Energy-Related Uncertainty and Idiosyncratic Return Volatility: Implications for Sustainable Investment Strategies in Chinese Firms

Abstract

This study examines the impact of energy-related uncertainty on idiosyncratic volatility (IVOL) in Chinese firms, leveraging data from the Shanghai and Shenzhen stock exchanges between 2007 and 2022. Utilizing the Energy-Related Uncertainty Index (EUI) and the Fama–French five-factor model, we analyze a comprehensive dataset of 20,998 firm-year observations to understand how macroeconomic uncertainties specific to the energy sector influence firm-specific risk. Our findings reveal that a one-unit increase in the EUI is associated with a 5.1% rise in idiosyncratic volatility across all firms, underscoring the significant impact of energy-related uncertainty on firm-specific risks. The effect is more pronounced in energy-related firms, where a one-unit increase in the EUI leads to a 6.4% increase in IVOL, compared to a 3.7% increase in non-energy-related firms. By incorporating industry-wise, heterogeneity, and phase-based analyses, our findings reveal significant variations in the EUI’s impact across energy and non-energy sectors. State-owned enterprises, firms with high ownership concentration, and smaller firms are more vulnerable to energy uncertainties. Additionally, the effect of the EUI on IVOL is more pronounced during periods of high uncertainty. These insights have important implications for sustainable investment strategies, risk management, and policymaking, providing a deeper understanding of the intricate dynamics of energy markets in fostering sustainable economic growth and development.

1. Introduction

The volatility of stock returns, particularly idiosyncratic volatility (IVOL), has been a critical area of interest in finance due to its implications for portfolio management, asset pricing, and corporate finance [1]. As investors increasingly prioritize sustainability, understanding IVOL becomes essential in evaluating how sustainable practices and green investments impact firm-specific risks. Idiosyncratic volatility refers to the portion of a stock’s total volatility that is not explained by market-wide factors and is unique to the individual firm. Research suggests that sustainability practices, such as corporate social responsibility (CSR) initiatives, can influence idiosyncratic volatility. Firms engaging in sustainability efforts may experience unique volatility patterns due to the distinct risks and opportunities associated with their environmental and social initiatives [2,3,4]. By understanding the determinants of IVOL, investors can better align their portfolios with sustainable development goals, ensuring that investment decisions contribute to a more resilient and equitable economy [5].

In recent years, market dynamics, regulatory changes, and geopolitical considerations are just a few of the uncertainties that have caused substantial volatility in the energy sector. These uncertainties, in particular, energy-related uncertainty, have emerged as significant factors influencing sustainability in the energy sector. Energy prices are inherently volatile, influenced by factors such as geopolitical tensions, supply disruptions, and changes in demand [6]. This volatility can create substantial uncertainty for firms, especially those heavily reliant on energy inputs, affecting their profitability, investment decisions, and stock prices [7]. However, the Energy-Related Uncertainty Index (EUI), introduced by Dang et al. [8], provides a new way to quantify these uncertainties by taking into account changes in energy costs, policy dynamics, and advances in technology. This index is essential for understanding the intricate effects of energy-related uncertainties on idiosyncratic volatility (IVOL), particularly in the Chinese market. Although prior research [9,10,11] has focused on the role of macroeconomic uncertainties, such as policy uncertainty, geopolitical risks, and commodity price volatility, in driving idiosyncratic risk, there is a notable gap in studies specifically addressing how the EUI affects the idiosyncratic volatility of Chinese firms. This study aims to fill this gap.

The EUI is favored because it offers invaluable insights for sustainable investment strategies. It highlights the importance of understanding and managing energy-related uncertainties to promote sustainable economic practices. By quantifying the specific risks associated with energy markets, the EUI enables investors to make informed decisions that support long-term environmental and economic goals. The ability of the EUI to capture uncertainties from supply disruptions, policy changes, and geopolitical concerns provides investors with critical information to assess the sustainability of their investments. Unlike general uncertainty measures such as the economic policy uncertainty index [12], the financial uncertainty index [13], government policy uncertainty [10], and oil price uncertainty [14], the EUI offers specialized insights into the dynamics of the energy market, making it possible to evaluate energy-specific shocks directly. This makes it especially valuable for financial economics studies where comprehending the immediate effects of energy-related events on firm-specific risks is crucial.

We select China as the target country for our study because the Chinese market presents a unique context for examining the effects of energy-related uncertainty. As the world’s second-largest economy and a major consumer of energy, China is particularly vulnerable to energy price fluctuations [15]. The difference between China’s energy output and consumption since the 2000s, as seen in Figure 1, indicates the country’s increasing reliance on energy imports due to its fast industrialization and economic growth. This pattern, which shows a steady rise in energy consumption over output, emphasizes China’s rising demand and its susceptibility to changes in the world energy market. China’s intense industrial activity and urbanization are reflected in the growing disparity between energy production and demand, highlighting the need of researching how global energy uncertainties affect China’s corporate investment environment [8]. Furthermore, China’s energy market, characterized by substantial government intervention and regulation, presents unique challenges and opportunities from a sustainability and sustainable investment perspective. The heavy involvement of the government in shaping energy policies ensures that environmental considerations are often at the forefront of energy development strategies. This regulatory landscape aims to promote renewable energy sources, reduce carbon emissions, and achieve long-term environmental goals. However, this also adds a layer of complexity and uncertainty for firms operating within the market. Companies must navigate an intricate web of policies and compliance requirements that can change rapidly in response to shifting political priorities and international sustainability commitments. For sustainable investors, this dynamic environment requires careful analysis and strategic adaptation to align investments with both current regulations and future trends in sustainable energy development. The intricate balance of regulatory compliance and sustainability objectives underscores the importance of strategic foresight and flexibility in sustainable investment strategies within China’s energy sector [16,17]. Therefore, understanding how energy-related uncertainty impacts Chinese firms can provide valuable insights for both domestic and international investors.

This study aims to evaluate the influence of the Energy-Related Uncertainty Index (EUI) on the idiosyncratic volatility of corporations. Additionally, it investigates the diverse consequences of this influence on companies operating both within and outside the energy-related industry. The central research questions are as follows: (1) How does energy-related uncertainty impact idiosyncratic volatility in Chinese firms? (2) Do these effects vary for businesses that are involved in the energy sector versus those who are not? Using a panel regression model, we examine data covering 2332 Chinese non-financial listed companies between 2007 and 2022. The data consist of 20,998 firm-year observations (of which 13,575 are from the energy sector and 7423 are from non-energy).

Our estimation of IVOL is based on the Fama–French five-factor model, and the macroeconomic uncertainty associated with energy markets is captured by the EUI. Our regression analysis reveals several key findings. First, we find that a one-unit increase in the EUI is associated with a 5.1% increase in idiosyncratic volatility across all firms, highlighting the significant effect of energy-related uncertainty on firm-specific risks. Second, the effect is particularly noticeable in companies involved in the energy sector, where an increase in the EUI of one unit causes an increase in idiosyncratic volatility of 6.4%. In contrast, non-energy-related firms experience a smaller increase of 3.7% in idiosyncratic volatility per unit increase in the EUI. These results are derived from a robust panel regression model that controls for various firm-specific characteristics and macroeconomic factors. These insights are crucial for stakeholders in making informed decisions regarding risk management and strategic planning in an increasingly volatile energy market, thereby contributing to the broader goals of sustainability and sustainable development. Moreover, our findings show notable differences in the impact of the EUI across firms and industries by combining industry-wise, heterogeneity, and phase-based analysis. Energy uncertainty increases the vulnerability of state-owned organizations, firms with a large ownership concentration, and smaller firms. Furthermore, at times of high uncertainty, the impact of the EUI on IVOL is particularly prominent.

This study examines the widespread effect of energy-related uncertainty on firm-level idiosyncratic volatility, standing at the nexus of energy economics and corporate behavior. By meticulously employing the comprehensive Energy-Related Uncertainty Index, our work both aligns with and extends the scope of earlier research in the field of energy economics, which has traditionally focused on the impacts of energy policy and market uncertainties. More specifically, our study advances the knowledge of energy economics by assessing the impact of uncertainty on China’s idiosyncratic volatility using the recently developed EUI. Even though it is in the context of developing economies like China, to the best of our knowledge, no study has examined the relationship between the EUI and idiosyncratic volatility, thereby adding a substantial amount of knowledge. Our work is different from other studies that have looked at how uncertainties affect economic outcomes, such as oil-price [14,18], energy policy [19], and climate policy [20]. This is because our work applies the EUI. By applying the EUI, our research provides a more comprehensive framework for understanding the impact of energy-related uncertainty, which integrates several aspects of energy-related concerns. By focusing on firm-level effects rather than on more broad economic indicators, it builds on the work of Hou et al. [21] and supports findings from Adams et al. [22]. It also explores how marketization levels interplay with policy uncertainty in the energy sector. Additionally, the EUI’s comprehensiveness enables a thorough analysis that can effectively capture the complexity of energy dynamics, making it an invaluable tool for stakeholders navigating the volatile energy landscape [18,19].

Our second contribution is to establish the connection between idiosyncratic volatility and the EUI. Our research focuses on uncertainties particular to energy, while previous studies, e.g., [10,11], have highlighted the more general idea of uncertainty affecting stock returns. The results of Dang et al. [8], which emphasize the significance of energy uncertainty in corporate decision-making, are consistent with this study. By quantifying the positive effect of EUI on IVOL, our study offers empirical support for the theoretical assumptions of the Risk–Return Tradeoff Theory. The results imply that firms do indeed experience higher idiosyncratic volatility given heightened energy uncertainty. This emphasizes the relevance of our findings during a time of substantial market uncertainty in the energy markets and is consistent with a study by Bali et al. [23], which found that market uncertainties raise firm-specific risks.

Thirdly, we investigate how the EUI affects energy-related vs. non-energy-related firms differently. Although a number of studies, like those by Gulen and Ion [24], have shown how uncertainties affect business volatility generally, our analysis distinguishes between the relative impacts on different industries. Here, we contribute by showing how sector-specific risk dynamics can be significantly impacted by how sensitive an industry is to energy markets, which in turn improves our understanding of sector-specific risk dynamics. This work is noteworthy because it deepens our knowledge of the industry-specific effects of uncertainty measures, especially with regard to energy-intensive firms—a vital area of study that has drawn interest recently from numerous researchers, e.g., refs. [25,26], on the uncertainty of climate policy in Chinese energy companies. This study extends the existing literature by incorporating additional analyses to capture the influence of the EUI across different industries, phases of uncertainty, and firm characteristics. Our industry-wise analysis highlights significant variations in the EUI’s impact on IVOL across energy-related and non-energy-related sectors. The heterogeneity analysis reveals that state-owned enterprises, firms with high ownership concentration, and smaller firms are particularly sensitive to energy uncertainties. Furthermore, we analyze the effects of the EUI during periods of high and low uncertainty, providing comprehensive insights into how energy-related risks affect firm-specific volatility.

Finally, we contribute to the literature on the determinants of idiosyncratic volatility, which has identified several key factors, including firm size, growth potential, leverage, and profitability [27]. Larger firms typically have more diversified operations and stable cash flows, reducing their idiosyncratic risk [28]. High book-to-market firms, often classified as value firms, also tend to exhibit lower volatility due to their stable earnings and stronger financial health [29]. Leverage, on the other hand, increases financial risk and, consequently, idiosyncratic volatility [30]. By employing the Fama–French five-factor model to estimate IVOL and using the EUI as a macroeconomic factor, this research seeks to provide a nuanced understanding of how energy-related uncertainties affect firm-specific risks. Our study also extends this literature by incorporating energy-related uncertainty as a macroeconomic factor influencing IVOL. By comparing the impacts on energy-related and non-energy-related firms, we aim to provide a nuanced understanding of how sector-specific characteristics modulate the effects of macroeconomic uncertainty. This sectoral comparison is particularly relevant given the diverse nature of the Chinese economy and the varying degrees of energy dependence across industries [31].

Overall, these contributions provide relevant insights into the increasingly unpredictable energy market, while also improving our understanding of the complexity involved in idiosyncratic volatility in the context of energy uncertainty. The findings are expected to greatly impact the formulation of risk management strategies and policy interventions for investors, corporate managers, and policymakers. As they traverse the intricate dynamics of energy reliance and economic growth in China’s quickly changing economy, these insights will be extremely helpful.

The following is the format of the sections that follow in this paper: There is a thorough analysis of pertinent literature in Section 2. An overview of the dataset, variable definitions, and econometric model are provided in Section 3. The results are discussed in Section 4. The discussion on the robustness testing is provided in Section 5. Conclusions, limitations, and future directions are presented in Section 6.

2. Literature Review and Research Hypotheses

Our literature review is structured to first define idiosyncratic volatility (IVOL) and its significance in financial markets, particularly through the perspectives of informed arbitrage and noise trading. This foundation is critical for understanding how IVOL operates within the broader context of market behaviors and investor actions. Following this, we delve into the specific relationship between energy-related uncertainty (EUI) and IVOL, framing our discussion through the lenses of real options theory, information asymmetry, and behavioral finance. These theories provide a robust framework for analyzing how fluctuations in energy markets uniquely impact firm-specific risks. Additionally, we discuss key developments in China’s energy sector, offering a contextual background that underscores the relevance of studying the EUI in the Chinese market. By synthesizing these diverse strands of literature, we not only establish the theoretical basis for our hypotheses but also illustrate the multifaceted nature of IVOL and its determinants. This interconnected approach highlights our study’s contribution to understanding the nuanced impacts of the EUI on IVOL, particularly in a dynamic and rapidly evolving market like China.

