Quantile Domain Connectedness Between Climate Risks and Cryptocurrency Classes

Abstract

This research article explores whether the climate transition risk (CTR) and climate physical risk (CPR) transmit greater shocks towards the sustainable, gold-backed, energy-related and Sharia-compliant cryptocurrencies during bullish market conditions as compared with the normal and bearish market conditions. We employ the novel quantile vector auto-regression (QVAR)-based connectivity framework. Overall findings suggested that CPR and CTR transmitted greater shocks towards cryptocurrency classes during extremely high and lower quantiles as compared with the median quantile. This U-shaped and non-linear climate risks shock transmission indicates that Sharia-compliant, energy-related and gold-backed cryptocurrencies become more vulnerable during extreme market conditions (higher and lower quantiles) and may not consistently serve as reliable hedging or diversification instruments, particularly during periods of heightened climate uncertainty. Overall findings suggested that both the CPR and CTR transmitted greater shocks towards energy-related, gold-backed, and Sharia-compliant cryptocurrencies as compared with the sustainable cryptocurrencies, across all the quantiles. Therefore, sustainable cryptocurrencies, particularly those with energy-efficient consensus mechanisms such as Stellar, Cardano and Ripple, exhibited resilience to climate risks and can therefore function as stabilizing core holdings in diversified portfolios. Fund managers should incorporate a rebalancing strategy that increases allocation to these climate-resilient, sustainable digital assets during periods of elevated climate risk. Fund managers should integrate CPR and CTR into the quantile-domain forecasting frameworks for predicting digital asset market returns to enhance financial stability. Portfolio managers should undertake dynamic and quantile-contingent climate risk hedging strategies that account for tail-risk exposure rather than relying on average market behavior.

1. Introduction

The evolving regulatory frameworks and technological changes aimed at curbing carbon-intensive activities increase compliance and adjustment costs and lead towards heightened climate transition risk (CTR). CTR may cause an accelerative shift in consumer demand toward environmentally sustainable alternatives. Whereas, economic and financial damages resulting from acute weather shocks and long-term climatic shifts are due to the climate physical risk (CPR), which can impair infrastructure, disrupt business operations, and reduce profitability (Bartolini et al. 2025). The cryptocurrency market had expanded significantly, with more than 22,000 digital assets traded across over 500 exchanges by the last quarter of 2022. The total cryptocurrency market capitalization has approached $3 trillion by the last quarter of 2021, with Bitcoin accounting for more than half of this value (Gambarelli et al. 2023). The climate-related risks elevate volatility in financial markets (Enriquez-Perales et al. 2026), thereby prompting panic trading and momentum-based reallocations. Since cryptocurrency markets are dominated by retail investors and exhibit lower informational efficiency (Aslam et al. 2023; Zargar and Kumar 2019), the climate-related transition and physical shocks can propagate rapidly, intensifying volatility spillovers across sustainable, gold-backed, Sharia-compliant and energy-related cryptocurrency classes. Another justification for the higher shock transmission from climate-related transition and physical risk (CPR and CTR) towards the cryptocurrencies is based upon the fact that elevated uncertainty increases investors’ risk aversion, leading to delayed investment decisions (Bloom 2009) and abrupt asset reallocation. Therefore, climate uncertainty arising from unpredictable climate policies, regulatory transitions, and extreme physical events amplifies macro-level ambiguity, which transmits shocks to speculative and high-volatility assets such as cryptocurrencies. For instance, Ding et al. (2025) stated that cryptocurrencies have impacted investment decisions and risk management because of their evaluation from speculative instruments into integral elements of the global financial system. Moreover, climate-related disruptions significantly affect cryptocurrency markets, which are also central to portfolio construction and risk control.

Cryptocurrencies constitute environmentally unsustainable financial instruments, and there may be the possibility that signals of environmental deterioration will prompt sustainability-oriented investors to reduce their exposure to non-sustainable cryptocurrencies (Clark et al. 2023). Furthermore, the shock transmission mechanism between climate-related policy uncertainty, physical and transition risk and cryptocurrency returns is based upon the fact that cryptocurrencies’ energy-intensive mining and transaction processes remain heavily dependent on fossil fuels. Therefore, an increase in climatic disruptions may led towards prolong reliance on carbon-intensive energy resources and may pose additional regulatory risk (Sarker et al. 2023). Moreover, growing environmental concerns and negative perceptions of Bitcoin’s carbon footprint can weaken the demand for conventional non-sustainable cryptocurrencies. Traditional cryptocurrencies tend to increase with regional emissions but decline with global emissions, reflecting environmental concerns, while energy-focused cryptocurrencies are also affected by environmental factors but in the opposite pattern (Alshammari et al. 2025). On the contrary, cryptocurrencies rely on blockchain technology, but their creation through energy-intensive mining raises environmental concerns (Clark et al. 2023). Sarker et al. (2023) suggested that climate uncertainty increases regulatory risk and reinforces dependence on fossil fuel-intensive energy sources, whereas growing environmental concerns may reduce demand and shift investors toward greener alternatives (sustainable cryptocurrencies). This may have reduced Bitcoin’s market value and market share, whereas Clark et al. (2023) and Gutsche and Ziegler (2019) argued that an increase in environmental awareness leads sustainable investors to reward green firms and may penalize conventional ones. El Ouadghiri et al. (2021) also debated that rising public environmental attention also shifts traditional investors toward sustainable assets, negatively affecting conventional stock returns while improving sustainable stock performance. Therefore, the strong interconnection between climate policies and cryptocurrency markets is due to the cryptocurrency’s sensitivity to climate policy shifts due to energy-intensive operations (see Wu and Ding 2023). This may have prompted investors to favor greener alternatives and potentially reducing conventional cryptocurrencies’ market share. This motivates us to pose an important research question whether the shocks within the climate related physical and transitional disruptions (CPR and CTR) causes higher shock transmission towards the gold-backed, Sharia-compliant and energy-related cryptocurrencies, as compared with the sustainable cryptocurrencies, across different quantiles.

Therefore, as a result of stricter environmental policies, i.e., carbon taxation and heightened energy regulations along with technological advancements, CTR and CPR cause reorientation of investor preferences toward low-emission systems. For instance, Thampanya and Wu (2026) stated that climate risk significantly influences short-term stock returns, particularly when ESG priorities are emphasized and behave like conventional risk measures. This furthermore motivates us to incorporate the climate-related risk in the decision-making models because of heightened investors’ reaction to climate-related disruptions. The influence of climate-related disruptions is dependent upon the assets’ energy consumption intensity, governance structure, and the nature of their underlying assets. For example, energy-linked cryptocurrencies are highly susceptible to climate-related disruptions because the valuation model of specific cryptocurrencies’ return dynamics is closely interrelated with developments in energy markets (Jiang et al. 2025). Moreover, climate policy disruptions played a contributory role in triggering energy market fear (Xiao and Liu 2023), thereby affecting the profitability of energy-related ventures. This has also challenged the sustainability of associated ecosystems, thereby intensifying the transmission of shocks towards the energy-related cryptocurrencies. Climate-related disruptions also played a contributory role in affecting the extraction and compliance costs in mining activities (Qarahasanlou et al. 2024) and the financial performance of mining industries (Sun et al. 2020), thereby influencing gold supply conditions and the valuation of gold-backed cryptocurrencies. For instance, Zhu et al. (2023) also highlighted the existence of heightened physical risk of climatic disruptions on the gold price volatility forecast and observed an inverse association between gold price volatility and climate physical risk. Sharia-compliant cryptocurrencies may also tend to exhibit a moderate degree of sensitivity but in an indirect manner because climatic risk may influence the broader economic and financial sectors (Dong and Yoon 2023) underpinning their asset-backed structures. However, the adherence of Islamic financial asset classes (e.g., Islamic cryptocurrencies) to ethical investment criteria (Sheikh et al. 2025; Tabash et al. 2025b) may reduce exposure to sectors most susceptible to climate-related disruptions. However, sustainable green cryptocurrencies are least susceptible to climatic disruptions because of the employment of the energy-efficient consensus mechanisms, such as proof-of-stake, and are aligned with ecological objectives. The increase in CPR and CTR may increase the appeal of sustainable cryptocurrencies by redirecting investor interest toward assets that are consistent with environmental sustainability standards.

The climate-related uncertainties effectively predict cryptocurrency volatility in a symmetric manner, which is important for portfolio construction and risk management (Ding et al. 2025). Prior studies have not explored the time-varying as well as static asymmetric shock transmission mechanism from CPR and CTR towards the returns of different cryptocurrency classes (energy-related, gold-backed, Sharia-compliant and sustainable) across bearish $(\tau =0.05)$, bullish $(\tau =0.95)$ and moderate $(\tau =0.50)$ quantiles. For instance, existing studies only employ symmetrical or linear econometric approaches and take into account the role of equity market fluctuations for affecting the cryptocurrency volatility (Nguyen 2022), shock transmission from uncertainty pertaining to central bank digital currencies towards the cryptocurrency returns (Akin et al. 2023), role of geopolitical risk on the time-varying shock transmission between cryptocurrency and equity returns (Xu and Kinkyo 2023) and shocks transmitting from economic as well as climatic uncertainty towards the Bitcoin (Li et al. 2024). Whereas, Clark et al. (2023) explored whether environmental factors, such as temperature anomalies, play a contributory factor in affecting the cryptocurrency volatility and found that environmental factors yield an insignificant effect for explaining the fluctuations in stable coins (prominent cryptocurrencies) by employing the linear OLS regression. Therefore, this is the first research article to explore the quantile-domain shock transmission from climate transition risk (CTR) and climate physical risk (CPR) towards the sustainable, energy-related, gold-backed and Sharia-compliant cryptocurrencies during the bearish, bullish and moderate market conditions. Furthermore, prior studies only highlighted that cryptocurrencies are now integral to the global financial system and are only symmetrically or linearly affected by climate uncertainty shocks (Bouri et al. 2022; Jin and Yu 2023; Alshammari et al. 2025). Therefore, prior studies emphasize the mean-based symmetrical econometric approaches such as TVP-VAR (Li et al. 2024), OLS regressions (Clark et al. 2023) and Mixed Data Sampling-based GARCH model (Jin and Yu 2023) in order to explore the linear shock transmission mechanism from climate-related policy uncertainties and environmental factors towards a particular class of traditional cryptocurrency’s price dynamics and Bitcoin’s energy consumption (Zribi et al. 2023). Moreover, prior studies only show that major cryptocurrencies respond homogeneously to both positive and negative climate uncertainty shocks (Jin and Yu 2023).

For instance, in the existing literature, Li et al. (2024) explore the symmetrical and linear shock transmission channel from varied uncertainty factors pertaining to economic, climate, oil and equity markets towards the cryptocurrency by employing the TVP-VAR approach. Findings suggested an insignificant hedging capability of digital assets against climate risks, whereas Clark et al. (2023) only utilized the OLS-based regression framework and observed the symmetrical association between cryptocurrencies and temperature anomaly. Findings suggested the insignificant effect of temperature anomaly on Bitcoin. Ding et al. (2025) also explored the symmetrical shock transmission mechanism from climate uncertainty towards only particular cryptocurrencies, i.e., Ethereum, Bitcoin and Ripple. Similarly, Jin and Yu (2023) also observed whether the fluctuations in the climate uncertainty lead towards higher volatility in Bitcoin, Litecoin and Ripple by employing only the MIDAS-GARCH approach, ignoring the potential asymmetries due to the different shock transmission from climate risk towards the different classes of cryptocurrencies, across varied quantiles. In a similar manner, Alshammari et al. (2025) only compared the responses of conventional cryptocurrencies with those of energy-related digital assets against the regional and global environmental factors by employing the OLS-based regression approach.

The objectives of this research article are twofold. The first objective is to explore the quantile-domain asymmetric shock transmission mechanism from climate transition risk (CTR) and climate physical risk (CPR) to the return series of different cryptocurrencies across bearish, bullish, and median quantiles by employing the quantile-based vector auto-regression (QVAR) approach of Ando et al. (2022) and Chatziantoniou et al. (2021). These cryptocurrency classes include gold-backed cryptocurrencies (Tether Gold (XAUt) and PAX Gold (PAXG)), Sharia-compliant cryptocurrencies (X8X and Hello Gold (HGT)), energy-related cryptocurrencies (Power Ledger (POWR) and Energy Web Token (EWT)), and sustainable (greenest) cryptocurrencies (Cardano (ADA), Stellar (XLM), and Ripple (XRP)).1 The second objective is to examine quantile-domain time-varying shock spillovers between these cryptocurrency classes, CTR, and CPR during prominent physical and transitional climate-disruptive events. These include episodes of severe heat and prolonged drought during 2020 and 2021, drought-related damages amounting to approximately EUR 9 billion during 2022–2023, the rare triple-dip La Niña event, and historic rainfall in California during the last quarter of 2022 and the first quarter of 2023. Furthermore, quantile-domain shock spillovers between cryptocurrency returns and climate physical and transition risks are also examined during the torrential rains and severe flooding in eastern Spain in the last quarter of 2024, extreme heatwaves that caused more than 16,500 heat-related deaths in 2025, and Storm Boris in the third quarter of 2024. Therefore, from the perspective of portfolio optimizers, the study’s findings inform portfolio allocation and hedging decisions in an environment where time-varying climate physical risks such as droughts, heatwaves, floods, hurricanes and storms are increasingly priced into financial markets. This is executed through the identification of whether the relative resilience of sustainable cryptocurrencies outperforms the energy-related, gold-backed, and Sharia-compliant cryptocurrencies to climate-related shocks, across different quantiles (lower, higher and median).

