The Path to Sustainable Stability: Can ESG Investing Mitigate the Spillover Effects of Risk in China’s Financial Markets?

Abstract

In the context of a low-carbon economic transition and escalating uncertainties in financial markets, understanding the relationship between the long-term benefits of ESG (Environmental, Social, and Governance) investments and the stability of China’s financial markets emerges as a critical issue. This paper analyzes the risk contagion mechanisms within China’s financial system from the perspective of volatility spillovers associated with ESG investments. Initially, the study employs the Time-Varying Parameter Vector Autoregression (TVP-VAR) model to calculate the variance decomposition spillover index, contrasting the dynamics and risk transmission mechanisms of market volatility between portfolios composed of ESG and conventional stocks. Building upon the analysis of risk spillover relations among financial sub-markets, the study utilizes the generalized forecast error variance decomposition method to construct a complex network of financial system risk spillovers, investigating the risk contagion characteristics within both financial systems through network topology. Empirical findings indicate a significant reduction in the risk and net spillover effects of China’s financial system when ESG stock indices replace conventional stock indices, with a notable mutation in the volatility spillover network structure during extreme risk events and even more substantial changes during the COVID-19 pandemic. Furthermore, based on volatility spillover analysis, the study computes optimal weights and hedging strategies for portfolios incorporating the ESG volatility index and other market volatility indices. The conclusions of this research are instrumental for regulatory authorities in establishing early warning mechanisms and for investors in avoiding financial investment risks.

1. Introduction

In recent years, the frequent occurrence of “green swan” events has exacerbated climate change risks, having significant impacts on global financial and economic stability. As extreme weather events have become more common, Environment, Social, and Governance (ESG) responsible investing has gained widespread recognition among investors, emerging as a predominant investment strategy in global capital markets. Data from the Global Sustainable Investment Alliance reveal that by early 2020 global ESG investments had reached USD 35.3 trillion, growing at an annual compound rate of 13.02%, a pace that substantially exceeds the global asset annual growth rate of 6.01%. Looking forward, ESG investments are projected to surpass USD 50 trillion by 2025, representing about one-third of the USD 140 trillion in managed global assets.

China is currently undergoing a significant economic development transformation, shifting from high pollution, high energy consumption, and high carbon emissions to a model characterized by low pollution, low energy consumption, low carbon emissions, and high productivity. In the government work report of the 2021 National Two Sessions, it was explicitly stated that there is a necessity to continuously reduce major pollutant emissions and the energy consumption per unit of GDP. Achieving set targets at critical stages of pollution control is crucial to ensure environmental sustainability, with an eye toward reaching the ambitious goal of carbon neutrality by 2060. The “Opinions on Comprehensively Implementing the New Development Concept to Promote Carbon Peak and Carbon Neutrality”, released in September of the same year, outlined a clear path for the green transformation of the economy and society.

To meet China’s “dual carbon” goals, advancing the development of green finance, responsible ESG investing, and sustainable finance is imperative. Establishing comprehensive capital market mechanisms for ESG information disclosure and investment evaluation has become crucial. Globally, climate issues have escalated to become a primary concern for all nations, and the financial gap needed to address climate change and achieve carbon reduction goals is immense, with investments in the hundreds of trillions of dollars required to achieve global carbon neutrality in the 21st century. Against this backdrop, ESG investing not only provides the capital needed to address these issues but also introduces new growth opportunities and innovative momentum to the global financial market, thus playing a pivotal role in the transition to a low-carbon economy. Scholars, both domestic and international, have extensively explored ESG investing from various perspectives, with some studying the risk–return of ESG investments [1,2], analyzing the impact of ESG factors on corporate valuation [3], incorporating ESG information into investment portfolios [4,5], and comparing the financial performance levels of different ESG rating companies [6,7].

Amid extreme risk events such as financial crises, geopolitical conflicts, and the COVID-19 pandemic, significant volatility has been observed in global asset prices. The spillover effects of monetary policies from major developed economies, along with the complexity and severity of external environments, pose substantial challenges to the high-quality development of China’s economy and the stability of its capital markets. In 2020, the COVID-19 pandemic delivered a severe shock to the global economic structure and financial markets, highlighting the crucial role of responsible investments that prioritize both economic and social benefits. This crisis has provoked a global re-evaluation of traditional development models, leading to a consensus on focusing on a green recovery.

Consequently, the connection between financial market stability and the sustainability of the economic and social environment has become a focal point for policymakers. Scholars, both domestically and internationally, have asserted that strong ESG performance positively impacts the economy. As a form of non-financial investment, ESG investing not only enhances a company’s social image but also broadens its risk management strategies and tools by mitigating downside risks and guarding against specific legal risks [8,9]. Academics have revealed the contributions of ESG investing [10], encouraging industry firms to undertake ESG responsibilities and highlighting the positive role of ESG investments in hedging financial risks [11,12], thus attracting external capital to ESG-compliant companies.

However, some scholars have questioned the ability of ESG investments to hedge financial risks. Liu (2022) [13] found that ESG assets do not reduce risk levels in extreme market conditions. Feng et al. (2022) [14] noted that in developing countries, ESG disclosures lack uniform standards and legal institutional support, leading to ESG investment development lagging behind that of developed nations. From 2016 to 2022, although the disclosure of corporate social responsibility reports by listed companies on China’s A-share market showed an increasing trend, the disclosure rate remained below 30%, and the quality of ESG reporting also fell short of that in developed countries. Consequently, investors are unable to obtain information from listed companies’ ESG disclosures that would facilitate informed ESG investment decisions.

ESG integrates economic growth, environmental protection, and social equity into a comprehensive sustainability framework, aligning closely with China’s “dual carbon” goals. However, ESG investing in China started late, and enterprises lack intrinsic motivation to enhance their sustainability capabilities, facing increasing challenges in maintaining financial market stability. Amid global economic uncertainty and significant financial market volatility, ESG investments must be integrated with the “dual carbon” framework to mitigate systemic financial risks and stabilize the Chinese financial market.

The impact of ESG investments on the stability of China’s financial markets is examined in this study through an analysis of risk contagion mechanisms and volatility spillovers. Particular attention is given to the comparative influence of ESG investments versus conventional investments in mitigating systemic financial risks. The research focuses on three core questions: To what extent do ESG investments reduce volatility and risk contagion in China’s financial markets? What are the unique risk spillover characteristics associated with ESG and traditional financial systems? How can ESG investment strategies be refined to enhance systemic resilience and promote sustainable stability amid China’s economic transformation? By addressing these questions, the study seeks to provide critical insights for policymakers and investors regarding the role of ESG investments in stabilizing financial systems within the broader context of global economic and environmental challenges.

This paper examines the role of ESG investments in stabilizing China’s financial market. The marginal contributions of this study are threefold. First, the paper explores the impact of ESG investments from the perspective of financial market stability, expanding the boundaries of research on the economic consequences of ESG investments and revealing their role in mitigating financial market volatility. This provides empirical support for market participants in financial investment and regulatory bodies in risk management. Second, the influence of ESG investments on specific financial markets, such as stock markets or corporate financial performance [15,16], is extended to the entire financial system, including money markets, capital markets, commodities markets, foreign exchange markets, gold markets, and real estate markets. Third, using a TVP-VAR model’s variance decomposition spillover index, the study quantifies the spillover effects of volatility between financial systems containing ESG and conventional stocks and constructs a complex network of sub-markets within the financial system to analyze the contagion characteristics of financial risks between them. Finally, the paper utilizes the DCC-GARCH t-copula model to analyze asset allocation issues in portfolio investment between ESG stock markets and other market indices, calculating optimal investment strategies.

2. Literature Review and Research Hypothesis

2.1. Literature Review

2.1.1. Defining the Concept of ESG Investing

As a comprehensive investment strategy that encompasses Environmental, Social, and Governance (ESG) factors, ESG investing represents a broader category of responsible investment methods. This approach diverges from traditional strategies focused solely on maximizing shareholder profits, prioritizing a balanced development of economic benefits, social responsibility, and environmental sustainability, aligning with globally advocated sustainable development strategies. The evolution of ESG investing has progressed through various stages, including ethical investing, socially responsible investing [17], social investing [18], responsible investing [19], and sustainable investment [20]. The origins of ESG investing trace back to the concept of socially responsible investment (SRI) from the 1960s, which integrated ethical, social, and environmental values into investment decisions. Building on the foundations of SRI, it evolved into a theory of responsible investment, emphasizing environmental concerns, social responsibility, and corporate governance.

