Copernican Journal of Finance & Accounting e-ISSN 2300-3065 p-ISSN 2300-12402025, volume 14, issue 1 Date of submission: February 20, 2025; date of acceptance: May 23, 2025. * Contact information (corresponding author): m_haa60@yahoo.com, Lecturer of Business Administration, Sadat Academy for Management Sciences, Cairo, Egypt, phone: 002 0 109 919 5459; ORCID ID: https://orcid.org/0000-0002-0531-6325. ** Contact information: khloudmagdy@gmail.com, Master of Business Adminis- tration, Sadat Academy for Management Sciences, Cairo, Egypt, phone: 002 0 122 307 8488; ORCID ID: https://orcid.org/0009-0003-5308-9555. *** Contact information: ma7gw8marsafwat@gmail.com, Master of Business Admin- istration, Sadat Academy for Management Sciences, Cairo, Egypt, phone: 002 0 110 416 5280; ORCID ID: https://orcid.org/0009-0003-5378-3125. **** Contact information: nada.nour.shawky2002@gmail.com, Master of Business Ad- ministration, Sadat Academy for Management Sciences, Cairo, Egypt, phone: 002 0 106 556 4332; ORCID ID: https://orcid.org/0009-0009-6995-5001. Abd-Allah, M.H., Mohamed, K.M.K., Mahmood, M.S.M., & Mohamed, N.N.S. (2025). Herding Behav- iour and Stock Market Efficiency: An Empirical Study in the Egyptian Stock Market. Copernican Journal of Finance & Accounting, 14(1), 9–28. http://dx.doi.org/10.12775/CJFA.2025.001 Mustafa Hussein abd-allaH* Sadat Academy for Management Sciences KHloud Magdy KHalaf MoHaMed** Sadat Academy for Management Sciences MaryaM safwat MoHaMMed MaHMood*** Sadat Academy for Management Sciences nada nour sayed MoHaMed**** Sadat Academy for Management Sciences HERDING BEHAVIOR AND STOCK MARKET EFFICIENCY: AN EMPIRICAL STUDY IN THE EGYPTIAN STOCK MARKET Keywords: herding behavior, stock market efficiency, Egyptian stock exchange, EGX100 index. M. H. Abd-Allah, K. M. K. Mohamed, M. S. M. Mahmood, N. N. S. Mohamed1010 J E L Classification: G23, M14, O16. Abstract: This paper examines the impact of herding behavior on market efficiency in the Egyptian stock market. It does so by testing the efficiency of two separate stock portfolios: one that exhibits statistically significant herding, and another in which herding is not significant. The efficiency was tested using the Runs test and the Aug- mented Dickey-Fuller test for Unit Root. Hwang and Salmon (2004) methodology was also applied to measure herding behavior. Our comprehensive analysis revealed sev- eral key findings regarding market efficiency at the weak form. Firstly, Shapiro-Wilk and Kolmogorov-Smirnov tests indicated that stock prices for both portfolios are not normally distributed, suggesting a deviation from the assumptions of weak-form effi- ciency. Secondly, the Runs test conclusively showed that the time series for both port- folios are non-random, further demonstrating the market’s inability to achieve a weak level of efficiency. Finally, the Augmented Dickey-Fuller (ADF) test confirmed the sta- tionarity of the return series for both portfolios, unequivocally rejecting the presence of a unit root and reinforcing the conclusion of market inefficiency at the weak level. Collectively, these results strongly reject the random walk hypothesis and demonstrate that the Egyptian stock market does not achieve weak-form efficiency, regardless of the presence or absence of herding behavior. This finding has significant implications for investors: it suggests that investors may be able to exploit past price information to po- tentially generate abnormal returns, as market prices do not fully reflect all available historical data.  Introduction Introduction The debate between financial market efficiency and behavioral finance has been a topic of controversy for decades. While the theory of market efficiency argues that investors act rationally and that stock prices rapidly adjust to re- flect all available information, behavioral finance challenges this notion. It ex- plores the influence of psychology on investor behavior, suggesting that mar- kets are not always rational and that investors often make irrational decisions, leading to potential mispricing of assets. In the 1960s and early 1970s, there was disagreement over whether stock prices followed random movement or not. The tests by Fama (1965) and Sam- uelson (1965a) were conducted to answer this question; they concluded that most of the evidence appeared to be consistent with the market efficiency hy- pothesis. The stock market is considered efficient when it is efficient to process information, as Fama (1976) suggests. In an efficient stock market, stock prices fully incorporate all available information; therefore, no investor can achieve abnormal returns through past stock prices or private sources. HERDING BEHAVIOR AND STOCK MARKET EFFICIENCY… 1111 Regarding herding behavior, Hwang and Salmon (2004) explained that it occurs when investors choose to follow others’ decisions in the market instead of relying on their own beliefs and information. In such cases, investors disre- gard their own insights in favor of imitating others, regardless of whether the behavior is rational. Herding behavior is also a phenomenon that occurs during periods of stress. When significant price changes occur, individual investment returns align with the market. In the presence of herding behavior, all invest- ments follow the same trajectory as the market portfolio. As a result, investors engaging in herding behavior cause these returns to converge with the market return, thereby reducing their deviation from it. Through this study, we are trying to identify the impact of herding behav- ior on the efficiency of the Egyptian stock market by analyzing the efficiency of a portfolio of stocks showing herd behavior among investors and the efficiency of a portfolio of stocks that does not show herding behavior. Theoretical Background And Previous StudiesTheoretical Background And Previous Studies Financial Market EfficiencyFinancial Market Efficiency The market efficiency hypothesis is based on the “random walk theory” for both Kendall (1953) and Fama (1965), which is used to characterize the stock price chain, where each subsequent price change is random and independent of previous prices. The “random walk theory” suggests that when information flows freely and is instantly incorporated into stock prices, tomorrow’s price changes will solely reflect new information from that day, independent of to- day’s price movements. Consequently, stock prices are unpredictable, mean- ing price changes occur randomly and unexpectedly. As a result, all new in- formation is fully reflected in prices, ensuring that even uninformed investors can achieve high returns similar to experts. Cootner (1964) also stated that the stock market does not retain any memory, as the way the stock price behaves in the past does not help determine its future behavior. Samuelson (1965b) also indicated that with respect to the “random walk theory”, there were no pat- terns that investors could employ to achieve abnormal returns; if any excess returns are achieved, they should be credited to luck or seen as compensation for the additional risks taken by the investor. In other words, outperforming the market is not possible. M. H. Abd-Allah, K. M. K. Mohamed, M. S. M. Mahmood, N. N. S. Mohamed1212 Fama (1970) noted that all the information that can be gathered concerning the stock has already been incorporated into the stock price, where stock pric- es reflect all the stock’s information. Fama (1970) used the term “efficient mar- ket”. According to the term “efficient market”, competition will instantly coun- teract the full impact of new information on actual prices. Financial market efficiency levels are three, with Fama (1970) classifying them as weak, semi- strong, and strong. The assumption of a lower-level efficient stock market suggests that current stock prices fully incorporate all available historical price information, making past prices ineffective for predicting future price movements; therefore, no in- vestor can achieve abnormal returns through previous stock prices. The semi- strong form of financial market efficiency is a model in which current stock prices incorporate all publicly available information, including economic fac- tors, financial reports of the firm and other publicly available information in addition to previous stock prices. Finally, the strong form of efficiency assumes that stock prices fully incorporate all available information, both public and private, preventing any investor from gaining abnormal returns