This study extends the current scholarship on the interplay between climate risk and cryptocurrency price dynamics in the following ways.

First, this research extends the existing literature on environmental financial risk management through the exploration of the quantile-domain asymmetric and nonlinear dynamic as well as static linkages between various cryptocurrency categories and climate-related transition and physical risks under multiple behavioral market conditions, namely bearish, bullish, and median quantiles. Prior studies overlook the leptokurtic nature of time-series observations on conventional cryptocurrencies and financial markets (Elsayed et al. 2022), as the violation of independence and identical distribution assumptions in leptokurtic financial time-series observations may result in biased and inconsistent estimations (Sheikh et al. 2024; Tabash et al. 2024) and are not able to explore the hidden co-integration between variables (Suleman et al. 2022). Therefore, this research article contributes by evaluating how different sustainable, gold-backed, Sharia-compliant, and energy-linked cryptocurrency classes absorb climate, physical and transition risk shocks across varying market conditions. This approach facilitates a deeper understanding of asymmetric transmission mechanisms across lower, median, and upper quantiles. The findings underscore the nonlinear and regime-dependent characteristics of climate risk spillovers, highlighting the critical role of tail risks and extreme market conditions in capturing heterogeneous exposure to climate-related vulnerabilities across diverse cryptocurrency segments. This theoretically aligns with financial market microstructure and behavioral finance theories, emphasizing that shocks may affect investor behavior differently in extreme market conditions (Tabash et al. 2025a). Therefore, exploring the quantile-domain shock transmission from climate transition and physical risk towards different classes of cryptocurrencies broadens the scope of climate–finance literature by including digital assets with heterogeneous energy, safe-haven (gold-backed) Sharia-compliant and environmental footprints. However, prior studies only examine the linear shock transmission mechanism across cryptocurrencies and energy-related commodities, geopolitical risk, economic uncertainty, oil uncertainty and equity market volatility, by taking into account the mean-based econometric procedures like a generalized VAR-based framework.

Second, this research article also contributes to the existing literature by broadening the conceptual scope of climate-related risk analysis through exploring the shock transmission from both the daily climate transition risk and climate physical risk towards green, gold, energy and Islamic cryptocurrencies, rather than focusing solely on shock transmission from monthly climate policy uncertainty towards financial and non-financial asset classes (Chen et al. 2023; Shahbaz et al. 2024). Hence, the quantile-domain time-varying shock transmission from CPR and CTR contributes by moving beyond the narrow focus on monthly CPU indices as utilized in prior studies, highlighting the importance of multidimensional daily climate related physical and transitional risks in shaping the return dynamics of asset-backed, ethical and sustainable cryptocurrencies and provides practical ramification for fund managers to incorporate higher-frequency climate-related physical and transitional disruptions for the digital asset pricing forecasting framework. Furthermore, the research challenges the traditional assumption of homogeneous market responses by emphasizing the differential sensitivity of cryptocurrency classes to distinct climate risk channels. The quantile-domain CPR and CTR shock transmission received by green, gold, energy and Islamic cryptocurrencies enable a more nuanced understanding of how cryptocurrencies’ structural characteristics, i.e., ethical compliance, energy dependence and asset backing, shape heterogeneous responses to climate-related transitional and physical risks. This enriches asset pricing and risk transmission theories in the context of digital finance. However, prior studies only take into account the conventional cryptocurrency (Bitcoin) return dynamics in order to observe the impact of higher fluctuations within the monthly climate policy and global uncertainties (Li et al. 2024). This research article also contributes to the existing literature on climate–finance by indigenizing digital assets’ technological infrastructure and blockchain design as a conduit through which climate transition and physical risks propagate into asset returns, across varied quantiles. This asymmetrical and quantile-dependent climate risk shock transmission aspect is largely absent from existing studies focusing on climate-financial risk models (Li et al. 2024; Clark et al. 2023; Ding et al. 2025; Jin and Yu 2023; Alshammari et al. 2025).

The significance of exploring the quantile-domain shock transmission mechanism from climate-related risks to different classes of cryptocurrencies’ returns lies in formulating quantile-domain risk mitigation strategies for hedging sustainable, Sharia-compliant, gold-backed, and energy-related cryptocurrencies against climate transition and physical risks across bearish, bullish, and moderate climatic stress and cryptocurrency market regimes. By exploring quantile-domain shock spillovers between different climate risks and cryptocurrency classes, this study moves beyond conventional average- or mean-based econometric procedures (see Alshammari et al. 2025; Clark et al. 2023; Ding et al. 2025; Jin and Yu 2023) to reveal how climate risks are transmitted in a non-homogeneous manner during market crashes (bearish quantile), market booms (bullish quantile) and normal market conditions (median quantile). Because Bitcoin is often characterized as digital gold (Baur et al. 2024), this research empirically tests whether that reputation holds for sustainable, gold-backed, energy-related and Sharia-compliant (Islamic) cryptocurrencies, when faced with the systemic threat of climate transition and physical risks across varied quantiles. Therefore, due to the global economic transition toward energy-efficient sources, this study provides a critical roadmap by comprehensively examining whether sustainable cryptocurrencies can withstand climate transition risks (CTR), such as regulatory shifts, and climate physical risks (CPR), such as extreme weather events, or whether energy-related, gold-backed, and Sharia-compliant cryptocurrencies remain the least susceptible to these climatic disruptions across multiple market states (bearish, bullish, and moderate). Therefore, this research article presents a comprehensive overview of how climate-related shocks are permeated into different classes of cryptocurrencies across different market states.

The rest of the article is structured as follows. Section 2 explains the literature review along with the economic rationality, whereas Section 3 and Section 4 deal with data and the econometric framework. Section 5 reports the results along with practical implications, and Section 6 concludes the overall article.

2. Literature Review

The literature review section is divided into three parts. The first section explains the association between climate risk and cryptocurrencies, whereas the second section explains the relationship between climate-related risks and financial markets, as well as hedging strategies between crypto asset classes. The third section explains the existing studies using the quantile-based asymmetrical connectedness approaches.

2.1. Climate Risk and Cryptocurrencies

In the existing literature, Haq et al. (2025) only observed the pairwise connectedness between different cryptocurrency classes and public climate change news by employing the wavelet coherence econometric procedure for the frequency domain connectedness. Overall findings suggested that sustainable cryptocurrencies may play a contributory role in optimizing portfolio performance as they remain resilient to climate change concerns. However, non-sustainable cryptocurrencies’ returns are inversely associated with climate concerns, and this may imply higher traders’ trust in green digital asset classes. Ding et al. (2025) also investigated whether fluctuations in climate-related uncertainty indices can predict variances within the three prominent cryptocurrencies, namely Bitcoin, Ethereum, and Ripple. Overall findings suggested a substantial role of changes in climate-related policy indices in predicting cryptocurrency variance. Alshammari et al. (2025) observed the association between regional and global environmental factors and sustainable cryptocurrencies by employing the symmetrical OLS regression approach. Overall findings suggested that regional environmental concerns, such as carbon emissions, play a contributory role in positively (negatively) affecting traditional (energy-related) cryptocurrencies. In the existing literature, Li et al. (2024) explored the time-varying symmetrical association between Bitcoin and various uncertainty indices, such as climate uncertainty, economic policy uncertainty, oil price uncertainty, equity market uncertainty, and geopolitical risk, by employing symmetrical Granger causality and the TVP-VAR approach. Overall findings suggested that investment in Bitcoin can hedge geopolitical, oil, and economic uncertainty indices, but is not able to attain higher risk mitigation against climate-related risk.

Patel et al. (2024a) employed the wavelet quantile correlational approach and investigated the time-varying correlational characteristics between the cryptocurrency environmental attention index, sustainable cryptocurrencies and financial asset classes. Overall findings observed the lower time-varying correlation between the cryptocurrency environmental attention index and green cryptocurrencies as well as financial asset classes. Zribi et al. (2023) observed the predictability of climate-related disasters and climate uncertainty events for the ecological footprints and environmental performance of conventional cryptocurrency (Bitcoin) by employing the Bayesian TVP-SVAR approach. Overall findings suggested the pertinent role of conventional cryptocurrencies’ environmental attention in reducing the conventional cryptocurrency’s carbon footprint. Clark et al. (2023) explored whether fluctuations in environmental factors, such as climate physical risk (temperature anomalies), can effectively predict cryptocurrency volatility by employing symmetrical predictive OLS-based regressions. Overall findings suggested an effective role of carbon emissions in predicting Bitcoin volatility, but an insignificant association was reported between Bitcoin volatility and environmental factors. Sarker et al. (2023) explored whether appreciation and depreciation in monthly climate-related uncertainty yield a similar effect on Bitcoin returns by employing the NARDL approach proposed by Shin et al. (2014). Overall findings suggested only a short-term positive effect, whereby an increase in climate-related uncertainty significantly and positively impacts Bitcoin prices. Jin and Yu (2023) explored whether extreme climate uncertainty events cause higher volatility in cryptocurrencies such as Bitcoin, Litecoin, and Ripple by employing the GARCH-MIDAS framework. Overall findings suggested a substantial role of extreme climate-related policy uncertainties in positively driving volatility indices within these cryptocurrencies.

The existing studies on the shock transmission between climate change and cryptocurrencies have mainly relied upon the symmetric, mean-based connectedness frameworks (Ding et al. 2025; Alshammari et al. 2025; Li et al. 2024; Clark et al. 2023), which implicitly assume uniform shock transmission across market conditions. These mean-based econometric approaches are unable to capture the non-homogeneous and nonlinear nature of climate–cryptocurrency connectedness dynamics (see Kayani et al. 2024). Furthermore, prior studies on the environmental effect on the cryptocurrency price dynamics ignored the role of quantile-based spillovers, therefore unable to explore whether climate transition and physical risks may exert differential impacts across bullish, median and bearish market conditions. Furthermore, prior studies also explored the symmetrical frequency-domain connectedness between public climate-related news, climate policy uncertainty and cryptocurrency markets, emphasizing time-scale variations (Haq et al. 2025). However, these studies ignored the direct quantile-domain asymmetrical shock transmission from high-frequency (daily) climate risk indicators. Another notable gap is the narrow focus on conventional cryptocurrencies, particularly Bitcoin (Zribi et al. 2023; Sarker et al. 2023), without accounting for the structural diversity across alternative cryptocurrency classes such as sustainable, gold-backed, and energy-related assets. Building upon these limitations, our research article contributes by exploring the time-varying and quantile-contingent asymmetrical connectedness framework by capturing shock spillovers from daily climate transition and physical risks to the green, energy, gold and Islamic cryptocurrencies. Our research article contributes by uncovering not only the nonlinear and quantile-dependent climate risk spillovers but also provides a more comprehensive and granular understanding of heterogeneous risk exposure across different types of digital assets and advances the empirical discourse in the Fintech literature.

2.2. Climate Risk, Financial Markets and Hedging Behavior Across Different Cryptocurrency Classes

Gonzalez et al. (2026) investigated the role of gold-backed, sustainable and conventional cryptocurrencies against the inflationary pressure and interest rate hikes. Overall findings suggested that gold-backed cryptocurrency formulated an inverse significant association with the interest rate and portfolios can be optimized by incorporating both the gold-backed and sustainable cryptocurrencies during an increase in interest rates across different time scales. Furthermore, findings also highlighted the safe-haven role of conventional cryptocurrency against an increase in inflationary pressure and in the wake of interest rate hikes during geopolitical risk. Hoque et al. (2024) investigated the role of gold-backed cryptocurrencies against the U.S. financial stress by employing the dynamic conditional correlation and quantile coherency approach. Overall findings suggested that gold-backed cryptocurrencies (PAXG and Tether) have proven to be effective safe-havens against U.S. financial stress and categorical financial stress of lower equity market valuation. Yousaf et al. (2024) investigated the hedging capacity of Sharia-compliant Islamic cryptocurrencies against the metal commodities during the COVID-19 pandemic, by employing the DCC-GARCH-t copula approach and TVP-VAR framework. Overall findings suggested that incorporating Islamic gold-backed cryptocurrencies into the metal portfolio can provide maximum risk reduction during the uncertainty period, i.e., COVID-19. Patel et al. (2024b) evaluated the quantile-domain shock transmission between healthcare cryptocurrencies during the uncertainty periods like COVID-19 and heightened geopolitical risk by employing the QVAR approach. Overall findings suggested that shock spillovers are increased during these uncertainty periods. Furthermore, findings also suggested that minimum variance portfolios outperform the minimum connectedness portfolio and incorporating the health care cryptocurrencies within the portfolio may cause improved portfolio diversification. As a contrast to these studies, Ankam and Pattanayak (2025) investigated the hedging capability of conventional cryptocurrency indices against the commodity and equity market implied volatility indices as well as economic and geopolitical uncertainty indices by employing the wavelet and quantile correlation methods. Overall findings reported the non-homogeneous hedging behavior of conventional cryptocurrencies against different uncertainty periods.

