Numerous papers have examined global su-pply chain in dynamic environments, with a particular focus on manufacturers, platforms, finance-related linkages, capital exports, and trade credit (e.g., Dubey et al., 2017; Öberg, 2021; Moretto and Caniato, 2021; Lin and Zhu, 2025; Liu et al., 2025). Pankratz and Schiller (2024) demonstrate that companies change the make-up of their supply chains as suppliers encounter greater climate risks, such as extreme heat or flooding. Lin and Zhu (2025) observe that diversification of supply chains enhances productivity and raises the resilience of renewable energy companies. Liu et al. (2025) stress that supply chain and inflation shocks combined with consumption shocks could hamper green technology, and digitalization increases the degree of adaptability. Agrawal et al. (2024) assert that technologies from Industry 5.0 can help alleviate climate-induced supply chain disruptions and promote sustainability. The focus is primarily on factors causing global su-pply chain rather than the potential impact of other aspects after the outage and possible measures to address the outage. As markets globalize and the operating environment becomes more dynamic, managers pursue lean supply chain management practices (Blackhurst et al., 2005; Min et al., 2019), where supply chains are optimized into more economical and responsive industrial networks. These trends present opportunities for supply chain development (Ghadge et al., 2020), placing significant pressure on a stable operating environment and increasing the risk of vulnerability and disruption (Katsaliaki et al., 2021). Global su-pply chain indicates that a company cannot meet the supply or demand required for normal operations (Hendricks and Vinod, 2005). While revealing the sudden disruption, Wilson (2007) explains global su-pply chain from the perspective of logistics and transportation. He defines global su-pply chain as events in which disruptions to logistics cause the movement of goods to stop suddenly. The global supply chain is characterized by unplanned abruptness and sudden events that can disrupt the expected flow of materials, information, and components (Skipper and Hanna, 2009; Butt, 2021).

Extreme weather, sudden disasters, and policy factors often lead to global su-pply chain. The U.S. National Oceanic and Atmospheric Administration (NOAA) has recorded 212 disasters since 1980, causing approximately $1.2 trillion in damage (Katsaliaki et al., 2021). Natural disasters such as the 2004 Indonesian tsunami and the 2011 Great East Japan Earthquake severely affected the supply chains of multiple products for companies such as Apple, Samsung, and Toyota, with the fragile chains immediately disrupted, negatively affecting the reputations and earnings of these companies (Chongvilaivan, 2012). The C19P outbreak in 2020 severely restricted factory production around the world, cut off logistics routes, and disrupted basic supply chain operations (Araz et al., 2020). Statistics from the Federal Emergency Management Agency (FEMA, 2015) show that approximately 50% of small and medium-sized businesses find it difficult to reopen after a disaster. In human factor disruptions, the risk of production failures for just-in-time carmakers and other manufacturers with similar operations has risen following Brexit (Banker, 2019). Recently, RUC has caused an imbalance in the global economic and political order, unstable energy and resource supply, blocked or interrupted supply chain networks to varying degrees, and a decline in the health index of residents (Mariotti, 2022; Malchrzak et al., 2022; Piccoli et al., 2022).

According to Luo and Kwok, 2020, C19P takes a 40% negative hit to China’s supply chain. Benigno et al. (2022) developed a new barometer to measure different dimensions of GSCPI. It offers data on the United States. In addition, it embeds 27 different metrics, including logistics networks, transportation, container shipping costs, and Purchase Manager Index (PMI) surveys. Investigate the link between the environment and GSCPI using panel quantiles and document a strong association between the global su-pply chain and environmental degradation. Scholars have highlighted the problem of global su-pply chain in the literature. This topic increasingly challenges the stability of product supply chains and the efforts of core companies to consolidate their supply and demand relationships. Supply chains cross each other into networks, and chains are interdependent. Disruptions can snowball, with severe consequences for all relevant supply chain echelons. Although the Global Supply Chain Pressure Index (GSCPI) is frequently used in the literature, in this study, we rely instead on a China-specific measure (GSCH). This choice reflects the focus on China’s domestic disruptions and their international spillovers.

