Calendar anomalies in the Gulf Cooperation Council stock markets

Emerging Markets Review 12 (2011) 293–307

Contents lists available at ScienceDirect

Emerging Markets Review j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / e m r

Calendar anomalies in the Gulf Cooperation Council stock markets Rima Turk Ariss a, Rasoul Rezvanian b,⁎, Seyed M. Mehdian c,1 a b c

Lebanese American University, P. O. Box 13-5053, Chouran Beirut, 1102 2801, Lebanon Northeastern Illinois University, 5500 N. St. Louis Avenue, Chicago, IL 60625, United States University of Michigan-Flint, Flint, MI 48502, United States

a r t i c l e

i n f o

Article history: Received 4 June 2010 Received in revised form 2 April 2011 Accepted 4 April 2011 Available online 12 April 2011 JEL classification: G14 G15 Keywords: Calendar anomalies Stock market efficiency Gulf Cooperation Council

a b s t r a c t We examine calendar anomalies in Gulf Cooperation Council (GCC) stock markets and document a Friday-type effect that occurs on the last trading day of the week and which we call “Wednesday effect”, since Wednesday is the last day before the weekend in the leading market for the region. This effect, however, is more pronounced outside the month of Ramadan. We also find a statistically significant positive December effect, contrary to the January effect documented in Western countries. The presence of such anomalies may provide money managers with opportunities to optimally time their trades based on daily and monthly price fluctuations. © 2011 Elsevier B.V. All rights reserved.

1. Introduction The Gulf Cooperation Council (GCC) region is gaining increasing attention on a global scale.2 The region has weathered well the global financial crisis, and the leading economic GCC country, Saudi Arabia, which is a member of the G20, has announced the largest fiscal stimulus plan as a percentage of GDP. Other developments have also positioned this region on a global scale. Prior to 2008, GCC member countries had witnessed an unprecedented economic boom ignited by high oil prices and resulting in excess liquidity in the form of petrodollars since the turn of the century. GCC Sovereign Wealth Funds have become predominant on the international arena, undertaking strategic investments in different crisis-afflicted industries throughout the world. Additionally, GCC countries have set the course for regional economic and

⁎ Corresponding author. Tel.: + 1 773 442 6163. E-mail addresses: [email protected] (R. Turk Ariss), [email protected] (R. Rezvanian), seyed@umflint.edu (S.M. Mehdian). 1 Tel.: + 1 810 762 3318. 2 GCC countries include Bahrain, Kuwait, Oman, Qatar, Saudi Arabia, and United Arab Emirates. 1566-0141/$ – see front matter © 2011 Elsevier B.V. All rights reserved. doi:10.1016/j.ememar.2011.04.002

294

R. Turk Ariss et al. / Emerging Markets Review 12 (2011) 293–307

monetary integration among member states, increasing their potential collective role as a global economic player. More importantly, all GCC countries have gained accession to the World Trade organization (WTO) and they are gradually opening up their markets to foreign investors. Broad-ranging structural reforms and regulations have recently provided greater foreign investors access to GCC financial markets, which had previously remained unknown to institutional investors. A substantial body of evidence analyzing correlation in returns across regional and international markets provides evidence that GCC markets might provide an important venue for international portfolio diversification (Al Janabi et al., 2010; Assaf, 2003; Bley and Heng Chen, 2006; Hammoudeh and Aleisa, 2004; Hammoudeh and Choi, 2005, 2007).3 In this paper, we examine whether GCC markets are weakly efficient through the presence of seasonal patterns that can be employed to realize abnormal returns.4 We focus on two major calendar anomalies that are reported in different stock markets across the globe, the day-of-the-week or Monday effect and the month-of-the-year or January effect. The Monday effect happens when returns are lower or negative on Mondays in comparison with the returns on other days of the week. The month-of-the-year or January effect, which is generally a characteristic of small stocks, occurs when returns are abnormally higher in January compared to other months of the year. Our motivation is that GCC stock markets are unique in their trading days and holidays. Specifically, they observe different trading days and non-overlapping holiday periods compared to most countries so that calendar anomalies, if any, are likely to be different from those reported for other economies. Monday does not represent the first day of the week in any of the GCC countries and there is no basis for international investors to expect negative returns on that day similar to those documented for other stock markets. Also, the last trading day of the week for the leading regional market or Tadawul in Saudi Arabia falls on Wednesday, and events affecting that market on a particular day may be expected to spill over to other GCC countries, especially in light of increased regional integration and projected monetary union (Bley and Heng Chen, 2006). If the day-of-the-week effect reported for other stock markets holds for the GCC region, we expect positive (negative) returns on the last (first) trading day of the week, or Wednesdays (Saturdays). Furthermore, religious holidays and activities for GCC nationals follow the Hijri calendar year that is based on lunar months, notwithstanding the use of the Gregorian calendar by businesses and governments. The observed holidays do not coincide with other non-Muslim religious holidays, and one example is Eid-El-Fitr that marks the end of the holy month of Ramadan. Ramadan bears a special significance to GCC investors because Muslims are required to fast during this month and to show care for the less fortunate in society, in addition to exercising spirituality. As a result, the trading activity of GCC investors is likely to be reduced during Ramadan, and a month-of-the-year effect, if present, may be associated with the occurrence of this month. The study of calendar anomalies in the context of GCC countries has received little attention in the academic literature, generally focusing on a single member country and not the region as a whole. For example, Seyyed et al. (2005) investigate seasonality in return volatility associated with the month of Ramadan for the Saudi Tadawul stock market, and they document a systematic declining pattern of volatility that reflects reduced trading activity during the fasting period. Al-Saad and Moosa (2005) report a July effect for the Kuwait Stock Exchange which they characterize as a “summer holiday effect”; Al-Saad (2004) also examines the Kuwait Stock Exchange but his results do not show the presence of any holiday effect, although stock returns are significantly higher post liberation of Kuwait following the Iraqi invasion. More recently, Al-Khazali (2008) finds no evidence in support of a day-of-the-week effect in the United Arab Emirates (UAE) stock markets, suggesting that these markets are to a certain extent efficient. Al-Barrak (2009) examines the day-of-the-week effect in three GCC markets (Saudi Arabia, Kuwait, and Dubai) and reports that a daily anomaly exists only in Kuwait stock market. Finally, Bley and Saad (2010) examine three market anomalies in the context of GCC countries; namely day-of-the-week, month-of-the-year, and holiday effects. They find evidence of a day-of-the-week effect in most GCC stock markets, but occurring on days that are different from Friday and Monday. However, they report

3 In contrast, Al-Khazali et al. (2006) do not find that GCC equity investments provide portfolio diversification gains to investors with long-term horizons, although short-term benefits are likely. 4 A recent study by Bley (2011) tests for the weak-form efficiency in GCC stock markets and rejects the random walk hypothesis, suggesting that foreign institutional investors might have had a role to play in increasing serial correlation in returns.