through pri- vate insights. Shiller (1999) noted that although the theory of financial market efficiency remains the basis for the development of many tools and models used to under- stand how financial markets operate, it failed to explain investors’ behaviors. Yao, Ma and He (2014) find that the theory of financial market efficiency failed to learn the mechanisms for the functioning of financial markets due to cogni- tive biases and complex human responses. In the early 1960s and mid-1970s, the concept of market efficiency was broadly embraced by academic financial economists, as evidenced by works such as Fama (1970), Fama and MacBeth (1973), Black, Jensen and Scholes (1972), and Jensen (1978). There was a general belief that stock markets were highly efficient in their ability to reflect stock information at the individual or the stock market level as a whole. Meanwhile, Fama, Fisher, Jensen and Roll (1969) found that in an active market involving many knowledgeable and intel- ligent investors, stock prices will accurately reflect all available information. Rationality is one of the fundamental pillars of the classical theory of finance, where Barberis and Thaler (2003) indicate that when new information reaches the market, investors make decisions based on Baez’s law, aligning their choic- es with their expected personal benefit, since, under conditions of uncertainty, they choose the decision that delivers the greatest expected self-benefit. HERDING BEHAVIOR AND STOCK MARKET EFFICIENCY… 1313 After the mid-1970s, an increase in research challenged the assumption of market efficiency. More specifically, numerous researchers observed that stock prices exhibited certain unexplained patterns that could not be accounted for by the asset pricing models of the time. These patterns, known as “anomalies”, posed a significant challenge for emerging markets, as they could not be pre- dicted by existing asset pricing models. Anomalies sparked extensive research that challenged the assumption of investor rationality, a fundamental princi- ple of stock market efficiency. This led many researchers to explore what later became known as behavioral finance, drawing on cognitive psychology to ar- gue that human behavior is not always rational. Studies have documented vari- ous behavioral and cognitive biases that influence decision-making and devi- ate from logical judgments. These findings present a significant challenge to market efficiency, suggesting that markets are not entirely homogeneous, and investors are not always fully rational, as documented by Rozeff and Kinney (1976), French (1980), and Shiller (1999). There are several other applied studies that have explored market efficien- cy. Ozkan (2021) aimed to investigate the impact of the recent coronavirus (COVID-19) pandemic on stock market efficiency for six hard-hit developed countries. The study’s population and sample included daily stock market data from the United States (US), Spain, the United Kingdom (UK), Italy, France, and Germany. The time period of the analysis spanned from July 29, 2019, to Janu- ary 25, 2021. To measure market efficiency. The key finding indicated that all stock markets examined deviated from market efficiency during some peri- ods of the pandemic, with deviations being more pronounced in the US and UK stock markets during the outbreak. This suggests an increasing chance for stock price predictions and abnormal returns during the COVID-19 pandemic. El Mosallamy and Gamal (2024) focused on the Egyptian stock market, spe- cifically analyzing data from the EGX30 and EGX100 indices, with the primary goal of investigating the impact of currency devaluation events on the market’s efficiency. Their study’s time period covered the significant devaluation events in Egypt during 2016 and 2022. To measure efficiency, they employed both par- ametric and non-parametric unit root tests for stationarity, along with ARCH and EGARCH models. The key finding indicated that the Egyptian market’s re- turns showed stationarity regardless of these major economic events, suggest- ing a certain level of efficiency even amidst such turbulence. M. H. Abd-Allah, K. M. K. Mohamed, M. S. M. Mahmood, N. N. S. Mohamed1414 Herding Behavior in Financial MarketsHerding Behavior in Financial Markets In the late 19th century, psychologist and sociologist Gustave Le Bon (1895) ex- plored crowd psychology and drew intriguing conclusions about how individu- als behave when they become part of a crowd; these include the absence of re- sponsibility and infection. The absence of responsibility means that when an individual joins a crowd, their sense of responsibility that governs behavior di- minishes, and thus the individual carries out actions that he would not have cho- sen without his presence in the crowd. Infection means that when part of the crowd, the individual disregards their personal beliefs and interests, conform- ing instead to the beliefs of the crowd; thus, individuals change attitudes and of- ten behave very differently from their basic behavior in the crowd or group than if they acted individually. Both Bikhchandani and Sharma (2001) defined herding behavior as inad- vertently or intentionally copying the behavior of other investors by some in- vestors. Hirshleifer and Teoh (2003) have noted that herding behavior involves similar behavior, and the causes of this similarity cannot always be deciphered, as it can be attributed to a variety of motivations with both psychological and rational backgrounds. Hirshleifer (2001) noted that “conformity” is one of the most important factors related to psychological background in which peo- ple feel more comfortable matching the work that others do, If an investment method has been successful in the past and is widely adopted by investors, it is likely that others will follow the same approach, selecting stocks based on the characteristics outlined by that method. Thus, maintaining the same pat- tern will lead to herding behavior. Devenow and Welch (1996) explained that investors considered it logical to mimic others if this would allow them to ben- efit from other person’s accumulated potential information experiences. The reason for this is that the investor may not have any information or may have information but is unsure of its decisions, perhaps because the investor’s infor- mation is of low quality, considers its information processing capabilities insuf- ficient, or sees others as better informed than him. Buckner (1965) explained that herding behavior may be caused by rumors. Schindler (2007) found that investors are frequently affected by rumors, either through social interactions or exposure to media reports, which are often un- reliable. Individual investors tend to be more susceptible to such rumors, par- ticularly during times of crisis when fear and uncertainty dominate the mar- HERDING BEHAVIOR AND STOCK MARKET EFFICIENCY… 1515 ket, while professionals often benefit better from these rumors and make gains from them. Guney, Kallinterakis and Komba (2017) also found that herding behavior exists strongly in African markets. Due to limited transparency that weakens the quality of their information landscape, investors rely on herding behavior to derive insights from the trading activities of their peers. Information asym- metry is also one of the most important factors leading strongly to herding be- havior, as Kauffman and Li (2003) indicate that due to incomplete and incon- sistent information in the stock market, decision-makers attempt to obtain higher value information by monitoring other people’s procedures for obtain- ing information, to overcome inaccurate information for the decision-maker. In their study of 1,155 Chinese companies from 1999 to 2004, Bo, Li and Sun (2013) found that when companies operate in an environment of high uncer- tainty or significant information asymmetry, managers in these circumstances gather in crowds because this is considered safer for them. Gavriilidis (2013) found that information was the most important element of successful investment; however, the process of collecting accurate informa- tion consumes a lot of time and money, so investors lacking sufficient informa- tion tend to replicate the decisions of those they perceive as more knowledgea- ble. Even those with dedicated media resources for data collection and analysis often opt to follow other investors when they believe their information is supe- rior. This behavior, known as a “free ride”, can ultimately