The Ramadan effect: Illusion or reality?

Sonjaya, Azwar Ramadhana; Wahyudi, Imam Article The Ramadan effect: Illusion or reality? Arab Economic and Business Journal Provided in Cooperation with: Holy Spirit University of Kaslik Suggested Citation: Sonjaya, Azwar Ramadhana; Wahyudi, Imam (2016) : The Ramadan effect: Illusion or reality?, Arab Economic and Business Journal, ISSN 2214-4625, Elsevier, Amsterdam, Vol. 11, Iss. 1, pp. 55-71, https://doi.org/10.1016/j.aebj.2016.03.001 This Version is available at: https://hdl.handle.net/10419/187531 Standard-Nutzungsbedingungen: Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden. Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen. Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. Terms of use: Documents in EconStor may be saved and copied for your personal and scholarly purposes. You are not to copy documents for public or commercial purposes, to exhibit the documents publicly, to make them publicly available on the internet, or to distribute or otherwise use the documents in public. If the documents have been made available under an Open Content Licence (especially Creative Commons Licences), you may exercise further usage rights as specified in the indicated licence. https://creativecommons.org/licenses/by-nc-nd/4.0/ The Ramadan effect: Illusion or reality? Azwar Ramadhana Sonjaya, Imam Wahyudi * Department of Management, Faculty of Economics and Business, Universitas Indonesia, Jakarta, Indonesia abstract Empirical tests of the efficient market hypothesis (EMH) have been repeated by many researchers with varying results, with several supporting it and others finding no clear evidence for it. One of the results that weakens the EMH is the study of such anomalies as the Ramadan effect. Anomaly studies are also denied by recent studies that demonstrated proof of the weakening and even disappearance of anomalous effects. This research tests the persistence of the Ramadan effect in the stock returns in 10 Muslim-majority countries. We have found that the Ramadan effect is present, but it is not persistent. This finding is consistent with the finding from the test of efficient market form, which indicated that the markets of all 10 Muslim-majority countries are not efficient. When economic crisis is considered as an influencing factor, the Ramadan effect is still not persistently present. ã2016 The Authors. Production and hosting by Elsevier B.V. on behalf of Holy Spirit University of Kaslik. This is an open access article under the CC BY-NC-ND license (http:// creativecommons.org/licenses/by-nc-nd/4.0/). article info Article history: Received 24 November 2015 Received in revised form 9 January 2016 Accepted 1 March 2016 Keywords: Efficient market Ramadan effect Anomaly Abnormal return Crisis JEL classification: D84 E44 G02 G14 1. Introduction There are two types of calendar anomalies: religious-related anomalies, such as Christmas and Good Friday effects (Cadsby & Ratner, 1992), Jewish High Holy Days effects (i.e., Rosh HaShanah and Yom Kippur) (Frieder & Subrahmanyam, 2004), or the Easter week holiday effect (Pantzalis & Ucar, 2014), and Ramadan effect (Bialkowski, Etebari, & Wisniewski, 2012); and non-religiousrelated, such as the January effect, Wednesday effect, and weekend effect (Schwert, 2003). For the Ramadan effect in particular, in addition to the significant impact of the moving calendar on abnormal returns (Alper & Arouba, 2001), a combination of factors not found in other religious-calendar anomaliesalso impact abnormal returns during Ramadan. Thesefactors include investor health due to Ramadan fasting (Rosen & Wu, 2004; Saleh, Elsharouni, Cherian, & Mourou, 2005), social empathy (positive social mood) with the poor due to the hunger experienced while fasting (Bialkowski et al., 2012), feeling happy and peaceful (Lakonishok & Smidt, 1988), investors’positive moods (Cadsby & Ratner, 1992), and the encouragement to do good deeds and prevent evil deeds Peer review under responsibility of Holy Spirit University of Kaslik. *Corresponding author. E-mail address: [email protected] (I. Wahyudi). http://dx.doi.org/10.1016/j.aebj.2016.03.001 2214-4625/ã2016 The Authors. Production and hosting by Elsevier B.V. on behalf of Holy Spirit University of Kaslik.This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). arab economic and business journal 11 (2016) 55–71 Available online at www.sciencedirect.com ScienceDirect journal homepage: www.elsevier.com/locate/aebj (Bialkowski et al., 2012). The positive mood that is the hallmark of the Ramadan effect is different from the Easter week holiday effect, which is marked by investor distraction, causing delayed responses to firm news (Pantzalis & Ucar, 2014), or the Jewish High Holidays effects, when Jewish investors’sentiments cause significant decreases in dollar volume on both named holidays (stock returns on Rosh HaShanah are significantly positive versus significantly negative on Yom Kippur). Bialkowski et al. (2012) suspect that the emotion and mood factors of investors play a significant role in their judgment and decision making, especially that related to the buying and selling of stock, the preference for risk and return, and the response to uncertainty. These psychological factors and the reality that the investors require significant funds during Ramadan, especially toward the end of Ramadan and Idul Fitri, to meet their religious needs (zakat, infaq and shadaqah) and celebrate the Eid (buying clothes, banquet foods, etc.), prompt investors to behave rationally by buying stocks at the beginning of Ramadan and selling them at the end of Ramadan, or shortly after the Eid-ul-Fitr (Bialkowski et al., 2012; Al-Khazali, 2014). In addition to individuals’behavior, during Ramadan, Islam teaches that believers must empathize and share with the poor through giving and worship in order that the positive social mood will increase the social and spiritual orientation (Bialkowski et al., 2012). This positive personal mood will encourage investors to be happier and more optimistic (Beit-Hallahmi & Argyle, 1997; Gavriilidis, Kallinterakis, & Tsalavoutas, 2015). In addition to the investor psychology, Bialkowski et al. (2012) also states that the health factor of investors when they fast during Ramadan, affects their physical and mental health (Saleh et al., 2005), and then affects their positive valuation of stock prices in the market (Rosen & Wu, 2004). Specifically, Gavriilidis et al. (2015) cites certain impacts of religious factors on the economic and financial environment that affect the propensity to save, the decision to invest in stocks (Renneboog & Spaenjers, 2012), the risk-attitude of the investor (Kumar, Page, & Spalt, 2011), and economic growth (Barro & McCleary, 2003). As with other studies of EMH anomalies, there is evidence for