R. Turk Ariss et al. / Emerging Markets Review 12 (2011) 293–307

295

Table 1 GDP growth (percentage annual change) of GCC countries. Source: Economist Intelligence Unit.

Bahrain Kuwait Oman Qatar Saudi Arabia UAE GCC average MENA average

2001

2002

2003

2004

2005

2006

7.1 3.2 10.1 5.8 2.4 6.0 5.8 5.4

7.6 7.0 4.1 9.0 1.3 4.4 5.6 4.6

9.4 15.9 4.4 5.7 9.2 14.3 9.8 6.0

9.8 9.4 8.4 24.3 7.5 19.6 13.2 8.9

12.5 11.5 9.2 9.5 11.9 20.7 12.5 9.6

10.1 9.7 10.9 15.9 6.5 18.6 11.9 9.2

no month-of-the year or January effect as is the case for a number of Western countries. Bley and Saad (2010) also find that the magnitude of the holiday effect depends on both the culture and religion of the market and of its participants. We also investigate the presence of a day-of-the-week and month-of-the-year effect, but we extend our coverage of calendar anomalies to all six GCC stock markets, which operate under a unique environment that is different from advanced and other emerging markets (as discussed in more detail in Section 2). For example, these stock markets have unique settlement procedures, trading days, and they prohibit short selling due to religious underpinnings.5 Trading activity is still dominated by small home country investors, and there are resilient restrictions on non-GCC ownership of shares, whether individual or institutional. Further, all GCC countries operate under a no-income tax and no-capital gain tax setting; therefore, the taxinduced January effect that prevails in the US should not exist in GCC markets. In light of these specificities, it is unlikely that the traditional Monday and January effects prevail in those markets. Should other calendar anomalies be present, they also cannot be attributed to the explanations previously provided by the literature for other countries. We find that returns are significantly positive during the last trading day of the week in GCC stock markets, similar to the results reported for other markets. However, that day coincides with Wednesday and not Friday, suggesting the presence of a “Wednesday effect” for GCC capital markets. However, the “Wednesday effect” is more pronounced outside the month of Ramadan. We also test for the presence of a month-of-the-year effect, but do not find evidence to suggest that investors' behavior is altered for any Gregorian or Hijri month of the year, despite a significant drop in market volatility during Ramadan. The results of this study are valuable to money managers, institutional investors, and sophisticated investors since the knowledge of anomalous behavior of stock returns is a vital piece of information when adopting strategies to structure optimal portfolios. These strategies include market timing, short selling, and tax savings that are all based on taking advantage of the seasonality behavior of stock prices. The rest of the paper is organized as follows. Section 2 presents an overview of GCC stock markets. Section 3 describes our data and methods. Section 4 discusses the empirical results, and Section 5 summarizes and concludes. 2. Overview of GCC stock markets Since the turn of the century, the GCC countries have witnessed an unprecedented economic boom ignited by high oil prices and excess liquidity, and which are fuelling the growth of capital markets. Table 1 presents the annual growth in GDP for the GCC countries and the average for the Middle East and North Africa (MENA) region,6 showing that GDP growth in the GCC region has consistently and by far exceeded

5 Short-selling is considered as haram (a forbidden activity) under Shariah or the Islamic jurisprudence because of the legal rule that “one cannot sell what one does not own”. 6 The MENA region covers the following countries: Algeria, Bahrain, Djibouti, Egypt, Iran, Jordan, Kuwait, Lebanon, Libya, Mauritania, Morocco, Oman, Qatar, Saudi Arabia, Sudan, Syria, Tunisia, United Arab Emirates and Yemen.

296

R. Turk Ariss et al. / Emerging Markets Review 12 (2011) 293–307

Table 2 Market capitalization ($mn) of GCC stock markets, 1994–2008. Source: Arab Monetary Fund.

1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008

Abu Dhabi securities market

Bahrain stock exchange

Doha securities market

Dubai financial market

Kuwait stock exchange

Muscat securities market

Saudi capital market

Total

– – – – – – – –

5129 4707 5019 7826 6772 7161 6624 6601 7716 9702 13,513 17,364 21,122 26,796 19,955

– – – – – – – – 10,567 26,702 40,435 87,143 60,905 95,518 76,657

– – – – – – – –

10,967 14,400 20,600 27,245 18,424 19,599 19,848 26,662 35,099 59,528 73,581 123,893 105,950 193,513 113,527

1856 1971 2753 7313 4537 4303 3518 2634 5268 7246 9318 12,062 13,037 22,767 15,643

38,693 40,904 45,856 59,378 42,631 60,953 67,166 73,201 74,851 157,306 306,256 646,121 326,852 522,721 246,810

56,646 61,982 74,228 101,761 72,363 92,015 97,157 109,099 163,347 305,131 533,683 1,130,989 695,506 634,881 308,697

