The impact of the financial turmoil between
February 2006 and the end of 2015 on the turn of
the month effect in France, Germany, Greece,
Italy, Spain and The Netherlands
Name: Joeri Verhagen Student number: 10274871
Supervisor: Dr. Tanju Yorulmazer Date: 29th June 2016 Bachelor thesis
2
Abstract
This thesis focuses on the turn of the month effectin6 European countries; France, Germany, Greece, Italy, Spain and The Netherlands. The time period of this research runs from February 2006 up until the end of 2015. In these years, there were periods of quite some financial turmoil like The Great Depression and The European sovereign debt crisis. These specific periods makes it interesting to investigate the influence of these events on the turn of the month effect. To get a good estimation of the time periods in which these events took place within the above-mentioned countries we used their periods of recession as an indicator. The results show that the turn of the month effect issignificant for Greece in the total time frame of February 2006 up until the end of 2015. The results also show that the periods of recession had an impact on the turn of the month effect. In times when these 6 countries were in recession, the turn of the month effect became significant for 4 of them. However, Italy and Spain didn’t show a turn of the month effect at all despite being in recession periods as well. The results show that in periods of no recession the turn of the month effect vanishes. This is in line with the efficient market hypothesis which states that a market anomaly like the turn of the month effect won’t appear in an efficient market. By removing the inefficient periods like the events mentioned above, this research shows that the turn of the month effect indeed appears to have dissolved.
Statement of Originality
This document is written by Student Joeri Verhagen who declares to take full responsibility for the contents of this document.
I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it. The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.
3
Contents
I Introduction ... 4
II Literature review ... 5
2.1 Turn of the month ... 5
2.2 The efficient market hypothesis ... 8
2.3 Financial crisis ... 9
IV Results ... 13
V Conclusion ... 16
VI Discussion ... 16
4
I
Introduction
Market anomalies have always been interesting for investors and researchers; for investors due to the arbitrage possibilities and for researchers for their curiosity of why they happen (Zwergel, 2010). Are these anomalies just statistical errors or is there something peculiar going on that can be explained by economic theory? The efficient market hypothesis states, that a market anomaly can only happen by chance or because of misjudgment of information (Fama, 1998). Although a few market anomalies exist, like the weekend effect, the turn of the year effect and others, the anomaly that will be discussed in this thesis will be the turn of the month effect (Schwert, 2003). The turn of the month effect was first described by Robert A. Ariel (1987). He found that the first half of the month showed a concentration of returns in comparison to the second half of the month. He found that over 18 years the S&P 500 index made most of its return in the first nine business days of the month. The last nine business days of the month showed no significantly different return than zero percent. First described in 1987, the investigation into the turn of the month effect is still ongoing and the latest research is published in 2014 (Sharma & Narayan, 2014). During these 27 years, this effect hasn’t dissolved yet, which is in contradiction with Schwert (2003). Schwert wrote that a market anomaly should dissolve over time after being documented and described, because of either arbitrage possibilities or because the market anomaly was a statistical error and not an anomaly in the first place (Schwert, 2003). The turn of the month effect has been proved by a lot of researchers, therefore the possibility of it being a statistical error seems to be very small. So one of the questions that come up is: Is the turn of the month effect still present in France, Germany, Greece, Italy, Spain and The Netherlands? To answer this question we have to investigate this effect in these countries. All countries have been researched for the turn of the month effect in the past, with mixed results.
The countries that will be researched within this thesis are France, Germany, Greece, Spain, Italy and The Netherlands. These countries have been chosen because they also fit the second question that this thesis tries to answer. This question is: Did the financial turmoil that happened in the European Union, within the chosen time frame, have an impact on the turn of the month effect? During financial turmoil, the efficiency in markets is reduced. This is due to investors showing more biases in their decision making (Shefrin, 2002). Will this reduction in efficiency due to financial turmoil affect the turn of the month effect? The
countries that have been chosen show a difference in the impact of the financial distress; for example, the 3 North European countries suffered less from the financial distress during the European sovereign crisis compared to the 3 South European countries. The 6 countries that will be investigated for the turn of the month effect, have been investigated in the past
5 (McConnell and Xu, 2008). The latest research proved the existence of the turn of the month effect up until February 2006 for 5 out of 6 countries. I also included Italy, which is a country that in the past got mixed results when tested for the turn of the month effect (Barone, 1990), (McConnell and Xu, 2008). After2006, several big events took place that affected the
markets efficiency. In the efficient market hypothesis a crisis shouldn’t even be able to occur (Abreu and Brunnermeier, 2003). This thesis will thus attempt to analyze the market after 2006 for the existence of the turn of the month effect, as well as the possible impact the financial turmoil after 2006 had on this market anomaly.
