Cryptocurrencies, green finance, and sectoral CO₂ emissions: A nonlinear time-frequency analysis

Abstract

This study examines how cryptocurrency markets, green finance, energy prices, and sectoral CO₂ emissions interact across time horizons and market conditions. Using daily data for 2019–2025, it links Bitcoin, Ethereum, the Green Bond index, and WTI crude oil to industrial, power, transportation, and residential emissions. The analysis applies wavelet quantile correlation and quantile-on-quantile regression to capture nonlinear, asymmetric, and frequency-specific dependence that conventional linear models may obscure. The sectoral and regime-based design tests whether environmental effects arise from broad crypto-market dynamics or from specific energy-intensive transmission channels. The findings show that Bitcoin’s emission effects are mainly medium-run and concentrated at moderate-to-upper emission quantiles, whereas Ethereum’s effects are systematically weaker, consistent with the environmental implications of its proof-of-stake transition. The Green Bond index is mildly and positively associated with industrial and power emissions, suggesting a short- to medium-run green-paradox mechanism, while WTI oil remains the most consistent driver across sectors. Post-COVID relationships are also stronger than pre-COVID dynamics. These results imply that climate and financial regulation should be horizon-aware and protocol-sensitive. Policy should distinguish proof-of-work from proof-of-stake assets, strengthen verified green-bond standards, and align mining disclosure, grid-management, and carbon-pricing instruments with the medium-term investment cycle through which digital-asset markets affect emissions.