Calendar Effects and the Foreign Exchange

Markets: A Matter of Efficiency?

Master Thesis – June 2012

Simon Kruithof – 2104563

VRIJE UNIVERSITEIT AMSTERDAM

DEPARTMENT: BUSINESS ADMINISTRATION – FINANCIAL MANAGEMENT
SUPERVISOR: PROF. DR. P.A. STORK
Thesis – Calendar Effects and the Foreign Exchange Markets: a matter of efficiency? Simon Kruithof

Abstract
This study examines calendar effects within efficient and weak-form efficient currency
exchange rates. The Efficient Market Hypothesis and the Random Walk Theory proclaim that
calendar effects should not exist in an efficient market. This research finds evidence on existence of
day-of-the-week effects, turn-of-the-month effects, pre-holiday effects, Friday-the-thirteenth effects
and January effects. These effects are mainly present in weak-form efficient markets, such as the
Taiwan Dollar, Thai Baht and Singapore Dollar. Additionally, no significant effects are present in an
efficient market, such as the Japanese Yen / US Dollar exchange rate. These findings support the
Efficient Market Hypothesis introduced by Fama, however, it cannot be seen as an affirmative result
for the Random Walk Theory.

2
Thesis – Calendar Effects and the Foreign Exchange Markets: a matter of efficiency? Simon Kruithof

1. Introduction

Unusual patterns in returns on financial markets have been of interest to researchers and
investors alike. According to Fama (1970) and his Efficient Market Hypothesis, anomalies should not
exist and will disappear once the market understands they exist. Prices always reflect all available
information and, thus, the fair value of the product. According to the Random Walk Hypothesis
(Malkiel 1973), stocks or other financial products have an equally even chance of decreasing as well
as increasing in price. However, generally the price will increase over time. When taking the Random
Walk Hypothesis and Efficient Market Hypothesis into account, anomalies in returns should not exist.
Nevertheless, numerous studies have been performed on calendar effects, anomalies on certain days
or times and, occasionally, showed the existence of anomalies or proved that an anomaly was not
present or disappeared over time. Are these anomalies the cause of behavioural finance or are they,
just as some researchers (i.e. Fama 1998) proclaim, the product of extensive data mining or wrongly
performed studies?

Calendar effects and anomalies have been researched extensively within stock market
returns across the world. An overview presented in Appendix 1 generated by Marquering et al.
(2006), illustrates some of the empirical findings on calendar effects. Identified calendar effects
include the day-of-the-week effect (sometimes mentioned as the weekend effect), the turn-of-the-
month effect, the holiday effect and the January effect. Some less known effects are the Halloween
effect, Friday-the-thirteenth-effect and effects for Asian countries such as the Chinese New Year
effect and Lunar calendar effects. In paragraph 2 a closer look on outcomes of these studies on stock
markers and some foreign currency exchange markets is presented on these calendar effects.

As mentioned earlier, research on calendar effects or anomalies have been performed much
more on the stock market than on the foreign exchange markets. Foreign exchange markets are the
most liquid markets in the world. Furthermore, foreign exchange markets are the largest and most
dynamic segment of financial markets. According to Vieira et al. (2007) foreign exchange markets are
the first to demonstrate the characteristics of integrated markets and thus reflect an efficient
market. According to Faruqee et al. (1994) exchange rates on the Yen and the Dollar are determined
by the following long-run determinants; “The structural factors underlying each country's net trade
and net foreign asset positions determine the long-run path for the real value of the yen and dollar
exchange.” This underlines the relation between a country’s financial position and the value of its
currency.

3
Thesis – Calendar Effects and the Foreign Exchange Markets: a matter of efficiency? Simon Kruithof

Most research on calendar effects within foreign exchange rates focus on daily patterns or
the-day-of-the-week effect, Yu et al. (2006), Thatcher et al. (1998), Hilliard et al. (1992), Aydogan et
al. (2003), Yamori (2003), McFarland et al. (1982), Ke et al. (2007), Berument et al. (2007). An
exception to this is a research done on anomalies on the time of the day and intraday trading in
foreign exchange markets, Ronaldo (2009). This shows that the amount of research performed on
calendar effects on foreign exchange markets is relatively limited. Furthermore, most of the time it
focuses on just one calendar effect, which is the day-of-the-week effect (sometimes referred to as
the weekend effect; high returns on Friday and abnormal negative returns on Monday).

To determine whether there are still significant calendar effects present, one should look at
currency exchanges in an efficient market versus a weak form or non-efficient market. Additionaly,
this research adds value on whether there are more calendar effects present in currency cross rates
apart from the day-of-the-week effect, such as a holiday effect, a January effect, turn-of-the-month
effect or a Friday-the-thirteenth effect. The outcome of this research can also add value on whether
calendar effects are present in foreign exchange rates markets and, additionally, if they are a matter
of efficiency in the market or whether they are a result of behavioural finance. Paragraph 2 of the
literature review describes several calendar effects and the outcomes identified in previous studies.
Paragraph 3 gives a brief description of the foreign exchange rates markets. Results on efficiency of
markets are discussed in Paragraph 4. Furthermore, selection criteria, data, methodology and results
of this research are presented in paragraph 5.

4
Thesis – Calendar Effects and the Foreign Exchange Markets: a matter of efficiency? Simon Kruithof

2. Calendar Effects

There are several calendar effects and lots of research has been done in order to see
whether these anomalies exist. Sometimes calendar effects have been successfully identified and
sometimes these findings were disproved. In this paragraph an overview of the found calendar
anomalies is presented in previous research, on stock markets and foreign exchange rates markets.

2.1 The day-of-the-week effect

Probably the most investigated calendar effect is the day-of-the-week effect. This effect can
consist of abnormal returns on a single day. Furthermore, most of the studies focus on the volatility
of the returns on the weekdays. This volatility says something about the risks, which is present during
that particular day. The volatility could be extra evidence on results, because it shows how much risk
one takes. A part of the day-of-the-week effect is the weekend effect. The weekend effect is the
puzzle, that returns on Friday are higher in comparison to returns on Monday. This is quite puzzling,
because one would expect that returns on Monday cover three days of return. An illustrative quote
from Choudry (2000) “According to French (1980) stock prices should rise higher on Mondays than
on other days, because the time between the close of trading on Friday and the close on Monday is
three days. Accordingly, Monday returns should be three times higher than other weekday returns.
However, other evidence prove: highest on Friday and lowest on Monday.”

Several studies have shown a significant lower result on Monday on stock markets, a few of
these studies are Berument et al. (2001), Choudry (2000), Agrawal et al. (1994) and Kenourgios et al.
(2008). The outcome on volatility measured differs per study. Studies focusing on currency exchange
rates exhibit a similar pattern for the weekend effect. Yamori et al. (2003) finds a weekend effect for
the Yen/Dollar market between 1973 and 1989. However, the effect disappears in the 1990s. The
possible explanation for the disappearance of this effect is the financial deregulation in Japan, which
has made foreign currency markets more efficient. Berument et al. (2007) and Hilliard et al (1992)
distinguish a significant lower return on Monday for the Turkish foreign exchange market, interbank
foreign exchange market and listed currency options market. Thatcher et al. (1998) finds an even
more significant return on Wednesdays on the Dollar/Sterling market. Aydogan et al. (2003) adds the
Tuesday to the Wednesday with abnormal higher returns, but only on the free market.

5
Thesis – Calendar Effects and the Foreign Exchange Markets: a matter of efficiency? Simon Kruithof

Finally, Yu et al. (2006) does not identify a weekend effect as well, however, Yu et al. (2006) does
detect significant higher returns on Tuesdays, which is the same for Aydogan et al. (2003).

A contradicting result on the weekend effect is identified by McFarland et al. (1982), which is
a significant higher average price change on Mondays and Wednesdays, instead of significant lower
price change on Monday. So (1987) has re-examined McFarland et al.’s outcomes and comes to an
affirmative result. However, So (1987) does not see this as an anomaly, but as a difference in risk
identities of each currency and country. Ke et al. (2007) also found a higher return on Monday,
however, depending on the number of trading days it can also be Friday, which seems to be a very
contradictory outcome.

2.2 The turn-of-the-month effect

The turn-of-the-month effect is an anomaly in which returns in the last trading days of the
month and the first two or three days of the next month show a significant abnormal positive return.
The first one to identify this effect was Ariel (1987) and according to Marquering et al. (2006) this
effect has increased since then. Favourable evidence of this effect has been provided by Kunkel et al.
(2003). This study re-examined the turn-of-the-month effect on stock markets in 19 different
countries between 1988 and 2000. It showed that 16 out of 19 countries had a significant turn-of-the
month effect and outperformed the rest of the month. Interestingly, the turn-of-the-month effect
was significantly negative in Japan and between 1994 and 2000 the effect disappeared in the United
States. Agrawal et al. (1994) identified significant better results in 9 countries on the last trading days
of the month, however, these results were present in the 1970’s, but did not exist in the 1980’s.

The turn-of-the-month effect in exchange rates has been researched by Aydogan et al.
(2010). This study shows a lower change in exchange rates in the week before the turn of the month
and shows that the changes are significantly higher during the first five days of the new month.
Aydogan et al. (2010) researched the Turkish Lira with the Deutsche Mark and US Dollar. Yamori et
al. (2003) identified a turn-of-the-month effect for the first day of the month, but not the last day.
However, until the 1990’s a turn-of-the month effect existed for the Yen/Dollar market.

6
Thesis – Calendar Effects and the Foreign Exchange Markets: a matter of efficiency? Simon Kruithof

2.3 Holiday effect

The holiday effect is the effect of abnormal returns prior to the start of a holiday such as
Christmas, Easter, New Year’s day or any other local holidays. Besides holidays with a fixed date on
the Gregorian calendar there are also holidays which have a patterns according to the Lunar
calendar. These lunar calendar holidays, such as Chinese New Year, have a different date each year
on the Gregorian calendar. Holiday effects have already been identified in the 1980s by Ariel (1990),
Lakonishok et al. (1988), Liano et al. (1992) on the US stock market. Ariel (1990) finds no evidence
that pre-holiday effects are the cause of other calendar anomalies, such a day-of-week effect or a
turn-of-the-year effect. More recent studies that identified pre-holiday effect on stock markets are
Meneu et al. (2004), Cadsby et al. (1992) and Mills et al. (1995). Cadsby et al. (1992 finds evidence of
a pre-holiday effect in the Hong Kong, Japanese, Australian, and Canadian markets, but not on a
European market. Nonetheless, Meneu et al. (2004) does find evidence of significant pre-holiday
effects on stocks in the Spanish stock exchange. Mills et al. (1995) support this by finding evidence of
a pre-holiday effect on several European stock markets.

Yamori et al. (2003) identified a pre and post-holiday effect for the Japanese currency market
in the 1980s, but the anomaly disappeared in the 1990s. Aydogan et al. (2010) finds a negative pre-
holiday return within the Turkish foreign exchange market. The holiday effect has not been
extensively researched on the foreign exchange rate markets.

2.4 Friday-the-thirteenth effect

Superstition has existed for a long time. One can find it everywhere. For example, never walk
under a ladder, black cats bring misfortune or never proclaim something on yourself without
knocking on wood. Friday the thirteenth is one of these superstitions, originating from myths and
predominantly present in the Judaeo-Christian world. Friday the thirteenth is associated with bad
luck. One might think that besides other behavioural aspects, superstitions influence the returns on a
stock market or foreign exchange market. Investors or traders believing in a Friday-the-thirteenth
effect point out on the downfall of the Dow Jones on Friday the thirteenth 1989, when the market
made a crash of 6.8%. Other “evidence” is the downfall on Friday the thirteenth 2000 when 69% of
value of the Heartland Mutual fund disappeared. These are extreme examples on the stock market,
but they are used by investors and traders to reconfirm superstition towards Friday the thirteenth

7
Thesis – Calendar Effects and the Foreign Exchange Markets: a matter of efficiency? Simon Kruithof

(Zweig 2009). Zweig (2009) adds that Friday the thirteenth usually does not give a better or worse
performance on average in comparison to other days. Moreover, Zweig states that investors or
traders would be foolish to act upon a superstition.

There is contradicting evidence for the existence of the Friday-the-thirteenth effect on stock
markets. Kolb et al. (1987) found an average negative mean return for Friday the thirteenth.
Chamberlain et al.(1991) and Dyl et al.(1988) disproved this and found an actual positive return on
Friday the thirteenth. Lucey et al. (2000) found that even though in some cases there is no Friday-
the-thirteenth effect, however, internationally there are many cases where there is a significant
Friday-the-thirteenth effect. A more recent study by Patel (2009) does not find any evidence of a
Friday-the-thirteenth effect on the US stock market between 1950 and 2007, which is contradicting
to what Keef et al. (2011) found. Keef et al. (2011) found between 2000 and 2008 an effect on Friday
the thirteenth when the preceding day has a negative return.

2.5 January effect

The January effect is the effect of higher abnormal returns in the month of January. Most of
the times experienced before the middle of the month of January. This gives an arbitrary opportunity
to investors. The arbitrary opportunity could consist of buying in the months previous to January and
selling in January. There is doubt whether the January effect still exists. Mehdian et al. (2002)
conclude that the January effect has diminished in the US stock market after the stock crash in 1987.
Furthermore, Fountas et al. (2002) affirm the fact that the January effect is not present in 18
emerging stock markets. However, a recent study by Moller et al. (2007) on the January effect on
stock markets in the USA, found that the effect is still present, but changed over time. Moller et al.
(2007) identified that the peak of higher abnormal returns in January are earlier in the years between
1995-2004, than for the years 1965-1994. Moreover, Moller et al. (2007) find significant negative
abnormal returns for the latter half of January.
January effects in currency exchange rates are not widely researched which is acknowledged
by Dumas et al. (1993). Dumas et al. (1993) did find a significant role for the month of January in the
returns of the Japanese Yen and the Deutsche Mark.

In conclusion, it becomes evident that anomalies were of interest to a lot of researchers.

Sometimes calendar effects were identified and sometimes disproved or they just disappeared.

8
Thesis – Calendar Effects and the Foreign Exchange Markets: a matter of efficiency? Simon Kruithof

From the review, it becomes apparent that efficient markets should not have anomalies or they
should be arbitraged. Furthermore, it is evident that apart from the day-of-the-week effects, not a lot
of anomalies have been studied on the foreign exchange market. Is this because investors see foreign
exchange markets as efficient or are investors still working with trading patterns on this market?
Moreover, the above reviewed calendar effects are not extensive. Other calendar effects
consist of a size effect (for stocks), turn-of-the-year effect, half-month effects, Halloween effect and
several superstition effects. (For example Chinese superstition; 4,5,6 and 7 are unlucky numbers and
2,3 and 8 are lucky numbers in the Chinese culture.)

3. Foreign exchange markets

The average daily turnover of foreign exchange markets is estimated at 2 trillion dollar per
day (BIS, 2010) and the foreign exchange market is one of the fastest growing markets. Together
London and New York are the most dominant players in the foreign exchange markets. Emerging
countries such as Brazil, Thailand, Turkey and India have doubled the annual turnover in foreign
currency trading since the 1990’s (King et al. 2011). The Dollar is the most traded currency, followed
by the Euro, Yen and UK Pound. Trading in currencies consists of foreign exchange swaps, spot
trading and forwards.

As the most liquid market in the world, foreign exchange markets have impact on the world
economy. For example exchange rates:

 Affect inflation through the cost of imports and commodity prices

 Affect international capital flows through the risks and returns of different assets
 Are a major focus for policymakers, the public and the media (King et al. 2011).

According to research (Thomas et al. 2008) 90 percent of traders on the foreign exchange
market tend to trade based on technical rules. These technical rules can consist of patterns, such as a
positive effect is likely to be followed by another positive effect. Some other technical rules examples
consist of Fibonacci number sequence, “double bottoms” or “head and shoulders” patterns in foreign
exchange rates (Thomas et al. 2008). Subsequently it can be noted, that traders do believe in some
sort of trend or patterns in a currency exchange rate.

9
Thesis – Calendar Effects and the Foreign Exchange Markets: a matter of efficiency? Simon Kruithof

4. Efficient Market vs. Non-Efficient Market

From research, the Efficient Market Hypothesis and Random Walk Hypothesis it is apparent
that anomalies should not exist in efficient markets. What is efficiency and which markets are still not
strong-form efficient?