Current research defines ESG investing from multiple perspectives, encompassing “E+S+G,” “CSR+G,” and methodological viewpoints. From the “E+S+G” perspective, Leins (2020) [21] views ESG as an investment approach focused on environmental, social, and governance issues. Gao et al. (2021) [22] consider ESG strategies as comprehensive considerations of environmental impact, social responsibility, and corporate governance by economic entities. From the “CSR+G” perspective, Gerard (2019) [23] highlights that ESG includes the environmental and social elements of corporate social responsibility (CSR) combined with the efficacy of corporate governance. From a methodological standpoint, Landi and Sciarelli (2018) [24] note that ESG investing covers a wide range of investment types, emphasizing the importance of ESG factors in enhancing efficiency, output, long-term risk management, and operational improvements. Bebbington (2001) [25] argues that adopting an ESG investment approach demonstrates a company’s commitment to sustainable development, creating value for society and all stakeholders through responsible operations.

2.1.2. Digital Transformation in Upstream and Downstream Enterprises and Innovation in Midstream Enterprises

ESG investing emphasizes the necessity for enterprises to comprehensively consider the impacts of their financing and investment decisions on the ecological environment, social ethics, and long-term value creation [26], and is broadly an extension of corporate social responsibility. Numerous studies both domestically and internationally have enriched the understanding of the motives and economic effects of corporate social responsibility. Within the framework of ESG investing, two distinct motivational and behavioral patterns have been identified in environmental protection: “shareholderism” and “managerialism”.

“Shareholderism” stresses the maximization of shareholder value. Enterprises might invest in environmental protection, such as adopting eco-friendly production technologies, to enhance long-term profitability and market competitiveness, thereby yielding broader social benefits. Based on this theory, the academic community has extensively studied the positive economic effects of ESG investments, including shaping brand image, alleviating information asymmetry and financing constraints, and enhancing corporate financial performance [27]. Silva (2022) [28] explored the impact of ESG disclosure on market crash risk from an economic perspective of information disclosure, finding that it effectively reduces market crash risk by decreasing the opacity and asymmetry of information between management and external investors. Xie et al. (2019) [29] noted that by increasing the transparency of environmental information, enterprises can improve the accuracy of profit forecasts. This enhancement stems from a clearer understanding of environmental risks, allowing timely measures to optimize financial performance. Other scholars have studied the impact of ESG investing on corporate branding and financing constraints. Li and Zhang (2010) [30] found that corporate social responsibility enhances brand reputation, customer satisfaction, and loyalty. Konar and Cohen (2001) [31] argued that companies willing to undertake environmental responsibilities enjoy a good environmental reputation, facilitating investor support and improving financial performance. In terms of financing constraints, Cheng et al. (2014) [32] discovered that corporate social responsibility contributes to increased profitability, thus easing financing constraints. Wang et al. (2022) [33] not only found that corporate social responsibility behavior mitigates financing constraints but also reduces the risk of corporate misconduct.

“Managerialism” is exemplified by some companies enhancing their reputation through “greenwashing” strategies, which may involve publicizing environmental commitments and actions, but lack substantive environmental governance efforts in practice. The primary aim of such strategies is often to respond to public and market pressures and to meet external image requirements, rather than stemming from genuine environmental protection motives. Within the framework of “managerialism”, some scholars have pointed out the potential negative economic impacts of ESG investments. According to Friedman (1970) [34], corporate social responsibility activities are often centered around management interests, with shareholders ultimately bearing the related risks and costs. In some cases, management might use the guise of social responsibility to conceal negative information or unethical behavior, increasing opacity [35]. Duque-Grisales and Aguilera-Caracuel (2021) [36] studied 104 multinational companies and found that higher ESG scores correlated with lower levels of corporate financial performance.

2.1.3. ESG Investment and Financial Markets

The contagion theory of financial crises highlights the rapid spread of shocks across markets due to the interconnectedness of economic sectors and the openness of financial markets. Initially posited by Forbes and Rigobon (2002) [37], this theory underscores the increase in market risk correlation and spillover effects in response to external shocks. The extent of risk spillovers can be considered a potential indicator of emerging risks. The susceptibility of one financial market to quickly influence another and cause significant risk spillovers may precipitate systemic financial crises. Research by Aloui et al. (2011) [38] further affirmed the fragility of financial market interdependencies during crises, emphasizing that such interconnectedness could escalate local market turbulences into global financial crises.

To this end, Diebold and Yilmaz (2012) [39], Antonakakis et al. (2020) [40], and Balcilar et al. (2021) [41] proposed methods to measure the spillover effects and risk correlations of financial markets, thereby capturing the dynamics of systemic financial risks. However, the existing literature predominantly examines the stock market in studies on the impact of ESG investment on financial market stability. Kumar et al. (2021) [42] assessed the performance of stocks from 157 US-listed ESG index companies and 809 unlisted companies, finding that stocks incorporating ESG factors exhibited lower volatility and higher returns compared to their industry peers. Feng et al. (2022) [14] initially studied the impact of ESG on Chinese stock market risks, uncovering a negative relationship between ESG ratings and stock crash risk. Zarafat et al. (2022) [43] further investigated the influence of ESG ratings on asymmetric volatility behaviors, noting that higher ESG ratings resulted in lower asymmetry before the COVID-19 outbreak but increased asymmetry afterward.

For the bond market, ESG-related green bond markets experienced more significant volatility clustering compared to traditional bond markets, with ESG-related assets displaying greater uncertainty. Pham (2016) [44] found that ESG investment did not reduce uncertainty in the bond market, and shocks to traditional bond markets could spill over into the green bond market. Liu (2022) [13] observed that financial instruments with green attributes did not reduce financial market risk levels under extreme conditions, and the volatility in the green bond market was mainly driven by the uncertainty in the traditional fixed-income market.

In conclusion, existing research has not reached a consensus on the economic consequences of ESG investment, with corporate-related studies primarily focusing on the impact of ESG investment on corporate financial performance indicators. However, since the international financial crisis of 2008, an increasing number of scholars have recognized the importance of systemic risk early warning and prevention [45]. Yet, the existing literature is limited to specific markets, overlooking the complex relationship between the broader economic environment and financial stability. ESG investment does not consistently serve as a haven for investors. Therefore, a more in-depth and comprehensive analysis of the role of ESG investment in maintaining financial market stability is warranted.

2.2. Research Hypothesis

The relationship between ESG investments and financial market stability has emerged as a critical area of focus in academic and policy discussions. ESG investments seek to balance financial returns with sustainable development objectives by incorporating environmental stewardship, social responsibility, and governance efficiency into decision-making processes. These principles have been linked to reduced risk and enhanced resilience during periods of financial market stress. Nevertheless, the mechanisms by which ESG investments influence financial market dynamics—particularly their role in mitigating systemic risks and volatility spillovers—remain insufficiently understood. Within the context of China’s economic transformation and its commitment to “dual carbon” goals, the stabilizing effects of ESG investments on financial markets merit rigorous investigation.

Volatility spillovers, which represent the transmission of risk across financial sub-markets, exacerbate systemic instability during periods of stress. In highly interconnected financial systems like China’s, minor disruptions in one market can cascade into significant disturbances across others. ESG investments, emphasizing long-term sustainability and ethical governance, are hypothesized to mitigate such effects. Previous studies have indicated that ESG investments attract stable capital flows and encourage disciplined risk management practices, functioning as buffers against the propagation of volatility. By reducing exposure to high-risk sectors and enhancing transparency, ESG investments are expected to limit the spread of shocks throughout the financial system. This study examines the hypothesis that ESG investments significantly reduce overall volatility spillovers in China’s financial markets compared to conventional investments.

Hypothesis 1.

ESG investments significantly reduce the overall volatility spillover effects in China’s financial markets compared to conventional investments.

Systemic risks in financial markets often emerge from the interconnectedness of sub-markets, where disruptions in one sector can rapidly propagate across the entire system. This contagion effect becomes particularly pronounced during extreme market events, such as financial crises or global pandemics. ESG investments, by emphasizing diversification and fostering robust governance practices, are hypothesized to modify these risk transmission pathways, thereby contributing to a more resilient market structure. The focus on sustainability and long-term performance is posited to disrupt traditional contagion channels, reducing both the speed and magnitude of systemic risk transmission. This hypothesis examines the potential of ESG investments to reshape the topology of risk contagion networks within China’s financial system, ultimately limiting the spread of systemic risks.

