Journal of Asset Management and Financing (Sep 2025)
Analysis of Herd Behavior in Industry Groups of the Tehran Stock Exchange
Abstract
This study aims to investigate herd behavior across nine industry groups within the Tehran Stock Exchange from 2015 to 2023. Utilizing the methodology proposed by Chang et al. (2000), we analyze daily and weekly fluctuations in the market to assess herd behavior during periods of market volatility. The findings reveal significant herd behavior in nearly all industry groups, suggesting a pervasive phenomenon across the overall market. Notably, herd behavior is predominantly observed during upward trends within both daily and weekly time frames, potentially contributing to stock market surges, such as the notable rise from 2018 to mid-2019, which led to substantial price bubbles. Furthermore, by segmenting the study period into downward and upward phases, we explore the symmetry of herd behavior, revealing asymmetries in many industry groups.Keywords: Behavioral Finance, Herd Behavior, Total Index, Equal-Weighted Index, Tehran Stock Exchange IntroductionBarber and Odean (1999) introduced behavioral finance as a framework that elucidates irrational investor behaviors and enhances our understanding of inefficiencies in financial markets. A key concept within this domain is herd behavior, which emerged in the literature during the early 1990s. For instance, Banerjee (1992) explored herd behavior in abstract settings, illustrating how once a certain number of brokers favored a particular option, subsequent brokers tended to imitate this choice while disregarding their own information. The investigation of herd behavior in the Tehran Stock Exchange (TSE) has become increasingly relevant in recent years due to the market's experience with various currency crises and multiple upward and downward trends. Over the past decade, the TSE has undergone significant growth, marked by a dramatic increase in the number of active trading codes and a sharp rise in their trading values, underscoring the necessity of examining herd behavior in this context. Consequently, this study aims to investigate herd behavior within the TSE across nine distinct industry groups. Analyzing herd behavior by industry is particularly important given the high correlation among firms within each group and the simultaneous influence of macroeconomic news on companies operating in the same sector. Materials & MethodsChristie and Huang (1995) were pioneers in the empirical study of herd behavior in financial markets, employing an econometric approach to illustrate that the decision-making processes of market participants are influenced by prevailing market conditions. Building on their work, Chang et al. (2000) introduced a new model for identifying herd behavior, positing that investors often lose confidence during stressful periods, such as market bubbles or downturns, which leads them to follow prevailing market trends. The methodology for detecting herd behavior as proposed by Chang et al. (2000) is encapsulated in Equation 3. 𝐶𝑆A𝐷𝑡 = 𝛾0 + 𝛾1 |𝑅𝑚,t | + 𝛾2 R 2𝑚,t + 𝜀𝑡 (1)In Equation 3, CSADt represents the cross-sectional absolute deviation, which is utilized to measure the dispersion of stock returns relative to the average market return. The calculation of the cross-sectional absolute deviation is detailed in Equation 4. (2) In whick, Rm,t and Ri,t denote the average market return and the return of stock i at time t, respectively. N refers to the number of companies within the relevant industry selected for estimating herd behavior in that sector. The variable t represents the time frame used for calculating returns and absolute deviations; this study employs both daily and weekly intervals to analyze herd behavior. Additionally, it is important to note that the total index return and the equal-weighted index return are utilized as measures of the average market return in the current analysis. FindingsThe results of the analysis of herd behavior within industry groups, utilizing both the total index and the equal-weighted index, are summarized in Table 1. Table (1): Summary of results related to herd behavior in industry groupsEqual-weighted index Industry GroupTotal daily periodTotal weekly periodDaily upwardWeekly upwardDaily downwardWeekly downward Pharmaceutical companies●●○●●○ Basic Metals companies●○○○●○ Sugar production companies●○○○●○ Food Production companies●○○○●○ Automobile manufacturing companies●●●●●○ Oil refining companies○○●●●○ Investment companies●○○●●○ Chemical companies●●○●●○ Cement companies●●○●●○ Percentage of herd behavior presence894422671000 Total index Industry groupTotal daily periodTotal weekly periodDaily upwardWeekly upwardDaily downwardWeekly downward Pharmaceutical companies●●●●●○ Basic Metals companies●○●○●○ Sugar production companies●○●●●○ Food Production companies●●●●●○ Automobile manufacturing companies●●●●●○ Oil refining companies●○●○●● Investment companies●○●●●○ Chemical companies●●●●●○ Cement companies●●●●●○ Percentage of herd behavior presence100561007810011 Indicates the presence of herd behavior and ○ indicates the absence of herd behavior.According to the results reported in Table 1, evidence of herd behavior is observed in nearly all studied groups at least during one period, suggesting a prevalent presence of herd behavior across the Tehran Stock Exchange. When using the equal-weighted index as the average market return, the most significant herd behavior was identified in the group of automobile and parts manufacturers, followed by pharmaceuticals, cement, and chemical companies. Notably, no herd behavior was detected in any group during the downward trend of the weekly period. In contrast, when employing the total index as the average market return, the groups that exhibited the most herd behavior included pharmaceuticals, food, automobile and parts manufacturing, chemicals, and cement. Additionally, herd behavior was only observed in the oil refining group during the downward trend of the weekly period. Overall, both indices indicated that herd behavior was more pronounced in the daily time frame compared to the weekly time frame. The separation of the entire period into bullish and bearish phases revealed that when using the overall index as the average market return, herd behavior was more prevalent in both bullish and bearish daily time frames, indicating symmetry in the occurrence of herd behavior during this interval. In contrast, employing the equal-weighted index did not yield this symmetry; instead, more intense herd behavior was noted during the bearish daily time frame. Furthermore, across both indices, a greater degree of herd behavior was observed in the bullish phase compared to the bearish phase within the weekly time frame. Discussion & conclusionThe results demonstrate significant herd behavior across nearly all industry groups, indicating its pervasiveness within the overall market. The findings reveal that herd behavior is most pronounced during upward trends in both daily and weekly time frames, which may contribute to stock market increases, such as the notable rise from 2018 to mid-2019 that led to substantial price bubbles. Additionally, the analysis identifies the highest levels of herd behavior within the automobile manufacturing sector, while the oil refining sector exhibits the lowest. The results further indicate that herd behavior is more prevalent in the daily time frame compared to the weekly time frame, aligning with the emotional contagion aspect of this phenomenon. By partitioning the entire study period into upward and downward phases, the investigation of symmetry in herd behavior reveals asymmetry in many groups. This asymmetry may be attributed to market one-sidedness or cognitive factors such as loss aversion, as suggested by Kahneman and Tversky's Prospect Theory.
Keywords