2.1. Idiosyncratic Volatility Defined

Idiosyncratic volatility represents variations in a firm’s stock returns that are not attributed to the fundamental market characteristics that are usually included in asset pricing models. This form of volatility is crucial as it encapsulates the firm-specific uncertainties that do not correlate with the broader market movements. IVOL is crucial as it represents a firm-specific risk that is not explained by market movements, and it has implications for portfolio diversification, asset pricing, and risk management. Research in this area is substantial, providing two prevailing interpretations: the informed arbitrage and noise-trading perspectives.

2.1.1. The Informed Arbitrage Perspective

The perspective of informed arbitrage proposes that heightened idiosyncratic volatility signals that stock prices incorporate a greater amount of firm-specific information. According to this view, the connection implies that idiosyncratic volatility rises as external investors leverage firm-specific data to predict expected cash flows, aiming for arbitrage opportunities.

Transparency in information greatly reduces uncertainty and the cost of acquiring information, which makes informed arbitrage possible. The theory suggests that more transparent information environments encourage investors to engage in arbitrage that more closely aligns stock prices with their fundamental values. Research indicates that stronger property rights, which enhance the security of information, lead to more active informed arbitrage and subsequently higher idiosyncratic volatility. This correlation is evident in diverse markets, highlighting the role of legal and institutional frameworks in influencing stock market behavior [32]. Some studies, such as Jin and Myers [33], indicate that increased information opacity, characterized by limited availability or obscured firm-specific details, tends to decrease idiosyncratic volatility. This implies an inverse relationship between the accessibility of transparent information and idiosyncratic risk. According to Ferreira and Laux [9], institutional investors are discouraged from obtaining personal data for arbitrage trading by anti-acquisition protections. Their findings show that companies with strong anti-acquisition protection mechanisms see less idiosyncratic risk. Conversely, the presence of block ownership, often seen as a proxy for greater transparency and a concentrated monitoring mechanism, has been linked with increased idiosyncratic volatility. This relationship underscores the importance of ownership structure in enhancing firm-specific information flow and arbitrage activities. Moreover, diverse board compositions, such as those with gender diversity, have been associated with enhanced firm-specific information transparency and higher idiosyncratic volatility, suggesting that governance structures impact information dynamics and market behavior [34].

In the similar context, a recent study by Gu et al. [35] on the Chinese stock market highlights how trading constraints, framed as limits of arbitrage, significantly influence idiosyncratic volatility. They find that these constraints exacerbate the negative idiosyncratic volatility return premium, illustrating the complex interaction between market structure and volatility. Another study by Li et al. [36] investigates the impact of firm-specific news on idiosyncratic volatility in China, revealing that the association between information flow and idiosyncratic risk is nuanced by the nature of the news and the broader economic context.

The informed arbitrage argument, substantiated by extensive historical and recent empirical studies [32,33,35,36], offers a sturdy framework for comprehending the positive correlation between uncertainty and idiosyncratic volatility. This association underscores the pivotal influence of factors such as information transparency, legal frameworks, institutional structures, and market mechanisms on the dynamics of idiosyncratic volatility. Ongoing research in various global contexts further enhances this comprehension, affirming the significance of informed arbitrage in modern financial markets.

2.1.2. The Noise-Trading Perspective

In the context of financial markets, the noise-trading perspective asserts a substantial role for idiosyncratic volatility beyond merely conveying market information. It also highlights the pervasive influence of noise trading, which complicates the process of informed arbitrage. This theory aligns with the foundational viewpoint presented by De Long et al. [37], suggesting that return volatility encompasses risks associated with both noise and informational signals.

The core argument of noise trading suggests that idiosyncratic volatility reflects not only the actions of informed traders but also the erratic movements caused by noise traders—those who operate without relying on fundamental information. This duality creates a market environment where the clarity of information is perpetually challenged by the chaos of speculation and rumor. As outlined by Morck and Yeung [38], idiosyncratic volatility encapsulates this dual role, serving as the battleground where noise-driven mispricing clashes with informed speculative adjustments.

As this discourse progresses into contemporary research, it unveils nuanced perspectives. Aabo et al. [39], for instance, explore the impact of heightened idiosyncratic volatility on equity mispricing, shedding light on the significant influence of noise trading in distorting market efficiency. They highlight how this noise element not only impedes informed arbitrage but also actively fosters mispricing, intensifying the difficulty for investors who rely on fundamental analysis.

From the standpoint of deteriorating earnings quality, Rajgopal and Venkatachalam [28] illustrate that as firms experience a decline in the clarity and reliability of their earnings reports, their stocks typically encounter increased idiosyncratic volatility. This relationship suggests that poorer earnings quality, a direct consequence of heightened noise in financial reporting, contributes to more volatile firm-specific price movements. Expanding this framework, recent studies have provided more depth. For instance, research by Gu et al. [35] examines how trading constraints act as limits of arbitrage, intensifying the impact of idiosyncratic volatility in the Chinese stock market. This study emphasizes the pervasive effects of market structures and regulatory frameworks on noise-trading activities. Moreover, Li et al. [36] have investigated the impact of firm-specific news on idiosyncratic volatility, further elucidating how news quality, whether perceived as informative or noisy, significantly alters stock price volatility patterns. These findings align with the broader narrative that noise trading significantly impacts market dynamics, influencing not only how information is perceived but also how it affects stock valuation and volatility.

Collectively, these insights enrich our understanding of the complex interplay between information, noise, and market behavior, demonstrating that idiosyncratic volatility is not merely a reflection of firm-specific risk but also a manifestation of broader market inefficiencies driven by noise trading. These dynamics significantly challenge and complicate the efforts of informed traders to capitalize on arbitrage opportunities, underscoring the intricate nature of modern financial markets.

2.2. Energy-Related Uncertainty and Idiosyncratic Volatility

The relationship between uncertainty and idiosyncratic volatility is a well-established concept in financial economics, which posits that increased uncertainty within a firm or industry leads to higher idiosyncratic volatility. This relationship is grounded in the premise that idiosyncratic volatility reflects firm-specific risks that are not explained by broader market movements. In a similar context, the relationship between energy-related uncertainty and idiosyncratic volatility has become increasingly relevant in the context of energy firms, where fluctuations in energy prices, regulatory changes, and technological advancements play pivotal roles in determining the stock price variations independent of the overall market. Studies indicate that higher idiosyncratic volatility in energy firms is often a result of the unique, firm-specific information that these uncertainties unveil, supporting the informed arbitrage view Morck et al. [32]. This perspective suggests that informed arbitrage becomes more prevalent in environments where firm-specific information about future cash flows due to energy-related factors is crucial for investment decisions. This section synthesizes the theoretical and empirical underpinnings that explain how the EUI influences idiosyncratic volatility, drawing on insights from financial economics, real options theory, information asymmetry, and behavioral finance.

2.2.1. Real Options Theory

Real options theory provides a framework for understanding how firms make investment decisions under uncertainty. According to this theory, investments can be viewed as options, where the firm has the flexibility to delay, expand, or abandon a project in response to changing market conditions [40]. Higher uncertainty increases the value of waiting or deferring investment decisions, as firms prefer to gather more information before committing resources. This strategic delay can lead to increased idiosyncratic volatility (IVOL) as market participants speculate on the firm’s future actions and potential adjustments to their investment plans.

Empirical studies have supported the link between uncertainty and increased IVOL. For example, Pastor and Veronesi [10] found that policy uncertainty leads to higher stock price volatility, particularly for firms with significant investment opportunities. Similarly, Gulen and Ion [24] demonstrated that firms delay investment during periods of high economic policy uncertainty, resulting in increased IVOL as investors reassess the risk profiles of these firms. These findings suggest that energy-related uncertainties, captured by the EUI, can similarly increase IVOL by influencing firms’ investment timing and strategies. Additionally, Kang et al. [31] demonstrated that oil price shocks lead to increased stock return volatility, especially for firms in the energy sector. These findings suggest that energy-related uncertainties, captured by the EUI, can similarly increase IVOL by influencing firms’ investment timing and strategies.

2.2.2. Information Asymmetry

Information asymmetry theory posits that significant uncertainty, especially regarding firm-specific or industry-specific developments, exacerbates the information gap between insiders and outsiders [41]. When insiders have more information about the firm’s prospects or potential market changes than outside investors, this discrepancy can cause higher idiosyncratic volatility. Market participants attempt to price in the potential impacts of undisclosed or partially understood developments, leading to greater fluctuations in stock prices.

Empirical evidence supports the idea that increased information asymmetry is associated with higher IVOL. Bali et al. [23] showed that stocks with higher levels of information asymmetry exhibit greater idiosyncratic volatility. Jin and Myers [33] found that firms with less public information disclosure experience higher stock return volatility, highlighting the role of information asymmetry in driving IVOL. Applying this to energy-related uncertainties, firms facing significant energy price volatility or regulatory changes may experience increased information asymmetry, thereby elevating their idiosyncratic volatility. Some recent empirical evidence also highlights the relationship between information asymmetry and IVOL. For example, Bharath et al. [42] found that firms with higher levels of information asymmetry exhibit greater stock price volatility. Similarly, Gu et al. [43] showed that increased information asymmetry, measured by analyst forecast dispersion, leads to higher idiosyncratic volatility. Applying this to energy-related uncertainties, firms facing significant energy price volatility or regulatory changes may experience increased information asymmetry, thereby elevating their idiosyncratic volatility.

2.2.3. Behavioral Finance

Behavioral finance theory suggests that uncertainty can exacerbate investor biases, such as overreaction and underreaction to news, leading to increased idiosyncratic volatility [44]. Under conditions of high uncertainty, investors’ psychological biases and differing interpretations of available information can result in greater discrepancies in their beliefs and expectations about future firm performance. This divergence can cause more pronounced firm-specific stock price movements.

Empirical studies in behavioral finance have documented how investor sentiment and biases influence IVOL. Barberis et al. [44] found that psychological biases such as overconfidence and overreaction to past returns contribute to increased stock return volatility. Bloom [45] highlighted that heightened uncertainty can lead to more erratic investor behavior, thereby increasing firm-specific volatility. In the context of energy-related uncertainties, fluctuating energy prices and an unpredictable regulatory environment can amplify investor biases and reactions, leading to higher idiosyncratic volatility for firms exposed to these risks. Recent studies in behavioral finance have documented how investor sentiment and biases influence IVOL. Huang et al. [46] found that investor sentiment significantly affects stock return volatility, with higher sentiment leading to increased IVOL. Baker et al. [12] highlighted that periods of heightened uncertainty lead to more erratic investor behavior, thus increasing firm-specific volatility. In the context of energy-related uncertainties, fluctuating energy prices and unpredictable regulatory environments can amplify investor biases and reactions, leading to higher idiosyncratic volatility for firms exposed to these risks.

In conclusion, there are strong theoretical justifications for the positive correlation between energy-related uncertainty and idiosyncratic volatility in real options theory, information asymmetry, and behavioral finance. Higher uncertainty—whether from macroeconomic policies, knowledge gaps, or investor behaviors—consistently results in increased firm-specific risk, according to empirical studies examining these hypotheses. Our work improves the knowledge of how energy market dynamics particularly affect idiosyncratic volatility in Chinese enterprises by applying these theories to the Energy-Related Uncertainty Index (EUI).

2.3. Key Developments in China’s Energy Sector

Between 2007 and 2022, China’s energy sector underwent profound changes, driven by aggressive policy reforms, technological advancements, and increasing global interactions, which have significantly impacted both local and international energy landscapes.

The Medium and Long-term Energy Conservation Plan, which focused on energy efficiency and conservation, was first introduced by the Chinese government in 2007, setting the foundation for future efforts. The Renewable Energy Law was then improved in 2009, introducing feed-in tariffs and other measures to encourage the expansion of renewable energy [47]. This formative era was essential in determining the trajectory of ensuing energy policies. With the launch of the 12th Five-Year Plan in 2010, a major phase centered on lowering carbon emissions, enhancing energy efficiency, and increasing the proportion of renewable energy in the energy mix was completed. Significant investments were also made during this time in the clean energy sector, especially in wind and solar energy, which sped up the growth of the sector.

In order to promote a more market-driven pricing system, China liberalized its energy sector, particularly in electricity, between 2013 and 2016. This resulted in other reforms being adopted throughout that time. The “Made in China 2025” project, which sought to improve the technological capabilities of the renewable energy sector, was a complement to this [48]. The trade conflicts between the United States and China, however, presented further difficulties in 2017 and 2018, which raised expenses and disrupted supply chains. International trade disputes caused a great deal of market anxiety at this time [49].

Global commitments to sustainable energy practices and significant policy developments defined the final years of 2019–2022. China modified its renewable energy subsidies in 2020 with the goal of reducing state support and promoting a financially sustainable and market-driven renewable energy sector. Innovation and investment propelled the renewable energy sectors, particularly wind and solar, to sustained growth in spite of these reductions in subsidies [50]. China made a historic commitment in 2021 to attain carbon neutrality by 2060, which reflected a strategic shift in the country’s industrial and national policies in favor of sustainability. The results of these promises became apparent by 2022 as China rapidly increased its renewable energy capacity in spite of geopolitical and economic obstacles on a worldwide scale, such as the fallout from Russia’s invasion of Ukraine.

Overall, these developments underscore China’s strategic shift towards a sustainable energy future, highlighting its role as a global leader in the transition towards renewable energy and sustainability amidst ongoing global environmental challenges.