result in a pattern of interconnected trading among investors. The concept of information cascading has been widely explored in research on herding behavior, initially introduced by Banerjee (1992) and Bikhchanda- ni, Hirshleifer and Welch (1992). They identified the information asymmetry among investors as a potential source of herding behavior and called it the “in- formational cascading”. This information relay occurs when an investor de- cides to ignore her information collection and relies on information about oth- er investors, as he works to mimic previous investors because he does not trust the information he possesses and this is thus reflected in his actions. Informa- tion cascading can negatively impact market efficiency, as it prevents all avail- able information from being accurately reflected in stock prices. Hirshleifer and Teoh (2003) assumed that the information cascading phe- nomenon caused information constraints, as not all investor information was transferred to the market. Bikhchandani, et al. (1992) explained that informa- tion cascading leads to increased herding behavior among investors. Informa- M. H. Abd-Allah, K. M. K. Mohamed, M. S. M. Mahmood, N. N. S. Mohamed1616 tion cascading is also initiated by a small number of individuals who have limit- ed information so that the company that follows the crowd completely ignores its information or may not assign it the proper significance and blindly follows the decisions of others, which makes valuable information lost. Many studies have tried to measure herding behavior through accurate sci- entific measures by which to assess the existence of herding behavior with- in financial markets and determine its proportion. Christie and Huang (1995) introduced a method known as Cross-Sectional Standard Deviation (CSSD) to measure dispersion and assess herding behavior in the market. This approach examines how individual stock returns deviate from the overall market portfo- lio return. If investors exhibit herding behavior, individual stock returns should closely align with market returns. The study analyzed market pressure during periods of significant market return fluctuations, assuming that investors set aside their personal beliefs and based their decisions on market consensus. Us- ing dispersion analysis, the study evaluated daily and monthly NYSE and AMEX data from 1962 to 1988. The findings revealed that dispersion significantly in- creased during large price movements, aligning with rational asset pricing mod- els and ultimately failing to confirm the presence of herding behavior. Chang, Cheng and Khorana (2000) then introduced an innovative approach built upon the relationship between Cross Sectional Absolute Deviation (CSAD) and the market’s unusual circumstances. The research investigated herding behavior across various global stock markets and found mixed results. While no evidence of herding was detected in developed markets, the findings indi- cated its presence in emerging markets like South Korea and Taiwan. Hwang and Salmon (2004) also introduced an alternative approach in their study of the American and South Korean markets, utilizing the Cross-Sectional Standard Deviation of Betas. This model assumes that when investors exhib- it behavioral biases – meaning their decisions are not entirely rational – their perception of the relationship between asset returns and risk becomes skewed. Consequently, if herding behavior is present in the market, the returns on all in- vestments will align with the market portfolio, causing CAPM betas to diverge from their equilibrium values. In the context of applied studies on market efficiency, a recent investigation by Litimi (2017) examined herding behavior within the French stock market. The study specifically aimed to assess this behavior’s influence on idiosyncrat- ic conditional volatility at the sectoral level. The research sample encompassed all firms listed on the French stock market, categorized by their respective sec- HERDING BEHAVIOR AND STOCK MARKET EFFICIENCY… 1717 tors. The analytical period spanned from 2000 to 2016, encompassing four dis- tinct economic crisis periods: the stock market downturn of 2002, the global fi- nancial crisis, the European sovereign debt crisis, and the Brexit event. Abd-Alla (2020) investigated the presence of herding behavior in the Egyp- tian stock market during the COVID-19 crisis. The research utilized firm-level data and further examined herding at the portfolio level, categorized by size and value factors. To measure herding behavior, the study employed two dis- tinct methodologies: the nonlinear model proposed by Chang et al. (2000) and the state-space model developed by Hwang and Salmon (2004). The findings presented a notable contrast between the two models. The nonlinear model by Chang et al. (2000) yielded results indicating evidence of herding during the COVID-19 crisis. Conversely, the state-space model by Hwang and Salmon (2004) revealed no evidence of herding behavior during the same period. At the portfolio level, herding was detected exclusively when applying the Chang et al. (2000) methodology, specifically in portfolios comprising stocks with low and high book-to-market (B/M) ratios and in the portfolio of large-cap stocks dur- ing the COVID-19 crisis. In their investigation into investor behavior dynamics, a study conducted by Hudson, Yan and Zhang (2020) explored whether investor sentiment influences fund managers’ herding behavior. The research utilized two composite indices as indicators of investor sentiment: the UK market-wide sentiment index and the UK institutional sentiment index. Herding behavior itself was quantified through the cross-sectional dispersion of factor sensitivity, specifically using beta. To examine the association between institutional herding and investor sentiment, the authors employed the Vector Autoregression (VAR) rule. The es- timations provided evidence suggesting that herding behavior may not solely be driven by market fundamentals but also by investor sentiment. The find- ings distinctly indicated that both UK market-wide and institutional sentiment significantly impact the herding tendencies of UK open-ended fund managers. These observed effects demonstrated a unidirectional influence, thus support- ing the proposition that investor sentiment serves as a key driver of herding. Abd El-Razcek and Abd Elbaseet (2023) addressed the issue of herding be- havior in the Egyptian stock market. The study aimed to test the existence of this behavior across various time periods, including the revolutionary era and the COVID-19 pandemic. Utilizing daily stock price data, the research employed an empirical analysis. Results generally indicated an absence of evidence for collective herding behavior in the Egyptian market. However, the models used M. H. Abd-Allah, K. M. K. Mohamed, M. S. M. Mahmood, N. N. S. Mohamed1818 revealed the presence of “adverse herding”, characterized by a non-linear rela- tionship. The study also noted that herding, when analyzed within sub-periods, represents a transient phenomenon linked to market classifications as bullish or bearish, and to changes in volatility levels. Conversely, in aggregated mod- els, herding was observed in bullish markets with high volatility, while adverse herding emerged in bearish markets with low volatility. Messaoud and Ben Amar (2025) investigated the relationship between herding behavior and investor sentiment in emerging stock markets (ESMs). The research aimed to determine the impact of sentiment on herding and the causal direction between them. Using daily data from several major emerging markets, the findings indicated that higher investor sentiment is generally as- sociated with lower herding. However, during market downturns, elevated sen- timent led to increased herding. The study also revealed a one-way causal re- lationship from investor sentiment to herding behavior. These results suggest that investor sentiment can contribute to irrational herding, potentially in- creasing market volatility and the likelihood of bubbles, underscoring the need for regulators to monitor sentiment to predict herding and stabilize markets. Data And MethodologyData And Methodology DataData The study analyzed the impact of herding behavior on the efficiency of the Egyptian stock market using daily and monthly data from a 60-month peri- od from 1/1/2014 to 31/12/20181. It examined a sample of 50 stocks from the EGX100 index, which represents the top 100 active firms in the market. The data included stock closing prices, EGX100 index values, and monthly beta cal- culations for each stock. The sample stocks were selected based on specific criteria: they must be traded in Egyptian pounds, exclude banks and financial institutions due to their unique characteristics, have continuous price data for 60 months, and be listed on the EGX100 index. 