Fama (1970) describes an efficient market as a market that always fully reflects available
information. In the case of the foreign exchange markets there is a strong belief initialized by
Friedman (1953) that exchange rates fully reflect long-run determinants of currency values. On the
other hand, some researchers believe that the market and in particular the foreign exchange market
cannot be considered efficient, such as Froot et al.(1990). Froot et al. (1990) found evidence of
inefficiency in currency exchange markets. This inefficiency, or anomalies, could not be explained by
volatility or risk that was experienced during these anomalies. A more recent study on exchange rate
bubbles by Jarrow et al. (2010) finds evidence for exchange rate bubbles in the US Dollar/Euro
market. Jarrow et al. (2010) states that one of their insights is that exchange rate bubbles are caused
by the price level bubbles in either one of the currencies. To the contrary, Ibrahim et al. (2011) finds
no evidence of inefficiency or a rejection of the random walk hypothesis in a number of foreign
exchange rates.
Market efficiency comes in three forms according to the Efficient Market Hypothesis:
it begins with weak-form efficiency where historical market prices are fully reflected. Secondly,
Semistrong-form efficient market where all public information is reflected in the price. And lastly, a
strong form market where all information (even private information) is reflected in the price
(Fama,1970).

Al-Khazali et al. (2011) performed a study on the market efficiency in the Asian-Pacific region.
From their study they found a few weak-form efficiency currencies. These are Korean Won,
Malaysian Ringgit, Singapore Dollar, Thai Baht and Taiwan Dollar. They excluded the Chinese Yuan
and the Hong Kong Dollar, because the Chinese government heavily regulates them. Azad (2009)
found the same results on these currencies, however, Azad (2009) states that the Malaysian Ringgit
cannot be seen as a free-floating currency due to government regulation. Chiang et al. (2010) even
finds evidence for the Taiwan Dollar to be completely inefficient and that the other currencies
mentioned above are weak-form efficient markets.

10
Thesis – Calendar Effects and the Foreign Exchange Markets: a matter of efficiency? Simon Kruithof

Oh et al. (2007) affirmed the conclusion that Asian foreign exchange markets, except for Japan, are
less efficient than North American and European foreign exchange markets.

The most traded currencies like the Dollar, Euro and Yen for instance are identified as strong
efficient currencies (Azad 2009). According to Thomas at al. (2008) less liquid currencies tend to
move towards more efficiency in the past ten years, such currencies include the Singapore Dollar,
Taiwan Dollar and Thai Baht. Additionally, Thomas et al. (2008) states that the Japanese Yen and US
Dollar can be considered strong efficient currencies.

Conclusion Literature Review

From the literature review it can be concluded, that apart from the day-of-the-week effect
not a lot of effects have been researched within the foreign exchange rates. The effects noted most
of the times disappear with the years. One of the most common explanations for the disappearance
of calendar effects in recent decades is the integration of the markets throughout the World. This is
in line with Fama’s Efficient Market Hypothesis. A few of the effects, such as Friday-the-thirteenth
effect, have not been researched on the foreign exchange market at all.

Furthermore, from paragraph 3 and 4 it becomes evident that even though the foreign
exchange market is highly liquid there are still weak-form efficient currency markets. The most
traded and efficient currencies are the Dollar, Yen, Euro and Pound. Asian foreign exchange markets
are developing quickly, however, they still show signs of weak-form efficiency. In addition, traders on
the foreign exchange market still make use of technical rules to predict market developments. These
traders still believe in a pattern, which should not be the case if a trader would be fully rational and
would acknowledge the random walk of a currency. This concludes that anomalies and in particular
calendar anomalies could still be present in foreign exchange rates.

The following consideration arises: Will there be a difference between strong and weak-form
efficient currency markets in terms of existing calendar anomalies? This consideration gives the
following main research question:

“Do calendar effects noted within stock exchanges exist within efficient and weak-form efficient
currency exchange rates?”

11
Thesis – Calendar Effects and the Foreign Exchange Markets: a matter of efficiency? Simon Kruithof

Sub-questions:
 Does a day-of-the-week effect exist within currency exchange rates?
 Does a turn-of-the-month effect exist within currency exchange rates?
 Does a holiday effect exist within currency exchange rates?
 Does a Friday-the-thirteenth effect exist within currency exchange rates?
 Does a January effect exist within currency exchange rates?

12
Thesis – Calendar Effects and the Foreign Exchange Markets: a matter of efficiency? Simon Kruithof

5. Data and Methodology

5.1 Data

To investigate calendar effects on foreign exchange rates within weak-form efficient markets
and strong or efficient markets there needs to be a selection in currencies. These currencies need to
be traded in the past 15 years and should be considered as a weak-form efficient currency and on the
opposite currencies that are considered to be strong efficient. To be able to compare the results
better, currencies from the Asian market have been selected. From paragraph 4 it can be concluded
that the Thai Baht, Singapore Dollar and Taiwan Dollar are to be considered as a weak-from currency.
The Dollar and the Yen are both currencies that are the most traded around the world (1st and
second place). In two-thirds of all foreign exchange transactions the Dollar is involved. Therefore, the
Dollar and Yen are selected as the efficient or strong-form efficient currencies. Table 1 presents the
selected currency exchange rates ranging from weak-form to strong-form or to be considered
efficient.

Table 1 Selected Currency Exchange Rates

Weak form Efficient Strong-form Efficient
Singapore Dollar / Thai Baht Taiwan Dollar/ US Dollar Japanese Yen/ US Dollar
Singapore Dollar / 100 Taiwan Singapore Dollar/ US Dollar
Dollar
Thai Baht / US Dollar
Singapore Dollar/ 100 Yen

These exchange rates are daily spot rates retrieved from DATASTREAM, which have been
taken from the Bankers Trust, Bank of Tokyo and DBS Bank Singapore. These fixed exchange rates
which were obtained by DATASTREAM daily, respectively at 6 pm New York time, 10 am Tokyo time
and 12 pm Singapore time. The time period is selected from 1995-2011, which is just before the
Asian financial crisis of 1997.
The daily spot rates have been adapted to represent the daily return of each currency
exchange rate. The returns are calculated as follows:
( ⁄ )

13
Thesis – Calendar Effects and the Foreign Exchange Markets: a matter of efficiency? Simon Kruithof

The fluctuations of the currency exchange rates selected seem to follow a rather similar
pattern, which can be seen in figure 1. From figure 1 it can be seen that there is a steep
appreciation/depreciation in 1997. Especially the Thai Baht depreciates, which shows in the Thai Baht
/ US Dollar and Singapore Dollar / Thai Baht. These two currency exchange rates seem to mirror each
other. The Asian crisis is less strong for the Taiwan Dollar and the Singapore Dollar and Japanese Yen
versus the US Dollar. Between 2000 and 2006 there are no dramatic depreciations or appreciations
to be seen within the currency exchange rates. The financial crisis in 2007 and 2008 is clearly visible
and had a negative impact on the value of the US Dollar.
The descriptive statistics of each currency exchange rate presented in table 2 have a few
similarities between each other. Firstly, the kurtosis of each currency exchange rate is higher than 3,
which indicates leptokurtosis. This leptokurtosis is an indicator that it is more likely that the currency
exchange rate returns will exhibit large and very small price changes. This is also referred to as a
distribution with fat tails. This leptokurtosis is in line with the conclusions from figure 1. There it is
visible that the Thai Baht/ US Dollar and Singapore Dollar / Thai Baht have a very explosive change
during the Asian crisis which can be seen in the highest kurtosis of all exchange rates as well as the
highest maximum and minimum. This excess kurtosis is also clearly visible in the QQ-plots which
exhibits for all currency exchange rates an S-shape, which can be found in Appendix 2. The QQ-plot
for Thai Baht/ US Dollar has a somewhat different pattern, which displays that there are quite some
returns located around 0. Secondly, all currency exchange rates are not normally distributed which
indicates the Jarque-Bera test, where the null hypothesis of a normal distribution is rejected with 1
percent significance.
The standard deviation is the highest for Thai Baht / US Dollar and the lowest for Taiwan
Dollar / US Dollar. In Appendix 3 the returns are plotted, which gives a similar overview to figure 1,
where the Asian crisis is visible as well as the financial crisis in the United States. Furthermore, the
returns do no exhibit a stationary pattern.
The Asian crisis started in July 1997 (IMF,1998) and affected the economy of Thailand spread
around other countries, such as the South-Korea, Philippines, Indonesia, Malaysia and even Japan.
This resulted in the collapse of the Thai Baht and forced the Thai government to impose restrictions
onto foreign exchange transactions. Before 1997 the Thai Baht was pegged to the US Dollar,
however, the Thai government decided to let the Thai Baht float.
The data has been divided into three periods; these are 1995-2000, 2001-2006 and 2007-
2011. This has been done for two reasons; firstly, because there are two abnormal periods to be seen
in the data in the form of a crisis. Secondly, because it is also interesting to see whether the calendar
effects are present in one period and disappear in the subsequent period.

14
Thesis – Calendar Effects and the Foreign Exchange Markets: a matter of efficiency? Simon Kruithof

Figure 1 Return per currency exchange rate

15
Thesis – Calendar Effects and the Foreign Exchange Markets: a matter of efficiency? Simon Kruithof

Table 2 Descriptive Statistics per Currency Exchange Rate 1995-2011

Mean Median Maximum Minimum Std. Dev. Skewness Kurtosis Jarque-Bera Probability
Taiwan Dollar /US Dollar 0.001408 0.000000 1.453841 -1.17896 0.128696 0.607095 19.58209 51072.27 0.000000 ***

Singapore Dollar/US Dollar -0.00114 -0.00244 1.298344 -1.88881 0.163965 -0.25281 13.94419 22175.72 0.000000 ***

Thai Baht/US Dollar 0.002244 0.00000 8.154315 -4.79182 0.326387 3.306077 126.2637 2815156 0.000000 ***

Singapore Dollar/ Thai Baht -0.00341 0.00000 3.078033 -6.90809 0.289593 -3.23778 94.51407 1554996 0.000000 ***

Singapore Dollar / 100 Taiwan Dollar -0.00255 0.00000 2.34402 -2.59563 0.188164 -0.66912 31.83594 153952.6 0.000000 ***

Japanese Yen / US Dollar -0.00258 0.002824 2.694093 -3.01828 0.310391 -0.38198 9.047783 6865.184 0.000000 ***

Singapore Dollar /100 Japanese Yen 0.001386 0.000000 3.012245 -3.05393 0.307136 0.53976 12.91363 18372.56 0.000000 ***

*** statistically significant at 1%

16
Thesis– Calendar Effects and the Foreign Exchange Markets: a matter of efficiency? Simon Kruithof

5.2 Methodology

From the descriptive analysis it became clear that the time series for the currency exchange
rate exhibit leptokurtosis and are not normally distributed. The Ljung-Box Q statistic is significant for
all currency exchange rates, which indicate autocorrelation. Furthermore, it can be noted that
heteroskedasticity can be expected within a time series and especially within currency exchange
rates. Therefore, a simple OLS regression which has been used by many researchers in the past to
detect calendar effects will not be sufficient. Additionally, an OLS regression does not provide
information on the volatility of a certain day or period.
The GARCH (Generalized Autoregressive Conditional Heteroskedasticity) originally by Baillie
and Bollerslev (1989) is capable of assuming an error variance as well as accounting for all
shortcomings of the data such as heteroskedacity, identified autocorrelation and leptokurtosis.
A GARCH model or ARCH is conditional and time-varying. The model consists of conditional q lagged
errors and p lags on conditional variance. The GARCH model is preferred over the ARCH model,
because an ARCH needs a high level of q, which can make the model over fitting.
A GARCH(1,1) model will be used instead of a higher order GARCH or ARCH model, such as
GARCH(2,2). Higher ARCH models can make the model over fitting and through research GARCH(1,1)
it has proven to be a suitable model (Brooks et al. 2009). Equation 1 presents the model which will
be used as a standard for all other models to measure the calendar effects:

Eq. 1

is the log return on spot exchange rates on time t. The is the intercept in this case
representing Tuesdays. …. are the coefficients of the excess returns or loss of the other
weekdays in comparison to the intercept, which represents Tuesday in this case.
… are the dummies where represents Monday…. and represents Friday.
Furthermore, the lagged return ( ) is estimated by an Autoregressive order or AR
variable. An AR is preferred over a lagged return, because it takes into account the serial
dependence. After extensive testing and analysis of the Akaike’s Information Criterion an AR of an
order 5 (AR(5)) achieves the best results on average for all currency exchange rates.

17
Thesis– Calendar Effects and the Foreign Exchange Markets: a matter of efficiency? Simon Kruithof

Formula for the error variance:

Eq. 2

∑ ∑

( )
is the conditional variance and represents the constant. The ARCH coefficient represents
the development in short-term volatility and or GARCH coefficient represents the development
in long-term volatility. All these parameters have the restriction to be positive.
The aim of this paper is to examine whether there are calendar effects present in foreign
exchange rate markets between weak-form and efficient markets. Besides a pattern in the mean
returns it is also interesting to measure this in the volatility level. Therefore, the variance equation
needs to be expanded with the days of the week.

Eq. 3

∑ ∑

A Ljung-Box test will be performed with lags 5, 10 and 20 to test for any remaining autocorrelation in
the residuals at α=0.. Furthermore, to test for remaining heteroskedacity a Lagrange Multiplier test
will be performed with lags 5, 10 and 20.
To verify whether the calendar effects found in the currency exchange rates are noticeably
different from 0, a Wald test is implemented.

6. Empirical Results

In the following paragraph the empirical results on the day-of-the-week effect, turn-of-month
effect, Holiday effect, Friday-the-thirteenth effect and the January effect are presented.

6.1 The day-of-the-week effect

The descriptive statistics for each day and each period are presented in Appendix 4. Overall
the standard deviation of the period between 1995-2000 exhibits the highest standard deviation of
all three periods. There is one exception, which is the Singapore Dollar/ Japanese Yen currency

18
Thesis– Calendar Effects and the Foreign Exchange Markets: a matter of efficiency? Simon Kruithof

exchange rates where the last period between 2007-2011 has the biggest standard deviation.
Another irregularity in the first period is the extreme high standard deviation on Wednesday of the
Singapore Dollar / Thai Baht. The lowest standard deviation is observed in the second period of 2001-
2006, which appears to be a rather stable period. There are no distinctive patterns to be noted
within the means of each day. Therefore, nothing can be concluded about a possible weekend effect
or significant returns on a particular day.

The equation 1 for the day-of-the-week effect has dummy variables for each day of the week.
The dummy variable for Tuesday has been suppressed in order to avoid a dummy trap. This means
that the constant α represents the return on a Tuesday, subsequently the dummy coefficients
…. represent the excess return or loss versus the constant. The following Wald test has been
performed whether found significant effects in the coefficients significantly differ from 0. The Wald
tests the following:

Eq. 4

1995-2000

In the first period (1995-2000) there are only a few significant returns measured, which can
be seen in table 3. A highly significant return on Friday is measured for the Taiwan Dollar / US Dollar
exchange rate, this result is confirmed by the Wald test. Other Fridays which are significant are the
Friday return on the Japanese Yen / US Dollar exchange rate. Other abnormal returns in the first
period are a significant return on Thursday Singapore Dollar / US Dollar and less significant within the
Japanese Yen / US Dollar exchange rate, which was also found by Yamori et al.(2003). The abnormal
return on Wednesday on the Thai Baht / US Dollar exchange market has been found earlier by
Yamori et al. (2003).
Volatility is significant for quite a few days throughout the week for all currency exchange
rates in the first period. For four out of seven exchange rates the volatility on all days is lower in
comparison to Tuesday. All currency exchange rates with the US Dollar exhibit a lower volatility on
Monday.
Explosiveness is identified in a few exchange rates: Taiwan Dollar / US Dollar, Singapore
Dollar / Taiwan Dollar and Singapore Dollar / US Dollar. This can be due to the Asian financial crisis,
which is not completely taken into account in the model.