20,376 30,363 55,490 132,413 80,745 112,160 61,888

9470 14,284 35,091 111,993 86,895 138,698 65,218

the average for the whole MENA region.7 The buoyant growth in the GCC region is mainly attributed to high oil and non hydrocarbon prices, since energy is the main economic resource and export product in this region. Unlike the oil bonanza in the 1973–74 and 1979–80 periods when the financial surpluses of the governments and institutions were recycled in Western countries, the billions of petrodollars resulting from the 21st century oil boom are invested mainly at home (Billmeier and Massa, 2009). After September 11, 2001, the fear that potential financial controls will freeze Arab assets led to capital repatriation from the US and Europe. The GCC governments are using the oil windfall revenue to repay public debt and support long term economic and social development programs through expansionary fiscal policies (IMF, Middle East Economic Outlook Regional, 2007).8,9 Capital market growth in the GCC region is also driven by excess demand and oversubscription in Initial Public Offerings (IPOs). Between 1995 and 2003, there were a total of nine IPOs only. In 2004 alone, there were twelve IPOs and the number reached twenty four in 2005 (Gulf Capital Group, 2006). The average dollar amount of IPOs almost doubled in 2005 to reach USD 248 million, up from USD 128 million in 2004.10 In compliance with the World Trade Organization (WTO) accession requirements of which all GCC countries are now members, several economic reforms were adopted to liberalize and deregulate capital markets in order to attract foreign investments. Some of the liberalization programs privatized government stakes in businesses, increased foreign ownership in local companies, and reduced the income tax for foreign corporations. It is important to note that GCC stock markets are recent compared to other emerging stock markets, and that some were not even in existence prior to the year 2000 (e.g. Abu Dhabi Securities Market, Dubai Financial Market, and Doha Securities Market). Nonetheless, GCC countries are making significant strides to develop proper legal and regulatory frameworks as well as to improve prudential standards in order for 7 Our sample period extends till June 2008, and thus does not cover the unfolding of the global financial crisis following the failure of Lehman Brothers in September 2008. It is interesting to note, however, that the GCC region has shown greater resilience to the global financial turmoil, with growth figures by the Economist Intelligence Unit (EIU) for the GCC region reaching 10.5 and 10% in 2007 and 2008 respectively. The first serious worry occurred in December 2009 after Dubai World announced a standstill on its debt payments. 8 Government spending on infrastructure projects encouraged the establishment of large real estate companies such as Emaar Properties, Al-Nakheel, Al-Ittihad, and Jumeirah. 9 According to the World Wealth Report (2006) published by Merrill Lynch and Capgemini, the millionaires of the Middle East have increased in 2005 by 9.8% from 2004, totaling 300,000 people having a total wealth of USD 1.2 trillion. 10 In November 2006, the Dubai Financial Market (DFM) IPO collected over USD 50 billion for its USD 435 million IPO, an oversubscription of 300 times in 11 days.

R. Turk Ariss et al. / Emerging Markets Review 12 (2011) 293–307

297

Table 3 Trading days, hours, and settlement cycles. Stock market

Trading days

Trading hours

Settlement cycle

Oman UAE 1,2 (Dubai and Abu Dhabi) Bahrain Kuwait Saudi Tadawul Qatar

Saturday–Thursday Saturday–Thursday Sundays–Thursdays Sunday–Thursday Saturday–Wednesday Sunday–Thursday

10:00–13:00 10:00 am–2:00 pm 9:30 am 3–12:30 noon 9:00 am–12:30 pm 11:00 am–3:30 pm 10:00 am 2–12:30 pm

T+3 T+3 T+2 T+0 T+1 –

1 2 3

Starting March 22, 2009, trading was extended by 3 h every day till 5:00 pm. Including a pre-opening period of 30 min starting at 9:30 am. Including a pre-opening period of 15 min starting from 9:15 am.

securities markets to emerge and grow. Consequently, financial systems are being upgraded and strengthened to play a wider role in financing economic activity (Bolbol and Omran, 2005). Table 2 shows the evolution of market capitalization for each of the GCC stock markets over the period 1994–2008. GCC capital markets have developed exponentially over this period, with market capitalization rising more than ten times from $97 billion in 2000 to $1.13 trillion in 2005. However, following a series of market corrections in 2006, market capitalization declined to less than $700 billion by year end and, similar to the trend for most world stock markets, it further dropped to $309 billion by mid of 2008 due to the global financial meltdown. The rising importance of GCC capital markets is also evident through the rising number and value of traded securities, in addition to the doubling of the number of listed firms. By June 2008, there were 661 listed firms on all six GCC capital markets, with more than 253.4 million shares traded valued at $781.5 million, a significant decline from $1.6 trillion value traded in 2006 (Arab Monetary Fund). The Saudi Capital Market (SCM) or Tadawul is the largest in the Gulf region in terms of market capitalization, followed by the Kuwait Stock Exchange (KSE).11 However, when the two UAE markets are combined together, their size exceeds the second largest market in region, showing the rapidity with which these markets have established themselves as major regional players. Table 2 also indicates that the Bahrain Stock Exchange (BSE) and Muscat Securities Market (MSM) are the smallest in terms of market capitalization. Table 3 provides information on the unique structures of GCC markets in terms of different trading days, schedules, and settlement practices. Security trades in the GCC region are conducted from Sunday to Thursday in Bahrain, Kuwait, Oman, and Qatar, though at different times, while markets are open from Saturday to Thursday in the UAE (both Dubai and Abu Dhabi exchanges) and from Saturdays to Wednesdays in the case of Saudi Arabia. Therefore, the only common closing day across all GCC markets falls on Friday, and all exchanges are closed for two consecutive days of the week except for the UAE stock markets which operate for six days a week. Also, the end of the week occurs on Wednesday for Saudi Arabia, while markets are open on Thursdays for all other GCC markets. In Table 3, the short or nonexistent settlement cycles prevailing in Kuwait and Saudi Arabia rule out the Settlement Procedure hypothesis as a possible explanation for a day-of-the-week effect (Gibbons and Hess, 1981; Lakonishok and Levi, 1982).12 3. Data and methods 3.1. Data We use the daily closing values of all GCC market indices from inception until June 2008 from Global Financial Data.13 In order to isolate any bias from the global financial crisis and not to contaminate our 11

The KSE is the largest market in the GCC region in terms of number of listed shares. According to the Settlement Procedures hypothesis and since settlement and payment take place several business days after the securities trade, longer settlement periods provide an incentive for investors to pay higher (lower) prices on Friday (Monday). 13 For a study on intra-day seasonality for a stock market in the region, see Bildik (2001). 12

298

R. Turk Ariss et al. / Emerging Markets Review 12 (2011) 293–307

GCC Market Indices

1,000

Saudi Capital Market

900

Dubai Financial Market

800 Doha Securities Market

700 Kuwait Stock Exchange

600 AMF Composite Index

500

Abu Dhabi Securities Market Bahrain Stock Exchange

400

Muscat Securities Market

300 200 100 0 1994

1995

1996

1997

1998

1999

2000

2001

2002

2003

2004

2005

2006

Fig. 1. Evolution of GCC market indices. Source: Arab Monetary Fund.

analysis with related spillovers, we do not consider data after June 2008.14 We compute the daily stock returns Rit for all seven GCC stock market indices as follows15,16: Rit = lnðIit = Iit−1 Þ  100