1 This period was selected for this study primarily to ensure the availability of complete and reliable data for the Egyptian stock market. This timeframe also represents a period of relative stability in market conditions, which facilitated a more accurate and robust analysis of herding behavior. HERDING BEHAVIOR AND STOCK MARKET EFFICIENCY… 1919 MethodologyMethodology Herding Behavior MeasurementHerding Behavior Measurement The Hwang and Salmon (2004) model was applied to measure herding behavior in the Egyptian stock market. To calculate herding behavior values, the Hwang and Salmon (2004) model was based on the application of the State Space Mod- el. Through the model, herding behavior can be estimated by logarithm Cross Sectional Standard Deviation of Betas Log [ active firms in the market. The data included stock closing prices, EGX100 index values, and monthly beta calculations for each stock. The sample stocks were selected based on specific criteria: they must be traded in Egyptian pounds, exclude banks and financial institutions due to their unique characteristics, have continuous price data for 60 months, and be listed on the EGX100 index. Methodology Herding Behavior Measurement The Hwang and Salmon (2004) model was applied to measure herding behavior in the Egyptian stock market. To calculate herding behavior values, the Hwang and Salmon (2004) model was based on the application of the State Space Model. Through the model, herding behavior can be estimated by logarithm Cross Sectional Standard Deviation of Betas Log [ 𝑆𝑆𝑆𝑆𝑆𝑆�(𝛽𝛽���� )], and to measure herding behavior, the following steps have been followed: Calculating the daily stock return as well as the daily EGX 100 return by using the following equations: Daily stock return 𝑅𝑅��= ������������� )1 ( Where 𝑅𝑅�� represents the return of stock i on day t, while 𝑃𝑃� and 𝑃𝑃��� denote the closing prices of stock i on days t and t-1, respectively. Daily EGX 100 return 𝑅𝑅��= ����������� ����� )2 ( Where 𝑅𝑅�� represents the return of EGX 100 index on day t, while 𝑃𝑃�� and 𝑃𝑃���� denote the closing prices of EGX 100 index on days t and t-1, respectively. Calculating Excess Daily Return for each stock as well as the EGX 100 index: The Excess Daily Return is calculated by deducting the risk-free rate from the daily return of each stock, as well as from the daily return of the EGX 100 market index. Monthly Beta is calculated using the Ordinary Least Squares method (OLS) by the following regression formula: ], and to measure herding behavior, the following steps have been followed: Calculating the daily stock return as well as the daily EGX 100 return by us- ing the following equations: Daily stock return active firms in the market. The data included stock closing prices, EGX100 index values, and monthly beta calculations for each stock. The sample stocks were selected based on specific criteria: they must be traded in Egyptian pounds, exclude banks and financial institutions due to their unique characteristics, have continuous price data for 60 months, and be listed on the EGX100 index. Methodology Herding Behavior Measurement The Hwang and Salmon (2004) model was applied to measure herding behavior in the Egyptian stock market. To calculate herding behavior values, the Hwang and Salmon (2004) model was based on the application of the State Space Model. Through the model, herding behavior can be estimated by logarithm Cross Sectional Standard Deviation of Betas Log [ 𝑆𝑆𝑆𝑆𝑆𝑆�(𝛽𝛽���� )], and to measure herding behavior, the following steps have been followed: Calculating the daily stock return as well as the daily EGX 100 return by using the following equations: Daily stock return 𝑅𝑅��= ������������� )1 ( Where 𝑅𝑅�� represents the return of stock i on day t, while 𝑃𝑃� and 𝑃𝑃��� denote the closing prices of stock i on days t and t-1, respectively. Daily EGX 100 return 𝑅𝑅��= ����������� ����� )2 ( Where 𝑅𝑅�� represents the return of EGX 100 index on day t, while 𝑃𝑃�� and 𝑃𝑃���� denote the closing prices of EGX 100 index on days t and t-1, respectively. Calculating Excess Daily Return for each stock as well as the EGX 100 index: The Excess Daily Return is calculated by deducting the risk-free rate from the daily return of each stock, as well as from the daily return of the EGX 100 market index. Monthly Beta is calculated using the Ordinary Least Squares method (OLS) by the following regression formula: (1) Where Rit represents the return of stock i on day t, while Pt and Pt–1 denote the closing prices of stock i on days t and t-1, respectively. Daily EGX 100 return active firms in the market. The data included stock closing prices, EGX100 index values, and monthly beta calculations for each stock. The sample stocks were selected based on specific criteria: they must be traded in Egyptian pounds, exclude banks and financial institutions due to their unique characteristics, have continuous price data for 60 months, and be listed on the EGX100 index. Methodology Herding Behavior Measurement The Hwang and Salmon (2004) model was applied to measure herding behavior in the Egyptian stock market. To calculate herding behavior values, the Hwang and Salmon (2004) model was based on the application of the State Space Model. Through the model, herding behavior can be estimated by logarithm Cross Sectional Standard Deviation of Betas Log [ 𝑆𝑆𝑆𝑆𝑆𝑆�(𝛽𝛽���� )], and to measure herding behavior, the following steps have been followed: Calculating the daily stock return as well as the daily EGX 100 return by using the following equations: Daily stock return 𝑅𝑅��= ������������� )1 ( Where 𝑅𝑅�� represents the return of stock i on day t, while 𝑃𝑃� and 𝑃𝑃��� denote the closing prices of stock i on days t and t-1, respectively. Daily EGX 100 return 𝑅𝑅��= ����������� ����� )2 ( Where 𝑅𝑅�� represents the return of EGX 100 index on day t, while 𝑃𝑃�� and 𝑃𝑃���� denote the closing prices of EGX 100 index on days t and t-1, respectively. Calculating Excess Daily Return for each stock as well as the EGX 100 index: The Excess Daily Return is calculated by deducting the risk-free rate from the daily return of each stock, as well as from the daily return of the EGX 100 market index. Monthly Beta is calculated using the Ordinary Least Squares method (OLS) by the following regression formula: (2) Where Rmt represents the return of EGX 100 index on day t, while Pmt and Pmt–1 denote the closing prices of EGX 100 index on days t and t-1, respectively. Calculating Excess Daily Return for each stock as well as the EGX 100 index: The Excess Daily Return is calculated by deducting the risk-free rate from the daily return of each stock, as well as from the daily return of the EGX 100 market index. Monthly Beta is calculated using the Ordinary Least Squares method (OLS) by the following regression formula: 𝑟𝑟��� = 𝛼𝛼��� + 𝛽𝛽���� 𝑟𝑟��� + 𝜀𝜀��� (3) Where 𝑟𝑟��� and 𝑟𝑟��� represent the daily return of stock i and the EGX100 index during month t respectively, while 𝛽𝛽���� denotes the monthly beta for each stock. The regression analysis was performed 60 times per stock to obtain monthly Beta per stock, thus the regression analysis was performed 3,000 times for all stocks. The Hwang and Salmon (2004) methodology is based on the Cross-Sectional Standard deviation of betas. as follows: … �� ������ ������� = 𝛽𝛽���� =𝛽𝛽���  ℎ�� (𝛽𝛽���1) (4) Where 𝐸𝐸�� represents the expected return of stock i under behavioral bias, while 𝐸𝐸��𝑟𝑟��� denotes the expected return of the market portfolio. ℎ�� represents the Herding behavior Coefficient, if (ℎ��= 0) herding behavior is absent. Conversely, if (ℎ��= 0), herding behavior is fully present, meaning individual stock returns move in sync with market portfolio returns. Hwang and Salmon (2004) proposed the State Space Model to measure herding behavior using the change in stock sensitivity to change in market return through the following two equations: Log [ 𝑆𝑆𝑆𝑆𝑆𝑆�(𝛽𝛽���� )] = 𝜇𝜇� + 𝐻𝐻�� + 𝜐𝜐�� (5) 𝐻𝐻�� = 𝜑𝜑� 𝐻𝐻�,��� + 𝜂𝜂�� (6) Where: 𝜂𝜂��~ iid (0,𝜎𝜎�,�� ) The model was estimated using the Kalman filter methodology within EViews software. The presence of herding behavior is inferred from the significant