tax-loss.All in all, there is enough evidence to show the Januaryeffect exists however; there is also enough research to say the January effectin fact, does not exist. Whilst I do my final dissertation and collect more evidenceto agree or refute against this seasonal anomaly and I am able to followthrough my methodology, I will able to get a more definite answer of whether Ithink the January effect still exists.