Hypothesis 2.

The inclusion of ESG investments in China’s financial system alters risk contagion pathways, thereby limiting the transmission of systemic risks under extreme market conditions.

Hedging strategies are critical for mitigating financial risks, particularly during periods of heightened market uncertainty. Traditional portfolios typically manage risk through diversification within conventional assets. However, ESG investments, with their focus on sustainability, long-term performance, and reduced exposure to high-risk practices, are posited to provide superior hedging capabilities. Empirical studies suggest that ESG investments attract long-term investors and encourage corporate responsibility, thereby enhancing portfolio resilience. This hypothesis investigates whether portfolios incorporating ESG investments achieve better risk-adjusted returns, greater hedging effectiveness, and lower systemic risk exposure compared to portfolios dominated by traditional assets.

Hypothesis 3.

Portfolios incorporating ESG investments achieve superior hedging effectiveness and lower systemic risks compared to portfolios focused on traditional assets.

Extreme market events impose significant stress on financial systems, often exposing structural weaknesses and amplifying systemic risks. ESG investments, with their emphasis on governance and sustainability, are hypothesized to respond uniquely to such conditions. For instance, during the COVID-19 pandemic, companies with strong ESG practices demonstrated enhanced resilience, which attracted investor confidence and contributed to market stabilization. This hypothesis examines whether ESG investments influence the structure and dynamics of volatility spillovers during extreme market events, potentially improving financial market stability.

Hypothesis 4.

The dynamics of volatility spillovers associated with ESG investments undergo significant structural changes during extreme market events, such as financial crises or the COVID-19 pandemic.

3. Construction of Financial Risk Spillover Models

The methodology adopted in this study employs advanced econometric models to analyze the role of ESG investments in mitigating financial risk spillovers and stabilizing China’s financial markets. The analysis is structured in three stages. First, the Time-Varying Parameter Vector Autoregression (TVP-VAR) model is applied to calculate variance decomposition spillover indices, offering a quantitative assessment of the intensity and directionality of risk transmission within the financial system. This approach ensures that dynamic relationships among financial sub-markets are accurately captured over time.

Second, a complex network model is developed using generalized forecast error variance decomposition. Financial sub-markets are represented as nodes, with their risk spillovers modeled as weighted edges, enabling the visualization and analysis of risk propagation topology within China’s financial system. This network-based framework identifies critical nodes and pathways that contribute to systemic risk.

Third, the DCC-GARCH t-copula model is utilized to evaluate asset allocation and hedging strategies involving ESG investments and other market indices. This model accounts for the nonlinear and time-varying correlations between ESG assets and conventional financial instruments, facilitating the identification of optimal portfolio compositions and effective hedge ratios.

3.1. Financial Risk Spillover Index

Diebold and Yilmaz (2012) [39] calculated the spillover index using the generalized forecast error variance decomposition method, which contrasts with the traditional Cholesky decomposition approach by being unaffected by the ordering of variables. To more accurately capture the dynamic spillover effects, this paper adopts the Time-Varying Parameter (TVP) approach, replacing the conventional moving window technique. The TVP-VAR method, when conducting large-scale Bayesian computations, effectively preserves essential information from the original data without resulting in information loss. The spillover index generated from the variance decomposition using the TVP-VAR model not only quantitatively assesses the directionality of risk spillovers within the financial system but also quantifies their intensity, which is crucial for a deeper understanding of the risk propagation mechanisms within the financial system. The process of the TVP-VAR model is defined as follows:

$$y_t=c_0+C_1{y}_{t-1}+\cdots +C_p{y}_{t-p}+{\mu }_t$$

(1)

Here, the volatility of N financial markets is denoted by the vector y_t, while the intercept vector is represented by c₀, and $C_1,...,C_p$ serves as the coefficient matrix for $n\times n$. The disturbance vector ${\mu }_t$ consists of components that are independent and identically distributed.

Defining ${\alpha }_t=vec(c_0^{\prime },C_1^{\prime },\cdots ,C_p^{\prime })$ and $x_t=I\otimes (1,y_{t-1}^{\prime },...,y_{t-p}^{\prime })$, the coefficient vector ${\alpha }_t$ is assumed to follow an autoregressive process of order one (AR(1)):

$$\begin{gathered} y_{\iota }={\alpha }_t{x}_{\iota }+{\mu }_{\iota } \\ {\alpha }_t={\alpha }_{t-1}+{\eta }_t \end{gathered}$$

(2)

where ${\eta }_t\sim N(0,\Omega )$. This paper employs Bayesian sampling to estimate such high-dimensional state space models, following the methodology developed by Chan and Eisenstat (2018) [46].

When the variable y_t experiences an external shock, the proportion of its h-step forecast error variance explained by y_t itself, denoted as $d_{ij}\left(h\right)$, is termed the variance contribution. This proportion reflects the extent to which changes in y_t are driven by its own dynamics or by other variables within the system. The calculation of these proportions of forecast error variance serves as the foundational basis for constructing variance decomposition spillover indices.

$$d_{ij}(\begin{array}{cc}h)& ={\displaystyle \frac{{\sigma }_{ii}^{-1}{\sum }_{h=0}^H({\begin{array}{c}{e}_i{C}_i\Sigma e_j)\end{array}}^2}{{\sum }_{h=0}^H({\begin{array}{c}{e}_i{C}_i\Sigma C_i{e}_i)\end{array}}^2}}\end{array}$$

(3)

where $\Sigma$ represents the covariance matrix of the disturbance term ${\mu }_t$, ${\sigma }_{ii}$ denotes the standard deviation of ${\mu }_t$, and $e_i$ is a unit vector in which the i-th element is 1, with all other elements being 0. The matrix $D_{ij}\left(h\right)$, composed of elements from $d_{ij}\left(h\right)$, serves as a variance decomposition matrix of dimensions $N\times N$. This matrix is utilized to represent the risk spillover effects among non-ventilated markets.

$$D(\begin{array}{ccc}h)& =& \left(\begin{array}{cccc}{d}_{11}& d_{12}& \cdots & d_{1N}\\ d_{21}& d_{22}& \cdots & d_{2N}\\ ⋮& ⋮& \ddots & ⋮\\ d_{N1}& d_{N1}& \cdots & d_{NN}\end{array}\right)\end{array}$$

(4)

In the forecast error variance decomposition matrix $D_{ij}\left(h\right)$, the off-diagonal elements ($i\ne j$) reveal the extent of risk spillover from market i to market j. Accordingly, the sum of elements in the i-th row reflects the total risk spillover received by market i from all other markets ($j=1,...,N$). Similarly, the sum of elements in column j indicates the level of risk spillover that market j generates towards all other markets. The net spillover effect for a market is defined as the difference between its risk output (i.e., risks transmitted to others, labeled “To”) and risk inputs (i.e., risks received from other markets, labeled “From”). The average of all elements in the matrix represents the overall degree of risk transmission among the markets.

To effectively analyze the spillover effects of volatility, it is necessary to ensure that the row sums of the variance decomposition table $D\left(h\right)$ equal 1. Consequently, $d_{ij}\left(h\right)$ is normalized to $l_{ij}\left(h\right)$, facilitating the computation of the overall financial system’s risk spillover index $S\left(h\right)$, which measures the impact of volatility spillovers across markets on the entire financial system.

$$S(h)={\displaystyle \frac{{\sum }_{i,j=1,i\ne j}^N{l}_{ij}\left(h\right)}{{\sum }_{i,j=1}^N{d}_{ij}\left(h\right)}}\times 100$$

(5)

3.2. Construction of the Financial Risk Spillover Network

Within the financial system, the complex interplay of volatility spillovers among markets constitutes a network structure of risk spillover. In this study, the various sub-markets of the financial system are treated as nodes within the network, comprising N nodes in total, while the risk spillovers between markets form the edges that connect these nodes (with edge A_ij indicating the direction from node i to node j). The adjacency matrix of the volatility spillover network is constructed using variance contribution rates, thereby establishing a model of the risk spillover network within China’s financial system. This model, by analyzing the network’s topological structure, reveals the characteristics of risk propagation within the Chinese financial system. The volatility spillover network constructed in this paper is a weighted directed network, where the directional risk spillover indices represent the out-degree and in-degree of the network nodes, and the variance contribution rates signify the weights of the network connections (${\omega }_{ij}$). Consequently, the weighted degree $d\left(i\right)$ of node i is defined as follows:

$$d(i)=d_{in}(i)+d_{out}(i)={\sum }_{j=1}^N{\omega }_{j,i}+{\sum }_{j=1}^N{\omega }_{i,j}$$

(6)

where $d_{in}(i)$ is the weighted in-degree of node i and $d_{out}\left(i\right)$ is the weighted out-degree of node i.