ð1Þ

where Rit is the daily percentage return of stock index i on day t, and Iit and Iit − 1 are the closing values of the index on day t and t − 1 respectively. Classical time series analysis requires the data being used to be stationary, which we test for by conducting augmented Dickey–Fuller unit root tests on each GCC stock market return series. The results, which are not reported, show that the stock price indexes are stationary in their first differences.17 Fig. 1 graphs the evolution of all seven GCC market indices in addition to the Arab Monetary Fund (AMF) composite index since 1994. Prior to the turn of the century, all GCC markets exhibited a relatively stable performance, followed by a sharp rising trend fueled by increased liquidity resulting from high energy prices. The increased oil wealth led to irrational investment enthusiasm and speculative behavior among investors causing an upsurge in equity prices. The shortage in the supply of stocks failed to absorb the high liquidity of the GCC region, and in February 2006, GCC markets reached a turning point after which a series of corrections took place all throughout the rest of the year. The Saudi Tadawul first plummeted, triggering declines in all other GCC markets. The high degree of contagion among these markets raises concerns about 14 The decision to focus on the period prior to the crisis is supported by a preliminary examination of raw data on stock market indices that shows that they indiscriminately exhibited a downward trend for a number of periods following the crisis. 15 We exclude from the analysis index values after June 2008, concentrating on the period prior to the global financial crisis and meltdown. 16 In our analysis, we consider the two UAE markets separately, resulting in a total of seven market indices for the six GCC countries. 17 These findings of the stationarity of daily stock returns are consistent with the results obtained for major stock market indices of other countries (see for instance Bachman et al., 1996).

R. Turk Ariss et al. / Emerging Markets Review 12 (2011) 293–307

299

Table 4 Summary statistics of daily returns for GCC countries.

Nb of obs. Mean Max Min Range Std Dev.

Abu Dhabi

Bahrain

Kuwait

Muscat

Qatar

Saudi Arabia

Dubai

970 0.0785 8.2500 −8.6484 16.8984 1.4588

4307 0.0361 6.2971 −5.3462 11.6433 0.5674

2704 0.1965 154.1779 −44.4351 198.6130 3.6272

3594 0.0679 16.5514 −13.5602 30.1117 0.9893

2048 0.1146 9.8825 −8.0741 17.9567 1.2590

4144 0.0540 58.8694 −36.9789 95.8483 1.7178

1416 0.1851 46.4940 −46.0645 92.5585 2.9404

inherent structural weaknesses and inadequate corporate governance mechanisms. Most listed firms do not exhibit much variation in their ownership structure, and governments hold many of the shares of the listed companies, which are also small in number and mostly family-owned businesses (MENA-OECD Investment Programme, 2005). For example, the Chairman of the Board of Directors of SABIC, the largest listed company on Tadawul, is a royal family member whereas the other directors include government officials and only two private sector representatives. Further, markets are not transparent; corporations generally do not abide by international accounting standards nor do they publish timely and accurate financial information; they also lack credit ratings thus aggravating information asymmetry issues. In fact, herd behavior is predominant in the GCC stock markets. Individual investors free ride on information and buy and sell following trends, market sentiments, and intuitions rather than using investment fundamentals, in the absence of important market players such as institutional investors. Only lately have the exchanges started to undergo major reforms to address such concerns. 3.2. Method of measuring the day-of-the-week effect Following Gibbons and Hess (1981), Peterson (1990), and Solnik and Bousquet (1990), we examine whether there is a day-of-the-week effect in the GCC region by estimating the following model18: Rit = α1i D1i + α2i D2i + α3i D3i + α4i D4i + α5i D5i + εti

ð2Þ

where Rit is the daily return of stock index i as defined earlier, D1i through D5i are dummy variables for each of the five trading days of the week for index i and εt is a random error term. For any particular stock market, if day t is a Monday, then D1 takes the value 1 and otherwise the value 0, if t is a Tuesday then D2 takes the value 1 otherwise the value 0, and so forth. We estimate the α parameters using Ordinary Least Squares (OLS) with robust standard errors.19 If a given stock index i exhibits, for example, a traditional Monday effect, then the estimated coefficient α1i is expected to be significantly negative or significantly lower than the estimated coefficients for the rest of the week. Alternatively, since the trading days in GCC capital markets differ from those in other countries and in order to infer another calendar day anomaly, we expect the corresponding coefficient to be significantly different compared to the other days. 3.3. Method of measuring the month-of-the-year effect We test for the presence of a month-of-the-year effect using a framework similar to Eq. (2) as20: Rit = α1i D1i + α2i D2i + α3i D3i + ……:: + α10i D10i + α11i D11i + α12i D12i + εti

ð3Þ

18 For robustness, we estimate an alternative equation to (2) that includes a constant and four dummy variables instead of five for Tuesday through Friday, with the constant denoting the omitted dummy variable for Monday. We notice that the coefficient of the constant is identical to α1. The other α coefficients differ in size but not in sign or significance, so that our main results hold. 19 We test for heteroskedasticity in returns for each market index using the Cook–Weisberg test procedure. We then apply the Huber–White sandwich estimator to obtain robust t-values in the presence of heteroskedasticity. 20 For robustness, we also estimate an alternative equation to (3) that includes a constant and in which the number of dummy variables is eleven instead of twelve for all months except July. Our main results hold.

300

R. Turk Ariss et al. / Emerging Markets Review 12 (2011) 293–307

Table 5 Correlation of returns among GCC countries.