II
Literature review 2.1 Turn of the month
Robert A. Ariel (1987) found, that from the day before the turn of the month, throughout the first half of the month, the returns were significantly positive. He also found that during the second half of the month the returns were not significantly different from zero. His research was done on the S&P 500 index and his data covered the time span of 1963 till 1981.
This effect was further described by Lakonishok and Smidt (1988). They investigated the effect of the turn of the week, turn of the month, turn of the year and the stock return around the holidays. The focus of their research was finding evidence to confirm the
existence of these anomalies, as well as confirming that they weren’t caused by each other. This was based on analyzing 90 years of data from the Dow Jones Industrial Average. Their research confirmed the research already done by Ariel. They also tried to find the cause of the turn of the month effect within the turn of the year effect. They did this by excluding the periods, the turn of the year effect occurs. Their results came out negative and showed no proof that the turn of the month effect has caused the turn of the year effect. They also tested other possible relations but couldn’t find enough support to prove these. They were able to narrow the turn of the month effect days and found that most of the return happened the day before the turn of the month until the 3rd day of the month.
In the 1990’s Barone published his research about the turn of the month effect. He was the first to investigate if the turn of the month effect was also to be found in other markets. He did this by analyzing the Italian Milan Stock Exchange. His results were in line with the results of the research done in America and confirmed the presence of the turn of the month effect in Italy, making it an international market anomaly.
6 this phenomenon but instead tried to explain the cause of the effect. His paper explained the turn of the month effect to be caused by the standardized payments in the United States at the end of the month and the thereby caused surge on the stock market. These standardized payments occurred mostly around the turn of the month and included wages, dividends, interest, principal payments and other liabilities. This theory about standardized payments is one of the most accepted theories about the turn of the month effect origins.
Ziemba (1991) did a research about indices in Japan. He found that within Japan the turn of the month effect takes place on other days than for example the United States of America, where the turn of the month effect goes from the day before the turn of the month until the third day of this month. The days that the turn of the month effect takes place in Japan are from the fifth day before the turn of the month till the second day of the month. Ziemba (1991) states that this is due to differences in the standardized payments dates between Japan and the United States of America. In Japan, the payment is concentrated on the 25th of the month and this creates buying pressure 5 days before the turn of the month. In the USA, the standardization of payments has led to the payment at either the last business day or the first business day of a calendar month of interest, dividends and principal
payments Ogden (1990). This difference in the standardization of payments dates explains the shift in days of the turn of the month effect in Japan compared with the USA. This result provided extra prove for the claim of Ogden (1990) that the turn of the month effect is caused by standardized payments.
Cadsby and Ratner (1992) showed that the turn of the month effect wasn’t only present in the USA and Italy, but that it was also significantly present in Canada, the UK, Australia, Switzerland, and West Germany.
Booth, Kallunki and Martikainen (2001) did a research of the turn of the month effect within Finland. They tried to find support for Ogden’s claim that the turn of the month effect was caused by standardized payments. They researched the Helsinki Stock Exchange for the years 1991 till 1997. Their results were that the FIM volume, the share volume and the number of trades which measure liquidity, were positively influenced by the turn of the month effect and thus confirming Ogden’s (1990) claim.
Kunkel, Compton and Beyer (2003) examined 19 country stock market indices for evidence of the turn of the month effect. They found this effect in 16 out of 19 countries and they used data from 1988 to 2000. They proved that the turn of the month effect wasn’t only present in western countries but also in less developed countries like Brazil, Mexico and South-Africa. They also used multiple tests to see if the mild violations that occur in daily stock return caused any wrong results. They concluded that in large samples these violations won’t affect the turn of the month effect results. The method of regression used by Kunkel, Compton and Beyer (2003) is also used for the research of this thesis.