Yousaf et al. (2023b) also investigated the symmetrical and linear shock transmission from conventional cryptocurrency (Bitcoin) and energy-related cryptocurrencies in order to observe the time-varying spillovers during pertinent economic uncertainty periods like COVID-19, geopolitical risk and oil price spikes. Overall findings suggested the decline in hedging and portfolio risk diversification incorporating both the conventional and energy-related cryptocurrencies due to the heightened shock spillovers during these time frames. Mensi et al. (2025) also investigated the quantile-domain shock spillovers between commodity markets, sustainable financial asset classes and green digital assets (sustainable cryptocurrencies) by employing the QVAR approach. Overall findings suggested that sustainable cryptocurrencies are more vulnerable to extreme tail risk and commodity markets, while sustainable financial assets remain the least susceptible to such risk. Overall findings also suggested the higher shock spillovers across extreme quantiles, thereby the diversification benefits of incorporating the sustainable financial asset classes for hedging the conventional cryptocurrencies and commodity markets may be compromised at extreme higher quantiles. Guo and Zhong (2025) highlighted the shock spillovers between financial asset classes, green and non-green cryptocurrencies, by employing the TVP-VAR methodological approach. Furthermore, the hedging effectiveness of green and non-green cryptocurrencies is also examined through the use of the minimum connectedness and minimum variance portfolio. Overall findings suggested that portfolios with sustainable and non-green cryptocurrencies provide higher hedging effectiveness and risk mitigation against global financial assets as compared with portfolios excluding the non-green and green cryptocurrencies. Korsah et al. (2024) examined the hedging capacity of the South African stock market, conventional cryptocurrency (Bitcoin) and gold market returns by employing the cross quantilogram and TVP-VAR methodologies. Overall findings suggested that precious metal and conventional cryptocurrency exhibited the hedging effectiveness of similar intensity, whereas the Johannesburg equity market exhibited the least susceptibility to shocks and remain resilience to shock spillovers from other variables. Lamine (2025) investigated the non-homogeneous quantile domains shock spillovers between gold-backed cryptocurrencies, conventional cryptocurrencies, AI firms and G7 financial markets using economic uncertainty periods like geopolitical risk and COVID-19. Overall findings suggested time-varying fluctuations in the shock spillovers during uncertainty periods and extreme quantiles. Moreover, AI stocks and conventional cryptocurrencies have proved to be efficient hedging tools during periods of technological and economic disruptions.

This research article advances the existing literature by predominantly focusing on the quantile-domain interactions between climate risks and different cryptocurrency classes, as prior studies only focus on assessing the hedging properties of cryptocurrencies against macroeconomic and non-macroeconomic fundamentals such as inflation, interest rates, movements in commodity and equity markets, as well as shock spillovers between conventional and energy-related cryptocurrencies (Yousaf et al. 2024; Ankam and Pattanayak 2025; Yousaf et al. 2023b; Mensi et al. 2025; Guo and Zhong 2025; Korsah et al. 2024; Lamine 2025). Few other studies only explore the effective portfolio risk mitigation and hedging effectiveness of different cryptocurrencies under conditions of financial stress, whereas Patel et al. (2024b) only explore the hedging effectiveness of healthcare-related cryptocurrencies against commodity forex and financial asset classes during the economic and geopolitical uncertainty period. Furthermore, existing studies only explore the time-varying shock spillovers between equities, cryptocurrencies, and commodities, particularly during periods of economic or geopolitical uncertainty, by relying upon the mean-based connectedness approach, i.e., TVP-VAR-based connectivity network approach (Yousaf et al. 2023b, 2024; Guo and Zhong 2025). Therefore, prior studies relying upon the symmetrical and mean-based connectivity framework embedded within either the GVAR or TVP-VAR approaches overlook the non-homogeneous risk transmission across bearish, bullish and moderate quantiles. As a contrast to these studies, our research article explores the time-varying asymmetrical shock spillovers from CPR and CTR towards the gold-backed, energy-related, Islamic and Sharia-compliant cryptocurrencies, across bearish, bullish and median quantiles. Therefore, this research article provides deeper insights into how extreme bearish, normal, and bullish market states differentially influence the transmission of climate-related risks, whereas Tabash et al. (2025a) only explore whether multiple quantiles of equity market returns of developed economies intervene in the transmission of macroeconomic and non-macroeconomic fundamentals. This approach not only enriches the understanding of asymmetric dependencies but also offers a more comprehensive perspective on the role of green, gold, Islamic and energy cryptocurrencies in the context of climate risk, thereby significantly extending the scope of prior literature.

2.3. Quantile Domain Asymmetrical Connectedness Approach

As compared with the mean-based generalized vector auto-regression (VAR) and TVP-VAR connectedness approaches, the implementation of quantile vector auto-regression (QVAR) can be able to explore the complicated and non-homogeneous nature of financial market data, particularly in the context of cryptocurrencies (Kyriazis et al. 2024). This is generally because of the fact that the QVAR-based connectedness approach possesses an inherent capability to capture the heterogeneous shock spillovers across different market states (bearish, bullish and moderate) (Kyriazis et al. 2024). Financial time-series observations may exhibit non-independence and non-identical distribution, thereby possessing the non-linearity and asymmetrical tendency (Sheikh et al. 2024). Therefore, utilization of symmetrical econometric procedures like VAR for the interrelationship between financial variables may lead to spurious findings and estimates and is not able to untangle the hidden co-integration between variables. Financial markets often exhibit fat tails and asymmetries in return distributions, which standard linear models may fail to address adequately (Suleman et al. 2022; Benedetti et al. 2026). The QVAR approach, with its capacity to model conditional distributions at various quantiles, can effectively account for such characteristics, enhancing the robustness and reliability of our findings (Kyriazis et al. 2024). As opposed to traditional mean-based connectedness approaches, the QVAR model can be able to explore the quantile-contingent connectivity between different financial asset classes across different quantiles (bearish, bullish and moderate) of the conditional distribution, offering a more granular analysis (Tabash et al. 2024). This is especially relevant for the cryptocurrency markets, as the crypto market behavioral states may differ extensively, encompassing episodes of extreme market conditions such as booms and crashes and median volatility (Kyriazis et al. 2024). As opposed to VAR and TVP-VAR-based connectivity approaches, the quantile-domain econometric estimation techniques are particularly adept at analyzing different behavioral states of financial asset classes due to the fluctuations within the climatic disruptive events (Khalfaoui et al. 2022). For instance, Bitcoin’s response to changes within the exogenous variables may be different in periods of market stability (median quantile) compared to periods of extreme volatility (bearish and bullish quantiles) (Kyriazis et al. 2024). Furthermore, the QVAR model is flexible in handling the asymmetrical and non-homoscedastic properties inherent in financial data, making it a superior choice for our analysis (see Kayani et al. 2024).

In the existing literature, Kyriazis et al. (2024) investigated the quantile-domain interrelationship between conventional cryptocurrency (Bitcoin), energy commodities, as well as industrial and precious metals by employing the QVAR-based connectivity approach. The QVAR approach plays a critical role in exploring the shock spillovers between variables by incorporating dynamic behavioral differentials across bearish, bullish and moderate quantiles, presenting evidence of dynamic shifts in traditional interlinkages amongst the selected financial assets (see Kyriazis et al. 2024). Khalfaoui et al. (2022) examine the quantile-based shock spillovers between green commodities, digital asset classes and financial markets by employing the quantile-based vector auto-regression (QVAR) connectivity approach. The QVAR approach aids in examining the dynamics of information transmission across the lower, median, and upper segments of the quantile distribution, thereby capturing periods of market distress, equilibrium, and expansion, respectively. Benedetti et al. (2026) explored the asymmetric shock transmission between conventional cryptocurrencies (e.g., Ethereum) and diversified stablecoins using the QVAR framework, which enables the estimation of connectedness across the entire distribution, including tails. This is crucial for cryptocurrency markets, which exhibit asymmetric behavior and extreme volatility. By capturing heterogeneous shock dynamics across quantiles, the approach offers deeper insights into connectedness under bearish, neutral, and bullish conditions, and enhances understanding of systemic risk during crises and speculative booms. Nasir et al. (2026) investigated quantile-based connectedness among commodity markets, cryptocurrencies, and financial assets using a quantile VAR (QVAR) model. This approach identifies spillover intensity, risk sources, and transmission channels. Its sensitivity reveals net transmitters and receivers of shocks across quantiles, enabling detection of spillover origins and supporting tailored risk mitigation strategies to limit contagion.

3. Data and Descriptive Statistics

3.1. Data

In order to explore the shock transmission mechanism from daily climate-related transition risk (CTR) and climate physical risk (CPR) to multiple classes of cryptocurrencies across quantiles, we take into account the daily climate-related risk indices developed by Bua et al. (2024). The daily time-series data for CTR and CPR are incorporated for the period from 1 April 2020 to 1 July 2025. These daily climate-related risk indices are obtained from the Policy Uncertainty data repository (https://www.policyuncertainty.com/Climate_Risk_Indexes.html (accessed on 1 August 2025)). The climate-related transition and physical risk indices (CTR and CPR) are constructed using text-based methods and authoritative sources, where the vocabularies provide relevance-ranked phrases for physical and transition risks. The PRI and TRI capture daily innovations in climate risk based on Reuters news. Consequently, spikes in the CTR and CPR indices reflect unexpected discussions related to physical hazards, adaptation and mitigation policies, and net-zero carbon emission targets. In the existing literature, Banerjee et al. (2024) utilized climate-related news-based uncertainty and climatic risk to examine shock spillovers between clean and brown energy ETFs. Chang et al. (2025) also employed the climate transition and physical risk indices developed by Bua et al. (2024) to examine whether heightened tensions between China and the U.S. led to higher climatic risks (transition and physical). Zhang (2022) also used news-based climate transition and physical risk indices and identified a negative response of financial market returns to climatic disruptions.

In order to explore the quantile-domain interactions between climate transition risk (CTR), climate physical risk (CPR), and sustainable, gold-backed, Sharia-compliant, and energy-related cryptocurrency returns, daily data on sustainable green cryptocurrencies such as Cardano (ADA), Stellar (XLM), and Ripple (XRP) are obtained from the data repositories of CoinMarketCap (https://coinmarketcap.com/about/ (accessed on 1 August 2025)) and www.investing.com (accessed on 1 August 2025). The daily natural logarithmic returns of these sustainable (green), gold-backed, Sharia-compliant, and energy-related cryptocurrencies are calculated as $r_t=ln\left(p_t\right)- ln \left(p_{t-1}\right)$, for the time series spanning from 1 April 2020 to 1 July 2025. Furthermore, in the existing literature, Patel et al. (2024a) observed shock spillovers between sustainable cryptocurrencies such as ADA, XLM, and XRP and the cryptocurrency environmental attention index and documented a decoupling between these cryptocurrencies and crypto-based environmental attention indices. These sustainable cryptocurrencies are incorporated into the QVAR framework because they require significantly lower energy inputs due to their reliance on low-power validation mechanisms, such as Proof-of-Stake, along with protocol frameworks used by networks like Ripple and Stellar (Patel et al. 2024a). Whereas Ha (2025) also explored time-varying shock transmission from global geopolitical disruptions to sustainable cryptocurrencies such as ADA, XLM, and XRP, and observed the effects of geopolitical events, including the Russia–Ukraine war and COVID-19, on shock spillovers. Furthermore, sustainable cryptocurrencies are aligned with environmental objectives by promoting environmentally friendly practices and fostering ethically grounded technological progress within the digital finance ecosystem (Ha 2025).

The rationale for incorporating specific energy-related cryptocurrencies, such as Power Ledger (POWR) and Energy Web Token (EWT), in the analysis is consistent with existing studies (Haq et al. 2025; Mbarek 2025). Haq et al. (2025) explored time-varying symmetrical shock propagation from climate concern to the energy-related cryptocurrency Power Ledger (POWR). Whereas, Mbarek (2025) examined the quantile dynamics between the electric vehicle sectoral stock index and energy-related cryptocurrencies such as Power Ledger (POWR) and Energy Web Token (EWT).

Moreover, the decision to include specific Sharia-compliant cryptocurrencies such as X8X and Hello Gold (HGT), as well as gold-backed cryptocurrencies, including Tether Gold (XAUt) and PAX Gold (PAXG), is aligned with existing studies on financial stress and gold-backed cryptocurrencies (Hoque et al. 2024), Sharia-compliant cryptocurrencies and GCC equities (Ali et al. 2024), and the hedging effectiveness of Sharia-compliant gold-backed cryptocurrencies during COVID-19 (Wasiuzzaman et al. 2023). For instance, Wasiuzzaman et al. (2023) characterized X8X and Hello Gold as Sharia-compliant cryptocurrencies adhering to Islamic principles due to their gold-backed features and documented their hedging effectiveness against financial turmoil. Whereas, Hoque et al. (2024) also observed the hedging effectiveness of a Sharia-compliant cryptocurrency such as X8X against financial stress.

