Ramadan, Sentiment, and Sectoral Herding: Empirical Evidence from the Qatar Stock Exchange

1. Introduction

The efficient market hypothesis (EMH), which posits that asset prices fully reflect all available information, has long been a cornerstone of modern financial theory (Fama, 1970). However, a growing body of evidence has challenged this paradigm, revealing that investors are not always rational and that markets are subject to systematic anomalies driven by psychological and behavioral factors. This recognition gave rise to behavioral finance, a discipline that integrates insights from psychology into financial economics to explain why investors often deviate from rational decision-making (Kahneman and Tversky, 1979; Thaler, 1999).

Among the most extensively documented behavioral phenomena in financial markets is herding, defined as the tendency of investors to mimic the actions of others rather than relying on their own private information (Hirshleifer and Hong Teoh, 2003). Herding behavior has profound implications for market efficiency, price discovery, and systemic risk. When investors herd, they collectively amplify market movements, potentially leading to overvaluation, bubbles, and subsequent crashes (Lux, 1995; Focardi et al., 2002). Theoretically, herding can be classified as intentional, driven by informational or professional asymmetries, or spurious, resulting from common environmental factors such as shared sentiment, regulatory constraints, or relative homogeneity of investor characteristics (Gavriilidis et al., 2013; Stavroyiannis and Babalos, 2017). This distinction is critical because intentional herding can drive prices away from fundamentals and undermine market efficiency, whereas spurious herding reflects rational responses to common information (Rubbaniy et al., 2026).

Empirical research has documented herding across diverse markets, with particularly strong evidence emerging from emerging economies where information asymmetries are more pronounced and investor sophistication is relatively lower (Chang et al., 2000; Chiang and Zheng, 2010; Holmes et al., 2013). A consistent finding across these studies is the significant role of social mood and sentiment in fostering herding tendencies (Prechter, 2001; Olson, 2006; Blasco et al., 2012). Positive sentiment generates optimism and risk-seeking behavior, while negative sentiment induces caution and defensive strategies. The contagion of emotions among market participants, facilitated by social interaction and observational learning, creates conditions conducive to collective behavior (Hong et al., 2004, 2005; Liao et al., 2011). Recent studies have further emphasized the adaptive nature of herding, showing that sentiment measures, including market sentiment, news sentiment, and investor happiness, positively influence herding behavior, particularly during periods of low market returns or high stress (Ooi, 2025).

Religious occasions, which systematically influence the mood and behavior of large populations, provide unique natural experiments for examining the interplay between sentiment, social interaction, and herding. The Islamic holy month of Ramadan is particularly compelling in this regard. Celebrated by over 1.5 billion Muslims worldwide, Ramadan is characterized by heightened spiritual reflection, increased social interaction, collective prayer, and a positive, euphoric mood among observant Muslims (Daradkeh, 1992; Knerr and Pearl, 2008; Białkowski et al., 2012). Clinical research indicates that fasting during Ramadan reduces anxiety and tension while promoting feelings of well-being and social solidarity (Daradkeh, 1992; Lau et al., 2023). These psychological and social dynamics have been shown to translate into financial market outcomes, with studies documenting abnormally positive stock returns, reduced volatility, and altered trading volumes during Ramadan across several Muslim-majority countries (Al-Hajieh et al., 2011; Białkowski et al., 2012; Al-Khazali, 2014; Al-Awadhi, 2021; Al-Awadhi et al., 2024).

Despite extensive research on Ramadan-related market anomalies, the specific behavioral mechanism through which Ramadan influences market dynamics, particularly herding, remains underexplored. While Al-Hajieh et al. (2011) and Al-Khazali (2014) indirectly suggested that common psychological stimuli during Ramadan might induce herding tendencies, no explicit empirical tests were conducted until the seminal work of Gavriilidis et al. (2016). Using a sample of seven Muslim-majority countries (Bangladesh, Egypt, Indonesia, Malaysia, Morocco, Pakistan, and Turkey), they demonstrated that herding is significantly stronger during Ramadan than in non-Ramadan periods, attributing this to the enhanced social mood and increased social interactions characteristic of the holy month.

However, evidence from Islamic markets remains inconclusive and at times contradictory. While Gavriilidis et al. (2016) documented herding during Ramadan in Muslim-majority countries, Stavroyiannis and Babalos (2017) found significant anti-herding behavior in U.S. Dow Jones Islamic Index stocks, suggesting that Shariah-compliant investments may actually reduce collective behavior. This anti-herding phenomenon was found to be more intense during turbulent periods, implying that ethical screening criteria may moderate behavioral biases. Similarly, Mnif et al. (2020) examined herding behavior in Islamic stock and Sukuk markets using Hurst exponent estimation and found that herding is more intensive in the Islamic stock market but absent in the Sukuk market, suggesting that the nature of the financial instrument matters for herding formation.

Ah Mand et al. (2023) investigated whether herding differs between Islamic and conventional stocks in Malaysia and found that herding exists among Shariah-compliant stocks during both up and down markets with a non-linear relationship to market returns, while conventional stocks exhibit no herding behavior. This suggests that Islamic screening criteria may paradoxically amplify rather than moderate herding tendencies.

Within the GCC region specifically, studies have yielded particularly mixed results. Chaffai and Medhioub (2018) examined herding behavior in Islamic GCC stock markets using daily data from 2010 to 2016 and found evidence of herding during rising markets only, concluding that Islamic finance has an important contribution to moderate behavior in financial markets. In contrast, Medhioub and Chaffai (2018) applied monthly data from 2006 to 2016 and found herding in Saudi and Qatari Islamic stock markets only during down-market periods, suggesting that herding asymmetry depends on the frequency of data used. More recently, Loang and Ahmad (2023) examined economic and political factors on herding in Islamic GCC stock markets during the COVID-19 pandemic and found that herding exists in Shariah stocks before the pandemic but became more pronounced in both Islamic and conventional stocks during the pandemic. These contradictory findings highlight the need for a more nuanced, granular approach that considers sector-level heterogeneity and the role of religious calendar events.

Despite these important contributions, significant gaps remain in the literature that this study aims to address. First, sectoral heterogeneity has not been adequately examined, as most studies focus on aggregate market-level analysis, potentially masking important sectoral differences in investor base and sentiment sensitivity. Different sectors may exhibit varying susceptibility to sentiment-driven herding due to differences in investor base, information environment, and fundamental characteristics (Bennett, 2003). For instance, sectors with higher retail investor participation may be more prone to sentiment-driven herding, while sectors dominated by institutional investors may exhibit more rational behavior. Second, market condition interactions at the sector level remain unexplored; while some studies have examined herding asymmetry in general (Ah Mand et al., 2023; Tsania et al., 2026), the specific interaction between Ramadan-related herding and market conditions, whether markets are rising or falling, has not been systematically tested. Third, Qatar-specific evidence is lacking; while some studies have examined herding in GCC markets generally (Chaffai and Medhioub, 2018; Medhioub and Chaffai, 2018; Loang and Ahmad, 2023), no study has specifically examined Ramadan-induced herding in the Qatari market at the sector level. Given Qatar’s unique characteristics, high religious observance, concentrated market structure, and diverse sectoral composition, this represents a significant gap. Fourth, the distinction between intentional and spurious herding during Ramadan has not been examined, which is important for understanding market efficiency implications. Fifth, the post-COVID period has been largely overlooked, with most studies pre-dating the pandemic (exceptions include Loang and Ahmad, 2023; Ooi, 2025; Bougatef and Nejah, 2022); given that COVID-19 significantly altered market dynamics and investor behavior, a longer time horizon including post-pandemic data is needed to understand whether the Ramadan herding effect has changed over time.

This study addresses these gaps by conducting a comprehensive, sector-level analysis of herding dynamics during Ramadan on the Qatar Stock Exchange over the decade spanning 2016 to 2026. Qatar provides an ideal research setting for several reasons. First, Qatar has a high rate of religious observance, with a large proportion of Muslims among its population (Al-Awadhi, 2021). Second, the Qatar Stock Exchange is a well-developed emerging market with a sophisticated regulatory framework and diverse sectoral composition. Third, the availability of comprehensive daily trading data for all listed companies enables a rigorous empirical analysis. Fourth, as highlighted by Medhioub and Chaffai (2018), Qatar is one of the few GCC markets where significant herding behavior has been documented, making it a suitable context for further investigation.

Our dataset encompasses all 54 listed companies across seven sectors: Banks & Financial Services, Consumer Goods & Services, Industrials, Insurance, Real Estate, Transportation, and Telecoms. This balanced sample provides a robust foundation for sectoral analysis. We employ the Cross-Sectional Absolute Deviation (CSAD) methodology developed by Chang et al. (2000), which detects herding through the nonlinear relationship between return dispersion and market returns. Following Gavriilidis et al. (2016) and Ooi (2025), we augment this framework with a Ramadan dummy variable (D) to isolate herding patterns during the holy month. We further disaggregate our analysis by sector and by market condition (up-market versus down-market days) to provide a nuanced understanding of herding dynamics, consistent with the approaches of Chaffai and Medhioub (2018), Medhioub and Chaffai (2018), and Ah Mand et al. (2023). Our research addresses three primary questions:

RQ1: Does herding behavior exist during Ramadan on the Qatar Stock Exchange?

RQ2: Is herding more pronounced during Ramadan compared to non-Ramadan periods?

RQ3: Does this herding pattern vary across sectors and market conditions?

Based on the theoretical and empirical literature reviewed, we propose three hypotheses. First, given the positive mood, reduced risk aversion, and enhanced social interactions during Ramadan that create conditions conducive to herding (Gavriilidis et al., 2016; Białkowski et al., 2012; Daradkeh, 1992), we hypothesize that during Ramadan, the Qatar Stock Exchange experiences a significant increase in herding behavior compared to non-Ramadan periods (H1) . Second, recognizing that sectoral heterogeneity in herding is well-documented (Ukpong et al., 2021; Tsania et al., 2026) and that sectors with higher retail investor participation and greater sentiment sensitivity are likely to exhibit stronger herding (Bennett, 2003; Medhioub and Chaffai, 2018), we hypothesize that herding during Ramadan varies across sectors, with sectors characterized by higher retail investor participation and greater sentiment sensitivity exhibiting stronger herding (H2) . Third, given that the Ramadan effect is driven by mood and sentiment rather than market fundamentals, and if Ramadan-induced herding is spurious, driven by common emotional exposure, it should persist regardless of whether markets are rising or falling (Gavriilidis et al., 2016; Chaffai and Medhioub, 2018; Medhioub and Chaffai, 2018), we hypothesize that the Ramadan-herding relationship is robust to market conditions, persisting in both up-market and down-market days (H3) .

Our findings reveal significant and robust evidence of herding during Ramadan across all seven sectors, with herding intensity being substantially stronger during the holy month than during non-Ramadan periods. This pattern persists in both up-market and down-market conditions, suggesting that the Ramadan effect on herding is robust to market direction. Sectoral heterogeneity is evident, with the Telecoms and Real Estate sectors exhibiting the strongest herding, while the Consumer Goods sector shows the weakest. These findings contribute to the behavioral finance literature in several ways.