19
Thesis– Calendar Effects and the Foreign Exchange Markets: a matter of efficiency? Simon Kruithof

Table 3 Empirical Results day-of-the-week effect

Singapore Dollar / Thai Baht Singapore Dollar / Taiwan Dollar Taiwan Dollar / US Dollar Singapore Dollar / US Dollar
1995-2000 2001-2006 2007-2011 1995-2000 2001-2006 2007-2011 1995-2000 2001-2006 2007-2011 1995-2000 2001-2006 2007-2011
α 0.013 -0.001 0.011 -0.010 -0.001 -0.017 0.004 ** 0.001 0.013 ** 0.013 ** -0.002 -0.004
Monday -0.016 0.013 -0.034 * -0.012 0.008 0.010 0.003 0.000 -0.020 ** -0.014 * 0.012 -0.001
Wednesday 0.015 -0.012 -0.003 0.002 -0.016 0.025 * -0.003 -0.005 -0.029 *** -0.004 -0.012 -0.010
Thursday -0.012 0.009 -0.021 0.009 0.002 0.022 -0.005 -0.006 -0.018 ** -0.028 *** -0.009 -0.003
Friday -0.039 -0.003 -0.011 0.009 -0.009 0.008 -0.017 *** -0.005 -0.009 -0.005 -0.005 -0.017
Return(t-1) -0.001 -0.134 *** -0.157 *** -0.006 -0.178 *** -0.217 *** -0.003 -0.026 0.035 -0.021 -0.017 -0.001

Variance equation
ω 0.132 *** 0.000 -0.002 0.001 0.000 0.003 0.010 *** -0.004 *** 0.000 0.003 *** -0.001 0.001
λ 0.284 *** 0.092 *** 0.104 *** 0.407 *** 0.140 *** 0.065 *** 0.524 *** 0.068 *** 0.115 *** 0.070 *** 0.050 *** 0.044 ***
ϴ 0.551 *** 0.845 *** 0.897 *** 0.671 *** 0.711 *** 0.922 *** 0.647 *** 0.929 *** 0.880 *** 0.933 *** 0.897 *** 0.950 ***
Monday -0.116 *** 0.001 0.016 *** 0.004 *** 0.003 -0.004 -0.018 *** 0.008 *** 0.001 -0.005 *** 0.002 0.003
Wednesday -0.026 -0.001 -0.003 0.014 *** 0.004 ** 0.001 -0.017 *** 0.003 *** 0.002 * -0.004 *** 0.001 0.002
Thursday -0.193 *** 0.001 0.005 0.009 *** 0.006 *** -0.002 -0.009 *** 0.004 *** 0.004 *** -0.004 *** 0.004 *** -0.001
Friday -0.130 *** 0.002 -0.003 -0.015 *** 0.000 -0.007 ** 0.006 *** 0.004 *** -0.004 *** -0.002 *** -0.001 -0.007 ***

Tests
ARCH-LM(5) 2.153 4.144 1.000 1.106 5.416 5.341 1.803 3.060 3.122 10.452 3.126 5.043
ARCH-LM(10) 8.197 7.940 5.098 1.649 12.043 9.113 3.252 10.703 5.878 12.282 4.661 7.194
ARCH-LM(20) 31.957 * 16.297 17.540 2.808 25.755 19.328 6.069 18.333 19.183 18.657 18.297 19.204
Ljung-Box Q(5) 2.188 2.180 5.756 2.773 5.395 8.260 5.029 16.421 3.729 1.131 6.214 3.234
Ljung-Box Q(10) 8.981 12.838 15.811 5.719 12.971 13.124 5.916 25.356 16.397 * 6.961 10.699 13.053
Ljung-Box Q(20) 37.048 * 19.441 21.159 15.885 26.019 19.478 30.112 * 33.814 * 24.796 15.027 18.458 30.092

Wald 0.186 0.056 1.525 0.091 0.219 2.322 4.299 ** 0.552 8.741 *** 3.898 ** 0.246 0.582

Log likelihood -200.1 1040.9 123.7 477.8 990.5 552.6 1436.4 1403.1 917.7 993.9 1134.9 567.6
AIC 0.273 -1.314 -0.170 -0.595 -1.250 -0.828 -1.821 -1.778 -1.388 -1.255 -1.435 -0.851
BIC 0.317 -1.270 -0.118 -0.550 -1.206 -0.776 -1.777 -1.733 -1.336 -1.211 -1.390 -0.799

Standard errors are removed from this table for visibility reasons and are available upon request, * statistically significant at 10%, ** statistically significant at 5%, ***
statistically significant at 1%. The null hypothesis for the Wald test is that all dummies for the day-of-the-week effect are equal

20
Thesis– Calendar Effects and the Foreign Exchange Markets: a matter of efficiency? Simon Kruithof

Table 3 (continued) Empirical results day-of-the-week effect

Thai Baht / US Dollar Singapore Dollar / Japanese Yen Japanese Yen / US Dollar
1995-2000 2001-2006 2007-2011 1995-2000 2001-2006 2007-2011 1995-2000 2001-2006 2007-2011
α -0.026 0.000 -0.001 -0.004 0.004 0.022 0.010 -0.012 -0.031 *
Monday 0.029 -0.003 0.008 -0.015 -0.007 -0.013 -0.012 0.021 0.025
Wednesday 0.101 * 0.000 -0.005 0.008 -0.031 -0.055 * 0.007 0.018 0.044 *
Thursday 0.057 -0.019 ** -0.002 -0.011 0.000 -0.025 -0.043 * 0.001 0.027
Friday 0.038 -0.006 0.007 -0.017 -0.019 -0.023 0.047 * 0.022 0.027
Return(t-1) 0.065 ** 0.050 * -0.058 * -0.013 -0.048 * -0.073 ** -0.014 -0.015 -0.038

Variance equation
ω 0.181 *** 0.002 ** -0.001 *** -0.007 0.006 * 0.028 *** 0.007 0.004 0.008
λ 0.261 *** 0.156 *** 0.582 *** 0.148 *** 0.076 *** 0.111 *** 0.043 *** 0.023 *** 0.090 ***
ϴ 0.589 *** 0.776 *** 0.684 *** 0.786 *** 0.866 *** 0.876 *** 0.951 *** 0.950 *** 0.855 ***
Monday -0.181 *** -0.001 0.001 ** 0.041 *** -0.015 *** -0.041 *** -0.024 ** -0.019 ** -0.017
Wednesday -0.021 -0.004 ** 0.003 *** 0.004 0.000 -0.029 ** 0.016 0.000 0.019
Thursday -0.268 *** -0.002 0.001 0.009 -0.010 ** -0.010 -0.026 *** -0.002 0.002
Friday -0.186 *** 0.002 0.001 0.018 ** 0.006 -0.044 *** 0.004 0.011 -0.017 *

Tests
ARCH-LM(5) 8.473 0.909 0.897 11.216 11.841 9.611 * 15.712 * 2.021 1.919
ARCH-LM(10) 22.021 ** 2.699 1.587 15.180 17.120 13.707 17.824 6.280 4.122
ARCH-LM(20) 24.532 11.680 2.722 26.611 27.191 26.631 29.052 17.371 6.487
Ljung-Box Q(5) 5.674 9.345 * 8.609 * 4.400 2.743 4.781 1.411 1.802 3.850
Ljung-Box Q(10) 15.500 * 12.542 13.677 6.217 6.311 8.329 7.832 5.018 12.463
Ljung-Box Q(20) 25.812 23.188 17.771 11.717 11.078 22.397 24.463 12.815 24.975

Wald 1.637 1.007 0.090 0.227 1.428 1.621 0.000 1.061 2.364

Log likelihood -321.2 1038.0 484.3 -341.0 354.2 -385.0 -447.4 -22.4 -295.5
AIC 0.428 -1.311 -0.723 0.453 -0.436 0.610 0.589 0.045 0.473
BIC 0.472 -1.266 -0.671 0.498 -0.392 0.662 0.634 0.090 0.525

Standard errors are removed from this table for visibility reasons and are available upon request, * statistically significant at 10%, ** statistically significant at 5%, ***
statistically significant at 1%. The null hypothesis for the Wald test is that all dummies for the day-of-the-week effect are equal

21
Thesis– Calendar Effects and the Foreign Exchange Markets: a matter of efficiency? Simon Kruithof

2001-2006

This period does not exhibit any daily patterns except for a Thursday effect within the Thai
Baht / US dollar exchange rate. Monday exhibits a lower significant volatility within Japanese Yen /
US Dollar and Singapore Dollar / Japanese Yen. In addition, volatility is most of the times higher
throughout the week in comparison to Tuesday.

In comparison to period 1 there is no explosiveness left. Furthermore, the models seem to fit
the best in the second period, which can be seen through the higher values of the Log Likelihood,
Akaike Information Criterion(AIC) and Bayesian Information Criterion(BIC).

2007-2011

In the last years during the financial crisis a few day effects are identified. Firstly, the
Wednesday appears to have a significant abnormal return in four currency exchange rates, which is
within Singapore Dollar / Japanese Yen, Japanese Yen / US Dollar, Singapore Dollar/ Taiwan Dollar,
and Taiwan Dollar / US Dollar. Furthermore, an abnormal return on Monday is identified twice as
well as two times an abnormal return on Thursday. The Taiwan Dollar / US Dollar demonstrate
abnormal returns on all weekdays except for Friday. The Wald test is rejected for the Taiwan Dollar /
US Dollar therefore, it can be concluded that these effects are present.
A significant lower volatility is measured on Friday for almost all currency exchange rates
with an exception for the Thai Baht / US Dollar and Singapore Dollar / Thai Baht. Monday’s volatility
is significantly lower in the Singapore Dollar/ Japanese Yen and Japanese Yen / US Dollar. In contrast
to Monday’s volatility of the Thai Baht / US Dollar and Singapore Dollar / Thai Baht which is
significantly higher.

There are some issues with autocorrelation in the residuals which is evident form the Ljung-
Box test. This autocorrelation is significant at α=0.10. However, the individual statistics for
autocorrelation and partial correlation did not show any big changes from the 0. Therefore, it is
extremely likely that the left autocorrelation is significant due to the large sample and
autocorrelations gathered around null. Mild explosive conditional variance is measured in some
variance equation, which can be a result of a shock which is not integrated well in the model; this has
a negative effect on the interpretation of the variance coefficients.

22
Thesis– Calendar Effects and the Foreign Exchange Markets: a matter of efficiency? Simon Kruithof

Overall it appears that day-of-the-week effects, if present, are more likely to occur during a
period of financial crisis. The period without a financial crisis, however, after the 9/11 attacks, does
not exhibit any significant daily patterns. The Taiwan Dollar / US Dollar displays the most abnormal
returns on weekdays, this was also found by Ke et al. (2007). Furthermore, the exchange rates for the
Japanese Yen both have the same lower volatility on Monday.

6.2 The turn-of-the-month effect

To measure the turn-of-the-month effect a dummy is added to the original equation for the day-
of the-week effect. The turn-of-the-month dummy is created with the last prior to the turn of the
month and the first three trading days of the new month. Other research (Kunkel et al. 2003)
(Lakonishok et al. 1988) (Cadsby et al. 1992) (Ariel , 1987) use the same period to examine the turn-
of-the-month effect. To account for the found day-of-the-week effects, the daily dummies have been
added to the equation. The model to be estimated for the turn-of-the-month effect is as follows:

Eq. 5

And for the conditional variance:

Eq. 6

∑ ∑

For these models the same restrictions apply as for the models for the day-of-the-week effect. A
Wald test is performed to determine whether the coefficients are significantly differ from null..

The descriptive statistics for the turn-of-the-month effect is present in Appendix 5. It is clearly
visible that the turn of the month 4 day period gives a very different mean from the rest of the
month. In a lot of currency exchange rates the direction of the mean changes in comparison to the
rest of the month as well as the whole sample. Even though the mean differs from the rest of the
month, the risk represented as the standard deviation does not.

23
Thesis– Calendar Effects and the Foreign Exchange Markets: a matter of efficiency? Simon Kruithof

Table 4 displays the estimated parameters derived from the model in equation 5. A turn-of-the-
month effect in the mean is only present in three currency exchange rates. These are the Singapore
Dollar/ Taiwan Dollar, Taiwan Dollar / US Dollar and Thai Baht / US Dollar. For the Taiwan Dollar / US
Dollar there is a significant abnormal return in the second and third period. The Thai Baht / US Dollar
has a significant return on the turn of the month within the first and last period and as for the
Singapore Dollar / Taiwan Dollar it is exhibited in the second period. For all these turn-of-the-month
effects in the mean, the Wald test null hypothesis is rejected. A turn-of-the-month effect is not
present in the Japanese Yen / US Dollar exchange rate, which was also not identified by Yamori et al.
(2003)
The turn of the month’s volatility is displayed significantly in a lot more currencies. In all cases
where the turn of the month’s volatility turns out significant, it provides a higher volatility with an
exception for the Taiwan Dollar / US Dollar and Singapore Dollar / Thai Baht, whereas respectively
the second and third period gives a lower volatility for the turn of the month.

24
Thesis– Calendar Effects and the Foreign Exchange Markets: a matter of efficiency? Simon Kruithof

Table 4 Empirical Results turn-of-the-month effect

Singapore Dollar / Thai Baht Singapore Dollar / Taiwan Dollar Taiwan Dollar / US Dollar Singapore Dollar / US Dollar
1995-2000 2001-2006 2007-2011 1995-2000 2001-2006 2007-2011 1995-2000 2001-2006 2007-2011 1995-2000 2001-2006 2007-2011
α 0.018 -0.002 0.018 -0.009 -0.004 -0.020 ** 0.003 0.011 ** 0.017 *** 0.011 ** 0.000 -0.001
Monday -0.020 0.013 -0.038 -0.009 0.008 0.010 0.008 ** -0.007 -0.020 ** -0.010 0.012 -0.003
Wednesday -0.055 * -0.012 0.003 0.004 -0.016 0.025 * -0.001 -0.010 -0.029 *** -0.002 -0.012 -0.011
Thursday -0.020 0.009 -0.028 0.016 0.002 0.022 * -0.003 -0.012 * -0.020 ** -0.026 *** -0.010 -0.007
Friday -0.053 * -0.003 -0.007 0.007 -0.009 0.008 -0.021 *** -0.012 * -0.009 -0.002 -0.004 -0.016
Turn of the Month -0.006 0.005 0.008 -0.011 0.016 ** 0.015 0.001 -0.014 *** -0.020 *** -0.008 -0.007 -0.011
Return(t-1) 0.030 -0.134 *** -0.167 *** -0.013 -0.180 *** -0.222 *** 0.020 -0.023 0.028 -0.025 -0.017 -0.002

Variance equation
ω 0.131 *** 0.000 0.069 *** 0.001 0.000 0.003 0.006 *** -0.003 *** 0.000 0.003 *** -0.001 0.001
λ 0.374 *** 0.093 *** 0.233 *** 0.384 *** 0.149 *** 0.063 *** 0.532 *** 0.102 *** 0.115 *** 0.091 *** 0.049 *** 0.041 ***
ϴ 0.541 *** 0.844 *** 0.558 *** 0.686 *** 0.692 *** 0.924 *** 0.646 *** 0.862 *** 0.881 *** 0.910 *** 0.899 *** 0.954 ***
Monday -0.128 *** 0.001 -0.031 * 0.003 ** 0.003 -0.004 -0.014 *** 0.008 *** 0.001 -0.004 *** 0.002 0.001
Wednesday -0.029 -0.001 -0.071 *** 0.012 *** 0.004 ** 0.000 -0.010 *** 0.005 *** 0.001 -0.004 *** 0.001 0.001
Thursday -0.195 *** 0.001 -0.046 *** 0.008 *** 0.006 *** -0.003 -0.006 *** 0.000 0.004 *** -0.004 *** 0.004 *** -0.003
Friday -0.130 *** 0.003 -0.039 ** -0.014 *** 0.000 -0.006 ** 0.010 *** 0.003 *** -0.004 *** -0.003 *** -0.001 -0.006 ***
Turn of the Month 0.007 *** 0.000 -0.028 *** 0.002 *** 0.000 0.001 0.000 *** -0.001 *** 0.000 0.001 *** 0.000 0.003 ***

Tests
ARCH-LM(5) 1.646 4.068 5.556 1.119 5.810 5.124 1.405 1.960 3.530 7.148 2.999 5.262
ARCH-LM(10) 6.145 7.870 6.264 1.609 12.399 8.672 2.745 6.279 6.555 9.960 18.635 7.318
ARCH-LM(20) 21.382 16.364 27.307 * 2.729 26.785 20.034 4.945 19.759 19.977 17.318 18.635 20.001
Ljung-Box Q(5) 2.341 2.304 6.584 2.684 7.404 8.945 4.115 4.363 3.802 0.921 6.310 3.132
Ljung-Box Q(10) 8.946 13.226 13.909 5.846 11.109 14.137 12.224 13.138 14.696 6.675 10.786 12.924
Ljung-Box Q(20) 23.921 19.788 20.049 15.213 25.598 20.380 25.811 24.437 26.628 15.100 18.540 29.682 *

Wald 0.133 0.413 0.434 1.339 5.021 ** 2.586 0.047 7.704 *** 7.190 *** 1.712 1.020 0.811

Log likelihood -152.208 1041.169 -30.155 481.734 993.294 554.646 1452.441 1397.706 921.501 996.785 1135.476 572.268
AIC 0.214 -1.312 0.069 -0.597 -1.251 -0.828 -1.839 -1.768 -1.390 -1.256 -1.433 -0.855
BIC 0.265 -1.261 0.129 -0.546 -1.200 -0.768 -1.788 -1.717 -1.331 -1.205 -1.381 -0.795

Standard errors are removed from this table for visibility reasons and are available upon request, * statistically significant at 10%, ** statistically significant at 5%, ***
statistically significant at 1%. The null hypothesis for the Wald test is that all dummies for the day-of-the-week and turn-of-th-month effects are equal