7 McConnell and Xu (2008) did an extensive research about the turn of the month effect. They researched for the possibility of the turn of the month effect, being caused by the standardized payments of countries as well as the possibility of the mutual fund boosts that takes place at the end of the quarter. The possibility that it was caused by standardized payments was tested by looking at trading volume. The trading volume on the days of the turn of the month effect turned out to be the days that the volume of trading was the smallest. This is contradicting to what would be required to support the hypothesis of the standardized payments. The mutual fund boosts were also tested by looking at transactions made to mutual funds, but these didn’t appear to increase on the turn of the month effect days. These transactions even appeared to decrease on the turn of the month days. They also tried to look at the possibility of the turn of the month effect being caused by other market anomalies, like the turn of the year effect and the turn of the quarter effect. All these results came out negative and showed no correlation between the turn of the month effect and the other market anomalies. They also tested the turn of the month effect in other countries to explain this anomaly by some characteristic of the US trading system. They tested 35 countries and in 31 of these countries this effect was proved. This proved that the effect isn’t only taking place in the USA and therefore eliminated the possibility that it’s caused by the US trading systems. Their research extended the research about the turn of the month effect up until 31 January 2006. Within the 35 countries, McConnell and Xu (2008) researched; they found that in The Netherlands, Germany, Greece, France and Spain the turn of the month effect was still present. They couldn’t prove the turn of the month effect for Italy.
Zwergel (2010) did a research about the exploitability of the turn of the month effect. He did this to find out if the market anomaly might not dissolve due to transaction costs and such to become impossible to dissolve totally. The research showed that the turn of the month can be exploited and could be profitable after deducting transaction costs and
slippage. The paper also showed that the turn of the month effect is also present in the future of corresponding indices.
Sharma and Narayan (2014) tested if the turn of the month effect also affects firm returns and their return volatility depending on sector and size. This research is the last research that has been published about the turn of the month effect.
8
2.2 The efficient market hypothesis
Eugene F. Fama (1969) founded the theory of efficient capital markets. He provided three possibilities to categorize the importance of information to stock prices and to model the world. The three variations are the strong form, the semi-strong form and the weak form of information availability. They portray three situations in which information is incorporated into the stock price. The strong form states that a normal trader has all the information that is available, the semi-strong form states that some information is not publicly available (inside information) and the weak form states that the price of stocks is based on the information of historical pricing data. This would mean that under all forms of market efficiency the historical stock information is incorporated in the pricing of stock and thus the turn of the month effect should be incorporated into the price of the stocks and thus should dissolve after time (Schwert, 2003).
Within the efficient market hypothesis, a crisis is an economic event that shouldn’t be able to occur (Abreu and Brunnermeier, 2003) because the origin of a crisis is a bubble and a bubble is a value for a stock that isn’t the value it should have under the strong form of the efficient market hypothesis. In times of strong financial turmoil and the big changes that can occur within short time make it so that investors become more afraid to invest and pass up on profit opportunities, because they become more risk averse (Shefrin, 2002). This behavior makes markets less efficient during periods of financial turmoil than during periods of
financial stability.
In periods when markets become less efficient, market anomaly’s that are reliant on the inefficiency of the market, become more apparent. Some of the markets that will be researched were, before the financial turmoil, already inefficient. This is proved by the work of Maria R. Borges (2010). She investigated the efficient market hypothesis on the European stock market. In this article, she tests stock markets by variance ratio test. She tested six European countries; France, Germany, Greece, Portugal, Spain and the UK. She concluded that France, Greece, Portugal and the UK were not efficient markets. For Greece and Portugal were not efficient markets, due to the first-order positive autocorrelation in their returns. France and the UK were not found efficient, due to the presence of mean reversion in weekly data. In her research, she tested two-time frames in order to find out if market efficiency increases over time. She found out that for Greece and Portugal the efficiency increases over time. For France and the UK, she found out that the efficiency was
decreasing. For Germany and Spain, the tests came out positive as being efficient markets. It showed that Spain was the most efficient market in Europe. In the conclusion of the article the author states that when a country becomes a more developed market, the market
9 efficiency increases as well. This would mean that if the countries that I investigate have improved their market efficiency, the turn of the month effect could have disappeared. This is in line with the statement about the market efficiency hypothesis of Schwert (2003). This statement says that a market anomaly should dissolve over time after being documented and described, because of either arbitrage possibilities or because the market anomaly was a statistical error and not an anomaly in the first place.