For the purpose of synchronizing daily time-series observations on climate-related transition risk and physical risk with the green (sustainable), gold-backed, energy-related and Islamic cryptocurrency classes, we utilize the VLOOKUP function within Microsoft Excel. The utilization of Microsoft Excel’s VLOOKUP can ensure the temporal alignment of climate-related risks with the time-series observations on different cryptocurrency classes by matching values based on a common date column. This may create a consistent panel where each observation on climate risk and cryptocurrency classes corresponds to the same trading day. Both the cryptocurrencies’ daily time-series observations and climate-related risks contain missing observations because of their incorporation from different sources. The daily time-series data synchronization through VLOOKUP may avoid asynchronous bias, connectedness distortion and robust spillover analysis across quantiles and also improves the integrity of connectedness estimates. The synchronization of time-series observations on climate-related risks and cryptocurrencies may ensure that both the variables are compared on identical time points, and in order to avoid time-varying interrelationships that are purely driven by mismatched timing.

3.2. Descriptive Statistics

Table 1 shows the descriptive statistics of climate transition risk (CTR), climate physical risk (CPR), and the return series of different cryptocurrency classes. These include sustainable cryptocurrencies (Ripple (XRP), Stellar (XLM), and Cardano (ADA)), energy-related cryptocurrencies (Power Ledger (POWR) and Energy Web Token (EWT)), Islamic cryptocurrencies (X8X and Hello Gold (HGT)), and gold-backed cryptocurrencies (Tether Gold (XAUt) and PAX Gold (PAXG)). Figure 1 graphically illustrates the fluctuations in the return series of different cryptocurrency classes, as well as CPR and CTR. Table 1 shows that almost all classes of cryptocurrencies exhibited positive average returns, except the Sharia-compliant cryptocurrencies, namely X8X and HGT, which recorded average negative returns of −0.00082 and −0.00043, respectively. Furthermore, these Sharia-compliant cryptocurrencies, namely X8X and HGT, also exhibited higher standard deviation values of 0.10 and 0.094, respectively, compared with the rest of the cryptocurrencies. This indicates more frequent deviations of these Islamic cryptocurrencies from their mean values and higher volatility clustering. The sustainable cryptocurrencies XRP, XLM, and ADA exhibited the highest average positive returns of 0.0013, 0.00092, and 0.001529, respectively, compared with the other cryptocurrency classes.

Table 1. Descriptive statistics of the climate transition and physical risk, as well as the return series of cryptocurrency classes.

Descriptive Statistics	Climate Risk		Gold Backed		Sharia Compliant		Energy Cryptocurrency		Sustainable Cryptocurrency
	CPR	CTR	Tether Gold	PAX Gold	X8X	Hello Gold	Energy Web Token	Power Ledger	Ripple	Stellar	Cardano
Mean	−0.00354	−0.00339	0.000353	0.000357	−0.00082	−0.00043	1.90 × 10−5	0.00063	0.001333	0.000922	0.001529
Median	−0.00569	−0.00401	0.000198	0.000208	0.000144	0	−0.00332	0.001441	0.000972	0.000755	0
Maximum	0.104382	0.100684	0.036766	0.06454	1.056856	1.298278	0.673847	0.57979	0.548118	0.553585	0.544745
Minimum	−0.06315	−0.07508	−0.0487	−0.07444	−1.25312	−0.61067	−0.60918	−0.68323	−0.54101	−0.36408	−0.30958
Std. Dev.	0.021651	0.022982	0.007779	0.008829	0.103547	0.094351	0.066258	0.06375	0.054524	0.052047	0.051953
Skewness	0.701049	0.394937	−0.29955	−0.27319	−0.64671	2.527274	0.881773	0.160999	0.797603	1.735955	0.814039
Kurtosis	4.306874	4.075063	6.70823	10.14601	39.38984	43.22332	15.70652	21.47465	22.51464	22.43185	12.50746
Jarque–Bera	293.4445	142.1504	1127.028	4102.699	105,905.8	131,271.7	13,144.69	27,270.63	30,621.35	31,123.37	7431.757
Probability	0.00000	0.00000	0.00000	0.00000	0.00000	0.00000	0.00000	0.00000	0.00000	0.00000	0.00000
Sum	−6.78195	−6.50165	0.677515	0.68452	−1.57129	−0.81378	0.036418	1.208178	2.556182	1.767734	2.931587
Sum Sq. Dev.	0.89812	1.011963	0.115933	0.149365	20.54345	17.05642	8.411451	7.786804	5.696063	5.190187	5.171445
Observations	1917	1917	1917	1917	1917	1917	1917	1917	1917	1917	1917
Unit root test (at level)
ADF	−11.48 ***	−11.39 ***	−43.73 ***	−47.31 ***	−54.77 ***	−33.04 ***	−49.50 ***	−33.77 ***	−45.41 ***	−45.71 ***	−46.52 ***
PP	−39.19 ***	−38.59 ***	−43.84 ***	−48.01 ***	−53.22	−58.66 ***	−50.01 ***	−45.71 ***	−44.15 ***	−45.21 ***	−46.61 ***

Note: This table explains the descriptive statistics of return series of different cryptocurrencies, climate transition risk (CTR) and climate physical risk (CPR). Furthermore, the asterisk sign of *** signifies the rejection of the null hypothesis of “non-stationarity” at 1% level of significance.

Figure 1. Graphical presentation of climate transition risk (CTR), climate physical risk (CPR), and return series of different cryptocurrencies.

Table 1 also shows that among the energy-related cryptocurrencies, POWR exhibited the highest positive average return of 0.00063, compared with EWT (0.000019), and gold-backed cryptocurrencies such as XAUt and PAXG, which recorded average positive returns of 0.000353 and 0.000357, respectively. Table 1 further shows that, after the Sharia-compliant cryptocurrencies, the energy-related cryptocurrencies EWT and POWR exhibited higher standard deviation values of 0.066 and 0.063, respectively, compared with gold-backed and sustainable cryptocurrencies. Table 1 also indicates that all classes of cryptocurrencies and climate transition and physical risk (CTR and CPR) exhibit higher excess kurtosis, portraying the presence of leptokurtic data distributions due to large outliers. The Sharia-compliant cryptocurrencies X8X and HGT exhibited the highest kurtosis values of 39.3 and 43.2, respectively, followed by sustainable, energy-related, and gold-backed cryptocurrencies (see Table 1).

Table 1 also shows that CPR and CTR not only exhibit leptokurtic data distributions but also higher standard deviation values of 0.021 and 0.022. This indicates that leptokurtic behavior, due to excess kurtosis, implies that the return distributions of cryptocurrency classes and climate transition and physical risk exhibit fat tails and a higher probability of extreme outcomes than implied by the normal distribution. Therefore, this reflects the frequent occurrence of tail events accompanied by large shocks and abrupt price adjustments. This justifies the use of a QVAR-based methodology to explore quantile-domain shock transmission mechanisms, as traditional mean-based VAR models, which focus on average dynamics, are inefficient and non-robust in capturing such extreme behavior. Furthermore, Table 1 also confirms the presence of non-normal distributions based on the Jarque–Bera (JB) non-normality test. This violates a key assumption underlying standard linear VAR frameworks, as the daily time-series observations of cryptocurrency returns and climate-related physical and transition risks are heavy-tailed. Therefore, parameter estimates from mean-based models may be inefficient and misleading. Based upon the unit root test statistics of augmented Dickey–Fuller (ADF) unit root test by Dickey and Fuller (1981) and Phillips–Perron (PP) test by Phillips and Perron (1988), Table 1 also shows that all variables, including the returns of all cryptocurrency classes (sustainable, energy, gold-backed, and Sharia-compliant) and climate transition and physical risk, are stationary at levels, i.e., I(0). The unit root test confirms that the data exhibit stationarity characteristics due to the rejection of the null hypothesis of “non-stationarity” at the 1% and 5% levels of significance.

Table 2 shows the estimated coefficients of the Brock–Dechert–Scheinkman (BDS) test of Brock et al. (1996), used to examine whether the underlying data-generating process in CPR, CTR, and cryptocurrency returns follows an independent and identically distributed (i.i.d.) structure. The BDS test of non-normality confirms the presence of dependence and non-identical data structures due to hidden asymmetries. Therefore, the rejection of the BDS test null hypothesis provides further evidence of nonlinear dependence and dynamic complexity in the climate risk and cryptocurrency return series. The presence of non-independent and non-identical time-series data within the variables (cryptocurrency returns and climate risk) warrants the need for asymmetrical and quantile-dependent interactions between the variables, Suleman et al. (2022). Consequently, mean-based VAR frameworks are unable to capture nonlinear and asymmetric dependence across different market conditions (Sheikh et al. 2024) due to the violation of the classical assumptions of normality, independence, and identical data distributions. The rejection of the BDS null, therefore, strengthens the motivation for employing a QVAR approach, which is specifically designed to model heterogeneous dynamics and tail-risk transmission across different points of the conditional distribution. Kayani et al. (2024) also observed the heterogeneous responses of U.S. sectoral stock returns to both the good and bad shocks in the climate uncertainty, thereby detecting the asymmetrical shock propagation capability of climate-related risk towards the financial market returns.

Table 2. The BDS test of non-linearity.

	CPR			CTR			Tether Gold			PAX Gold
Dimension	BDS Statistic	Std. Error	z-Statistic	BDS Statistic	Std. Error	z-Statistic	BDS Statistic	Std. Error	z-Statistic	BDS Statistic	Std. Error	z-Statistic
2	0.0351 ***	0.0017	20.2610	0.0399 ***	0.0017	23.9966	0.0087 ***	0.0023	3.7608	0.0121 ***	0.0022	5.4567
3	0.0560 ***	0.0028	20.3213	0.0648 ***	0.0026	24.5974	0.0133 ***	0.0037	3.6333	0.0217 ***	0.0035	6.1564
4	0.0637 ***	0.0033	19.4602	0.0760 ***	0.0031	24.2934	0.0144 ***	0.0044	3.2955	0.0277 ***	0.0042	6.6031
5	0.0638 ***	0.0034	18.7050	0.0771 ***	0.0033	23.7175	0.0141 ***	0.0046	3.0842	0.0298 ***	0.0044	6.7882
6	0.0605 ***	0.0033	18.4388	0.0730 ***	0.0031	23.3569	0.0149 ***	0.0044	3.3679	0.0301 ***	0.0042	7.1047
	X8X			Hello Gold			Power Ledger			Energy Web Token
	BDS Statistic	Std. Error	z-Statistic	BDS Statistic	Std. Error	z-Statistic	BDS Statistic	Std. Error	z-Statistic	BDS Statistic	Std. Error	z-Statistic
2	0.0294 ***	0.0028	10.6604	0.0549 ***	0.0029	18.7713	0.0240 ***	0.0022	11.0116	0.0260 ***	0.0022	11.8796
3	0.0520 ***	0.0044	11.8270	0.1053 ***	0.0047	22.5931	0.0448 ***	0.0035	12.9403	0.0459 ***	0.0035	13.2175
4	0.0685 ***	0.0053	13.0151	0.1439 ***	0.0056	25.7898	0.0560 ***	0.0041	13.5585	0.0576 ***	0.0041	13.9174
5	0.0786 ***	0.0055	14.2566	0.1700 ***	0.0058	29.0981	0.0625 ***	0.0043	14.4771	0.0633 ***	0.0043	14.6742
6	0.0832 ***	0.0053	15.5832	0.1834 ***	0.0057	32.3741	0.0632 ***	0.0042	15.1590	0.0646 ***	0.0042	15.5325
	Cardano			Stellar			Ripple
	BDS Statistic	Std. Error	z-Statistic	BDS Statistic	Std. Error	z-Statistic	BDS Statistic	Std. Error	z-Statistic
2	0.0232 ***	0.0022	10.4047	0.0252 ***	0.0022	11.2439	0.0312 ***	0.0024	12.8962
3	0.0446 ***	0.0035	12.6073	0.0500 ***	0.0036	14.0457	0.0603 ***	0.0038	15.6923
4	0.0586 ***	0.0042	13.9169	0.0662 ***	0.0042	15.6023	0.0783 ***	0.0046	17.0847
5	0.0674 ***	0.0044	15.3677	0.0757 ***	0.0044	17.1145	0.0901 ***	0.0048	18.8311
6	0.0700 ***	0.0042	16.5802	0.0800 ***	0.0043	18.7365	0.0933 ***	0.0046	20.1919

Note: This table reports the estimates of the BDS test of non-linearity developed by Brock et al. (1996). The asterisk sign of *** implies the rejection of the null hypothesis of independence and identical distribution at 1% level of significance.