value of the variance of the error term 𝜂𝜂�� (𝜎𝜎�,�� ), and the significance of the persistence parameter �𝜑𝜑��. Cross-Sectional Standard deviation of betas is calculated by the following formula: 𝑆𝑆𝑆𝑆𝑆𝑆�(𝛽𝛽��) = ������ �������� ������������� ��� (7) STATISTICAL RESULTS OF HERDING BEHAVIOR The study sample was divided into two different portfolios, as follows: (3) M. H. Abd-Allah, K. M. K. Mohamed, M. S. M. Mahmood, N. N. S. Mohamed2020 Where ritd and rmtd represent the daily return of stock i and the EGX100 in- dex during month t, respectively, while 𝑟𝑟��� = 𝛼𝛼��� + 𝛽𝛽���� 𝑟𝑟��� + 𝜀𝜀��� (3) Where 𝑟𝑟��� and 𝑟𝑟��� represent the daily return of stock i and the EGX100 index during month t respectively, while 𝛽𝛽���� denotes the monthly beta for each stock. The regression analysis was performed 60 times per stock to obtain monthly Beta per stock, thus the regression analysis was performed 3,000 times for all stocks. The Hwang and Salmon (2004) methodology is based on the Cross-Sectional Standard deviation of betas. as follows: … �� ������ ������� = 𝛽𝛽���� =𝛽𝛽���  ℎ�� (𝛽𝛽���1) (4) Where 𝐸𝐸�� represents the expected return of stock i under behavioral bias, while 𝐸𝐸��𝑟𝑟��� denotes the expected return of the market portfolio. ℎ�� represents the Herding behavior Coefficient, if (ℎ��= 0) herding behavior is absent. Conversely, if (ℎ��= 0), herding behavior is fully present, meaning individual stock returns move in sync with market portfolio returns. Hwang and Salmon (2004) proposed the State Space Model to measure herding behavior using the change in stock sensitivity to change in market return through the following two equations: Log [ 𝑆𝑆𝑆𝑆𝑆𝑆�(𝛽𝛽���� )] = 𝜇𝜇� + 𝐻𝐻�� + 𝜐𝜐�� (5) 𝐻𝐻�� = 𝜑𝜑� 𝐻𝐻�,��� + 𝜂𝜂�� (6) Where: 𝜂𝜂��~ iid (0,𝜎𝜎�,�� ) The model was estimated using the Kalman filter methodology within EViews software. The presence of herding behavior is inferred from the significant value of the variance of the error term 𝜂𝜂�� (𝜎𝜎�,�� ), and the significance of the persistence parameter �𝜑𝜑��. Cross-Sectional Standard deviation of betas is calculated by the following formula: 𝑆𝑆𝑆𝑆𝑆𝑆�(𝛽𝛽��) = ������ �������� ������������� ��� (7) STATISTICAL RESULTS OF HERDING BEHAVIOR The study sample was divided into two different portfolios, as follows: denotes the monthly beta for each stock. The regression analysis was performed 60 times per stock to obtain month- ly Beta per stock, thus the regression analysis was performed 3,000 times for all stocks. The Hwang and Salmon (2004) methodology is based on the Cross- Sectional Standard deviation of betas, as follows: 𝑟𝑟��� = 𝛼𝛼��� + 𝛽𝛽���� 𝑟𝑟��� + 𝜀𝜀��� (3) Where 𝑟𝑟��� and 𝑟𝑟��� represent the daily return of stock i and the EGX100 index during month t respectively, while 𝛽𝛽���� denotes the monthly beta for each stock. The regression analysis was performed 60 times per stock to obtain monthly Beta per stock, thus the regression analysis was performed 3,000 times for all stocks. The Hwang and Salmon (2004) methodology is based on the Cross-Sectional Standard deviation of betas. as follows: … �� ������ ������� = 𝛽𝛽���� =𝛽𝛽���  ℎ�� (𝛽𝛽���1) (4) Where 𝐸𝐸�� represents the expected return of stock i under behavioral bias, while 𝐸𝐸��𝑟𝑟��� denotes the expected return of the market portfolio. ℎ�� represents the Herding behavior Coefficient, if (ℎ��= 0) herding behavior is absent. Conversely, if (ℎ��= 0), herding behavior is fully present, meaning individual stock returns move in sync with market portfolio returns. Hwang and Salmon (2004) proposed the State Space Model to measure herding behavior using the change in stock sensitivity to change in market return through the following two equations: Log [ 𝑆𝑆𝑆𝑆𝑆𝑆�(𝛽𝛽���� )] = 𝜇𝜇� + 𝐻𝐻�� + 𝜐𝜐�� (5) 𝐻𝐻�� = 𝜑𝜑� 𝐻𝐻�,��� + 𝜂𝜂�� (6) Where: 𝜂𝜂��~ iid (0,𝜎𝜎�,�� ) The model was estimated using the Kalman filter methodology within EViews software. The presence of herding behavior is inferred from the significant value of the variance of the error term 𝜂𝜂�� (𝜎𝜎�,�� ), and the significance of the persistence parameter �𝜑𝜑��. Cross-Sectional Standard deviation of betas is calculated by the following formula: 𝑆𝑆𝑆𝑆𝑆𝑆�(𝛽𝛽��) = ������ �������� ������������� ��� (7) STATISTICAL RESULTS OF HERDING BEHAVIOR The study sample was divided into two different portfolios, as follows: (4) Where 𝑟𝑟��� = 𝛼𝛼��� + 𝛽𝛽���� 𝑟𝑟��� + 𝜀𝜀��� (3) Where 𝑟𝑟��� and 𝑟𝑟��� represent the daily return of stock i and the EGX100 index during month t respectively, while 𝛽𝛽���� denotes the monthly beta for each stock. The regression analysis was performed 60 times per stock to obtain monthly Beta per stock, thus the regression analysis was performed 3,000 times for all stocks. The Hwang and Salmon (2004) methodology is based on the Cross-Sectional Standard deviation of betas. as follows: … �� ������ ������� = 𝛽𝛽���� =𝛽𝛽���  ℎ�� (𝛽𝛽���1) (4) Where 𝐸𝐸�� represents the expected return of stock i under behavioral bias, while 𝐸𝐸��𝑟𝑟��� denotes the expected return of the market portfolio. ℎ�� represents the Herding behavior Coefficient, if (ℎ��= 0) herding behavior is absent. Conversely, if (ℎ��= 0), herding behavior is fully present, meaning individual stock returns move in sync with market portfolio returns. Hwang and Salmon (2004) proposed the State Space Model to measure herding behavior using the change in stock sensitivity to change in market return through the following two equations: Log [ 𝑆𝑆𝑆𝑆𝑆𝑆�(𝛽𝛽���� )] = 𝜇𝜇� + 𝐻𝐻�� + 𝜐𝜐�� (5) 𝐻𝐻�� = 𝜑𝜑� 𝐻𝐻�,��� + 𝜂𝜂�� (6) Where: 𝜂𝜂��~ iid (0,𝜎𝜎�,�� ) The model was estimated using the Kalman filter methodology within EViews software. The presence of herding behavior is inferred from the significant value of the variance of the error term 𝜂𝜂�� (𝜎𝜎�,�� ), and the significance of the persistence parameter �𝜑𝜑��. Cross-Sectional Standard deviation of betas is calculated by the following formula: 𝑆𝑆𝑆𝑆𝑆𝑆�(𝛽𝛽��) = ������ �������� ������������� ��� (7) STATISTICAL RESULTS OF HERDING BEHAVIOR The study sample was divided into two different portfolios, as follows: represents the expected return of stock i under behavioral bias, while Et (rmt ) denotes the expected return of the market portfolio. hmt repre- sents the Herding behavior Coefficient, if (hmt = 0) herding behavior is absent. Conversely, if (hmt = 0), herding behavior is fully present, meaning individual stock returns move in sync with market portfolio returns. Hwang and Salmon (2004) proposed the State Space Model to measure herding behavior using the change in stock sensitivity to change in market re- turn through the following two equations: 𝑟𝑟��� = 𝛼𝛼��� + 𝛽𝛽���� 𝑟𝑟��� + 𝜀𝜀��� (3) Where 𝑟𝑟��� and 𝑟𝑟��� represent the daily return of stock i and the EGX100 index during month t respectively, while 𝛽𝛽���� denotes the monthly beta for each stock. The regression analysis was performed 60 times per stock to obtain monthly Beta per stock, thus the regression analysis was performed 3,000 times for all stocks. The Hwang and Salmon (2004) methodology is based on the Cross-Sectional Standard deviation of betas. as follows: … �� ������ ������� = 𝛽𝛽���� =𝛽𝛽���  ℎ�� (𝛽𝛽���1) (4) Where 𝐸𝐸�� represents the expected return of stock i under behavioral bias, while 𝐸𝐸��𝑟𝑟��� denotes the expected return of the market portfolio. ℎ�� represents the Herding behavior Coefficient, if (ℎ��= 0) herding behavior is absent. Conversely, if (ℎ��= 0), herding behavior is fully present, meaning individual stock returns move in sync with market portfolio returns. Hwang and Salmon (2004) proposed the State Space Model to measure herding behavior using the change in stock sensitivity to change in market return through the following two equations: Log [ 𝑆𝑆𝑆𝑆𝑆𝑆�(𝛽𝛽���� )] = 𝜇𝜇� + 𝐻𝐻�� + 𝜐𝜐�� (5) 𝐻𝐻�� = 𝜑𝜑� 𝐻𝐻�,��� + 𝜂𝜂�� (6) Where: 𝜂𝜂��~ iid (0,𝜎𝜎�,�� ) The model was estimated using the Kalman filter methodology within EViews software. The presence of herding behavior is inferred from the significant value of the variance of the error term 𝜂𝜂�� (𝜎𝜎�,�� ), and the significance of the persistence parameter �𝜑𝜑��. Cross-Sectional Standard deviation of betas is calculated by the following formula: 𝑆𝑆𝑆𝑆𝑆𝑆�(𝛽𝛽��) = ������ �������� ������������� ��� (7) STATISTICAL RESULTS OF HERDING BEHAVIOR The study sample was divided into two different portfolios, as follows: (5) 𝑟𝑟��� = 𝛼𝛼��� + 𝛽𝛽���� 𝑟𝑟��� + 𝜀𝜀��� (3) Where 𝑟𝑟��� and 𝑟𝑟��� represent the daily return of stock i and the EGX100 index during month t respectively, while 𝛽𝛽���� denotes the monthly beta for each stock. The regression analysis was performed 60 times per stock to obtain monthly Beta per