In the current era of deepening globalization, informatization, and ecological processes, as well as major changes in geopolitics and international environmental politics, it is urgent to re-examine the importance of climate technology transfer in the reshaping of global su-pply chain and environmental governance, as well as its face new challenges and opportunities, and actively explore innovations in technology transfer models (Petricevic and Teece, 2019; Collins et al., 2021; Anderson, 2022). CTCH is a specific technological innovation designed to reduce the impact of product production on the natural environment, covering processes, products, services, and business management updates (Razzaq et al., 2021; Irfan et al., 2022). Information technology has made tremendous progress in the past decade and has penetrated all aspects of daily life (Guo et al., 2020). Among them, CTCH meets the needs of social progress and business development without compromising climate and natural resources (Yap et al., 2021). Research on CTCH mainly focuses on sustainable economic development, energy conservation, manufacturing technology upgrades, and their impacts on the natural environment (Shan et al., 2021; Tan et al., 2021; Li et al., 2022).

Among the observed associations between CTCH and sustainability issues in different regions, Razzaq et al. (2021) find that green innovations mitigate carbon emissions, particularly at higher emissions quantiles, by examining the asymmetric interdependence between carbon emissions and green innovation for BRICS economies from 1996 to 2017. Shan et al. (2021) consider the STIRPART model and find that innovation in climate technology plays an important role in achieving the SDG by keeping production processes on track with minimal negative impact on the environment in Turkey. Chien et al. (2021) demonstrate that information and communication technology (ICT) can help improve supply chain effectiveness and significantly reduce environmental degradation when considering the SDG framework. With the help of the bootstrap autoregressive distributed lag (BARDL) technique from 1990 to 2018, Meirun et al. (2021) believe that despite the remarkable economic growth achievements in Singapore, it still faces severe environmental-related problems, and technological innovation is an effective way to achieve environmental sustainability. In addition, many studies have explored the dynamic causal relationship between CTCH and environmental issues and have largely recognized their positive role in supply chain management, energy conservation, and emission reduction (Du et al., 2019; Jiao et al., 2020; Yang et al., 2020; Yin et al., 2020; Hao et al., 2021).

Given the uniqueness of environmental and climate issues (e.g., externalities, noncompetitiveness, transboundary, complex, temporal, and spatial heterogeneity, irreversible consequences, etc.), there is an urgent need to strategically promote and manage the invention, innovation, dissemination, and transfer of environmental technologies and use (Good et al., 2019; Gupta et al., 2021). Research on the relationship between green technology or CTCH and China’s economy or supply chain is rarely mentioned. China’s role is changing from a mere recipient of environmental technology transfer from developed countries to a technology supplier to other developing countries (Pandey et al., 2022). As significant emerging economies, China and developing countries face numerous challenges in economic transformation and supply chain upgrading, and they share many relevant development experiences. Learning from all parties involved and finding solutions makes CTCH critical to the effectiveness of supply chain environments.

Recent literature emphasizes that the U.S.-China trade war has revolutionized the global supply chains through direct tax shocks and general trade policy risk. Mao and Görg (2020) show that the cumulative and indirect effects of tariffs employed during the trade war had a wide-reaching impact on bilateral trade and third parties, negatively affected by the embedded involvement in trade in global value chains. Wu et al. (2021) build on this work using the OECD Inter-Country Input-Output model to develop a methodology for quantifying the cumulative tariff impact of both direct and indirect contributions. Their results indicate that the U.S. and China bore the highest indirect costs regarding the ripple effect, with the U.S. and China incurring $10 billion and $6.5 billion, respectively. In contrast, other third-party economies such as the EU, Canada, and Mexico also had to bear significant spillover effects, which would increase by 30%–70% due to full tariff pass-through. Similarly, Benguria and Saffie (2024) also illustrate how firms re-purposed exports towards substitute destinations following tariff shocks, with the re-allocation constrained by financial conditions and prior supply linkages. Fan et al. (2022) offer firm-level evidence on how U.S. firms with deep sourcing connections to China experienced deteriorating operating performance, especially those with complex sourcing networks. Kong et al. (2024) found that the cost of innovation for Chinese firms was quite significant when facing U.S. tariff exposure. However, some strategically increased innovation so as not to fall behind. From an environmental standpoint, Yuan et al. (2023) reveal how supply chain restructuring led to an increase in emissions worldwide as production moved to carbon-intensive economies. Recent works by Padhi et al. (2024) and Tsao et al. (2024) highlight how Industry 4.0 technologies can mitigate supply chain risk under increasing global uncertainty. Similarly, as Blessley and Mudambi, (2021) observed, resilience relies on resource reconfiguration and strategic alliances, as well as, further, as Handfield et al. (2020) followed, recurring disturbances (i.e., tariffs) and permanent changes (i.e., pandemics) are speeding up the transformation of GVCs into regionalized and adaptive GVCs. These studies highlight that the U.S.–China trade war has significantly reconfigured supply chains with implications for innovation, sustainability, and resilience in global industries.