4. Methodology2

For the purpose of exploring the asymmetrical and quantile-domain shock spillovers from the climate transition and physical risk towards the cryptocurrency classes’ returns, we take into account the quantile-domain vector auto-regression (QVAR) by Ando et al. (2022) and Chatziantoniou et al. (2021). For the purpose of exploring the quantile-domain shock spillovers between climate transition risk, climate physical risk and different cryptocurrency classes, we take into account the lag order of 1, based on the minimum values of the Akaike Information Criterion (AIC) and the Schwarz Criterion (SC). This selection of lag order of 1 based upon the minimum AIC and SC values is in line with (Tabash et al. 2024; Chatziantoniou et al. 2022) and incorporates an optimum equilibrium between model fit and parsimony. The selection of an appropriate lag based upon the minimum value of AIC may avoid overfitting, as taking into account a higher lag order may cause complexity, and recent past information is sufficient to explain current movements within the adjusting nature of cryptocurrency markets. The utilization of a 20-day h-step-ahead forecast horizon for the estimation of quantile-domain shock spillovers between CPR, CTR and cryptocurrency classes is consistent with the existing studies (Chatziantoniou et al. 2021, 2022). This allows the analysis to capture how shocks propagate over a practically relevant period and reflects the persistence of spillovers beyond immediate reactions.

For the purpose of extreme shock spillovers between climate-related risk and cryptocurrency classes, the appropriate quantiles such as $\tau =$0.05, 0.50, and 0.95 are selected based upon the existing studies (Chatziantoniou et al. 2022; Jain et al. 2023). The selection of bearish, bullish and moderate quantiles for the asymmetrical shock spillovers between climate risk and cryptocurrency classes allows us to move beyond average (mean-based) climate–cryptocurrency relationships and capture nonlinear and asymmetric dynamics. Furthermore, the selection of a 100-day rolling window and 20-day ahead forecasting horizons for the quantile-domain shock spillovers from CPR and CTR is consistent with the study of (Benedetti et al. 2026). Benedetti et al. (2026) only observed the quantile-domain shock spillovers between traditional cryptocurrencies and stablecoins and investigated whether the overall connectedness between these variables intensified during economic uncertainty periods.

According to Ando et al. (2022), the quantile-domain vector auto-regression estimated at ${\tau }^{th}$ conditional quantile is expressed as follow,

$$y_t= u_{\left(\tau \right)}+ {\sum }_{j=1}^p{\mathsf{\Phi }}_{j\left(\tau \right)}{y}_{t-j}+ {\mathsf{\Lambda }}_{\left(\tau \right)}{f}_t+v_{t \left(\tau \right)}$$

(1)

where τ ∈ (0, 1) is a given quantile index. Following Koenker and Xiao (2006), we assume that the optimal lag order for the conditional mean model remains valid at every conditional quantile. Assuming that $Q_{\tau }\left(y_t\right|{\mathcal{F}}_{t-1})$ = 0, where ${\mathcal{F}}_{t-1}$ denotes the information set available at time t − 1.

$$Q_{\tau }\left(y_t\right|{\mathcal{F}}_{t-1})= u_{\left(\tau \right)}+ {\sum }_{j=1}^p{\mathsf{\Phi }}_{j\left(\tau \right)}{y}_{t-j}+ {\mathsf{\Lambda }}_{\left(\tau \right)}{f}_t$$

(2)

For the purpose of illustrating the quantile regression procedure, the $ith$ equation of 1 can be written as,

$$y_{it}=B_{i\left(\tau \right)}^{\prime }{z}_t+v_{it\left(\tau \right)}$$

(3)

For each $i=\mathrm{1,2},\dots \dots m$ denote a vector of dimension (mp + f + 1) × 1 that encompasses all explanatory variables, including the constant term. The parameter vector ${\beta }_i$(τ) represents the associated coefficients estimated at the τ-th conditional quantile. The imposition of the usual assumption of correct specification of the conditional quantile model is provided as

$$E\left({\psi }_{\tau }\left(v_{it \left(\tau \right)}\right)|z_t\right)=0$$

(4)

Let the function ${\psi }_{\tau }\left(z\right)$ be defined as ${\tau -1}_{[\mathrm{z}\le 0]}$. Under this specification, it follows that ${\int }_{-\infty }^{{\beta }_i\left(\tau \right)z_t}{f}_{y_{it}|z_t}\left(t|z_t\right)dt=\tau$. Whereas, $f_{y_{it}|z_t}\left(t|z_t\right)$ denotes the conditional density function of $y_{it}$ given $z_t$. For a given quantile level τ, the corresponding one-step quantile regression coefficients can then be estimated in the following manner.

$${}_{{\beta }_i\left(\tau \right)}{}^{min}\sum _{t=1}^T{\xi }_{\tau }(y_{it}-{\beta }_{i\left(\tau \right)}^{\prime }{z}_t)$$

(5)

Let ${\xi }_{\tau }\left(z\right)$ denote the check (loss) function, which is specified as ${\xi }_{\tau }\left(z\right)$= ${z(\tau -1}_{[\mathrm{z}\le 0]}$]), following the formulation introduced by Koenker and Hallock (2001). According to Ando et al. (2022), the quantile-domain forecast error variance decomposition can be mentioned as,

$$y_t= {\sum }_{j=1}^p{\mathsf{\Phi }}_{j\left(\tau \right)}{y}_{t-j}+ {{\mathsf{\Lambda }}^{*}}_{\left(\tau \right)}{f^{*}}_t+v_{t \left(\tau \right)}$$

(6)

whereas, in the above equation, ${{\mathsf{\Lambda }}^{*}}_{\left(\tau \right)}$ equal to $(u_{\left(\tau \right),}$ ${\mathsf{\Lambda }}_{\left(\tau \right)})$, ${f_t}^{*}=(1,f_t^{\prime }{)}^{\prime }$. The Wold representation of Equation (6) can be framed as,

$$Q_{\tau }\left(y_t|{\mathcal{F}}_{t-1}\right)={\sum }_{j=0}^{\infty }{B}_{j\left(\tau \right)}{v}_{t-j\left(\tau \right)}+{\sum }_{j=0}^{\infty }{C}_{j\left(\tau \right)}{f^{*}}_{t-j}$$

(7)

In the above equation, $B_{j\left(\tau \right)}={\mathsf{\Phi }}_{1\left(\tau \right)}{B}_{j-1\left(\tau \right)}+ {\mathsf{\Phi }}_{2\left(\tau \right)}{B}_{j-2\left(\tau \right)}+\cdots \dots \dots for j=\mathrm{1,2},3\dots \dots with B_{0\left(\tau \right)}={\mathit{I}}_m$ and $B_{j\left(\tau \right)}$ = 0 for $\mathcal{l}<0, whereas, C_{j\left(\tau \right)}=B_{j\left(\tau \right)}{{\mathsf{\Lambda }}^{*}}_{\left(\tau \right)}$.

There is a presumption that the selected quantile level remains constant over the entire forecasting horizon within the framework of multivariate prediction using dynamic quantile regression models. Given this premise, the resulting vector of forecast deviations for $y_{t+h }$ conditioned on the information set available at time $t-1$ as well as the underlying common factors, can be expressed as follows.

$$u_{t+h\left(\tau \right)}={\sum }_{\mathcal{l}=0}^h{\beta }_{\mathcal{l}\left(\tau \right)}+v_{t+h-\mathcal{l}\left(\tau \right)}$$

(8)

Whereas, the overall forecast error variance is estimated as,

$${COV(u}_{t+h\left(\tau \right)})={\sum }_{\mathcal{l}=0}^h{\beta }_{\mathcal{l}\left(\tau \right)}+\mathsf{\Omega }{B}_{\mathcal{l}\left(\tau \right)}^{\prime }$$

(9)

Next, consider the variance–covariance structure of the prediction errors arising from forecasting $y_{t+h}$, conditional on the sequence of innovations affecting the $i-th$ equation, namely $v_{it\left(\tau \right),} v_{i,t+1\left(\tau \right),\dots \dots \dots }{v}_{i,t+h\left(\tau \right)}$ as

$${u^{\left(i\right)}}_{t+h\left(\tau \right)}={\sum }_{\mathcal{l}=0}^h{B}_{\mathcal{l}\left(\tau \right)}(v_{t+h-\mathcal{l}\left(\tau \right)}-E(v_{t+h-\mathcal{l}\left(\tau \right)}|v_{i,t+h-\mathcal{l}\left(\tau \right)}))$$

(10)

Taking into account the assumption that the disturbance term $v_{t\left(\tau \right) }$ follows an independent and identically distributed process with zero mean. The covariance matrix Ω is a diagonal matrix given by diag(${\omega }_{11},{\omega }_{22}\dots \dots \dots \dots {\omega }_{mm}$), we proceed accordingly.

$$E\left(v_{t+h-\mathcal{l}\left(\tau \right)}|v_{i,t+h-\mathcal{l}\left(\tau \right)}\right)=\left({\omega }_{ii}^{-1}\mathsf{\Omega }{e}_i\right)v_{i,t+h-\mathcal{l}\left(\tau \right)}= e_i{v}_{i,t+h-\mathcal{l}\left(\tau \right)}$$

(11)

Here, $e_i$ denotes an m × 1 indicator vector in which the $i-th$ component equals one while all remaining entries are zero. Moreover, it holds that ${\omega }_{ii}^{-1}\mathsf{\Omega }{e}_i=e_i$. Incorporating this relationship into Equation (10) yields the following expression.

$${u^{\left(i\right)}}_{t+h\left(\tau \right)}={\sum }_{\mathcal{l}=0}^h{B}_{\mathcal{l}\left(\tau \right)}(v_{t+h-\mathcal{l}\left(\tau \right)}-{(\omega }_{ii}^{-1}\mathsf{\Omega }{e}_i)(v_{i,t+h-\mathcal{l}\left(\tau \right)}))$$

(12)

By taking into account the unconditional expectation yield,

$$COV\left(u_{t+h\left(\tau \right)}^{\left(i\right)}\right)={\sum }_{\mathcal{l}=0}^h{\beta }_{\mathcal{l}\left(\tau \right)}\mathsf{\Omega }{B}_{\mathcal{l}\left(\tau \right)}^{\prime }- {\omega }_{ii}^{-1}{\sum }_{\mathcal{l}=0}^h{\beta }_{\mathcal{l}\left(\tau \right)}\mathsf{\Omega }{e_{i}e}_i^{\prime }\mathsf{\Omega }{B}_{\mathcal{l}\left(\tau \right)}^{\prime }$$

(13)

Accordingly, the reduction in the h-step-ahead prediction error variance of $y_t$, attributable to conditioning on anticipated shocks in the $ith$ equation, can be expressed as follows.

$${∆}_{ih\left(\tau \right)}=COV(u_{t+h\left(\tau \right)}-u_{t+h\left(\tau \right)}^{\left(i\right)})= {\omega }_{ii}^{-1}{\sum }_{\mathcal{l}=0}^h{\beta }_{\mathcal{l}\left(\tau \right)}\mathsf{\Omega }{e_{i}e}_i^{\prime }\mathsf{\Omega }{B}_{\mathcal{l}\left(\tau \right)}^{\prime }$$

(14)

By normalizing the jth diagonal component of ${∆}_{ih\left(\tau \right)}$, specifically ${e_j^{\prime }∆}_{ih\left(\tau \right)}{e}_j$, with respect to the h-step-ahead forecast error variance of the jth element of $y_t$, we derive the following expression.

$$FEVD\left(y_{jt};u_{it\left(\tau \right)},h\right)={\displaystyle \frac{{\omega }_{ii}^{-1}{\sum }_{\mathcal{l}=0}^h{e}_j^{\prime }\left({\beta }_{\mathcal{l}\left(\tau \right)}\mathsf{\Omega }{e_{i}e}_i^{\prime }\mathsf{\Omega }{B}_{\mathcal{l}\left(\tau \right)}^{\prime }\right)e_j}{{\sum }_{\mathcal{l}=0}^h{e}_j^{\prime }{\beta }_{\mathcal{l}\left(\tau \right)}\mathsf{\Omega }{B}_{\mathcal{l}\left(\tau \right)}^{\prime }{e}_j}}= {\displaystyle \frac{{\omega }_{ii}^{-1}{\sum }_{\mathcal{l}=0}^h{{(e}_j^{\prime }{\beta }_{\mathcal{l}\left(\tau \right)}\mathsf{\Omega }{e}_i)}^2}{{\sum }_{\mathcal{l}=0}^h{e}_j^{\prime }{\beta }_{\mathcal{l}\left(\tau \right)}\mathsf{\Omega }{B}_{\mathcal{l}\left(\tau \right)}^{\prime }{e}_j}}$$

(15)

For ℓ = 0,1,…,h and $i,j$= 1, ……… m, the vector $e_j$ identifies the variable being forecasted, whereas $e_i$ is characterized as the originating shock. Accordingly, the forecast error variance decomposition, $FEVD\left(y_{jt};u_{it\left(\tau \right)},h\right)$ quantifies the share of the h-step-ahead prediction error variance of the j-th component of $y_t$ that is attributable to the idiosyncratic disturbance $u_{it\left(\tau \right)}$ associated with the i-th equation.

