Climate change represents one of the most significant systemic risks to global financial stability, yet traditional asset pricing models inadequately capture the complex, non-linear dynamics of climate-related financial risks. This paper develops an integrated theoretical framework that synthesizes Explainable Artificial Intelligence (XAI), Game Theory, and advanced Machine Learning techniques for climate-aware financial decision-making. We introduce novel mathematical formulations including Robust SHAP Values for uncertainty-aware feature attribution, a three-player game-theoretic model capturing strategic interactions among investors, regulators, and firms, and multi-temporal portfolio optimization with climate constraints. The framework extends the Capital Asset Pricing Model to incorporate climate beta and establishes theoretical conditions for Nash equilibrium existence in climate-constrained markets. Our empirical methodology employs a hierarchical ensemble architecture combining XGBoost, LSTM networks, Graph Neural Networks, and FinBERT for comprehensive climate risk assessment across multiple data modalities. Results demonstrate superior predictive performance with enhanced model interpretability, providing practitioners with actionable tools for sustainable investment strategies while maintaining regulatory transparency. This research contributes to the emerging field of climate finance by bridging the gap between computational sophistication and decision-maker comprehension.
Keywords : Climate Risk, Asset Pricing, Explainable AI, Game Theory, Machine Learning, SHAP Values, Portfolio Optimization, Sustainable Finance
Citation: Massimiliano F. (2026). Climate Risk and Uncertainty: Environmental Sustainability and Asset Pricing through Explainable AI and Game-Theoretic Frameworks. J AI & Mach Lear., 2(2):1-4.
DOI : https://doi.org/10.47485/3069-8006.1011