Abu Dhabi Bahrain Kuwait Muscat Qatar Saudi Arabia Dubai

Abu Dhabi

Bahrain

1 (0.1043)⁎⁎⁎ 0.0019 (0.1671)⁎⁎⁎

1 1 (0.124)⁎⁎⁎ 0.0000 0.0598)⁎⁎⁎

Kuwait

Muscat

Qatar

Saudi Arabia

Dubai

1

0.0000 (0.1693)⁎⁎⁎ 0.0000 (0.2362)⁎⁎⁎ 0.0000 (0.1866)⁎⁎⁎

0.0008 (0.0862)⁎⁎⁎ 0.0001 (0.0572)⁎⁎⁎

(0.0661)⁎⁎⁎ 0.0022 (0.1184)⁎⁎⁎ 0.0000 (0.1813)⁎⁎⁎

(0.129)⁎⁎⁎ 0.0000 (0.0779)⁎⁎⁎

0.0000 (0.5094)⁎⁎⁎ 0.0000

0.0010 (0.0679)⁎⁎ 0.0200

0.0000 (0.046) 0.1306

0.0000 (0.0876)⁎⁎⁎ 0.0029

1 1 (0.0874)⁎⁎⁎ 0.0002 (0.0867)⁎⁎⁎ 0.0037

1 (0.0507)⁎ 0.0829

1

***,**,* indicates significant at 0.01, 0.05 and 0.1 levels, respectively.

where Rit is the daily return of stock index i as defined above, D1i ….. D12i are dummy variables for each month of the year for index i. We assign the value 1 to the dummy variable D1i if day t falls in January and the value 0 otherwise, and so on. The αs are coefficients to be estimated, and εti is a random error term. The traditional January effect is present in stock index i if the estimated α1i is statistically significantly positive and/or statistically significantly higher than estimated coefficients for the rest of the months of the year. Alternatively, since holidays in the GCC region do not follow the Gregorian calendar, we use Eq. (3) to test for the presence of a different month-of-the-year effect that derives from the Hijri lunar calendar year. We start by matching the trading days of the Gregorian calendar to the days of the moving lunar calendar. Next, we generate new dummy variables to represent each of the 12 Hijri months related to the lunar calendar year.21 The month of Ramadan is the ninth month of the Hijri calendar, and it starts earlier each year by about ten days. If GCC investors modify their trading behavior during Ramadan, then a Ramadan effect would be detected if the corresponding dummy variable coefficient is significantly different from the parameter estimates for other months. 4. Empirical findings Table 4 displays summary statistics of the daily returns for each of the GCC market indices where a great disparity across different countries is recognizable. Over the sample period, the Kuwait stock market index has generated the highest daily mean return and standard deviation, whereas the Bahrain index has produced the lowest mean return and standard deviation among all GCC countries. We further present the correlation of daily returns among GCC countries in Table 5. These figures suggest a relatively low but statistically significant positive correlation among a number of GCC markets. The highest positive correlation of returns is between the two UAE markets, Abu Dhabi and Dubai, perhaps due to the geographic proximity and the strong economic ties between the two emirates. We investigate the presence of calendar anomalies in GCC stock markets by estimating Eqs. (2) and (3) for each index using ordinary least squares with robust standard errors. Table 6 presents the results for the day-of-the-week analysis, from which we draw three main inferences. First, in all GCC markets, Wednesday returns are positive and statistically significant. Second, the positive Wednesday returns are also the highest compared to the returns for other days of the week for five market indices (Abu Dhabi, Bahrain, Qatar, Saudi Arabia, and Dubai). For the other two GCC stock markets, Kuwait and Muscat, Wednesday returns are ranked second to the highest returns for the rest of the week. Third, the persistent positive and significant Wednesday returns are maintained on the following day (or Thursday) for three out of the five stock indices where Thursday is the last trading day (Kuwait, Qatar, and Dubai). These findings provide evidence to indicate the presence of an end-of-week or a Friday-type effect in all GCC stock markets but which occurs on Wednesdays. It is interesting to note the similarity in the 21 These months are: Muharram, Safar, Rabi' Al-Awwal, Rabi' Al-Thani, Jamadi Al-Awwal, Jamadi Al-Thani, Rajab, Sha'aban, Ramadan, Shawwal, Dhu Al-Qi'dah, and Dhu Al-Hijjah.

R. Turk Ariss et al. / Emerging Markets Review 12 (2011) 293–307

301

Table 6 Day-of-the-week effect regression results.

Monday Tuesday Wednesday Thursday Saturday Sunday

ADSE

BSE

KSE

MSE

QSE

SASE

DSE

0.09 (0.1070) 0.103 (0.1060) 0.183 (0.108) ⁎

0.043 (0.019) ⁎⁎ 0.041 (0.019) ⁎⁎

0.052 (0.1550) 0.175 (0.1550) 0.363 (0.156) ⁎⁎

0.04 (0.0370) 0.158 (0.037) ⁎⁎⁎

0.07 (0.0620) 0.153 (0.062) ⁎⁎

0.14 (0.037) ⁎⁎⁎ 0.05 (0.0500)

0.187 (0.063) ⁎⁎⁎ 0.167 (0.063) ⁎⁎⁎

0.071 (0.0640) 0.096 (0.0640) 0.175 (0.064) ⁎⁎⁎

0.134 (0.1760) 0.142 (0.1750) 0.377 (0.175) ⁎⁎

−0.03 (0.0650) 0.021 (0.0650)

0.34 (0.183) ⁎ −0.205 (0.2600) 0.199 (0.2490)

0.149 (0.1080) −0.052 (0.1970) −0.101 (0.1100)

0.085 (0.019) ⁎⁎⁎ 0.016 (0.0190)

−0.011 (0.0200)

2.089 (0.409) ⁎⁎⁎

0.092 (0.1580)

0.002 (0.0370)

0.012 (0.0630)

We estimate OLS with robust p-values; applying the Huber–White sandwich estimator to correct for heteroscedasticity. Standard errors in parentheses. ⁎ p b 0.1. ⁎⁎ p b 0.05. ⁎⁎⁎ p b 0.01.

end-of-the-week calendar anomaly in returns for Western and GCC markets. While the literature for Western countries documents positive and significant returns on the last trading day of the week or Friday, we find similar results for GCC markets which observe an end of the week that predominantly falls on Wednesdays.22 Given that Wednesday is the last trading day for the leading exchange in the region, and in light of the highly significant and positive Wednesday returns, we conclude that an endof-the-week effect is prevalent in GCC markets.23 Fig. 2 displays the bar charts for weekday returns for all GCC stock markets. The graphs show that there exists a common trend across all GCC equity markets related to the Wednesday effect. Specifically, daily returns exhibit a rising trend between Monday and Wednesday for all markets except Oman, and this pattern is more pronounced for the leading Tadawul stock market. We conduct further analyses to confirm the presence of a Wednesday “effect”. We calculate normalized or risk-adjusted daily average returns and find that Wednesday normalized returns are highest compared to the returns of other days of the week for each market index, in line with our previous OLS findings.24 One interesting byproduct of these findings is that the common explanations usually provided for theday-of-week effect reported for developed markets are not applicable for the GCC region. For example, the “short selling hypothesis” is not valid since short selling activities are not permitted in these countries.25 Nor can the Settlement Procedure Hypothesis be used to explain the presence of such a calendar anomaly due to the very short settlement periods prevalent. Rather, it is likely that the significant and positive returns observed on the last trading day are caused by the optimism and merriness state of mind of GCC investors right before the weekend, rendering them more willing to buy and less willing to sell on Wednesdays (Gondhalekar and Mehdian, 2003). This “investors' mood hypothesis” connects the day-of-the-week effect to the investors' mental disposition to buy and/or sell in a particular day of the week and month of the year.