According to Ando et al. (2022), based upon the formulation of the quantile-based forecast error variance decomposition presented in the above equation, the extension of the Diebold and Yilmaz (2012) connectedness approach from the conditional mean framework to a conditional quantile context becomes readily achievable. Accordingly, the h-step-ahead m × m spillover matrix for $y_t$, evaluated at the τ-th conditional quantile, can be expressed as follows.

$${\mathbb{A}}_{\left(\tau \right)}^{\left(h\right)}=\left[\begin{array}{ccc}{\theta }_{1\leftarrow 1,\left(\tau \right)}^{\left(h\right)}& {\theta }_{1\leftarrow 2,\left(\tau \right)}^{\left(h\right)}\dots \dots \dots & {\theta }_{1\leftarrow m,\left(\tau \right)}^{\left(h\right)}\\ {\theta }_{2\leftarrow 1,\left(\tau \right)}^{\left(h\right)}& {\theta }_{2\leftarrow 2,\left(\tau \right)}^{\left(h\right)}\dots \dots \dots & {\theta }_{2\leftarrow m,\left(\tau \right)}^{\left(h\right)}\\ .& .& .\\ .& .& .\\ .& .& .\\ {\theta }_{m\leftarrow 1,\left(\tau \right)}^{\left(h\right)}& {\theta }_{m\leftarrow 2,\left(\tau \right)}^{\left(h\right)}& {\theta }_{m\leftarrow m,\left(\tau \right)}^{\left(h\right)}\end{array}\right]$$

(16)

For the purpose of streamlining the notation, let ${\theta }_{j\leftarrow i,\left(\tau \right)}^{\left(h\right)}\equiv$ $F{EVD}_{y_{jt};u_{it\left(\tau \right)},}\left(h\right)$ takes into account idiosyncratic disturbance transmission originating from variable i to variable j. Unlike the normalization procedure adopted by Diebold and Yilmaz (2012), it is pertinent to highlight that row-wise standardization is unnecessary in this setting because the diagonal structure of the covariance matrix Ω guarantees, by construction, that ${\sum }_{i=1}^m{\theta }_{j\leftarrow i,\left(\tau \right)}^{\left(h\right)}=1,j=\mathrm{1,2},3,\dots \dots \dots ,m$ Therefore, using the matrix ${\mathbb{A}}_{\left(\tau \right)}^{\left(h\right)}$, one can subsequently construct a range of aggregate indicators that summarize the network structure at the τ-th conditional quantile. Therefore, the directional “OWN” variance shares, “TO”, “FROM” and “NET” shock spillovers across quantiles can be mentioned as

$${Own}_{i\leftarrow i \left(\tau \right)}^{\left(h\right)}={\theta }_{i\leftarrow i. \left(\tau \right)}^{\left(h\right)}$$

(17)

$${From}_{i\leftarrow ∎ \left(\tau \right)}^{\left(h\right)}={\sum }_{j=1,j\ne 1}^m{\theta }_{i\leftarrow j\left(\tau \right)}^{\left(h\right)}$$

(18)

$${TO}_{i\to ∎ \left(\tau \right)}^{\left(h\right)}={\sum }_{j=1,j\ne i}^m{\theta }_{j\leftarrow i\left(\tau \right)}^{\left(h\right)}$$

(19)

$${Net}_{i\leftarrow i \left(\tau \right)}^{\left(h\right)}={TO}_{i\to ∎ \left(\tau \right)}^{\left(h\right)}- {From}_{i\leftarrow ∎ \left(\tau \right)}^{\left(h\right)}$$

(20)

According to Ando et al. (2022), the directional “TO” shock spillovers explain the overall shocks transmitted by a variable “$i$” towards all others, across quantile, whereas directional “FROM” shock spillovers explains the overall shocks received by a variable “$i$” due to the shocks from all others, across quantile ($\tau$). The directional net spillover measure represents the difference between the shocks transmitted to other variables (“TO”) and those received from them (“FROM”). A positive value indicates that variable $i$ exerts a stronger influence on the system by sending more shocks to other variables than it absorbs, thereby identifying it as a net transmitter of shocks. The TCI, as the aggregated measure of the forecast error variances due to the spillover of shocks within the entire QVAR system, can be formulated as

$${TCI}_{\left(\tau \right)}^{\left(h\right)}=m^{-1}{\sum }_{i=1}^m{F}_{i\leftarrow ∎,\left(\tau \right)}^{\left(h\right)}$$

(21)

5. Results and Discussion with Practical Implications for Crypto Traders, Fund Managers and Speculators

Table 3 explains the quantile-domain shock transmission from climate transition risk (CTR), climate physical risk (CPR) towards the sustainable, energy, gold-backed and Sharia-compliant cryptocurrency returns. The TCI explains the overall aggregated measure of forecast error variances due to the overall spillover of shocks within the entire QVAR system between CTR, CPR and cryptocurrency returns across bearish ($\tau =0.05$), bullish $(\tau =0.95)$ and moderate $(\tau =0.50)$ quantiles. Figure 2a graphically illustrates the quantile-based time-varying shock spillovers between CTR, CPR, as well as sustainable, energy, gold-backed and Sharia-compliant cryptocurrency returns, across higher, middle and lower quantiles. Furthermore, Figure 2b shows the quantile-domain connectedness heat map, and this depicts the evolution of quantile-domain time-varying total connectedness across all variables and quantiles. Darker color (dark golden heat waves) intensities indicate stronger and more pronounced interlinkages between cryptocurrencies’ return series, climate-related physical and transition risk. However, the lighter hues (light golden mild heatwaves) reflect weaker and more loosely associated relationships among the variables, signaling a gradual attenuation of overall connectedness within the QVAR system.

Table 3. The quantile shocks spillovers from climate transition and physical risk towards the cryptocurrency classes by using the “Time” based QVAR approach.

$\mathbf{Panel} \mathbf{A}: \mathit{\tau } = 0.05$	Ripple (XRP)	Cardano (ADA)	Stellar (XLM)	Energy Web Token (EWT)	Power Ledger (POWR)	Hello Gold (HGT)	X8X	PAX Gold (PAXG)	Tether Gold (XAUt)	CTR	CPR	FROM
Ripple (XRP)	12.99	10.67	11.67	6.97	9.94	9.73	7.15	7.86	7.75	7.87	7.41	87.01
Cardano (ADA)	10.4	12.57	10.99	7.19	10.07	9.94	7.39	7.93	7.91	7.99	7.62	87.43
Stellar (XLM)	11.32	10.76	13.18	6.97	9.96	9.65	7.28	7.76	7.8	7.8	7.51	86.82
Energy Web Token (EWT)	9.68	9.75	9.95	11.01	9.48	9.3	7.87	8.35	8.43	8.18	8	88.99
Power Ledger (POWR)	9.85	10.15	10.19	7.26	13.04	9.82	7.77	8.04	8.07	8.07	7.74	86.96
Hello Gold (HGT)	9.79	10.17	10.16	7.05	10	12.81	7.77	8.15	8	8.21	7.89	87.19
X8X	9.08	9.34	9.5	7.51	9.31	9.47	12.29	8.38	8.28	8.5	8.36	87.71
PAX Gold	8.76	8.89	9.24	7.42	8.85	8.99	8.07	12.41	11.04	8.31	8.03	87.59
Tether Gold	8.87	8.94	9.11	7.46	8.75	8.76	7.98	11.18	12.45	8.44	8.05	87.55
CTR	8.79	8.78	8.95	7.37	8.76	8.66	7.97	8.33	8.28	14.13	9.99	85.87
CPR	8.49	8.83	8.95	7.42	8.8	8.67	7.82	8.32	8.37	10.37	13.97	86.03
TO	95.03	96.26	98.7	72.62	93.92	93	77.06	84.3	83.93	83.72	80.59	959.12
NET	8.02	8.83	11.88	−16.37	6.96	5.81	−10.64	−3.29	−3.61	−2.14	−5.44	TCI = 87.19%
Panel B: $\tau =0.50$
Ripple (XRP)	38.63	16.39	23.27	0.65	8.69	8.28	0.74	1.07	1.09	0.63	0.56	61.37
Cardano (ADA)	16.23	37.52	19.14	1	10.97	11.24	0.55	0.97	0.87	0.8	0.72	62.48
Stellar (XLM)	22	18.18	36.23	0.84	9.76	8.65	0.77	1.13	1.11	0.76	0.58	63.77
Energy Web Token (EWT)	9.75	11.86	11.18	44.66	8.45	7.54	1.08	1.3	1.43	1.47	1.28	55.34
Power Ledger (POWR)	10.19	12.56	11.91	1.18	49.36	9.91	0.71	1.08	1.17	1.05	0.88	50.64
Hello Gold (HGT)	9.65	13.27	10.6	0.83	9.69	51.41	0.83	0.9	0.94	0.93	0.95	48.59
X8X	7.19	8.91	7.83	1.1	7	9.6	54.05	1.13	1.19	0.92	1.09	45.95
PAX Gold	2.09	2.58	2.38	1.14	1.94	2.41	1.32	50.71	32.99	1.37	1.08	49.29
Tether Gold	1.98	1.9	2.08	1.01	1.78	1.89	1.26	34.09	52.27	0.88	0.86	47.73
CTR	2.11	2.07	2.03	1.38	1.98	2.01	1.42	1.75	1.38	71.23	12.64	28.77
CPR	1.97	2.2	1.97	1.56	2.1	1.99	1.34	1.77	1.39	14.57	69.13	30.87
TO	83.15	89.92	92.39	10.69	62.35	63.52	10.02	45.19	43.56	23.4	20.63	544.81
NET	21.79	27.44	28.61	−44.65	11.71	14.93	−35.93	−4.1	−4.18	−5.37	−10.24	TCI = 49.53%
Panel C: $\tau =0.95$
Ripple (XRP)	10.61	9.39	10.11	8.92	9.17	9.63	7.52	8.57	8.18	8.76	9.15	89.39
Cardano (ADA)	9.59	10.51	9.81	8.96	9.34	9.61	7.54	8.45	8.13	8.67	9.39	89.49
Stellar (XLM)	9.87	9.5	10.91	8.98	9.17	9.62	7.48	8.41	8.17	8.61	9.27	89.09
Energy Web Token (EWT)	9.06	9.08	9.41	10.66	9.24	9.27	7.78	8.73	8.36	8.9	9.51	89.34
Power Ledger (POWR)	9.22	9.32	9.52	9.15	10.94	9.58	7.65	8.48	8.15	8.85	9.14	89.06
Hello Gold (HGT)	9.18	9.32	9.54	8.98	9.21	10.93	7.67	8.68	8.25	8.88	9.39	89.07
X8X	8.86	8.89	9.27	9.21	9.15	9.43	9.89	8.6	8.33	9.03	9.35	90.11
PAX Gold	8.96	8.88	9.15	9.18	8.8	9.27	7.67	10.4	9.43	8.91	9.35	89.6
Tether Gold	8.9	8.87	9.13	9.07	8.82	9.3	7.71	9.66	9.99	9.07	9.48	90.01
CTR	8.84	8.76	9.09	9.3	8.79	9.04	7.79	8.47	8.35	11.36	10.21	88.64
CPR	8.83	8.92	9.14	9.31	8.98	9.15	7.8	8.69	8.34	9.55	11.28	88.72
TO	91.31	90.93	94.18	91.05	90.68	93.88	76.62	86.74	83.69	89.22	94.22	982.53
NET	1.92	1.44	5.1	1.72	1.62	4.81	−13.49	−2.86	−6.32	0.58	5.5	TCI = 89.32%

Note: This table reports the connectedness estimates based upon the quantile-based vector auto-regression (QVAR) approach by (Ando et al. 2022).

Figure 2. (a) Graphical representation of the time-varying quantile-domain shock spillovers between CTR, CPR and cryptocurrencies’ returns. (b) The heat map for the QVAR-based overall shock spillovers between all the variables (CTR, CPR and cryptocurrencies’ returns across all the quantiles).

Table 3 shows that the quantile-domain average total connectedness index (TCI) between cryptocurrency returns and climate transition and physical risks is higher at the extreme lower (τ = 0.05) and upper (τ = 0.95) quantiles compared with the median (τ = 0.50) quantile. This indicates stronger shock transmission between climate-related risks and the return series of all cryptocurrency classes during bullish and bearish market conditions than in the median market state. For instance, the TCI, which represents the aggregated share of forecast error variances attributable to spillover effects between climate risk and cryptocurrency returns within the entire QVAR system, reaches 87.19% and 89.32% at the extreme lower and upper (τ = 0.05, 0.95) quantiles, respectively, compared with 49.5% at the median quantile (τ = 0.50). Moreover, Figure 2a graphically demonstrates that overall connectedness between climate-related risks and all cryptocurrencies intensifies at the extreme lower and upper quantiles relative to the median quantile. In addition, Figure 2b visually displays darker color intensities (dark golden heat waves) at the extreme lower and upper quantiles, reflecting strong and amplified interlinkages between cryptocurrency return series and climate-related physical and transition risks. Table 3 also shows that a shock in CPR and CTR leads to higher error variances of 80.59% and 83.72% at the lower quantile, respectively, in forecasting the 20-day-ahead returns of all other cryptocurrencies, as compared with the median quantile. Whereas, a shock in CPR and CTR also causes higher contributions of shocks of 94.22% and 89.22%, respectively, toward all other cryptocurrency classes’ returns at the higher quantile, as compared with the median quantile (20.63% and 23.4%).