2.3 Financial crisis
The Great Depression and the European sovereign debt crisis are events that start when the information about mispricing becomes widespread and therefore leads to the bursts of a bubble. So with these said events the starting point of these events is pretty clear to indicate. The end of the collapse of markets is harder to pinpoint. Due to this problem with the timeframe, this thesis will be using the periods of recession. A recession can be
appointed to a fixed timeframe. Therefore the problem of pinpointing the ending date of these events can be avoided. This is necessary to get a more precise picture of countries dealing with a financial crisis. Also, these events had a different impact on different countries. For example, the impact on Greece and Germany appears to be quite big. Due to a recession being able to be strictly connected to the country’s own data we can use different timeframe’s for each of the regressions that will be done in this thesis. Therefore it’s important to be precise in the definition of a recession. The IMF’s definition: “There is no official definition of recession, but there is a general recognition that the term refers to a period of decline in economic activity. Very short periods of decline are not considered recessions. Most commentators and analysts use, as a practical definition of recession, two consecutive quarters of decline in a country's real (inflation adjusted) gross domestic product (GDP)—the value of all goods and services a country produces (see "Back to Basics," F&D, December 2008). Although this definition is a useful rule of thumb, it has drawbacks. A focus on GDP alone is narrow, and it is often better to consider a wider set of measures of economic activity to determine whether a country is indeed suffering a recession. Using other indicators can also provide a timelier gauge of the state of the economy.” (Claessens & Kose, 2009). Since the limitations I have, as a writer of this thesis, it will be impossible to focus on the wider set of measures of economic activity to determine whether a country is suffering from a
recession. Therefore within this thesis, a recession will be defined as a period in which a country is having a negative real gross domestic product growth for a minimum of two consecutive quarters.
10 tested throughout. Some of the countries that will be investigated have been known to show the turn of the month effect up until 31 January 2006. After 2006 the economic situation has changed within the countries that are being investigated. First, there was The Great
Depression and soon after The European sovereign debt crisis followed. They both had a huge impact on stock values within this region and therefore had an impact on the efficiency of the markets within this region. The research question that this thesis will try to answer is if the turn of the month effect is still present in France, Germany, Greece, Italy, Spain and The Netherlands. The follow-up question is if the financial turmoil that happened in the European Union, within the chosen time frame, had an impact on the turn of the month effect?
III
Methodology and Data
To be able to answer the questions of this thesis we need a model to run our regressions upon. The model that will be used is one that is used in the article by Kunkel, Compton and Beyer (2003). This article is one of the more recently published articles on this subject and tried to find a good model for testing the turn of the month effect by using several methods. They found that doing an ordinary OLS regression worked fine for finding the turn of the month effect. This OLS regression is also fairly robust to mild violations of
assumptions, especially in large samples. This model is based on the methods of Lakonishok and Smidt (1988), Pettengill and Jordan (1988), and others. The model of Kunkel, Compton and Beyer (2003) is as follows:
Rt= α + βDTOM + εt
They ran their OLS regression on this specific model, where Rt stands for the return on day t; α is an intercept for the mean return of the rest of the month; DTOM is a dummy variable for the days around the turn of the month; β is the difference between the mean return of the rest of the month compared with the turn of the month return; and εtis the error term. So the requirements for this model are: daily mean return categorized on calendar date, a
specification of the turn of the month effect days, a specification of the rest of the month and lastly the model requires a timeframe which will be tested for the presence of the turn of the month effect. So now we specified how to analyze a market for the presence of the turn of the month effect.
The daily mean return categorized on calendar date will be retrieved from
DataStream. This will be done for the following markets: France, German, Greece, Italy, Spain and The Netherlands. To invest the turn of the month effect on these markets, the
11 research of this thesis will be based on the daily closing quotes of the major indices in these countries, which hopefully gives a good reflection of the total country. These indices are the AEX-25, DAX-30, CAC-40, IBEX-35, MIB-40 and the FTSE-20. The data used for this research covers the timeframe February 2006 until the end of 2015 and will be divided into three parts. The first part is the whole set of data from February 2006 till the end of 2015, which will almost be 10 years of daily closing data to do a regression upon. This timeframe hasn’t been researched yet, which makes it interesting to research if the turn of the month effect is still present. The second part will be all time periods between February 2006 and December 2015, during which periods the countries in question were dealing with a
recession. And the third part will be all time periods between February 2006 and December 2015, during which periods these countries in question were not dealing with a recession. The timeframe for the periods of recession within a country will be based on the real gross domestic product of the country in question. The data for the real gross domestic growth of a country is publicly available and will be downloaded from the Organization for Economic Co-operation and Development website. The data they provide is indirect data from Eurostat and was available until December 2015. This data provided the basis for analyzing if a country in question is in recession or not. Within the literature review, it was stated that a country is in recession if the country has at least two quarterly declining real gross domestic growth stats. After specifying the timeframes of the recessions within each country, the distinction for a country in recession and the same country not being in recession could be made clear. Table 1 reflects the time periods the countries were in recession. As well as the total amount of months spent in a recession.