3.3. Construction of the DCC-GARCH t-Copula Model

The financial system addressed in this paper includes diverse markets such as the money, capital, commodity, foreign exchange, real estate, and gold markets. These varied markets offer a plethora of asset allocation options for investors with different risk preferences. Investors can effectively hedge financial asset risks by appropriately allocating assets across these markets. Drawing on the methodology of Antonakakis et al. (2018) [47], this study not only analyzes the volatility spillover effects between markets but also explores portfolios of volatility indices across different markets. This research provides both policy references for regulatory bodies in risk management and practical investment advice for asset allocation.

In the financial system, the characteristics of risk transmission among sub-markets are pronounced. Investors holding stock volatility indices tend to adopt diversified investment strategies in constructing portfolios that include stock volatility indices, to minimize risk. Such strategies involve investments across different markets and categories to enhance hedging effectiveness. Considering the nonlinear correlations among specific volatility variables, this paper employs the DCC-GARCH t-copula model to analyze hedging strategies for volatility indices. Patton (2006) [48] extended the static copula to a dynamic copula model, resulting in a t-copula characterized by shape parameters and dynamic conditional correlations as follows.

Within the financial system, the characteristics of risk transmission among sub-markets are significantly pronounced. Investors holding stock volatility indices often utilize diversified investment strategies when constructing multi-asset portfolios that include these indices to mitigate risk. Such strategies entail investing across various markets and categories to enhance the effectiveness of hedging. Considering the nonlinear correlations between specific volatility variables, this paper employs the DCC-GARCH t-copula model to analyze hedging strategies for volatility indices. Patton (2006) [48] expanded the static copula to a dynamic copula model, resulting in a t-copula characterized by the shape parameter $\upsilon$ and dynamic conditional correlations $R_t$, as follows:

$$c(\begin{array}{cc}{u}_1,\cdots ,u_N|R_t,\upsilon )& ={\displaystyle \frac{f\left({\mathrm{F}}_1^{-1}\right(\begin{array}{c}{u}_1\mid \upsilon )\end{array},\cdots ,F_N^{-1}(\begin{array}{c}{u}_N\mid \upsilon )\end{array}\mid R_t,\upsilon )}{{\prod }_{n=1}^N{f}_n(\begin{array}{c}{F}_n^{-1}(\begin{array}{c}{u}_N\mid \upsilon )\end{array}\mid \upsilon )\end{array}}}\end{array}$$

(7)

In analyzing the time-varying nonlinear correlations between different variables, the bivariate DCC-GARCH model demonstrates its effectiveness in estimating correlation matrices and exploring nonlinear relationships among variables, as outlined by Engle (2002) [49]. The formulation of the bivariate DCC-GARCH model is as follows:

$$H_t=D_t{R}_t{D}_t$$

(8)

where H_t is expressed as decomposing the covariance matrix into the conditional variance D_t and conditional correlation coefficient R_t, and the dynamic conditional correlation coefficient matrix R_t is as follows:

$$\begin{gathered} R_{\iota }=\left(Q_{\iota }^{*}{)}^{-1}{Q}_{\iota }\right(Q_{\iota }^{*}{)}^{-1} \\ {Q}_{\iota }=(1-\alpha -\beta )\bar{Q}+\alpha \left(z_{\iota -1}{z}_{\iota -1}^{\prime }\right)+\beta Q_{\iota -1} \end{gathered}$$

(9)

where $Q_t$ represents the covariance matrix, $Q_t^{*}$ denotes the diagonal matrix obtained by taking the square roots of the diagonal elements of $Q_t$, and $\bar{Q}$ is the unconditional variance of the standardized residuals $z_t$. The parameters $\alpha$ and $\beta$ are associated with dynamic conditional correlations ($\alpha +\beta <1$).

Furthermore, $D_{\iota }$ is constituted by the diagonal elements derived from the time-varying conditional standard deviations calculated using the GARCH model:

$$\begin{gathered} D_{\iota }=\mathrm{diag}\{({\begin{array}{c}{h}_{11t})\end{array}}^{1/2},({\begin{array}{c}{h}_{NNt})\end{array}}^{1/2}\} \\ {h}_{it}=w_i+{\sum }_{\rho =1}^{p_i}{\alpha }_{ip}(z_{it-p}{)}^2+{\sum }_{q=1}^{q_i}{\beta }_{iq}{h}_{it-q} \end{gathered}$$

(10)

where $h_{it}$ is the conditional variance generated by the $GARCH(p,q)$ model.

4. Empirical Results

This paper builds upon the classification of financial system sub-markets by Liu et al. (2017) [50] and Gong and Xiong (2020) [45], further extending the categorization to secondary sub-markets to construct a comprehensive index of China’s financial system, as shown in

Table 1. Index system of the Chinese financial market.

Primary Market	Secondary Market	Variable Name	Abbreviation
Money Market	Interbank Lending Market	Interbank 7-Day Repo Rate	Shibor
	Repurchase Market	Interbank 7-Day Weighted Repo Rate	DR007
Capital Market	Stock Market	CSI 300 Index	Stock
		ESG 300 Index	ESG300
	Interbank Bond Market	China Bond Composite Index	CGBI
	Exchange Bond Market	China Govt Bond Index	CGBNI
Foreign Exchange Market		Mid-rate USD to CNY	USD_CNY
Commodity Market	Metals Market	Nan Hua Metals Index	NMI
	Energy Market	Nan Hua Energy Chemistry Index	NCI
	Agricultural Market	Nan Hua Agricultural Products Index	NAI
Gold Market		Spot Gold Price	AU
Real Estate Market		Real Estate Industry Index	House

. Considering data availability and the inclusion of significant crisis events, the sample period selected spans from 10 August 2010 to 31 September 2023. This period encompasses indicators from 12 secondary financial sub-markets, with all data sourced from the WIND database. For the real estate sector, indices related to real estate from the WIND database are used, with daily data frequency. The study utilizes P to denote the closing price of market i on the t-th trading day, while R represents the natural logarithm of returns for the market on the trading day. Through this approach, the daily returns on risk assets in the Chinese financial market were estimated.

4.1. Risk Spillover Effect Analysis

To calculate the risk interconnectedness within China’s financial markets, characteristic volatilities were extracted from each market to construct a TVP-VAR model. The generalized forecast error variance method was applied, resulting in a spillover effect table (see

Table 2. Financial market risk spillover table.