2.4. Research Gap and Objectives

The existing body of research clearly establishes the significant role of uncertainty, particularly energy-related uncertainty (EUI), in shaping strategic planning and corporate performance, e.g., [51,52]. Studies have explored how fluctuations in energy prices, regulatory changes, and technological advancements influence corporate decision-making and risk assessment strategies. However, there remains a notable gap in the research concerning the impact of the EUI on the idiosyncratic volatility of firms. Idiosyncratic volatility, defined as the variance in a firm’s stock returns not explained by market-wide movements, serves as a crucial indicator of firm-specific risk and investor sentiment. While the broader impacts of the EUI have been acknowledged in strategic contexts, its direct relationship with idiosyncratic volatility, especially across different sectors within the Chinese market, has not been extensively studied. In order to bridge this gap, this study aims to compare the effects of the EUI on idiosyncratic volatility in the energy and non-energy sectors.

• Objective 1—Assess the General Impact of Energy-Related Uncertainty on Idiosyncratic Volatility: This research is aimed at examining the influence of the Energy-related Uncertainty Index (EUI), as introduced by Dang et al. [8], on the idiosyncratic volatility of listed firms in China.
• Objective 2—Examine Sector-Specific Effects of the EUI: The second objective is to investigate whether the impact of the EUI on idiosyncratic volatility differs between energy and non-energy sectors, focusing on the intensity and magnitude of these effects in each sector.

The objectives laid out here aim to give us a comprehensive understanding of how energy uncertainty, as measured by the EUI [8], affects idiosyncratic volatility across different industries and corporate settings. More specifically, our study addresses the following research questions:

• How do idiosyncratic volatility and energy-related uncertainty relate to each other for all Chinese firms?
• What differences exist between the energy and non-energy sectors in the way that energy-related uncertainty affects idiosyncratic volatility?

2.5. Hypotheses Development

In exploring the impact of the EUI on the idiosyncratic volatility of firms, this study aims to understand the broader implications of sector-specific risks on corporate behavior, with a particular focus on Chinese firms. The hypotheses are developed on the basis of a theoretical framework that combines the theories of informed arbitrage and noisy trading, evaluating the varied effects on the energy and non-energy sectors. Specifically, this study will explore the dual impact of informed decision-making and noise trading on the idiosyncratic volatility of energy firms, offering insights into how energy-specific uncertainties shape firm-specific risk profiles. This approach not only aligns with the informed arbitrage theory by highlighting the role of precise, firm-specific information in arbitrage decisions but also acknowledges the significant influence of noise trading as posited by behavioral finance theories.

Hypothesis 1 

(H1). Energy-related uncertainty (EUI) is positively associated with the idiosyncratic volatility of Chinese firms.

Rationale: This hypothesis posits that as energy-related uncertainties increase—driven by factors such as fluctuations in energy prices, changes in energy policies, and innovations in energy technology—there will be a corresponding increase in the idiosyncratic volatility of firms. This association is expected due to the significant role that energy costs and policies play in the operational and financial aspects of firms, influencing investor perceptions and behaviors towards these firms. The relationship is supported by past research, such as Morck et al. [32], who found that environmental and regulatory factors enhance stock price volatility in emerging markets due to heightened uncertainty.

The rationale for this hypothesis is grounded in a robust body of research that explores various dimensions of how uncertainties, particularly those related to energy, impact firm-specific volatility across different contexts and stages of firm development. For example, Yu [53] focuses on fundamental uncertainties, such as demand shifts and innovations, particularly in high-tech industries. These factors, akin to energy-related uncertainties, can cause significant swings in firm-specific volatility due to their unpredictable nature and substantial impact on a firm’s operational landscape. In addition to this, Gaspar and Massa [54] explore another dimension by linking market competition to idiosyncratic volatility. They argue that competitive pressures and market power significantly influence how a firm’s specific risks are perceived and priced by the market. This perspective is crucial in understanding how energy-related uncertainties, which can alter competitive dynamics (especially in energy-intensive industries), might exacerbate idiosyncratic volatility.

Hasan and Habib [55] provide foundational insights into how idiosyncratic volatility fluctuates across a firm’s lifecycle, influenced by uncertainties in cash flows and stock returns. These fluctuations are particularly pronounced during stages of the lifecycle characterized by high uncertainty, such as during expansion or entry into new markets. This variability in performance and information availability as firms evolve supports the notion that energy-related uncertainties could similarly lead to increased idiosyncratic volatility, as they add another layer of complexity to firm-specific forecasting and strategic planning. In a similar context, Dutta et al. [56] extend this understanding by examining the volatility within the energy sector itself, noting how shifts in this sector, signaled by the energy sector implied volatility index (VXXLE), adversely affect clean energy assets. This study underscores the direct impact of sector-specific volatility on related industries and highlights how energy-related uncertainties propagate through the market, affecting firm valuations and investor perceptions. Moreover, Cao et al. [57] specifically investigate how international oil price volatility affects investment decisions in renewable energy firms in China. Their findings highlight the sensitivity of investments to fluctuations in energy prices, suggesting that such volatility not only affects operational decisions but also has a profound impact on investment strategies, thereby influencing the idiosyncratic volatility of these firms.

Together, these studies form a cohesive rationale suggesting that energy-related uncertainty, whether due to price fluctuations, policy changes, or sectoral shifts, significantly influences the idiosyncratic volatility of firms. Each piece of research contributes to a broader understanding that, as the EUI increases, so does the complexity of the environment in which firms operate, thereby elevating the level of idiosyncratic risk associated with their stock prices. This relationship is particularly pertinent in the context of Chinese firms, which are often at the intersection of rapidly changing energy policies, aggressive growth trajectories, and intense global and domestic market pressures.

Hypothesis 2 

(H2). The positive relationship between energy-related uncertainty and idiosyncratic volatility is stronger for firms in the energy sector relative to the non-energy sector.

Rationale: Firms within the energy sector are directly affected by changes in the energy landscape, making them more sensitive to energy-related uncertainties. The idiosyncratic volatility of these firms is expected to be more responsive to shifts in energy policies, price volatility, and technological changes in energy production and management. This heightened sensitivity is anticipated due to the direct impact of energy variables on the operational costs, revenue streams, and strategic decisions in energy firms. This direct exposure likely results in a stronger linkage between the EUI and idiosyncratic volatility for these firms, as evidenced by Sadorsky [58], who demonstrated that energy price volatility significantly affects the stock returns of energy companies due to their direct dependence on energy resources. Moreover, Dutta et al. [56] explored the impact of volatility within the energy sector on clean energy assets, finding that higher volatility in the energy sector leads to negative impacts on related assets, emphasizing the strong connection between sector-specific uncertainties and idiosyncratic volatility in energy firms. Moreover, Szczygielski et al. [59] noted that the COVID-19 pandemic, which introduced significant uncertainty, had a pronounced effect on energy sector stocks, highlighting the susceptibility of energy indices to external shocks.

3. Methodology

In this section, we outline the methodology employed in our study, which is divided into several key components. First, we describe the Sample and Data, detailing the sources of our data and the criteria used for sample selection. Next, we introduce the Main Variables utilized in the analysis, including the definitions and measurements of key variables. This is followed by a discussion of the Summary Statistics, where we provide an overview of the central tendencies and dispersion of the variables. We then present the Econometric Model, explaining the models and techniques used to test our hypotheses. Finally, we address the Diagnostic Tests conducted to ensure the robustness and validity of our findings. Additionally, we employed ChatGPT (version GPT-4), an AI language model developed by OpenAI, to assist in refining the English language and improving the clarity and coherence of the manuscript. The tool provided suggestions for rephrasing sentences, improving word choice, and correcting minor grammatical errors. Importantly, while ChatGPT aided in language polishing, the authors carefully reviewed and edited the AI-generated suggestions to maintain the accuracy and integrity of the scientific content. All substantive content, data analysis, and interpretations were entirely conducted by the authors.

3.1. Sample and Data

This section outlines the variables used in our research. In 2007, China implemented new accounting standards, prompting us to exclude data from before this year in our primary analysis [14,24]. The study aims to investigate the effect of energy-related uncertainty on idiosyncratic volatility, using data from publicly listed firms on the Shanghai and Shenzhen stock exchanges between 2007 and 2022. Firm-level data, including control variables, are sourced from the China Stock Market and Accounting Research (CSMAR) database.

To ensure robust data quality, the study applies several exclusion criteria: (1) firms listed for less than a year are not considered; (2) firm-year observations with missing information or zero total assets are discarded; (3) observations are excluded if they involve firms under non-standard listing conditions (e.g., with ST or *ST status) or if the stock was traded for fewer than 90 days in the year; (4) firms in the financial sector are excluded; and (5) firms involved in specific transfer processes are also excluded. The study uses an unbalanced panel data approach due to fluctuations in the number of listed and delisted firms over the period analyzed. Continuous variables are winsorized at the 1% and 99% levels to mitigate the effects of outliers. The final dataset includes 2332 unique firms, resulting in 20,998 firm-year observations.

This study investigates how energy-related uncertainty affects idiosyncratic volatility across various industries. Adopting the methodology of Chen et al. [14], Xie et al. [26], the sample is split into two categories. The first category, representing the energy-related sector, includes industries such as Energy, Materials, Industrial, Consumer Discretionary, and Utilities, comprising 1475 firms. The second category examines the non-energy-related sector, which includes industries like Consumer Staples, Health Care, Real Estate, Information Technology, and Telecommunication Services, totaling 857 firms.

3.2. Main Variables

3.2.1. Measuring Idiosyncratic Volatility

Following the past studies, e.g., [60,61,62], we utilize the Fama and French [63] five-factor model to measure the idiosyncratic volatility (IVOL) of individual stock returns through the time-series modeling given in Equation (1). In addition to the Fama–French five-factor model, other models have been employed in the literature to estimate IVOL. These include the Capital Asset Pricing Model (CAPM), a foundational model that explains the relationship between expected return and risk, used extensively in earlier studies [64]. Another model includes the Fama–French three-factor model [29], which is an extension of CAPM that includes size and value factors, providing a more nuanced understanding of market anomalies. The Carhart four-factor model [65] is also an alternative that further extends the Fama–French three-factor model by incorporating momentum as an additional factor, offering insights into short-term stock price movements.

The decision to use the Fama and French five-factor model in this study is driven by its superior ability to account for a broader spectrum of risk factors compared to its predecessors. The model includes two additional factors—profitability and investment patterns—which are particularly pertinent in the context of the Chinese market, characterized by rapid industrial growth and varying firm investment behaviors. These additional factors enhance the model’s explanatory power, making it a more robust tool for analyzing the impact of energy-related uncertainty on idiosyncratic volatility.

$$\begin{array}{cc}\hfill R_{i,t}-R_{f,t}& ={\alpha }_i+{\beta }_i(R_{m,t}-R_{f,t})+s_{i}SM{B}_t+h_{i}HM{L}_t\hfill \\ \hfill & +r_{i}RM{W}_t+c_{i}CM{A}_t+{\epsilon }_{i,t},\hfill \end{array}$$

(1)

where $R_{i,t}$ indicates the return on asset i during period t, while $R_{f,t}$ denotes the risk-free rate for the same period. ${\beta }_i$ represents the beta associated with asset i and $R_{m,t}$ stands for the market portfolio during period t. Moreover, SMB and HML refer to the size and value factors, respectively. The model further incorporates two supplementary factors: RMW, indicating profitability, and CMA, representing investment. Lastly, ${\epsilon }_{i,t}$ symbolizes the zero-mean residual associated with asset i during period t. The construction of the Fama–French factors involves sorting firms into portfolios based on specific characteristics such as market capitalization and book-to-market ratios. The market factor (RMRF) is calculated as the excess return of the market portfolio over the risk-free rate, proxied by the CSI 300 Index for Chinese firms. The size factor (SMB) and value factor (HML) are derived by comparing returns between small-cap and large-cap firms and between high book-to-market and low book-to-market firms, respectively. The profitability factor (RMW) is based on the difference in returns between firms with robust and weak profitability, while the investment factor (CMA) is based on differences between firms with conservative and aggressive investment policies. In our analysis, we reference conservative and aggressive investment policies to illustrate how firms’ strategic approaches impact their exposure to energy-related uncertainty. These terms are defined consistently with their usage in the context of US stocks, as conceptualized in the Fama and French models.

Conservative investment policies generally refer to strategies that prioritize stability and risk aversion, focusing on maintaining a steady growth trajectory with minimal exposure to high-risk ventures. These policies are characterized by lower leverage, higher liquidity reserves, and investments in stable, lower-yielding assets. Conversely, aggressive investment policies are characterized by higher risk-taking, with firms engaging in extensive borrowing, investing in high-growth opportunities, and maintaining lower liquidity reserves.

The applicability of these definitions to Chinese stocks is well-supported by existing literature. Studies such as Liu et al. [60], Tabatabaei Poudeh et al. [61], and Liu et al. [62] have utilized similar frameworks to analyze Chinese firms, demonstrating that the risk–return dynamics and strategic categorizations observed in US markets are also relevant in the Chinese context. Moreover, the Fama and French models, which incorporate factors like size, value, profitability, and investment patterns, have been effectively applied to Chinese markets to capture these investment behaviors.

Thus, our use of the terms ‘conservative’ and ‘aggressive’ investment policies aligns with established financial theories and empirical studies in both US and Chinese markets, ensuring consistency and relevance in our analysis.

To estimate idiosyncratic volatility, the parameters of the Fama–French five-factor model given in Equation (1) are first estimated using ordinary least squares (OLS) regression on daily return data. The residuals (${\epsilon }_{i,t}$) are then computed for each trading day. Given that our main analysis is based on annual data, we follow the literature, e.g., [5,60], and compute the standard deviation of the residuals over the course of a year, as given in Equation (2). This process aggregates the daily fluctuations in the unexplained part of a stock’s return into a single measure of volatility for the entire year. This methodology effectively captures the unique risk associated with individual stocks beyond the common risk factors.

$${\mathrm{IVOL}}_{i,\mathrm{annual}}=\mathrm{Stdev}\left({\epsilon }_{i,t}\right)\times 100$$

(2)

For robustness, we also use annualized volatility, which is determined by multiplying by the square root of the number of trading days in a year (typically 252 for the Chinese market), and our key findings remain unchanged.

