This study examined the dynamics of climate change and renewable energy in the GCC region. First, climate change is represented by regional temperature data from the GCC countries. Renewable energy is measured by the wilder hill clean energy index (ECOI), which comprises publicly listed companies that benefit significantly from the transition to cleaner energy and conservation. The regional temperature data of the GCC countries were used as a proxy to capture climate variability that directly affects energy demand, particularly in power-intensive sectors requiring cooling (Yuan et al., 2024). The ECOI is selected to represent renewable energy performance due to its broad coverage of publicly traded clean energy firms with global operations, including exposure to markets relevant to the GCC’s diversification strategies (Hossain et al., 2024, Korkut et al., 2024). morgan stanley capital international (MSCI) country indices for GCC members are used to capture financial market performance in a standardized, comparable format across the region, ensuring consistency in measuring cross-country spillovers (Eleuch et al., 2025). Additionally, the GCC-MSCI index data were sourced from datastream, covering individual MSCI indices for the GCC markets from August 29, 2014, to December 2023. The choice of this time frame is driven by the fact that MSCI data for the largest economy in the region, Saudi Arabia, are available only from August 29, 2014, onwards. Hence, selecting this period ensured consistency and completeness across the GCC markets. Furthermore, the GCC-MSCI index is equally weighted and adjusted, making it a robust representation of stock market dynamics in GCC economies.

Table 1b presents the dataset, which consists of two panels. panel A focuses on the level indices, while panel B concentrates on returns for variables related to climate change, renewable energy, and MSCI prices of the GCC countries. From Table 1b, panel A (level indices), the mean temperature (TEMP) is 300.89 K, which indicates the average regional temperature in the GCC. The mean renewable energy index is 80.28, reflecting the average stock price of companies benefiting from the transition to cleaner energy. The mean MSCI prices for individual GCC countries vary, with Saudi Arabia having the highest at 1038.80 and Bahrain the lowest at 98.74. The standard deviation for temperature is 5.44, indicating relatively low variability in temperature, whereas the renewable energy index has a higher standard deviation of 48.30, suggesting more significant fluctuations in stock prices. Among the GCC countries, Saudi Arabia exhibits the highest standard deviation of 231.28, indicating considerable price volatility in its market, whereas Bahrain has a lower standard deviation of 23.34, suggesting greater stability. The maximum and minimum temperatures recorded were 308.83 and 289.32 K, respectively. The renewable energy index shows a maximum of 281.44, highlighting potential peaks in renewable energy stock performance, while the minimum is 36.53, indicating possible downturns in performance. The renewable energy index displayed a positive skewness of 1.71, suggesting a distribution with more extremely high values. The kurtosis of 5.45 indicates a distribution with heavy tails, implying that extreme outcomes are more likely.

From Table 1b, panel B (return), the mean return for renewable energy index is −0.002, indicating a slight average decline in stock performance. The mean returns for the GCC countries range from −0.007 (United Arabi Emirates) to 0.003 (Kuwait), suggesting mixed performance across the markets. The renewable energy index exhibited a standard deviation of 1.018, indicating high variability in returns. In contrast, Kuwait has a standard deviation of 0.530, indicating more stable returns than those of other GCC markets. The maximum return for renewable energy index is 10.041, while the minimum is −9.531, highlighting the extreme volatility in performance. For the GCC countries, returns range from a maximum of 4.719 (Oman) to a minimum of −9.820 (Kuwait). The skewness for the renewable energy index is −0.152, indicating a slight negative skew, whereas GCC countries, such as Bahrain, have a skewness of −1.968, suggesting more extreme negative returns. The kurtosis values, particularly for Oman (99.71), imply a distribution with extreme outliers that affect overall market stability. The augmented Dickey-Fuller (ADF) and Phillips-Perron (PP) test results indicate that all variables are stationary at first differences, as evidenced by significant test statistics, which is crucial for subsequent time series analyses.

