The effects of attention to climate change on carbon, fossil energy and clean energy markets: Based on causal network learning algorithms

Climate-driven market interdependencies critically influence emission reduction strategies, yet the causal mechanisms linking carbon, fossil, and clean energy markets remain poorly quantified -a knowledge gap impairing risk-aware policymaking and sustainable investments. This study uses the Peter-Clark Momentary Conditional Independence (PCMCI) method in causal network learning algorithms to systematically quantify the multidimensional dynamic causal effects of climate attention, yield, and volatility across these three markets. We construct an investor climate attention indicator using 47 climate-related keywords and analyze daily price data from seven carbon, four clean energy, and three fossil energy markets. Empirical results demonstrate that structural heterogeneity exists in the price transmission mechanism, and attention to climate change disturbances trigger cross-market chain reactions. In the yield dynamics, an expectation-driven asymmetry emerges with a multiplier amplification mechanism. The volatility transmission follows a hierarchical path. Evolutionary analysis reveals the transformation of the energy system and the core position of the carbon market. This study offers investors quantitative guidance for constructing climate-resilient portfolios, highlighting the pivotal role of climate attention within the green financial system.