and against the existence of the Ramadan effect. Bialkowski et al. (2012) usestheeventstudyonthecumulativeabnormal returns and finds thatstockreturnsduringRamadanarehigherthanin the other 11 months in 11 Muslim countries (from a total sample of 14 Muslim countries). Their study argued that the Ramadan effect is purely caused by investor psychology, executing an investment strategy of buying stocks before Ramadan arrives and then selling them at the end ofRamadan, or shortly after Eid-ul-Fitr(see also Al-Khazali, 2014).Al-Mudhaf(2012) finds empirical evidence ofthe Ramadan effect in only four Muslim countries from a sample of 12, while Al-Hajieh, Redhead, and Rodgers (2011) uses run tests and finds the Ramadan effect in 6 out of 8 markets. In the Karachi stock market, Mustafa (2011) finds the Ramadan effect, while in other empirical studies using the GARCH model, the Ramadan effect is found not to significantly affect mean returns in some countries, including Pakistan (Husain, 1998), Saudi Arabia (Seyyed, Abraham, & Al-Hajji, 2005) and Indonesia (Rainly, 2006), although they document a decrease in return volatility during Ramadan. Al-Khazali (2014) uses the non-parametric stochastic dominant method and finds evidence of the Ramadan effect in the 15 Muslim countries, although the results were not sufficient to conclude that the returns during Ramadan outperform the return beyond Ramadan. Unfortunately, there has not yet been a satisfactory explanation of the phenomenon of the Ramadan effect (Al-Khazali, 2014), especially related to the violation of the trading hypothesis during Ramadan. As a result, the discussions tend to focus on the accuracy of the method used. Al-Khazali (2014) criticized the use of the mean-variance method, as performed by Husain (1998), Seyyed et al. (2005),Rainly (2006), and Bialkowski et al. (2012). According to Al-Khazali (2014), the use of parametric statistic (mean and variance) and GARCH often relies on the normality assumption, which is difficult to fulfill in emerging markets, including Muslim countries. Second, the use of mean and variance ignores the effect of positive and negative skewness that represents the risk preference of the investor. Third, the mean-variance method requires the use of the quadratic-utility function. Al-Khazali (2014) also admits, however, that any group of methods, such as the mean-variance method and stochastic dominant, can be complementary. Regardless of the debates over the most appropriate method for examining the Ramadan effect, in this research, we use index data from 10 Muslim-majority countries to first test whether their markets are efficient. Logically, when the market is efficient, no investors can persistently exploit abnormal returns (Fama, 1965) because the price fully reflects the information equilibrium in the market. When the market is efficient, however, and some investors are still able to exploit market information to gain abnormal returns, there is assumed to be an anomaly in the market (Stulz & Williamson, 2003). Further, when the market is not efficient and the investor can exploit the information, it is argued that behavioral finance (Thaler, 1993), which is typically driven by psychological factors, such as herding, emotions, mood, and the investors’religious beliefs (Bialkowski et al., 2012), is to blame. Fama (1970) and Jensen (1978) state that if market is not efficient (even in a weak form), the presence of abnormal returns is to be expected as the price does not immediately reflect the available information in the market. We then examine the existence of the Ramadan effect in 10 Muslim-majority countries and test the persistenceof the Ramadan effectby dividing the observation period into 5 sub-periods.Identifyingtheexistenceof Ramadan effect, we conduct a comparative analysis by computing the annualized returns between Ramadan and the 11 other months in the Muslim lunar calendar in order to obtain the abnormal returns during Ramadan. Then, we divide the observation period into several sub-periods to examine whether the Ramadan effect persistently exists in each sub-period, conduct t-tests on the significant cumulative abnormal return, and observe the pattern of the Ramadan effect. In this stage, using the regression analysis, we also examine the coincidence of the Ramadan effect with other calendar effects, such as the January effect, the weekend (Friday) effect, and the Christmas effect. According to Bialkowski et al. (2012), during Ramadan, the positive societal mood will affect the positive effect on the stock price, while there is simultaneouslyno coincidence between the Ramadaneffect and the Gregorian calendareffects. In this stage, wealso examine the impact of local and global crises on the Ramadan effect. Al-Khazali (2014) finds that the magnitude of the Ramadan 56 arab economic and business journal 11 (2016) 55–71 effect decreases during global crises. Previously, Bialkowski et al. (2012) have cited the Asian crisis as the reason that the Ramadan effect does not appear in the Indonesian stock market. Like Frieder and Subrahmanyam (2004), our results show that Ramadan is the religious event that can affect investors’moods and investment decisions (Bialkowski et al., 2012) through the combination of two herding factors that prompts less risk-averse action and enhances social interaction among investors (Gavriilidis et al., 2015). We find that the Ramadan effect is persistently present in three countries (Kuwait, Oman and Tunisia), and the magnitude is always larger than in other months. Five other countries (Indonesia, Malaysia, Jordan, Morocco and Qatar) also experience the Ramadan effect, but it is not persistent. Finally, two countries (Bahrain and Saudi Arabia) have never experienced the Ramadan effect, showing negative annualized returns during Ramadan in all periods. As with the other EMH anomalies, we find that the Ramadan effect tends to disappear and is not persistent, except in the three countries cited above (Kuwait, Oman and Tunisia). The regression analysis shows that the Ramadan effect does not significantly affect annualized stock returns. From the perspective of Islamicfinance, our findings, similar to those from earlier studies (e.g., Al-Khazali, 2014; Bialkowski et al., 2012; Gavriilidis et al., 2015), provide practical implications for investors to exploit the abnormal stock returns in some Muslim countries (Kuwait, Oman and Tunisia) during Ramadan. Moreover, investors must be cautious regarding the destabilizing potential of herding behavior that promotes systemic risk in the market. The remainder of the paper is organized follows: Section 2provides a literature review. Section 3presents the data and methodology.Section4contains the results and analyses. Section 5provides the study’sconclusionsandmanagerialimplications, while Section 6gives suggestions for further research. 