22 In contrast to the results for US markets, we do not find any significant negative return for the first trading day of the week, whether it is Saturday or Sunday. 23 One could arguably identify the last trading day for the GCC region as Thursday since all markets except the Saudi Tadawul are open on that day. However, as discussed earlier, the Saudi market is by far the largest and most active stock market in the region. Also, the series of market corrections of 2006 have established the leading role of the Saudi stock market in the region, since contagion effects worked their way from this market and onto other GCC exchanges afterwards. 24 The results can be requested from the authors. 25 According to the “short selling hypothesis”, short sellers attempt to close their short positions before weekends, causing excess demand for equity hence generating positive returns on Fridays (Chen and Singal, 2003). It follows that short sellers' attempt to create short positions after the weekend (particularly on Monday) causes an excess supply of securities and results in negative returns on Mondays.

302

R. Turk Ariss et al. / Emerging Markets Review 12 (2011) 293–307

As discussed previously, the month of Ramadan bears a special significance to GCC investors, and it is possible that investor behavior is modified during this month. To assess whether the month of Ramadan has a significant bearing on the reported Wednesday effect in GCC countries, we partition our data into two sets, Ramadan days and non-Ramadan days, and re-estimate Eq. (2) for each sub-sample separately. We present the results in Table 7 Panel A for non-Ramadan days and in Panel B for Ramadan days. The figures in Panel A generally support our earlier finding reported in Table 6 regarding the presence of a

AbuDhabi Stock Index Average Returns 0.2 0.15 0.1 0.05 0 -0.05

Sunday

Monday

Tuesday

Wednesday

Thursday

Saturday

-0.1 -0.15

Bahrain Composite Stock Index Average Returns 0.4 0.35 0.3 0.25 0.2 0.15 0.1 0.05 0 -0.05

Sunday

Monday

Tuesday

Wednesday

Thursday

Saturday

Kuwait Stock Index Average Returns 2.25 2 1.75 1.5 1.25 1 0.75 0.5 0.25 0 -0.25

Sunday

Monday

Tuesday

Wednesday

Thursday

Fig. 2. Daily average returns for the days of the week.

Saturday

R. Turk Ariss et al. / Emerging Markets Review 12 (2011) 293–307

303

Muscat Stock Index Average Returns 0.2 0.15 0.1 0.05 0 Sunday

Monday

Tuesday

Wednesday

Thursday

Saturday

-0.05 -0.1

Qatar Stock Index Average Returns 0.3 0.2 0.1 0 -0.1

Sunday

Monday

Tuesday

Wednesday

Thursday

Saturday

-0.2 -0.3 -0.4 -0.5 -0.6 -0.7 -0.8

Saudi SE Stock Index Average Returns 0.2

0.15

0.1

0.05

0 Sunday

Monday

Tuesday

Wednesday

Thursday

Saturday

-0.05 Fig. 2 (continued).

statistically significantly positive Wednesday effect for GCC countries during non-Ramadan days, except for Abu Dhabi where the Wednesday returns are positive but not statistically significant. In testing for the day-of-the-week effect during the month of Ramadan in Panel B, we again detect a statistically significant positive Wednesday effect, but only in four out of the seven examined markets. To sum, these results reveal the presence of a Friday-type effect taking place on Wednesdays in the GCC region, and this effect is more

304

R. Turk Ariss et al. / Emerging Markets Review 12 (2011) 293–307

UAE Stock Index Average Returns 0.5 0.4 0.3 0.2 0.1 0 Sunday

-0.1

Monday

Tuesday

Wednesday

Thursday

Saturday

-0.2 -0.3 Fig. 2 (continued).

pronounced during non-Ramadan days of the year as compared to the less prevalent Wednesday effect during the month of Ramadan. Finally, we investigate the presence of a month-of-the year effect following Eq. (3) and display the estimated coefficients for each stock market index in Table 8. These results for a possible month-of-theTable 7 Day-of-the week effect. ADSE Panel A: non-Ramadan days Monday 0.064 (0.1160) Tuesday 0.101 (0.1160) Wednesday 0.171 (0.1170) Thursday 0.139 (0.1170) Saturday −0.044 (0.2140) Sunday −0.107 (0.1200) Panel B: Ramadan days Monday 0.343 (0.178) ⁎ Tuesday 0.127 (0.1780) Wednesday 0.305 (0.184) ⁎ Thursday Saturday Sunday

0.266 (0.1900) −0.128 (0.3280) −0.05 (0.1780)

BSE

KSE

MSE

QSE

SASE

DSE

0.033 (0.0200) 0.055 (0.020) ⁎⁎⁎

0.037 (0.1660) 0.162 (0.1670) 0.384 (0.167) ⁎⁎

0.039 (0.0390) 0.152 (0.039) ⁎⁎⁎

0.078 (0.0670) 0.156 (0.066) ⁎⁎

0.123 (0.039) ⁎⁎⁎ 0.033 (0.0530)

0.166 (0.067) ⁎⁎ 0.141 (0.067) ⁎⁎

0.065 (0.0690) 0.1 (0.0690) 0.179 (0.069) ⁎⁎⁎

0.126 (0.1620) 0.167 (0.1610) 0.31 (0.170) ⁎ 0.214 (0.1610) 0.103 (0.2430) 0.211 (0.2330)

0.097 (0.020) ⁎⁎⁎ 0.012 (0.0200)

2.434 (0.457) ⁎⁎⁎

−0.013 (0.0200)

0.085 (0.1700)

0.001 (0.0390)

0.016 (0.0670)

−0.035 (0.0700) 0.031 (0.0700)

0.16 (0.069) ⁎⁎ −0.122 (0.069) ⁎

0.247 (0.2130) 0.324 (0.2000) 0.098 (0.2050) −0.018 (0.3960)

0.056 (0.0990) 0.23 (0.099) ⁎⁎

−0.029 (0.1400) 0.126 (0.1400) 0.444 (0.142) ⁎⁎⁎

0.134 (0.1230) 0.044 (0.1220) 0.128 (0.1220)

0.226 (1.1500) −0.145 (1.1500) 2.402 (1.177) ⁎⁎

0.032 (0.1250) −0.104 (0.1240)

0.682 (1.1770) −2.932 (1.496) ⁎ 0.095 (1.4420)

−0.052 (0.0700) 0.057 (0.0710)

0.013 (0.0690)

0.175 (0.2160)

0.35 (0.103) ⁎⁎⁎ 0.253 (0.137) ⁎

0.013 (0.0990)

0.491 (0.142) ⁎⁎⁎

−0.031 (0.1380)

We estimate OLS with robust p-values; applying the Huber–White sandwich estimator to correct for heteroscedasticity. Standard errors in parentheses. ⁎ p b 0.1. ⁎⁎ p b 0.05. ⁎⁎⁎ p b 0.01.