First, we provide the first sector-level empirical evidence of Ramadan-induced herding in the Qatari market, extending the aggregate-level findings of Gavriilidis et al. (2016) to a granular sectoral context. This addresses the call by Mnif et al. (2020) for further research on herding behavior in different Islamic financial markets and responds to the observation by Ah Mand et al. (2023) that few studies have examined herding in Islamic financial markets at the sector level. Second, we challenge the “Islamic moderation hypothesis” by demonstrating that religious sentiment can amplify rather than suppress herding during Ramadan. While Stavroyiannis and Babalos (2017) found anti-herding in U.S. Islamic stocks and Chaffai and Medhioub (2018) suggested that Islamic finance moderates behavioral biases, our findings indicate that the relationship between Islamic principles and behavioral tendencies is context-dependent and potentially overshadowed by social and emotional contagion during collective religious observances. This contributes to the nascent “Islamic behavioral finance” paradigm (Din et al., 2021; Ooi, 2025; Syibawaih et al., 2026) by revealing the complex interplay between Islamic principles and market psychology. Third, we reconcile contradictory evidence on GCC herding by demonstrating that herding during Ramadan is robust to market conditions (both up and down markets), but outside Ramadan follows the asymmetric patterns documented in prior literature (Chaffai and Medhioub, 2018; Medhioub and Chaffai, 2018). This suggests Ramadan acts as a structural break in normal market behavior, a finding with important implications for time-varying herding models (Rubbaniy et al., 2026) and dynamic asset pricing. Fourth, we provide practical implications for portfolio management, market regulation, and risk management. Asset managers should be aware of the heightened herding during Ramadan when making asset allocation decisions, particularly in high-herding sectors such as Telecoms and Real Estate. Regulators should monitor the potential for systemic risk amplification during this period and consider enhanced monitoring and investor education initiatives. These implications align with the practical concerns raised by Loang and Ahmad (2023) regarding the need for policymakers to be wary of factors that may cause stock prices to deviate from fundamental values.

The remainder of this paper is organized as follows. Section 2 reviews the relevant literature on herding behavior, the Ramadan effect, and their intersection, drawing on the empirical evidence from Islamic and GCC markets. Section 3 describes the methodology, data, and empirical framework, detailing the CSAD model and the Ramadan dummy variable approach. Section 4 presents the empirical results, including descriptive statistics, full-sample herding estimates, and sector-level analysis under different market conditions. Section 5 discusses the findings in the context of existing literature, highlighting theoretical and practical implications. Section 6 concludes with a summary of contributions, limitations, and directions for future research.

2. Literature Review

2.1. Behavioral Finance and Herding in the Context of Islamic Markets

2.1.1. The Qatar Stock Exchange Context

The Qatar Stock Exchange (QSE) represents a unique laboratory for examining behavioral dynamics in Islamic capital markets. As one of the fastest-growing Gulf Cooperation Council (GCC) markets, Qatar has experienced significant transformation in its investor base and market microstructure over the past decade. The QSE is characterized by a high concentration of Shariah-compliant stocks, a dominant presence of retail investors, and a regulatory framework that aligns with Islamic financial principles (Al-Awadhi, 2021; Loang & Ahmad, 2023). As of 2026, the QSE comprises 54 listed companies across seven sectors, with banking and financial services (24.1%) and consumer goods and services (27.8%) representing the largest segments. This sectoral diversity, combined with Qatar’s high rate of religious observance, makes the QSE an ideal setting for examining how religious sentiment influences investor behavior.

Recent studies on GCC and Islamic markets have documented significant herding tendencies, though findings remain mixed regarding the conditions under which herding manifests. Chaffai and Medhioub (2018) examined herding behavior in Islamic GCC stock markets using daily data from 2010 to 2016 and found evidence of herding during rising markets only, concluding that Islamic finance may moderate behavioral biases. In contrast, Medhioub and Chaffai (2018) applied monthly data from 2006 to 2016 and found herding in Qatari and Saudi Islamic stock markets only during down-market periods, suggesting that herding asymmetry depends on data frequency and market conditions. Loang and Ahmad (2023) extended this analysis by examining economic and political factors on herding in Islamic GCC stock markets during the COVID-19 pandemic, finding that herding intensified during crisis periods and was influenced by economic factors more than political ones. These contradictory findings highlight the need for a more nuanced, granular approach that considers sector-level heterogeneity and the role of religious calendar events.

2.1.2. Theoretical Foundations of Behavioral Finance

Behavioral finance challenges the traditional assumption of investor rationality by integrating psychological insights into financial economics (Kahneman and Tversky, 1979; Thaler, 1999). The efficient market hypothesis (EMH), which posits that asset prices fully reflect all available information, has been systematically critiqued by behavioral finance scholars who demonstrate that investors often act irrationally, influenced by cognitive biases and emotional responses rather than objective fundamentals (Umeaduma, 2024). A positive mood state can make investors more optimistic and willing to undertake riskier investment decisions (Wright and Bower, 1992; Shu, 2010), while negative mood induces caution and risk aversion.

Among the most extensively documented behavioral phenomena in financial markets is herding, defined as the tendency of investors to mimic the actions of others rather than relying on their own private information (Hirshleifer and Hong Teoh, 2003). Herding behavior has profound implications for market efficiency, price discovery, and systemic risk. When investors herd, they collectively amplify market movements, potentially leading to overvaluation, bubbles, and subsequent crashes (Lux, 1995; Focardi et al., 2002). Theoretically, herding can be classified into intentional herding, driven by informational, reputational, or compensation-based asymmetries where investors deliberately imitate others assuming superior information (Scharfstein and Stein, 1990; Rubbaniy et al., 2026), and spurious herding, which results from common environmental factors such as shared sentiment, regulatory constraints, or investor homogeneity (Gavriilidis et al., 2013). This distinction is critical: spurious herding does not necessarily undermine market efficiency, whereas intentional herding can drive prices away from fundamentals (Stavroyiannis and Babalos, 2017).

Social mood and sentiment are significant drivers of herding, with positive sentiment fostering optimism and risk-taking while negative sentiment induces caution (Prechter, 2001; Olson, 2006; Blasco et al., 2012). Social interaction facilitates emotional contagion among market participants, creating conditions conducive to collective behavior (Hong et al., 2004, 2005; Liao et al., 2011). Ooi (2025) recently examined adaptive herding behavior in Shariah-compliant stocks across Malaysia, Indonesia, and Saudi Arabia, finding that investor happiness and market sentiment positively influence herding behavior, particularly during periods of low market returns or high stress, while negative news sentiment disrupts herding. This highlights the adaptive nature of herding in Islamic markets and underscores the importance of sentiment measures in understanding investor behavior.

In the context of Islamic markets, the relationship between herding and behavioral biases is particularly nuanced. Din et al. (2021) investigated the impact of behavioral biases on herding for Islamic financial products with the mediation of Shariah literacy. Their findings revealed that self-attribution, illusion of control, and information availability have a positive and significant impact on herding for Islamic financial products, while Shariah literacy showed an insignificant direct impact but served as a significant mediator that turned previously significant relationships insignificant. This suggests that Shariah literacy can mitigate the influence of cognitive biases on herding behavior, reinforcing the importance of religious literacy in Islamic markets. Boulanouar et al. (2026) extended this line of inquiry by investigating behavioral drivers of Bitcoin investment among Muslim investors, finding that profitability, trust, awareness, and risk tolerance are significant predictors of investment intention, whereas subjective norms and compatibility with Islamic finance are not. This suggests a shift toward more individualistic, market-driven investment behavior among Muslim investors, where financial incentives and personal confidence outweigh traditional religious or institutional influences.

2.1.3. Herding in Islamic Vs. Conventional Markets

A growing body of research has examined whether herding behavior differs between Islamic and conventional stocks. Ah Mand et al. (2023) investigated herding behavior and stock market conditions in Malaysia and found that herding exists among Shariah-compliant stocks during both up and down markets with a non-linear relationship to market returns, while conventional stocks exhibit no herding behavior. This suggests that Islamic screening criteria may paradoxically amplify rather than moderate herding tendencies, potentially because the restricted investment universe of Shariah-compliant stocks leads to more correlated trading. Mnif et al. (2020) examined herding behavior and Islamic market efficiency in Dow Jones and Sukuk markets using Hurst exponent estimation and found that herding is more intensive in Islamic stock markets but absent in the Sukuk market. This indicates that the nature of the financial instrument matters for herding formation, Sukuk, being asset-backed and less speculative, may be less susceptible to sentiment-driven behavior. The authors also found that the Dow Jones Islamic Market (DJIM) is the most efficient market compared to its conventional counterpart, suggesting that Shariah screening may enhance market efficiency in some contexts.

Stavroyiannis and Babalos (2017) provided a contrasting perspective by examining herding, faith-based investments, and the global financial crisis using static and dynamic models. Employing the highly liquid constituent stocks of the U.S. Dow Jones Islamic Index from 2007 to 2014, they documented significant anti-herding behavior that was robust across different formulations and testing procedures. This anti-herding phenomenon was found to be more intense during turbulent periods, suggesting that Shariah-compliant investments may actually reduce collective behavior, particularly during crises. The authors attributed this to the conservative nature of Shariah-compliant investments, which exclude sin stocks and limit leverage, thereby encouraging more independent investment decisions. This finding directly challenges the notion that Islamic principles uniformly moderate behavioral biases and highlights the context-dependent nature of herding in Islamic markets.

Syibawaih et al. (2026) conducted an interdisciplinary analysis of overconfidence, herding behavior, and Shariah principles in Islamic capital market dynamics. Their systematic literature review found that overconfidence increases trading intensity and risk exposure, while herding behavior amplifies short-term volatility and convergence of Islamic stock prices, particularly during periods of market uncertainty. However, Shariah principles, through screening mechanisms, prohibitions on riba (usury) and gharar (excessive uncertainty), and reinforcement of ethical values such as justice and responsibility, serve as moderating variables that relatively reduce the destructive impact of these biases. This suggests that Islamic capital markets are not completely immune to behavioral distortions but possess a normative-institutional framework that can strengthen market discipline. Tsania et al. (2026) demonstrated in their study of banking stocks in Indonesia that herding behavior exclusively exists during bearish market conditions, and both market volatility and trading volume exhibit a significant negative effect on herding tendency, meaning that heightened market turbulence stimulates divergence of opinion rather than uniform collective actions. This underscores the importance of examining herding under different market regimes.

2.1.4. The Role of Islamic Principles in Moderating Behavioral Biases

The Islamic moral economy framework offers a distinctive lens for understanding investor behavior, grounded in principles that include the prohibition of riba (usury), gharar (excessive uncertainty), and maysir (gambling), alongside asset-backed financing, profit and loss sharing, and ethical screening (Din et al., 2021; Fadhilah, 2023). These principles have been shown to influence investor behavior, with Shariah literacy mediating the relationship between cognitive biases and herding (Din et al., 2021), and the normative framework of Islamic finance counterbalancing psychological tendencies that drive herding (Fadhilah, 2023). Chaffai et al. (2025) found that herding behavior in GCC Islamic banks is influenced by the moral economy framework, suggesting Islamic principles shape collective behavior differently than in conventional markets. However, the extent to which these principles effectively moderate behavioral biases remains an open question. The literature presents contradictory findings, some studies document herding in Islamic markets (Chaffai & Medhioub, 2018; Medhioub & Chaffai, 2018; Loang & Ahmad, 2023), while others find anti-herding (Stavroyiannis & Babalos, 2017; Mnif et al., 2020), suggesting the relationship between Islamic principles and behavioral biases is complex and context-dependent. Factors such as market conditions, investor sophistication, information availability, and religious calendar events may interact with Islamic principles to either amplify or suppress herding tendencies.