The role of EPU in achieving the SDG, environmental governance, and supply chain resilience has been a key issue in academic research (Ali et al., 2025; Cheng et al., 2023; Kim et al., 2024). Economists have found that increased uncertainty can inhibit the flow of capital in supply chains and thus affect the normal functioning of supply chains, causing shocks.

On the one hand, EPU can cause companies to delay or cancel new supply chain network relationships that require upfront investment. Existing empirical research corroborates this statement. For example, establishing new supplier links requires start-up costs, while closing existing plants or supply chains can incur high separation costs such as severance, environmental cleanup, asset write-downs, and, in some cases, financial punishment (Cohen and Hau, 2020). In a cross-country study, Julio and Youngsuk, (2012) show that companies reduced capital expenditures and slowed the development of new markets and supply chains around elections. Based on the analysis of real options theory and the irreversibility of investment, the increase in EPU inhibits the capital output of enterprises in supply chains. This inhibitory effect is more obvious in enterprises with a higher level of investment irreversibility and a higher reliance on government spending (Gulen and Ion, 2016; Alfaro et al., 2018). Akron et al. (2020) conducted empirical research using data from companies in America’s hospital industry. They found that EPU is negatively correlated with trade credit provided upstream in the supply chains of hospitals. Therefore, in response to unpredictable EPU, if companies cannot stabilize existing supply chains, they may terminate ongoing relationships and explore other supply chains (Charoenwong et al., 2022).

On the other hand, higher EPU may prompt companies to mitigate potential future shocks by expanding new relationships and building stronger supply chains, thereby preemptively reducing operational risk (Sting and Arnd, 2014; Chaturvedi and de‐Albéniz, 2016). Thus, more perceived uncertainty incentivizes firms to increase the number of high-quality supplier relationships and reduce the number of low-quality suppliers, leading to disruptions in abandoned supply chains. Still, this behavior is only effective for diversifiable risks (Ang et al. al. 2017). The research of Marcus (1981) points out that in the EPU environment, it is difficult for enterprises as microeconomic entities to evaluate the benefits and risks they will face, so it is difficult to judge the impact of EPU on enterprise innovation. Some economists have found that the increase in EPU inhibits supply chains’ R&D or innovation. Xu (2020) argues that the increase in EPU will increase the financing cost of enterprises, thereby reducing the overall innovation ability of the supply chains. Firms with severe financial constraints and firms that rely more on supply chain finance are more negatively impacted by EPU (Nodari, 2014; Phan et al., 2021).

Most scholars believe that the increase of EPU will significantly reduce enterprises’ financing, investment, and innovation capabilities in supply chains, causing trouble in supply chain operations. From another perspective, the decline in corporate investment and innovation has impacted the supply chain, increasing the risk of global su-pply chain. As a result, EPU may affect global su-pply chain through financing, investment, and innovation channels.

Taken together, the literature indicates that these four strands, supply-chain pressures, energy prices, EPU, and U.S.–China geopolitical tensions, are not isolated but interconnected through identifiable spillover mechanisms. Supply-chain shocks (e.g., logistics disruptions, black-swan events) amplify Brent volatility through transportation and input costs, while oil price swings feed back into supply-chain stress by raising production and shipping expenses. EPU operates both directly, by altering investment flows and financing conditions, and indirectly, by magnifying volatility in oil and clean-tech markets; for instance, policy uncertainty can delay clean-energy deployment, slowing its stabilizing effect on supply chains. UCT shapes trade flows and technology transfer, disrupting both GSPCI and clean-energy innovation channels. Importantly, these channels are likely to differ by quantile regimes: in lower quantiles (tranquil states), shocks may be absorbed, whereas in higher quantiles (stress regimes), spillovers intensify and cross-domain amplification dominates. This framework motivates our empirical tests, which explicitly measure quantile-dependent connectedness among the four domains.