2. Literature review 2.1. Ramadan and its effect on investor behavior Ramadan is the 9th month of the Hijriyah, the Muslim lunar calendar. At the present, many Muslim-majority countries use two calendars (Hijriyah and the Gregorian calendar). The Gregorian calendar is used for business and administrative purposes, while the Hijriyah calendar is used for religious observance, including the month of fasting during Ramadan and Eid-ul-Fitr (the first day of the Syawal month). The Hijriyah calendar moves every year and is approximately 10–11 days faster than the Gregorian calendar. This drift of the Hijriyah calendar compared to the Gregorian calendar is used as an argument to show that the presence of abnormal returns in Ramadan is truly a calendar effect, unlike the January effect, which is a moving calendar effect. A moving calendar event can significantly affect economic and financial variables, including abnormal returns (Alper & Arouba, 2001). During Ramadan, Muslims fast, bearing hunger and thirst as well as other desires from just before dawn and until sunset. Many Muslims honor Ramadan by reducing the number of hours they work per day to maintain performance and aid observance. Fasting during Ramadan has been clinically proven to improve health through a natural detoxification process (Fuhrman, 1998) and the reduction of weight, cholesterol (Saleh et al., 2005), blood pressure, and anxiety (Daradkeh, 1992). The collective feeling of religious observance as well as more intense social interaction compared to other days also contributes to this improvement in health (Bialkowski et al., 2012). Other than fasting, Ramadan is also a time that is highly anticipated by Muslims, both in the context of the vertical relationship with God as well as the horizontal relationship with their fellow people. During Ramadan, Muslims are more social and religious as they expect their good deeds to be multiplied during the holiday, as promised in the Quran (Quran 97:3). As during Ramadan, the celebration of religious holidays such as Islam’s Eid-ul-Fitr increases the happiness of religious believers (Muslims), which has been proven to significantly affect the capital market in countries with large numbers of the religious believers (Frieder & Subrahmanyam,2004;Lakonishok & Smidt, 1988). Thissocialand religious orientation generates positivemoodsaffectinginvestor psychology individually and collectively, and it also affects an investor’s self-esteem in taking risks and investing (Lucey & Fig. 1 –The behavior of Muslim investors during Ramadan. arab economic and business journal 11 (2016) 55–71 57 Dowling, 2005; Rosen and Wu, 2004). Those positive moods encourage investors to take risks, thus increasing the proportion of risky assets in their investment portfolio, which will raise stock prices and create the possibility of abnormal returns (Fig. 1). 2.2. EMH anomaly: the calendar effects Previous studies on the anomalies of efficient markets have provided a collection of empirical proofs that there are certain deviations in the movement of price(s) from what can be expected if the efficient market hypothesis holds. A market is efficient when the price reflects all relevant information in the market, under the assumptions that the investors are rational and have homogenous expectations and that given this equilibrium between price and information, it is impossible for investors to persistently gain abnormal returns (Fama, 1970). In reality, however, some events are able to trigger under and over reactions, causing shifts in stock prices that enable abnormal returns to be earned, indicating market inefficiency and thus a hole in the EMH (Stulz & Williamson, 2003). The occurrence of various events that can trigger abnormal returns is a starting point of academic interest in the study of EMH anomalies, such as the January effect (Branch, 1977), neglected firms effect (Arbel & Strebel, 1983), exchange listing effect (Ritter, 1991), and size effect (Reinganum, 1992). Fundamentally, an anomaly is defined as the relative deviation from a model of normal return behavior (Schwert, 2003) where normal return behavior is based on different levels of market efficiency, as has been expressed by Fama (1970) through three market forms: the weak-form, semi-strong-form and strong-form efficiency. If the market is not efficient, even in the weakest form, then the presence of abnormal returns is still reasonable because the price that is formed does not reflect the available information (Jensen, 1978). Additionally, the term “anomaly”itself is sometimes misused and misapplied. Anomaly is interpreted as a failure of the EMH as a paradigm of modern finance theory in explaining price movement, the same dissatisfaction that fueled the development of behavioral finance. The original meaning of anomaly is irregularity, deviation from the common order or natural order, and as such it is possible for an anomaly to exist simply because there is an unexpected mismatch between what is expected to happen according to theory and what actually happens. It is a “puzzle,”if you will, and if that puzzle is solved, then what had seemed to be an anomaly before can be understood, with the result that the anomaly no longer exists (Kuhn, 1977). To help maintain neutrality, the use of the word “effect”is well recognized in academic literature to illustrate the “puzzle”of EMH theory (Frankfurter & McGoun, 2001). Interestingly, recent studies show that several famous and well-documented anomalies in the academic literature have actually weakened in effect and some has even disappeared over time. Anomalies, such as the weekend effect and the January effect, are found to be present only inconsistently in the entire sample period, and several other anomalies, such as the size effect and value effect, have been found to disappear since they were first documented (Schwert, 2003). 