The higher shock transmission mechanism across extreme quantiles, i.e., bearish and bullish, is consistent with existing studies. For instance, Wang et al. (2023) also found higher shock spillover mechanisms across extremely high and low quantiles between sustainable financial asset classes, energy-related commodities, and climate uncertainty indices. Yin et al. (2025) also explored quantile-domain shock spillovers between agricultural, energy, financial, and precious metal classes and observed that overall shock spillovers between these diversified asset classes are intensified across extremely high and lower quantiles as compared with the median. Ha (2025) explored the role of exogenous factors, such as geopolitical risk, in transmitting shocks toward sustainable and non-sustainable cryptocurrencies across varied quantiles and found that the overall shock transmission mechanism is quantile-dependent and intensified at extremely high and lower quantiles as compared with the median quantile, whereas Yousaf et al. (2023a) explored the quantile-domain interactions between meme stocks and tokens, as well as precious metals, energy-related commodities, and financial asset classes. Overall findings suggested that during bearish and bullish quantiles, the time-varying shock spillovers between these variables are intensified and strengthened. Furthermore, the quantile-domain non-homogeneous shock transmission between different cryptocurrency classes and climate-related risk is inconsistent with existing studies focusing only on the symmetrical shock propagation mechanism across traditional cryptocurrencies and decentralized financial asset classes, using the generalized vector auto-regression based connectedness approach (see Mensi et al. 2024).

Table 3 shows that both climate transition risk (CTR) and climate physical risk (CPR) transmit higher contributions of shocks toward energy-related, gold-backed, and Sharia-compliant cryptocurrency classes across all quantiles as compared with sustainable cryptocurrencies. However, the transmission of shocks from CPR and CTR increases at the extreme lower and higher quantiles toward the energy, gold-backed, Islamic, and sustainable cryptocurrencies’ returns as compared with the median quantile. This shows not only the heterogeneous responses of cryptocurrencies’ returns to climate risk shock transmission across quantiles but also the resilience of sustainable cryptocurrencies to climate transition and physical risk across bearish and median quantiles. For instance, at the lower quantile (τ = 0.05), a shock in CPR (CTR) leads to higher contributions of shocks of 8%, 7.74%, 7.89%, 8.36%, 8.03%, and 8.05% (8.18%, 8.07%, 8.21%, 8.5%, 8.31%, and 8.44%) toward the cryptocurrency returns of Energy Web Token (EWT), Power Ledger (POWR), Hello Gold (HGT), X8X, PAX Gold (PAXG), and Tether Gold (XAUt), respectively, as compared with the median quantile (τ = 0.50). However, the sustainable cryptocurrencies Ripple (XRP), Cardano (ADA), and Stellar (XLM) received the lowest shock contributions of 7.87%, 7.99%, and 7.8% (7.41%, 7.62%, and 7.51%) from CPR and CTR (see Table 3). The higher resilience of sustainable (green) cryptocurrency classes to climate transition and physical risk contrasts with existing studies, such as Bouri et al. (2022), who reported that monthly climate policy uncertainty acts as a favorable determinant of sustainable financial markets as compared with non-sustainable brown financial market returns. Furthermore, Pástor et al. (2021) also stated that higher upward fluctuations in climate concern prompt investors to reallocate their portfolios away from carbon-intensive firms toward environmentally sustainable firms. This results in superior performance of green equities relative to brown equities. In the wake of higher climate-related risk and uncertainty, investors and traders are more inclined to divest from higher carbon-intensive firms toward environmentally sustainable equities (see Choi et al. 2020).

However, the higher resilience of sustainable cryptocurrencies’ returns (ADA, XLM, and XRP) is inconsistent with existing studies that find an increase in volatility dynamics of sustainable global bonds, green equity markets, and sustainable energy-related indices due to climate policy disruptions (Raza et al. 2024). Similarly, the higher resilience of sustainable cryptocurrencies against climate transition and physical risk is justified, as Dong et al. (2023) found that green (sustainable) bonds played a contributory role in hedging climate uncertainty and outperformed conventional counterparts (dirty bonds). Whereas, one plausible justification for the higher resilience of sustainable cryptocurrencies to climate transition and physical risk is that investors derive non-pecuniary benefits from holding sustainable financial investments, and because these assets provide insurance against climate-related risks (see Pástor et al. 2021). Furthermore, performance benchmarks in sustainable financial asset classes improve markedly in the wake of favorable innovations in ESG factors. This is reflected by fluctuations in consumer and investor preferences for sustainable goods and investments. More specifically, Table 3 also shows that among sustainable cryptocurrencies (XRP, XLM, and ADA), XRP and XLM received the lowest shock contributions from both CTR and CPR at the lower and median quantiles, as compared with ADA and all other cryptocurrency classes’ returns.

Table 3 also shows that at the higher quantile, a shock in climate transition risk (CTR) and climate physical risk (CPR) causes the highest shock transmission contributions toward energy-related, gold-backed, and Sharia-compliant cryptocurrencies, as compared with sustainable cryptocurrencies’ returns. Furthermore, the shock transmission mechanism at the higher quantile due to climate-related risk is also higher in intensity and magnitude as compared with the median quantile (see Figure 2a,b). For instance, at the higher quantile, a shock in CTR leads to higher error variances of 8.9%, 8.85%, 8.88%, 9.03%, 8.91%, and 9.07% in forecasting the 20-day-ahead returns for EWT, POWR, HGT, X8X, PAX, and XAUt, respectively. Whereas, a shock in CPR also causes higher contributions of shocks of 9.51%, 9.14%, 9.39%, 9.35%, 9.35%, and 9.48% toward EWT, POWR, HGT, X8X, PAX, and XAUt, respectively, at the higher quantile. However, the shocks omitted from CTR and CPR are lower in intensity at the median quantile, and most sustainable cryptocurrencies’ returns receive lower shocks from both climate-related disruptions across all quantiles. Furthermore, the sustainable cryptocurrency returns of Ripple (XRP) and Stellar (XLM) receive the lowest shocks of 8.76% and 8.61% from CTR. These sustainable cryptocurrencies (XRP and XLM) also receive the lowest shocks of 9.15% and 9.27% from CPR at the higher quantile, as compared with energy-related, gold-backed, and Sharia-compliant cryptocurrencies’ returns. Similarly, at the median quantile, all gold-backed, Sharia-compliant, and energy-related cryptocurrencies receive greater shocks from climate-related disruptions, but the sustainable cryptocurrencies’ returns of ADA, XLM, and XRP receive the lowest shock spillovers.

The asymmetrical U-shaped shock propagation across climate-related disruptions and different cryptocurrency classes is in contrast with the findings of Khalfaoui et al. (2022), as the study reported higher shock transmission between sustainable and environmentally friendly asset classes with Bitcoin at the extremely high and lower quantiles, as compared with the median quantile. However, the findings regarding quantile-dependent heterogeneous shock transmission from climate transition and physical risk toward sustainable, energy-related, gold-backed, and Islamic (Sharia-compliant) cryptocurrencies are inconsistent with the existing literature, as Li et al. (2024) only explored symmetrical shock transmission from climate-related policy risk toward conventional Bitcoin. Overall findings suggest that Bitcoin cannot serve as an efficient hedge against climate uncertainty. In a similar manner, Alshammari et al. (2025) only symmetrically observed the association between cryptocurrency markets and environmental factors, such as carbon emissions. Findings based on the symmetrical or linear OLS framework suggest that Bitcoin and Ethereum are adversely affected by global carbon emissions but show positive co-movement with regional carbon emissions as a prominent environmental factor. Whereas other studies only explore the symmetrical or linear impact of climate risk on the volatility of conventional cryptocurrencies and observe that increases in climate risk raise the volatility of major cryptocurrencies (Bitcoin and Ethereum). Furthermore, rather than exploring quantile-domain shock transmission between daily climate transition and physical risk and different cryptocurrency classes, Ding et al. (2025) only observed the responses of Bitcoin and Ripple conditional volatility to monthly climate policy fluctuations in a symmetrical fashion. Similarly, Jin and Yu (2023) only investigated symmetrical shock propagation between climate uncertainty and Bitcoin and observed the favorable role of only climate policy-related uncertainty indices in driving Bitcoin price volatility.

Our findings also suggest that at the higher quantiles, climate physical risk causes higher error variances in forecasting the 20-day-ahead returns in all cryptocurrency classes (energy-related, gold-backed, Islamic, and sustainable), as compared with transition risk. These findings contrast with existing studies, as Ouyang et al. (2025) analyze the dynamic shock transmission mechanism between non-ferrous as well as energy sectors, climate transition, and physical risk. Overall findings suggest that climate physical risk leads to higher shock propagation toward energy-related commodities and non-ferrous sectors.

5.1. Practical Implications and Discussion

The results discussed above yield multiple actionable insights that are highly relevant for portfolio managers, cryptocurrency market participants, and trading professionals, in the wake of climate-related transition and physical risks.

First, the shareholders and fund managers should optimally adopt the quantile-domain dynamic hedging strategies for cryptocurrency classes because shocks transmitted from climate transition and physical risk are more intensified in both the extremely high and low tails as compared with the median quantile. Figure A1 also shows that the asymmetrical and U-Shaped transmission of climate-related shocks, because climate-related disruptions are more intensely connected with the cryptocurrencies’ returns across bearish and bullish quantiles, as compared with the median. Therefore, traditional mean–variance portfolio models, which assume symmetric information transmission, may underestimate risk exposure to climate shocks in crypto. Fund managers should consider stress testing for portfolio optimization under extreme climate-risk scenarios in order to get a more realistic assessment of potential losses and gains. Portfolio managers should adopt quantile-centric risk management strategies, such as conditional value-at-risk at extremely high quantiles, rather than relying solely on mean-based forecasting frameworks. The shareholders, fund managers and portfolio optimizers can use derivative instruments like cryptocurrency options to hedge against tail-specific climate-disruptive shocks. Fund managers should adopt a buying out-of-the-money put option strategy during periods of higher and lower climate transition and physical risks in order to protect cryptocurrency portfolios from sharp downside movements. Figure A2 also shows that climate, physical and transition risk are mainly shock transmitters and transmit more shocks towards all other cryptocurrency classes’ returns at the bullish quantile. Figure A1 and Figure A2 show that most of the gold-backed and Sharia-compliant cryptocurrencies mainly receive greater shocks from all others at both the higher and lower quantiles. Since climate shocks affect extreme quantiles more intensely, traditional diversification benefits may break down during tail events, particularly in cryptocurrencies, which are already highly volatile. The quantile-specific impact suggests that early detection of climate-related events can preempt extreme losses or capitalize on extreme gains. Fund managers should implement time-varying real-time monitoring of climate transition signals, such as regulatory announcements and policy shifts, and physical risk events such as flooding and hurricanes, in order to get higher cryptocurrency portfolio optimization.

Second, the higher shock transmission from CTR and CPR towards the cryptocurrency returns across higher and lower quantiles means that investors may experience higher tail risk during periods of climate policy tightening or extreme weather events. Therefore, fund managers should incorporate a higher risk premium, tighter position limits, and climate-conditioned expected shortfall (ES) or tail-centric value at risk (tail-VaR) measures when allocating capital to these crypto classes. This is particularly under a heightened climate transition and physical risk events. The short-term speculators can exploit higher CTR and CPR shock transmission towards the energy-related, gold-backed and Sharia-compliant cryptocurrencies as relevant predictive trading signals rather than treated as exogenous noise. The increases in CPR and CTR are likely to generate sharp price adjustments and volatility clustering in Sharia-compliant, gold-backed and energy-related cryptocurrencies, creating profitable opportunities for event-driven and volatility-based strategies. This includes the short-term momentum trades, options straddles, or tactical short positions in the wake of climate transition and physical risks. Nevertheless, the higher CTR and CPR shock transition across extreme quantiles also causes elevated liquidation risk, implying that leverage and holding periods must be actively adjusted when climate risk indicators signal stress.

Third, the least chock reception capacity of sustainable cryptocurrencies due to the shocks in CTR and CPR has presented several important allocation consequences from the portfolio construction perspective. For instance, sustainable cryptocurrencies exhibit lower tail dependence with climate risk factors. This makes them efficient hedging instruments for climate-resilient diversification within all the cryptocurrency classes. All the sustainable cryptocurrencies received lower shock spillovers from CPR and CTR but the extreme vulnerability of the particular two sustainable cryptocurrencies, i.e., Ripple (XRP) and Stellar (XLM), suggests that their energy-efficient consensus mechanisms and limited exposure to carbon-intensive infrastructure significantly. These green cryptocurrencies should be incorporated as stabilizing core holdings, especially during periods of elevated climate risk, while reallocating away from climate-sensitive assets, i.e., gold-backed, energy-related and Sharia-compliant cryptocurrencies to preserve portfolio resilience. For institutional and ESG-oriented investors, climate risk should be explicitly integrated into digital asset investment mandates. The higher allocations to gold-backed, Sharia-compliant, and energy-linked cryptocurrencies should be consistently reduced during upside (bullish quantile) and downside (lower quantile) climate risks, while sustainable cryptocurrencies can serve as long-term strategic holdings aligned with environmental risk management objectives. Therefore, fund managers and financial analysts should incorporate quantile-domain climate, physical and transition-related risk metrics into stress testing, scenario analysis, and disclosure frameworks. This allows fund managers to better anticipate extreme losses and improve transparency for stakeholders.