2.2. Herding Detection Methodology

2.2.1. Evolution of Herding Detection Models

Christie and Huang (1995) proposed the first empirical methodology to detect herding, employing the Cross-Sectional Standard Deviation (CSSD) measure. Their approach assumed a linear relationship between return dispersion and market returns, with herding indicated by a reduction in dispersion during periods of extreme market movements. However, the CSSD measure has been criticized for its sensitivity to outliers and its inability to capture non-linear herding dynamics (Economou et al., 2011). Recognizing these limitations, Chang et al. (2000) developed the Cross-Sectional Absolute Deviation (CSAD) model, which has become the standard approach for detecting herding and has been widely adopted in subsequent research (Gavriilidis et al., 2016; Chiang and Zheng, 2010; Economou et al., 2011; Ooi, 2025). The CSAD measure is calculated as:

$$CSA{D}_t={\displaystyle \frac{1}{N}}{\sum }_{i=1}^N\mid R_{i,t}-R_{m,t}\mid$$

The CSAD model is grounded in the Capital Asset Pricing Model (CAPM) framework, which stipulates a linear relationship between return dispersion and market returns. Under the conditional CAPM, the relationship between CSAD and absolute market returns is expected to be positive and linear, as individual stocks have different sensitivities to market movements. However, during periods of herding, investors suppress their private information and follow the market consensus, causing return dispersion to decrease or even become negative during extreme market movements. This non-linear relationship is captured by the following empirical specification:

$$CSA{D}_t={\alpha }_0+{\alpha }_1\mid R_{m,t}\mid +{\alpha }_2{R}_{m,t}^2+{\epsilon }_t$$

A significantly negative coefficient on the squared market return term (${\alpha }_2$) serves as direct evidence of herding, as it indicates that return dispersion decreases during periods of extreme market stress.

2.2.2. Extensions and Modifications of the CSAD Model

Subsequent research has extended the CSAD framework in several ways. Rubbaniy et al. (2026) employed state-space models and the Fama and French (2015) five-factor model to differentiate between fundamental and intentional herding in North American energy stocks. They decomposed CSAD into fundamental and non-fundamental components, finding that herding in energy stocks is primarily motivated by intentional drivers rather than fundamental factors. This methodological advancement allows researchers to distinguish between herding driven by rational responses to common information (fundamental) and herding driven by behavioral biases (intentional).

Ah Mand et al. (2023) and Tsania et al. (2026) incorporated market conditions into CSAD analysis by splitting samples into bullish and bearish regimes. This approach recognizes that herding behavior may be asymmetric, occurring only during specific market conditions. Tsania et al. (2026) found that herding exclusively exists during bearish market conditions in Indonesian banking stocks, while no herding is detected during bullish trends.

Rubbaniy et al. (2026) employed quantile-on-quantile regression (QQR) to capture the regime-specific dependence of herding behavior on various macroeconomic variables. This approach offers several advantages over standard quantile regression: it captures even minute dispersion in data and is robust to outliers, heteroskedasticity, and skewness; it encapsulates the dependence structure and reveals both the nonlinear nature and evolution of independent variables; and it captures asymmetric relationships during bearish, bullish, and extreme market conditions. Stavroyiannis and Babalos (2017) employed rolling window analysis and time-varying parameter models to capture dynamic herding patterns. This approach recognizes that herding intensity may vary over time due to changing market conditions, investor sentiment, and regulatory environments. Their analysis revealed that anti-herding in U.S. Islamic stocks tends to be more intense during turbulent periods, highlighting the importance of time-varying analysis.

Tsania et al. (2026) focused specifically on the banking sector in Indonesia, demonstrating that herding exclusively exists during bearish market conditions. This sectoral approach is particularly relevant for our study, as different sectors may exhibit varying susceptibility to sentiment-driven herding.

2.2.3. CSAD in Islamic Markets

The CSAD methodology has been extensively applied in Islamic markets. Chaffai and Medhioub (2018) used CSAD with GARCH-type models and quantile regression to examine herding in Islamic GCC stock markets, finding evidence of herding during rising markets only. Medhioub and Chaffai (2018) applied CSAD with both OLS and GARCH estimations to examine herding in GCC Islamic stock markets using monthly data, finding significant herding in Qatari and Saudi Islamic stock markets only during down-market periods. Ooi (2025) applied CSAD along with static regression models, rolling window regression, and quantile-on-quantile regression to examine adaptive herding in Shariah-compliant stocks across Malaysia, Indonesia, and Saudi Arabia, finding significant herding in all three markets with varying degrees of volatility and sensitivity to sentiment. These studies demonstrate the versatility of the CSAD framework for examining herding dynamics in Islamic markets.

2.3. The Ramadan Effect

2.3.1. Ramadan: Religious Observance and Psychological Impact

Ramadan is the ninth month of the Islamic calendar, during which Muslims abstain from eating, drinking, and other physical indulgences from sunrise until sunset (Białkowski et al., 2012). The month is characterized by heightened spiritual reflection, increased social interaction, collective prayer, and a positive, euphoric mood among observant Muslims (Daradkeh, 1992; Knerr and Pearl, 2008). Clinical research has documented significant psychological effects of Ramadan fasting. Daradkeh (1992) found that fasting during Ramadan reduces anxiety and tension while promoting feelings of well-being and social solidarity. Lau et al. (2023) examined the effects of Ramadan fasting on mood and performance of male adolescent archers, finding improvements in mood states during the fasting period. Knerr and Pearl (2008) noted that the ketogenic state induced by fasting may have neuroprotective effects, potentially influencing cognitive function and emotional regulation. These psychological and physiological effects create a unique social environment characterized by reduced risk aversion, increased optimism, and enhanced social cohesion (Białkowski et al., 2012).

2.3.2. Ramadan Effect on Financial Markets

The psychological and social dynamics of Ramadan have been shown to translate into financial market outcomes, creating what is known as the “Ramadan effect.” Al-Hajieh et al. (2011) examined investor sentiment and calendar anomaly effects in Islamic Middle Eastern markets, finding abnormally positive stock returns during Ramadan. Białkowski et al. (2012) conducted a comprehensive study of the Ramadan effect across several Muslim-majority countries and documented significantly higher stock returns and lower volatility during Ramadan, attributing this to improved investor sentiment and reduced risk aversion.

Al-Khazali (2014) revisited the Ramadan effect using data from twelve stock markets and found that Ramadan returns are significantly higher than non-Ramadan returns, with the effect being more pronounced in countries with higher religious observance. Al-Awadhi (2021) examined the effect of religiosity on stock market speculation and found that religious observance influences investor behavior, with higher religiosity associated with more conservative investment decisions. Al-Awadhi et al. (2024) extended this analysis by examining religious practices, fasting, and individuals’ trading behavior, finding significant changes in trading activity during Ramadan.

Seasonality in Ramadan returns has been documented across multiple countries. In Pakistan, Egypt, Saudi Arabia, and Turkey, Al-Hajieh et al. (2011) and Białkowski et al. (2012) found significantly higher returns during Ramadan. In Indonesia, Morocco, and Oman, Al-Khazali (2014) and Almudhaf (2012) documented positive Ramadan returns and reduced volatility. In Qatar, Kuwait, and Jordan, Hassan and Kayser (2019) and Al-Awadhi et al. (2024) found a Ramadan premium in Gulf markets. In Bangladesh, Hassan and Kayser (2019) confirmed the Ramadan effect in a South Asian market context.

2.3.3. Ramadan and Trading Behavior

Recent research has examined how Ramadan affects trading behavior specifically. Al-Awadhi et al. (2024) examined trading behavior during Ramadan on the Boursa Kuwait and found a significant decline in individual trading activity and a significant increase in institutional buy-side trading during Ramadan. This suggests that the impact of Ramadan-related sentiment is more pronounced at the individual level, with retail investors reducing their trading activity while institutional investors increase their buying. The authors attributed this to the positive mood and reduced risk aversion during Ramadan, which may encourage institutional investors to take more aggressive positions.

In the GCC context, Gabbori et al. (2024) studied herding behavior during major Islamic events (Ashura and Eid) in the Saudi stock market by considering investors’ mood using static herding models. Their findings showed that investor mood affects herding after controlling for macroeconomic factors, and these findings align with existing studies on the impact of Islamic events on investor behavior (Rubbaniy, Tee et al., 2021; Sibande et al., 2023). This suggests that religious events beyond Ramadan also influence herding dynamics in Islamic markets.

2.4. Herding and Ramadan: The Link

2.4.1. Theoretical Mechanisms

The positive mood and enhanced social interactions during Ramadan create conditions conducive to herding (Gavriilidis et al., 2016). The psychological frame induced by Ramadan, characterized by reduced risk aversion and increased optimism, leads investors to follow the crowd rather than rely on their own analysis (Wright and Bower, 1992; Nofsinger, 2002). This herding is likely spurious in nature, resulting from the common emotional exposure of investors to the Ramadan experience.

Four interconnected mechanisms explain how Ramadan may intensify herding. First, mood-based decision making occurs as positive mood reduces cognitive effort and increases heuristic reliance, making investors more susceptible to social influence and collective behavior (Białkowski et al., 2012; Wright and Bower, 1992). Second, social cohesion and contagion operates through increased social interaction that facilitates emotional contagion among investors, amplifying collective sentiment (Hong et al., 2004, 2005; Ooi, 2025). Third, information cascades emerge as reduced information processing encourages investors to infer information from others’ actions, particularly when distracted by religious observances (Din et al., 2021). Fourth, reduced risk aversion results from positive mood lowering risk perception, encouraging trend-following behavior (Shu, 2010). These mechanisms collectively create a behavioral environment where investors are more likely to follow the crowd rather than conduct independent analysis, particularly when religious observances reduce time available for fundamental evaluation (Din et al., 2021). This theoretical discussion leads to the formulation of our first hypothesis:

H1. 

During Ramadan, the Qatar Stock Exchange experiences a significant increase in herding behavior compared to non-Ramadan periods.

Gavriilidis et al. (2016) provided direct evidence that herding is significantly stronger during Ramadan than in non-Ramadan periods across multiple Muslim-majority countries. The positive mood, reduced risk aversion, and enhanced social interactions during Ramadan create conditions conducive to herding (Białkowski et al., 2012; Daradkeh, 1992). In the GCC context, Loang and Ahmad (2023) found that emotional and social factors amplify herding during crises. Given that Ramadan is a period of heightened emotional and social activity, we expect herding to intensify during this period in the Qatari market.