2.3. Ramadan effect: investor mood and herding behavior The effect of Ramadan is proved to affect the positive mood and trading behavior of investors, which is reflected in the significantly higher stock returns during Ramadan compared to non-Ramadan days (Al-Hajieh et al., 2011; Al-Khazali, 2014; Bialkowski et al., 2012). The positive mood resulting from fasting increases the investor’s health (Saleh et al., 2005) and the spirit of communal worship promotes the herding behavior through optimism and enhanced social interaction (positive social mood) (Bialkowski et al., 2012). Further, Gavriilidis et al. (2015) find that herding is significant within Ramadan in most (five out of seven) sample markets and the magnitude is greater during Ramadan. Based on social norm theory, individuals follow behavioral norms, beliefs and/or activity of the community members (Akerlof, 1980), here, religious social norms (Gavriilidis et al., 2015). The religious aspect of Ramadan plays a significant role in the risk-taking behavior of investors (Bialkowski et al., 2012; Hilary & Hui,2009). More than 1.6 billion Muslims across theworld celebrate this month (Al-Khazali, 2014)byfasting,refrainingfrom eating, drinking, smoking and sexual activity from dawn until sunset (Gavriilidis et al., 2015). Clinically, fasting causes the investor to become healthier (Knerr & Pearl, 2008; Saleh et al., 2005), decreases the level of anxiety, enhances optimism and social interaction, and cultivates a positive mood during Ramadan (Bialkowski et al., 2012). This psychological factor prompts the investor to herding behavior through a combination of two factors that encourage investors to be less risk-averse (Nofsinger, 2002; Wright & Bower, 1992). Enhanced social interaction during Ramadan also promotes common behavior among investors (Gavriilidis et al., 2015). Historically, higher returns during Ramadan induce investors to underestimate risk and increase their optimism. Enhanced social interaction among investors during Ramadan ensures this behavior is followed by other investors in the market (herding) due to the shared emotion and positive social mood. Recent studies have sought to examine the relationship between the mood-effect of Ramadan and investment behavior during Ramadan. Using the GARCH model, Husain (1998) examines the Ramadan effect in the Pakistani stock market and finds that there is a decrease in stock return volatility in Ramadan; however, the mean returns are not significantly different during Ramadan versus the rest of the year. Seyyed et al. (2005) uses stock market data from Saudi Arabia during the period from 1985 to 2000 and finds that there is no significant change in stock returns during and after Ramadan; however, they document a decrease in return volatility during Ramadan. Al-Hajieh et al. (2011) use a data from 8 Middle Eastern countries for the period from 1992 to 2007 and find significant positive abnormal returns during Ramadan in 6 of 8 countries. They attribute the findings to the positive investor mood during Ramadan and enhanced social interaction among investors. In the Karachi stock market, Mustafa (2011) also finds evidence of the Ramadan effect. Bialkowski et al. (2012) find evidence of the Ramadan effect in 11 of 14 Muslim countries in their 58 arab economic and business journal 11 (2016) 55–71 sample. They suspect that Ramadan promotes feelings of social empathy and social solidarity, hence enhancing the optimism of investors and affecting the investment decision. Al-Mudhaf (2012), however, finds the Ramadan effect in only four of 12 countries analyzed. Al-Khazali (2014), using 15 Muslim countries with various time periods as the sample, confirms the findings of Bialkowski et al. (2012) that stock returns during Ramadan are higher than those the rest of the year; however, the effect disappears during the global financialcrisis. Finally, Gavriilidis etal.(2015) find evidence of herdingbehavior duringn Ramadan in fiveof seven Muslim countries in the sample, and further find that herding behavior is more prevalent during Ramadan. 3. Data and methodology 3.1. Data This study is based on previous studies of the Ramadan effect, especially those by Bialkowski et al. (2012),Al-Khazali (2014) and Gavriilidis et al. (2015). Here, we perform a comprehensive analysis of 10 Muslim-majority countries, improving on previous studies on Ramadan effect examining a single country, e.g., Pakistan (Husain, 1998), Saudi Arabia (Seyyed et al., 2005), and Indonesia (Rainly, 2006). Some of the studies limited to one country conclude that the Ramadan effect on stock returns is insignificant, while Bialkowski et al. (2012),Al-Khazali (2014), and Gavriilidis et al. (2015) conclude that the Ramadan effect is significant based on a comparison of multiple countries. In several countries, however, such as Bahrain, Saudi Arabiam and Indonesia, Bialkowski et al. (2012) also find that the Ramadan effect is insignificant due to data irregularity (Bahrain and Saudi Arabia) or because it occurred at approximately the same time as a crisis (Indonesia). This research uses stock market index data taken from the Morgan Stanley Capital International (MSCI) Datastream. Sampled countries must have a Muslim population of more than 50% (based on the CIA World Factbook 2013). Ten countries are included in this research: Bahrain, Indonesia, Jordan, Malaysia, Morocco, Kuwait, Oman, Qatar, Saudi Arabia, and Tunisia. The research period covers 139 Ramadan events through the period from1989 to 2013. The sample period is then divided further intofive sub-periods to observe the persistence of the Ramadan effect: 1989–1993, 1994–1998, 1999–2003, 2004–2008, and 2009–2013. The determination of the beginning and ending of Ramadan in the Gregorian calendar is based on information from each country’s official website that states the official beginning and ending of Ramadan. Data taken from the MSCI Index is already USD-denominated, and as such there is no stock price bias due to inflation in each country. 3.2. Model specification Stock price data are processed into returns using the formula for simple returns, R t =(P t /P t 1 )1. To compare average returns, the returns are annualized, during Ramadan as well as in other months. The first step in this research is to test which efficient market form is taken by the capital market in the observed countries. The efficient market test is performed at the lowest level of EMH, the weak-form efficiency, which is tested using the autocorrelation test (Reilly & Brown, 2012). The results of the weak-form efficiency test, along with the abnormal returns test using the event study method and cumulative abnormal return (CAR), are used to conclude whether the Ramadan effect is purely psychological or if it can be explained by EMH. The event study method is applied to the Ramadan effect analysis to show the effects of the presence of abnormal return in the monitoredevent.Abnormalreturns (AR it )aredefinedastheactual return oftheindexintheeventwindow perioddeductedfrom the normal return, which is the expected return, without including the event window period in the estimation value (MacKinlay, 1997): ARit ¼Rit EðRitjXitÞ(1) where AR it is the abnormal return of index iin period t, and EðRitjXitÞis the expected value of normal return. This research also uses three models to estimate normal returns: the market, constant-mean and market-adjusted models (Brown & Warner, 1985; MacKinlay, 1997). In the market model, the expected normal return can be estimated with the equation: Rit ¼aiþbiRmt þit (2) where R it is the index return of the event window periods, R mt is the global market return (MSCI world index), ^ Rit is an estimation of normal returns using