25
Thesis– Calendar Effects and the Foreign Exchange Markets: a matter of efficiency? Simon Kruithof

Table 4 (continued) Empirical results turn-of-the-month effect

Thai Baht / US Dollar Singapore Dollar / Japanese Yen Japanese Yen / US Dollar
1995-2000 2001-2006 2007-2011 1995-2000 2001-2006 2007-2011 1995-2000 2001-2006 2007-2011
α -0.034 0.003 0.000 0.002 0.004 0.025 0.007 -0.009 -0.034 *
Monday 0.038 -0.003 0.018 ** -0.013 -0.008 -0.017 -0.010 0.020 0.029
Wednesday 0.099 * 0.000 -0.007 0.006 -0.031 ** -0.053 * 0.006 0.018 0.044
Thursday 0.052 -0.020 ** 0.002 -0.013 -0.001 -0.024 -0.045 * 0.000 0.026
Friday 0.049 -0.006 0.006 -0.020 -0.020 -0.023 0.047 * 0.021 0.023
Turn of the Month 0.057 ** -0.013 -0.022 *** -0.026 0.001 -0.013 0.015 -0.011 0.014
Return(t-1) 0.101 *** 0.048 * -0.119 *** -0.017 -0.047 * -0.079 *** -0.016 -0.015 -0.039

Variance equation
ω 0.180 *** 0.002 ** -0.002 *** -0.008 0.006 * 0.031 *** 0.003 0.003 0.009 *
λ 0.208 *** 0.156 *** 0.499 *** 0.141 *** 0.076 *** 0.105 *** 0.044 *** 0.023 *** 0.096 ***
ϴ 0.590 *** 0.778 *** 0.721 *** 0.799 *** 0.866 *** 0.884 *** 0.947 *** 0.951 *** 0.844 ***
Monday -0.180 *** -0.001 0.002 *** 0.033 *** -0.015 *** -0.049 *** -0.020 * -0.019 ** -0.018 **
Wednesday -0.023 -0.003 ** 0.003 *** 0.007 0.000 -0.037 *** 0.021 * 0.000 0.017
Thursday -0.268 *** -0.002 0.001 * 0.008 -0.010 ** -0.015 -0.026 *** -0.002 -0.005
Friday -0.187 *** 0.002 0.001 * 0.019 ** 0.005 -0.046 *** 0.011 0.011 -0.019 **
Turn of the Month 0.006 ** 0.000 0.001 ** 0.008 *** -0.001 0.006 ** 0.007 *** 0.001 0.010 ***

Tests
ARCH-LM(5) 8.119 1.007 0.720 11.371 9.357 * 10.313 6.878 2.029 2.151
ARCH-LM(10) 9.680 2.743 1.318 15.648 14.467 14.699 10.295 6.568 4.379
ARCH-LM(20) 11.795 12.332 2.215 27.757 24.543 27.341 20.804 17.754 7.579
Ljung-Box Q(5) 5.443 9.147 * 6.401 4.505 2.736 4.340 1.277 1.824 3.624
Ljung-Box Q(10) 12.229 12.531 13.130 6.304 6.321 7.860 7.910 5.004 12.624
Ljung-Box Q(20) 27.149 23.646 20.818 11.582 11.097 21.498 23.464 12.706 25.430

Wald 3.998 ** 2.582 5.239 ** 1.618 0.012 0.430 0.578 0.418 0.348

Log likelihood -346.255 1039.713 483.382 -337.083 354.481 -383.504 -444.401 -22.131 -292.461
AIC 0.462 -1.310 -0.718 0.451 -0.434 0.611 0.588 0.047 0.472
BIC 0.514 -1.259 -0.659 0.502 -0.383 0.671 0.639 0.099 0.531
Standard errors are removed from this table for visibility reasons and are available upon request, * statistically significant at 10%, ** statistically significant at 5%, *** statistically
significant at 1%. The null hypothesis for the Wald test is that all dummies for the day-of-the-week and turn-of-the-month effects are equal

26
Thesis– Calendar Effects and the Foreign Exchange Markets: a matter of efficiency? Simon Kruithof

6.3 Holiday effect

To measure a pre-holiday effect correctly within the selected currency exchange rates, it is of the
utmost importance to look at the holidays of each currency independently. Every nation has its own
local holidays which could be of influence. For example Taiwan has most of its holidays based on the
lunar calendar, which is used within the Chinese culture. Other countries such as Thailand
(Buddhism), Japan (No religious holidays) and Singapore (combination of Christian, Chinese and
Muslim holidays) do not have similar days, therefore, creating different dummies for each currency
holidays, is of importance. This method has been used before by Cadsby et al. (1991). The national
holidays in Thailand, Singapore, Taiwan, Japan and the United States which were used for the
dummy variables can be found in Appendix 6. The added dummy , which represents the day
before the local holidays of the numerator, such as Thai Baht within the Thai Baht / US Dollar
exchange rate. represents the day before the local holidays of the denominator, which is in
this case the US Dollar. Lastly, is added to measure the impact of pre-holiday days which is
the same in both countries, for example New Year’s day. To account for the day-of-the-week effect,
daily daummies have been added to the equation. The mean equation and the conditional variance
equation for the holiday effect with the added dummies are as follows:
Eq. 7

Eq. 8

∑ ∑

The descriptive statistics in Appendix 7 show a few interesting facts. The mean is predominantly
negative for all holidays. The standard deviation within the Singapore Dollar/ Thai Baht of the dummy
variable Holiday A, which in this case represents the local pre-holidays of Singapore, is twice as high
as the other dummies between 2007-2011. Within the period 1995-2000, the standard deviation of
the Taiwanese pre-holidays in the Singapore Dollar / Taiwan Dollar is 4 times as high as the standard
deviation of the Singapore pre-holidays and the combined pre-holidays.

27
Thesis– Calendar Effects and the Foreign Exchange Markets: a matter of efficiency? Simon Kruithof

Table 5 Empirical Results Holiday effect

Singapore Dollar / Thai Baht Singapore Dollar / Taiwan Dollar Taiwan Dollar / US Dollar Singapore Dollar / US Dollar
1995-2000 2001-2006 2007-2011 1995-2000 2001-2006 2007-2011 1995-2000 2001-2006 2007-2011 1995-2000 2001-2006 2007-2011
α 0.032 0.000 0.011 -0.014 ** -0.001 -0.017 * 0.003 0.005 0.013 *** 0.007 0.000 -0.004
Monday 0.001 0.012 -0.038 * -0.009 0.007 0.009 0.001 -0.004 -0.018 ** -0.005 0.010 -0.001
Wednesday -0.080 -0.011 -0.002 0.011 -0.017 0.024 -0.001 -0.005 -0.028 *** 0.001 -0.014 * -0.010
Thursday 0.015 0.008 -0.021 0.018 * 0.001 0.020 -0.007 * -0.009 -0.018 ** -0.020 * -0.010 -0.003
Friday -0.045 -0.002 -0.016 0.009 -0.010 0.006 -0.018 *** -0.010 -0.010 0.009 -0.007 -0.017
Holiday A 0.005 -0.018 0.033 -0.014 -0.013 0.024 0.020 0.006 -0.012 0.064 ** 0.010 0.006
Holiday B -0.007 -0.024 0.008 -0.160 *** 0.011 0.002 -0.005 0.007 -0.004 -0.007 0.007 -0.027
Holiday Both 0.081 0.019 -0.046 0.160 *** -0.004 -0.021 -0.152 -0.088 ** -0.002 -0.124 *** -0.037 -0.038
Return(t-1) -0.021 -0.134 *** -0.163 *** -0.026 -0.176 *** -0.218 *** -0.032 -0.011 0.028 -0.041 -0.016 -0.007

Variance equation
ω 0.139 *** 0.001 -0.002 0.002 *** 0.000 0.004 * -0.003 *** 0.007 *** -0.001 * 0.027 *** 0.004 *** 0.002
λ 0.186 *** 0.091 *** 0.112 *** 0.499 *** 0.143 *** 0.055 *** 0.489 *** 0.188 *** 0.119 *** 0.353 *** 0.063 *** 0.038 ***
ϴ 0.594 *** 0.841 *** 0.890 *** 0.590 *** 0.705 *** 0.931 *** 0.575 *** 0.689 *** 0.873 *** 0.582 *** 0.553 *** 0.959 ***
Monday -0.107 *** 0.000 0.016 *** 0.003 ** 0.003 -0.006 * 0.000 -0.002 ** 0.002 ** -0.027 *** 0.002 0.001
Wednesday -0.016 -0.001 -0.005 -0.003 ** 0.004 ** -0.001 0.002 *** -0.007 *** 0.002 *** -0.031 *** 0.003 ** 0.001
Thursday -0.080 ** 0.001 0.005 0.023 *** 0.006 *** -0.002 0.003 *** -0.009 *** 0.005 *** -0.029 *** 0.004 *** -0.004 **
Friday -0.057 * 0.003 -0.005 -0.016 *** 0.000 -0.008 *** 0.022 *** -0.005 *** -0.003 *** -0.024 *** 0.001 -0.008 ***
Holiday A -0.076 *** -0.002 0.000 -0.007 *** 0.001 0.005 ** 0.000 -0.003 *** 0.006 *** -0.003 *** 0.001 0.004 **
Holiday B -0.078 *** 0.001 0.006 ** 0.101 *** -0.004 ** 0.003 0.000 -0.002 ** -0.002 *** -0.006 *** -0.006 *** -0.008 ***
Holiday Both -0.052 0.003 -0.010 ** -0.095 *** 0.005 -0.007 * 0.085 0.004 -0.003 0.010 *** -0.005 0.002

Tests
ARCH-LM(5) 3.309 3.643 0.855 2.476 5.375 5.137 3.240 2.343 3.056 1.362 2.954 3.904
ARCH-LM(10) 13.027 8.114 4.665 3.863 12.345 10.068 9.152 9.947 5.496 4.082 7.750 6.026
ARCH-LM(20) 22.534 17.036 18.716 6.071 26.778 21.050 14.227 24.575 16.964 22.495 30.422 19.257
Ljung-Box Q(5) 5.613 2.065 5.937 3.737 5.341 3.740 1.653 4.564 3.575 3.415 5.522 3.896
Ljung-Box Q(10) 11.956 11.948 15.396 7.051 13.176 6.930 5.941 12.737 14.175 9.892 10.221 13.646
Ljung-Box Q(20) 20.641 18.069 26.354 19.564 25.431 13.319 13.210 29.072 * 22.083 20.504 17.755 32.731 **

Wald 2.958 * 0.195 0.013 0.461 0.021 0.027 0.171 4.118 ** 0.715 5.936 ** 4.684 ** 0.639

Log likelihood -610.454 1043.674 127.611 507.328 993.128 555.735 1475.230 1344.555 924.286 865.689 1137.410 580.523
AIC 0.805 -1.310 -0.167 -0.625 -1.246 -0.823 -1.863 -1.695 -1.388 -1.083 -1.430 -0.861
BIC 0.871 -1.245 -0.091 -0.560 -1.181 -0.748 -1.798 -1.630 -1.313 -1.018 -1.365 -0.786
Standard errors are removed from this table for visibility reasons and are available upon request, * statistically significant at 10%, ** statistically significant at 5%, *** statistically
significant at 1%. The null hypothesis for the Wald test is that all dummies for the holiday effect are equal.

28
Thesis– Calendar Effects and the Foreign Exchange Markets: a matter of efficiency? Simon Kruithof

Table 5 (continued) Empirical Results Holiday effect

Thai Baht / US Dollar Singapore Dollar / Japanese Yen Japanese Yen / US Dollar
1995-2000 2001-2006 2007-2011 1995-2000 2001-2006 2007-2011 1995-2000 2001-2006 2007-2011
α 0.039 -0.005 -0.004 -0.003 0.008 0.021 0.013 -0.011 -0.029 *
Monday -0.110 ** 0.001 0.041 -0.014 -0.009 -0.021 -0.016 0.018 0.009
Wednesday 0.000 0.006 -0.021 0.007 -0.034 * -0.053 * 0.006 0.018 0.038
Thursday 0.037 -0.016 0.005 -0.010 -0.002 -0.025 -0.046 * 0.002 0.022
Friday 0.048 0.003 0.027 -0.009 -0.012 -0.031 0.046 * 0.016 0.022
Holiday A 0.018 -0.010 -0.011 -0.042 -0.066 *** 0.030 -0.044 0.022 0.028
Holiday B -0.093 0.004 0.038 -0.033 -0.016 0.002 0.049 0.022 0.011
Holiday Both 0.009 -0.038 -0.061 -0.086 -0.025 -0.037 -0.006 -0.051 -0.203
Return(t-1) 0.182 *** 0.053 -0.022 -0.014 -0.047 -0.071 ** -0.012 -0.013 -0.032

Variance equation
ω 0.178 *** 0.017 *** 0.090 *** -0.005 0.024 *** 0.030 *** 0.009 0.027 *** 0.014 **
λ 0.269 *** 0.102 *** 0.151 *** 0.147 *** 0.118 *** 0.107 *** 0.040 *** 0.033 * 0.082 ***
ϴ 0.542 *** 0.565 *** 0.588 *** 0.795 *** 0.593 *** 0.883 *** 0.953 *** 0.593 *** 0.864 ***
Monday -0.175 *** -0.005 -0.008 0.039 *** -0.022 *** -0.041 *** -0.025 ** -0.016 ** -0.025 ***
Wednesday -0.157 *** -0.008 -0.049 0.002 -0.017 ** -0.034 *** 0.013 -0.003 0.016
Thursday -0.162 *** -0.006 -0.029 0.009 -0.018 *** -0.013 -0.030 *** -0.001 -0.006
Friday -0.081 *** -0.005 -0.020 0.015 * -0.005 -0.048 *** 0.002 0.009 -0.023 **
Holiday A -0.060 *** -0.009 * -0.080 *** 0.001 -0.008 * -0.003 0.000 -0.015 ** -0.017 ***
Holiday B -0.064 *** -0.011 * 0.001 -0.005 -0.007 -0.016 *** 0.006 -0.039 *** -0.027 ***
Holiday Both -0.011 -0.007 -0.072 -0.004 -0.011 ** 0.098 *** -0.020 * 0.045 * 0.076 **

Tests
ARCH-LM(5) 8.948 0.577 9.038 11.570 3.752 10.862 * 6.895 3.814 1.610
ARCH-LM(10) 10.681 2.252 9.141 15.566 8.776 15.239 10.631 11.156 4.005
ARCH-LM(20) 24.448 9.462 13.549 27.074 18.566 27.489 21.694 29.141 * 6.776
Ljung-Box Q(5) 5.535 3.637 3.733 3.925 1.674 4.654 1.305 2.648 4.510
Ljung-Box Q(10) 12.115 6.320 15.103 * 5.924 5.625 8.314 8.482 6.323 13.627
Ljung-Box Q(20) 25.242 16.626 18.445 11.972 10.004 22.524 24.763 14.365 25.863

Wald 0.220 2.415 0.332 4.542 ** 9.782 *** 0.000 0.000 0.004 1.064

Log likelihood -475.293 883.131 -270.729 -337.839 351.057 -379.426 -444.974 -24.444 -281.209
AIC 0.632 -1.105 0.444 0.457 -0.425 0.611 0.594 0.056 0.460
BIC 0.698 -1.040 0.520 0.522 -0.360 0.686 0.659 0.121 0.536
Standard errors are removed from this table for visibility reasons and are available upon request,, * statistically significant at 10%, ** statistically significant at 5%, ***
statistically significant at 1%. The null hypothesis for the Wald test is that all dummies for the holiday effect are equal

29
Thesis– Calendar Effects and the Foreign Exchange Markets: a matter of efficiency? Simon Kruithof

Table 5 presents the outcome of the estimated model for pre-holiday effects. The local holidays
of Singapore ( ) have a significant abnormal return within the Singapore Dollar / Japanese Yen
exchange rate and Singapore Dollar / US Dollar exchange rate. Within the Singapore Dollar / Taiwan
Dollar exchange rate the local Taiwanese holidays display a significant abnormal return as well as a
combined holidays ( ) significant abnormal return. Taiwan and Singapore both have the
Chinese New Year as a public holiday. The combined holiday dummy for Singapore Dollar / US Dollar
exhibits an abnormal return as well. Singapore and the United States share New Year’s day and
Christmas as a holiday.
The local US holidays within the currency exchange rates exhibit a lower volatility when they are
significant. This significant lower volatility is also exposed with the local Japanese Holidays. Higher
significant volatilities due to local holidays can be found in the period 2007-2011. Local holidays that
increase volatility in this period are the Taiwanese, Singaporean and Thai holidays.