stock, thus the regression analysis was performed 3,000 times for all stocks. The Hwang and Salmon (2004) methodology is based on the Cross-Sectional Standard deviation of betas. as follows: … �� ������ ������� = 𝛽𝛽���� =𝛽𝛽���  ℎ�� (𝛽𝛽���1) (4) Where 𝐸𝐸�� represents the expected return of stock i under behavioral bias, while 𝐸𝐸��𝑟𝑟��� denotes the expected return of the market portfolio. ℎ�� represents the Herding behavior Coefficient, if (ℎ��= 0) herding behavior is absent. Conversely, if (ℎ��= 0), herding behavior is fully present, meaning individual stock returns move in sync with market portfolio returns. Hwang and Salmon (2004) proposed the State Space Model to measure herding behavior using the change in stock sensitivity to change in market return through the following two equations: Log [ 𝑆𝑆𝑆𝑆𝑆𝑆�(𝛽𝛽���� )] = 𝜇𝜇� + 𝐻𝐻�� + 𝜐𝜐�� (5) 𝐻𝐻�� = 𝜑𝜑� 𝐻𝐻�,��� + 𝜂𝜂�� (6) Where: 𝜂𝜂��~ iid (0,𝜎𝜎�,�� ) The model was estimated using the Kalman filter methodology within EViews software. The presence of herding behavior is inferred from the significant value of the variance of the error term 𝜂𝜂�� (𝜎𝜎�,�� ), and the significance of the persistence parameter �𝜑𝜑��. Cross-Sectional Standard deviation of betas is calculated by the following formula: 𝑆𝑆𝑆𝑆𝑆𝑆�(𝛽𝛽��) = ������ �������� ������������� ��� (7) STATISTICAL RESULTS OF HERDING BEHAVIOR The study sample was divided into two different portfolios, as follows: (6) Where: 𝑟𝑟��� = 𝛼𝛼��� + 𝛽𝛽���� 𝑟𝑟��� + 𝜀𝜀��� (3) Where 𝑟𝑟��� and 𝑟𝑟��� represent the daily return of stock i and the EGX100 index during month t respectively, while 𝛽𝛽���� denotes the monthly beta for each stock. The regression analysis was performed 60 times per stock to obtain monthly Beta per stock, thus the regression analysis was performed 3,000 times for all stocks. The Hwang and Salmon (2004) methodology is based on the Cross-Sectional Standard deviation of betas. as follows: … �� ������ ������� = 𝛽𝛽���� =𝛽𝛽���  ℎ�� (𝛽𝛽���1) (4) Where 𝐸𝐸�� represents the expected return of stock i under behavioral bias, while 𝐸𝐸��𝑟𝑟��� denotes the expected return of the market portfolio. ℎ�� represents the Herding behavior Coefficient, if (ℎ��= 0) herding behavior is absent. Conversely, if (ℎ��= 0), herding behavior is fully present, meaning individual stock returns move in sync with market portfolio returns. Hwang and Salmon (2004) proposed the State Space Model to measure herding behavior using the change in stock sensitivity to change in market return through the following two equations: Log [ 𝑆𝑆𝑆𝑆𝑆𝑆�(𝛽𝛽���� )] = 𝜇𝜇� + 𝐻𝐻�� + 𝜐𝜐�� (5) 𝐻𝐻�� = 𝜑𝜑� 𝐻𝐻�,��� + 𝜂𝜂�� (6) Where: 𝜂𝜂��~ iid (0,𝜎𝜎�,�� ) The model was estimated using the Kalman filter methodology within EViews software. The presence of herding behavior is inferred from the significant value of the variance of the error term 𝜂𝜂�� (𝜎𝜎�,�� ), and the significance of the persistence parameter �𝜑𝜑��. Cross-Sectional Standard deviation of betas is calculated by the following formula: 𝑆𝑆𝑆𝑆𝑆𝑆�(𝛽𝛽��) = ������ �������� ������������� ��� (7) STATISTICAL RESULTS OF HERDING BEHAVIOR The study sample was divided into two different portfolios, as follows: The model was estimated using the Kalman filter methodology within EViews software. The presence of herding behavior is inferred from the signifi- cant value of the variance of the error term 𝑟𝑟��� = 𝛼𝛼��� + 𝛽𝛽���� 𝑟𝑟��� + 𝜀𝜀��� (3) Where 𝑟𝑟��� and 𝑟𝑟��� represent the daily return of stock i and the EGX100 index during month t respectively, while 𝛽𝛽���� denotes the monthly beta for each stock. The regression analysis was performed 60 times per stock to obtain monthly Beta per stock, thus the regression analysis was performed 3,000 times for all stocks. The Hwang and Salmon (2004) methodology is based on the Cross-Sectional Standard deviation of betas. as follows: … �� ������ ������� = 𝛽𝛽���� =𝛽𝛽���  ℎ�� (𝛽𝛽���1) (4) Where 𝐸𝐸�� represents the expected return of stock i under behavioral bias, while 𝐸𝐸��𝑟𝑟��� denotes the expected return of the market portfolio. ℎ�� represents the Herding behavior Coefficient, if (ℎ��= 0) herding behavior is absent. Conversely, if (ℎ��= 0), herding behavior is fully present, meaning individual stock returns move in sync with market portfolio returns. Hwang and Salmon (2004) proposed the State Space Model to measure herding behavior using the change in stock sensitivity to change in market return through the following two equations: Log [ 𝑆𝑆𝑆𝑆𝑆𝑆�(𝛽𝛽���� )] = 𝜇𝜇� + 𝐻𝐻�� + 𝜐𝜐�� (5) 𝐻𝐻�� = 𝜑𝜑� 𝐻𝐻�,��� + 𝜂𝜂�� (6) Where: 𝜂𝜂��~ iid (0,𝜎𝜎�,�� ) The model was estimated using the Kalman filter methodology within EViews software. The presence of herding behavior is inferred from the significant value of the variance of the error term 𝜂𝜂�� (𝜎𝜎�,�� ), and the significance of the persistence parameter �𝜑𝜑��. Cross-Sectional Standard deviation of betas is calculated by the following formula: 𝑆𝑆𝑆𝑆𝑆𝑆�(𝛽𝛽��) = ������ �������� ������������� ��� (7) STATISTICAL RESULTS OF HERDING BEHAVIOR The study sample was divided into two different portfolios, as follows: , and the significance of the persistence parameter 𝑟𝑟��� = 𝛼𝛼��� + 𝛽𝛽���� 𝑟𝑟��� + 𝜀𝜀��� (3) Where 𝑟𝑟��� and 𝑟𝑟��� represent the daily return of stock i and the EGX100 index during month t respectively, while 𝛽𝛽���� denotes the monthly beta for each stock. The regression analysis was performed 60 times per stock to obtain monthly Beta per stock, thus the regression analysis was performed 3,000 times for all stocks. The Hwang and Salmon (2004) methodology is based on the Cross-Sectional Standard deviation of betas. as follows: … �� ������ ������� = 𝛽𝛽���� =𝛽𝛽���  ℎ�� (𝛽𝛽���1) (4) Where 𝐸𝐸�� represents the expected return of stock i under behavioral bias, while 𝐸𝐸��𝑟𝑟��� denotes the expected return of the market portfolio. ℎ�� represents the Herding behavior Coefficient, if (ℎ��= 0) herding behavior is absent. Conversely, if (ℎ��= 0), herding behavior is fully present, meaning individual stock returns move in sync with market portfolio returns. Hwang and Salmon (2004) proposed the State Space Model to measure herding behavior using the change in stock sensitivity to change in market return through the following two equations: Log [ 𝑆𝑆𝑆𝑆𝑆𝑆�(𝛽𝛽���� )] = 𝜇𝜇� + 𝐻𝐻�� + 𝜐𝜐�� (5) 𝐻𝐻�� = 𝜑𝜑� 𝐻𝐻�,��� + 𝜂𝜂�� (6) Where: 𝜂𝜂��~ iid (0,𝜎𝜎�,�� ) The model was estimated using the Kalman filter methodology within EViews software. The presence of herding behavior is inferred from the significant value of the variance of the error term 𝜂𝜂�� (𝜎𝜎�,�� ), and the significance of the persistence parameter �𝜑𝜑��. Cross-Sectional Standard deviation of betas is calculated by the following formula: 𝑆𝑆𝑆𝑆𝑆𝑆�(𝛽𝛽��) = ������ �������� ������������� ��� (7) STATISTICAL RESULTS OF HERDING BEHAVIOR The study sample was divided into two different portfolios, as follows: . Cross-Sectional Standard deviation of betas is calculated by the following formula: 𝑟𝑟��� = 𝛼𝛼��� + 𝛽𝛽���� 𝑟𝑟��� + 𝜀𝜀��� (3) Where 𝑟𝑟��� and 𝑟𝑟��� represent the daily return of stock i and the EGX100 index during month t respectively, while 𝛽𝛽���� denotes the monthly beta for each stock. The regression analysis was performed 60 times per stock to obtain monthly Beta per stock, thus the regression analysis was performed 3,000 times for all stocks. The Hwang and Salmon (2004) methodology is based on the Cross-Sectional Standard deviation of betas. as follows: … �� ������ ������� = 𝛽𝛽���� =𝛽𝛽���  ℎ�� (𝛽𝛽���1) (4) Where 𝐸𝐸�� represents the expected return of stock i under behavioral bias, while 𝐸𝐸��𝑟𝑟��� denotes the expected return of the market portfolio. ℎ�� represents the Herding behavior Coefficient, if (ℎ��= 0) herding behavior is absent. Conversely, if (ℎ��= 0), herding behavior is fully present, meaning individual stock returns move in sync with market portfolio returns. Hwang and Salmon (2004) proposed the State Space Model to measure herding behavior using the change in stock sensitivity to change in market return through the following two equations: Log [ 𝑆𝑆𝑆𝑆𝑆𝑆�(𝛽𝛽���� )] = 𝜇𝜇� + 𝐻𝐻�� + 𝜐𝜐�� (5) 𝐻𝐻�� = 𝜑𝜑� 𝐻𝐻�,��� + 𝜂𝜂�� (6) Where: 𝜂𝜂��~ iid (0,𝜎𝜎�,�� ) The model was estimated using the Kalman filter methodology within EViews software. The presence of herding behavior is inferred from the significant value of the variance of the error term 𝜂𝜂�� (𝜎𝜎�,�� ), and the significance of the persistence parameter �𝜑𝜑��. Cross-Sectional Standard deviation of betas is calculated by the following formula: 𝑆𝑆𝑆𝑆𝑆𝑆�(𝛽𝛽��) = ������ �������� ������������� ��� (7) STATISTICAL RESULTS OF HERDING BEHAVIOR The study sample was divided into two