the market model with aiand bias model parameters, and it is the error factor of the model with the assumption of white noise ðEðitÞÞ ¼ 0;varðitÞ¼s3 i(MacKinlay, 1997). In the constant-mean model, normal returns come from expected value (mean) Rit in the non-event period (estimation window). The estimation of normal returns both using the market model and the constant-mean model spans 200 days before the event window. In the constant-mean model, the expectation of normal returns is estimated with the equation: Rit ¼EðRitÞþit;where :ðEðitÞÞ ¼ 0 and varðitÞ¼s3 i(3) arab economic and business journal 11 (2016) 55–71 59 In the market-adjusted model, the expectation of normal returns is the value of the market index (MSCI world index), which is: Rit ¼Rmt (4) The cumulative abnormal return (CAR) is found by adding the average abnormal return of each index iin period tto T (MacKinlay, 1997): AAR ¼X N t¼1 ARit (5) CARðn1;n2Þ¼X T t¼1 AAR (6) tðCARðn1;n2ÞÞ ¼ CARðn1;n2Þ sðARÞxffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi n2n1þ1 p(7) where CAR(n 1 ,n 2 ) is the cumulative abnormal return in the eventwindow from n 1 to n 2 ,s(AR) is the standard deviation of the mean abnormal return calculated through time series in event window n 1 to n 2 , while n 1 and n 2 are the beginning and end of the event window period, respectively. The statistical significance test for CAR uses the parametric t-test (Kothari & Warner, 2007). The statistical significance of the event study method only generates abnormal return results that are contemporaneously correlated (Bialkowski et al., 2012). It is thus necessary to perform portfolio return regressions that are formed from the indices of the sample countries to the Ramadan dummy variable and the MSCI world index as the reference market return: returnportfoliot¼atþb:Ramadantþg:returnworldtþt(8) Bialkowski et al. (2012) states that one of the reasons for the insignificance of the Ramadan effect in Indonesia is the economic crisis factor. To prove this statement, testing the significance of the Ramadan effect on stock returns will include the economic crisis factor as a dummy variable: returnportfoliot¼atþb:Ramadantþg:returnworldtþd:crisistþt(9) 4. Results and analysis 4.1. Efficient market testing The weak-form efficient market is tested using the autocorrelation test (Reilly & Brown, 2012) on the index return in each country and each period. In the full period from 1989 to 2013, Table 4.1 displays the efficient market test results showing that the market is not efficient in all countries, even when the weak form is being tested, except for Saudi Arabia. This is due to inadequate data in Saudi Arabia, contributing to the inaccuracy of the test result. Several countries in some sub-periods actually demonstrated that their market is already efficient in the weak form, even if it is not always consistent across all sub-periods. This finding indicates that even the weak form of the efficient market is not steady in those countries (Frankfurter & McGoun, 2001; Reilly & Brown, 2012). Table 4.1 –Efficient market test. Indonesia Jordan Malaysia Morocco Bahrain Kuwait Oman Qatar Saudi Tunisia 1989–2013 Not efficient Not efficient Not efficient Not efficient Not efficient Not efficient Not efficient Not efficient Efficient Not efficient 1989–1993 Not efficient Efficient Not efficient 1994–1998 Not efficient Efficient Not efficient Not efficient 1999–2003 Not efficient Efficient Not efficient Not efficient 2004–2008 Not efficient Not efficient Not efficient Not efficient Efficient Not efficient Not efficient Efficient Efficient Not efficient 2009–2013 Efficient Not efficient Not efficient Efficient Not efficient Efficient Not efficient Not efficient Not efficient Not efficient 60 arab economic and business journal 11 (2016) 55–71 The efficiency level of the capital market is closely related to a country’s economic condition. High growth of the capital market index’sreturnisone of theindicatorsthatacountry’s economy is experiencinghigh growth(Kannan&Henry, 2008). High growthis also one of the differentiating factors between developed markets and emerging markets (Kvint, 2009). Over the last 10 years, the growth rate of developed markets has remained under 3.00%,as in Japan (1.32%), the US (2.56%), and Great Britain (2.17%), while the growth of emerging market economies has exceeded 6.00%, as in Indonesia (6.11%) and Malaysia (6.15%) (World Bank, 2013). Theresultsof the efficient markettest(weak-form)show that noneoftheobservedcountries have achieved anefficient market, not even a weak-form one. Based on the classification of economic growth, all the observation countries can also be categorized into emerging markets and frontier markets, as displayed in Fig. 4.1. The results of this test are consistent with several previous studies that attempted to test the market efficiency level of emerging and frontier markets, as in the studies of Chang, Lima, and Tabak (2004),Worthington and Higgs (2006), and Segot and Lucey (2006). They found that no emerging and frontier markets are significant in the random walk test (variance ratio test), which indicates the markets in all these countries are not efficient, not even weak-form efficient. 4.2. Presence of Ramadan effect 4.2.1. Annualized stock return during Ramadan The Ramadan effect can be defined as the occurrence of abnormal returns enjoyed by investors due to psychological-religious effect that drives the movement of stock market prices (Bialkowski et al., 2012). The presence of the Ramadan effect can be observed through two methods: (1) comparison of annualized returns between Ramadan and other months, or (2) observing the presence of abnormal returns before, during, and after Ramadan. The comparison of annualized returns can be used to find out whether the Ramadhan effect can provide larger returns relative to returns in other months (Bialkowski et al., 2012). The identification of abnormal return can provide a guide for investors to track the movement pattern of the price and understand the possibilities for exploiting the abnormal returns offered by the Ramadan effect (Al-Ississ, 2010). This study found three country groups, each with a different type of relationship between the Ramadan annualized returns and the annualized returns from other months. The first group shows a positive Ramadan annualized return that is larger than the annualized returns of other months. This group is stated to have experienced the Ramadan effect, where there is a potential to gain higher average returns during Ramadan compared to the rest of the year. This group consists of six countries: Tunisia, Qatar, Jordan, Kuwait, Oman, and Morocco (Fig.4.2). The second group comprises countries with a positive Ramadan annualized return that is nonetheless smaller than the annualized returns of other