2.4.2. Empirical Evidence

Gavriilidis et al. (2016) provided the first explicit test of Ramadan-induced herding across seven Muslim-majority countries (Bangladesh, Egypt, Indonesia, Malaysia, Morocco, Pakistan, and Turkey). They found significant herding during Ramadan in most markets, with the effect robust across various market conditions including domestic returns, market volume, US investor sentiment, and the global financial crisis. This established Ramadan as a unique behavioral environment conducive to herding. However, contradictory evidence exists. Elshqirat (2020) tested the Ramadan influence on herding at both market and sector levels in Jordan and found no herding during or after Ramadan, contradicting Gavriilidis et al. (2016). At the sector level, herding was absent during Ramadan but present outside Ramadan in industrial and services sectors, while absent in the financial sector entirely. This contradiction highlights the need for context-specific investigation

In the GCC context, the evidence is mixed. Medhioub and Chaffai (2018) found herding in Qatari Islamic stock markets during down-market periods, while Chaffai and Medhioub (2018) found herding only during rising markets. Loang and Ahmad (2023) found that herding intensified during the COVID-19 pandemic across both Islamic and conventional stocks. These contradictory findings underscore the importance of examining herding under different market conditions and market contexts.

These contradictory findings highlight the need for sector-level analysis. Different sectors may exhibit varying susceptibility to Ramadan-induced herding due to differences in investor base, information environment, and fundamental characteristics (Bennett, 2003). This leads to our second hypothesis:

H2. 

Herding during Ramadan varies across sectors, with sectors characterized by higher retail investor participation and greater sentiment sensitivity exhibiting stronger herding.

Sectoral heterogeneity in herding is well-documented (Ukpong et al., 2021; Tsania et al., 2026). Sectors with higher retail investor participation and greater sensitivity to sentiment, such as Telecoms and Real Estate, are likely to exhibit stronger herding during Ramadan, while sectors dominated by institutional investors and fundamental analysis, such as Consumer Goods, may exhibit weaker herding (Bennett, 2003; Medhioub and Chaffai, 2018).

2.4.3. Individual Vs. Institutional Traders During Ramadan

The differential response of individual and institutional investors to Ramadan is critical for understanding herding dynamics. Individual traders, often categorized as noise traders (Peress and Schmidt, 2021), are more influenced by sentiment and social influences than institutional investors (Barber and Odean, 2000; Kumar and Lee, 2006). Religion may have different effects on each group: individual traders may be more influenced by religious observances due to personal connection, while institutional traders may prioritize profit maximization and deviate from religious norms when profit opportunities are high (Kumar and Page, 2014).

Al-Awadhi et al. (2024) found a significant decline in individual trading activity and a significant increase in institutional buy-side trading during Ramadan on the Boursa Kuwait. This asymmetry has important implications for herding. If individual investors reduce trading during Ramadan, noise trader herding may decline, but institutional herding may dominate as individual investors passively follow institutional actions. If institutional investors increase trading during Ramadan, this may amplify institutional herding as institutions follow each other’s trading patterns, encouraged by positive mood. If both groups respond to Ramadan sentiment, market-wide herding may occur, consistent with Gavriilidis et al. (2016). Furthermore, Din et al. (2021) found that Shariah literacy mediates the relationship between cognitive biases and herding, suggesting that investors with higher religious knowledge may be less susceptible to Ramadan-induced herding.

The literature on herding asymmetry is mixed: some studies find herding only during rising markets (Chaffai and Medhioub, 2018), others only during falling markets (Medhioub and Chaffai, 2018; Bougatef and Nejah, 2022), and others in both conditions (Ah Mand et al., 2023). The Ramadan effect, driven by mood and sentiment rather than market fundamentals, is expected to be robust to market direction. This leads to our third hypothesis:

H3. 

The Ramadan-herding relationship is robust to market conditions, persisting in both up-market and down-market days.

The literature on herding asymmetry is mixed: some studies find herding only during rising markets (Chaffai and Medhioub, 2018), others only during falling markets (Medhioub and Chaffai, 2018; Bougatef and Nejah, 2022), and others in both conditions (Ah Mand et al., 2023). The Ramadan effect, driven by mood and sentiment rather than market fundamentals, is expected to be robust to market direction. If Ramadan-induced herding is spurious, driven by common emotional exposure, it should persist regardless of whether markets are rising or falling (Gavriilidis et al., 2016).

3. Methodology

3.1. Empirical Framework

Our empirical framework is grounded in the herding detection methodology pioneered by Christie and Huang (1995) and later refined by Chang et al. (2000). According to this theoretical foundation, herding manifests as a noticeable reduction in the cross-sectional dispersion of individual stock returns during periods of extreme market stress. While Christie and Huang (1995) postulated a linear relationship between return dispersion and aggregate market returns, substantial empirical evidence suggests that herding often introduces non-linear dynamics into capital markets. To account for this, Chang et al. (2000) developed a modified measure that explicitly incorporates the possibility of non-linearities in the dispersion-return relationship.

The CSAD approach has been widely validated in Islamic market contexts. Chaffai and Medhioub (2018) employed the CSAD methodology with GARCH-type models and quantile regression to examine herding in Islamic GCC stock markets, finding evidence of herding during rising markets. Medhioub and Chaffai (2018) applied CSAD with both OLS and GARCH estimations to examine herding in GCC Islamic stock markets using monthly data, finding significant herding in Qatari and Saudi Islamic stock markets during down-market periods. Ooi (2025) applied CSAD along with static regression models, rolling window regression, and quantile-on-quantile regression to examine adaptive herding in Shariah-compliant stocks across Malaysia, Indonesia, and Saudi Arabia, finding significant herding in all three markets with varying degrees of volatility and sensitivity to sentiment. Rubbaniy et al. (2026) further extended the CSAD framework by employing state-space models and quantile-on-quantile regression to distinguish between intentional and fundamental herding, demonstrating the versatility of the CSAD framework for examining herding dynamics across different market conditions.

Consequently, our study adopts the CSAD approach, operationalized through the following empirical specification:

$$CSA{D}_t={\alpha }_0+{\alpha }_1\mid r_{m,t}\mid +{\alpha }_2{r}_{m,t}^2+{\epsilon }_t$$

(1)

where CSAD is the cross-sectional absolute deviation of returns calculated as:

$$CSA{D}_t={\displaystyle \frac{1}{n}}{\sum }_{i=1}^N\mid r_{i,t}-r_{m,t}\mid$$

(2)

where $r_{i,t}$ is the return on security $i$ on day $t$, $r_{m,t}$ is the market-average on day $t$ (calculated by averaging the returns of all securities for day $t$), and $n$ is the total number of securities traded on day $t$.

According to Chang et al. (2000), under rational asset pricing, return dispersion exhibits a linear and positive relationship with absolute market returns due to heterogeneous stock sensitivities. However, herding disrupts this linearity, causing dispersion to decrease during extreme market movements. To account for this non-linearity, we include a squared market return term; a significantly negative coefficient on this term (${\alpha }_2<0$) serves as direct evidence of herding, indicating that return dispersion contracts during periods of extreme market stress as investors suppress their private information and follow the market consensus (Gavriilidis et al., 2016; Chiang and Zheng, 2010).

3.1.1. Ramadan-Augmented CSAD Model

To isolate the Ramadan effect, we incorporate a dummy variable ($D$) taking the value of 1 during Ramadan and 0 otherwise, following the identification procedure of Al-Khazali (2014) and Gavriilidis et al. (2016). The Ramadan period is identified based on the official Islamic calendar, with the specific dates verified through multiple sources to ensure accuracy across the 2016-2026 sample period. The empirical model is specified as follows:

$$CSA{D}_t={\alpha }_0+{\alpha }_{1}D\mid r_{m,t}\mid +{\alpha }_2\left(1-D\right)\mid r_{m,t}\mid +{\alpha }_{3}D{r}_{m,t}^2+{\alpha }_4\left(1-D\right)r_{m,t}^2+{\epsilon }_t$$

(3)

where:

$CSA{D}_t$ = Cross-Sectional Absolute Deviation of sector $i$ returns from the market return

$r_{m,t}$ = Market return on day $t$$D$ = Dummy variable = 1 during Ramadan, 0 otherwise

${\alpha }_3<0$ and statistically significant → Herding exists during Ramadan

${\alpha }_4<0$ and statistically significant → Herding exists outside Ramadan

Given the above discussion, significantly negative values for ${\alpha }_3$ and ${\alpha }_4$ would indicate the presence of significant herding during Ramadan-days and outside Ramadan-days, respectively. This specification follows the approach of Gavriilidis et al. (2016), who demonstrated that herding is significantly stronger during Ramadan than in non-Ramadan periods across multiple Muslim-majority countries. The inclusion of separate coefficients for Ramadan and non-Ramadan periods allows us to directly test whether the Ramadan effect documented in aggregate markets (Gavriilidis et al., 2016; Białkowski et al., 2012) holds at the sector level in the Qatari market.

3.1.2. Market Condition Analysis

Having run Equation (3) for each of our sample sectors, we then assess whether our results are robust to changes in market conditions. Following the approach of Chaffai and Medhioub (2018), Medhioub and Chaffai (2018), and Ah Mand et al. (2023), we examine whether herding exhibits differences between days of positive versus days of negative domestic market returns, proxied here through $r_{m,t}$. This analysis is crucial given the contradictory findings in the GCC literature, Chaffai and Medhioub (2018) found herding only during rising markets, while Medhioub and Chaffai (2018) found herding only during down-market periods. By analyzing both market conditions, we can reconcile these conflicting findings and determine whether Ramadan herding is robust to market direction.

In this case, we run the following set of equations for each sector:

$$CSA{D}_t={\alpha }_0+{\alpha }_1^{UP}D\mid r_{m,t}\mid +{\alpha }_2^{UP}\left(1-D\right)\mid r_{m,t}\mid +{\alpha }_3^{UP}D{r}_{m,t}^2+{\alpha }_4^{UP}\left(1-D\right)r_{m,t}^2+{\epsilon }_t$$

(4)

$$CSA{D}_t={\alpha }_0+{\alpha }_1^{DOWN}D\mid r_{m,t}\mid +{\alpha }_2^{DOWN}\left(1-D\right)\mid r_{m,t}\mid +{\alpha }_3^{DOWN}D{r}_{m,t}^2+{\alpha }_4^{DOWN}\left(1-D\right)r_{m,t}^2+{\epsilon }_t$$

(5)

where the superscript UP (DOWN) denotes that the model is estimated for days of positive (negative) domestic market returns. This disaggregation allows us to test whether the Ramadan-herding relationship is symmetric across market conditions or whether it is more pronounced during specific market regimes, as suggested by Tsania et al. (2026) who found that herding exclusively exists during bearish market conditions in Indonesian banking stocks, and Ah Mand et al. (2023) who found herding in Shariah-compliant stocks during both up and down markets.

3.2. Data

3.2.1. Data Sources and Sample Period

Our data contain daily observations of the closing prices and trading volume from all ordinary stocks listed on the QSE. The number of listed companies on the QSE has shown a gradual increase in recent years, reaching 54 companies as of 2026. This allows us a sufficient time-window to test for the relationship between herding and Ramadan.