	Stock	DR007	Shibor	CGBI	CGBNI	USD_CNY	NMI	NCI	NAI	AU	House	From
Stock	36.62	0.69	0.77	2.01	1.89	2.21	7.43	6.56	3.73	1.45	36.65	63.38
	(35.22)	(0.7)	(0.72)	(2.15)	(2.01)	(2.39)	(8.18)	(7.25)	(3.82)	(1.42)	(36.15)	(64.78)
DR007	1.08	68.45	17.47	3.59	2.89	2.52	0.73	0.68	1.03	0.61	0.97	31.55
	(1.17)	(68.22)	(17.63)	(3.5)	(2.82)	(2.59)	(0.74)	(0.7)	(1.02)	(0.63)	(0.98)	(31.78)
Shibor	1.37	10.94	77.22	1.27	1.11	3.27	0.91	0.83	1.03	1.02	1.03	22.78
	(1.42)	(10.6)	(77.54)	(1.25)	(1.13)	(3.28)	(0.88)	(0.82)	(1.02)	(1.02)	(1.03)	(22.46)
CGBI	2.81	2.46	1.37	5.36	74.5	1.47	3.14	2.76	1.75	1.97	2.41	94.64
	(3.08)	(2.51)	(1.3)	(5.26)	(74.43)	(1.48)	(3.13)	(2.71)	(1.73)	(1.97)	(2.41)	(94.74)
CGBNI	2.74	2.12	1.28	75.2	4.9	1.41	3.26	2.84	1.89	2.03	2.32	95.1
	(2.98)	(2.15)	(1.2)	(75.05)	(4.94)	(1.43)	(3.24)	(2.8)	(1.88)	(2.02)	(2.32)	(95.06)
USD_CNY	2.33	5.57	4.53	1.93	1.77	77.48	1.13	1.09	0.82	1.57	1.78	22.52
	(2.59)	(5.74)	(4.54)	(1.9)	(1.76)	(77.08)	(1.13)	(1.06)	(0.81)	(1.57)	(1.8)	(22.92)
NMI	6.42	0.47	0.59	1.97	2.04	0.99	38.76	29.35	11.63	3.18	4.6	61.24
	(7.1)	(0.46)	(0.55)	(2.06)	(2.12)	(0.98)	(38.53)	(29.02)	(11.5)	(3.12)	(4.54)	(61.47)
NCI	5.67	0.47	0.65	1.61	1.66	0.93	29.33	36.56	15.63	3.16	4.32	63.44
	(6.3)	(0.47)	(0.62)	(1.66)	(1.72)	(0.92)	(29.02)	(36.43)	(15.48)	(3.12)	(4.26)	(63.57)
NAI	3.18	0.59	0.71	1.27	1.29	0.69	11.3	15.46	60.95	2.12	2.44	39.05
	(3.32)	(0.6)	(0.74)	(1.19)	(1.23)	(0.69)	(11.24)	(15.44)	(61.05)	(2.09)	(2.42)	(38.95)
AU	1.57	0.58	0.8	2.01	2.03	1.64	3.9	3.77	2.64	80.13	0.93	19.87
	(1.67)	(0.58)	(0.8)	(2)	(2.02)	(1.64)	(3.85)	(3.73)	(2.6)	(80.17)	(0.93)	(19.83)
House	38.93	0.7	0.76	1.92	1.82	1.36	5.72	5.45	3.02	0.92	39.39	60.61
	(39.54)	(0.7)	(0.74)	(1.91)	(1.81)	(1.36)	(5.76)	(5.48)	(3)	(0.91)	(38.82)	(61.18)
To	66.1	24.59	28.93	92.8	91.01	16.49	66.84	68.77	43.16	18.04	57.45	574.17
	(69.18)	(24.51)	(28.84)	(92.68)	(91.04)	(16.76)	(67.15)	(69.02)	(42.88)	(17.87)	(56.82)	(576.75)
Net	2.72	−6.96	6.15	−1.85	−4.09	−6.02	5.6	5.33	4.11	−1.83	−3.16	52.2
	(4.4)	(−7.27)	(6.37)	(−2.06)	(−4.02)	(−6.16)	(5.68)	(5.45)	(3.93)	(−1.96)	(−4.36)	(52.43)

). The number of lags in the spillover analysis was set to one, based on the Akaike information criterion (AIC) provided by the VAR model. Additionally, following Diebold and Yilmaz (2012) [39], a 200-day rolling sample was utilized to estimate the spillover index.

Table 2 presents the static spillover effects in China’s financial markets calculated via the TVP-VAR method. “From” indicates the degree to which markets receive risk spillovers from other financial markets, whereas “To” represents the extent of risk spillovers transmitted to other markets. For comparative purposes, Table 2 also displays the risk spillover scenarios with the ESG stock index replacing the traditional stock index; parentheses contain the spillover values for the conventional stock market. According to Table 2, risk propagation in China’s financial markets exhibits high connectivity, demonstrating significant volatility and directional asymmetry. The diagonal values in the spillover table reflect the impact of lag effects on the current market, with the gold market showing an 80.17% influence from its own previous states, indicating a significant self-influence.

The total risk spillover for China’s financial markets was 52.43%. When ESG stocks replaced conventional stocks, the overall spillover effect decreased by 0.44 percentage points from 52.43% to 52.2%. The stock market’s risk spillover received from other markets decreased by 2.16 percentage points, from 64.78% to 63.38%, and the risk spillover transmitted to other markets also decreased by the same amount. Concurrently, the directed risk spillover effects transmitted and received between markets diminished, such as the foreign exchange market’s risk spillover effects to and from other markets, which declined by 1.64 and 1.74 percentage points, respectively. In addition to total and directed risk spillovers, the paired market risk spillovers also decreased, particularly the bidirectional spillovers between the stock market and other markets. For example, the stock market’s risk spillovers transmitted to and received from the metals market decreased by 9.58% and 9.17%, respectively. However, changes in the foreign exchange market’s risk transmissions to and from the metals market were not significant.

From the perspective of risk spillover direction in financial markets considering ESG investments (i.e., the difference between the market’s risk spillover “To” and risk absorption “From”), it is observed that the net spillover effects are negative for the repurchase market, interbank bond market, exchange bond market, foreign exchange market, gold market, and real estate market, with respective values of −6.96%, −1.85%, −4.09%, −6.02%, −1.83%, and −3.16%. Conversely, the net spillover effects are positive for the ESG stock market, interbank lending market, metals market, energy market, and agricultural products market, recorded at 2.72%, 6.15%, 5.6%, 5.33%, and 4.11%, respectively.

The stock market, interbank lending market, and commodities market exhibit relatively higher external risk spillovers, while the repurchase, bond, foreign exchange, gold, and real estate markets are more impacted by risk from other markets. This indicates that China’s capital market is a source of financial system instability. The interbank and foreign exchange markets, subject to strict scrutiny by monetary authorities, are generally more stable. When considering external risk spillovers further, the conventional stock market exhibits the largest external risk spillover effect at 69.18%. However, when ESG stocks replace conventional stocks, the stock market’s external spillover level decreases to 66.1%, and it no longer represents the primary source of systemic risk within China’s financial markets. Instead, the energy market holds the highest position at 68.77%, assuming a leading role in the dissemination of information within China’s financial markets.

Additionally, in recent years, due to accelerated urbanization and increasing population inflow, the demand for real estate has continually risen. Government reliance on land lease revenues and adjustments in mortgage policies have also propelled housing prices. Investors seeking asset diversification have increasingly entered the real estate market, aiming for stable investment returns. This has made the bidirectional spillover effects between the real estate market and other markets particularly prominent.

In the analysis of pairwise market risk volatility spillover effects within the Chinese financial system, the stock market and the real estate market are significant components. A comparison of the risk spillover effects between these two markets reveals that the spillover from the stock market to the real estate market (38.93%) exceeds the spillover from the real estate market to the stock market (36.65%). Furthermore, the national trade structure and fiscal balance are significantly influenced by the foreign exchange and commodity markets. The risk spillovers from the commodity market to the foreign exchange market, recorded at 1.13%, 1.09%, and 0.82%, demonstrate a relatively high capacity for risk transmission.

Figure 1, generated using the TVP-VAR method, illustrates the dynamic total risk spillover effects in the Chinese financial market. As shown in Figure 1, the red line represents ESG investments, while the black segment denotes conventional stocks. The trends in total risk spillovers, both with and without specific consideration of ESG investments, appear similar; however, there are notable differences in specific values. The COVID-19 pandemic in 2020, which induced turbulence in the Chinese financial markets, resulted in greater total risk spillover effects associated with conventional stocks. Given this paper’s focus on the dynamic instability of China’s financial markets, subsequent analyses will disregard the minor differences in risk spillover effects between scenarios with and without ESG investment, focusing instead on the dynamic risk spillover of ESG stocks.

During the sample period, the risk spillover index for China’s financial markets fluctuated between 40% and 70%, exhibiting uncertainty and volatility. Overall, the total risk spillover index showed a downward trend, with significant increases during periods of supply–demand imbalance in 2013, stock market turbulence in 2015, and the outbreak of the COVID-19 pandemic in 2020, demonstrating the model’s effectiveness in capturing systemic risk changes. In 2015, following a rapid bull market and subsequent significant corrections in the Chinese stock market, financial deleveraging was reintroduced in 2016 to reduce financial system risk, ensure robust economic growth, and promote economic structural optimization. This led to a significant drop in the spillover risk index. Small increases occurred in 2017 and 2018 amid debt defaults and the crisis in internet finance. Thus, the risk spillover index calculated in this study accurately reflects the systemic risk level of China’s financial markets and provides regulatory authorities with an effective tool for dynamic risk monitoring.