months. For this reason, the single country in this group, Malaysia, does not experience the Ramadan Indonesia Malaysia Jordania Maroko Kuwait Oman Qatar Tunisia Bahrain Saudi -20% -15% -10% -5% 0% 5% 10% 15% 20% 25% 14%12%10%8%6%4%2%0% Return Index Growth Economic Growth Emerging Markets Frontier Markets Fig. 4.1 –The classification of emerging and frontier markets. Fig. 4.2 –Annualized return of indexes in ten Muslim countries. arab economic and business journal 11 (2016) 55–71 61 effect. The third group contains countries with negative Ramadan annualized return. This country group is stated not to experience the Ramadan effect, because on average, these countries actually experience negative returns during Ramadan. There are three countries in this group: Indonesia, Bahrain, and Saudi Arabia. Malaysia exhibits positive Ramadan annualized returns, but its returns behave differently than those of the other six countries whose Ramadan annualized returns are also positive. The other six countries display the normal, expected behavior of their Ramadan annualized returns in that it is positive, and the anualized returns for the rest of the year are smaller than the total annualized returns, signifying that the Ramadan effect does exist and affects the total annualized return of the aforementioned market. This behavior, however, is nonexistent in Malaysia. The lack of a Ramadan effect in Malaysia can be analyzed in light of the Southeast Asian economic crisis of 1997–1998, faced by both Indonesia and Malaysia (Hunter, Kaufman, & Krueger, 1999). The effect of the Southeast Asian crisis in Malaysia was not sizeable, as it was not accompanied by a political crisis, as occurred in Indonesia. Malaysia still managed to generate a positive Ramadan annualized return, while Indonesia experienced a negative Ramadan annualized return because of an economic and political crisis that took place at approximately the same time as Ramadan. The countries in the third group (Indonesia, Saudi Arabia and Bahrain) have a negative Ramadan annualized return, unlike the other seven countries. This supports the previous two studies of the Ramadan effect in Saudi Arabia by Seyyed et al. (2005) and in Indonesia by Rainly (2006), which found that the Ramadan effect does not significantly affect the stock returns in those two countries. The insignificance of the Ramadan effect in Bahrain is caused by fundamental factors, where the movement of the country’s stock market index continuously decreases (Fig. 4.3). The stock market index is downward sloping year after year; thus both in general and on a monthly basis, the value of the returns on Bahrain’s index tends to move toward zero. In the case of Saudi Arabia, the insignificance of the Ramadan effect is affected by the lack of data due to the cessation of the stock market index provided by the capital market in Saudi Arabia (Table 4.2). 4.2.2. Pattern of abnormal returns during Ramadan The fluctuations of stock prices during Ramadan are inseparable from the price fluctuations in the market. Using the theory of behavioral finance, Bialkowski et al. (2012) attempt to explain the fluctuation of stock prices during Ramadan based on psychological-religious factors affecting the investors, who drive the stock price to attractive levels based on their mood. This Fig. 4.3 –Movement of the stock market index. Source: Thomson Reuters Datastream Table 4.2 –The annualized return of indexes in 10 Muslim countries. No. Country Rest of the year annualized return Ramadan annualized return Total annualized return Events 1 Tunisia 1.69% 31.77% 3.98% 8 2 Qatar 0.24% 28.64% 2.03% 8 3 Jordan 1.25% 23.89% 0.71% 26 4 Kuwait 4.20% 20.17% 2.31% 8 5 Oman 5.11% 13.92% 3.61% 8 6 Morocco 5.67% 7.11% 5.76% 19 7 Malaysia 5.93% 4.01% 5.78% 26 8 Indonesia 3.48% 2.70% 2.86% 26 9 Saudi A. 9.81% 7.70% 9.64% 2 10 Bahrain 20.62% 26.74% 21.17% 8 Total observation (year) 139 62 arab economic and business journal 11 (2016) 55–71 December2008),asdefined by Calomiris, Love, and Peria(2012).Inthis study,crisis is thereforepresentinthe sub-periods of 1994 to 1998 and 2004 to 2008. The impact of crisis on the Ramadan effect is analyzed in two ways: (i) by constructing the same regression model on the previous model with the addition of a crisis dummy variable, and (ii) by constructing a regression model that excludes returns of crisis periods. The two regression models above provide consistent results such that Ramadan variable still has an insignificant effect on the dependent variable (return portfolio) in all sample periods, while other independent variables (R-world and crisis) significantly affect the return portfolio. The F-test and R-squared test of the two models are also significant, showing that the model is capable of explaining the relationship between the dependent variable with the independent variable. As such, the insignificance of the Ramadan effect is due to the variable itself having an insignificant effect on the return portfolio. 4.5. Robustness To check the robustness of the Ramadan effect, we conduct the estimation of ARMA and GARCH models and find that with any combinations of pand q, the results are same; i.e., the Ramadan effect does not significantly affect the return portfolio. Similarly, when we enter the other Gregorian calendar effects, the results are consistent: there is no Ramadan effect on the return portfolio. To check whether the Ramadan effect is an anomaly that occurs independently and is not influenced by other calendar effects, we then perform the regression analysis as shown inTable 4.11. The tested calendar effects includethe January effect, weekend effect, and Christmas effect. Simultaneously, we regress these effects as the dummy variables. In the full period, we find that only the weekend effect is significantly not independent from the Ramadan effect. In all subperiods and over the full period, however, we find that only the Christmas effect is persistently independent from the Ramadan effect, while the January and weekend effects are seemingly not persistent and independent from the Ramadan effect. 