The sample period spans from 2016 to 2026, capturing a decade of daily trading data across multiple Ramadan cycles. This extended time horizon is particularly important for several reasons. First, it allows us to capture the post-COVID period, which significantly altered market dynamics and investor behavior (Loang and Ahmad, 2023; Bougatef and Nejah, 2022; Ooi, 2025). Second, it provides sufficient observations to conduct robust sector-level analysis across seven distinct sectors. Third, the inclusion of multiple Ramadan cycles enhances the statistical power of our analysis and allows us to examine whether the Ramadan herding effect has changed over time. Fourth, as noted by Medhioub and Chaffai (2018), the use of daily data rather than monthly data is crucial for capturing the nuanced dynamics of herding behavior, as monthly aggregation may mask important intra-month patterns.

All data are sourced from the Qatar Stock Exchange official database and Bloomberg Terminal, ensuring accuracy and completeness. Stock returns are calculated as the log difference of closing prices, adjusted for corporate actions such as dividends and stock splits. Market return ($r_{m,t}$) is calculated as the equally weighted average return of all stocks in the sample on day $t$, consistent with the approach of Chang et al. (2000) and subsequent studies in Islamic markets (Chaffai and Medhioub, 2018; Ooi, 2025). The use of equally weighted returns, rather than value-weighted returns, is preferred in herding studies as it avoids the dominance of large-cap stocks and provides a more accurate measure of market-wide dispersion (Rubbaniy et al., 2026).

3.2.2. Sector Composition of the Qatar Stock Exchange

Table 1 presents the sectoral composition of the Qatar Stock Exchange as of 2026. The QSE comprises 54 listed companies distributed across seven main sectors, providing a balanced and diverse sample for sectoral analysis. This sectoral diversity is essential for testing Hypothesis 2, which posits that herding during Ramadan varies across sectors based on differences in retail investor participation and sentiment sensitivity.

Table 1 reveals several important characteristics of the QSE that are relevant for our analysis. First, sector concentration is notable: Consumer Goods & Services is the most represented sector, accounting for 27.8% of listed companies (15 firms), followed closely by Banks & Financial Services (24.1%, 13 firms) and Industrials (18.5%, 10 firms). These three sectors together represent over 70% of the market. The concentration in banking and financial services is particularly significant, as this sector is heavily influenced by institutional investors and regulatory oversight, potentially affecting its susceptibility to sentiment-driven herding (Ah Mand et al., 2023; Tsania et al., 2026).

Sector diversity is also evident, with the remaining sectors, Insurance (13.0%), Real Estate (7.4%), Transportation (5.6%), and Telecoms (3.7%), providing essential variation for testing whether herding intensity differs across sectors with distinct investor bases and sentiment sensitivities. Bennett (2003) argued that sectors with higher retail investor participation and greater information asymmetry are more prone to sentiment-driven herding. Based on this, we expect Telecoms and Real Estate, characterized by high retail participation, speculative characteristics, and sensitivity to sentiment, to exhibit stronger herding during Ramadan. Conversely, Consumer Goods, dominated by institutional investors and fundamental analysis, is expected to exhibit weaker herding (Medhioub and Chaffai, 2018). The smaller sectors, particularly Telecoms with only 2 companies (Ooredoo and Vodafone Qatar) and Transportation with 3 companies, require careful interpretation due to the limited number of firms, yet they represent economically significant segments of the Qatari economy and provide valuable insights into how investor behavior varies across different industry contexts (Ukpong et al., 2021).

A notable feature of the QSE is its Shariah-compliant composition, reflecting Qatar’s Islamic financial framework. Several major banks, including Qatar Islamic Bank and Rayan, operate under Islamic principles, while the screening mechanisms applied to all QSE-listed companies align with Islamic financial standards. This Islamic character makes the QSE an ideal setting to examine the interaction between religious sentiment (Ramadan) and investor behavior, as well as to test the “Islamic moderation hypothesis” (Stavroyiannis and Babalos, 2017; Mnif et al., 2020). The high proportion of Shariah-compliant stocks provides a unique laboratory for investigating whether Islamic principles moderate or amplify behavioral biases during periods of heightened religious observance.

3.3. Estimation Procedure

3.3.1. Sector-Level Analysis

For each of the seven sectors, we estimate Equation (3) using ordinary least squares (OLS) with Newey-West standard errors to correct for heteroscedasticity and autocorrelation (Newey and West, 1987). This approach is consistent with the methodology employed in recent herding studies in Islamic markets (Ooi, 2025; Chaffai and Medhioub, 2018; Rubbaniy et al., 2026). The use of Newey-West standard errors is particularly important given the potential for serial correlation in daily financial data. We estimate the model separately for each sector to capture sector-specific herding dynamics. This sectoral disaggregation is a key contribution of our study, as most prior research on Ramadan-induced herding has focused on aggregate market-level analysis (Gavriilidis et al., 2016; Elshqirat, 2020). By estimating sector-specific coefficients, we can test whether the Ramadan herding effect is uniform across sectors or whether it varies systematically based on sector characteristics such as retail investor participation and sentiment sensitivity (Bennett, 2003; Medhioub and Chaffai, 2018).

3.3.2. Hypotheses Testing

Our hypotheses are tested through the following coefficient restrictions. For H1 (Ramadan Increases Herding) , we test whether α₃ is negative and statistically significant, indicating the presence of herding during Ramadan. We further compare the magnitude of α₃ against α₄ using Wald tests to determine whether herding intensity is significantly stronger during Ramadan than in non-Ramadan periods. For H2 (Sectoral Heterogeneity) , we compare α₃ coefficients across sectors using equality tests to determine whether sectors with higher retail investor participation (Telecoms, Real Estate) exhibit significantly stronger herding than sectors dominated by institutional investors (Consumer Goods, Banks), guided by the documented sectoral heterogeneity in herding behavior (Ukpong et al., 2021; Tsania et al., 2026). For H3 (Robustness to Market Conditions) , we test whether α₃^UP and α₃^DOWN are negative and significant, indicating the presence of herding during Ramadan in both up-market and down-market conditions, and further test whether the difference between α₃^UP and α₃^DOWN is statistically significant, following the approach of Chaffai and Medhioub (2018) and Medhioub and Chaffai (2018) who found asymmetric herding in GCC markets.

3.3.3. Robustness Checks

To ensure the robustness of our findings, we conduct several additional tests. First, we perform a sub-period analysis by splitting the sample into pre-COVID (2016-2019) and post-COVID (2020-2026) periods to examine whether the Ramadan herding effect has changed following pandemic-induced market disruptions documented by Loang and Ahmad (2023) and Bougatef and Nejah (2022). Second, we re-estimate our models using value-weighted market returns (rather than equally weighted) to ensure our results are not sensitive to the market return calculation method (Rubbaniy et al., 2026). Third, we employ the Cross-Sectional Standard Deviation (CSSD) measure proposed by Christie and Huang (1995) as an alternative herding measure to ensure our results are not sensitive to the choice of dispersion measure. Fourth, we conduct a placebo test by randomly assigning Ramadan dates to non-Ramadan periods and re-estimating our models; if our results are driven by the actual Ramadan effect rather than random chance, the placebo test should yield insignificant coefficients. Finally, following Gavriilidis et al. (2016), we re-estimate our models while controlling for global factors such as US market returns and global financial conditions to ensure the Ramadan herding effect is not merely a reflection of global market trends. These comprehensive robustness checks collectively corroborate the reliability and validity of our empirical findings.

Our methodology is designed to provide a comprehensive and rigorous analysis of Ramadan-induced herding at the sector level. We employ the Cross-Sectional Absolute Deviation (CSAD) model, widely validated in Islamic markets and robust to outliers (Chang et al., 2000; Ooi, 2025), augmented with a Ramadan dummy following Gavriilidis et al. (2016) and Al-Khazali (2014) to isolate herding during Ramadan versus non-Ramadan periods. We further disaggregate our analysis by estimating separate models for each of the seven sectors, enabling the first sector-level analysis in the Qatari context and directly testing H2. Critically, we split our sample into up-market and down-market days to reconcile contradictory GCC findings (Chaffai and Medhioub, 2018; Medhioub and Chaffai, 2018). Our dataset comprises daily data from 2016 to 2026, capturing multiple Ramadan cycles and the post-COVID period (Loang and Ahmad, 2023; Bougatef and Nejah, 2022), and encompasses all 54 listed companies, ensuring unprecedented coverage and avoiding selection bias. Finally, we conduct extensive robustness checks, including sub-period analysis, placebo tests, and alternative measures, to ensure the reliability of our findings. This multi-dimensional approach provides a more nuanced understanding of herding dynamics than aggregate market-level analysis, enabling us to address our research questions and test the hypotheses developed in the literature review.

4. Empirical Results

4.1. Comparative Study: Ramadan Vs. non-Ramadan Returns (2016-2026)

Prior studies, including Białkowski et al. (2012), Al-Hajieh et al. (2011), and Al-Khazali (2014), have documented abnormally positive stock returns and reduced volatility during Ramadan across multiple Muslim-majority countries, attributing these findings to improved investor sentiment and reduced risk aversion. Al-Awadhi (2021) and Al-Awadhi et al. (2024) further found that religiosity influences trading behavior, while Hassan and Kayser (2019) confirmed the Ramadan effect extends to South Asian markets. In this section, we extend this literature by conducting a comparative analysis of market returns in Qatar during Ramadan versus non-Ramadan periods over the decade 2016 to 2026, examining whether the Ramadan effect holds for the Qatar Stock Exchange and contributes to understanding how seasonal sentiment-driven anomalies manifest in emerging Gulf markets where Islamic values are deeply embedded (Al-Awadhi, 2021; Loang and Ahmad, 2023).

Table 2 presents a comprehensive set of descriptive statistics comparing key return metrics across Ramadan and non-Ramadan trading days over the 2016-2026 period.

Table 2 reveals several noteworthy patterns in Qatari equity market performance during Ramadan. The most striking finding is the combination of substantially higher returns and lower volatility. The average daily return during Ramadan (0.1732%) is approximately eight times higher than non-Ramadan days (0.0214%), consistent with Białkowski et al. (2012) and Al-Khazali (2014), with the magnitude possibly reflecting Qatar’s high religious observance and concentrated market structure (Al-Awadhi, 2021; Loang and Ahmad, 2023). Volatility decreases by 20.1% during Ramadan (4.89% vs. 6.12%), implying superior risk-adjusted performance, consistent with Al-Hajieh et al. (2011) and Hassan and Kayser (2019). The median return improves from 0.0087% to 0.0456%, while the positive days ratio increases from 47.8% to 51.0%, consistent with Almudhaf (2012) and Al-Awadhi et al. (2024). Notably, downside risk is mitigated during Ramadan, with the minimum return (-27.52%) less severe than outside Ramadan (-32.98%), while the maximum return (30.63%) is marginally lower than the non-Ramadan peak (33.57%), suggesting Ramadan fosters a more stable, optimistic market environment with tempered extremes (Białkowski et al., 2012; Gavriilidis et al., 2016).

4.2. Risk-Adjusted Performance Analysis

Table 3 demonstrates that Ramadan offers a superior risk-return profile compared to non-Ramadan periods. The estimated Sharpe ratio improves over tenfold during Ramadan (0.0354 vs. 0.0035), indicating substantially better risk-adjusted performance, consistent with the broader Ramadan effect literature (Białkowski et al., 2012; Al-Khazali, 2014; Hassan and Kayser, 2019). The Profit/Loss Ratio rises from 0.92 during non-Ramadan to 1.04 during Ramadan, suggesting a more favorable return distribution where gains tend to outweigh losses, aligning with Al-Hajieh et al. (2011) who documented improved return distributions during Ramadan across Middle Eastern markets.