This research aims to examine the connectedness of supply chain disruptions in China, considering the CTCH Index, EPU, commodity prices, and tension between the U.S. and China, using monthly data from April 2006 to April 2024. Supply-chain stress for China is proxied by the China Supply Chain Pressure Index (GSCH). GSCH follows the NY Fed methodology used for the GSCPI family and is China-specific. The Invesco WilderHill Clean Energy ETF is used in this study to track NEX as an index, which rates corporations based on their contribution to clean energy, technical influence, and climate change (Bouri et al., 2022; Wang Xiong et al., 2022b). The source of data is downloaded from the Bloomberg terminal. The following select index is the economic EPU, calculated by (Baker et al., 2022; Baker et al., 2016) as an uncertainty measure. This index is obtained from St Louis Fed Financial Stress. To measure the degree of tensions, both geopolitical and trade-related, we use the UCT (U.S.–China Tension Index) index. This index is constructed using keyword text analysis of the top five U.S. newspapers and measures bilateral tensions using the frequency and context of trade war and political conflict frames. It has been extensively used in the literature to proxy for trade policy uncertainty and geopolitical stress (Baker et al., 2016). The UCT data is pulled from the Economic Policy Uncertainty homepage maintained by the Economic Policy Uncertainty Project, which provides”searchable time-series, for analysis of the economic policy. Brent crude oil is considered in this work as a proxy for global energy and commodity market behaviour. Brent is one of the world’s most widely used oil benchmarks and is the reference price for energy pricing. Energy Information Administration (EIA), a credible source of global energy data and projections. The Brent crude series can be seen as an aggregate market for energy in the sense that it summarizes and eliminates partial demand and supply-driven movements in energy prices, and it has been widely employed in fundamental empirical analysis of the relationship that exists between the commodity markets and the macroeconomic or environmental uncertainty. The data is obtained from the official website of EIA. Summary statistics of the data are shown in Table 1.

For better estimation, all data transform to return measure, which can be described as follows:

To standardize all data series and obtain consistent comparative results, we convert each variable to a return form, as expressed in Equation 1. R it = ln ⁡ T it T t − 1 (1)

Where: T it : index at time  t T t − 1 : index at time  t − 1 R it : index return at the time

As a sensitivity analysis of the quantile model, we also fit a time-varying parameter VAR in which the AR coefficients and shock variances can smoothly evolve. We estimate the model via a Kalman filter/smoother with discount (forgetting-factor) specification, so that recent observations receive more weight and older information is down-weighted; initialization is based on an OLS fit to the early sample, but the discount setting implies an effective memory of, say, 50 months, in line with our rolling-window reports. We calibrate the lag order and the forecast horizon as in pictures (four lags; 10-step-ahead horizon), and we update the time-dependent shock covariance with a standard exponentially-weighted updating schedule. At each time, we establish impulse-response dynamics and an order-invariant, generalized variance-decomposition to summarize the information on how much of the one-step-ahead forecast error variance in each series can be traced back to shocks in any other series. We then calculate the identical totals from, to, and net connectedness metrics as described in QVAR, but over time, paths of measures indicating evolving dynamics. Results are relatively insensitive to sensible variations in the discount rate, horizon, or lag, and they reflect regime switches identified by the quantile analysis (in particular after 2020).

Table 2 shows summary statistics of all data variables. The mean for all data series is positive and close to zero except for UCT, which is 2.01, exhibiting a high return. In contrast, the SD of the EPU is higher. The series kurtosis coefficients show excessive kurtosis (with a kurtosis value of 3). The skewness values varied from zero, suggesting that the series is not symmetric. Additionally, the skewness coefficient for all series is positive except UCT, indicating that outlier data were established in recent years compared to the estimation of the early period. All series have significant results for the JB test and not normality, indicating that the series times do not follow a normal distribution.