5. Conclusions and implications Based on a series of analytical and statistical tests, this study found that the Ramadan effect is indeed a financial phenomenon present among various other anomalies in the financial market, especially capital markets in Muslim-majority countries. Similar to Rosh HaShanah and Yom Kippur, Ramadan is a religious event that can affect investors’moods and investment decisions (Bialkowski et al., 2012) through the combination of two herding factors that prompt risk-taking behavior and enhance social interaction among investors (Gavriilidis et al., 2015). Yet the effect is not strong enough to be the basis of an argument for the presence of the EMH anomaly or behavioral finance because the presence of the Ramadan effect is proven to be not persistent statistically in certain sub-periods, both in times of crisis and non-crisis. Even if the Ramadan effect is present, as described in previous studies (Bialkowski et al., 2012), this study shows that not all the countries observed have an efficient market, or even a weak-form efficient market. Thus, the presence of abnormal returns in particular periods has yet to be able to be justified as an EMH anomaly. In our results, the Ramadan effect is persistently present in only three countries: Kuwait, Oman, and Tunisia, and the magnitude is always larger than in other months. Moreover, other countries (i.e., Indonesia, Malaysia, Jordan, Morocco and Qatar) have experienced the Ramadan effect, but it is not persistently present in all sub-periods. Two final sample countries, Bahrain and SaudiArabia,havenever experiencedRamadan effect, displaying negative annualizedreturnswithin Ramadanin all periods. Akin to many EMH anomalies that have a weakened impact and even disappear after being documented in the literature (Schwert, Table 4.11 –Coincidence effect between the Ramadan and other Gregorian calendar effects. Period CRamadan effect Rworld January effect Weekend effect Christmas effect R 2 F-stat DW-stat 1989–1993 0.0005 * 0.0010 0.1827 *** 0.0006 0.0013 ** 0.0008 0.0228 0.0000 1.45 (1.6363) (1.0410) (5.2482) (0.6651) (2.0808) (0.1714) (6.6864) 1994–1998 0.0002 0.0003 0.5403 *** 0.0006 0.0016 * 0.0008 0.0784 0.0000 1.78 (0.3823) (0.2297) (9.9625) (0.4374) (1.7671) (0.0121) (18.5819) 1999–2003 0.00001 0.0007 0.0586 ** 0.0005 0.0012 0.0002 0.0086 0.0479 1.74 0.0442 (0.8222) (2.4786) (0.5465) (2.0226) (0.0369) (2.2424) 2004–2008 0.000008 0.0003 0.2309 0.0017 ** 0.0006 0.0046 0.0985 0.0000 1.77 (0.0325) (0.3313) (11.6151) (2.1964) (1.331) (1.0264) (28.3849) 2009–2013 0.000006 0.0006 0.2277 *** 0.0011 ** 0.0003 0.0011 0.1686 0.0000 2.03 (0.3490) (1.0469) (15.9849) (2.0678) (0.9085) (0.3646) (52.6719) 1989–2013 0.0002 * 0.0002 0.2164 *** 0.0003 0.0010 *** 0.0011 0.0522 0.0000 1.69 (1.3393) (0.3590) (17.3390) (0.6965) (3.5636) (0.4989) (61.0247) * Significant at 10%. ** Significant at 5%. *** Significant at 1%. arab economic and business journal 11 (2016) 55–71 69 2003), we find that the Ramadan effectalso has a non-persistent impact, except in three countries(Kuwait, Oman and Tunisia). The Ramadan effect is not significant when regression analysis is performed on portfolio returns. This research implies that investor psychology does affect investment behavior in the capital market, but relies on a single event that could affect price movement in the capital market. This study supports the argument that the Ramadan effect can still be explained by the efficient market hypothesis and does not necessarily need to be explained by behavioral finance. This study also finds the significance of the influence of the crisis factor on the Ramadan effect for both global and local crises. From the perspective of Islamic finance, these findings suggest the investment utility of exploiting the abnormal stock returns in some Muslim countries during Ramadan. Moreover, investors must be cautious about the effect of destabilizing potential of herding behavior that promotes systemic risk in the market. 6. Suggestions for further research The Ramadan effect is influenced by at least three factors: (i) the individual investor, (ii) the collective investors, and (iii) market fundamentals. In the literature, it is often assumed that the Ramadan effect can be impacted only by individual and collective investor factors arising directly from the fasting experience during Ramadan; the literature generally ignores the impact of economic and market fundamentals on the Ramadan effect. We therefore recommend the inclusion of these factors in future research on the Ramadan effect. In the study by Bialkowski et al. (2012), a significant Ramadan effect is found in 11 Muslim-majority countries, where one of the determinants of the occurrence of the Ramadan effect is suspected to be the psycho-religious factor of the investor, which is hypothesized to contribute to an increase in risk-taking behavior. Yet the use of cumulative abnormal returns cannot directly capture the impact of the Muslim investor’s risk-taking behavior unless other statistical measures are also used, i.e., skewness. Further research is recommended to utilize skewness to analyze risk-taking behavior in the Ramadan effect. Finally, other than risk-taking behavior, any study of the Ramadan effect should be sensitive to market liquidity as measured by transaction volume. Hopefully, future studies on the Ramadan effect will integrate the concept of market liquidity into the measurement of the Ramadan effect. Compliance with ethical standards This paper was presented at the conference on Finance and Development in Islamic Economies held at Bangor Business School, Bangor University on September 15, 2014, in collaboration with the Islamic Research Training Institute (IRTI) of the Islamic Development Bank (IDB). Conflicts of interest The authors declare that they have no conflicts of interest. REFERENCES Akerlof, G. (1980). A theory of social custom of which unemployment may be one consequence. Quarterly Journal of Economics,94(4), 749–775. Al-Hajieh, H., Redhead, K.,& Rodgers, T. (2011). Investor sentiment and calendaranomaly effects: A case study of theimpact of Ramadan on Islamic Middle Eastern markets. Research in International Business and Finance,25(3), 345–356. Al-Ississ, M. (2010). The impact of religious experience on financial markets. Working Paper, Harvard University. Al-Khazali,O. (2014).Revisitingfast profitinvestor sentimentand stockreturns duringRamadan. InternationalReviewof FinancialAnalysis,33, 157–170. Al-Mudhaf, F. (2012). The Islamic calendar effects: Evidence from twelve stock markets. International Research Journal of Finance & Economics, 87,185–191. Alper, C. E., & Arouba, S. B. (2001). Deseasonalizing macroeconomic date: A caveat to applied researchers in Turkey. ISE Review,5(18), 33–52. Arbel, A., & Strebel, P. (1983). Pay attention to neglected firms!. Journal of Portfolio Management,9(2), 37–42. Barro, R. J., & McCleary, R. M. (2003). Religion and economic growth across countries. American Sociological Review,68(5), 760–781. Bartholdy, J., Boyle, G. W., & Stover, R. D. (2004). Deposit insurance and the stock market: Evidence from Denmark. European Journal ofFinance, 10(6), 567–578. Beit-Hallahmi, B., & Argyle, M. (1997). The psychology of religious behavior: Belief and experience. London: Routledge. Bialkowski, J., Etebari, A., & Wisniewski, T. P. (2012). Fast profits: Investor sentiment and stock returns during Ramadan. Journal of Banking & Finance,36, 835–845. Binder, J. J. (1985). On the use of multivariate regression model in event studies. Journal of Accounting Research,23(1), 370–383. Binder, J. J. (1998). The event study methodology since 1969. Review of Quantitative Finance and Accounting,11(2), 111–137. Branch, B. (1977). A tax loss trading rule. Journal of Business,50(2), 198–207. Brown, R. I., & Warner, J. B. (1985). Using daily stock returns: The case of event studies. Journal of Financial Economics,14,3–31. 