Table 1 Country recession periods Total months spend in recession France Q2 2008-Q2 2009 Q1 2012-Q2 2012 21 Greece Q3 2007-Q4 2007 Q2 2008-Q1 2009 Q1 2010-Q2 2013 60 Germany Q2 2008-Q1 2009 Q4 2012-Q1 2013 18 Spain Q3 2008-Q4 2009 Q1 2011-Q2 2013 48 Italy Q2 2007-Q4 2007 Q2 2008-Q1 2009 Q1 2010-Q2 2013 63 Netherland s Q3 2008-Q2 2009 Q4 2011-Q1 2012 Q3 2012-Q4 2012 24
12 The days, when the turn of the month effect occurs, need to be specified in order for the turn of the month dummy variable to fit its purpose within the model. The days that the turn of the month occurs shall not be analyzed by making calculations based on each individual day, but rather on the literature that has already been written about this variable. The literature is mostly on the same side in terms of days that the turn of the month effect occurs. The literature starts with Robert A. Ariel (1987) with his discovery of the turn of the month effect. He discovered it as an anomaly that the first half of the month minus the second half of the month gave positive returns. He specified the first half of the month as the day before the turn of the month up until the eighth day of the month. The second half of the month was specified as the ninth day before the turn of the next month up until the second day before the turn of the month. To illustrate this, an example will be given for the turn of the month effect in the month of April in 2006: the first half of the month is then defined as the 31st of March up until the 12th of April. This period gives a total of 9 business days and 4 weekend days. The weekend days are not included in the sample pool since almost no financial activity takes place. The second half of the month is then defined as the 17th of April up until the 27th of April. The 22nd and the 23rd of April are weekend days as well as the 29th and 30th day and are not included in the second half. This gives us a total of 9 business days for the second half of the month. And thus the days of the month were defined by Robert A. Ariel (1987). Later research by Lakonishok and Smidt (1988) showed that the turn of the month was concentrated in 4 business days instead of the 9 business days that Robert A. Ariel (1987) proposed. Lakonishok and Smidt (1988) based this on analyzing 90 years of data. Their findings were also confirmed by Kunkel, Compton and Beyer (2003). Kunkel, Compton and Beyer (2003) also showed that the concentration of the turn of the month effect days was still on the same days for the countries that are being researched. Within this thesis, these four days will be used as the days of the turn of the month effect. The
specification of the days of the rest of the month will be all days that are not turn of the month effect days. In the example of April of 2006, this means that the turn of the month effect days are the 31st of March up until the 5th day of April. The rest of the month will then be defined as the 6th of April up until 27th of April. The 28th day of April will be used for the turn of the month effect in May.
So now it has been specified how all the components of the turn of the month effect model will be retrieved and/or specified, this thesis will now make the step into explaining the details of the results that came out of the multiple regressions that were run.
13
IV Results
To do the regression of the turn of the month effect, on the timeframe of February 2006 until the end of 2015, the daily returns needs to be calculated. The daily returns were based on the daily closing data of the major indices. After this was done the turn of the month effect days where turned into a dummy variable that indicated if the return on a specific date was a turn of the month day. When this was done the data was loaded into SPSS and the model regression was run. The results of the regression over the period of February 2006 until the end of 2015 are visible in table 2, in which the alpha and beta values are being displayed in percentages.
Table 2
Country alpha beta t-stat significance
France -0,001 0,049 0,65 0,515 Greece -0,111 0,241** 1,96 0,050 Germany 0,031 0,022 0,30 0,767 Spain -0,003 0,050 0,63 0,530 Italy -0,023 0,083 1,00 0,319 Netherlands 0,000 0,051 0,72 0,473 ** Significant at the 5% level
The table shows clearly that the turn of the month effect is still present for all countries, which is shown by the positive value for beta. The T-stat, that indicates if the results are significant, show that the turn of the month effect isn’t significant for France, Germany, Spain, Italy and The Netherlands. In Greece, it’s clear that the turn of the month effect is still present and significant. The turn of the month effect is even significant at the five percent level. This gives us the answer to the main research question of this thesis: The turn of the month effect is still present in most countries however it’s only relevant for Greece. For all other countries, the turn of the month return is not significantly different for zero percent.