3.2.2. Energy-Related Uncertainty

The primary variable of interest in this study is the China energy-related uncertainty index (EUI), which is obtained from Dang et al. [8] (Data sourced from https://www.policyuncertainty.com/ (accessed on 24 September 2023)). The EUI is a newly created method for quantifying energy market uncertainty. Since we collect our EUI data on a monthly basis, we turn it into an annual metric by averaging the monthly EUI values throughout the course of the year (Note that, in line with Gulen and Ion [24], Jing et al. [66], we also use the time-weighted EUI, which assigns a higher weight to recent months in order to reflect the most pertinent and recent trends in energy uncertainty).

Construction of the Energy-Related Uncertainty Index (EUI)

The EUI was constructed by Dang et al. [8] using a comprehensive methodology that incorporates several key steps to ensure its reliability and validity. The primary source of data for the EUI is the energy-related news articles published in major Chinese newspapers and online media outlets. These articles are systematically collected and processed using advanced text analysis techniques. The process begins with the identification of relevant keywords and phrases associated with energy market uncertainty. The index includes a predetermined list of energy-related terms. These keywords include, among other things, “energy supply disruption”, “renewable energy regulation”, and “oil price volatility”. More terms that capture energy shocks and crises, such as “energy embargo”, “nuclear energy crisis”, and “electricity shortage”, have been included to improve the index’s precision and relevance to current situations. Finding these keywords in the Economist Intelligence Unit’s (EIU) monthly country reports and evaluating the context in which they occur are the first steps in determining the degree of uncertainty they represent. This strategy makes sure that the index captures both expected fluctuations and unanticipated shocks, reflecting sentiments and movements in the energy market in real time.

These keywords are selected based on their frequency and relevance in the context of energy-related discussions. The next step involves the application of natural language processing (NLP) algorithms to parse and analyze the collected news articles. These algorithms quantify the presence of uncertainty by measuring the frequency and sentiment of the identified keywords within the articles.

The quantified data is then aggregated on a monthly basis to construct the EUI. The final index value for each month is normalized to ensure comparability over time. Specifically, as the index is derived with respect to a base year, comparisons of changes over time are made simple by the usual base index value of 100. The resulting EUI captures the fluctuations in energy-related uncertainty, providing a valuable tool for analyzing the impact of such uncertainty on various economic and financial variables. Please refer to Dang et al. [8] for a thorough explanation of the individual elements and their weights in the EUI computation.

Trends in the EUI

Figure 2 offers a comparative analysis of the Energy-Related Uncertainty Index (EUI) for China versus global energy uncertainty from 2007 to 2022, providing valuable insights into trends and fluctuations within both China and the global context. Initially, from 2007 to 2010, China’s EUI began around 15 and exhibited a slight decline, indicating relatively stable energy market conditions. However, 2011 saw a noticeable drop in the EUI to its lowest point of approximately 6.4, suggesting a period of reduced uncertainty. From 2012 to 2015, the EUI for China sharply increased, attributed to various factors such as economic policies, global market conditions, and domestic energy issues. Between 2015 and 2022, China’s EUI displayed significant fluctuations, reflecting a volatile energy market environment. Despite a decline during 2016 and 2017, there was a resurgence of energy-related uncertainty in China, peaking in 2020. In 2022, the index climbed back to approximately 22.1, likely due to global disruptions, including the COVID-19 pandemic, geopolitical tensions, and shifts in energy policies.

In comparison, the global EUI started at a higher level than China’s EUI, around 20 in 2007, and then gradually declined until 2010. From 2010 to 2012, the global EUI remained relatively stable but was generally higher than China’s EUI. The global EUI rose significantly between 2013 and 2015, reflecting increased global energy market uncertainties, though it peaked slightly earlier than China’s EUI. From 2016 to 2018, the global EUI showed a decline, indicating a period of relative stability in the global energy markets. However, from 2019 to 2022, the global EUI exhibited fluctuations similar to China’s EUI but generally remained higher, suggesting that global energy market uncertainties have been significant and persistent. The trends in China’s EUI closely follow the global EUI during certain periods, particularly from 2012 to 2015 and from 2019 to 2022, indicating that China’s energy market uncertainties are influenced by global trends. However, there are distinct periods where the patterns diverge, such as from 2008 to 2011 and from 2016 to 2018, highlighting that local factors also play a significant role in shaping China’s energy uncertainty.

These observations have several implications. For policymakers, understanding these trends helps in formulating policies that mitigate energy-related risks and uncertainties, especially during periods of high global and local uncertainty. For investors and businesses, monitoring the EUI can guide investment decisions and risk management strategies, as periods of high uncertainty might require more cautious and strategic planning. For researchers, the comparative analysis between China’s EUI and global EUI provides a valuable context for studying the impacts of energy uncertainty on economic and financial outcomes.

Overall, the graph offers a comprehensive view of how energy-related uncertainty has evolved over time in China compared to global trends, providing important insights into the dynamics of energy markets and their broader implications.

Applicability of the China Energy-Related Uncertainty Index (EUI)

The EUI is highly applicable in studying the effects of energy market uncertainty on economic and financial outcomes. Its construction methodology ensures that it accurately reflects the prevailing uncertainty in the energy market, making it a robust indicator for empirical research. The relevance of the EUI is particularly pronounced in the context of China, given the country’s significant role in the global energy market and its ongoing efforts to transition towards sustainable energy sources.

The applicability of the China Energy-Related Uncertainty Index (EUI) is well-supported by recent studies that demonstrate its relevance and utility in various economic and financial contexts. The EUI, developed by Dang et al. [8], provides a comprehensive measure of uncertainty within the energy market, capturing both domestic and global influences. This index is particularly valuable for understanding the impact of energy-related uncertainty on a wide range of economic activities. Their findings highlight the EUI’s ability to reflect the volatility in energy prices, policy changes, and market dynamics. This comprehensive approach ensures that the EUI is a reliable indicator of energy-related uncertainty, making it highly applicable for empirical research and policy analysis.

Salisu et al. [67] demonstrate the significant impact of the Energy-Related Uncertainty Index developed by Dang et al. [8] on international stock market volatility. Their study shows that higher levels of energy uncertainty lead to increased volatility in stock markets across 28 advanced and developing countries. The EUI’s ability to capture these fluctuations highlights its applicability in financial market analysis, where understanding the link between energy uncertainty and stock market behavior is crucial for investors and policymakers.

Xie et al. [26] investigate the influence of the Energy-Related Uncertainty Index of Dang et al. [8] on corporate investment decisions in China. Their findings reveal that higher energy uncertainty discourages corporate investment due to the increased risk and unpredictability associated with energy costs. This study supports the EUI’s relevance in corporate finance, where it serves as a critical factor in strategic decision-making processes. Companies can use the EUI to assess the potential risks and adjust their investment strategies accordingly.

The evidence from these studies confirms the EUI’s broad applicability and its importance as a tool for analyzing the effects of energy market uncertainty. Investors and corporate managers can use the EUI to make informed decisions, manage risks, and develop strategies that account for energy-related uncertainties. By incorporating the EUI into their analyses, researchers and policymakers can better understand the transmission mechanisms of energy-related shocks and develop strategies to mitigate their adverse effects.

In conclusion, the China Energy-Related Uncertainty Index (EUI) is a valuable and reliable measure for assessing energy market uncertainty. Its construction is grounded in rigorous text analysis methodologies, ensuring its accuracy and robustness. The EUI’s applicability is well-supported by existing literature, making it a critical tool for both academic research and policy analysis.

3.2.3. Control Variables

We include several control variables in our analysis to account for factors that have been shown to influence idiosyncratic volatility in prior literature. As shown in

Table 1. Variable definitions.

Variable	Symbol	Definition
Dependent Variable
Idiosyncratic Volatility	IVOL	Calculated as residuals from the Fama–French five-factor model.
Independent [Main]
Energy-related Uncertainty Index	EUI	Based on a text analysis of monthly country reports from the Economist Intelligence Unit (EIU).
Independent [Control]
Firm Size	FSIZE	The natural logarithm of the firm’s total assets.
Book-to-Market Ratio	BMR	Ratio of book value to market value of the firm.
Leverage	LEVG	Ratio of total debt to total assets.
Firm’s Listing Age	FLAGE	Natural logarithm of the number of years since the firm was listed.
Proportion of Tangible Assets	TANGA	Proportion of fixed assets relative to total assets.
Cash Holdings	CH	Sum of cash and tradable financial assets divided by total assets.
Return on Equity	ROE	Net income to average total equity ratio.
Sales Growth	SGR	Growth in sales over a year.
State-Owned Enterprise Dummy	SOE	Binary variable indicating whether or not the firm is a state-owned enterprise.
Institutional Shareholding Ratio	INSTHOLD	The percentage of institutional investors’ shares.
Managerial Shareholding Ratio	MHOLD	Shareholding ratio of managers.
Board Size	BSIZE	Natural logarithm of the number of board members.
Percentage of Independent Directors	PINDD	Percentage of independent directors on the board.
Supervisory Board Size	SUPBSIZE	Natural logarithm of the number of supervisory board members.
Market Beta	BETA	Stock’s sensitivity to market movements.
Stock Return	SRET	Stock return over the period.
Turnover Rate	TOVER	Ratio of trading volume to the number of shares outstanding.

, these controls can be categorized into three main groups: firm characteristics, equity and governance structures, and market performance variables.

First, idiosyncratic volatility, which serves as a measure of non-systemic risk, is closely tied to the fluctuations in future cash flows that are linked to a company’s core business fundamentals and profitability (Rajgopal and Venkatachalam [28], Cao et al. [68]). Therefore, we control for firm-size (${\mathrm{FSIZE}}_{i,t}$; logged), book value to market value ratio (${\mathrm{BMR}}_{i,t}$), financia leverage (${\mathrm{LEVG}}_{i,t}$), listing age of firm (${\mathrm{FLAGE}}_{i,t}$; logged), percentage of total tangible assets (${\mathrm{TANGA}}_{i,t}$), cash held by company (${\mathrm{CH}}_{i,t}$), return on equity (${\mathrm{ROE}}_{i,t}$), and annual growth in sales (${\mathrm{SGR}}_{i,t}$). These variables capture the fundamental aspects of the firms that may affect their idiosyncratic risk. For instance, larger firms with higher tangible assets and stable cash flows are likely to have lower idiosyncratic volatility compared to smaller, high-growth firms (Zhang [27], Rajgopal and Venkatachalam [28]).

Second, a firm’s information openness is influenced by equity and governance arrangements, and this is directly linked to the idiosyncratic volatility in stock prices (Ferreira and Laux [9], Gul et al. [69], Tan and Liu [70]). The variables that we control for are the proportion of independent directors (${\mathrm{PINDD}}_{i,t}$), supervisory board size (${\mathrm{SUPBSIZE}}_{i,t}$; logged), shareholding ratios of institutional shareholders (${\mathrm{INSTHOLD}}_{i,t}$), managers (${\mathrm{MHOLD}}_{i,t}$), and the state-owned enterprise dummy (${\mathrm{SOE}}_{i,t}$; binary). As they may affect the firm’s decision-making procedures and, consequently, its volatility, these governance-related variables are crucial (Ferreira and Laux [9], Tan and Liu [70]).

Third, we control for market performance variables as idiosyncratic volatility is closely tied to the market performance of individual stocks. Following existing literature such as Rajgopal and Venkatachalam [28], Hao and Xiong [71], we include market beta (${\mathrm{BETA}}_{i,t}$), stock return (${\mathrm{SRET}}_{i,t}$), and turnover rate (${\mathrm{TOVER}}_{i,t}$) in our model. These variables capture the stock’s sensitivity to market movements, past performance, and liquidity, which are crucial in understanding its idiosyncratic risk (Hao and Xiong [71], Brockman and Yan [72]).

3.3. Summary Statistics

The summary statistics in

Table 2. Summary statistics of variables.

Variable	Mean	St.dev	Min	Median	Max	Obsv.
IVOL	2.558	0.848	0.834	2.341	6.114	20,998
EUI	2.912	1.625	1.709	2.942	3.200	20,998
FSIZE	8.277	1.043	6.006	8.332	11.776	20,998
BMR	0.601	0.231	0.099	0.561	1.212	20,998
LEVG	0.411	0.211	0.051	0.376	0.912	20,998
FLAGE	2.077	0.801	0.711	2.182	3.522	20,998
TANGA	0.211	0.152	0.001	0.182	0.687	20,998
CH	0.152	0.131	0.012	0.119	0.591	20,998
ROE	0.049	0.155	−0.910	0.061	0.319	20,998
SGR	0.201	0.431	−0.513	0.119	2.568	20,998
SOE	0.278	0.461	0.000	0.000	1.000	20,998
INSHOLD	0.391	0.232	0.003	0.391	0.889	20,998
MHOLD	0.174	0.214	0.000	0.048	0.712	20,998
BSIZE	2.332	0.172	1.824	2.414	2.776	20,998
PINDD	0.388	0.061	0.322	0.387	0.589	20,998
SUPBSIZE	1.521	0.181	1.106	1.452	1.679	20,998
BETA	0.977	0.242	0.318	1.112	1.638	20,998
SRET	0.141	0.524	−0.563	0.016	2.331	20,998
TOVER	7.228	5.622	0.837	5.449	26.027	20,998

Note: This table presents summary statistics of the variables employed in this research. All variables are displayed with their mean, standard deviation, median, minimum, and maximum values. Please see Table 1 for further information on the variables and their measurement units.

provide an overview of the key variables used in this study, offering insights into their central tendencies and variability. The mean idiosyncratic volatility (IVOL) is 2.558, with a standard deviation of 0.848, indicating moderate variability across firms, ranging from 0.834 to 6.114. The Energy-Related Uncertainty Index (EUI) has a mean of 2.912 and a standard deviation of 1.625, reflecting substantial variability, with values spanning from 1.709 to 3.200.