R. Turk Ariss et al. / Emerging Markets Review 12 (2011) 293–307

305

year effect are generally murky. Four market indices exhibit negative returns in the month of January, though only one is statically significant, while the remaining three exchanges exhibit a positive effect during this month, and only one of them is significant. Here again, we cannot use the tax-loss selling hypothesis (Reinganum, 1983; Ritter, 1988) to explain our finding because taxes are practically nonexistent in the GCC region. However, the results suggest the presence of another month-of-the-year or December effect. Figures of Table 8 point to the presence of a statistically significant and positive December effect in all GCC countries. Indeed, December returns are significantly higher compared to the returns for other months of the year for five GCC market indices (Kuwait, Muscat, Qatar, Saudi Arabia, and Dubai). Therefore, instead of the positive January effect documented in Western countries, GCC markets exhibit a December effect. We also match the Gregorian trading days into moving Hijri months, and re-estimate Eq. (3) after matching the dummy variables with the lunar calendar months. The results (not reported) again provide mixed evidence on the presence of a month-of-the-year effect following the Hijri calendar. We do not find evidence to suggest that that returns are either significantly higher during the first month of the lunar calendar year (the month of Muharram) or lower during the corresponding last month (the month of Zul Hijjah). In addition, Ramadan returns are not significantly different from returns during other months of the Hijri year. This outcome is further corroborated when comparing the average returns for Ramadan to other months. Table 9 presents the results of t and F tests of differences in means (Panel A) and variance (Panel B) respectively across Ramadan and other Hijri months. The figures in this table imply that Ramadan mean returns are effectively not different compared to the average returns for other months. However, there is a significant decline in the volatility of returns between Ramadan and other non-Ramadan months, in line with the findings of Al-Saad (2004).

Table 8 Month-of-the-year effect.

January February March April May June July August September October November December

ADSE

BSE

KSE

MSE

QSE

SASE

UAESE

−0.065 (0.1620) 0.09 (0.1580) 0.329 (0.150) ⁎⁎ 0.126 (0.1540) 0.065 (0.1510) −0.057 (0.1580) −0.404 (0.174) ⁎⁎ 0.277 (0.1780) 0.073 (0.1850) 0.206 (0.1640) −0.015 (0.1590) 0.255 (0.1620)

−0.063 (0.030) ⁎⁎ −0.078 (0.031) ⁎⁎

0.156 (0.2520) 0.108 (0.2670) 0.347 (0.2420) 0.294 (0.2360) 0.202 (0.2280) 0.316 (0.2340) 0.099 (0.2380) 0.068 (0.2380) −0.024 (0.2420) 0.052 (0.2400) −0.052 (0.2460) 0.771 (0.244) ⁎⁎⁎

0.132 0.056) ⁎⁎ 0.008 (0.0590) 0.087 (0.0570) 0.069 (0.0560) 0.087 (0.0550) 0.102 (0.056) ⁎

0.013 (0.0980) 0.093 (0.1000) 0.165 (0.096) ⁎ 0.176 (0.097) ⁎ 0.113 (0.0910) 0.099 (0.0930) 0.19 (0.097) ⁎ 0.224 (0.098) ⁎⁎

−0.048 (0.0960) −0.062 (0.0970) 0.017 (0.0910) 0.001 (0.0940) −0.017 (0.0920) 0.095 (0.0910) 0.042 (0.0900) 0.178 (0.090) ⁎⁎

−0.042 (0.0970) 0.068 (0.0950) −0.071 (0.0980) 0.346 (0.097) ⁎⁎⁎

0.016 (0.0920) −0.018 (0.0910) 0.01 (0.0940) 0.406 (0.092) ⁎⁎⁎

−0.114 (0.2720) 0.016 (0.2860) 0.015 (0.2780) 0.423 (0.2980) 0.206 (0.2490) 0.163 (0.2520) −0.002 (0.2780) 0.359 (0.2690) 0.317 (0.2720) 0.174 (0.2630) 0.094 (0.2690) 0.596 (0.278) ⁎⁎

−0.034 (0.0300) 0.095 (0.030) ⁎⁎⁎ 0.055 (0.029) ⁎ 0.035 (0.0300) 0.079 (0.030) ⁎⁎⁎ 0.093 (0.029) ⁎⁎⁎ 0.064 (0.030) ⁎⁎ 0.091 (0.030) ⁎⁎⁎ 0.016 (0.0300) 0.071 (0.031) ⁎⁎

0.053 (0.0570) 0.035 (0.0560) 0.092 (0.0570) 0.02 (0.0570) −0.023 (0.0600) 0.143 (0.059) ⁎⁎

We estimate OLS with robust p-values; applying the Huber–White sandwich estimator to correct for heteroscedasticity. Standard errors in parentheses. ⁎ p b 0.1. ⁎⁎ p b 0.05. ⁎⁎⁎ p b 0.01.

306

R. Turk Ariss et al. / Emerging Markets Review 12 (2011) 293–307

Table 9 Ramadan vs. non-Ramadan returns. ADSE

BSE

KSE

MSE

QSE

SASE

DSE

0.1474 0.1852 0.1168 0.8828 0.4414

0.6236 0.1532 0.0038 0.533 0.2665

0.2103 0.18 0.0719 0.8334 0.4167

0.0509 0.0843 0.0665 0.9594 0.4797

0.3239 0.3239 0.0947 0.746 0.373

Panel B: Ramadan vs. non-Ramadan standard deviation of returns F statistic 0.1868 0.0435 0.0325 Ramadan days 0.6907 0.5791 1.2577 Non-Ramadan days 1.5983 2.775 6.98 Pr(ratio! = 1) 0.0000 0.0000 0.0000

0.1516 0.7815 4.1774 0.0000

0.0134 0.7747 6.7001 0.0000

0.0197 0.8976 6.3976 0.0000

0.4772 5.3006 7.6734 0.0000

Panel A: Ramadan vs. non-Ramadan average return t statistic 0.8908 −0.0381 Ramadan days 0.2085 0.0078 Non-Ramadan days 0.0544 0.0132 Pr(difference N 0) 0.3733 0.9696 Pr(difference! = 0) 0.1866 0.5152