The quantile vector autoregressive model, across extreme lower, higher, and median quantiles, enables us to capture connectivity metrics that reflect the behavior of economic relationships under both typical and extreme conditions (Diebold and Yilmaz, 2012, Diebold and Yilmaz, 2014). The transmission of shocks within the economic framework, particularly through the lens of the quantile vector autoregressive model, is a crucial area of research for understanding interconnected financial markets, macroeconomic fluctuations, and the impact of monetary policy. Early studies by Sadorsky, 2009, Henriques and Sadorsky, 2008were among the first to model the relationship between energy prices, clean energy stocks, and market volatility. The literature, including these early studies, has not adequately emphasized how quantile vector autoregressive can elucidate the mechanisms of shock transmission across diverse economic variables and markets. Similarly, the connectedness framework introduced by Diebold and Yilmaz, 2012, Diebold and Yilmaz, 2014 laid the groundwork for quantifying spillover effects across financial variables. These seminal contributions provide the theoretical scaffolding upon which this study expands, using quantile and wavelet-based innovations. However, unlike traditional models that primarily focus on average interactions, this approach provides insights into how variables interact at the tails of the distribution. By integrating the quantile-linkage method, we specifically analyze the spillover effects between temperature, renewable energy, and market returns, providing a comprehensive view of their interconnectedness. This dual emphasis enhances our understanding of risk transmission and interdependencies, offering deeper insights that standard methodologies may overlook. Ultimately, our approach enables a more nuanced assessment of economic dynamics within the context of GCC economies. quantile vector autoregression models offer an advancement over the traditional vector autoregression by allowing researchers to account for the dynamics of economic relationships across different quantiles of the conditional distribution of the variables involved. There are several reasons why quantile vector autoregression and wavelet quantile correlation approaches are particularly suitable for modeling financial and climate-induced shocks in the GCC. Contrary to standard vector autoregression or time-varying parameter vector autoregression approaches, which assume symmetric responses around the mean, quantile vector autoregression models allow for tail-dependent and nonlinear spillovers, which are particularly important in cases of extreme events (e.g., oil crashes, pandemics, and climate shocks). In addition, although generalized autoregressive conditional heteroskedasticity and multivariate – generalized autoregressive conditional heteroskedasticity type models can capture volatility clustering, they are not appropriate for studying directional spillovers or network-based connectedness among quantiles. Although cointegration or fully modified ordinary least squares / autoregressive distributed lag model models are robust for long-run equilibrium analysis, they are static and cannot capture time-varying linkages. In contrast, wavelet quantile correlation offers a frequency-domain view, allowing us to decompose short, medium, and long-term dependencies - important in disentangling transient shocks from structural changes (in particular, we refer to the transition towards clean energy). The quantile vector autoregression and the wavelet quantile correlation, taken together, offer a powerful modeling environment that is particularly suited for time-frequency and asymmetric spillovers, which are precisely the features observed in GCC stock, temperature, and renewable energy markets.

We apply the quantile vector autoregression model at the extreme lower, higher, and median quantiles, adhering to the quantile regression methodology of (Markowitz, 1952). Consequently, we can now observe the connectivity metrics at the extreme tails of the distribution. Beyond concentrating on median quantile models with means, this technique offers other characteristics and applications. Hence, we employ the quantile-linkages method proposed by Ando et al. (2022) to examine quantile spillover linkages between temperature, renewable energy, and market returns in GCC economies from August 29, 2014, to December 2023. Although a considerable body of literature has explored shock transmission, the emphasis on quantile vector autoregression models and their capabilities remains underexplored. The lag order for the quantile vector autoregression model was determined using the akaike information criterion (AIC), and lag one was found to be optimal, striking a balance between model fit and complexity. Before estimation, the statistical properties of all variables were examined concerning stationarity using the Phillips Perron unit root tests. All series were found to be stationary at first differences. To mitigate the endogeneity problem, the model employed the generalized forecast error variance decomposition method, which is independent of the ordering of the variables. We further used rolling window estimation to (a) incorporate time-varying effects and (b) mitigate sensitivity to initial conditions.

To test the robustness of our findings, we performed various sensitivity analyses using both rolling windows (250-day) and thresholds (τ = 0.25 and 0.75). The key findings—especially that Saudi Arabia, the United Arab Emirates, and Kuwait led the way—were unchanged. We cross-validated the quantile vector autoregression -based spillovers with wavelet multiple correlation outputs, which provided further support for the temporal and directional reliability of the realized associations. Robustness results are provided in the appendix. By employing quantile vector autoregression frameworks, researchers can more effectively capture the complexities of shock interactions and their diverse impacts on various economic variables and markets. This understanding is pivotal for developing more effective economic policies and frameworks that mitigate the adverse effects of shocks, enhance financial stability, and promote a resilient economic environment in the country. Defining the infinite-order-based vector moving average parameters of the quantile vector autoregression, the quantile spillover measures are computed as follows:using the generalised forecast error variance decomposition constructed as follows, and a prediction horizon of H (Eleuch et al., 2025, Diebold and Yilmaz, 2012).

 finds a zero-dimensional vector in which the ith position has the value 1. The following steps show how the elements are normalised according to the decomposition matrix:

The following explanations of the generalised forecast error variance decomposition -based spillover measurements are defined by (Hossain et al., 2024):

The term TO j, t represents the impact of variable j on variable i, indicating how j influences i. Conversely, FROM j,t reflects the effect of i on j, demonstrating an influence in the opposite direction. The net spillover, denoted as NET j,t, is the difference between TO and FROM, where negative values indicate that j is a net recipient of spillover, while positive values signify that it is a transmitter. Finally, TCIt represents the total connectedness index, summarizing the mean degree of overall linkages among the variables over time. This methodology offers several benefits for understanding the interconnected systems. First, it distinguishes between TO j,t and FROM j,t, providing insights into both the directional influences and the reverse effects. Second, the NET j,t calculation identifies the net transmitters and recipients of spillover effects, which is crucial for determining which variables drive changes. Third, the total connectedness index summarizes overall interconnectedness, enabling policymakers to assess resilience and vulnerability to shocks. Ultimately, the quantile analysis reveals the relationships across various conditions, enhancing risk management strategies related to climate change, renewable energy, and market dynamics. Furthermore, we consider the wavelet quantile approach for the analysis. This indicates the overall reliance caused by the combined time-frequency quantile ranges. It recognizes the dependency pattern over both quantiles and intervals of climate change. Renewable energy and GCC stock markets. Our analysis considered different frequencies and quantiles of the data.