5.2. Time-Varying Shock Spillovers from Climate-Related Risks (CPR, CTR) Towards the Sustainable, Energy-Related, Gold-Backed and Sharia-Compliant Cryptocurrencies Across Quantiles

Figure 3 graphically illustrates the time-varying directional “TO” shock spillovers and shows that a shock in climate transition risk (CTR) and climate physical risk (CPR) causes stronger shock transmission effects toward the returns of other cryptocurrencies at extreme quantiles, both upper and lower. This is particularly evident during the early quarters of 2020 and 2021. The shock transmission from CTR and CPR toward all other cryptocurrencies’ returns is further intensified in 2022, as pronounced upward fluctuations became evident during this period. Furthermore, an increase in the intensity of shock transmission is also observed at both extreme (higher and lower) tails of the distribution between 2022 and 2023. One justification for the higher shock transmission from CTR and CPR is due to the fact that Europe’s southern and western regions experienced recurring episodes of severe heat and prolonged drought during 2020 and 2021. This placed substantial pressure on water resources, agricultural production, and escalated heat-related mortality rates (EU Civil Protection Knowledge Network 2023). Between 2020 and 2021, drought-related damages amounted to approximately EUR 9 billion per year in the European region (Cammalleri et al. 2020). Furthermore, White et al. (2023) also stated that an unprecedented heatwave affected the U.S. Pacific Northwest and Western Canada during the second quarter of 2021. This resulted in extensive infrastructure damage and substantial economic losses. Another justification for the heightened shock transmission from climate-related risk during 2022 to 2023 is that an unusually prolonged and rare triple-dip La Niña event emerged, significantly influencing global climate patterns (National Oceanic and Atmospheric Administration 2023). Recurrent episodes of intensified atmospheric river events delivered historic rainfall to California and surrounding states, triggering severe flooding, substantial infrastructure destruction, economic losses of $5–7 billion (Guy Carpenter and LLC Company 2023), and loss of life during the last quarter of 2022 and the initial quarter of 2023. Figure 4 graphically shows the time-varying directional “FROM” spillover of shocks. This further indicates that cryptocurrencies experience higher shock transmission from all others, including CTR and CPR, during the initial quarters of 2020 and 2021, as well as during the second and third quarters of 2022 and 2023.

Figure 3. The graphical representation of the directional “${TO}_{jt}$” spillover of shocks, as the overall spillover of shocks towards all other variables due to the shock in variable $“j”$, across quantiles (bearish, bullish and moderate).

Figure 4. The graphical representation of the directional “${FROM}_{jt}$” spillover of shocks, as the overall spillover of shocks from all other variables, received by a variable $“j”$, across quantiles (bearish, bullish and moderate).

Furthermore, Figure 3 also shows that both climate-related risks, such as climate physical and transition risks, transmitted greater shocks to all other cryptocurrencies during the third quarter of 2023 and the initial quarter of 2024. Furthermore, a higher upsurge in climate-related shock transmission is also observed during the last quarter of 2024 and throughout 2025. Furthermore, Figure 4 also shows that all cryptocurrencies receive greater shocks from all others, including CTR and CPR, during the initial quarter of 2024 and throughout 2025. One justification is that torrential rains triggered severe flooding in Eastern Spain during the last quarter of 2024, causing about 90 fatalities and economic losses (Al Jazeera 2024). During the second and third quarters of 2025, the European region experienced extreme heatwaves, which caused more than 16,500 heat-related and global warming deaths (AFP 2025). The higher shock transmission from CPR and CTR during 2024 is based on the fact that there were varied incidents of extreme heatwaves and storm events, such as Storm Boris in the third quarter of 2024 (Athanase et al. 2024). This caused widespread flooding across major European regions. This resulted in losses worth billions of euros. Therefore, the economic disruption from flooding and heavy rainfall highlighted increasing climate hazard exposure.

5.3. Robustness Analysis

Figure A3 shows the time-varying connectedness between climate-related risks and different cryptocurrency classes by employing the symmetrical TVP-VAR-based connectedness approach of Antonakakis et al. (2020). Upon comparing Figure 2a with Figure A3, the utilization of the quantile-based VAR to explore time-varying spillovers across varied quantiles between climate risk and cryptocurrency classes offers a clear advantage over the TVP-VAR framework because it captures the heterogeneity of relationships across the entire conditional distribution rather than focusing solely on the mean. Figure A3 graphically shows a time-varying aggregated measure of the forecast error variances due to the overall shocks spillovers between all the variables (climate risks and cryptocurrencies), but remains restricted to average dynamics. Therefore, the TVP-VAR-based total connectivity metric ignores the asymmetric responses that typically arise during extreme market conditions (bearish and bullish cryptocurrency market behavior). However, the time-varying quantile-domain shock spillovers extracted through the QVAR-based connectivity approach (see Figure 2a) account for different quantile levels, enabling the analysis to distinguish between bearish market behavior, median (normal), and bullish market behavior regimes. Therefore, the implication of the quantile-based shock spillovers incorporates a comprehensive and nuanced understanding of total connectedness dynamics between cryptocurrency classes and climate risk for shareholders and fund managers by incorporating distributional asymmetries that the mean-based TVP-VAR methodology fails to capture. The exploration of quantile-domain shock spillover is important because cryptocurrencies may often exhibit nonlinear tendencies due to leptokurtic data distribution (Catania and Grassi 2022), and climate-related risks may affect the financial asset classes in an asymmetrical fashion (Kayani et al. 2024).

6. Conclusions with Policy Guidelines, Future Research Directions and Limitations

While prior studies explored the impact of climate-related policy uncertainties on the conventional cryptocurrency returns by using only the symmetrical (linear) forecasting frameworks, ignoring the leptokurtic, non-independent and non-identical distribution in the time-series data of financial asset classes, this is the first research article exploring the quantile-domain interactions between daily climate risk factors such as climate transition risk (CTR), climate-related physical risk (CPR) and returns of sustainable, gold-backed, Sharia-compliant and energy-related cryptocurrencies. Therefore, the objective of this research article is to examine whether the CPR and CTR transmit higher contributions of shocks towards different cryptocurrency classes at the median quantile as compared with the extreme higher (bullish) and lower (bearish) quantiles, thereby exploring the climate risk-related non-homogeneous shock transmission mechanism. For the said purpose, we utilized the novel quantile vector auto-regression (QVAR) domain connectedness methodology of Ando et al. (2022) and Chatziantoniou et al. (2021) in order to explore the asymmetrical shock transmission mechanism from CPR and CTR towards the sustainable, energy-related, gold-backed and Sharia-compliant cryptocurrencies. Furthermore, this research article also explores the time-varying shock transmission mechanism from CTR and CPR towards the multiple cryptocurrency asset classes during prominent climatic disruptive events like recurrent extreme climate events, i.e., severe heatwaves and prolonged droughts during 2020–2021, a rare triple-dip La Niña during 2022–2023, historic rainfall in California in late 2022 and early 2023, severe flooding in Eastern Spain in late 2024, and extreme heatwaves causing over 16,500 heat-related deaths in 2025.

Overall findings suggested that climate-related transition and physical risk transmitted greater shocks towards sustainable, energy-related, Sharia-compliant and gold-backed cryptocurrencies at the extremely high and lower quantiles as compared with the median quantile. This shows the U-shaped and non-homogeneous nature of quantile-dependent shock transmission from climate-related transition and physical risk. Therefore, cryptocurrency investors and portfolio optimizers should adopt quantile-based dynamic hedging strategies for cryptocurrency portfolios, as climate transition and physical risk shocks are significantly stronger in the upper and lower tails than at the median quantile. Consequently, traditional mean–variance portfolio models, which assume symmetric risk transmission, are likely to underestimate climate-related tail risk in cryptocurrencies. Portfolio optimization should therefore incorporate stress testing under extreme climate scenarios and rely on quantile-centric risk measures, such as expected shortfall at conditional value-at-risk at extreme quantiles, rather than the forecasts relying upon the mean-based econometric approaches.

Findings also suggested both the CPR and CTR transmit greater shocks towards the energy-related (Power Ledger, Energy Web Token), gold-backed (Tether Gold and PAX Gold) and Sharia-compliant (X8X and Hello Gold) cryptocurrencies at all the quantiles. However, the shock transmission from both the CTR and CPR at higher and lower quantiles is higher in intensity and magnitude as compared with the median quantile. Fund managers should therefore apply a higher risk premium, tighter position limits, and climate-conditioned tail-risk measures, such as expected shortfall or tail-VaR, particularly during heightened climate stress. Short-term speculators may exploit the pronounced CTR and CPR spillovers into energy-related, gold-backed, and Sharia-compliant cryptocurrencies as predictive trading signals. This is due to the fact that rising climate risk is associated with sharp price adjustments and volatility clustering, thereby creating opportunities for event-driven and volatility-based strategies, including momentum trades, options straddles, and tactical short positions. However, this may elevate liquidation risk, and it is also necessary to take into account active leverage and holding-period management. To mitigate tail-specific climate shocks, fund managers may employ cryptocurrency derivatives, particularly out-of-the-money put options, during periods of extreme upside fluctuations in the climate transition and physical risk. Therefore, early detection and real-time monitoring of climate policy signals and physical risk events are essential for effective risk management and portfolio optimization, as diversification benefits tend to weaken during tail events in highly volatile crypto markets.

Furthermore, findings also suggested that CTR and CPR transmitted greater shocks towards the gold-backed, Sharia-compliant and energy-related cryptocurrencies as compared with the sustainable cryptocurrencies’ returns, across all the quantiles. For instance, at extremely high (bullish) and lower (bearish) quantiles, a shock in the CPR and CTR leads towards the higher contributions of shocks in forecasting the 20-day-ahead returns in the gold-backed, Islamic and energy-related cryptocurrencies as compared with the sustainable cryptocurrencies. Amongst the sustainable cryptocurrencies, Ripple (XRP) and Stellar (XLM) received the lowest shocks from CPR and CTR across all the quantiles. The weak shock reception of sustainable cryptocurrencies indicates low tail dependence with climate risk factors, making them effective instruments for climate-resilient portfolio diversification. In particular, Ripple (XRP) and Stellar (XLM) exhibit the lowest exposure to CTR and CPR shocks, reflecting their energy-efficient consensus mechanisms and limited reliance on carbon-intensive infrastructure. The sustainable cryptocurrencies (ADA, XLM and XRP) should be treated as stabilizing core holdings during periods of elevated climate risk, while exposure to climate-sensitive cryptocurrencies (such as EWT, POWR, X8X, HGT, XAUt and PAXG) should be reduced.

This research article explores the quantile-domain shock transmission from the CTR and CPR towards the different cryptocurrency classes. Whereas, future studies should explore the shock transmission from climate-related risk towards the good and bad volatility of clean and brown energy exchange-traded funds (ETFs). The good (bad) volatility is constituted of the positive (negative) returns, as can be extracted from the daily price trends of the clean and brown energy ETFs. In the existing literature, good volatility in financial asset classes absorbs exogenous shocks differently than bad volatility, thereby signaling the asymmetry in the shock transmission (see Sheikh et al. 2025; Suleman et al. 2023; Tabash et al. 2024). Therefore, future studies should explore whether the lag and contemporaneous shock transmission mechanism from CTR and CPR towards the good volatility of clean and brown energy ETFs surpasses that of bad volatility. Therefore, researchers can explore the shock transmission from CPR and CTR during the bearish and bullish clean and brown energy ETFs’ volatility.

This research article explores the quantile-domain shock spillovers from climate-related risks towards the cryptocurrency classes by taking into account the daily time-series observations on climate-related transitional and physical risk as well as cryptocurrency classes. Because of daily time-series observations in order to explore quantile-contingent climate–cryptocurrency connectedness, this research is unable to capture intraday high-frequency market dynamics that occur within a single trading day. Therefore, this research article is not able to observe the climate risk effects on intraday cryptocurrency price adjustments, microstructure effects, and short-lived volatility spikes. Future studies should take into account the intraday data with a 5-min interval on the cryptocurrency classes and extract the positive and negative realized semi-variances of these cryptocurrency classes. Furthermore, future studies should also utilize the quantile vector auto-regression approach in order to explore the shock transmission from daily climate-related risk towards the positive and negative realized semi-variances (good and bad volatility). The shock transmission from daily CPR and CTR towards the intraday higher-frequency price fluctuations of the cryptocurrencies is important because financial time-series observations often react to new information almost instantaneously. Therefore, these reactions may dissipate before the daily closing price is recorded.