6.4 Friday -the-thirteenth effect

Friday the thirteenth has a superstition effect that has not been researched within currency
exchange rates. To measure the effect of a Friday the thirteenth, a model (equation 8 and equation
9) with a dummy for a regular Friday, excluding all Fridays the thirteenths. This is to avoid
multicollinearity in the parameters and a separate dummy for all Fridays the thirteenths.
Eq. 8

Eq. 9

∑ ∑

Friday the thirteenth appears to have increased the risk within the Thai currency in the period
between 1995-2000, which can be seen in Appendix 8 within both currency exchange rates, Thai
Baht / US Dollar and Singapore Dollar / Thai Baht. The standard deviation is at least twice as high in
comparison to a regular Friday. Another striking conclusion is, that all Fridays the thirteenths, except
for the second period of the Singapore Dollar / US Dollar and the third period of the Taiwan Dollar /
US Dollar, exhibit a larger positive or negative mean in comparison to a regular Friday.

30
Thesis– Calendar Effects and the Foreign Exchange Markets: a matter of efficiency? Simon Kruithof

Table 6 Empirical Results Friday-the-thirteenth Effect

Singapore Dollar / Thai Baht Singapore Dollar / Taiwan Dollar Taiwan Dollar / US Dollar Singapore Dollar / US Dollar
1995-2000 2001-2006 2007-2011 1995-2000 2001-2006 2007-2011 1995-2000 2001-2006 2007-2011 1995-2000 2001-2006 2007-2011
α 0.000 0.000 -0.002 -0.017 ** -0.003 -0.005 0.022 *** 0.002 -0.003 0.010 *** -0.004 -0.007
Friday -0.010 -0.001 -0.005 0.020 ** -0.005 -0.003 -0.044 *** -0.005 0.008 -0.004 -0.003 -0.013
Friday13th -0.056 ** -0.063 0.070 -0.001 -0.006 0.010 0.062 ** -0.005 -0.011 ** -0.025 -0.001 -0.018
Return(t-1) -0.135 *** -0.137 *** -0.163 *** -0.061 -0.174 *** -0.214 *** 0.230 *** -0.011 0.051 * -0.047 -0.014 -0.001

Variance equation
ω 0.005 *** 0.000 0.001 * 0.040 *** 0.003 *** 0.002 *** -0.001 *** 0.011 *** 0.002 *** 0.005 *** 0.001 ** 0.002 ***
λ 0.236 *** 0.092 *** 0.104 *** 0.313 *** 0.139 *** 0.055 *** 0.300 *** 0.162 *** 0.173 *** 0.300 *** 0.053 *** 0.045 ***
ϴ 0.789 *** 0.849 *** 0.901 *** 0.590 *** 0.711 *** 0.933 *** 0.640 *** 0.039 0.796 *** 0.636 *** 0.892 *** 0.949 ***
Friday -0.013 *** 0.003 ** -0.003 -0.088 *** -0.003 ** -0.006 *** 0.026 *** -0.003 *** -0.006 *** -0.011 *** -0.001 -0.008 ***
Friday13th -0.017 *** 0.006 -0.009 -0.091 *** 0.009 -0.016 *** 0.006 *** -0.012 *** -0.009 *** -0.014 *** 0.002 -0.013 ***

Tests
ARCH-LM(5) 0.747 4.851 0.779 2.679 6.976 5.563 0.959 2.174 2.746 1.776 3.017 4.763
ARCH-LM(10) 1.261 8.303 4.578 4.605 15.725 9.289 2.443 8.042 5.604 2.894 4.402 6.924
ARCH-LM(20) 2.299 16.124 17.567 17.824 27.947 21.572 6.336 20.578 13.908 21.177 16.978 18.818
Ljung-Box Q(5) 1.079 2.751 6.484 6.485 6.329 2.623 6.729 4.407 2.137 5.221 7.167 3.127
Ljung-Box Q(10) 10.908 13.514 17.162 ** 10.187 9.998 5.436 8.316 13.814 11.600 15.876 * 11.578 13.194
Ljung-Box Q(20) 22.606 20.750 32.346 ** 21.493 27.777 * 11.119 24.580 26.067 19.980 27.009 17.894 29.070 *

Wald 5.546 ** 1.726 1.633 0.161 0.071 0.061 0.390 1.173 0.134 1.264 0.018 1.574

Log likelihood 172.38 1038.25 114.92 245.31 986.47 554.11 1208.29 1285.64 905.90 863.63 1128.71 569.25
AIC -0.209 -1.316 -0.162 -0.302 -1.250 -0.836 -1.535 -1.633 -1.376 -1.094 -1.432 -0.859
BIC -0.178 -1.285 -0.127 -0.272 -1.219 -0.800 -1.504 -1.602 -1.340 -1.063 -1.401 -0.824
Standard errors are removed from this table for visibility reasons and are available upon request,, * statistically significant at 10%, ** statistically significant at 5%, ***
statistically significant at 1%. The null hypothesis for the Wald test is that all dummies for Friday and Friday the thirteenth are equal.

31
Thesis– Calendar Effects and the Foreign Exchange Markets: a matter of efficiency? Simon Kruithof

Table 6 (continued) Empirical Results Firday-the-thirteenth effect

Thai Baht / US Dollar Singapore Dollar / Japanese Yen Japanese Yen / US Dollar
1995-2000 2001-2006 2007-2011 1995-2000 2001-2006 2007-2011 1995-2000 2001-2006 2007-2011
α -0.001 -0.006 ** -0.002 -0.008 -0.006 0.000 -0.001 -0.001 -0.007
Friday 0.004 -0.001 0.010 -0.016 -0.008 0.003 0.056 *** 0.010 -0.004
Friday13th -0.118 0.041 -0.055 -0.021 -0.022 0.053 0.114 0.035 -0.027
Return(t-1) -0.001 0.046 * -0.035 -0.014 -0.052 * -0.077 ** -0.014 -0.017 -0.034

Variance equation
ω 0.160 *** 0.001 ** 0.000 ** 0.005 *** 0.001 0.009 *** 0.003 ** -0.001 0.010 ***
λ 0.244 *** 0.152 *** 0.471 *** 0.137 *** 0.075 *** 0.102 *** 0.044 *** 0.021 *** 0.086 ***
ϴ 0.577 *** 0.777 *** 0.716 *** 0.797 *** 0.868 *** 0.877 *** 0.949 *** 0.955 *** 0.871 ***
Friday -0.307 *** 0.003 *** 0.000 0.016 ** 0.005 -0.026 *** -0.009 0.010 * -0.024 ***
Friday13th -0.147 *** 0.005 * -0.002 -0.032 ** 0.002 -0.054 *** -0.031 *** 0.011 -0.056 ***

Tests
ARCH-LM(5) 4.158 0.903 0.599 10.208 * 5.009 6.799 11.059 * 1.712 2.270
ARCH-LM(10) 13.249 2.634 1.177 13.926 11.018 11.220 13.095 6.401 4.724
ARCH-LM(20) 14.181 9.994 2.040 26.462 22.140 23.062 24.349 17.940 7.195
Ljung-Box Q(5) 5.633 8.473 4.207 4.604 2.777 4.369 1.150 2.101 3.546
Ljung-Box Q(10) 15.586 * 11.226 12.316 5.822 6.089 8.503 8.062 5.123 12.846
Ljung-Box Q(20) 26.539 * 21.272 20.398 11.266 11.432 24.150 24.742 13.084 25.788

Wald 0.595 1.237 0.551 0.233 0.244 0.466 2.442 0.457 0.140

Log likelihood -696.17 1033.63 479.48 -345.31 346.43 -389.55 -454.29 -26.32 -299.97
AIC 0.902 -1.310 -0.722 0.453 -0.431 0.611 0.593 0.045 0.474
BIC 0.933 -1.279 -0.686 0.484 -0.401 0.647 0.624 0.076 0.510
Standard errors are removed from this table for visibility reasons and are available upon request,, * statistically significant at 10%, ** statistically significant at 5%, ***
statistically significant at 1%. The null hypothesis for the Wald test is that all dummies for Friday and Friday the thirteenth are equal.

32
Thesis– Calendar Effects and the Foreign Exchange Markets: a matter of efficiency? Simon Kruithof

Friday the thirteenth exhibits a significant abnormal return in two currency exchange rates. The
first currency exchange rate, which has a significant effect, is the Taiwan Dollar / US Dollar exchange
rate. Within this exchange rate Friday the thirteenth has an abnormal return in the first period and
last period during both financial crises. However, the Wald test is not rejected, therefore, these
Friday-the-thirteenth effects do not differ significantly from null. The other identified abnormal
return on Friday the thirteenth appears within the Singapore Dollar / Thai Baht exchange rate, which
was already prominent in the descriptive statistics table.
Observed volatility for Friday the thirteenth is all of the times significant between 1995 and 2000.
Friday the thirteenth’s significant volatility is lower in all currency exchange rates, except for the first
period of the Taiwan Dollar / US Dollar and the second period of the Thai Baht / US Dollar when the
volatility was higher.

6.5 January effect

Researchers measure the January effect differently. Some focus on half-month effects and others
on the whole month at once. More researchers have focussed on the month of a January as a whole,
therefore, it has been decided to take create one dummy for the month of January to measure its
effect on currency exchange rates.
The equation to measure the January effect is as follows:

Eq. 10

Eq. 11

∑ ∑

Throughout all sub-samples January generates on average a higher mean in comparison to

the rest of the year, which can be seen in Appendix 9. Furthermore, standard deviations generated in
January do not differ greatly from the rest of the year. There is one exception which is the standard
deviation of the Thai Baht / US Dollar between 1995 and 2000. This standard deviation is 22 times
higher than the average standard deviation the rest of the year. This abnormal high standard
deviation in January is likely to be due to the Asian financial crises which hit Thailand’s currency the
Thai Baht the most with reaching its lowest value in January 1998.

33
Thesis– Calendar Effects and the Foreign Exchange Markets: a matter of efficiency? Simon Kruithof

Table 7 Empirical Results January Effect

Singapore Dollar / Thai Baht Singapore Dollar / Taiwan Dollar Taiwan Dollar / US Dollar Singapore Dollar / US Dollar
1995-2000 2001-2006 2007-2011 1995-2000 2001-2006 2007-2011 1995-2000 2001-2006 2007-2011 1995-2000 2001-2006 2007-2011
α -0.007 ** -0.002 -0.004 -0.013 ** -0.006 ** -0.004 -0.001 -0.001 -0.004 0.000 -0.004 -0.010 **
January 0.043 0.026 *** 0.010 0.019 * 0.020 * -0.002 0.008 -0.018 ** 0.007 0.023 *** -0.003 0.010
Return(t-1) -0.141 *** -0.138 *** -0.160 *** -0.002 -0.179 *** -0.220 *** 0.074 ** -0.005 0.026 -0.036 -0.014 -0.005

Variance equation
ω 0.001 *** 0.001 *** 0.000 *** 0.004 *** 0.003 *** 0.000 *** 0.000 *** 0.000 *** 0.000 *** 0.000 *** 0.001 *** 0.000 ***
λ 0.227 *** 0.089 *** 0.106 *** 0.350 *** 0.129 *** 0.058 *** 0.663 *** 0.052 *** 0.094 *** 0.071 *** 0.052 *** 0.037 ***
ϴ 0.813 *** 0.855 *** 0.898 *** 0.686 *** 0.719 *** 0.929 *** 0.741 *** 0.944 *** 0.904 *** 0.930 *** 0.894 *** 0.959 ***
January 0.006 *** 0.000 0.001 *** -0.002 *** 0.000 0.000 0.000 0.000 ** 0.000 0.000 *** 0.000 0.000 ***

Tests
ARCH-LM(5) 0.810 4.962 0.620 0.381 6.803 4.766 0.214 2.271 3.507 6.984 2.440 4.423
ARCH-LM(10) 1.585 7.970 4.171 0.651 15.444 9.186 0.866 20.928 ** 6.460 9.128 3.672 7.018
ARCH-LM(20) 4.174 16.331 17.144 1.223 27.902 19.483 1.236 27.474 15.734 13.509 15.562 19.736
Ljung-Box Q(5) 1.103 2.666 1.939 4.489 8.987 3.150 2.160 3.718 3.819 1.134 7.339 3.545
Ljung-Box Q(10) 11.990 14.297 11.720 7.627 11.732 5.822 12.006 13.315 14.522 8.031 11.768 14.155
Ljung-Box Q(20) 23.927 22.050 25.827 16.923 28.148 11.804 24.415 25.274 24.052 15.958 18.164 32.070 **

Wald 2.159 6.801 *** 0.272 2.863 * 3.668 * 0.059 2.243 6.221 ** 0.348 6.933 ** 0.079 0.756

Log likelihood 219.875 1038.674 115.556 404.551 984.118 545.771 1354.074 1381.957 900.937 984.819 1127.780 561.711
AIC -0.272 -1.319 -0.166 -0.509 -1.250 -0.826 -1.724 -1.758 -1.371 -1.251 -1.433 -0.851
BIC -0.248 -1.295 -0.139 -0.485 -1.226 -0.799 -1.700 -1.734 -1.343 -1.227 -1.409 -0.823
Standard errors are removed from this table for visibility reasons and are available upon request,, * statistically significant at 10%, ** statistically significant at 5%, ***
statistically significant at 1%. The null hypothesis for the Wald test is that the dummy for the January effect does not difer from 0.

34
Thesis– Calendar Effects and the Foreign Exchange Markets: a matter of efficiency? Simon Kruithof

Table 7 (continued) Empirical Results January effect

Thai Baht / US Dollar Singapore Dollar / Japanese Yen Japanese Yen / US Dollar
1995-2000 2001-2006 2007-2011 1995-2000 2001-2006 2007-2011 1995-2000 2001-2006 2007-2011
α 0.005 *** -0.004 0.000 -0.009 -0.006 0.000 0.005 0.000 -0.007
January 0.001 -0.019 0.008 -0.020 -0.016 0.000 0.044 0.010 0.001
Return(t-1) 0.004 0.045 -0.032 -0.012 -0.052 ** -0.079 ** -0.022 -0.019 -0.040

Variance equation
ω 0.000 *** 0.001 *** 0.000 *** 0.007 *** 0.002 *** 0.003 *** 0.001 ** 0.001 ** 0.005 ***
λ 0.157 *** 0.147 *** 0.485 *** 0.129 *** 0.075 *** 0.104 *** 0.040 *** 0.018 *** 0.085 ***
ϴ 0.880 *** 0.781 *** 0.709 *** 0.811 *** 0.863 *** 0.877 *** 0.956 *** 0.960 *** 0.870 ***
January 0.000 0.000 0.002 *** -0.001 0.000 0.000 0.001 0.000 -0.003 **

Tests
ARCH-LM(5) 6.711 0.724 0.577 9.174 7.550 5.545 5.235 1.414 2.410
ARCH-LM(10) 7.973 2.660 1.166 12.657 11.876 9.523 13.127 6.080 4.666
ARCH-LM(20) 22.408 9.879 2.047 24.364 22.044 21.931 23.902 18.011 7.386
Ljung-Box Q(5) 6.715 9.277 7.100 4.440 2.973 4.362 1.046 2.241 3.293
Ljung-Box Q(10) 12.129 12.001 13.578 6.295 6.222 9.054 9.244 5.304 13.144
Ljung-Box Q(20) 27.161 22.722 20.673 11.080 11.828 24.616 25.589 13.464 25.366

Wald 0.055 2.594 0.323 0.626 0.989 0.000 2.537 0.193 0.003

Log likelihood 612.360 1031.026 480.521 -350.699 345.651 -394.778 -459.036 -27.610 -302.741
AIC -0.775 -1.309 -0.726 0.458 -0.433 0.616 0.596 0.044 0.475
BIC -0.751 -1.286 -0.698 0.482 -0.409 0.644 0.620 0.068 0.503
Standard errors are removed from this table for visibility reasons and are available upon request, * statistically significant at 10%, ** statistically significant at 5%, *** statistically
significant at 1%. The null hypothesis for the Wald test is that the dummy for the January effect does not difer from 0.

35
Thesis– Calendar Effects and the Foreign Exchange Markets: a matter of efficiency? Simon Kruithof

The observed abnormal returns in January seen in table 7 are only present between
1995-2000 and 2001-2006. January is highly significant within the Singapore Dollar / US Dollar
exchange rate (period 1) and the Singapore Dollar / Thai Baht exchange rate (period 2). Less
significant abnormal returns are present in the first and second period of the Singapore Dollar /
Taiwan Dollar exchange rate. Lastly, an abnormal return is measured within the Singapore Dollar /
Thai Baht exchange rate between 2001 and 2006.

A significant volatility for January is mostly measured between 1995-2000 and last 2007-
2011, except for a higher volatility in the second period within the Taiwan Dollar / US Dollar
exchange rate. Some explosiveness is measured in the variance equation of the Singapore Dollar /
US Dollar and Singapore Dollar / Thai Baht; this can be a negative influence to the interpretation of
the variance regression parameters.

