Document Type : Research Paper

Authors

• Iman Dadashi ¹
• Vahid Omidi ²

¹ Assistant Professor, Department of Accounting, Faculty of Economics Sciences and Administrative, University of Qom, Qom, Iran

² Assistant Professor, Department of Economics, Faculty of Economics Sciences and Administrative, University of Qom, Qom, Iran

Abstract

Considering the impact of global variables on stock market industries, the present study relied on the Quantile-on-Quantile Connectedness (QQC) and Structural Vector Autoregression (SVAR) to examine the impact of geopolitical risk fluctuations on the volatility of the petroleum products, chemical products, metal ores, and basic metals sectors in the Tehran Stock Exchange. The analysis focused on the period from January 1, 2020, to December 24, 2024. The results from the QQC model revealed that fluctuations in geopolitical risk exhibited the strongest correlation with the volatility of the petroleum products industry index at extreme deciles, indicating a significant impact. In other industries, the highest susceptibility to geopolitical risk fluctuations had occurred when their volatility was in the 9th and 10th deciles. In addition, the SVAR model results indicated that the immediate response of industry index volatility to geopolitical risk shocks was positive across all cases. Over 360 periods, this response converged to a positive value, reflecting the persistence of the shock. The cumulative response analysis further demonstrated an exponential increase in all industries, suggesting a rising trend in the effect of geopolitical risk over time. Specifically, after 360 periods, the volatility of the petroleum products industry index increased by 0.34, chemical products by 0.06, metal ores by 0.03, and basic metals by 0.06.

Introduction

Recently, the Tehran Stock Exchange (TSE) has been grappling with various risk factors, including the government budget, uncertainties in domestic and foreign policies, the Al-Aqsa Storm and Promise Fulfilled operations, interest rates, the exchange rate, and inflation. Notably, the TSE has not consistently mirrored the behavior of global markets across different periods. For instance, at the height of the COVID–19 pandemic, when most global stock markets experienced significant downturns, the TSE reached historic record highs. Conversely, at times when global markets were on the rise and commodity prices increased, the TSE entered a decline. This divergence was primarily due to internal risks unique to the TSE, which prevented the domestic market from benefiting from global market growth. The present study aimed to examine the impact of geopolitical risk fluctuations on the price index volatility of selected industries listed in the TSE. The industries were selected based on their specific characteristics and their sensitivity to geopolitical risks.

Materials and Methods

The study employed the Quantile-on-Quantile Connectedness (QQC) model to examine the relationship between the overall stock index and Islamic Treasury Bonds (Sukuk). To this end, the QVAR(P) model, which enables the estimation of relationships across different quantiles, is utilized as follows:
                                                                  (1)
In this equation,   and  represent the vector of endogenous variables with a  dimension of . The vector τ denotes the quantiles within the range [0,1], while P indicates the lag order of the QVAR model. Additionally, μ(τ) is the  vector of conditional means,  is the  coefficient matrix, and  is the  vector of error terms. Subsequently, the Generalized Forecast Error Variance Decomposition (GFEVD) for an F-step-ahead forecasting, which represents the impact of a shock in series j on series i, is expressed as follows:
                                           (2)
In this equation,  denotes the  variance-covariance matrix of the error terms. The vector  is the standard basis vector or unit vector of dimension , with its the i-th element equal to one and all other elements set to zero.
In this case, the rows of  do not sum to one. Therefore,  is standardized to obtain the scaled GFEVD:
                                                              (3)
Using this, the overall adjusted connectedness index (quantile-to-quantile) is calculated as follows:
                               (4)
In Equation (4), the higher the Total Connectedness Index (TCI), the higher the market risk.
The analysis also used the Structural Vector Autoregression (SVAR) model. In the QQC model, the volatility of geopolitical risk was analyzed in relation to each of the other variables in the model, with results extracted accordingly. The SVAR model followed the same principle. Consequently, four models were estimated.
The VAR model in this study is represented in its general form as follows:
                                                           (5)
Where  is a vector containing the volatility of geopolitical risk and the index of each industry analyzed individually. The matrices  to  contain the coefficients of the lagged variables, and  represents the residuals, which follow a normal distribution with zero mean and covariance ​. However, the shocks derived from Model (5) are not structural. To address this, the following model is used, allowing constraints to be imposed on matrices A and B:
                                                                  (6)
In Equation (10),  represents the structural error terms. The relationship between the VAR and SVAR models is expressed as .

Results and Discussion

The results indicated that geopolitical risk had a significant and varying impact on different industries within the TSE. This impact is influenced not only by each industry’s volatility level but also by the distribution of risk quantiles and industry indices. The QQC results revealed that the petroleum products industry was the most sensitive to geopolitical risk, particularly in extreme quantiles, where its connection to geopolitical risk reaches its peak. This finding suggests that during periods of high volatility, risk transmission accelerates. Similarly, in the chemical, metal ore, and basic metals industries, increased volatility heightened their susceptibility to geopolitical risk shocks. Notably, when these industries experience higher volatility quantiles, their connection to geopolitical risk strengthens across all levels. Structural shock analysis using the SVAR model indicated that all industries exhibited a positive immediate response to geopolitical risk volatility shocks. This reaction is strongest in the short term and gradually weakens over time. Among the industries analyzed, the petroleum products sector displayed the highest sensitivity, with an increase of 1 unit, while the impact on the chemical products, metal ore, and basic metals industries was 0.6, 0.3, and 0.5 units, respectively.

Conclusion

According to the findings, the relationship between geopolitical risk and the petroleum products industry is strongest in extreme quantiles. For other industries, the QQC model identifies two key patterns: first, when geopolitical risk volatility is in the 9th and 10th quantiles, it has the greatest impact on these industries; second, when the industries’ own volatility is in the 9th and 10th quantiles, they show the highest susceptibility to geopolitical risk across all quantiles. In addition, the results from the SVAR model indicated that the impact of geopolitical risk shocks on these industries would remain positive even after 360 periods. In other words, geopolitical risk shocks have a lasting effect on the volatility of the industries analyzed in this study.

Keywords

• Geopolitical Risk
• Industry Indices
• Petroleum Products
• Chemical Products
• Metal Ores
• Basic Metals
• QQC
• SVAR

Main Subjects

• Risk