Firm size (FSIZE), measured as the natural logarithm of total assets, averages 8.277, with a standard deviation of 1.043, indicating a diverse sample that includes both small and large firms. The book-to-market ratio (BMR) has a mean of 0.601 and a standard deviation of 0.231, suggesting a balanced distribution of value-oriented firms. Leverage (LEVG) averages 0.411, with a standard deviation of 0.211, showing moderate levels of leverage across firms. The firm’s listing age (FLAGE), logged, has a mean of 2.077 and a standard deviation of 0.801, reflecting a wide range of listing durations. The proportion of tangible assets (TANGA) averages 0.211, with a standard deviation of 0.152, indicating considerable variation in asset structures. Cash holdings (CH) have a mean of 0.152 and a standard deviation of 0.131, suggesting most firms maintain moderate cash reserves.

Return on equity (ROE) averages 0.049, with a standard deviation of 0.155, indicating that while most firms are profitable, some exhibit significant losses. Sales growth (SGR) has a mean of 0.201 and a standard deviation of 0.431, reflecting a substantial variation in sales performance. Approximately 27.8% of the firms are state-owned, as indicated by the state-owned enterprise (SOE) dummy variable. Institutional shareholding (INSHOLD) averages 0.391, with a standard deviation of 0.232, and managerial shareholding (MHOLD) has a mean of 0.174, with a standard deviation of 0.214, indicating varying levels of ownership. Board size (BSIZE), logged, averages 2.332, with a standard deviation of 0.172, while the percentage of independent directors (PINDD) has a mean of 0.388 and a standard deviation of 0.061, suggesting high levels of board independence.

The supervisory board size (SUPBSIZE), logged, averages 1.521, with a standard deviation of 0.181. Market beta (BETA) has a mean of 0.977 and a standard deviation of 0.242, reflecting varying levels of systematic risk. Stock return (SRET) averages 0.141, with a standard deviation of 0.524, indicating a substantial variation in stock performance. The turnover rate (TOVER) has a mean of 7.228 and a standard deviation of 5.622, highlighting significant trading activity variability.

These summary statistics provide a foundational understanding of the data, highlighting the diversity and range of firm characteristics and financial metrics within the sample, essential for interpreting the study’s regression results and empirical analysis.

3.4. Econometric Model

We employ the two-way fixed effects model for regression analysis in accordance with the body of evidence already available on the relationship between uncertainty and IVOL, e.g., [9,11].

$$\begin{array}{cc}\hfill IVO{L}_{i,t}& =b_0+b_{1}log\left(EU{I}_{t-1}\right)+\beta C{V}_{i,t-1}+{\eta }_t+{\phi }_i+{ϵ}_{i,t},\hfill \end{array}$$

(3)

where ${\mathrm{IVOL}}_{i,t}$ denotes the idiosyncratic volatility of company i in the year t. The energy-related uncertainty index is denoted by ${\mathrm{EUI}}_{t-1}$. A vector of control variables is indicated by $C{V}_{i,t-1}$, as Table 1 summarizes. The year-fixed effect, which helps lessen the impact of macroeconomic factors, is denoted by ${\eta }_t$, while the firm-fixed effect is indicated by $u_i$. The last term denotes the unobservable exogenous error component, which is ${ϵ}_{i,t}$. While firm-specific control variables are lagged by one period, IVOL is quantified contemporaneously, in accordance with previous research [14,73]. The idea behind this strategy is that endogeneity problems might be lessened by using firm-level control variables from year $t-1$ [73].

3.5. Diagnostic Tests

One of our main concerns when performing our panel regression investigation is the possibility of multicollinearity between the explanatory variables that may cause the regression coefficients’ variance to increase, producing unstable and inconsistent estimates that are susceptible to small modifications in the model. We performed a variance inflation factor (VIF) analysis, a crucial diagnostic procedure to identify multicollinearity among predictors, in order to address issue. All of the VIF scores, according to our study, were below the 5-point cutoff, suggesting no significant multicollinearity [74]. This criterion is far lower than the widely accepted cut-off value of 10, which numerous academics recommend using as a guideline when identifying severe multicollinearity [75].

We found endogeneity-related issues in our initial analysis, which might have skewed our findings. Endogeneity in our model may result from the concurrent link between the $\left(EU{I}_t\right)$ or other independent variables and idiosyncratic volatility $\left(IVO{L}_t\right)$, where the error term and these factors might be correlated. We employed the Durbin–Wu–Hausman test to test for endogeneity, which helps determine the connection between endogenous regressors and the error terms and suggests that estimates using ordinary least squares (OLS) might not be consistent. After using this test to confirm endogeneity, we employed lagged variables $\left(X_{t-1}\right)$ in accordance with literature guidelines [14,73,76].

By presuming that there is less likelihood of a correlation between the current error term and the predictors’ lag values, our method strengthens the validity of our causal inference. This methodological change was necessary to successfully address the endogeneity problem, and it was further confirmed by rerunning the Durbin–Wu–Hausman test after the modification [77]. Even though the $R^2$ values decreased when lagged predictors were included, our causal analysis’s reliability was much increased. Accepting lower $R^2$ values in favor of more impartial and accurate estimations is in line with the aims of econometric research, which prioritize the goodness of fit over causal relationship accuracy [78].

Furthermore, as non-stationary data might produce erroneous regression findings, stationarity is a crucial consideration in panel data analysis. We ran unit root tests on every panel dataset to make sure our conclusions were sound. We used the Levin–Lin–Chu (LLC) tests, which are an expansion of the augmented Dickey–Fuller (ADF) test for panel data, in accordance with Levin et al. [79]. It takes into account the variability and cross-sectional dependence of the panel’s individual units. The LLC test is appropriate for panel datasets to verify that all of the model’s variables are stationary since it considers both the common temporal trend and the impact of each individual in the data. This procedure ensures that the relationships found in our regression models represent true associations between the variables rather than the random walk properties of non-stationary data. According to test findings, the EUI series exhibits unit roots, a characteristic common to economic time series, necessitating further transformation in order to achieve stationarity. We use a log transformation to address this and refine the data into a stationary series.

Heteroscedasticity and autocorrelation are two common issues with fixed-effects panel data analysis. When the variances of the error components differ among data, this is known as heteroscedasticity and can result in inaccurate and skewed estimates of standard errors. Our robust standard errors account for this discrepancy in order to correct for heteroscedasticity, guaranteeing that our coefficient estimations hold true in the presence of non-constant error variance. It is also essential to take into account the within-firm correlation of the error terms because, given the panel format of our data, observations made inside the same firm over consecutive years are probably not independent. By clustering our standard errors at the firm level—a technique that has been well explored in the literature—we resolve this problem [80]. Making this adjustment is essential for removing any within-group error correlation and delivering more accurate and credible inference. These techniques follow the best principles in econometric analysis for handling panel data and greatly improve the reliability of our statistical evaluations. To obtain more precise and trustworthy estimates, especially when heteroscedasticity and autocorrelation are present, it is advised to include robust standard errors and cluster these errors at the firm level [81].

Through the application of these corrective steps, our study conforms to strict econometric guidelines, guaranteeing the validity and reliability of our results. These procedures are critical for improving the statistical inference’s quality as well as the validity and relevance of our conclusions to actual economic occurrences. For brevity, the results pertaining to diagnostic analysis have not been given here, but they are available upon request.

4. Results and Discussion

4.1. Impact of Energy-Related Uncertainty Index (EUI) on Idiosyncratic Volatility

In this research, we employ a panel regression model for our data analysis (We apply the Hausman [82] test to verify that the panel regression model with fixed effects is appropriate for our empirical study. Although the comprehensive test results are not displayed here, they are available upon request). The regression results presented in

Table 3. Regression results of the effect of energy-related uncertainty on idiosyncratic volatility.

Variable	(1)	(2)	(3)
EUI	0.062 ***	0.051 ***
	[0.000]	[0.000]
FSIZE		−0.024 ***	−0.025 ***
		[0.000]	[0.000]
BMR		−0.651 ***	−0.672 ***
		[0.000]	[0.000]
LEVG		0.346 ***	0.351 ***
		[0.000]	[0.000]
FLAGE		0.056 ***	0.057 ***
		[0.000]	[0.000]
TANGA		−0.061 *	−0.062 *
		[0.092]	[0.091]
CH		−0.122 ***	−0.123 ***
		[0.000]	[0.000]
ROE		−0.701 ***	−0.703 ***
		[0.000]	[0.000]
SGR		0.016 *	0.017 *
		[0.086]	[0.087]
SOE		−0.078 ***	−0.077 ***
		[0.000]	[0.000]
INSHOLD		0.478 ***	0.489 ***
		[0.000]	[0.000]
MHOLD		0.256 ***	0.257 ***
		[0.000]	[0.000]
BSIZE		−0.031	−0.030
		[0.112]	[0.113]
PINDD		−0.029	−0.028
		[0.762]	[0.773]
SUPBSIZE		−0.018	−0.017
		[0.874]	[0.889]
BETA		0.243 ***	0.245 ***
		[0.000]	[0.000]
SRET		0.563 ***	0.564 ***
		[0.000]	[0.000]
TOVER		0.064 ***	0.065 ***
		[0.000]	[0.000]
Constant	0.954 ***	0.975 ***	0.977 ***
	[0.000]	[0.000]	[0.000]
Firm FE/Time FE	Yes/Yes	Yes/Yes	Yes/Yes
No. of obs.	20,998	20,998	20,998
Adjusted ${\mathit{R}}^{\mathbf{2}}$	0.274	0.445	0.528

Note: The estimated effect of the Energy-Related Uncertainty Index (EUI) on idiosyncratic volatility is summarized in the table by Dang et al. [8]. Every column represents a separate regression model, and every row has a different array of variables. Idiosyncratic volatility and the EUI are compared in Model (1) in the absence of any control variables. The EUI is not included in Model (2), but control variables are. Model (3) includes control variables in addition to the EUI. The inclusion of firm and time-fixed effects in the models is indicated by the presence of ‘Yes’ under ‘Firm FE’ and ‘Time FE’, respectively. The term ‘No. of obs.’ indicates how many observations were utilized in each regression, which reflects the size of the sample. After adjusting for the degrees of freedom in the model, the ‘Adjusted $R^2$’ shows the percentage of the dependent variable’s variance that the independent variables account for. Significance levels are denoted by *** at the 1% level, and * at the 10% level.

provide an insightful analysis of the effect of energy-related uncertainty (EUI) on idiosyncratic volatility (IVOL) of firms listed on the Shanghai and Shenzhen stock exchanges. Three specifications are provided in Table 3, along with the regression results. No control variables are used in the estimation of Model 1. Without taking into account any potential confounding factors, it sheds light on the relationship between the dependent variable (idiosyncratic volatility) and the primary explanatory variable (EUI). The EUI is not present in Model 2, but control variables are. The purpose of adding control variables is to take into consideration additional variables that might affect idiosyncratic volatility. To evaluate the effect of control variables without the EUI’s influence, the EUI is purposefully left out. Both the EUI and the control variables are included in Model 3. It offers an extensive investigation by taking into account the simultaneous effects of the EUI and control variables on the idiosyncratic volatility. This makes it possible to comprehend how various factors, including uncertainty connected to energy, interact to effect firm idiosyncratic volatility more thoroughly.

In Model (1), the coefficient for the EUI is 0.062, which is statistically significant at the 1% level. This positive relationship indicates that higher energy-related uncertainty is associated with higher idiosyncratic volatility. This finding is consistent with existing theories that suggest uncertainty in energy markets increases firm-specific risk due to unpredictable changes in energy costs and supply disruptions [14,24]. The significant positive coefficient suggests that firms may face greater operational and financial challenges during periods of heightened energy uncertainty, leading to increased volatility in their stock prices.

Model (2) includes a range of control variables but excludes the EUI. The control variables provide a nuanced understanding of other factors influencing IVOL. The negative and significant coefficients for firm size (FSIZE, −0.024) and book-to-market ratio (BMR, −0.651) indicate that larger firms and those with higher book-to-market ratios exhibit lower idiosyncratic volatility. These results are consistent with the theories that larger firms benefit from economies of scale and more stable cash flows, reducing their idiosyncratic risk [27,28]. Similarly, firms with higher book-to-market ratios, often considered value firms, tend to have more predictable earnings, which can lower idiosyncratic volatility. Leverage (LEVG) shows a positive and significant relationship with IVOL (coefficient = 0.346), highlighting that firms with higher debt levels are more prone to idiosyncratic risk due to financial distress and potential solvency issues. This finding is in line with the trade-off theory of capital structure, which suggests that higher leverage increases the risk of bankruptcy, thus leading to higher idiosyncratic volatility [83].

The positive coefficient for listing age (FLAGE, 0.056) suggests that older firms might have higher IVOL, potentially due to legacy issues or slower adaptation to market changes. This aligns with the notion that older firms may face more entrenched inefficiencies and rigidities that can lead to higher volatility in response to shocks. The control variables related to firm fundamentals, such as the proportion of tangible assets (TANGA), cash holdings (CH), return on equity (ROE), and sales growth (SGR), also demonstrate significant relationships with IVOL. These findings support the resource-based view of the firm, which posits that firm-specific resources and capabilities, such as liquidity and profitability, are crucial determinants of firm performance and risk [84]. These findings are also consistent with previous studies that highlight the importance of firm fundamentals in determining idiosyncratic risk [9,70].

