Relation exploration between clean and fossil energy markets when experiencing climate change uncertainties: substitutes or complements?

The study addresses how extreme climate change and associated uncertainties affect interactions among clean and fossil energy markets, asking whether these energies act as substitutes or complements across time horizons. Extreme weather disrupts energy supply chains and power systems, generating price volatility and potential systemic risk. With many economies entering a critical energy transition, understanding cross-market connectedness and volatility spillovers is vital for risk management and portfolio construction. Prior evidence shows energy markets are interconnected and subject to external shocks; however, how climate change influences clean energy markets across different time scales remains underexplored. This paper proposes to analyze spillovers across wind, solar, oil, and gas from 2015–2022 and to decompose dynamics by time frequency to inform short- and long-term policy and investment decisions.

Research documents strong connectedness and volatility spillovers across energy and broader financial markets, with shocks transmitting to commodities, equities, precious metals, and cryptocurrencies. Traditional energy markets are often more sensitive to external shocks. Growing focus on renewables finds significant linkages between clean and traditional energy markets and within clean energy segments (e.g., wind–solar), and spillovers to food, metals, and crypto. Extreme events (financial crises, COVID-19) exacerbate spillovers and systemic risk. Climate-related factors—investor attention to climate change, climate policy uncertainty, and extreme weather—link to energy and financial market volatility and can heighten spillover networks, including between fossil, renewable, and carbon markets. Methodologically, GARCH-family models (including BEKK-GARCH) are widely used to identify spillovers; EMD has been applied to decompose nonlinear, nonstationary signals. Yet, few studies examine climate change impacts on clean energy markets at different time frequencies, motivating this paper’s time-frequency decomposition and event-focused analysis.

The paper constructs a Temporal Volatility Spillover Decomposition (TVSD) framework that integrates Empirical Mode Decomposition (EMD) with a multivariate BEKK-GARCH model. Steps: (1) Model conditional means with a simple specification to obtain residuals and use BEKK-GARCH(1,1) to capture conditional variance-covariance dynamics and cross-market ARCH/GARCH effects, identifying spillover directions via parameter significance and Wald tests. (2) Apply EMD to each market’s time series to decompose into Intrinsic Mode Functions (IMFs) and a residual trend, capturing high- and low-frequency components. (3) Use sample entropy to classify IMFs: higher-than-original entropy IMFs form the high-frequency (short-term, sentiment/speculation) component; lower-than-original entropy IMFs form the low-frequency (event-driven, medium/long-term) component; the residual is the trend (long-term). (4) Reconstruct three subsequences—high-frequency, low-frequency, and trend—and estimate BEKK-GARCH spillovers for each to assess time-scale-specific connectedness. Data: daily log returns for wind and solar indices (NASDAQ OMX Green Economy indices: GRNWIND and GRNSOLAR), WTI crude oil futures, and Bloomberg Natural Gas from 2015–2022. Event analysis: three intervals with pronounced low-frequency fluctuations aligned with notable climate activity—2015.09–2016.09 (super El Niño and global heat), 2019.01–2020.01 (European snowstorms; Amazon forest fires), and 2021.04–2022.04 (bomb cyclone in US, unusually warm European winter, Tonga eruption). Robustness: re-estimation with optimal lag order (BIC, lag 4) confirms main results.

• Full-sample (2015–2022) original series: Clean energy markets (wind–solar) show clearer connectedness than oil; gas often transmits volatility. Oil appears relatively independent in the aggregate original series.
• Decomposition: Spillovers are predominantly significant in low-frequency and trend terms (medium/long-term), while high-frequency spillovers are often weaker or insignificant among clean markets. This indicates event-driven and persistent long-term connectedness.
• Climate events and time-scale effects: In selected climate-intensive intervals, spillovers become stronger and more pervasive, including for oil. This suggests frequently occurring climate issues can trigger or amplify oil’s spillovers with other energy markets.
• Short vs long term: As climate uncertainties rise, markets exhibit heterogeneous short-term reactions (weaker or selective high-frequency links), but synchronized trends and stronger spillovers in low-frequency/trend components over the long run.
• Substitutes vs complements: Evidence supports that clean and fossil energies behave as substitutes in the short term (limited short-run spillovers among some pairs) and as complements in the long run (synchronized low-frequency/trend spillovers across markets).
• Lagged responses: Low-frequency fluctuations often lag extreme climate events by several months, indicating delayed transmission of climate-related risk to energy prices.
• Robustness: Results are consistent under alternative lag specifications (BIC-selected lag 4).

The analysis directly addresses whether clean and fossil energy markets are substitutes or complements under climate uncertainty. By decomposing market dynamics into high-, low-frequency, and trend components, the study shows that short-term market reactions are heterogeneous and less interconnected—consistent with substitutability when shocks or speculation affect markets unevenly. In contrast, the persistent co-movements and significant spillovers in low-frequency and trend terms indicate stronger long-run integration, consistent with complementarity of energy markets over longer horizons. Climate-related events increase the likelihood and magnitude of spillovers, including involving oil, which appears more isolated in the full-sample original series but becomes connected during climate-intensive periods. These findings imply that climate shocks can reconfigure cross-market linkages temporarily and that long-run energy system dynamics are jointly influenced by shared macro, policy, and climate factors. Policy and portfolio implications include: in short-run climate shocks, focus on vulnerable markets and consider substitution strategies; over the long run, coordinated policies and diversified portfolios should account for the complementary, system-wide co-movements driven by climate and transition forces.

The paper proposes a TVSD framework combining EMD and BEKK-GARCH to identify volatility spillovers across wind, solar, oil, and gas at different time scales from 2015–2022 and during climate-intensive events. It finds that clean markets are more interconnected than oil in the aggregate, but decomposition reveals that significant spillovers concentrate in low-frequency and trend components, and climate-related periods strengthen cross-market linkages, including oil’s. The central conclusion is temporal: energy markets resemble substitutes in the short term but complements in the long run, as evidenced by weak high-frequency and strong low-frequency/trend spillovers, respectively. These insights aid policymakers and investors in managing climate-induced volatility, designing energy portfolios, and crafting coordinated transition strategies. Future work should more precisely quantify climate event impacts and mechanisms and examine dynamic correlations across frequencies.

The study does not explicitly quantify the intensity or direct causality of specific climate events on market spillovers; attribution remains indirect. The mechanisms through which climate events affect different energy markets are not fully delineated. Future research should develop objective measures of climate event impacts, clarify transmission mechanisms, and incorporate time-varying correlations across frequencies to enhance causal inference and generalizability.