36
Thesis– Calendar Effects and the Foreign Exchange Markets: a matter of efficiency? Simon Kruithof

Table 8 Overview Identified Calendar Effects

Singapore Dollar / Thai Baht Singapore Dollar / Taiwan Dollar Taiwan Dollar / US Dollar Singapore Dollar / US Dollar Thai Baht / US Dollar Singapore Dollar / Japanese Yen Japanese Yen / US Dollar
Effect Effect Effect Effect Effect Effect Effect

Day-of-the-week effect

1995 - 2000 Tuesday + Monday - Wednesday - Thursday -

Friday ++ Tuesday + Friday -
Thursday ++
2001 - 2006 Wednesday - Thursday +
2007 - 2011 Monday - Monday + Wednesday - Tuesday -
Tuesday + Wednesday -
Wednesday ++
Thursday +

Turn-of-the-month effect

1995 - 2000 TOM + TOM +

2001 - 2006 TOM ++
2007 - 2011 TOM ++ TOM ++

Holiday effect

1995 - 2000 Taiwan ++ Singapore ++

Both ++ Both +
2001 - 2006 Both + Singapore ++
2007 - 2011

Friday-the-thirteenth effect

1995 - 2000 Fri 13th + Fri 13th +

2001 - 2006
2007 - 2011 Fri 13th +

January effect

1995 - 2000 January - January -

2001 - 2006 January ++ January - January -
2007 - 2011

- represents a weak effect, + represents a strong effect and ++ represents a very strong effect

37
Thesis– Calendar Effects and the Foreign Exchange Markets: a matter of efficiency? Simon Kruithof

Conclusion

The Efficient Market Hypothesis has been of interest to researchers and investors since it was
published by Fama (1970). Abnormal returns in efficient markets should not exist. This paper has
examined whether there are still calendar effects present within currency exchange markets that are
considered efficient or inefficient. Table 8 presents an overview of the identified calendar effects in
the mean within this research and are arranged from the least efficient currency exchange rate to
the most efficient currency exchange rate.

Day-of-the-week effects are present, but mainly between the years 1995 till 2000 and 2007
and 2011. Steady periods, such as the one between 2001 and 2008 with steady economic growth, do
not have calendar effects. Furthermore, turn-of-the month effects can be found in weak-form
efficient currency, such as the Singapore Dollar / Taiwan Dollar. Asian pre-holiday effects have been
identified for the Singapore holidays and the Taiwan holidays. Friday-the-thirteenth effects are only
present and significantly different from null within the Singapore Dollar / Thai Baht currency
exchange rate. Finally, a January effect is only between 1995 and 2006 and is to be considered a
weak effect.

Overall the calendar effects measured are highly inconsistent across the various sub-samples
and currency exchange rates. The Taiwan Dollar / US Dollar exhibits most calendar effects of all
currency exchange rates. These calendar effects can be explained by the fact that The Taiwan Dollar
has been identified as inefficient currency and market by Chiang et al (2010). Though, according to
Yamori et al. (2003) the effects in the Taiwan Dollar / US Dollar exchange rate should disappear
within time. This is contradicting to the findings in this study concerning the Taiwan Dollar / US
Dollar.

It can be concluded, that most traded and efficient currency exchange rate, the Japanese Yen
/ US Dollar, does not exhibit any calendar effects apart from a few feeble day-of-the-week effects.
This fully endorses Fama’s Efficient Market Hypothesis, where all information is reflected in the price
and no calendar effects can be found. Nevertheless, how inconsistent the effects may be, they are
still measured within the other currency exchange rates. Investors and traders can benefit from
these effects by focusing on Taiwan Dollar and Singapore Dollar. For example, trade on a day before
a Singapore holiday.

38
Thesis– Calendar Effects and the Foreign Exchange Markets: a matter of efficiency? Simon Kruithof

There is no profound evidence for a relation between the volatility and a calendar effects.
Many of the volatilities are significant, but do not provide evidence on more risks that were taken.
For example, a Friday the thirteenth does not entail a higher volatility. There is some remaining
autocorrelation, heteroskedacity and explosiveness In the residuals and outcomes, which can be
detrimental for the interpretation of the estimations within the variance and mean equations.

This paper provides evidence that weak-form or inefficient markets do exhibit more calendar
effects, especially, during a crisis. Efficient currencies do not exhibit calendar effects in the past 10
years.

The researcher recommended future research to entail:

- Half-months effects for January
- Chinese superstition effects on the Taiwan and Singapore Dollar exchange rates
- Fibrenacci numbers.

39
Thesis– Calendar Effects and the Foreign Exchange Markets: a matter of efficiency? Simon Kruithof

References

Agrawal, A., Tandon, K., | 1994 | Anomalies or illusions? Evidence from stock markets in eighteen
countries | Journal of International Money and Finance 13 pp 083-106

Al-Khazali, O.M., Leduc, G., Pyun, C.S. | 2011 | Market efficiency of floating exchange rate systems:
Some evidence from Pacific-Asian countries | Global Finance Journal vol 22 pp. 154-168

Ariel, A.R. | 1987 | A monthly effect in stock returns | Journal of Financial Economics vol 18 pp. 161-
174

Ariel, R.A. | 1990 | High stock returns before holidays: existence and evidence on possible causes |
The Journal of Finance vol 7 pp. 1611-1626

Aydogan, K., Booth, G. | 2010 | Calendar anomalies in the Turkish foreign exchange markets |
Applied Financial Economics vol 13:5 pp. 353-360

Azad, A.S.M.S. | 2009 | Random walk and efficiency tests in the Asia-Pacific foreign exchange
markets: Evidence from the post-Asian currency crisis data | Research in International Business and
Finance vol 23 pp. 322-338

Baker, H.K., Rahman, A., Saadi, S. | 2008 | The day-of-the-week effect and conditional volatility:
sensitivity of error distributional assumptions | Review of Financial Economics vol 17 pp. 280-295

Berument, H., Coskun, M.N., Sahin, A. | 2007 | Day-of the-week effect on foreign volatility: Evidence
from Turkey | Research in International Business and Finance vol 21 pp. 87-97

Berument, H., Kiymaz, H., | 2001 | The Day of the Week Effect on Stock Market Volatility | Journal of
Economics and Finance vol. 25 Nr 2

BIS | 2010 | Triennial Central Bank survey: foreign exchange and derivatives market activity in 2010 |
Bank for International Settlements

40
Thesis– Calendar Effects and the Foreign Exchange Markets: a matter of efficiency? Simon Kruithof

Cadsby, C.B., Ratner, M.T. | 1992 | Turn-of-the-month and pre-holiday effects on stock returns |
Journal of Banking and Finance vol 16 pp. 497-509

Chamberlain, T.W., Cheung, C.S., Kwan, C.C.Y. | 1991 | The Friday the thirteenth effect: myth or
reality? | Quarterly journal of Business and Economics vol 30 pp. 111-118

Chan, M.W.L., Khantavit, A., Thomas, H. | 1996 | Seasonality and cultural influences on four Asian
stock markets | Asia Pacific Journal of Management vol 13:2 pp. 1-24

Chiang, S.M., Lee, Y.H., Su, H.M., Tzou, Y.P. | 2010 | Efficiency tests of foreign exchange markets for
four Asian Countries | Research in International Business and Finance vol. 24 pp.284-294

Choudry, T. | 2000 | Day of the week effect in emerging Asian stock markets: evidence from the
GARCH model | Applied Financial Economics, 10:3, pp. 235-242

Dumas, B., Solnik, B. | 1993 | The World price of foreign exchange risk | National Bureau of
Economic Research working paper no. 4459

Dyl, E., Maberly, E. | 1988 | The anomaly that isn’t there: a comment on Friday the thirteenth |
Journal of Finance vol 43 pp. 1285-1286

Fama, E.F. | 1970 | Efficient Capital Markets: A review of theory and empirical work | Journal of
Finance vol 25:2 pp. 383-417

Fama, E.F. | 1998 | Market efficiency, long-term returns, and behavioural finance | Journal of
Financial Economics vol 49 pp. 283-306

Faruqee, H. |1994| Long-Run Determinants of the Real Exchange Rate: A Stock-Flow Perspective |
IMF working paper, SSRN

Friedman, M. | 1953 | The Case for Flexible Exchange Rates | Positive Economics, essay pp. 157-203

Froot, K.A., Thaler, R.H. | 1990 | Anomalies: Foreign Exchange | The Journal of Economic
Perspectives vol 4:3 pp. 179-192

41
Thesis– Calendar Effects and the Foreign Exchange Markets: a matter of efficiency? Simon Kruithof

Hansen, P.R., Lunde, A., Nason, J.M. | 2005 | Testing the significance of calendar effects | Federal
Reserve Bank of Atlanta, working paper 2005-2

Hilliard, J.E., Tucker, A.L. | 1992 | A note on weekday, intraday, and overnight patterns in the
interbank foreign exchange and listed currency options market | Journal of Banking and Finance vol
16 pp. 1159-1171

Holden, K., Thompson, J., Ruangrit, Y. | 2005 | The Asian Crisis and Calendar Effects on Stock Returns
in Thailand | European Journal of Operational Research Vol 163 pp. 242-252

Ibrahim, J., Long, Y., Ghani, H.A. | Weak-Form Efficiency of Foreign Exchange Market in the
Organisation for Economic Cooperation and Development Countries: Unit Root Test | International
Journal of Business and Management Vol 6:6

IMF | 1998 | World Economic and Financial Survey: International Capital Markets Developments,
Prospects and Key Policy Issues | International Monetary Fund vol September

Jarrow, R.A., Protter, P. | 2011 | Foreign Currency Bubbles | Review of Derivatives Research Vol 14:1
pp. 67-83

Ke, M.C., Chiang, Y.C., Liao, T.L. | 2007 | Day-of-the-week effect in the Taiwan foreign exchange
market | Journal of Banking & Finance Vol 31 pp. 2847-2865

Keef, S.P., Khaled, M.S. | 2011 | The Friday the thirteenth effect in stock prices: International
evidence using panel data | Victoria University of Wellington, working paper

Kenourgios, D., Samitas, A. | 2008 | The day of the week effect patterns on stock market return and
volatility: evidence for the Athens stock exchange | International Research Journal of Finance and
Economics vol 15

King, M.R., Osler, C., Rime, D. | 2011 | Foreign exchange market structure, players and evolution |
Norges Bank Working Paper October 2011

Kolb., R.W., Rodriguez, R.J. | 1987 | Friday the Thirteenth: ‘Part VII’- a note | The Journal of Finance
vol 42:5 pp. 1385-1387

42
Thesis– Calendar Effects and the Foreign Exchange Markets: a matter of efficiency? Simon Kruithof

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 vol 12 pp. 207-221

Lakonishok, J., Smidt, S. | 1988 | Are seasonal anomalies real? A ninety year perspective | The
Review of Financial Studies vol 1 pp. 403-425

Liano, K., Marchand, P.H., Huang, G.C. | 1992 | The holiday effect in stock returns: evidence from
OTC market | Review of Financial Economics vol 2 pp. 45-54

Lucey, B.M. | 2000 | Friday the 13th and the philosophical basis of financial economics | Journal of
Economics and Finance vol 24:3 pp. 294-301

Malkiel, B.G. | 1973 | A random walk down Wall Street | W.W. Norton & Company Limited
edition 10

Marquering, W. , Nisser, J. , Valla, T. | 2006 | Disappearing anomalies: a dynamic analysis of the

percistence of anomalies | Applied Financial Economics, 16:4, 291-302

McFarland, J.W., Richardson Pettit, R., Sung, S.K. | 1982 | The distribution of foreign exchange price
changes: trading day effects and risk measurement | The Journal of Finance Vol 37:3 pp. 693-715

Meneu, V., Pardo, A. | 2004 | Pre-holiday effect, large trades and small investor behaviour | Journal
of Empirical Finance vol 11 pp. pp. 231-246

Mills, T.C., Coutts, J.A. | 1995 | Calendar effects in the London stock exchange FT-SE indices |
European Journal of Finance vol 1 pp. 79-93

Moller, N., Zilca, S. | 2007 | The evolution of the January effect | Journal of Banking and Finance vol
32 pp. 447-457

Oh, G., Kim, S., Eom, C. | 2007 | Market efficiency in foreign exchange markets | Physica A vol 382
pp. 209-212

43
Thesis– Calendar Effects and the Foreign Exchange Markets: a matter of efficiency? Simon Kruithof

Patel, J.B. | 2009 | Recent evidence on Friday the thirteenth effect in U.S. Stock returns | Journal of
Business and Economics Research vol 7:3 pp. 55-58

Ronaldo, A. | 2009 | Segmentation and time-of-day patterns in foreign exchange markets | Journal
of Banking and Finance vol 33 pp. 2199-2206

So, J. C. | 1987 | The distribution of foreign exchange price changes: trading day effects and risk |
The Journal of Finance Vol 42:1 pp 181-188

Sun, Q., Tong, W.H.S. | 2010 | Risk and the January effect | Journal of Banking and Finance vol 34 pp.
965-974

Thatcher, J.S., Blenman, L.P. | 1998 | Synthetic trades and calendar day patterns: The case of the
Dollar/Sterling markets | The Financial Review vol 33 pp. 177-200

Thomas, L.R., Pukthuanthong-Le, K. | 2008 | Weak-Form Efficiency in Currency Markets | Financial

Analysts Journal vol 64:3 pp. 31-51

Vieira, C., Vieira, I. | 2007 | Seasonalities in Eastern Foreign Exchange Markets: A Barrier to Euro
Adoption | Transition Studies Review 14:2 pp. 283-294

Yamori, N., Kurihara, Y. | 2004 | The day-of-the-week effect in foreign exchange markets: multi-
currency evidence | Research in International Business and Finance vol 18 pp. 51-57

Yamori, N., Mourdoukoutas, P. | 2003 | Does the day-of-the-week effect in foreign currency markets
disappear? Evidecen from the Yen/Dollar Market | International Finance Review vol 4 pp. 443-459

Yu, H.C., Chiou, I., Wagner, J.J. | 2006 | Does the weekday effect of the Yen/Dollar spot rates exist in
Tokyo, London and New York? An analysis of panel probability distributions | Applief Economics Vol
40 pp. 353-375

Zweig, J. | 2009 | Friday the 13th phobia? For investors, it’s usually a good day | Online Article Wall
Street Journal

44
Thesis– Calendar Effects and the Foreign Exchange Markets: a matter of efficiency? Simon Kruithof

Appendix 1 - Overview on calendar effects by Marquering et al.