Model (3) integrates both the EUI and the control variables. The coefficient for the EUI remains positive (0.051) and highly significant, reaffirming the earlier finding that energy-related uncertainty increases idiosyncratic volatility. The persistence of this relationship across different models underscores the robustness of the impact of the EUI on IVOL.

The control variables maintain their significance and direction, suggesting consistent effects on IVOL. For instance, the negative relationship between state ownership (SOE) and IVOL (−0.078) indicates that state-backed firms may have lower firm-specific risks due to government support, which is consistent with the political economy theory that state-owned enterprises benefit from softer budget constraints and government bailouts [85]. It is also in line with the empirical findings of [14].

Similarly, the positive coefficients for institutional shareholding (INSHOLD) and managerial shareholding (MHOLD) suggest that higher ownership by these entities can lead to increased idiosyncratic volatility. This can be linked to the agency theory, which posits that institutional and managerial ownership can induce more aggressive risk-taking behaviors, thereby increasing firm-specific risk [86]. This argument is also supported by empirical studies, e.g., [71].

The inclusion of firm and time-fixed effects in all models accounts for unobserved heterogeneity and temporal shocks, enhancing the reliability of the estimates. An important finding from Models (1) and (3) is the presence of positive coefficients for EUIs $0.062$ and $0.051$, respectively. These coefficients are significant at the 1% level and show a statistically significant direct relationship between idiosyncratic volatility and energy-related uncertainty, where a 0.051% increase in energy-related uncertainty corresponds to a 1% increase in idiosyncratic volatility, holding all other factors constant. The adjusted $R^2$ values increase from Model (1) to Model (3), indicating that the inclusion of control variables improves the explanatory power of the models. This implies that a more detailed understanding of the factors impacting idiosyncratic volatility is provided by the inclusion of control variables and the EUI, as evidenced by the Adjusted $R^2$ values of $0.528$ for Model (3). Compared to the other models (0.274 for Model 1, 0.445 for Model 2), Model 3 explains a larger percentage of the variance in idiosyncratic volatility.

The findings of this study contribute to the existing literature by highlighting the significant role of energy-related uncertainty in influencing idiosyncratic volatility. The positive relationship between the EUI and IVOL suggests that firms face greater challenges during periods of heightened energy uncertainty, leading to increased firm-specific risk. This finding is consistent with the theoretical perspectives that emphasize the impact of macroeconomic uncertainty on corporate risk profiles [8,24,87].

The significant effects of control variables such as FSIZE, BMR, and LEVG align with prior empirical studies, validating the importance of firm characteristics and financial structure in determining idiosyncratic volatility. The robustness of these relationships across different models underscores the need to consider both macroeconomic factors and firm-specific attributes when analyzing idiosyncratic risk.

In conclusion, the regression results provide strong evidence that energy-related uncertainty significantly impacts idiosyncratic volatility, and this relationship is robust to the inclusion of various control variables. The findings enhance our understanding of the interplay between macroeconomic uncertainty and firm-specific risk, offering valuable insights for policymakers and investors.

4.2. Impact of the EUI on Energy-Related Firms

The regression results presented in

Table 4. Results of the regression analysis: energy-related versus non-related.

Variable	Overall	Energy-Related	Non Energy-Related
EUI	0.051 ***	0.064 ***	0.037 ***
	[0.000]	[0.000]	[0.000]
FSIZE	−0.025 ***	−0.027 ***	−0.024 ***
	[0.000]	[0.000]	[0.000]
BMR	−0.672 ***	−0.688 ***	−0.615 ***
	[0.000]	[0.000]	[0.000]
LEVG	0.351 ***	0.367 ***	0.346 ***
	[0.000]	[0.000]	[0.000]
FLAGE	0.057 ***	0.059 ***	0.056 ***
	[0.000]	[0.000]	[0.000]
TANGA	−0.062 *	−0.065 *	−0.060 *
	[0.091]	[0.087]	[0.099]
CH	−0.123 ***	−1.128 ***	−1.201 ***
	[0.000]	[0.000]	[0.000]
ROE	−0.703 ***	−0.711 ***	−0.687 ***
	[0.000]	[0.000]	[0.000]
SGR	0.017 *	0.019 *	0.012
	[0.087]	[0.084]	[0.114]
SOE	−0.077 ***	−0.081 ***	−0.074 ***
	[0.000]	[0.000]	[0.000]
INSHOLD	0.489 ***	0.491 ***	0.487 ***
	[0.000]	[0.000]	[0.000]
MHOLD	0.257 ***	0.259 ***	0.256 ***
	[0.000]	[0.000]	[0.000]
BSIZE	−0.030	−0.029	−0.031
	[0.113]	[0.110]	[0.116]
PINDD	−0.028	−0.026	0.031
	[0.773]	[0.784]	[0.749]
SUPBSIZE	−0.017	−0.016	−0.018
	[0.889]	[0.911]	[0.885]
BETA	0.245 ***	0.246 ***	0.244 ***
	[0.000]	[0.000]	[0.000]
SRET	0.564 ***	0.565 ***	0.563 ***
	[0.000]	[0.000]	[0.000]
TOVER	0.065 ***	0.067 ***	0.063 ***
	[0.000]	[0.000]	[0.000]
Constant	0.977 ***	0.932 ***	0.986 ***
	[0.000]	[0.000]	[0.000]
Firm FE/Time FE	Yes/Yes	Yes/Yes	Yes/Yes
No. of obs.	20,998	13,575	7423
Adjusted ${\mathit{R}}^{\mathbf{2}}$	0.076	0.124	0.081

Note: The table presents the estimated findings of the effect of energy uncertainty on idiosyncratic volatility for corporation that are tied to energy and those that are not. Table 1 provides further information about the variables. The p-values are given in square brackets beneath each variable’s coefficient. According to the coefficients, significance levels are denoted by *** at the 1% level, and * at the 10% level.

offer a detailed comparison of the impact of energy-related uncertainty (EUI) on idiosyncratic volatility (IVOL) between energy-related and non-energy-related firms. The findings from this analysis provide significant insights into how different sectors respond to energy-related uncertainties.

The overall regression results indicate that the EUI has a statistically significant positive effect on IVOL, with a coefficient of 0.051 (significant at the 1% level). This positive relationship suggests that increased energy-related uncertainty leads to higher idiosyncratic volatility across all firms. This result aligns with the broader literature that highlights the impact of macroeconomic uncertainty on firm-specific risk. For example, Gulen and Ion [24] and Chen et al. [14] have documented that policy and oil price uncertainties respectively increase firm-level risk, which is consistent with our findings.

For energy-related firms, the coefficient for the EUI is 0.064, which is statistically significant at the 1% level. This higher coefficient compared to the overall sample suggests that energy-related firms are more sensitive to energy-related uncertainties. This heightened sensitivity can be attributed to the direct impact of energy price fluctuations and supply disruptions on these firms’ operations and profitability. The positive relationship between the EUI and IVOL for energy-related firms is consistent with the resource dependence theory, which posits that firms reliant on external resources (in this case, energy) are more vulnerable to uncertainties in those resources [88].

The control variables for energy-related firms also show significant relationships with IVOL. Firm size (FSIZE), book-to-market ratio (BMR), and cash holdings (CH) have negative coefficients, indicating that larger, value-oriented firms with more cash reserves tend to have lower idiosyncratic volatility. These findings support the notion that firm characteristics such as size and financial stability play crucial roles in mitigating firm-specific risks, as suggested by Rajgopal and Venkatachalam [28] and Zhang [27].

For non-energy-related firms, the coefficient for the EUI is 0.037, which is also statistically significant at the 1% level but lower than that for energy-related firms. This suggests that while non-energy firms are affected by energy-related uncertainties, the impact is less pronounced compared to energy firms. This difference may be due to the indirect effects of energy prices on non-energy firms, as they are less dependent on energy as a primary input.

The control variables for non-energy firms show similar patterns to those observed in the overall and energy-related samples. Firm size (FSIZE), book-to-market ratio (BMR), and return on equity (ROE) all exhibit significant negative relationships with IVOL, indicating that larger, value-oriented firms with higher profitability tend to have lower firm-specific volatility. These results are consistent with the literature, which emphasizes the importance of firm fundamentals in determining idiosyncratic risk [9,70].

The comparison between energy-related and non-energy-related firms reveals that energy-related firms are more susceptible to energy-related uncertainties, as evidenced by the higher EUI coefficient. This finding underscores the importance of sector-specific analyses when examining the impact of macroeconomic factors on firm-level outcomes. The differential impact of the EUI on IVOL between the two sectors highlights the need for tailored risk management strategies that account for sector-specific vulnerabilities.

Furthermore, the significant role of control variables such as firm size, book-to-market ratio, and leverage across both sectors suggests that firm-specific characteristics remain critical determinants of idiosyncratic volatility. These findings reinforce the importance of considering both macroeconomic and firm-level factors in comprehensive risk assessments.

In conclusion, the regression results demonstrate that energy-related uncertainty significantly increases idiosyncratic volatility, with a more pronounced effect on energy-related firms. The findings align with existing theories and empirical studies, providing robust evidence for the critical role of energy-related uncertainty in shaping firm-specific risk profiles. This study contributes to the literature by highlighting the sector-specific impacts of macroeconomic uncertainty and underscores the need for nuanced risk management approaches.

4.3. Industry-Wise Analysis

Understanding how energy-related uncertainty affects idiosyncratic volatility (IVOL) in various industries is important for a number of reasons. First of all, it makes it possible to pinpoint industry-specific vulnerabilities, which is important for decision-making by investors and policymakers. Second, by shedding light on how various industries react to outside shocks, such analysis advances our knowledge of the mechanisms by which shocks are transmitted throughout the economy. Lastly, because different businesses may show varying sensitivities to energy-related uncertainty, it helps in the development of focused solutions for risk management and mitigation. Based on this, we perform industry-wise analysis (the results are presented in

Table 5. Results of the industry-wise analysis.

Variable	A. Energy-Related					B. Non-Energy Related
	A.1	A.2	A.3	A.4	A.5	B.1	B.2	B.3	B.4	B.5
EUI	0.073 ***	0.065 ***	0.061 ***	0.070 ***	0.051 **	0.021 *	0.061 ***	0.024 **	0.003	0.047 ***
	[0.000]	[0.000]	[0.000]	[0.000]	[0.049]	[0.076]	[0.000]	[0.000]	[0.249]	[0.000]
Control Variables	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Firm FE	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Time FE	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes
No. of obs.	1886	4682	2599	1954	2454	1174	2303	2083	889	974
Adjusted ${\mathit{R}}^{\mathbf{2}}$	0.142	0.128	0.124	0.138	0.109	0.057	0.127	0.094	0.054	0.103

Note: The table presents the estimated effects of the Energy-Related Uncertainty Index (EUI) on idiosyncratic volatility across various industries in China, classified according to the Global Industry Classification Standard (GICS). Following Chen et al. [14] and Xie et al. [26], the sectors are divided into two main panels: A. Energy-related (Energy, Industrials, Materials, Utilities, Consumer Discretionary for A.1–A.5, respectively) and B. Non-energy related (Consumer Staples, Health Care, Information Technology, Real Estate, Telecommunication Services for B.1–B.5, respectively). The p-values are provided in square brackets beneath each variable’s coefficient, with significance levels denoted by *** at the 1% level, ** at the 5% level, and * at the 10% level.

) that examines the impact of the Energy-Related Uncertainty Index (EUI) on IVOL across different sectors in China, classified according to the Global Industry Classification Standard (GICS) [89].

For energy-related sectors (Panel A), the EUI coefficients are significantly positive across all industries, indicating that increased energy-related uncertainty leads to higher idiosyncratic volatility. Specifically, the coefficients for Energy (A.1), Industrials (A.2), Materials (A.3), and Utilities (A.4) are 0.073, 0.065, 0.061, and 0.070, respectively, all significant at the 1% level. Consumer Discretionary (A.5) also shows a positive and significant relationship, though at the 5% level, with a coefficient of 0.051. These findings align with previous studies by Chen et al. [14] and Xie et al. [26], which highlight the heightened sensitivity of energy-intensive industries to energy market uncertainties.

For non-energy-related sectors (Panel B), the results are more varied. Health Care (B.2) and Information Technology (B.3) show significant positive relationships with the EUI, with coefficients of 0.061 and 0.024, respectively, both significant at the 1% and 5% level, respectively. This is consitent with the findings of Caporale et al. [90], who investigated the relationship between oil price uncertainty and stock market volatility in China. They found that sectors sensitive to government policies, such as energy and healthcare, exhibited higher volatility during periods of increased uncertainty. This aligns with our findings where energy-intensive firms in China show significant sensitivity to the Energy-Related Uncertainty Index (EUI). Moreover, Consumer Staples (B.1) and Telecommunication Services (B.5) also exhibit significant positive relationships, though Consumer Staples is significant at the 10% level, with a coefficient of 0.021, and Telecommunication Services at the 1% level, with a coefficient of 0.047. Real Estate (B.4) does not show a significant relationship, indicating that this sector may be less affected by energy-related uncertainties.

These findings are consistent with the literature, indicating that sectors with high energy consumption or reliance are more susceptible to energy-related uncertainties. The significant positive relationship between the EUI and IVOL in energy-related sectors is supported by studies such as Caporale et al. [90], Chen et al. [14], and Ren et al. [91], which show that energy market fluctuations have substantial impacts on firms’ risk profiles. In non-energy-related sectors, the mixed results suggest that while some industries are affected by energy uncertainties, others may have different primary risk factors.