different portfolios, as follows: (7) HERDING BEHAVIOR AND STOCK MARKET EFFICIENCY… 2121 Statistical results of herding behaviorStatistical results of herding behavior The study sample was divided into two different portfolios, as follows: The portfolio shows herding behavior among investors, where herding behav- ior in this portfolio is statistically significant and consists of 18 stocks. A sec- ond portfolio, comprising 32 stocks, shows no statistically significant herding behavior. Statistical Results of Herding BehaviorStatistical Results of Herding Behavior The stock sample is divided into two portfolios, each tested for herding behav- ior. Tables 1 and 2 present the statistical results for both portfolios. Table 1. Statistical Results of the Space Model for the Portfolios in which Herding Behavior is Statistically Significant Coefficient Std. Error z-Statistic Prob. Variable Estimate C(1) μm -0.001063 0.089924 -0.011821 0.9906 μm -0.001063 C(2) vmt -3.140787 0.348612 -9.009405 0.0000 σmv 0.20796333 C(3) φm 0.867566 0.194713 4.455622 0.0000 φm 0.86757 C(4) ηmt -5.541811 1.936191 -2.862223 0.0042 σmη 0.06260529 S o u r c e : authors’ construction. Table 1 shows the state space model results for a portfolio consisting of 18 stocks of the sample. Coefficient c(1) represents (μm) in the state space model. The results indicated that it is not significant at the level of 5%. Coeffi- cient c(2) represents the error term (υmt), and coefficients c(3) and c(4) refer to the persistence parameter (φm) and the standard deviation of the state-equa- tion error (ηmt), respectively. The results indicated that they are statistically significant at the level of 5%, which confirms the presence of herding behavior. M. H. Abd-Allah, K. M. K. Mohamed, M. S. M. Mahmood, N. N. S. Mohamed2222 Table 2. Statistical Results of the Space Model for the Portfolios in which Herding Behavior is not Statistically Significant Coefficient Std. Error z-Statistic Prob. Variable Estimate C(1) μm -0.147764 0.019455 -7.595299 0.0000 μm -0.147764 C(2) vmt -21.7119 1.48E+10 -1.47E-09 1.0000 σmv 1.9289E-05 C(3) φm 0.023413 8.687411 0.002695 0.9978 φm 0.023413 C(4) ηmt -4.21498 373.3079 -0.011291 0.9910 σmη 0.12154266 S o u r c e : authors’ construction. Table 2 presents the state-space model results for a portfolio of 32 stocks. The findings indicate that coefficients c(3) and c(4), representing the persistence parameter (φm) and the standard deviation of the state-equation error (ηmt), are not statistically significant at the level of 5%, confirming the absence of herding behavior. Efficiency TestingEfficiency Testing Market efficiency was analyzed in two portfolios: one exhibiting statistically significant herding behavior and another where herding was not significant. Efficiency was assessed based on the weak form of efficient markets, as defined by Fama (1970), which states that stock prices should be normally distributed, independent, and unpredictable. The evaluation process involved the follow- ing steps: ■ Normality Test: The Shapiro-Wilk and Kolmogorov-Smirnov tests were used to check if stock prices followed a normal distribution; ■ Dickey-Fuller Unit Root Test: Applied to determine whether the time se- ries data was stationary or non-stationary; ■ Runs Test: Used to assess whether stock price movements followed a random pattern. HERDING BEHAVIOR AND STOCK MARKET EFFICIENCY… 2323 Descriptive StatisticsDescriptive Statistics Table 3 presents the descriptive statistics for both portfolios: one where herd- ing behavior is statistically significant and another where it is not. The results indicate that the time series for both portfolios exhibit right-skewed distribu- tions. The skewness coefficient is 0.546 for the portfolio with herding behavior and 0.463 for the portfolio without it. Table 3. Descriptive Statistics The portfolio shows herding behavior The portfolio does not show herding behavior Mean 19.6767 Mean 21.5867 Median 16.6518 Median 20.6888 Variance 74.068 Variance 122.926 Std. Deviation 8.60626 Std. Deviation 11.08721 Minimum 6.94 Minimum 6.43 Maximum 36.71 Maximum 46.69 Range 29.77 Range 40.26 Skewness 0.546 Skewness 0.463 Kurtosis -1.042 Kurtosis -0.871 S o u r c e : authors’ construction. Testing the Normal Distribution of Time SeriesTesting the Normal Distribution of Time Series The Shapiro-Wilk and Kolmogorov-Smirnov tests were performed to assess whether the time series follows a normal distribution. The results in table 4 show that Sig. Value is at less than 5%. Therefore, stock prices do not normally distribute and are independent for both the portfolio that shows herding be- havior and the portfolio that does not show herding behavior. Therefore, the two portfolios do not achieve a weak level of efficiency. M. H. Abd-Allah, K. M. K. Mohamed, M. S. M. Mahmood, N. N. S. Mohamed2424 Table 4. Time-Series Natural Distribution Test The portfolio shows herding behavior The portfolio does not show herding behavior Statistic df Sig. Statistic df Sig. Kolmogorov-Smirnov 0.156 1220 0.000 0.099 1220 0.000 Shapiro-Wilk 0.905 1220 0.000 0.905 1220 0.000 S o u r c e : authors’ construction. Testing Random Behavior of the Time SeriesTesting Random Behavior of the Time Series This study analyzes the randomness of the time series using various statistical methods, including: The Runs Test The Runs Test The Runs test is used to examine the randomness of data. Table 5 shows the re- sults of the Runs test. The results show that the p-value is less than 5%, as the p-value is equal to zero for both the portfolio showing herding behavior and the portfolio does not showing herding behavior. Thus, the time series of each portfolio is non-random, which indicates that the two portfolios do not achieve a weak level of efficiency. Table 5. The Runs Test Results The portfolio shows herding behavior The portfolio does not show herding behavior Test value 16.65 20.69 Number of Runs 14 12 Z-statistic -34.198 -34.313 p-value 0.000 0.000 S o u r c e : authors’ construction. HERDING BEHAVIOR AND STOCK MARKET EFFICIENCY… 2525 Unit Root Test for Dickey-FullerUnit Root Test for Dickey-Fuller The Unit Root Test for Dickey-Fuller, or the so-called Augmented Dickey-Fuller test (ADF test), is used to study the stationary and stability of the time series. If the time series is static and does not move randomly, this means that the mar- ket is inefficient according to the weak level. Table 6 shows the results of the Unit Root test, which indicates that the ADF test statistic is smaller than the critical values at 1%, 5%, and 10%. The value of the p-value is also less than 5%, as it is equal to zero for both portfolios showing herding behavior and those that do not show herding behavior. Therefore, these results show that there is no unit root in the return series of the two portfolios, which indicates incompe- tence at the weak level of each portfolio. Table 6. Unit Root Test for Dickey-Fuller The portfolio shows herding behavior The portfolio does not show herding behavior critical value 1% -3.969876 -3.969845 5% -3.415595 -3.41558 10% -3.130037 -3.130028 ADF test statistic -19.0147 -26.59111 p-value 0.0000 0.0000 S o u r c e : authors’ construction.  Conclusion Conclusion This paper examined the impact of herding behavior on market efficiency in the Egyptian stock market. The main sample of the research was divided into two sub-samples: the first sample is a portfolio of stocks showing herding behavior and the second sample is a portfolio of stocks that does not show herding be- havior. The efficiency was tested using the Runs test and the Dickey-Fuller Unit Root Tests, in addition to applying the Hwang and Salmon (2004) methodology to test herding behavior. Through statistical results, the time-series prices of both the portfolio showing herding behavior and the portfolio does not show herding behavior, are not going randomly, as the results of the Runs test and M. H. Abd-Allah, K. M. K. Mohamed, M. S. M. Mahmood, N. N. S. Mohamed2626 the Dickey-Fuller Unit Root Tests are consistent and supportive of each other. Therefore, the two portfolios do not achieve a weak level of efficiency. Thus, we can conclude that herding behavior among investors in the Egyptian stock market does not affect market efficiency. Therefore, other factors, rather than herding behavior, may influence the efficiency of the Egyptian stock market. There may be a relationship between herding behavior and inefficient fi- nancial markets, but due to several limitations in this research, this relation- ship was not as clear as the results showed. One of the limitations is the small size of the sample, it is advisable to use a larger sample size to achieve different results more accurately. Additionally, the period of study was only 5 years. For future studies, it is advisable to consider a study period longer than 5 years, as well as utilizing different measures of herding behavior, which may yield var- ied findings.  