70 arab economic and business journal 11 (2016) 55–71 Cadsby, C. B., & Ratner, M. (1992). Turn-of-month and pre-holiday effects on stock returns: Some international evidence. Journal of Banking and Finance,16(3), 497–509. Calomiris, C. W., Love, I., & Peria, M. S. M. (2012). Stock returns’sensitivities to crisis shocks: Evidence from developed and emerging markets. Journal of International Money and Finance,31(4), 743–765. Chang, E. J., Lima, E. J. D., & Tabak, B. M. (2004). Testing for weak form efficiency in emerging markets. Emerging Market Review,3, 295–316. Corsetti, G., Pesenti, P., & Roubini, N. (1999). What caused the Asian currency and financial crisis? Working Paper, NBER. Daradkeh, T. K. (1992). Parasuicide during Ramadan in Jordan. Acta Psychiatrica Scandinavica,86(3), 253–254. Fama, E. F. (1965). Random walk in stock market prices. Financial Analysts Journal,51,55–59. Fama, E. F. (1970). Efficient capital markets: A review of theory and empirical work. Journal of Finance,25, 383–417. Fama, E. F. (1998). Market efficiency, long-term returns, and behavioral finance. Journal of Financial Economics,49, 283–306. Frankfurter, G. M., & McGoun, E. G. (2001). Anomalies in finance: What are they and what are they good for? International Review of Financial Analysis,10, 407–429. Frieder, L., & Subrahmanyam, A. (2004). Non-secular regularities in returns and volume. Financial Analysts Journal,60(4), 29–34. Fuhrman, J. (1998). Fasting and eating for health. A medical doctor’s program for conquering disease. New York: St. Martin's Press. Gavriilidis, K., Kallinterakis, V., & Tsalavoutas, I. (2015). Investor mood, herding and the Ramadan effect. Journal of Economic Behavior & Organization (In press). Han, S., Lerner, J. S., & Keltner, D. (2007). Feeling and consumer decision making: The appraisal-tendency framework. Journal of Consumer Psychology,17(3), 158–168. Hilary, G., & Hui, K. W. (2009). Does religion matter in corporate decision making in America? Journal of Financial Economics,93(3), 455–473. Hunter, W. C., Kaufman, G. G., & Krueger, T. H. (1999). The Asian financial crisis: Origins, implications and solutions. Kluwer Academic Publisher. Husain, F. (1998). A seasonality in the Pakistani equity market: The Ramadan effect. Pakistan Development Review,37(1), 77–81. Jensen, M. C. (1978). Some anomalous evidence regarding market efficiency. Journal of Financial Economics,6,95–101. Kannan, P., & Henry, P. B. (2008). Growth and return in emerging markets. In A.K. Rose, & T. Ito (Eds.), International financial issues in the Pacific Rim: Global imbalances, financial liberalization, and exchange rate policyNBER, The University of Chicago Press. Knerr, I., & Pearl, P. L. P. (2008). Ketogenic diet: Stoking energy stores and still posing questions. Experimental Neurology,211(1), 11–13. Kothari, S. P., & Warner, J. B. (2007). Econometrics of event studies. In E.B. Eckbo (Ed.), Handbook of corporate finance: Empirical corporate financeElsevier. Kuhn, T. S. (1977). The function of measurement in modern physical science. The essential tension. Chicago: University of Chicago Press165– 177. Kumar, A., Page, J. K., & Spalt, O. G. (2011). Religious beliefs, gambling attitudes and financial market outcomes. Journal of Financial Economics, 102(3), 671–708. Kvint, V. (2009). The global emerging market: Strategic management and economics. New York, USA: Routledge. Lakonishok, J., & Smidt, S. (1988). Are seasonal anomalies real? A ninety-year perspective. Review of Financial Studies,1(4), 403–425. Lucey, B., & Dowling, M. (2005). The role of feelings in investor decision-making. Journal of Economic Surveys,19(2), 211–239. MacKinlay, G. (1997). Event studies in economics and finance. Journal of Economic and Literature,35,13–39. Mustafa, K. (2011). The Islamic calendar effect on Karachi stock market. Pakistani Business Review 562–574. Nofsinger, J. R. (2002). Do optimists make the best investors? Corporate Finance Review,6(4), 11–17. Pantzalis, C., & Ucar, E. (2014). Religious holidays, investor distraction, and earnings announcement effects. Journal of Banking & Finance,47, 102–117. Rainly, A. (2006). Analysis of Ramadan effect in the return and volatility on the Indonesia stock exchange and the sectoral stock index for period 1996– 2005. Thesis, University of Indonesia. Reilly, F. K., & Brown, K. C. (2012). Analysis of investments and management of portfolios, 10th ed. South Western Cengage Learning. Reinganum, M. R. (1992). A revival of the small-firm effect. Journal of Portfolio Management,18(3), 55–62. Renneboog, L., & Spaenjers, C. (2012). Religion, economic attitudes and household finance. Oxford Economic Papers,64(1), 103–127. Ritter, J. R. (1991). The long run performance of initial public offerings. Journal of Finance,46,3–27. Rosen, H. S., & Wu, S. (2004). Portfolio choice and health status. Journal of Financial Economics,72(3), 457–484. Saleh, S. A., Elsharouni, S. A., Cherian, B., & Mourou, M. (2005). Effects of Ramadan fasting on waist circumference, blood pressure, lipid profile, and blood sugar on a sample of healthy Kuwaiti men and women. Malaysian Journal of Nutrition,11(2), 143–150. Schwert, G. W. (2003). Anomalies and market efficiency. Handbook of the Economics of Finance,1, 939–974. Seyyed, F., Abraham, A., & Al-Hajji, M. (2005). Seasonality in stock returns and volatility: The Ramadan effect. Research in International Business and Finance,19(3), 374–383. Segot, T. L., & Lucey, B. M. (2006). Efficiency in emerging markets–Evidence from the MENA region. International Financial Markets, Institutions and Money,18(1), 94–105. Shirai, S. (2009). The impact of the US subprime mortgage crisis on the world and East Asia-through analyses of cross-border capital movements. Working Paper, Keio University. Stulz, R. M., & Williamson, R. (2003). Culture, openness and finance. Journal of Financial Economics,70(3), 313–349. Thaler, R. (1993). Advances in behavioral finance. New York: Russell Sage Foundation. World Bank (2013). World Development Indicators. Washington DC, USA: World Bank. Worthington, A. C., & Higgs, H. (2006). Weak-form market efficiency in Asian emerging and developed equity markets: Comparative tests of random walk behavior. Accounting Research Journal,19(1), 54–63. Wright, W. F., & Bower, G. H. (1992). Mood effects on subjective probability assessment. Organizational Behavior and Human Decision Processes, 52(2), 276–291. arab economic and business journal 11 (2016) 55–71 71