Overall, the industry-wise analysis highlights the importance of considering sectoral characteristics when assessing the impact of external uncertainties on firm-level risk. These insights are valuable for developing sector-specific risk management strategies and for understanding the broader economic implications of energy-related uncertainties.

4.4. Heterogeneity Analysis

In this section, we perform a heterogeneity analysis that allows us to understand how different firm characteristics, such as state ownership, ownership concentration, and firm size, moderate the relationship between energy-related uncertainty and idiosyncratic volatility. By exploring these dimensions, we can provide more nuanced insights into the differential impacts of uncertainty on various types of firms, enhancing the practical relevance and robustness of our findings.

Prior research has highlighted the significance of heterogeneity in analyzing economic and financial phenomena. For instance, studies by La Porta et al. [92] and Claessens et al. [93] suggest that ownership structures can significantly influence firm behavior and performance. Furthermore, research by Demsetz and Lehn [94] and Shleifer and Vishny [95] emphasizes the role of ownership concentration in shaping corporate governance and market outcomes. Similarly, studies such as those by Rajan and Zingales [96] and Fama and French [97] highlight the importance of firm size in determining access to capital and overall financial stability. By incorporating these perspectives, our heterogeneity analysis presented in

Table 6. Results of the heterogeneity analysis.

Variable	Ownership		Own. Concentration		Firm Size
	SOEs	Non-SOEs	High	Low	Large	Small
EUI	0.102 ***	0.021 *	0.115 ***	0.013	0.045 ***	0.087 ***
	[0.000]	[0.073]	[0.000]	[0.235]	[0.000]	[0.000]
Control Variables	Yes	Yes	Yes	Yes	Yes	Yes
Firm FE/Time FE	Yes/Yes	Yes/Yes	Yes/Yes	Yes/Yes	Yes/Yes	Yes/Yes
No. of obs.	9796	11,202	10,499	10,499	10,499	10,499
Adjusted ${\mathit{R}}^{\mathbf{2}}$	0.091	0.062	0.092	0.078	0.095	0.102

Note: The table presents the estimated effects of the Energy-Related Uncertainty Index (EUI) on idiosyncratic volatility across different firm characteristics such as state ownership, ownership concentration, and firm size. Ownership type distinguishes between state-owned enterprises (SOEs) and non-state-owned enterprises (Non-SOEs). Ownership concentration is proxied by the Herfindahl index, calculated as the sum of the squares of the shareholdings of the firm’s top five largest shareholders. Firm size is measured by market value. Firms are separated into groups based on ownership concentration and size using the median as the criterion. The p-values are given in square brackets beneath each variable’s coefficient. Significance levels are denoted by *** at the 1% level and * at the 10% level.

aims to contribute to the broader literature on how firm-specific characteristics interact with external uncertainties.

4.4.1. Type of Ownership

The ownership structure of a firm can lead to diverse impacts on corporate strategies, primarily due to variations in their levels of political affiliation. Notably, state-owned enterprises (SOEs) tend to have stronger political ties with the government compared to non-SOEs. Specifically, political connections offer two advantageous aspects, making firms more appealing to financial institutions and enabling them to secure additional credit. Firstly, political affiliations provide firms with implicit guarantees from the government against default, which reduces their perceived risk of default and encourages financial institutions to extend loans [98]. Secondly, political connections grant firms privileged access to political insights and a better comprehension of policy developments [99]. As a result, firms with political affiliations tend to operate more effectively and surpass their counterparts without such ties, particularly during times of increased uncertainty. This improved performance makes them more appealing to financial institutions, thereby raising the probability of obtaining loans, as evidenced by the research of [100]. In conclusion, the existence or lack of political connections within firms can lead to varied outcomes within the framework of the relationship between energy-related uncertainty and corporate risk-taking behavior measured as idiosyncratic volatility.

The results in Table 6 show that state-owned enterprises (SOEs) exhibit a significant positive relationship between the EUI and idiosyncratic volatility ($\beta =0.102$, $p<0.01$). In contrast, non-SOEs demonstrate a weaker and marginally significant relationship ($\beta =0.021$, $p<0.10$). These findings align with previous studies, such as Chen et al. [14] and Ren et al. [91], which suggest that SOEs are more sensitive to policy uncertainties due to their closer ties to government policies and regulations.

Conventional wisdom holds that corporate risk-taking is adversely impacted by state ownership [101]. Environmental projects related to climate change, however, are exceedingly prevalent and necessitate taking into account a variety of interests, including long-term, aggregate, and public ones. Large capital expenditure, cutting-edge technology, seasoned operational expertise, and a long-term planning are all necessary. Many companies are discouraged by the lengthy payback period and these strict conditions. State-owned companies, in contrast with ordinary enterprises, rely on financial benefits and policy assistance while performing a variety of tasks, including social, political, economic, and environmental construction.

Despite being typically less financially constrained and having more access to information and resources, SOEs might be more sensitive to energy-related uncertainty for several reasons. First, their close ties to government policies mean that any changes or uncertainties in energy policies can have a direct and substantial impact on their operations. This is particularly true in industries where energy costs constitute a significant portion of operating expenses. Second, SOEs often have broader social and political objectives that extend beyond profit maximization, making them more reactive to policy changes that affect their strategic goals. Lastly, the expectation of government support during periods of uncertainty might lead SOEs to engage in riskier behavior, thereby increasing their sensitivity to fluctuations in energy-related policies. In short, there is pressure on SOEs to lead by example, positively contribute to the national campaign for pollution prevention and control, and strike a balance between the construction of ecological civilisation and economic growth.

4.4.2. Concentration of Ownership

Ownership concentration in an industry serves as a function for market competitiveness and, in turn, the willingness of companies to take risks during times of uncertainty, e.g., [14,91]. Based on this, we use the Herfindahl–Hirschman Index (HHI) as a metric for ownership concentration. The HHI is calculated by squaring the market share of each firm’s sales within an industry and then summing these squared values. This can be expressed with the following formula:

$$HH{I}_{j,t}=\sum _{i=1}^{N_j}{S}_{i,j,t}^2,$$

(4)

where $S_{i,j,t}$ denotes the market share of firm i in industry j for year t. The HHI ranges from zero to one, with higher values indicating a more concentrated industry. A higher HHI suggests that the industry is less competitive, as fewer firms hold larger market shares. We utilize the median as the standard to divide the firms into high and low categories; those with an HHI greater than the median value are regarded as highly concentrated.

The results in Table 6 show that firms with high ownership concentration (proxied by the Herfindahl index) show a significant positive relationship between the EUI and idiosyncratic volatility ($\beta =0.115$, $p<0.01$). Conversely, firms with low ownership concentration do not exhibit a significant relationship ($\beta =0.013$, $p>0.10$). This is consistent with the alignment of interests hypothesis, according to which, concentrated ownership can enhance asset quality and efficiently oversee management [95]. Increased ownership concentration keeps the firm’s risk-taking at an elevated level and aids in addressing the first kind of agency problem. Large owners do, however, have a greater impact on the firm’s policies, internal management, operational state, and long-term growth as equity concentration rises. This enables businesses to act swiftly and optimally in uncertain circumstances. Higher equity concentration thus provides more favourable conditions and chances for companies to take on more risk in the face of uncertain environmental policies. This is also consistent with some of the empirical findings, including studies by Chen et al. [14] and Ren et al. [91], which argue that concentrated ownership structures may lead to more pronounced responses to external uncertainties.

4.4.3. Firm Size

It is generally noted that larger firms tend to have higher collateral values than smaller firms, which reduces their perceived credit default risk. As a result, larger firms often have better access to credit loans. Ref. [102] found that larger firms usually have higher leverage ratios than smaller firms due to lower borrowing costs [103]. Additionally, ref. [104] have shown that firm size is a key predictor of financing constraints. This prompts us to examine the impact of the EUI by firm size to determine if the effect is greater on smaller firms or larger firms. We use market value as a proxy for firm size, and the median serves as the division between large and small groupings of companies.

The results in Table 6 show that both large and small firms show significant positive relationships between the EUI and idiosyncratic volatility. However, the effect is more pronounced in smaller firms ($\beta =0.087$, $p<0.01$) compared to larger firms ($\beta =0.045$, $p<0.01$). This finding is in line with Rajan and Zingales [96] and Fama and French [97], who highlight that smaller firms tend to be more vulnerable to external shocks due to their limited access to capital and resources. This is also in line with the findings of Ren et al. [91], which indicate that, when faced with uncertainty surrounding climate-related policies, large companies typically take less risk than smaller ones.

Overall, our findings corroborate the existing literature on the differential impacts of external uncertainties across various firm characteristics. In particular, ref. [91] investigate the effect of climate policy uncertainty on the idiosyncratic volatility of Chinese firms. According to the study, firms’ idiosyncratic volatility is increased by climate policy uncertainty, which primarily affects state-owned enterprises, corporations with a large equity concentration, and companies with smaller market capitalization. The consistency of our results with past studies highlights the robustness of our analysis and its relevance for policymakers, investors, and corporate managers.

4.5. Analysis of Energy-Related Uncertainty across Different Phases

Understanding the impact of energy-related uncertainty on idiosyncratic volatility across various time-periods is crucial for both academics and practitioners. This analysis not only highlights how firms react to varying levels of uncertainty but also provides insights into the resilience and adaptability of different types of firms. As shown in Figure 3, by classifying periods into high and low uncertainty based on the EUI, with values above the median considered high uncertainty and those below considered low, we can better comprehend how these fluctuations influence firm behavior.

Conducting such an analysis is supported by several studies that emphasize the importance of examining firm behavior under different economic conditions. For instance, studies such as Bloom [45] have shown that economic uncertainty can significantly impact firm decisions and market outcomes. Additionally, the use of median values for classification in panel data analysis has been employed in various research contexts, such as in the works of Gulen and Ion [24] and Brogaard and Detzel [105], to capture the effects of uncertainty on firm performance.

Our analysis presented in

Table 7. Results of the regression analysis: high EUI versus low EUI.

Variable	Overall	High EUI	Low EUI
EUI	0.051 ***	0.059 ***	0.042 ***
	[0.000]	[0.000]	[0.000]
FSIZE	−0.025 ***	−0.028 ***	−0.026 ***
	[0.000]	[0.000]	[0.000]
BMR	−0.672 ***	−0.689 ***	−0.682 ***
	[0.000]	[0.000]	[0.000]
LEVG	0.351 ***	0.356 ***	0.347 ***
	[0.000]	[0.000]	[0.000]
FLAGE	0.057 ***	0.061 ***	0.060 ***
	[0.000]	[0.000]	[0.000]
TANGA	−0.062 *	−0.063 *	−0.061 *
	[0.091]	[0.083]	[0.093]
CH	−0.123 ***	−1.129 ***	−1.223 ***
	[0.000]	[0.000]	[0.000]
ROE	−0.703 ***	−0.709 ***	−0.701 ***
	[0.000]	[0.000]	[0.000]
SGR	0.017 *	0.018 *	0.016
	[0.087]	[0.085]	[0.088]
SOE	−0.077 ***	−0.083 ***	−0.086 ***
	[0.000]	[0.000]	[0.000]
INSHOLD	0.489 ***	0.493 ***	0.491 ***
	[0.000]	[0.000]	[0.000]
MHOLD	0.257 ***	0.258 ***	0.257 ***
	[0.000]	[0.000]	[0.000]
BSIZE	−0.030	−0.031	−0.033
	[0.113]	[0.121]	[0.126]
PINDD	−0.028	−0.029	0.033
	[0.773]	[0.764]	[0.728]
SUPBSIZE	−0.017	−0.018	−0.016
	[0.889]	[0.911]	[0.885]
BETA	0.245 ***	0.247 ***	0.243 ***
	[0.000]	[0.000]	[0.000]
SRET	0.564 ***	0.561 ***	0.567 ***
	[0.000]	[0.000]	[0.000]
TOVER	0.065 ***	0.068 ***	0.064 ***
	[0.000]	[0.000]	[0.000]
Constant	0.977 ***	0.963 ***	0.974 ***
	[0.000]	[0.000]	[0.000]
Firm FE/Time FE	Yes/Yes	Yes/Yes	Yes/Yes
No. of obs.	20,998	10,919	10,079
Adjusted ${\mathit{R}}^{\mathbf{2}}$	0.076	0.101	0.082

Note: The table presents the estimated findings of the effect of energy uncertainty on idiosyncratic volatility for the periods of high EUI and low EUI based on the median value. Table 1 provides further information about the variables. The p-values are given in square brackets beneath each variable’s coefficient. According to the coefficients, significance levels are denoted by *** at the 1% level, and * at the 10% level.

reveals several key insights into how energy-related uncertainty impacts firms during periods of high and low uncertainty. In high-uncertainty periods, the effect of the EUI on idiosyncratic volatility is more pronounced, indicating that firms face greater challenges and risks when energy-related uncertainties are elevated. This is consistent with the findings of studies like Bloom [45], who noted that economic uncertainty can significantly influence firm behavior.

In periods of high energy-related uncertainty, we observe a stronger positive relationship between the EUI and idiosyncratic volatility ($\beta =0.059$, $p<0.01$), suggesting that firms are more reactive to energy-related risks during these times. Conversely, during periods of low energy-related uncertainty, the relationship, while still positive, is less pronounced ($\beta =0.042$, $p<0.01$). This differential impact highlights the importance of context in understanding how external uncertainties influence firm behavior, as supported by Brogaard and Detzel [105], who found similar patterns in their analysis of financial uncertainty.