The results are in line with the research of Maria R. Borges (2010) on the following points: She found out that Spain and Germany were developed and thus efficient markets. These countries would thus be expected not to show the turn of the month effect. She concluded that France and Greece were inefficient markets. Greece indeed seems to show the turn of the month effect but France, on the other hand, doesn’t show the presence as
14 well. This is not in line with the conclusion of the research of Maria R. Borges. France, in this case, was expected to show significantly the turn of the month effect.
To be able to do the two regressions, which involves the countries being in
recession or not, it was first necessary to analyze when a country was in recession. In Table 1 it becomes clear that the northern European countries spend less time in recession than the southern European countries. The northern European countries spend in total 63 months in recession and the southern European countries spend in total 171 months in recession. This indicates that The Great Depression and the European sovereign debt crisis had a bigger impact on the southern European countries. After the information about the periods of recession became specified, it was possible to split the data from February 2006 until the end of 2015 in two parts. The first part became the part where the countries were not in recession and the second part became the part where the countries were dealing with a recession. The first regression that will be shown is the regression about the countries not dealing with a recession. This regression was thus done on the time frame that these countries were not dealing with a recession. In Table 3 the results of this regression are shown.
Table 3
country alpha beta t-stat significance
France 0,035 -0,060 -0,83 0,407 Greece -0,028 0,168 1,07 0,286 Germany 0,065 -0,055 -0,78 0,437 Spain 0,009 0,031 0,35 0,725 Italy 0,036 -0,046 -0,47 0,639 Netherlands 0,028 -0,055 -0,86 0,390
It becomes clear that the turn of the month effect almost completely disappeared. France, Germany, Italy and The Netherlands show a negative beta value and a positive alpha value, which shows that the turn of the month effect became a negative effect, which is not in line with the theory about the turn of the month effect. For Greece and Spain, the turn of the month effect is still positive but not significant and thus could just as well be a statistical error. So in conclusion; if a country is not in a recession the turn of the month effect seems to have disappeared.
15 The regression results on the time frame, when countries were in recession, are shown in Table 4. If we follow the theory from the literature review it would be expected that countries would be less efficient during a crisis. This would mean that the turn of the month effect might increase in times of inefficiency of the market.
Table 4
country alpha beta t-stat significance
France -0,166 0,560** 2,14 0,033 Greece -0,192 0,314* 1,66 0,097 Germany -0,163 0,453* 1,69 0,092 Spain -0,021 0,079 0,53 0,599 Italy -0,080 0,210 1,56 0,119 Netherlands -0,114 0,471* 1,95 0,052 ** Significant at the 5% level
* Significant at the 10% level
For all researched countries the beta value is positive and thus there is a positive turn of the month effect. For 4 of the 6 countries, the turn of the month effect is significant. For Germany, Greece and The Netherlands, the turn of the month effect is significant at the 10 percent level. For France, the turn of the month effect is even significant at the 5 percent level. For Spain and Italy, the turn of the month effect is not significant. Since the results of both regressions are now visible we can answer the second question of this thesis: Was the turn of the month effect influenced by the financial turmoil that took place after 2006? Analyzing the results gives the impression that the turn of the month effect was indeed influenced by the financial turmoil. The removal of the time frame, that the countries in question were in recession, shows that the turn of the month effect vaporizes. If the time frame, when the countries were in recession, the turn of the month effect becomes more prominent in the results. This is the case for 4 of the 6 researched countries. Spain and Italy show no turn of the month effect that is significantly different from zero. This result is also in line with the research of Maria R. Borgos (2010) since she stated that Spain is the most developed country within her test sample. This would indicate that Spain should have an efficient market. She also stated that Germany had a developed and efficient market, but this is not shown in this regression. Italy has been known not to show the turn of the month effect in the past, which is confirmed by these regressions (McConnell and Xu, 2008).