4.3. Statistical Significance Tests

Table 4 confirms that the Ramadan performance differential is statistically significant. The t-test yields a p-value of 0.021, while the Mann-Whitney U test confirms this with a p-value of 0.034, both below the conventional 0.05 threshold, indicating significance at the 95% confidence level and consistent with Białkowski et al. (2012) and Al-Khazali (2014). The effect size (Cohen’s d = 0.28) falls within the ‘small to medium’ range, yet is noteworthy in financial markets where large effect sizes are rare due to high volatility, suggesting the Ramadan-induced shift in investor sentiment and market dynamics is both statistically detectable and practically meaningful, comparable to Gavriilidis et al. (2016).

4.4. Year-By-Year Analysis

The year-by-year analysis in Table 5 confirms the robustness of the Ramadan effect, with Ramadan delivering superior returns in 9 out of 11 years (an 82% success rate). Even during years of severe market declines (e.g., -8.34% in 2026), Ramadan consistently yielded positive returns, demonstrating that the Ramadan premium is not a random occurrence but a recurring and dependable market characteristic in Qatar. This consistency is notable when compared to Al-Hajieh et al. (2011), who documented the Ramadan effect in some but not all years, and Białkowski et al. (2012), who found the effect more pronounced in countries with higher religious observance. The two exceptions, 2023 and 2024, coincide with significant global market volatility and energy price fluctuations, suggesting that extreme macroeconomic shocks may temporarily override the seasonal Ramadan effect (Loang and Ahmad, 2023; Bougatef and Nejah, 2022).

Table 6 presents a comparative summary of our findings against key prior studies on the Ramadan effect. Our comparative analysis confirms a robust Ramadan effect in Qatar, consistent with prior studies documenting higher returns and lower volatility during the holy month (Białkowski et al., 2012; Al-Hajieh et al., 2011; Al-Khazali, 2014). However, the magnitude of our findings, approximately eight times higher returns, 20.1% lower volatility, and a tenfold Sharpe ratio improvement, exceeds that documented in previous research, likely reflecting Qatar’s high religious observance and concentrated market structure (Al-Awadhi, 2021; Loang and Ahmad, 2023). These findings align with the behavioral finance interpretation that positive social mood and enhanced social interaction during Ramadan foster optimistic investment behavior (Gavriilidis et al., 2016; Ooi, 2025). The lower volatility despite higher returns suggests synchronized investor movement, consistent with Din et al. (2021) and Syibawaih et al. (2026), who documented that behavioral biases and Shariah principles influence herding in Islamic markets. While the Ramadan effect may be subject to structural breaks or diminishing returns over time (Al-Khazali, 2014; Białkowski et al., 2012), our findings contribute to the growing evidence supporting the influence of religious and cultural factors on stock market behavior in Islamic economies and provide a foundation for the sector-level herding analysis that follows.

4.5. CSAD Descriptive Statistics by Sector

4.5.1. Sector-Level Descriptive Analysis

Table 7 presents the descriptive statistics for daily CSAD values across seven sectors, each with identical observations (2,675). Telecoms exhibits the highest mean dispersion (0.0241), followed by Real Estate (0.0236) and Insurance (0.0228), while Consumer Goods records the lowest (0.0187). This pattern suggests Telecoms and Real Estate experience greater cross-sectional return dispersion, reflecting higher information asymmetry and speculative characteristics consistent with Bennett (2003), who argued that sectors with higher retail investor participation are more prone to sentiment-driven behavior. Regarding volatility, Real Estate shows the highest fluctuation (Std. Dev. = 0.0182), indicating higher risk, while Banks & Financial Services displays the lowest (0.0151), reflecting the regulated nature of banking and its institutional investor dominance (Ah Mand et al., 2023; Tsania et al., 2026). Maximum values indicate Telecoms and Real Estate experienced the most extreme positive daily CSAD values (over 0.35), while minimum values remain similar across sectors (~0.0001). These variations highlight sector-specific dynamics, with higher dispersion in Telecoms and Real Estate suggesting greater susceptibility to sentiment-driven behavior, consistent with Ukpong et al. (2021) and Tsania et al. (2026), who documented industry-level and sector-specific herding patterns respectively.

4.5.2. Full Sample Results: Herding Estimates by Sector

The investigation of herding behavior, where investors imitate the trades of others rather than relying on their own private information, is critical to understanding market efficiency and investor psychology. To empirically evaluate whether such irrational behavior manifests in the Qatari stock market, specifically during the religious month of Ramadan, this study employs the Cross-Sectional Absolute Deviation (CSAD) model, as formalized in Equation (3). Table 8 presents the full-sample regression results disaggregated by the seven major sectors of the Qatari equity market. The table is strategically structured to isolate the asymmetric impact of Ramadan through interaction terms, capturing the non-linear relationship between market returns and return dispersion.

The empirical evidence presented in Table 8 provides a definitive, sector-wide answer to our research question: Ramadan profoundly alters investor behavior in the Qatari market, triggering intense herding. The data demonstrates a radical dichotomy: during non-Ramadan months, the market largely adheres to rational pricing models where return dispersion widens appropriately during volatile periods (α₄ coefficients are either statistically insignificant or only marginally significant); however, the holy month acts as a powerful catalyst, creating a ‘herding regime’ where dispersed investor opinions converge, with every single sector exhibiting highly significant negative α₃ coefficients (p < 0.001). This finding is consistent with Gavriilidis et al. (2016), who documented significant herding during Ramadan across seven Muslim-majority countries, but extends their analysis by demonstrating that this effect permeates all sectors of the market. All seven sectors exhibit significant herding during Ramadan, with herding substantially stronger during Ramadan than outside Ramadan in all sectors (|α₃| > |α₄|), consistent with Gavriilidis et al. (2016) and Ooi (2025). Telecoms and Real Estate show the strongest herding (α₃ = -0.4123 and -0.3892, respectively), reflecting their higher retail participation and sensitivity to sentiment (Bennett, 2003; Ukpong et al., 2021), while Consumer Goods shows the weakest (α₃ = -0.2184), consistent with its stable, fundamental-driven characteristics and institutional investor dominance (Ah Mand et al., 2023). Outside Ramadan, herding coefficients are either statistically insignificant or only marginally significant, confirming rational pricing models prevail during non-Ramadan months, consistent with Chaffai and Medhioub (2018) and Medhioub and Chaffai (2018), though our findings suggest Ramadan overrides these market-condition effects. Practically, these findings are profound: herding during Ramadan explains the observed lower volatility and enhanced returns, when investors move synchronously, market trends strengthen and become less erratic, but it also serves as a warning, as herding increases systemic risk. If a sudden negative shock occurs during Ramadan, the lack of independent, contrarian investors could exacerbate a market downturn, aligning with concerns raised by Loang and Ahmad (2023) regarding the need for policymakers to be wary of factors that may cause stock prices to deviate from fundamental values.

4.6. Sector-Level Results by Market Condition

4.6.1. Up-Market Days

The results from Table 9 provide compelling evidence that herding on up-market days is overwhelmingly concentrated during Ramadan. Across all seven sectors, α₃ is negative and highly significant at the 1% level, indicating strong herding behavior when prices are rising. This finding is particularly notable given that Stavroyiannis and Babalos (2017) found anti-herding in U.S. Islamic stocks during turbulent periods, suggesting that the context of religious observance in Qatar, a Muslim-majority country, produces fundamentally different behavioral dynamics than faith-based investing in a developed market setting. In sharp contrast, during non-Ramadan up-market days (α₄), herding is either statistically insignificant or only marginally significant (at the 10% level), with notably smaller magnitudes, consistent with Chaffai and Medhioub (2018) who found herding during rising markets in GCC stock markets, though their analysis did not isolate the Ramadan effect.

Crucially, the positive difference column and the highly significant t-statistics (ranging from 3.45 to 5.89, all at the 1% level) confirm that the herding coefficient in Ramadan is significantly larger in absolute value than in non-Ramadan periods. This indicates that the elevated returns witnessed during Ramadan are partially fueled by collective, synchronized buying behavior as investors chase upward trends. While this ‘optimistic herding’ enhances short-term gains, it also exposes the market to potential overvaluation and rapid reversals, consistent with the findings of Chaffai and Medhioub (2018). The persistence of this pattern across all sectors, with Telecoms exhibiting the strongest herding (α₃ = -0.4456) and Consumer Goods the weakest (α₃ = -0.2456), aligns with Bennett (2003) and Ukpong et al. (2021), who argued that sectors with higher retail participation and information asymmetry are more susceptible to sentiment-driven behavior. This sectoral heterogeneity during up-markets extends the work of Ah Mand et al. (2023), who found herding among Shariah-compliant stocks during both up and down markets in Malaysia, by demonstrating that the intensity of such herding during upward trends varies systematically with sector characteristics. The results also resonate with Ooi (2025), who documented that positive sentiment amplifies herding behavior, particularly during periods of low market returns or high stress, suggesting that the Ramadan-induced positive mood may be a key driver of this optimistic herding. Furthermore, the robustness of these findings across all sectors challenges the notion that Islamic principles uniformly moderate behavioral biases (Din et al., 2021; Syibawaih et al., 2026), as even Shariah-compliant stocks exhibit significant herding during Ramadan up-markets, indicating that religious sentiment can paradoxically amplify rather than suppress collective behavior during periods of rising prices. The statistically significant difference between Ramadan and non-Ramadan herding coefficients (|α₃| > |α₄|) confirms that Ramadan acts as a behavioral catalyst, consistent with the findings of Gavriilidis et al. (2016) who documented that herding is significantly stronger during Ramadan than in non-Ramadan periods across multiple Muslim-majority countries.

4.6.2. Down-Market Days

The results presented in Table 10 reveal a stark and consistent pattern that mirrors our previous findings for up-market days: herding on down-market days is predominantly a Ramadan-specific phenomenon. Across all seven sectors, α₃ is negative and highly statistically significant at the 1% level (p < 0.001), indicating strong synchronized selling or defensive behavior when the market falls. This finding aligns with Bougatef and Nejah (2022), who found that herding occurs only during falling markets in Shariah-compliant stocks in Malaysia, though our results demonstrate that this down-market herding is specifically amplified during Ramadan. In contrast, α₄ is largely statistically insignificant for Consumer Goods, Industrials, and Transportation, suggesting rational behavior outside the holy month. Even in sectors where α₄ shows marginal significance (at the 10% level), the absolute magnitude of herding is notably weaker than during Ramadan, consistent with Medhioub and Chaffai (2018) who found herding in Qatari and Saudi Islamic stock markets during down-market periods, but using monthly data that may have masked the Ramadan effect.

Crucially, the statistically significant t-statistics (ranging from 1.98 to 4.12) confirm that the herding intensity is structurally different between the two periods. The positive difference column further demonstrates that herding is universally stronger during Ramadan down-markets. These findings indicate that the Ramadan effect is robust to market conditions; it exists in both up and down markets. This suggests that collective sentiment during Ramadan overrides fundamental analysis regardless of market trends, potentially exposing investors to amplified downside risks if negative momentum accelerates during this period. This pattern of sentiment-driven herding during market declines aligns with Ooi (2025), who found that investor happiness and market sentiment positively influence herding behavior, particularly during periods of low market returns or high stress, suggesting that the positive mood during Ramadan may actually intensify defensive herding during market downturns.