Figure 2 and Figure 3, respectively, display the dynamics of risk influenced by external factors (“From”) and the dynamics of outward risk spillovers (“To”), while Figure 4 illustrates the direction of dynamic risk spillovers (i.e., the net value of “To” minus “From”). Figure 4 clearly shows that the commodity and stock markets are primary sources of volatility shocks, causing significant risk spillovers to other markets. In contrast, the banking and foreign exchange markets primarily absorb external risks, acting as net recipients of volatility shocks. Within the stock market, ESG investments exhibit less volatility in risk shocks compared to conventional stocks, particularly during the significant stock market surge in 2015, where the net spillover of risks in ESG investments only fluctuated slightly. In 2016, following adjustments in China’s stock market, governmental and regulatory measures such as the suspension of IPOs and the implementation of the “circuit breaker” mechanism were taken to stabilize the market. During this period, ESG investments primarily withstood volatility shocks.

The net risk spillover in the money market fluctuated frequently, with the interbank lending market typically presenting as a net emitter. In 2013, the significant risk spillover effects to other markets highlighted the notable influence of monetary policy through interest rates and liquidity. However, from 2016 onwards, the transmission effect of monetary policy interest rates was not significant, nor was its impact on other sub-markets noticeable. This may be due to the high leverage state of economic sectors in recent years in China. High leverage could limit firms’ willingness to finance through rate cuts or credit increases, thereby weakening the effectiveness of monetary policy to some extent.

Between 2013 and 2016, China’s foreign exchange market experienced phases of RMB appreciation and depreciation, along with intensified foreign exchange management and capital outflow controls. Particularly following the “8.11 Exchange Rate Reform” in 2015, which further opened China’s capital account, the foreign exchange market became more susceptible to external risk fluctuations. To curb speculative home buying and reduce mortgage risks, the government implemented restrictive purchase and loan policies starting in 2010 and further strengthened these measures in 2013 and 2017. The real estate market, most of the time, acted as a net recipient of volatility shocks, easily affected by external shocks, which in turn influenced the risk in other markets. The financial deleveraging in 2016 and the regulation of shadow banking significantly impacted the volatility in the real estate market.

4.2. Risk Spillover Network Analysis

Figure 5 displays the results of the volatility spillover network within China’s financial market. In this network, the size of nodes corresponds to net spillovers, while the direction and weight of edges represent the direction of risk spillovers and the net directed pairwise spillover effects, respectively. Figure 5 illustrates that during periods of extreme risk events, the density of the financial volatility spillover network increases significantly, as does the number and strength of connections. After the 2010–2013 financial crisis adjustment period, China underwent a series of policy shifts aimed at addressing challenges following the international financial crisis, including inflation, overheating of the real estate market, and deepening reforms of the financial markets, which resulted in tighter network connections. Subsequent periods include the 2015 stock market turbulence, US–China trade tensions, and the COVID-19 pandemic. During 2015, the money and foreign exchange markets were the most interconnected, while other markets were relatively isolated. This indicates that the People’s Bank of China’s adjustment of the RMB central parity pricing mechanism, which led to a rapid depreciation of the RMB against the dollar and significant capital outflows, impacted the liquidity of the money market and was closely linked to the foreign exchange market. Furthermore, the gold market shifted from being a net recipient to a net emitter of risk volatility following the financial crisis adjustment period. During the 2007–2009 financial crisis, the extreme uncertainty and turmoil in the financial markets drove many investors towards gold as a safe-haven asset, causing its price to rise sharply. As the market gradually recovered from the crisis, investors’ risk appetites began to normalize, reducing the demand for gold as a haven, thereby affecting the volatility in the gold market. In the volatility spillover network, financial sub-markets with greater propagative force have a broader impact, and the influence of volatility within the network is more pronounced.

To gain a deeper understanding of the time-varying characteristics of risk spillover effects in China’s financial markets and to evaluate the performance of ESG investments during extreme risk events, this study compares the risk inputs, outputs, and net spillover indices of stock markets that include ESG stocks versus traditional stocks across four risk events.

Table 3. Risk spillover in financial market network nodes.

	Post-Financial-Crisis Adjustment Period			Stock Market Turbulence Period			US–China Trade Friction Period			COVID-19 Pandemic Period
	From	To	Net	From	To	Net	From	To	Net	From	To	Net
Stock	70.14	70.2	0.06	81.85	82.49	0.64	73.91	83.6	9.69	74.09	86.82	12.73
	(73.4)	(76.7)	(3.28)	(84.4)	(85.2)	(0.82)	(74.3)	(84.6)	(10.3)	(76.1)	(89.1)	(13)
DR007	36.38	32.62	−3.76	88.3	95.48	7.18	26.5	14.21	−12.3	55.6	52.74	−2.87
	(36.3)	(32.1)	(−4.17)	(87.9)	(95)	(7.19)	(26.4)	(14.4)	(−12.06)	(55.6)	(52.4)	(−3.18)
Shibor	21.68	35.86	14.18	86.28	97.39	11.11	20.97	14.16	−6.8	51.1	36.71	−14.4
	(21.4)	(36)	(14.5)	(85.7)	(97)	(11.4)	(21.2)	(14.9)	(−6.3)	(51)	(37)	(−13.94)
CGBI	94.35	87.34	−7.01	93.01	85.12	−7.89	94.29	92.28	−2.01	96.57	81.64	−14.9
	(94.5)	(86.9)	(−7.59)	(93.2)	(85.2)	(−7.98)	(94.3)	(92.5)	(−1.8)	(96.7)	(82.1)	(−14.55)
CGBNI	95.23	86.78	−8.46	93.74	83.6	−10.1	94.45	88.18	−6.27	96.56	86.9	−9.66
	(95.3)	(86.4)	(−8.85)	(93.9)	(83.6)	(−10.26)	(94.5)	(87.3)	(−7.14)	(96.6)	(87.3)	(−9.3)
USD_CNY	24.77	11.03	−13.7	24.05	15.95	−8.1	14.51	30.78	16.28	34.08	46.31	12.23
	(24.8)	(11.1)	(−13.71)	(24.4)	(16.1)	(−8.36)	(14.9)	(30.6)	(15.8)	(33.7)	(45.9)	(12.2)
NMI	74.25	85.58	11.34	68.52	69.09	0.58	51.31	59.21	7.9	60.09	71.42	11.33
	(74.8)	(86.9)	(12.1)	(68.1)	(69.1)	(1)	(51.4)	(59.8)	(8.37)	(60.5)	(71.8)	(11.4)
NCI	75.07	85.15	10.08	73.28	75.24	1.97	51.42	51.82	0.4	50.31	57.02	6.71
	(75.2)	(85.9)	(10.7)	(73.2)	(75)	(1.88)	(51.5)	(51.8)	(0.29)	(49.9)	(56.8)	(6.94)
NAI	53.03	59.49	6.47	48.09	56.49	8.41	21.43	14.56	−6.87	44	52.44	8.44
	(53)	(58.8)	(5.78)	(48.2)	(56.5)	(8.37)	(21.6)	(14.3)	(−7.31)	(43.9)	(53.1)	(9.13)
AU	27.79	28.39	0.61	15.16	12.41	−2.75	18.13	17.82	−0.31	21.54	11.8	−9.74
	(27.6)	(28.2)	(0.57)	(15.1)	(12.5)	(−2.61)	(17)	(17.1)	(0.16)	(22.6)	(12.1)	(−10.48)
House	67.88	58.13	−9.76	82.33	81.33	−1	71.85	72.13	0.28	64.47	64.61	0.14
	(68.4)	(55.7)	(−12.69)	(84.4)	(83)	(−1.43)	(71.6)	(71.3)	(−0.28)	(67.6)	(66.4)	(−1.2)
Total Risk Spillover	58.23			68.6			48.98			58.95
	(58.6)			(68.94)			(48.97)			(59.46)

reveals significant differences in network node risk spillovers across various periods. After replacing the volatility of the conventional stock index with the ESG stock index, the post-financial-crisis adjustment period saw the most substantial changes in net spillover for the stock and real estate markets. Specifically, the stock market’s net spillover decreased dramatically from 3.28% to 0.06%, a decline of 98.17%, while the real estate market altered by 23.09%. During this period, in markets considering ESG investments, the highest levels of net risk spillover were observed in the money and metals markets, recorded at 14.52% and 12.13%, respectively, positioning them as net exporters of risk. Conversely, the foreign exchange and real estate markets exhibited net risk spillovers of −13.74% and −9.76%, marking them as net importers of risk.

During periods of stock market turbulence, significant changes in net risk spillovers were seen in the metals, real estate, and stock markets, with variations of 42%, 30.07%, and 21.95%, respectively. Notably, the stock market in the ESG-focused financial markets shifted from being a primary transmitter of volatility shocks to a secondary role, with the money market taking the lead, and the bond market emerging as a primary receiver of risk spillovers.