Overall, these results also align with some of the research on energy uncertainty and stock market volatility carried out outside of China. For example, Colacito et al. [106] examined the impact of energy market uncertainty on stock volatility in the U.S. Their study showed a strong relationship between energy shocks and financial market volatility, similar to our findings in China. This suggests that energy-related uncertainty significantly impacts stock market behavior, regardless of the geographical context. Moreover, Salisu and Gupta [107] investigated the effects of oil price shocks on stock market volatility in BRICS countries (Brazil, Russia, India, China, and South Africa). Their findings indicate that energy uncertainty leads to heightened stock market volatility in these emerging economies, supporting our conclusion that energy-related uncertainty increases idiosyncratic volatility. More recently, Salisu et al. [67] demonstrated that higher levels of energy uncertainty lead to increased volatility in international stock markets, specifically 28 developed and developing counties. This study emphasizes the global relevance of energy-related uncertainty and corroborates our results by showing similar patterns in other countries. It suggests that the relationship between the EUI and IVOL observed in China can be generalized to other international contexts.

In conclusion, our findings highlight the importance of considering the phase of uncertainty when assessing the impact of external shocks on firm behavior. This nuanced understanding can provide valuable insights for policymakers, investors, and corporate managers, helping them to better navigate periods of high uncertainty and mitigate associated risks.

5. Robustness Analysis

5.1. Alternative Measures of Energy-Related Uncertainty

Our work follows Chen et al. [14], Xie et al. [26] in addressing the potential endogeneity of the Energy-Related Uncertainty Index (EUI) by using an instrumental variable (IV) methodology. This entails the use of a lagged U.S. EUI variable that was chosen because of its strong correlation ($r=0.610$) with China’s EUI. Given the significant economic spillover effects of the United States on developing economies—as the literature at present has shown—the relationship between the EUI in the United States and the uncertainty inside China’s EUI is crucial. Our principal results, shown in Table 3, show a strong alignment with our empirical findings, as stated in

Table 8. Robustness analysis: instrumental variable.

Variable	(1)	(2)
EUI	0.068 ***	0.048 ***
	[0.000]	[0.000]
FSIZE		−0.026 ***
		[0.000]
BMR		−0.710 ***
		[0.000]
LEVG		0.349 ***
		[0.000]
FLAGE		0.061 ***
		[0.000]
TANGA		−0.059 *
		[0.089]
CH		−0.119 ***
		[0.000]
ROE		−0.711 ***
		[0.000]
SGR		0.016 *
		[0.087]
SOE		−0.081 ***
		[0.000]
INSHOLD		0.477 ***
		[0.000]
MHOLD		0.249 ***
		[0.000]
BSIZE		−0.029
		[0.114]
PINDD		−0.027
		[0.775]
SUPBSIZE		−0.016
		[0.878]
BETA		0.251 ***
		[0.000]
SRET		0.554 ***
		[0.000]
TOVER		0.066 ***
		[0.000]
Constant	0.962 ***	0.983 ***
	[0.000]	[0.000]
Wu–Hausman F test	0.138	0.138
Firm FE/Time FE	Yes/Yes	Yes/Yes
No. of obs.	20,998	20,998
Adjusted ${\mathit{R}}^{\mathbf{2}}$	0.259	0.498

Note: Using an instrumental variable approach, the table shows the estimated effects of the Energy-Related Uncertainty Index (EUI) on idiosyncratic volatility for Chinese companies. Table 1 contains a comprehensive definition for each variable. p-values are given in square brackets beneath the coefficient of each variable. The values that follow are the significance levels: * for 10% and *** for 1%.

. The utilization of U.S. EUI as an IV in this strategy mitigates potential endogeneity issues as supported by the Wu–Hausman test.

In conclusion, while resolving potential endogeneity problems, this methodological approach enables a more detailed analysis of the direct impact of the EUI on corporate idiosyncratic volatility. The main hypothesis of our research is supported by the recurring finding that a rise in the EUI is associated with a fall in corporate capital investment. This effect runs counter to the rise in investment that would be anticipated in more stable circumstances. These results support our original prediction about the negative effects of the EUI on idiosyncratic volatility, giving our research more substance and adding important information to the larger conversation about the dynamics of idiosyncratic volatility and energy uncertainty.

5.2. Placebo Tests

We carried out a placebo test to verify the validity of our results and reduce the influence of random factors. In order to do this, the actual EUI variable must be replaced with ∼$EUI$, which is chosen at random from the real EUI sample distribution. We repeated our baseline regression 100 times using this random allocation of the EUI throughout the whole dataset. The average coefficient estimates for ∼$EUI$ in unreported details are negligible and nearly zero ($0.0001$), suggesting that our main findings hold up to the placebo test.

6. Conclusion, Practical Significance, and Future Directions

6.1. Conclusions

This study investigates the impact of energy-related uncertainty on idiosyncratic volatility (IVOL) in Chinese firms listed on the Shanghai and Shenzhen stock exchanges over the period from 2007 to 2022. Utilizing the Fama–French five-factor model, we calculate IVOL and analyze the effects of energy-related uncertainty, measured by the Energy-Related Uncertainty Index (EUI), across different sectors. Our findings reveal significant sector-specific responses to energy-related uncertainties, highlighting the nuanced nature of macroeconomic impacts on firm-specific risks. Our regression analysis demonstrates that the EUI has a statistically significant positive effect on IVOL across all firms, with a coefficient of 0.051 (p < 0.01). This indicates that a one-unit increase in the EUI is associated with a 5.1% increase in idiosyncratic volatility. When examining the impact on energy-related firms specifically, the coefficient for the EUI is 0.064 (p < 0.01), suggesting that these firms experience a 6.4% increase in IVOL per unit increase in the EUI. In contrast, non-energy-related firms exhibit a coefficient of 0.037 (p < 0.01), indicating a 3.7% increase in IVOL with rising energy-related uncertainty.

The differential impact of the EUI on energy and non-energy firms underscores the importance of considering sector-specific factors when assessing the effects of macroeconomic uncertainties. Energy-related firms, which are directly influenced by fluctuations in energy prices and supply disruptions, exhibit a higher sensitivity to the EUI. This heightened sensitivity necessitates more robust risk management strategies to mitigate the adverse effects of energy-related uncertainties. Thus, our findings underscore the significant impact of energy-related uncertainty (EUI) on idiosyncratic volatility (IVOL) across various industries and firm characteristics. The industry-wise analysis demonstrates that energy-related sectors are more sensitive to the EUI, while non-energy sectors show varied responses. The heterogeneity analysis indicates that state-owned enterprises, firms with high ownership concentration, and smaller firms are more vulnerable to energy uncertainties. Additionally, the phase-based analysis shows that periods of high EUI exacerbate IVOL more than periods of low EUI. These results align with and extend previous research, offering valuable insights for policymakers and investors to develop targeted risk management strategies and understand the broader implications of energy-related uncertainties on the Chinese market.

Overall, our study contributes to the literature by providing empirical evidence on the sector-specific impacts of energy-related uncertainty on firm-specific risks. The findings highlight the critical role of macroeconomic factors in shaping idiosyncratic volatility and underscore the need for nuanced risk management approaches tailored to different sectors. As energy markets continue to evolve, understanding these dynamics will be crucial for both policymakers and investors seeking to navigate the complexities of macroeconomic uncertainties.

6.2. Practical Significance

The practical significance of this study lies in its comprehensive analysis of how energy-related uncertainty affects firm-specific risks, providing valuable insights for investors, corporate managers, and policymakers. By understanding the nuanced impacts of macroeconomic uncertainties on different sectors, stakeholders can make more informed decisions to mitigate risks and enhance firm performance.

• For Investors: Investors can use the findings of this study to better assess the risk profiles of their portfolios. The significant positive relationship between energy-related uncertainty and idiosyncratic volatility (IVOL) indicates that periods of heightened energy uncertainty can lead to increased stock price volatility, particularly in energy-related firms. By recognizing this pattern, investors can adjust their portfolios to reduce exposure to high-risk sectors during times of energy market volatility. For instance, diversifying investments across sectors less sensitive to energy-related uncertainties can help stabilize returns and reduce overall portfolio risk.
• For Corporate Managers: Corporate managers, especially those in the energy sector, can benefit from understanding the heightened sensitivity of their firms to energy-related uncertainties. The study’s findings suggest that energy-related firms experience a 6.4% increase in IVOL for every unit increase in the EUI. Managers can use this information to implement more robust risk management strategies, such as hedging against energy price fluctuations or diversifying their energy sources. Additionally, firms can focus on strengthening their financial fundamentals, such as increasing cash holdings and maintaining a healthy leverage ratio, to mitigate the impact of macroeconomic uncertainties on their operations. Emphasizing sustainable practices, such as adopting renewable energy sources, can also contribute to long-term stability and reduced volatility.
• For Policymakers: Policymakers can leverage the insights from this study to design policies that help stabilize the energy market and reduce uncertainty. By understanding that energy-related uncertainty has a more pronounced effect on firm-specific risks in certain sectors, policymakers can target interventions to these vulnerable industries. For example, providing subsidies or financial support to energy-dependent firms during periods of high energy price volatility can help stabilize these companies and, by extension, the broader economy. Moreover, policies aimed at promoting energy diversification and sustainability can reduce the overall sensitivity of the economy to energy-related shocks. Encouraging corporate social responsibility and sustainability reporting can further enhance market transparency and investor confidence.
• Sector-Specific Insights: The differential impact of the EUI on energy and non-energy firms highlights the importance of sector-specific risk assessments. Energy-related firms, which show a higher sensitivity to energy uncertainties, need to prioritize risk management practices tailored to their unique vulnerabilities. Non-energy firms, while also affected, experience a lower increase in IVOL, suggesting that their risk management strategies can be less intensive but still proactive. This distinction highlights the need for tailored approaches to managing idiosyncratic risks, aligned with each sector’s specific characteristics and sustainability goals.
• Strategic Planning and Risk Management: The study provides a framework for strategic planning and risk management in the face of macroeconomic uncertainties. By incorporating the insights on how energy-related uncertainties influence firm-specific risks, companies can develop more resilient business strategies. This includes diversifying their operations, enhancing supply chain robustness, and investing in technologies that reduce energy dependency. Understanding these dynamics allows firms to anticipate potential risks and take preemptive actions to safeguard their long-term viability. Implementing sustainability-focused initiatives can further strengthen resilience and contribute to overall corporate sustainability.

In summary, this study offers practical insights that can help investors manage portfolio risks, corporate managers enhance their risk management strategies, and policymakers design targeted interventions to stabilize vulnerable sectors. By highlighting the sector-specific impacts of energy-related uncertainty, the study underscores the need for tailored approaches to risk management and strategic planning. These findings are not only academically significant but also practically valuable in guiding real-world decision-making in an increasingly uncertain macroeconomic environment.

6.3. Research Limitations and Future Directions

While this study provides valuable insights into the impact of energy-related uncertainty on idiosyncratic volatility (IVOL) in Chinese firms, it is essential to acknowledge several limitations that may affect the interpretation and generalizability of the findings.

• Data Constraints: Our analysis starts from 2007, the year China adopted new accounting standards. This limitation means we cannot analyze data from before 2007, potentially missing long-term trends and historical context that could provide a deeper understanding of our findings. Additionally, our study applies several exclusion criteria, such as excluding firms listed for less than a year, those with incomplete data, and those undergoing special treatments. While these criteria ensure data quality, they may also introduce selection bias. Firms excluded from the analysis might have different characteristics and risk profiles, potentially affecting our findings. Future studies could explore the impact of these exclusion criteria in more detail.
• Sector-Specific Focus: Our study categorizes firms into broad sectors: energy-related and non-energy-related. While this classification offers useful insights, it does not consider the diversity within these sectors. For instance, the energy sector includes oil, gas, and renewable energy firms, each likely to react differently to energy-related uncertainties. Future research could benefit from a more nuanced approach that examines sub-sectors individually.
• Other Macroeconomic Factors: Although we focus on energy-related uncertainty, other macroeconomic variables, such as interest rates, inflation, and geopolitical risks, also influence idiosyncratic volatility but are not explicitly controlled for in our models. These factors might confound our results. Future studies could include a broader range of macroeconomic variables to isolate the specific effects of energy-related uncertainty more accurately.
• Other Uncertainty Measures: This study specifically examines the impact of energy-related uncertainty on idiosyncratic volatility while not considering the effects of other types of uncertainties, such as economic policy uncertainty. Economic policy uncertainty can also significantly impact firm-specific risks and overall market volatility. By not including these factors, our analysis may overlook important dimensions of how various uncertainties interact to influence firm behavior and market outcomes. Future research should consider a broader range of uncertainties to provide a more comprehensive understanding of their impacts.
• Geographic Focus: The study is limited to firms listed on the Shanghai and Shenzhen stock exchanges. This geographic focus means the findings might not be applicable to firms in other countries with different regulatory environments and market conditions. Expanding the research to include firms from various regions could help determine whether the observed effects are consistent globally.
• Measurement of Uncertainty: Our measure of energy-related uncertainty is based on an index that reflects broad market sentiments and macroeconomic conditions. While useful, this index might not fully capture firm-specific exposures to energy risks. Future research could develop alternative measures of energy-related uncertainty, potentially incorporating more detailed firm-level data on energy consumption and hedging practices.
• Unbalanced Panel Data: The use of unbalanced panel data, due to variations in the number of listed and delisted firms over the study period, introduces potential biases related to survivorship and selection effects. Firms that delist may do so due to financial distress or regulatory issues, which could skew our results. Future research could address these biases by using matched samples or other robust statistical techniques.

Recognizing these limitations helps provide a more balanced view of our study’s findings and highlights areas for future research. Despite these constraints, this study makes a significant contribution to understanding the sector-specific impacts of energy-related uncertainty on firm-specific risks, offering valuable insights for investors, corporate managers, and policymakers. Future research that addresses these limitations will help build on this foundation, providing even more robust and comprehensive insights.