References References Abd-Alla, M.H. (2020). Sentimental herding: The role of COVID-19 crisis in the Egyptian stock market. Copernican Journal of Finance & Accounting, 9(3), 9–23. https://doi. org/10.12775/CJFA.2020.009. Abd El-Razcek, M.H., & Abd Elbaseet, A.S. (2023). Does the Egyptian Exchange Market still have herd behavior? Scientific Journal for Financial and Commercial Studies and Research, 4(1), 183–228. Banerjee, A.V. (1992). A simple model of herd behavior. The Quarterly Journal of Economics, 107(3), 797–817. https://doi.org/10.2307/2118364. Barberis, N., & Thaler, R. (2003). A survey of behavioral finance. In Handbook of the Economics of Finance (Vol. 1, pp. 1053–1128). https://doi.org/10.3386/w9222. Bikhchandani, S., Hirshleifer, D., & Welch, I. (1992). A theory of fads, fashion, custom, and cultural change as informational cascades. Journal of Political Economy, 100(5), 992–1026. https://doi.org/10.1086/261849. Bikhchandani, S., & Sharma, S. (2000). Herd behaviour in financial markets. IMF Staff Papers, 47(3), 279–310. Black, F., Jensen, M.C., & Scholes, M. (1972). The capital asset pricing model: Some em- pirical tests. In Studies in the theory of capital markets (pp. 79–121). Bo, H., Li, T., & Sun, Y. (2013). Board attributes and herding in corporate investment: Evidence from Chinese-listed firms. The European Journal of Finance, 22(4–6), 432–462. https://doi.org/10.1080/1351847x.2013.788536. Buckner, H.T. (1965). A theory of rumor transmission. Public Opinion Quarterly, 29(1), 54–70. https://doi.org/10.1086/267297. https://doi.org/10.2307/2118364 https://doi.org/10.3386/w9222 https://doi.org/10.1086/267297 HERDING BEHAVIOR AND STOCK MARKET EFFICIENCY… 2727 Chang, E.C., Cheng, J.W., & Khorana, A. (2000). An examination of herd behaviour in equity markets: An international perspective. Journal of Banking & Finance, 24(10), 1651–1679. https://doi.org/10.1016/s0378-4266(99)00096-5. Christie, W.G., & Huang, R.D. (1995). Following the Pied Piper: Do individual returns herd around the market? Financial Analysts Journal, 51(4), 31–37. Cootner, H. (1964). The random character of stock market prices. MIT Press. Devenow, A., & Welch, I. (1996). Rational herding in financial economics. European Economic Review, 40(3–5), 603–615. https://doi.org/10.1016/0014-2921(95)00073-9. El Mosallamy, D.A., & Gamal, N. (2024). Empirical evidence on testing the efficient mar- ket hypothesis in Egypt: Case of currency devaluation. MSA-Management Science Journal, 3(2), 1–27. https://doi.org/10.21608/msamsj.2024.257624.1049. Fama, E.F. (1965). Random walks in stock market prices. Financial Analysts Journal, 21(5), 55–59. https://doi.org/10.2469/faj.v21.n5.55. Fama, E.F. (1970). Efficient capital markets: A review of theory and empirical work. The Journal of Finance, 25(2), 383–417. https://doi.org/10.2307/2325486. Fama, E.F. (1976). Efficient capital markets: Reply. The Journal of Finance, 31(1), 143– 145. https://doi.org/10.2307/2326404. Fama, E.F., Fisher, L., Jensen, M.C., & Roll, R. (1969). The adjustment of stock prices to new information. International Economic Review, 10(1), 1–21. https://doi.org/10.2307/ 2525569. Fama, E.F., & MacBeth, J.D. (1973). Risk, return, and equilibrium: Empirical tests. Journal of Political Economy, 81(3), 607–636. French, K.R. (1980). Stock returns and the weekend effect. Journal of Financial Economics, 8(1), 55–69. https://doi.org/10.1016/0304-405x(80)90021-5. Gavriilidis, C. (2013). Essays on collective investor’s behaviour (Doctoral dissertation, Durham University). Guney, Y., Kallinterakis, V., & Komba, G. (2017). Herding in frontier markets: Evidence from African stock exchanges. Journal of International Financial Markets, Institutions and Money, 47, 152–175. https://doi.org/10.1016/j.intfin.2016.11.001. Hirshleifer, D. (2001). Investor psychology and asset pricing. The Journal of Finance, 56(4), 1533–1597. https://doi.org/10.1111/0022-1082.00379. Hirshleifer, D., & Teoh, S.H. (2003). Herd behaviour and cascading in capital markets: A review and synthesis. European Financial Management, 9(1), 25–66. https://doi. org/10.1111/1468-036x.00207. Hudson, Y., Yan, M., & Zhang, D. (2020). Herd behaviour & investor sentiment: Evidence from UK mutual funds. International Review of Financial Analysis, 71, 101494. htt- ps://doi.org/10.1016/j.irfa.2020.101494. Hwang, S., & Salmon, M. (2004). Market stress and herding. Journal of Empirical Finance, 11(4), 585–616. https://doi.org/10.1016/j.jempfin.2004.04.003. Jensen, M.C. (1978). Some anomalous evidence regarding market efficiency. Journal of Financial Economics, 6(2–3), 95–101. https://doi.org/10.1016/0304-405X(78)90025-9. https://doi.org/10.1016/s0378-4266(99)00096-5 https://doi.org/10.1016/0014-2921(95)00073-9 https://doi.org/10.2469/faj.v21.n5.55 https://doi.org/10.2307/2325486 https://doi.org/10.2307/2326404 https://doi.org/10.2307/2525569 https://doi.org/10.2307/2525569 https://doi.org/10.1016/0304-405x(80)90021-5 https://doi.org/10.1016/j.intfin.2016.11.001 https://doi.org/10.1111/0022-1082.00379 https://doi.org/10.1111/1468-036x.00207 https://doi.org/10.1111/1468-036x.00207 https://doi.org/10.1016/j.irfa.2020.101494 https://doi.org/10.1016/j.irfa.2020.101494 https://doi.org/10.1016/j.jempfin.2004.04.003 https://doi.org/10.1016/0304-405X(78)90025-9 M. H. Abd-Allah, K. M. K. Mohamed, M. S. M. Mahmood, N. N. S. Mohamed2828 Kauffman, R.J., & Li, X. (2003). Payoff externalities, informational cascades and mana- gerial incentives: A theoretical framework for IT adoption herding. In Proceedings of the 2003 INFORMS Conference on IS and Technology, Atlanta, GA, October 2003. Kendall, M.G., & Hill, A.B. (1953). The analysis of economic time-series—Part I: Prices. Journal of the Royal Statistical Society. Series A (General), 116(1), 11–34. https://doi. org/10.2307/2980947. Le Bon, G. (1895). The crowd: A study of the popular mind. London. Litimi, H. (2017). Herd behavior in the French stock market. Review of Accounting and Finance, 16(4), 497–515. https://doi.org/10.1108/raf-11-2016-0188. Messaoud, D., & Ben Amar, A. (2025). Herding behaviour and sentiment: Evidence from emerging markets. EuroMed Journal of Business, 20(2), 552–573. https://doi. org/10.1108/EMJB-08-2023-0209. Ozkan, O. (2021). Impact of COVID-19 on stock market efficiency: Evidence from devel- oped countries. Research in International Business and Finance, 58, 101445. https:// doi.org/10.1016/j.ribaf.2021.101445. Rozeff, M.S., & Kinney, W.R. (1976). Capital market seasonality: The case of stock re- turns. Journal of Financial Economics, 3(4), 379–402. https://doi.org/10.1016/0304- 405x(76)90028-3. Samuelson, P.A. (1965a). Proof that properly anticipated prices fluctuate randomly. Industrial Management Review, 6(2), 41–49. https://doi.org/10.1142/97898145 66926_0002. Samuelson, P.A. (1965b). Rational theory of warrant pricing. Industrial Management Review, 6(2), 13–39. https://doi.org/10.1007/978-3-319-22237-0_11. Schindler, M. (2007). Rumors in financial markets: Insights into behavioral finance. Chichester, England: John Wiley & Sons. Shiller, R.J. (1999). Human behaviour and the efficiency of the financial system. In Handbook of macroeconomics, 1, 1305–1340. https://doi.org/10.1016/S1574-0048 (99)10033-8. Yao, J., Ma, C., & He, W.P. (2014). Investor herding behaviour of Chinese stock market. International Review of Economics & Finance, 29, 12–29. http://dx.doi.org/10.1016/j. iref.2013.03.002. https://doi.org/10.2307/2980947 https://doi.org/10.2307/2980947 https://doi.org/10.1108/raf-11-2016-0188 https://doi.org/10.1108/EMJB-08-2023-0209 https://doi.org/10.1108/EMJB-08-2023-0209 https://doi.org/10.1016/j.ribaf.2021.101445 https://doi.org/10.1016/j.ribaf.2021.101445 https://doi.org/10.1016/0304-405x(76)90028-3 https://doi.org/10.1016/0304-405x(76)90028-3 https://doi.org/10.1142/9789814566926_0002 https://doi.org/10.1142/9789814566926_0002 https://doi.org/10.1007/978-3-319-22237-0_11 https://doi.org/10.1016/S1574-0048(99)10033-8 https://doi.org/10.1016/S1574-0048(99)10033-8 http://dx.doi.org/10.1016/j.iref.2013.03.002 http://dx.doi.org/10.1016/j.iref.2013.03.002