In the period of US–China trade tensions, the gold and real estate markets experienced substantial changes in net risk spillover, at 293.75% and 200%, respectively. After accounting for ESG investments, the gold market shifted from being a transmitter to a receiver of risk, and the real estate market transitioned from a receiver to a transmitter. Both the bond and stock markets functioned as major sources and receivers of risk in the financial markets.

During the COVID-19 pandemic, significant variations occurred in the real estate market, the repo sector of the money market, and the agricultural products market, with changes of 111.67%, 9.75%, and 7.56%, respectively. The real estate market also transitioned from being a risk receiver to a risk transmitter, whereas the money market remained a receiver, and the agricultural products market became a transmitter of risk.

When ESG investments replace conventional stocks, the total risk spillover in China’s financial markets shows varying magnitudes during different periods. The largest change occurred during the COVID-19 pandemic in 2020 (0.86%), followed by the post-financial-crisis adjustment period (0.63%) and the stock market turbulence period (0.49%). During the US–China trade friction period, the risk spillover increased by 0.02% when considering ESG investments.

ESG (Environmental, Social, and Governance) investments focus on companies’ performance in environmental protection, social responsibility, and good corporate governance. These companies are often more robust and sustainable, potentially responding differently to risks compared to traditional stocks in certain economic and financial environments. Possible explanations for the observed trends include the following: The COVID-19 pandemic significantly disrupted global economies, causing business interruptions and financial difficulties for many companies. However, companies adhering to ESG principles, with better risk management and healthier financial conditions, exhibited greater resilience. This resilience likely contributed to the reduced risk spillover among financial markets when considering ESG investments. In the post-financial-crisis period, investors became more cautious, seeking stable and reliable investment opportunities. Consequently, during this adjustment period, ESG investments gained popularity due to their sustainability and focus on future risks, potentially helping to reduce market risk spillovers. Unlike the previous contexts, the US–China trade friction likely had a direct impact on all companies, regardless of their adherence to ESG principles. Therefore, although the risk spillover may have increased even with ESG investments, the magnitude of this increase was relatively small.

4.3. Risk Investment Strategies in China’s Financial Markets

Following the analysis of the volatility spillover effects of ESG investments, this paper employs the method of Antonakakis et al. (2018) [47] to conduct a portfolio analysis of the volatility indices of ESG stocks and other financial markets. The goal is to provide strategic guidance for investors in asset allocation. The volatility of financial asset returns often exhibits significant nonlinear correlations. For example, due to the high interconnectivity of the Shanghai and Shenzhen stock markets, their volatility is closely linked to that of other financial markets. Additionally, the analysis of volatility spillovers highlights the leading role of the stock market in propagating external risks.

By examining the time-varying relationships between the volatility indices of ESG stock markets and other markets, the study further explores strategies for diversified volatility index portfolios and calculates optimal hedge ratios and portfolio weights.

Table 4. Dynamic optimal hedge ratios.

	Mean	Std. Dev.	5%	95%
Stock/DR007	−0.11	1.3	−1.61	1.03
Stock/Shibor	0.11	0.98	−0.82	1.42
Stock/CGBI	−1.77	3.48	−8.14	3.48
Stock/CGBNI	−1.54	2.68	−6.34	2.69
Stock/USD_CNY	−1.82	7.47	−14.77	1.52
Stock/NMI	0.53	0.34	0.03	1.09
Stock/NCI	0.38	0.24	0.01	0.77
Stock/NAI	0.44	0.43	−0.25	1.18
Stock/AU	0.01	0.37	−0.65	0.56
Stock/House	0.55	0.17	0.24	0.79
DR007/Stock	0.01	0.09	−0.11	0.16
Shibor/Stock	0	0.09	−0.12	0.12
CGBI/Stock	−0.01	0.01	−0.03	0.01
CGBNI/Stock	−0.01	0.02	−0.04	0.02
USD_CNY/Stock	0	0.03	−0.06	0.04
NMI/Stock	0.22	0.13	0.02	0.46
NCI/Stock	0.27	0.16	0.01	0.52
NAI/Stock	0.1	0.1	−0.05	0.29
AU/Stock	0.01	0.13	−0.19	0.2
House/Stock	0.92	0.24	0.41	1.2

presents the basic statistics of the optimal hedge ratios between ESG investments and other volatility indices derived using the DCC-GARCH t-copula method. The mean hedge ratios between ESG long positions and short positions in other financial market volatilities range from −11 to 0.55. This indicates that, on average, ESG stocks have an inverse relationship with DR007, CGBI, CGBNI, and USD_CNY, reflecting a flight to safety from risk assets during periods of market risk aversion.

The USD_CNY exhibits a large standard deviation of 7.47, indicating high volatility in the hedge ratios with ESG stocks, which reflects substantial market risk. Conversely, the average hedge ratios of ESG stocks against Shibor, NMI, NCI, NAI, and House are positive, indicating a direct relationship between stock prices and these indices. These indices have smaller standard deviations, such as the real estate market’s 0.17, suggesting lower volatility and risk in the associated hedge ratios. Additionally, Table 4 indicates that these optimal hedge ratios are not static but vary over time, as shown by the standard deviations and the 5th and 95th percentiles. Figure 5 corroborates the time-varying nature of the optimal hedge ratios, indicating that investors should adopt dynamic rather than static hedging strategies. On the other hand, when ESG stocks (Stock) are used as hedging assets, their volatility is relatively lower. For instance, the hedge ratios between Stock/DR007 and Stock/CGBI show significant volatility, whereas DR007/Stock and CGBI/Stock exhibit the opposite trend. Thus, investors using ESG stocks for hedging can benefit from lower hedging costs and relatively stable hedge ratios.

Table 5. Dynamic optimal portfolio weights.

	Mean	Std. Dev.	5%	95%
Stock/DR007	0.2	0.23	0	0.72
Stock/Shibor	0.21	0.22	0	0.68
Stock/CGBI	0.01	0.01	0	0.04
Stock/CGBNI	0.02	0.02	0	0.04
Stock/USD_CNY	0.06	0.08	0	0.24
Stock/NMI	0.23	0.2	0	0.63
Stock/NCI	0.39	0.17	0.12	0.68
Stock/NAI	0.13	0.09	0	0.31
Stock/AU	0.25	0.11	0.1	0.44
Stock/House	0.9	0.13	0.63	1

displays the proportions of each asset in the portfolio under long-term averages, with the portfolio consisting of ESG stocks and other asset volatility indices. For example, in a unit Stock/House portfolio, 0.9 should be invested in ESG stocks and 0.1 in the real estate market (House). In most portfolios, ESG stocks do not occupy a high weight, with Stock/House being the only exception. Like the optimal hedge ratios, the optimal weight graph in Figure 6 and the standard deviations, the 5th and 95th percentiles in Table 5 exhibit time-varying characteristics. This dynamic nature of the strategy further confirms the need for active portfolio rebalancing rather than a static approach.

To further evaluate the strategy, two criteria need to be considered: hedge effectiveness and profitability. Figure 7 and Column (1) of

Table 6. Hedge effectiveness of diversified strategies.

	Dynamic Portfolio Weights		Dynamic Hedge Ratios
	HE	p-Value	HE	p-Value
Stock/DR007	0.74 ***	0	−7.55	0.88
Stock/Shibor	0.75 ***	0	−4.47	0.51
Stock/CGBI	0.99	0.19	−0.01	0.86
Stock/CGBNI	0.99	0.11	−0.01	0.91
Stock/USD_CNY	0.9 ***	0	−1670.9	0.77
Stock/NMI	0.57 ***	0	0.11 **	0.02
Stock/NCI	0.54 ***	0	0.09 **	0.02
Stock/NAI	0.79 ***	0	0.05	0.14
Stock/AU	0.72 ***	0	−0.01	0.72
Stock/House	0.02 ***	0	0.64 ***	0
DR007/Stock	0.66 ***	0	−0.01 ***	0
Shibor/Stock	0.65 ***	0	−0.02 ***	0
CGBI/Stock	0.04 ***	0	0.01 ***	0
CGBNI/Stock	0.04 ***	0	0.01 ***	0
USD_CNY/Stock	−0.41 ***	0	−0.01 ***	0
NMI/Stock	0.36 ***	0	0.09 ***	0
NCI/Stock	0.37 ***	0	0.09 ***	0
NAI/Stock	0.14 ***	0	0.06 ***	0
AU/Stock	0.33 ***	0	0.01 ***	0
House/Stock	0.39	0.53	0.61 ***	0

Note: ***, ** denote significance at 1%, 5% level, respectively.

show that risk can be reduced by using dynamic hedge ratio strategies and dynamic portfolio weights compared to unhedged positions in ESG stocks. Hedge effectiveness measures the relative efficiency of risk reduction in hedging transactions. High hedge effectiveness indicates a significant risk reduction in portfolio risk management at a lower cost. Column (2) of Table 6 presents the optimal weight hedge effectiveness of the portfolio, demonstrating that ESG stocks are significantly effective at the 1% level in hedging against indices other than bond market indices. Similarly, other indices, except for the real estate index, show significant hedge effectiveness against ESG stocks at the 1% level. Using the foreign exchange index (USD_CNY) and capital market indices, ESG investments achieve the highest effectiveness. In other financial markets, employing ESG investments for hedging purposes is also shown to be most efficient.