This section reports the net transmission of the spillover among the supply chain, CTCH index, US-China tension, first-tramp preference, Brent, and EPU returns, as shown in Figures 2A–D. We employ a sequence of different quantiles, starting with the 5th (lowest quantile) and ending with the 95th as the highest. The warmer in this figure varies between blue (net receipt) and red (net transmit). As shown, the supply chain confirmed the outcome in The index of GSCH shows an always-present transition from being a net receiver (blue) before 2016 to a net transmitter (red) after 2017, with the net transmitters dominating in periods of significant worldwide interruptions. In particular, the C19P (2020), U.S.–China trade tensions (2018–2019), and the eruption of the RUC (2022) coincide with greater connectedness at higher quantiles (0.8–0.95). The observation suggests that (supply chain) disruptions are responding to changes on the demand side and amplifying volatility during high uncertainty. The findings report that supply chain connectedness is higher during C19P than in other periods of crisis, explained by tens of thousands of jobs being lost, production being disrupted, numerous airports and ports closing, and the cost of the shipping industry rising. This finding confirms the previous studies (Chowdhury et al., 2021; Luo and Kwok, 2020). It is also connected to domestic lockdown measures (Bonadio and Huo, 2020). Moreover, strong evidence of the connectedness between the supply chain and its determinants at extreme quantile levels highlights the need to increase countries’ resilience to supply-chain-related shocks. The supply chain and climate change index are closely connected during crises and extreme quantiles. This result is consistent with Si et al. (2022), who found links between environmental degradation and the supply chain in advanced economies and emerging markets, including China. The degree of connectedness ranges between 20 and 50 at the median quantile, influenced by the quality of infrastructure associated with trade and transportation in China. It also demonstrated how logistics operations’ effectiveness and quality impact China’s economic success (Hong et al., 2019). Economic Policy Uncertainty acts largely as a net transmitter, especially at the upper quantiles in high-stress times (e.g., C19P, U.S.–China tension, RUC). The increase in the strength of red coloration since 2018, especially in the 0.85 quantiles and over, indicates EPU’s increase of relevance in transmitting macroeconomic shocks. But before 2016 and in lower quantiles (0.1–0.3), EPU presented itself as a net receiver, implying that it seems to absorb rather than create shocks in a near-constant policy environment. UCT Index presents the latter with a more cyclical spillover structure. In the first stage of U.S.–China trade tensions (2013–2016), the index indicates that the transmission is strong at intermediate to high quantiles (between 0.4 and 0.9). This is especially the case around 2015, when President Trump was on the campaign trail and making some of his initial trade threats. A second wave of strong red clusters appeared around 2018–2019 and again during the pandemic. There is some evidence that after 2020, the index exhibits more symmetry in terms of net receiver role at lower quantiles, which agrees with the diminished independent influence of Pakistan outside the peak of geopolitical stress. The Brent crude oil price index exhibits modest and period-specific spillovers. Notably, the sudden redshift at higher quantiles circa 2014–2015 and during the 2020 oil price crash (due to C19P lockdowns and the OPEC + dispute) indicates Brent’s role as a transmitter in the event of extreme market volatility. At lower quantiles and tranquil times, Brent acts more as a net receiver, responding to demand-side economic shocks instead of a source of system-wide stress. While we do not illustrate this in the present figure when examining the Climate Technology Index (e.g., NEX), it should act as a net recipient at the early stage of innovation diffusion while becoming a transmitter in recent periods, for example, green recovery agendas in the aftermath of the C19P era and response to changes in global climate policies (e.g., COP26–28). We anticipate more coupling at higher quantiles post-2020, as markets are sensitized to climate-investment signals. Since the COP26 declaration, the role of environmental CTCH, green trade, public finance for infrastructure, and supply chain management for technology and innovation has become increasingly important, in line with the SDG (Dwivedi et al., 2022). Furthermore, the direction arrow and transmission spillover from the supply chain to the technology and climate index can explain how new technologies present the need for the promising potential for improvement throughout the supply chain (Gupta et al., 2021). Blockchain technology can potentially decrease administrative expenses while increasing supply chain transparency and traceability. Based on the framework of the Sustainable Development Goals, we provide suggestions from the perspective of the CTCH, considering economic uncertainty. First, an enterprise is not just a production and manufacturing department. Managers should enhance their understanding of corporate social responsibility and quality to stay informed about policy trends while operating. Microscopic individuals can also achieve macroscopic sustainable development goals. Second, the industrial manufacturing sector may need to periodically upgrade its technology to maintain the environmental performance of its production processes. Third, policymakers should strengthen environmental regulations and legislation, and industrial production and manufacturing should be based on reducing natural resource consumption and protecting public property resources.