Wavelet quantile correlation is a statistical approach that combines wavelet decomposition with quantile correlation to examine the dependence between two time series across both time and distributional dimensions. Unlike traditional correlation measures that capture only linear relationships around the mean, wavelet quantile correlation allows researchers to explore how variables relate at different quantiles, such as in the tails or the center of the distribution, while simultaneously distinguishing between short-, medium-, and long-term dynamics through wavelet analysis. This dual framework makes wavelet quantile correlation particularly well-suited for detecting nonlinear and asymmetric dependencies, especially under extreme conditions. wavelet quantile correlation is computed as followsWhere:

 are detail (wavelet) coefficient of series X and Y at scale j (short to long-term dynamics).

Qcovτ is the quantile covariance between X and Y

 gives quantile-specific dependence at time scale j

Blue nodes represent net shock transmitters, and net shock receivers are depicted in yellow. The thickness of the connecting lines indicates the strength of pairwise directional connectedness, and the size of each node reflects its overall network impact, signifying the level of influence or dependence each economy or sector has on others. However, the GCC energy sector has low spillovers, whereas finance has the highest (Kapar et al., 2024). During periods of economic prosperity, represented by the extreme upper quantile (τ = 0.90), Saudi Arabia, the United Arab Emirates, and Kuwait act as primary shock transmitters. These major economies, with their substantial oil resources and diverse investments, have a significant influence on regional dynamics, particularly during periods of high economic activity. Conversely, Oman, Qatar, Bahrain, and temperature are shock receivers, demonstrating their dependence on the economic actions of more dominant regional players such as Saudi Arabia. During these boom times, renewable energy index remains more responsive to the broader economic climate, suggesting that conventional energy sectors continue to lead in driving growth. Under typical economic conditions, as depicted by the median quantile (τ = 0.50), the network structure becomes more evenly distributed. Saudi Arabia, the United Arab Emirates, and temperature emerge as shock transmitters, reflecting their increasing influence on regional economic activities. temperature's role as a transmitter indicates that environmental factors, such as climate risks, are becoming increasingly important in shaping economic policy and energy consumption patterns. Meanwhile, renewable energy index, along with Qatar, Oman, Bahrain, and Kuwait, remains a shock receiver, suggesting that these economies and the renewable energy sector are more affected by external conditions than they are drivers of regional economic shifts. In times of economic hardship, as indicated by the extreme lower quantile (τ = 0.05), Saudi Arabia, the United Arab Emirates, and temperature continue to play a primary role as shock transmitters, significantly impacting the wider region, consistent with (Kapar et al., 2024).

The renewable energy index and smaller economies, such as Oman, Qatar, and Bahrain, remain shock receivers, demonstrating their reliance on larger, oil-dependent economies and the emphasis on conventional energy sectors over renewable sources during challenging times. From an economic perspective, this network configuration underscores the significant influence of Saudi Arabia, the United Arab Emirates, and Kuwait in shaping the regional economy, primarily through their substantial oil production and diverse economic portfolios. These economies consistently send shocks throughout the region, underscoring their impact during both prosperous and challenging periods. Oman, Qatar, and Bahrain exhibit greater vulnerability and dependence on the economic activities of their larger neighbors, particularly during downturns. While growing in significance, the renewable energy sector remains largely influenced by broader economic conditions, with the renewable energy index often serving as a shock absorber. Similarly, investors turn to renewable energy and innovation for financial stability and risk management (Lorente et al., 2023).

For renewable energy index, correlations tend to strengthen over medium- and long-term periods in major markets such as the United Arab Emirates and Saudi Arabia. The short-term effects of renewable energy on market returns are relatively minimal, indicating that renewable energy investments and initiatives often require extended periods to have a noticeable impact on market performance. As renewable energy projects and policies gain traction in the medium term, their impact on economic performance becomes increasingly noticeable. The long-term integration of renewable energy into the broader economy is most apparent in countries actively diversifying away from oil. Conversely, Qatar and Oman exhibit weaker long-term impacts, suggesting that renewable energy is less integrated into their economic frameworks than their dependence on conventional energy sources.

This temporal analysis highlights that climate change has a more significant impact on GCC markets in the medium to long term, while the influence of renewable energy increases over extended periods, particularly in more diversified economies such as the United Arab Emirates and Saudi Arabia.