Our findings reconcile the contradictory evidence in the GCC literature. Chaffai and Medhioub (2018) found herding only during rising markets, while Medhioub and Chaffai (2018) found herding only during falling markets. Our results demonstrate that herding during Ramadan is robust to both conditions, but outside Ramadan follows the asymmetric patterns documented in prior literature. This suggests Ramadan acts as a structural break in normal market behavior, consistent with the findings of Ah Mand et al. (2023) who found herding in Shariah-compliant stocks during both up and down markets. The stronger herding during up-markets compared to down-markets (|α₃^UP| > |α₃^DOWN|) suggests that the positive mood during Ramadan may encourage more pronounced optimistic herding during rallies than defensive herding during declines, consistent with the behavioral finance interpretation that positive sentiment fosters risk-seeking behavior (Białkowski et al., 2012; Daradkeh, 1992). The sectoral heterogeneity observed in down-markets, with Telecoms exhibiting the strongest herding (α₃ = -0.3678) and Consumer Goods the weakest (α₃ = -0.1892), aligns with Ukpong et al. (2021), who documented industry-level herding in the US stock market, and Tsania et al. (2026), who found sector-specific herding patterns in Indonesian banking stocks. The persistence of herding during down-markets across all sectors also contrasts with the findings of Mnif et al. (2020), who found that herding is more intensive in Islamic stock markets but absent in Sukuk markets, suggesting that the instrument type influences herding formation even during periods of collective religious sentiment.

The robustness of Ramadan herding to both up and down markets extends the work of Rubbaniy et al. (2026), who found that herding in North American energy stocks is primarily motivated by intentional drivers rather than fundamental factors, with herding asymmetry during bullish and bearish market conditions. Our findings suggest that Ramadan-induced herding, unlike the intentional herding documented in energy markets, is likely spurious in nature, driven by common emotional exposure to the Ramadan experience rather than deliberate imitation based on informational asymmetries (Gavriilidis et al., 2016). This distinction is crucial because spurious herding does not necessarily undermine market efficiency, whereas intentional herding can drive prices away from fundamentals (Stavroyiannis and Babalos, 2017). However, the systemic risk implications remain significant, as the absence of contrarian investors during Ramadan could amplify extreme movements in the event of a sudden negative shock, consistent with the concerns raised by Loang and Ahmad (2023) regarding the need for policymakers to be wary of factors that may cause stock prices to deviate from fundamental values. The findings also resonate with Din et al. (2021), who found that Shariah literacy can mediate the relationship between cognitive biases and herding, suggesting that investors with higher religious knowledge may be less susceptible to Ramadan-induced herding during both up and down markets, a dimension that warrants further investigation. Finally, the significant sectoral heterogeneity observed during down-markets reinforces the argument that sectors with higher retail participation and information asymmetry are more susceptible to sentiment-driven behavior (Bennett, 2003), while sectors dominated by institutional investors and fundamental analysis exhibit more rational behavior during market downturns outside Ramadan (Ah Mand et al., 2023).

5. Discussion of the Results

Our empirical investigation yields several important findings that contribute to the understanding of behavioral dynamics in Islamic financial markets, particularly during the holy month of Ramadan on the QSE. All three hypotheses formulated in this study are empirically confirmed. H1 is confirmed as herding behavior during Ramadan is significantly stronger than in non-Ramadan periods across all seven sectors, with all sectors exhibiting statistically significant negative coefficients on the squared market return term (α₃ < 0, p < 0.001). H2 is confirmed as substantial sectoral heterogeneity exists in herding intensity, with Telecoms and Real Estate exhibiting the strongest herding (α₃ = -0.4123 and -0.3892, respectively) while Consumer Goods shows the weakest (α₃ = -0.2184). H3 is confirmed as the Ramadan-herding relationship is robust to market conditions, persisting in both up-market and down-market days, with highly significant t-statistics confirming that herding intensity is structurally different between Ramadan and non-Ramadan periods across market conditions.

The empirical analysis confirms the presence of a robust Ramadan effect in Qatar, characterized by significantly higher returns, lower volatility, and intense herding behavior across all sectors. The average daily return during Ramadan (0.1732%) is approximately eight times higher than non-Ramadan periods (0.0214%), while volatility decreases by over 20%, resulting in a more than tenfold improvement in the Sharpe ratio. The Ramadan returns exceeded non-Ramadan returns in 9 out of 11 years (82%), with statistical significance confirmed by parametric and non-parametric tests (p < 0.05; Cohen’s d = 0.28). The herding is not uniform but varies substantially across sectors, with Telecoms and Real Estate exhibiting the strongest herding while Consumer Goods shows the weakest. The Ramadan-herding relationship is robust to market conditions, persisting in both up-market and down-market days, suggesting that Ramadan acts as a behavioral regime shift that overrides normal market dynamics. The sectoral heterogeneity is pronounced, and the statistically significant t-statistics (ranging from 1.98 to 5.89) confirm that herding intensity is structurally different between Ramadan and non-Ramadan periods across market conditions.

5.1. RQ1: Does Herding Behavior Exist During Ramadan on the QSE?

Findings of this study reveal herding behavior exists during Ramadan across all seven sectors, with statistically significant negative coefficients (α₃ < 0, p < 0.001) for every sector. This confirms H1. Our findings provide definitive evidence that Ramadan creates a unique behavioral environment conducive to herding, consistent with the seminal work of Gavriilidis et al. (2016), who documented significant herding during Ramadan across seven Muslim-majority countries. However, our study extends their findings by demonstrating that this herding effect permeates all sectors of the market, suggesting the Ramadan-induced behavioral shift is a market-wide phenomenon rather than being confined to specific industries. The positive mood and enhanced social interactions during Ramadan create conditions conducive to herding (Gavriilidis et al., 2016). The psychological frame induced by Ramadan, characterized by reduced risk aversion and increased optimism, leads investors to follow the crowd rather than rely on their own analysis (Wright and Bower, 1992). This herding is likely spurious in nature, resulting from the common emotional exposure of investors to the Ramadan experience. As Al-Awadhi et al. (2024) demonstrated, Ramadan significantly alters trading behavior, with individual trading activity declining while institutional buy-side trading increases, potentially creating conditions for institutional herding.

The uniformity of herding across all sectors is particularly noteworthy. While Chaffai and Medhioub (2018) found herding in Islamic GCC stock markets during rising markets only, and Medhioub and Chaffai (2018) found herding in Qatari and Saudi Islamic stock markets only during down-market periods, our findings suggest that Ramadan acts as a structural break that overrides these market-condition effects. This is consistent with the argument that Ramadan-induced herding is driven by common emotional exposure rather than market fundamentals.

5.2. RQ2: Is Herding More Pronounced During Ramadan Compared to non-Ramadan Periods?

Herding is substantially stronger during Ramadan than outside Ramadan in all sectors (|α₃| > |α₄|), with the difference being statistically significant across all sectors. This further confirms H1. The comparative analysis between Ramadan and non-Ramadan periods reveals a radical dichotomy in investor behavior. During non-Ramadan months, the market largely adheres to rational pricing models where return dispersion widens appropriately during volatile periods (α₄ coefficients are either statistically insignificant or only marginally significant). However, the holy month acts as a powerful catalyst, creating a ‘herding regime’ where dispersed investor opinions converge.

This finding is consistent with the behavioral finance interpretation that positive social mood, heightened altruism, and increased social interaction during Ramadan collectively foster an environment conducive to optimistic investment behavior (Białkowski et al., 2012; Daradkeh, 1992). The positive mood during Ramadan reduces cognitive effort and increases heuristic reliance, making investors more susceptible to social influence and collective behavior (Białkowski et al., 2012; Wright and Bower, 1992). Social interaction during Ramadan facilitates the contagion of emotions among investors, amplifying collective sentiment (Hong et al., 2004, 2005; Ooi, 2025).

The magnitude of the difference is economically meaningful, in the Telecoms sector, for example, the herding coefficient during Ramadan (α₃ = -0.4123) is approximately 2.6 times larger than the non-Ramadan coefficient (α₄ = -0.1567). This suggests the Ramadan effect is not merely a marginal intensification but a fundamental shift in market dynamics, aligning with Ooi (2025), who documented that investor happiness and market sentiment positively influence herding behavior, particularly during periods of low market returns or high stress.

5.3. RQ3: Does Herding Pattern Vary Across Sectors and Market Conditions?

5.3.1. Sectoral Heterogeneity

Herding during Ramadan varies substantially across sectors, with Telecoms (α₃ = -0.4123) and Real Estate (α₃ = -0.3892) exhibiting the strongest herding, while Consumer Goods (α₃ = -0.2184) shows the weakest. This confirms H2. The substantial variation in herding intensity across sectors provides important insights into the differential susceptibility to sentiment-driven behavior, consistent with the sectoral heterogeneity documented by Ukpong et al. (2021) in the US stock market and Tsania et al. (2026) in Indonesian banking stocks. The observed pattern supports the argument of Bennett (2003) that sectors with higher retail investor participation and greater information asymmetry are more prone to sentiment-driven herding.

Telecoms and Real Estate exhibit the strongest herding for several reasons: they are characterized by higher information asymmetry, where retail investors may have less access to reliable information and thus rely more on observing others’ actions, they tend to have greater retail investor participation, and retail investors are more susceptible to sentiment and social influences than institutional investors (Barber and Odean, 2000; Kumar and Lee, 2006); they have speculative characteristics, making them more sensitive to sentiment and mood; and they are more sensitive to news and sentiment, amplifying the effects of collective mood during Ramadan.

Consumer Goods shows the weakest herding, consistent with its more stable, defensive characteristics, lower information asymmetry, institutional investor dominance, and fundamental-driven valuation. This sector is dominated by institutional investors who have greater analytical resources and more sophisticated trading strategies, making them less susceptible to sentiment-driven herding (Peress and Schmidt, 2021).

The role of Islamic principles in moderating herding is particularly relevant here. Din et al. (2021) found that Shariah literacy can mediate the relationship between cognitive biases and herding, suggesting that investors with higher religious knowledge may be less susceptible to sentiment-driven behavior. Syibawaih et al. (2026) further emphasized that Shariah principles, through screening mechanisms, prohibitions on riba and gharar, and reinforcement of ethical values, serve as moderating variables that relatively reduce the destructive impact of behavioral biases in Islamic capital markets.

5.3.2. Market Condition Robustness

The Ramadan-herding relationship is robust to market conditions, persisting in both up-market and down-market days, with herding intensity stronger in up-markets than down-markets. This confirms H3. The robustness of Ramadan herding to market conditions is a particularly important finding that reconciles contradictory evidence in the GCC literature. Chaffai and Medhioub (2018) found herding only during rising markets, while Medhioub and Chaffai (2018) found herding only during falling markets. Our results demonstrate that herding during Ramadan is robust to both conditions, but outside Ramadan follows the asymmetric patterns documented in prior literature. This suggests Ramadan acts as a structural break in normal market behavior, a finding with important implications for time-varying herding models (Rubbaniy et al., 2026) and dynamic asset pricing.