16
V
Conclusion
We can conclude that the turn of the month effect has weakened over time. If we look at the complete time frame of February 2006 until the end of December of 2015, the turn of the month effect is only significantly positive in Greece. For the other countries in this research, France, Germany, Italy, Spain and The Netherlands, the turn of the month effect is no longer in the researched period significant. Up until February 2006, the turn of the month effect turned out to be significantly present in France, Germany, Spain and The Netherlands (McConnell and Xu, 2008). This is in line with Schwert (2003) who stated that market
anomalies should disappear over time. The second question was if the financial turmoil that happened during February 2006 until the end of December of 2015, namely the Great Depression and the European Sovereign Debt crisis, had an effect on the turn of the month effect. It seems that if the timeframe of these events is being removed for each specific country, the turn of the month effect is not significantly there anymore. This gives an
indication that the market is more efficient in stable periods of time. If we only test the turn of the month effect for the time frame that the countries in research are dealing with a
recession, the turn of the month effect becomes more apparent. The turn of the month effect becomes more apparent. It shows that four out of the six countries become significant in the presence of the turn of the month effect. This indicates that under times of financial distress the market efficiency decreases, which is in line with the theory of the literature review. So in conclusion to the second question, we can state that the financial turmoil indeed had an impact on the turn of the month effect.
VI
Discussion
For this thesis, the decision was made to use recessions as a reflection of the time periods when the countries were dealing with financial turmoil. The question remains if this reflection gives a good representation of the period that financial turmoil is affecting the market. Besides this, the time frame of quarterly data might not be precise enough for the daily data that is required for the turn of the month effect analysis. Another issue is the daily closing data for Greece, since they show some extreme values, due to the intensity of the European sovereign debt crisis. During this period there was the risk of default in Greece, which had an impact on the stock market value. There were even days the stock market was closed. This is noticeable within the daily return data and has most likely influenced the
17 outcome of my regressions. In the end, this research could also be a statistical error as stated by Schwert (2003).
VII
Bibliography
Abreu, D., & Brunnermeier, M. K. (2003). Bubbles and Crashes. Econometrica, 173-204. Ariel, R. A. (1987). A monthly effect in stock returns. Journal of Financial Economics,
161-174.
Barone, E. (1990). The Italian stock market: Efficiency and calendar anomalies. Journal of Banking & Finance, 483-510.
Booth, G. G., Kallunki, J. P., & Martikainen, T. (2001). Liquidity and the turn-of-the-month effect: evidence from Finland. Journal of International Financial Markets, Institutions and Money, 137-146.
Borges, M. R. (2010). Efficient market hypothesis in European stock markets. The European Journal of Finance, 711-726.
Cadsby, C. B., & Ratner, M. (1992). Turn-of-month and pre-holiday effects on stock returns: Some international evidence. Journal of Banking & Finance, 497-509.
Claessens, S., & Kose, M. A. (2009, March). What Is a Recession? Retrieved from International Monetary Fund:
http://www.imf.org/external/pubs/ft/fandd/2009/03/basics.htm
Fama, E. F. (1970). Efficient Capital Markets: A Review of Theory and Empirical Work. The Journal of Finance, 383-417.
Fama, E. F. (1998, September 1). Market efficiency, long-term returns, and behavioral finance. Journal of Financial Economics, 49(3), 283–306.
Kunkel, R. A., Compton, W. S., & Beyer, S. (2003). The turn-of-the-month effect still lives: the international evidence. International Review of Financial Analysis, 207-221.
Lakonishok, J., & Smidt, S. (1988). Are Seasonal Anomalies Real? A Ninety-Year Perspective. The Review of Financial Studies, 403-425.
McConnell, J. J., & Wei, X. (2008). Equity Returns at the Turn of the Month. Financial Analysts Journal, 49-64.
18 Ogden, J. P. (1990). Turn-of-Month Evaluations of Liquid Profits and Stock Returns: A
Common Explanation for the Monthly and January Effects. The Journal of Finance, 1259-1272.
Schwert, G. (2003). Anomalies and market efficiency. In G. Schwert, Handbook of the
Economics of Finance (pp. 939–974). Cambridge, MA: National Bureau of Economic
Research.
Sharma, S. S., & Narayan, P. K. (2014). New evidence on turn-of-the-month effects. Journal of International Financial Markets, Institutions and Money, 92-108.
Shefrin, H. (2002). Beyond greed and fear: Understanding behavioral finance and the psychology of investing. New York: Oxford University Press on Demand.
Ziemba, W. T. (1991). Japanese security market regularities: Monthly, turn-of-the-month and year, holiday and golden week effects. Japan and the World Economy, 119-146. Zwergel, B. (2010). On the exploitability of the turn-of-the-month effect–an international