Source: Marquering et al. (2006)

45
 Thesis– Calendar Effects and the Foreign Exchange Markets: a matter of efficiency? Simon Kruithof

Appendix 2 – QQ plots
Daily returns QQ-plots (normal distribution) for the selected currency spot exchange rates.
1.2 .8 .6 .8

.6 .6
0.8 .4
.4 .4

Quantiles of Normal

Quantiles of Normal

Quantiles of Normal
Quantiles of Normal

0.4 .2
.2 .2

0.0 .0 .0 .0

-.2 -.2
-0.4 -.2
-.4 -.4
-0.8 -.4
-.6 -.6

-1.2 -.8 -.6 -.8

-8 -6 -4 -2 0 2 4 -3 -2 -1 0 1 2 3 -1.5 -1.0 -0.5 0.0 0.5 1.0 1.5 -2 -1 0 1 2

Quantiles of Singapore Dollar / Thai Baht Quantiles of Singapore Dollar / Taiwanese Dollar Quantiles of Taiwanese Dollar/US Dollar Quantiles of Singapore Dollar/ US Dollar
1.5 1.2 1.2

1.0 0.8 0.8

Quantiles of Normal
Quantiles of Normal

Quantiles of Normal
0.5 0.4 0.4

0.0 0.0 0.0

-0.5 -0.4 -0.4

-1.0 -0.8 -0.8

-1.5 -1.2 -1.2

-6 -4 -2 0 2 4 6 8 10 -4 -3 -2 -1 0 1 2 3 4 -4 -3 -2 -1 0 1 2 3

Quantiles of Thai Baht / US Dollar Quantiles of Singapore Dollar / Japanese Yen Quantiles of Japanese Yen / US Dollar

46
Thesis– Calendar Effects and the Foreign Exchange Markets: a matter of efficiency? Simon Kruithof

Appendix 3 – Plotted returns

Daily returns series plots for currency spot exchange rates.
Taiwan Dollar / US Dollar Singapore Dollar / US Dollar
1.5 1.5

1.0 1.0

0.5
0.5

0.0
0.0

-0.5
-0.5
-1.0

-1.0
-1.5

-1.5 -2.0
1996 1998 2000 2002 2004 2006 2008 2010 1996 1998 2000 2002 2004 2006 2008 2010
Thai Baht / US Dollar Singapore Dollar / Thai Baht
10 4

8
2
6

0
4

2 -2

0
-4
-2

-6
-4

-6 -8
1996 1998 2000 2002 2004 2006 2008 2010 1996 1998 2000 2002 2004 2006 2008 2010
Japanese Yen / US Dollar Singapore Dolar / Taiwan Dollar

3 3

2 2

1
1

0
0
-1
-1
-2

-2
-3

-4 -3
1996 1998 2000 2002 2004 2006 2008 2010 1996 1998 2000 2002 2004 2006 2008 2010
Singapore Dollar / Japanese Yen
4

-1

-2

-3

-4
1996 1998 2000 2002 2004 2006 2008 2010

47
Thesis– Calendar Effects and the Foreign Exchange Markets: a matter of efficiency? Simon Kruithof

Appendix 4 Descriptive Table Day-of-the-Week effect

Descriptives for Days of the Week effect
Monday Tuesday Wednesday Thursday Friday All

Mean 0.016 0.032 -0.046 -0.018 -0.035 -0.010

1995-2000
Std. Dev. 0.352 0.369 0.528 0.327 0.303 0.385
Skew. 0.615 2.162 -6.743 -2.683 -0.294 -3.406

Singapore Dollar / Thai Baht

Kurt. 13.181 16.303 97.074 27.716 12.378 78.648
Obs. 312 313 313 313 313 1564
Mean 0.016 0.002 -0.016 0.009 0.000 0.002
2001-2006 Std. Dev. 0.136 0.124 0.131 0.127 0.132 0.130
Skew. -0.828 -0.288 -1.249 -0.358 0.140 -0.523
Kurt. 8.499 3.592 10.957 5.781 4.413 6.852
Obs. 313 313 313 313 313 1565
Mean -0.039 0.022 0.022 -0.017 0.003 -0.002
2007-2011

Std. Dev. 0.385 0.300 0.254 0.238 0.276 0.295

Skew. -2.221 0.682 0.635 0.878 -2.875 -1.186
Kurt. 23.549 18.669 10.439 10.266 30.532 24.584
Obs. 261 261 261 261 261 1305

Mean -0.011 -0.017 0.022 -0.006 0.003 -0.002

1995-2000

Std. Dev. 0.257 0.223 0.241 0.234 0.225 0.237

Skew. -3.476 -2.543 3.563 -3.974 2.193 -0.951
Singapore Dollar / Taiwan Dollar

Kurt. 40.212 19.407 35.519 34.858 28.603 33.867

Obs. 312 313 313 313 313 1564
Mean 0.003 -0.001 -0.013 0.000 -0.004 -0.003
2001-2006

Std. Dev. 0.149 0.128 0.129 0.141 0.129 0.135

Skew. -0.999 0.037 -0.022 0.267 -0.290 -0.242
Kurt. 8.665 3.934 3.861 4.926 3.847 5.534
Obs. 313 313 313 313 313 1565
Mean -0.007 -0.018 0.008 -0.002 0.002 -0.003
2007-2011

Std. Dev. 0.187 0.171 0.186 0.184 0.154 0.177

Skew. -0.502 -0.321 1.202 -0.157 0.225 0.101
Kurt. 6.164 4.336 8.620 6.511 4.693 6.561
Obs. 261 261 261 261 261 1305

Mean 0.037 0.001 0.007 -0.011 -0.001 0.006

1995-2000

Std. Dev. 0.180 0.133 0.133 0.117 0.134 0.142

Skew. 1.960 -2.403 4.086 -2.040 1.982 1.297
Kurt. 20.984 27.240 41.461 15.831 38.260 29.948
Taiwan Dollar / US Dollar

Obs. 312 313 313 313 313 1564

Mean 0.002 0.006 -0.004 -0.005 -0.001 0.000
2001-2006

Std. Dev. 0.121 0.108 0.113 0.105 0.098 0.109

Skew. 0.616 0.382 0.165 -0.206 -0.103 0.227
Kurt. 7.883 8.746 6.163 5.677 4.947 7.065
Obs. 313 313 313 313 313 1565
Mean -0.010 0.016 -0.017 -0.009 0.008 -0.002
2007-2011

Std. Dev. 0.136 0.131 0.123 0.157 0.115 0.133

Skew. -0.439 0.500 -0.912 -0.303 0.469 -0.197
Kurt. 6.581 4.688 7.326 9.112 6.505 7.710
Obs. 261 261 261 261 261 1305

Mean 0.011 -0.002 0.007 -0.003 0.012 0.005

1995-2000

Std. Dev. 0.176 0.215 0.205 0.184 0.162 0.189

Skew. 0.541 -0.481 -2.784 1.543 -1.239 -0.641
Singapore Dollar / US Dollar

Kurt. 9.374 8.389 30.773 14.098 15.433 16.663

Obs. 312 313 313 313 313 1564
Mean 0.012 -0.001 -0.013 -0.009 -0.005 -0.003
2001-2006

Std. Dev. 0.120 0.112 0.115 0.128 0.118 0.119

Skew. 0.536 0.045 0.022 -1.141 0.379 -0.087
Kurt. 6.499 3.662 3.051 10.150 3.953 6.050
Obs. 313 313 313 313 313 1565
Mean -0.004 -0.002 -0.005 -0.009 -0.009 -0.006
2007-2011

Std. Dev. 0.183 0.179 0.189 0.177 0.159 0.177

Skew. -0.443 0.301 0.942 -0.381 0.858 0.241
Kurt. 5.989 5.471 10.038 8.705 5.567 7.477
Obs. 261 261 261 261 261 1305

48
Thesis– Calendar Effects and the Foreign Exchange Markets: a matter of efficiency? Simon Kruithof

Descriptives for Days of the Week effect

Monday Tuesday Wednesday Thursday Friday All

Mean 0.027 -0.035 0.051 0.035 -0.003 0.015

1995-2000
Std. Dev. 0.387 0.412 0.601 0.421 0.408 0.453
Skew. 1.124 -2.293 7.230 1.801 -1.272 3.303
Kurt. 16.623 18.503 110.573 13.811 15.923 77.826
Thai Baht / US Dollar Obs. 312 313 313 313 313 1564
Mean 0.000 -0.005 0.001 -0.016 -0.009 -0.006
2001-2006 Std. Dev. 0.138 0.143 0.135 0.125 0.136 0.135
Skew. 1.324 0.742 0.063 -0.863 1.047 0.548
Kurt. 10.895 7.848 4.851 11.387 12.018 9.379
Obs. 313 313 313 313 313 1565
Mean 0.02 -0.01 -0.04 -0.01 0.02 0.00
2007-2011

Std. Dev. 0.27 0.38 0.36 0.23 0.27 0.31

Skew. 3.36 7.00 -8.77 -2.40 4.21 0.46
Kurt. 37.51 105.78 117.35 22.11 34.65 104.19
Obs. 261 261 261 261 261 1305

Mean -0.001 -0.005 0.017 -0.041 0.049 0.004

1995-2000

Std. Dev. 0.304 0.343 0.389 0.346 0.343 0.347

Skew. -0.065 0.026 -1.918 -0.471 -1.135 -0.848
Kurt. 4.070 7.144 17.227 5.217 6.394 9.396
Japanese Yen / US Dollar

Obs. 312 313 313 313 313 1564

Mean 0.009 -0.011 0.007 -0.010 0.010 0.001
2001-2006

Std. Dev. 0.232 0.243 0.247 0.251 0.268 0.248

Skew. -0.165 -0.025 -0.286 -0.336 0.160 -0.115
Kurt. 3.640 3.137 4.377 4.906 4.076 4.114
Obs. 313 313 313 313 313 1565
Mean -0.029 -0.019 -0.016 -0.003 -0.006 -0.015
2007-2011

Std. Dev. 0.290 0.351 0.346 0.338 0.324 0.330

Skew. -0.547 1.850 -0.667 -0.186 -0.078 0.168
Kurt. 8.659 16.254 5.913 5.630 5.618 8.833
Obs. 261 261 261 261 261 1305

Mean 0.004 -0.005 0.018 -0.001 -0.012 0.001

1995-2000

Std. Dev. 0.341 0.336 0.335 0.351 0.304 0.334

Skew. 0.836 2.378 -0.311 1.601 0.788 1.093
Singapore Dollar / Japanese Yen

Kurt. 6.288 23.570 11.492 17.048 6.610 13.528

Obs. 312 313 313 313 313 1564
Mean 0.004 0.008 -0.022 0.006 -0.018 -0.004
2001-2006

Std. Dev. 0.202 0.188 0.215 0.195 0.215 0.203

Skew. 1.808 0.057 1.327 -0.088 -0.488 0.522
Kurt. 18.487 2.770 10.779 4.738 5.492 8.808
Obs. 313 313 313 313 313 1565
Mean 0.017 0.020 0.005 0.000 0.003 0.009
2007-2011

Std. Dev. 0.390 0.398 0.371 0.380 0.313 0.371

Skew. 0.217 -1.694 0.750 0.540 0.349 -0.054
Kurt. 7.672 16.159 5.595 5.995 6.042 9.082
Obs. 261 261 261 261 261 1305

49
Thesis– Calendar Effects and the Foreign Exchange Markets: a matter of efficiency? Simon Kruithof

Appendix 5 Descriptive Table Turn-of-the-Month Effect

Descriptive Statistics for the turn-of-the-month effect
ROM TOM All ROM TOM All

Mean -0.004 -0.038 -0.010 Mean 0.012 0.029 0.015

1995-2000

1995-2000
Std. Dev. 0.333 0.559 0.385 Std. Dev. 0.404 0.628 0.453
Singapore Dollar / Thai Baht Skew. 0.249 -6.226 -3.406 Skew. -0.603 7.483 3.303
Kurt. 22.043 83.136 78.648 Kurt. 18.234 101.237 77.826
Obs. 1277 287 1564 Obs. 1277 287 1564

Thai Baht / US Dollar

Mean 0.001 0.007 0.002 Mean -0.003 -0.019 -0.006
2001-2006

2001-2006
Std. Dev. 0.132 0.120 0.130 Std. Dev. 0.137 0.126 0.135
Skew. -0.563 -0.247 -0.523 Skew. 0.660 -0.162 0.548
Kurt. 7.145 4.542 6.852 Kurt. 10.050 4.571 9.379
Obs. 1280 285 1565 Obs. 1280 285 1565
Mean -0.004 0.010 -0.002 Mean -0.002 -0.014 -0.004
2007-2011

2007-2011
Std. Dev. 0.306 0.243 0.295 Std. Dev. 0.324 0.208 0.306
Skew. -0.851 -3.995 -1.186 Skew. 0.162 5.062 0.460
Kurt. 21.998 47.289 24.584 Kurt. 99.473 54.848 104.195
Obs. 1065 240 1305 Obs. 1065 240 1305

Mean -0.005 0.014 -0.002 Mean 0.005 -0.017 0.001

1995-2000

1995-2000
Std. Dev. 0.231 0.260 0.237 Std. Dev. 0.341 0.298 0.334
Skew. -1.390 0.403 -0.951 Skew. 1.222 0.151 1.093
Singapore Dollar / Taiwan Dollar

Singapore Dollar / Japanese Yen

Kurt. 37.414 22.646 33.867 Kurt. 14.529 4.477 13.528
Obs. 1277 287 1564 Obs. 1277 287 1564
Mean -0.005 0.008 -0.003 Mean -0.006 0.000 -0.004
2001-2006

2001-2006

Std. Dev. 0.137 0.127 0.135 Std. Dev. 0.201 0.212 0.203
Skew. -0.278 0.008 -0.242 Skew. 0.270 1.491 0.522
Kurt. 5.713 4.203 5.534 Kurt. 6.468 17.252 8.808
Obs. 1280 285 1565 Obs. 1280 285 1565
Mean -0.006 0.010 -0.003 Mean 0.013 -0.012 0.009
2007-2011

2007-2011

Std. Dev. 0.176 0.181 0.177 Std. Dev. 0.381 0.321 0.371
Skew. 0.139 -0.067 0.101 Skew. -0.136 0.477 -0.054
Kurt. 7.120 4.434 6.561 Kurt. 9.371 5.506 9.082
Obs. 1065 240 1305 Obs. 1065 240 1305

Mean 0.010 -0.010 0.006 Mean 0.003 0.008 0.004

1995-2000

1995-2000

Std. Dev. 0.142 0.139 0.142 Std. Dev. 0.345 0.356 0.347
Skew. 2.173 -2.829 1.297 Skew. -0.937 -0.491 -0.848
Kurt. 30.169 27.368 29.948 Kurt. 10.390 5.461 9.396
Taiwan Dollar / US Dollar

Japanese Yen / US Dollar

Obs. 1277 287 1564 Obs. 1277 287 1564

Mean 0.004 -0.020 0.000 Mean 0.002 -0.005 0.001
2001-2006

2001-2006

Std. Dev. 0.111 0.099 0.109 Std. Dev. 0.251 0.238 0.248
Skew. 0.317 -0.483 0.227 Skew. -0.171 0.169 -0.115
Kurt. 7.318 4.610 7.065 Kurt. 4.184 3.731 4.114
Obs. 1280 285 1565 Obs. 1280 285 1565
Mean 0.001 -0.019 -0.002 Mean -0.020 0.010 -0.015
2007-2011

2007-2011

Std. Dev. 0.130 0.149 0.133 Std. Dev. 0.339 0.289 0.330
Skew. 0.162 -1.178 -0.197 Skew. 0.164 0.286 0.168
Kurt. 7.037 8.617 7.710 Kurt. 9.147 5.318 8.833
Obs. 1065 240 1305 Obs. 1065 240 1305

Mean 0.008 -0.008 0.005

1995-2000

Std. Dev. 0.191 0.179 0.189

Skew. -0.680 -0.458 -0.641
Singapore Dollar / US Dollar

Kurt. 17.761 10.225 16.663

Obs. 1277 287 1564
Mean -0.002 -0.010 -0.003
2001-2006

Std. Dev. 0.119 0.117 0.119

Skew. -0.046 -0.289 -0.087
Kurt. 6.372 4.451 6.050
Obs. 1280 285 1565
Mean -0.006 -0.005 -0.006
2007-2011

Std. Dev. 0.179 0.169 0.177

Skew. 0.144 0.759 0.241
Kurt. 7.726 5.967 7.477
Obs. 1065 240 1305

50
Thesis– Calendar Effects and the Foreign Exchange Markets: a matter of efficiency? Simon Kruithof

Appendix 6 – Local Holidays per country in 2011

Singapore
New Year's Day 1 January
Lunar New Year 1 3 February
Lunar New Year 2 4 February
Good Friday 22 April
Labour Day OBS 2 May
Vesak Day 17 May
National Day 9 August
Hari Raya Puasa 30 August
Deepavali* 26 October
Hari Raya Haji 7 November
Christmas Day 26 December
New Year's Eve 31 December

Taiwan
New Year's Day 1 January
Lunar New Year's Eve 2 February
Lunar New Year 1 3 February
Lunar New Year 2 4 February
Lunar New Year 3 7 February
Peace Memorial Day 28 February
Children's Day 4 April
Ching Ming Festival 5 April
Labour Day OBS 2 May
Dragon Boat Festival 6 June
Mid-Autumn Festival 12 September
National Day 10 October

Thailand
New Year's Day 1 January
Makha Bucha Day 18 February
Shakri Day 6 April
Songkran Festival 1 13 April
Songkran Festival 2 14 April
Songkran Festival 3 15 April
Labour Day 1 May
Coronation Day 5 May
Vishaka Bucha Day* 17 May
Mid-year Bank Holiday 1 July
Asarnha Bucha Day 15 July
Queen's Birthday 12 August
King's Birthday 5 December
Constitution Day 12 December
New Year's Eve 31 December

51
Thesis– Calendar Effects and the Foreign Exchange Markets: a matter of efficiency? Simon Kruithof

Japan
New Year's Day 1 January
Bank Holiday 2 2 January
Bank Holiday 3 3 January
Coming of Age (Adults') 10 January
Day
National Founding Day 11 February
Vernal Equinox 21 March
Showa Day 29 April
Constitution Day 3 May
Greenery Day 4 May
Children's Day 5 May
Marine Day 18 July
Respect for the Aged Day 19 September
Autumn Equinox 23 September
Health-Sports Day 10 October
Culture Day 3 November
Labour Thanksgiving Day 23 November
Emperor's Birthday 23 December
New Year's Eve 31 December

United States of America

New Year's Day 1 January
Martin Luther King Jr. Day 17 January
Presidents' Day 21 February
Memorial Day 30 May
Independence Day 4 July
Labor Day 5 September
Columbus Day 10 October
Veterans' Day 11 November
Thanksgiving 24 November
Christmas 25 December