5. Summary and conclusions The GCC region is receiving increased attention from international investors, not only due to the excess of liquidity from petrodollars and ensuing economic growth, but also because of greater resilience that these countries have so far shown to the global financial crisis compared to other economies. In this paper, we argue that GCC stock markets have certain specificities that render previously reported calendar anomalies for other markets not applicable. The scant earlier work investigates individual GCC countries separately, reports declining volatility during the month of Ramadan, and does not find calendar anomalies similar to those documented for other stock markets. We find that returns are positive and significant on the last trading day of the week, in line with the literature for Western markets. This calendar anomaly, however, does not occur on Fridays, but rather on Wednesdays, and it is more pronounced during non-Ramadan days. We attribute this outcome to investors' buying mood right before the weekend begins in an environment where short selling is not allowed and trading in derivatives is not available. The investors' mood hypothesis may also explain the significant positive December effect that we find across GCC stock markets. While we do not find any significant difference in returns across the Hijri lunar calendar months, we note that the volatility of returns is significantly reduced during the month of Ramadan. This behavior coincides with the fact that economic activity generally slows down during Ramadan, and that trading in securities and speculation are also considerably reduced. Overall, our results provide evidence to suggest that, similar to international markets, calendar anomalies prevail in GCC stock markets. This outcome may be due to a number of factors such as relatively inactive markets, low volume of trading and liquidity, and perhaps the presence of few sophisticated investors so that possible intra week arbitrage opportunities are totally exploited. It could be that active investment strategies might generate abnormal returns in these markets. However, this opportunity may be restricted to GCC nationals only, since there are still some resilient regulatory restrictions on stock ownership by non-GCC investors in those markets.

References Al Janabi, M., Hatemi-J, A., Irandoust, M., 2010. An empirical investigation of the informational efficiency of the GCC equity markets: Evidence from bootstrap simulation. International Review of Financial Analysis 19, 47–54. Al-Barrak, A.M., 2009. Day-of-the-week effect in some of Gulf Cooperation Council (GCC) stock markets. Scientific Journal of King Faisal University 10 (2), 255–265. Al-Khazali, Osamah M., 2008. The impact of thin trading on the day-of-the-week effect: evidence from the United Arab Emirates. Review of Accounting and Finance 7 (3), 270–284. Al-Khazali, O., Darrat, A., Saad, M., 2006. Intra-regional integration of the GCC stock markets: the role of market liberalization. Applied Financial Economics 16 (17), 1265–1272. Al-Saad, K., 2004. Asymmetry in the price–volume relationship: evidence from the kuwait stock market. Journal of Accounting and Finance 3, 53–56. Al-Saad, K., Moosa, I., 2005. Seasonality in stock returns: evidence from an emerging market. Applied Financial Economics 15, 63–71.

R. Turk Ariss et al. / Emerging Markets Review 12 (2011) 293–307

307

Arab Monetory Fund, www.amf.org.ae. Assaf, A., 2003. Transmission of stock price movements: the case of GCC stock markets. Review of Middle East Economics and Finance 1, 171–189. Bachman, D., Choi, J., Jeon, B., Kopecky, K., 1996. Common factors in international stock prices: evidence from a cointegration study. International Review of Financial Analysis 5, 39–53. Bildik, R., 2001. Intra-day seasonalities on stock returns: evidence from the Turkish Stock Market. Emerging Markets Review 2, 387–417. Billmeier, A., Massa, I., 2009. What drives stock market development in emerging markets—institutions, remittances, or natural resources? Emerging Markets Review 10, 23–35. Bley, J., 2011. Are GCC Stock Markets Predictable? Emerging Markets Review 12 (3), 217–237 Available online 25 March 2011. Bley, J., Heng Chen, K., 2006. Gulf Cooperation Council (GCC) stock markets: the dawn of a new era. Global Finance Journal 17, 75–91. Bley, J., Saad, M., 2010. International differences in seasonality. International Review of Financial Analysis 19, 306–312. Bolbol, A., Omran, M., 2005. Investment and the stock market: evidence from Arab firm-level panel data. Emerging Markets Review 6, 85–106. Chen, H., Singal, V., 2003. Role of speculative short sales in price formation: the case of the weekend effect. The Journal of Finance 153 (2), 685–705. Gibbons, M., Hess, P., 1981. Day of the week effects and asset returns. Journal of Business 54 (4), 579–596. Gondhalekar, V., Mehdian, S., 2003. The blue-monday hypothesis: evidence based on NASDAQ stocks, 1971–2001. Quarterly Journal of Business and Economics 42, 73–89. Gulf Capital Group, 2006. Gulf capital report. Available online at http://www.Gulfcapital.com. Hammoudeh, S., Aleisa, E., 2004. Dynamic relationships among GCC stock markets and NYMEX oil futures. Contemporary Economic Policy 22, 250–269. Hammoudeh, S., Choi, K., 2005. Behavior of GCC stock markets and impacts of US oil and financial markets. Research in International Business and Finance 20, 22–44. Hammoudeh, S., Choi, K., 2007. Characteristics of permanent and transitory returns in oil-sensitive emerging stock markets: the case of GCC countries. International Financial Markets, Institutions and Money 17, 231–245. International Monetary Fund, Middle East Economic Outlook Regional, 2007. Lakonishok, J., Levi, M., 1982. Weekend effects on stock returns: a note. Journal of Finance 37 (3), 883–889. OECD (Organization for Economic Cooperation and Development), 2005. MENA-OECD Investment Programme. OECD Publishing. Peterson, D.R., 1990. Stock return seasonalities and earnings information. Journal of Financial and Quantitative Analysis 25, 187–201. Reinganum, M., 1983. The anomalous stock behavior of small firms in January: empirical tests for year-end tax effects. Journal of Financial Economics 12, 89–104. Ritter, J., 1988. The buying and selling behavior of individual investors at the turn of the year. Journal of Finance 43, 701–717. Seyyed, F., Abraham, A., Al-Hajji, M., 2005. Seasonality in stock returns and volatility: the Ramadan effect. Research in International Business and Finance 19, 374–383. Solnik, B., Bousquet, L., 1990. Day of the week effect on the paris bourse: a note. Journal of Banking and Finance 14, 461–468. World Wealth Report, 2006. Merrill Lyuch and Capgemini. Available at: www.capgemini.com/worldwealthreport.