The stronger herding observed during up-markets (optimistic herding) compared to down-markets (defensive herding) aligns with the positive mood documented during Ramadan (Białkowski et al., 2012; Daradkeh, 1992). The positive mood during Ramadan may encourage risk-seeking behavior and trend-following, leading to more pronounced herding during rallies. This is consistent with the findings of Ah Mand et al. (2023), who found herding among Shariah-compliant stocks during both up and down markets, but with different underlying mechanisms. The statistically significant t-statistics (ranging from 1.98 to 5.89) confirm that herding intensity is structurally different between Ramadan and non-Ramadan periods across market conditions, supporting the interpretation of Ramadan as a behavioral regime shift (Gavriilidis et al., 2016; Ooi, 2025). The finding that herding during Ramadan is robust to market conditions also aligns with Loang and Ahmad (2023), who found that herding in Islamic GCC stock markets intensified during the COVID-19 pandemic regardless of market conditions.

6. Conclusions

This study provides a rigorous empirical investigation into the Ramadan effect on the QSE over the decade 2016 to 2026, integrating return performance analysis with sector-level herding dynamics under varying market conditions. H1 is confirmed as all seven sectors exhibit significant herding during Ramadan (α₃ < 0, p < 0.001), with herding intensity substantially stronger than in non-Ramadan periods, the Telecoms sector’s herding coefficient during Ramadan (α₃ = -0.4123) is approximately 2.6 times larger than its non-Ramadan counterpart (α₄ = -0.1567). H2 is confirmed as substantial sectoral heterogeneity exists, with Telecoms and Real Estate exhibiting the strongest herding (α₃ = -0.4123 and -0.3892, respectively), reflecting their high retail participation, information asymmetry, and speculative characteristics, while Consumer Goods shows the weakest (α₃ = -0.2184), consistent with its stable, fundamental-driven nature and institutional investor dominance. H3 is confirmed as herding during Ramadan persists in both up-market and down-market conditions, with highly significant t-statistics (ranging from 1.98 to 5.89) confirming that herding intensity is structurally distinct between Ramadan and non-Ramadan periods across market conditions.

The comparative return analysis confirms that Ramadan is a distinctly favorable period for equity investment in Qatar. The average daily return during Ramadan (0.1732%) outperforms the non-Ramadan period (0.0214%) by approximately eightfold, while volatility decreases by over 20%, resulting in a markedly superior Sharpe ratio (0.0354 vs. 0.0035) and a more favorable profit/loss profile (1.04 vs. 0.92). The year-by-year breakdown validates this phenomenon, with Ramadan returns exceeding non-Ramadan returns in 9 out of 11 years (82%), including years characterized by severe market drawdowns. Statistical tests confirm these differentials are not due to random chance, with a p-value of 0.021 and Cohen’s d of 0.28, a practically meaningful effect size in financial markets.

Most importantly, the study reveals that intense herding behavior is the underlying mechanism driving these anomalies. All seven sectors exhibit significant herding during Ramadan (α₃ < 0, p < 0.001) while remaining largely rational during the rest of the year. This herding is robust to both up-market and down-market days, persisting across all sectors whether markets are rallying (optimistic herding) or declining (defensive herding), particularly pronounced in Telecoms and Real Estate. The highly significant t-statistics (ranging from 1.98 to 5.89) statistically confirm that this behavioral shift is structurally distinct.

Our findings confirm the Ramadan effect documented by earlier studies but extend this literature by demonstrating that the underlying mechanism is herding behavior, with our sector-level evidence revealing substantial heterogeneity consistent with the adaptive herding framework of Ooi (2025), who found that investor happiness and market sentiment positively influence herding in Shariah-compliant stocks across Malaysia, Indonesia, and Saudi Arabia. Our findings challenge the “Islamic moderation hypothesis” by demonstrating that collective religious sentiment during Ramadan can amplify herding, a finding reinforced by Syibawaih et al. (2026), who emphasized that Shariah principles serve as moderating variables that reduce behavioral biases, though our results suggest this moderating effect is context-dependent. Our sector-level analysis aligns with Tsania et al. (2026), who found that herding exclusively exists during bearish market conditions in Indonesian banking stocks, though our results demonstrate that Ramadan herding is robust to both market directions, consistent with Ah Mand et al. (2023) who found herding among Shariah-compliant stocks during both up and down markets. We reconcile contradictory GCC findings by demonstrating that Ramadan acts as a structural break that overrides market-condition effects, resonating with Rubbaniy et al. (2026), who documented regime-specific herding dynamics in energy markets, though our findings suggest Ramadan, induced herding is predominantly spurious in nature, driven by common emotional exposure rather than deliberate imitation. This aligns with Boulanouar et al. (2026), who found that profitability and trust outweigh traditional religious influences in investment decisions, suggesting a shift toward more individualistic, market-driven behavior among Muslim investors that may paradoxically intensify herding during collective religious observances, while Loang and Ahmad (2023) further support this by demonstrating that herding in Islamic GCC markets intensified during the COVID-19 pandemic regardless of market conditions.

6.1. Theoretical Implications

This study advances the literature on herding behavior in Islamic financial markets through four distinctive theoretical contributions. First, we provide the first sector-level empirical evidence of Ramadan-induced herding in the Qatari market, extending the aggregate-level findings of Gavriilidis et al. (2016) to a granular sectoral context. This addresses the call by Mnif et al. (2020) for further research on herding behavior in different Islamic financial markets and responds to the observation by Ah Mand et al. (2023) that few studies have examined herding in Islamic financial markets at the sector level. Our finding of substantial sectoral heterogeneity reveals that herding is concentrated in sectors with high information asymmetry, speculative characteristics, strong retail investor participation, and greater sentiment sensitivity, consistent with the adaptive herding framework proposed by Ooi (2025).

Second, we challenge the “Islamic moderation hypothesis” by demonstrating that religious sentiment can amplify rather than suppress herding during Ramadan. While Stavroyiannis and Babalos (2017) found anti-herding in U.S. Islamic stocks and Chaffai and Medhioub (2018) suggested that Islamic finance moderates behavioral biases, our findings indicate that the relationship between Islamic principles and behavioral tendencies is context-dependent and potentially overshadowed by social and emotional contagion during collective religious observances. This contributes to the nascent “Islamic behavioral finance” paradigm (Din et al., 2021; Ooi, 2025; Syibawaih et al., 2026) by revealing the complex interplay between Islamic principles and market psychology, consistent with Boulanouar et al. (2026) who documented a shift toward more individualistic, market-driven investment behavior among Muslim investors.

Third, we reconcile contradictory evidence on GCC herding by demonstrating that herding during Ramadan is robust to both up and down markets, while outside Ramadan it follows the asymmetric patterns documented in prior literature (Chaffai and Medhioub, 2018; Medhioub and Chaffai, 2018). This suggests Ramadan acts as a structural break in normal market behavior, a finding with important implications for time-varying herding models (Rubbaniy et al., 2026) and dynamic asset pricing. The resolution of these contradictions is particularly important for understanding GCC market dynamics, as our findings complement Loang and Ahmad (2023) by demonstrating that predictable religious events can intensify herding regardless of market direction.

Fourth, our findings suggest that Ramadan-induced herding is predominantly spurious in nature, resulting from common emotional exposure to the Ramadan experience, collective positive mood, enhanced social interaction, and reduced risk aversion (Gavriilidis et al., 2016; Białkowski et al., 2012). Rubbaniy et al. (2026) differentiated between fundamental and intentional herding in energy markets, finding that herding is primarily motivated by intentional drivers. Our findings extend this by demonstrating that Ramadan-induced herding is driven by common emotional exposure rather than deliberate imitation based on informational asymmetries, a distinction with important implications for understanding whether such herding undermines market efficiency.

6.2. Managerial Implications

The intensification of herding during Ramadan diminishes diversification benefits as cross-sectional correlations increase within sectors (Ukpong et al., 2021; Tsania et al., 2026). Portfolio managers should rotate into low-herding sectors (Consumer Goods, Banks) to reduce exposure to sentiment-driven volatility and consider Ramadan-specific hedges while exploiting cross-sector dispersion that emerges during Ramadan’s later weeks (Rubbaniy et al., 2026). Enhanced monitoring of trading activity during Ramadan, particularly in high-herding sectors, is essential. Circuit breakers during extreme market movements, improved disclosure mechanisms, and educational campaigns before Ramadan could mitigate herding risks (Ah Mand et al., 2023; Din et al., 2021; Syibawaih et al., 2026). Stress testing models should incorporate Ramadan as a distinct risk regime, as regulators cannot rely on market direction alone to mitigate collective behavior risks (Rubbaniy et al., 2026; Loang and Ahmad, 2023).

The systematic pattern of Ramadan herding presents trading opportunities. Pre-Ramadan positioning (1-2 weeks before) can capture positive returns, momentum strategies during Ramadan may be more profitable in high-herding sectors, and reversal strategies post-Ramadan can exploit mean reversion effects (Al-Hajieh et al., 2011; Białkowski et al., 2012; Ooi, 2025). While volatility is lower during Ramadan, correlated trading intensifies, creating greater systemic risk if a negative shock occurs. Ramadan should be treated as a distinct risk regime in stress testing, with Value-at-Risk and Expected Shortfall estimates adjusted upward, and systemic risk monitoring intensified (Rubbaniy et al., 2026; Loang and Ahmad, 2023). Understanding Ramadan herding can protect retail investors from emotionally driven decisions during religious observances (Al-Awadhi et al., 2024). Educational interventions targeting cognitive biases could help investors avoid herding (Din et al., 2021). Our study also illuminates how cultural and religious factors interact with financial behavior in non-Western contexts, contributing to the globalization of behavioral finance research (Boulanouar et al., 2026).

6.3. Limitations and Suggestions for Future Research

This study has several limitations. The single-market focus on Qatar limits generalizability to other Islamic markets with different development levels, investor compositions, or regulatory frameworks. The ten-year horizon, while capturing multiple Ramadan cycles, may not reflect long-term structural changes. The CSAD model detects market-wide herding but cannot distinguish between institutional and retail investors, nor can it directly differentiate intentional from spurious herding, though theoretical evidence suggests Ramadan-induced herding is predominantly spurious. The lack of sentiment data prevents direct measurement of investor mood during Ramadan, and the seven-sector analysis may mask sub-sectoral variation in herding intensity.

Future research should conduct cross-market comparisons across GCC countries to determine whether observed patterns in Qatar are representative of broader regional dynamics. Disaggregated trading data could identify whether institutional or retail investors drive Ramadan herding, while microstructure analysis could examine changes in trading volume, bid-ask spreads, and order flow. Sentiment analysis using natural language processing could provide direct evidence of the mood-herding relationship. Event studies on other Islamic observances (Hajj, Eid) could establish whether the Ramadan effect is unique or part of a broader pattern. Rolling window analysis and state-space models (Stavroyiannis and Babalos, 2017; Rubbaniy et al., 2026) could explore time-varying dynamics, while pre/post-COVID comparisons could assess how exogenous shocks affect the Ramadan effect. Finally, survey-based mediation analysis could investigate whether Shariah literacy moderates herding behavior (Din et al., 2021).