We cross-validate results from the TVP-VAR framework and a node-based Quantile VAR (QVAR) network analysis with different period rolling windows for robustness results.

Firstly, to verify our results, we apply the QVAR approach and analyze the network structure of the spillover links over a 70-month rolling window, with a forecast horizon of 20 and BIC-based automatic determination of lags. The analysis is carried out at three sample quantiles: low extreme (5th percentile), median (50th percentile), and high extreme (95th percentile). The resulting networks are shown in Figure 3, with blue nodes being net transmitters of shocks and yellow nodes being net receivers. The thickness of the edge reflects the degree of directional connectedness between the variables.

In the below quantile (0.05), which seizes extreme downside conditions, GSCH and EPU show up as the main shock transmitters. At the same time, BRENT and UCT are net receivers, which means that global logistics impediments and policy ambiguity are sources of market stress in the face of adverse market conditions.

The mainstream becomes more connected at a quantile of 0.50, indicating a closer interplay between variables. GSCH and BRENT both play a relatively stronger transmitter role. At the same time, UCT and EPU are still potential shock receivers, representing the evolution from exogenous risk to even more endogenous spreading in regular market periods. On the other hand, in the upper quantile (0.95), conditioned on extremely positive outcomes, both NEX (climate tech) and BRENT become significant transmitters of dominance, signaling that the positive effects of innovative clean technology and energy markets have a more substantial amplifying impact on the system. On the other hand, GSCH, EPU, and UCT become net recipients again, illustrating their passive reaction to favorable market conditions. The frequency-specific network results corroborate the nonlinear and asymmetric characteristics of systemic risk transmission among climate, energy, supply chain, and geopolitical aspects. The diversity of node roles in quantiles also adds evidence to the robustness of the quantile-based connectedness framework to capture heterogeneous spillover properties in different market states.

Secondly, Figure 4 shows the time-varying net pairwise connectedness between five essential variables, which are NEX, UCT, SCH, Brent crude oil, and EPU, under the -VAR model with a 50-month rolling window, 10-step-ahead forecast horizon, and four lags. A positive value of the impact index means the variable is a net transmitter of shocks, whereas a negative value signifies a net receiver. The findings demonstrate significant heterogeneity in transmission dynamics across time. The NEX index was in net receiving countrydom until 2020, when it became a strong net transmitter, when the world entered a new phase of post-pandemic recovery and increasing global climate action, notably with the COP26 summit. This change represents an increasing power of green technologies to generate systemic market spillovers. In comparison, the unithank index in the U.S.–China trade war period (2013–2019) exhibited a significant net transmission, which indicates an increase in the couple’s relationship problems. However, UCT reverted to a net receiver after 2020, meaning its effect became attenuated in the context of wider geopolitical and macroeconomic upheavals. There were variations throughout time for the GSCH index, which, after 2018, became a net transmitter (especially after the C19P and the related supply chain disruptions) and served as an intermediary for the diffusion of economic shocks. Brent oil prices, in sharp contrast, averaged near neutrality or, at most, a marginal net receiver over the whole period, with only short periods of net transmission coinciding with major oil market shocks in 2008, 2014, and 2020. Third, the contagion effect of the EPU index has transformed from a beneficiary in previous years (2000–2007), to a majority gainer since 2008, and significantly peaked during the C19P crisis, the RUC, and energy tension, demonstrating an increasing systemic impact of economic policy uncertainty in stormy periods. Overall, the resulting network maps four are consistent with the dynamics of the identified dynamic spillover roles from the TVP-VAR-based net PW connectedness. In sum, the consistency of findings on these two different but complementary approaches, TVP-VAR and quantile-based network analysis, offers strong empirical support for the robustness, nonlinear, and quantile-dependent spillover structures, guaranteeing the reliability and robustness of our main results.