5. Discussion

The findings of this study align with and expand upon existing research on the role of ESG investments in stabilizing financial markets, particularly in mitigating systemic risks and managing volatility. Previous studies, such as Kumar et al. (2021) [42], have demonstrated that ESG investments enhance market stability by attracting long-term capital and fostering responsible corporate governance. These findings are corroborated by this study, which reveals that ESG investments significantly mitigate volatility spillovers in China’s financial markets, effectively reshaping the structure of risk contagion pathways.

The results also support the conclusions of Feng et al. (2022) [14], who identified a strong link between high ESG ratings and reduced crash risks. By employing dynamic econometric models, including the TVP-VAR model and the DCC-GARCH t-copula, this study provides a deeper understanding of how ESG investments influence risk transmission, particularly during periods of extreme market stress, such as the COVID-19 pandemic. These findings underscore the critical role of ESG investments in enhancing market resilience, reaffirming the dual benefits of financial performance and sustainability highlighted by Friede et al. (2015) [27].

However, divergent views in the literature are acknowledged. Liu (2022) [13] argued that ESG strategies may underperform under extreme market conditions due to inconsistent standards and inadequate data quality, particularly in developing economies like China. While the findings of this study align with concerns about data quality challenges, they also demonstrate that ESG investments, when effectively integrated, can provide significant benefits in reducing systemic risks and optimizing portfolio performance.

This research advances prior network-based risk analysis approaches, such as those introduced by Diebold and Yilmaz (2012) [39], by employing a novel combination of spillover indices and complex network models to analyze risk contagion pathways. The results indicate that ESG investments significantly alter the structure and intensity of risk propagation within financial systems, offering valuable insights for policymakers aiming to enhance systemic stability.

Furthermore, by examining portfolio optimization strategies, this study complements the findings of Hoepner et al. (2016) [8], who emphasized the importance of ESG investments in achieving superior risk-adjusted returns. The results demonstrate that portfolios incorporating ESG assets exhibit improved hedging effectiveness and reduced systemic risks, strengthening the case for ESG integration in asset management.

In conclusion, this study reaffirms the stabilizing effects of ESG investments while offering novel insights into their impact on risk contagion and portfolio resilience. These findings carry significant implications for policymakers and investors, particularly within the context of China’s financial market reforms and sustainability objectives. Future research should explore the long-term systemic effects of ESG investments, focusing on challenges such as data standardization and cross-market comparability.

Despite its valuable contributions, this study has certain limitations that must be acknowledged. First, the analysis focuses exclusively on China’s financial markets, and while the findings offer important insights, they may not be fully generalizable to other regions with different regulatory frameworks and market structures. Second, the study relies on specific econometric models, including the TVP-VAR and DCC-GARCH t-copula models, which, although robust, are sensitive to parameter selection and data quality. Alternative modeling approaches could provide additional perspectives. Third, the availability and quality of ESG data pose significant challenges, particularly in emerging markets. Inconsistent reporting standards and limited disclosure practices may introduce biases into the results. Lastly, the study examines a defined timeframe, which may not fully capture long-term trends or the evolving nature of ESG investments in response to regulatory developments and global sustainability initiatives. Nonetheless, this study lays a strong foundation for understanding the impact of ESG investments on financial stability and offers valuable avenues for future research to address these limitations.

6. Conclusions

In the post-financial-crisis era, China has experienced significant economic changes driven by the “Four Trillion” stimulus plan, leading to the rapid expansion of the shadow banking system and increasing local government debt levels. High leverage and rising property prices have become two prominent challenges in the Chinese economy, fostering potential systemic risks. The 2015 stock market turbulence, frequent bond defaults in 2018, and the worsening US–China trade tensions have all exacerbated threats to financial stability, intertwining risks between financial markets and the real economy. Against the backdrop of transitioning to a low-carbon economy and heightened financial market uncertainty, the relationship between sustainability and stability has become a crucial issue for economic development.

This study examines the relationship between ESG investments and financial market risk spillovers in China, aiming to reduce financial market risk spillovers and maintain market stability. Using the Time-Varying Parameter Vector Autoregression (TVP-VAR) model, we constructed a variance decomposition spillover index to analyze the spillover effects among different markets within the Chinese financial system, capturing both the intensity and direction of these spillovers. To identify the relationship between ESG investments and the stability of China’s financial markets, we compared the spillover effects in financial markets focusing on ESG stocks versus conventional stocks. Additionally, from the perspective of risk volatility spillover information transmission, we constructed a risk network of the Chinese financial system under extreme risk events, analyzing the differences in risk transmission effects between stock markets that include ESG stocks and those that do not. Based on this, we used the DCC-GARCH t-copula model to calculate the optimal portfolio strategy within China’s financial market, including ESG stocks. The conclusions of this study are as follows:

Firstly, ESG investments significantly reduce overall systemic risk, net spillover effects, and pairwise net spillover effects in China’s financial markets, thereby enhancing market stability. This suggests that strengthening ESG investments in publicly listed companies can further stabilize China’s financial markets. This stability partially stems from the core principles of ESG investing, which encourage companies to engage more proactively in environmental, social, and governance activities. Such proactive management fosters a more sustainable and healthier socio-economic environment, enabling investors to better balance economic and non-economic factors in their decisions. This trend is likely to attract more capital into ESG investments and promote the development of long-term investment strategies, revealing and nurturing the potential value of ESG companies. China’s stock market is predominantly influenced by irrational individual investors rather than institutions, creating a challenging investment environment [51]. To address this excessive market volatility, promoting long-term investment strategies has become a key task for policymakers. This study finds that market instability is a major factor in China’s financial turbulence. By adopting ESG investments, investors are more likely to follow long-term strategies, thereby mitigating associated risks.

Secondly, the financial volatility spillover network incorporating ESG investments exhibits structural changes under extreme risk events. During the post-financial-crisis adjustment period, the 2015 stock market crash, the US–China trade tensions, and the COVID-19 pandemic, the financial network connections became tighter. The ranking of total risk spillover intensity is as follows: 2015 stock market crash > US–China trade tensions > COVID-19 pandemic. When ESG investments replace conventional stocks, the ranking of total risk spillover changes, as follows: COVID-19 pandemic > post-financial-crisis adjustment period > stock market turbulence period > US–China trade tensions. During the COVID-19 pandemic, significant changes occurred in the real estate, money, and agricultural markets. The real estate market shifted from being a risk receiver to a risk transmitter, the money market remained a risk receiver, and the agricultural market became a risk transmitter.

Extreme risk events typically have a broad impact, rapid spread, and strong shock effect, which traditional regulatory models may not fully address. Therefore, regulatory agencies should enhance proactive risk prevention for extreme events and promote financial stability strategies centered around ESG investments. By analyzing the volatility spillover network, risk isolation mechanisms can be established to reduce the pathways of systemic risk transmission.

Thirdly, the return volatility spillover among variables in China’s financial markets shows significant nonlinearity. To mitigate risk spillovers primarily originating from the stock market, active portfolio management measures should be adopted rather than relying solely on static strategies. When investors use ESG stocks for hedging, the costs are relatively low, and the hedge ratios are stable. In terms of paired asset hedging performance, ESG investments are most effective when paired with foreign exchange and capital market indices. Similarly, in other financial markets, the capital market achieves the greatest risk hedging effect by using ESG investments.