52
Thesis– Calendar Effects and the Foreign Exchange Markets: a matter of efficiency? Simon Kruithof

Appendix 7 Descriptive Table Holiday Effect

Descriptive statistics Pre-Holiday effect
Holiday A Holiday B Holiday Both All Holiday A Holiday B Holiday Both All

Mean 0.007 -0.011 -0.006 -0.010 Mean 0.043 -0.004 -0.017 0.018
1995-2000

1995-2000
Std. Dev. 0.228 0.247 0.087 0.385 Std. Dev. 0.256 0.287 0.255 0.496
Skew. 0.681 0.212 -0.661 -3.406 Skew. 0.644 1.516 -0.621 3.010
Singapore Dollar / Thai Baht

Kurt. 9.605 11.466 3.063 78.648 Kurt. 8.363 12.395 2.449 65.171
Obs. 42 54 11 1564 Obs. 45 43 5 1305

Thai Baht / US Dollar

Mean -0.014 -0.017 -0.018 0.002 Mean -0.027 -0.019 -0.074 -0.006
2001-2006

2001-2006
Std. Dev. 0.119 0.132 0.116 0.130 Std. Dev. 0.103 0.109 0.086 0.135
Skew. 0.748 0.306 1.352 -0.523 Skew. -0.305 -0.362 -0.258 0.548
Kurt. 3.746 3.735 4.244 6.852 Kurt. 3.506 3.391 1.871 9.379
Obs. 59 56 17 1565 Obs. 56 54 7 1565
Mean 0.012 0.008 -0.008 -0.002 Mean -0.006 0.009 -0.041 -0.004
2007-2011

2007-2011
Std. Dev. 0.362 0.165 0.155 0.295 Std. Dev. 0.129 0.179 0.117 0.306
Skew. 2.035 -0.144 -0.228 -1.186 Skew. 0.242 2.827 -0.992 0.460
Kurt. 17.632 3.110 2.235 24.584 Kurt. 4.120 16.272 2.569 104.195
Obs. 50 61 15 1305 Obs. 61 45 6 1305

Mean -0.009 -0.048 -0.036 -0.002 Mean -0.091 -0.029 -0.140 0.001
1995-2000

1995-2000
Std. Dev. 0.095 0.389 0.066 0.237 Std. Dev. 0.301 0.481 0.236 0.334
Skew. -0.493 -5.172 -0.327 -0.951 Skew. -0.242 3.466 0.836 1.093
Singapore Dollar / Taiwan Dollar

Singapore Dollar / Japanese Yen

Kurt. 4.370 35.789 2.430 33.867 Kurt. 5.557 22.996 3.884 13.528
Obs. 42 54 13 1564 Obs. 42 74 11 1564
Mean -0.016 0.003 0.002 -0.003 Mean -0.060 -0.033 -0.080 -0.004
2001-2006

2001-2006

Std. Dev. 0.122 0.108 0.102 0.135 Std. Dev. 0.192 0.203 0.107 0.203
Skew. -0.249 -0.230 0.488 -0.242 Skew. 0.415 0.030 -0.339 0.522
Kurt. 2.773 2.458 2.273 5.534 Kurt. 3.182 3.169 2.006 8.808
Obs. 59 53 12 1565 Obs. 59 78 11 1565
Mean 0.009 0.020 -0.002 -0.003 Mean -0.023 -0.009 -0.026 0.009
2007-2011

2007-2011

Std. Dev. 0.173 0.201 0.266 0.177 Std. Dev. 0.366 0.376 0.238 0.371
Skew. 0.430 0.235 0.765 0.101 Skew. 1.305 -0.085 -0.029 -0.054
Kurt. 4.892 3.498 3.649 6.561 Kurt. 8.187 4.847 2.060 9.082
Obs. 50 43 12 1305 Obs. 50 64 7 1305

Mean 0.005 -0.015 0.002 0.006 Mean -0.007 0.051 -0.025 0.004
1995-2000

1995-2000

Std. Dev. 0.087 0.153 0.033 0.142 Std. Dev. 0.303 0.310 0.290 0.347
Skew. -1.590 -1.840 -0.327 1.297 Skew. -0.770 -0.441 -1.126 -0.848
Kurt. 11.630 16.186 2.222 29.948 Kurt. 4.411 5.488 3.694 9.396
Taiwan Dollar / US Dollar

Japanese Yen / US Dollar

Obs. 54 51 10 1564 Obs. 74 51 10 1564

Mean -0.011 -0.022 -0.114 0.000 Mean 0.020 0.023 -0.010 0.001
2001-2006

2001-2006

Std. Dev. 0.107 0.108 0.097 0.109 Std. Dev. 0.262 0.185 0.225 0.248
Skew. -0.235 -0.191 -0.858 0.227 Skew. -0.373 -0.319 -1.122 -0.115
Kurt. 3.909 3.350 2.394 7.065 Kurt. 3.674 3.170 2.668 4.114
Obs. 53 54 9 1565 Obs. 78 54 8 1565
Mean -0.010 -0.011 -0.037 -0.002 Mean 0.022 -0.009 -0.111 -0.015
2007-2011

2007-2011

Std. Dev. 0.175 0.121 0.128 0.133 Std. Dev. 0.333 0.246 0.280 0.330
Skew. -1.703 -0.214 -1.094 -0.197 Skew. 0.192 -0.583 -0.128 0.168
Kurt. 8.408 5.677 3.589 7.710 Kurt. 3.574 3.541 1.423 8.833
Obs. 43 45 10 1305 Obs. 64 45 6 1305

Mean -0.016 -0.002 -0.053 0.005

1995-2000

Std. Dev. 0.158 0.182 0.106 0.189

Skew. -2.214 1.119 -1.417 -0.641
Singapore Dollar / US Dollar

Kurt. 11.217 9.281 4.847 16.663

Obs. 42 51 18 1564
Mean -0.004 -0.007 -0.030 -0.003
2001-2006

Std. Dev. 0.114 0.087 0.066 0.119

Skew. -0.187 0.683 0.112 -0.087
Kurt. 2.519 3.899 2.257 6.050
Obs. 59 54 18 1565
Mean -0.035 -0.044 -0.079 -0.006
2007-2011

Std. Dev. 0.153 0.127 0.072 0.177

Skew. 0.599 0.493 -0.020 0.241
Kurt. 4.327 4.631 2.312 7.477
Obs. 50 45 16 1305

53
Thesis– Calendar Effects and the Foreign Exchange Markets: a matter of efficiency? Simon Kruithof

Appendix 8 Descriptive Table Friday-the-thirteenth Effect

Descriptive statistics for the Firday-the-thirteenth effect
Friday Friday 13th All Friday Friday 13th All

Mean -0.033 -0.075 -0.010 Mean 6E-05 -8E-02 2E-02

1995-2000

1995-2000
Std. Dev. 0.283 0.701 0.385 Std. Dev. 4E-01 1E+00 5E-01
Skew. -0.966 1.529 -3.406 Skew. -9E-01 -1E+00 3E+00
Singapore Dollar / Thai Baht

Kurt. 11.850 5.233 78.648 Kurt. 1E+01 5E+00 8E+01

Obs. 303 10 1564 Obs. 303 10 1564

Thai Baht / US Dollar

Mean 0.001 -0.074 0.002 Mean -0.010 0.060 -0.006
2001-2006

2001-2006
Std. Dev. 0.130 0.132 0.130 Std. Dev. 0.135 0.155 0.135
Skew. 0.122 0.376 -0.523 Skew. 1.050 1.099 0.548
Kurt. 4.523 4.082 6.852 Kurt. 12.607 3.916 9.379
Obs. 302 10 1565 Obs. 302 10 1565
Mean 0.000 0.097 -0.002 Mean -0.010 0.060 -0.006
2007-2011

2007-2011
Std. Dev. 0.276 0.271 0.295 Std. Dev. 0.135 0.155 0.135
Skew. -2.953 -0.163 -1.186 Skew. 1.050 1.099 0.548
Kurt. 31.226 2.645 24.584 Kurt. 12.607 3.916 9.379
Obs. 253 8 1305 Obs. 302 10 1565

Mean 0.007 -0.138 -0.002 Mean -0.010 -0.064 0.001

1995-2000

1995-2000
Std. Dev. 0.222 0.280 0.237 Std. Dev. 0.307 0.209 0.334
Skew. 2.433 -0.688 -0.951 Skew. 0.781 0.243 1.093
Singapore Dollar / Taiwan Dollar

Singapore Dollar / Japanese Yen

Kurt. 30.580 2.484 33.867 Kurt. 6.555 2.675 13.528
Obs. 303 10 1564 Obs. 303 10 1564
Mean -0.005 0.011 -0.003 Mean -0.017 -0.050 -0.004
2001-2006

2001-2006

Std. Dev. 0.126 0.218 0.135 Std. Dev. 0.216 0.219 0.203
Skew. -0.324 -0.141 -0.242 Skew. -0.528 0.930 0.522
Kurt. 3.972 1.948 5.534 Kurt. 5.559 3.621 8.808
Obs. 302 10 1565 Obs. 302 10 1565
Mean 0.000 0.068 -0.003 Mean 0.002 0.037 0.009
2007-2011

2007-2011

Std. Dev. 0.154 0.144 0.177 Std. Dev. 0.315 0.255 0.371
Skew. 0.217 0.977 0.101 Skew. 0.369 -0.809 -0.054
Kurt. 4.723 2.868 6.561 Kurt. 6.056 2.648 9.082
Obs. 253 8 1305 Obs. 253 8 1305

Mean -0.002 0.039 0.006 Mean 0.046 0.127 0.004

1995-2000

1995-2000

Std. Dev. 0.134 0.133 0.142 Std. Dev. 0.346 0.228 0.347
Skew. 1.986 2.268 1.297 Skew. -1.117 -1.424 -0.848
Kurt. 39.268 6.897 29.948 Kurt. 6.323 4.487 9.396
Taiwan Dollar / US Dollar

Japanese Yen / US Dollar

Obs. 303 10 1564 Obs. 303 10 1564

Mean -0.001 0.013 0.000 Mean 0.009 0.057 0.001
2001-2006

2001-2006

Std. Dev. 0.099 0.063 0.109 Std. Dev. 0.265 0.369 0.248
Skew. -0.106 0.728 0.227 Skew. 0.238 -1.028 -0.115
Kurt. 4.921 4.094 7.065 Kurt. 4.031 4.252 4.114
Obs. 302 10 1565 Obs. 302 10 1565
Mean 0.009 -0.008 -0.002 Mean -0.007 0.036 -0.015
2007-2011

2007-2011

Std. Dev. 0.116 0.099 0.133 Std. Dev. 0.326 0.244 0.330
Skew. 0.494 -1.386 -0.197 Skew. -0.079 0.579 0.168
Kurt. 6.491 4.416 7.710 Kurt. 5.591 3.170 8.833
Obs. 253 8 1305 Obs. 253 8 1305

Mean 0.014 -0.064 0.005

1995-2000

Std. Dev. 0.159 0.239 0.189

Skew. -1.223 -0.872 -0.641
Singapore Dollar / US Dollar

Kurt. 16.750 2.437 16.663

Obs. 303 10 1564
Mean -0.005 0.000 -0.003
2001-2006

Std. Dev. 0.116 0.182 0.119

Skew. 0.456 -0.441 -0.087
Kurt. 3.896 3.135 6.050
Obs. 302 10 1565
Mean -0.010 0.035 -0.006
2007-2011

Std. Dev. 0.159 0.165 0.177

Skew. 0.867 0.652 0.241
Kurt. 5.686 2.157 7.477
Obs. 253 8 1305

54
Thesis– Calendar Effects and the Foreign Exchange Markets: a matter of efficiency? Simon Kruithof

Appendix 9 Descriptive Table January Effect

Descriptive Statistics for the January Effect
rest of the year January All rest of the year January All

Mean -0.009 -0.026 -0.010 Mean 0.012 0.049 0.015

1995-2000

1995-2000
Std. Dev. 0.379 0.445 0.385 Std. Dev. 0.444 0.548 0.453
Skew. -3.876 -0.071 -3.406 Skew. 3.747 0.499 3.303
Singapore Dollar / Thai Baht

Kurt. 89.462 10.632 78.648 Kurt. 89.661 13.855 77.826

Obs. 1433 131 1564 Obs. 1433 131 1564

Thai Baht / US Dollar

Mean 0.000 0.028 0.002 Mean -0.003 -0.033 -0.006
2001-2006

2001-2006
Std. Dev. 0.129 0.137 0.130 Std. Dev. 0.133 0.159 0.135
Skew. -0.601 0.100 -0.523 Skew. 0.690 -0.182 0.548
Kurt. 7.268 2.988 6.852 Kurt. 10.259 3.819 9.379
Obs. 1431 134 1565 Obs. 1431 134 1565
Mean -0.001 -0.011 -0.002 Mean -0.006 0.019 -0.004
2007-2011

2007-2011
Std. Dev. 0.279 0.437 0.295 Std. Dev. 0.308 0.282 0.306
Skew. -0.825 -2.008 -1.186 Skew. 0.553 -0.822 0.460
Kurt. 22.397 20.694 24.584 Kurt. 110.073 10.005 104.195
Obs. 1195 110 1305 Obs. 1195 110 1305

Mean -0.004 0.022 -0.002 Mean 0.002 -0.008 0.001

1995-2000

1995-2000
Std. Dev. 0.226 0.330 0.237 Std. Dev. 0.332 0.350 0.334
Skew. -0.551 -2.379 -0.951 Skew. 1.134 0.706 1.093
Singapore Dollar / Taiwan Dollar

Singapore Dollar / Japanese Yen

Kurt. 28.650 38.761 33.867 Kurt. 14.236 7.039 13.528

Obs. 1433 131 1564 Obs. 1433 131 1564
Mean -0.005 0.015 -0.003 Mean -0.003 -0.021 -0.004
2001-2006

2001-2006

Std. Dev. 0.134 0.154 0.135 Std. Dev. 0.202 0.215 0.203
Skew. -0.287 -0.009 -0.242 Skew. 0.617 -0.305 0.522
Kurt. 5.730 3.985 5.534 Kurt. 9.451 3.065 8.808
Obs. 1431 134 1565 Obs. 1431 134 1565
Mean -0.004 0.006 -0.003 Mean 0.006 0.038 0.009
2007-2011

2007-2011

Std. Dev. 0.179 0.146 0.177 Std. Dev. 0.375 0.323 0.371
Skew. 0.082 0.584 0.101 Skew. -0.101 0.848 -0.054
Kurt. 6.426 8.268 6.561 Kurt. 9.214 5.099 9.082
Obs. 1195 110 1305 Obs. 1195 110 1305

Mean 0.006 0.009 0.006 Mean 0.000 0.041 0.004

1995-2000

1995-2000

Std. Dev. 0.138 0.177 0.142 Std. Dev. 0.348 0.326 0.347
Skew. 1.787 -1.380 1.297 Skew. -0.936 0.375 -0.848
Kurt. 31.368 20.149 29.948 Kurt. 9.623 5.450 9.396
Taiwan Dollar / US Dollar

Japanese Yen / US Dollar

Obs. 1433 131 1564 Obs. 1433 131 1564

Mean 0.002 -0.023 0.000 Mean 2E-05 0.013 0.001
2001-2006

2001-2006

Std. Dev. 0.107 0.130 0.109 Std. Dev. 0.248 0.259 0.248
Skew. 0.473 -1.083 0.227 Skew. -0.165 0.348 -0.115
Kurt. 6.990 5.989 7.065 Kurt. 4.171 3.490 4.114
Obs. 1431 134 1565 Obs. 1431 134 1565
Mean -0.004 0.010 -0.002 Mean -0.014 -0.022 -0.015
2007-2011

2007-2011

Std. Dev. 0.134 0.127 0.133 Std. Dev. 0.335 0.275 0.330
Skew. -0.201 -0.115 -0.197 Skew. 0.172 0.034 0.168
Kurt. 7.970 4.096 7.710 Kurt. 8.969 3.572 8.833
Obs. 1195 110 1305 Obs. 1195 110 1305

Mean 0.003 0.024 0.005

1995-2000

Std. Dev. 0.182 0.251 0.189

Skew. -0.968 0.691 -0.641
Singapore Dollar / US Dollar

Kurt. 17.575 10.254 16.663

Obs. 1433 131 1564
Mean -0.003 -0.007 -0.003
2001-2006

Std. Dev. 0.120 0.112 0.119

Skew. -0.095 0.014 -0.087
Kurt. 6.274 2.778 6.050
Obs. 1431 134 1565
Mean -0.007 0.013 -0.006
2007-2011

Std. Dev. 0.180 0.144 0.177

Skew. 0.212 1.085 0.241
Kurt. 7.387 7.609 7.477
Obs. 1195 110 1305

55