Artlficial intelligence, clean, and dirty energy markets: a quantile-on-quantile connectedness analysis of volatility

ECONOMIC CHANGE AND RESTRUCTURING, cilt.59, ss.1-59, 2026 (SSCI)

The growing interaction between artificial intelligence (AI), clean energy, and dirty

energy markets has created new channels of systemic risk that cannot be captured

by linear or mean-based frameworks. This study examines the nonlinear and quan-

tile-dependent volatility spillovers across these markets over the period 2018–2025

using a Quantile-on-Quantile Connectedness (QQC) approach, which allows spill-

overs to vary across market states and over time. The results reveal a strongly asym-

metric and state-dependent transmission structure. AI-related assets are associated

with the strongest net transmitting positions in volatility connectedness, particularly

in upper-tail and high volatility regimes. Clean energy markets are more frequently

observed in net receiving positions in the short run, but their connectedness pro-

files become more net transmitting over longer horizons as green-transition dynam-

ics strengthen. Dirty energy assets are more often associated with net receiving

and weaker outward spillover positions during turbulent periods while generating

relatively weaker feedback effects under normal conditions. Dynamic evidence fur-

ther shows that connectedness intensifies sharply during major global disruptions,

especially the COVID-19 pandemic and the Russia-Ukraine conflict, confirming

that tail-related spillover patterns are more pronounced within the connectedness

structure. Overall, the findings show that volatility linkages between innovation

and energy markets are nonlinear, time-varying, and highly regime-specific. These

results provide important implications for portfolio diversification, hedging strate-

gies, and policy coordination in an increasingly interconnected financial system.