Game-Theoretic Market Dynamics and Machine Learning-Informed Demand: Joint Effects of RPS, Subsidies, and Green Certificate Prices

Decarbonizing the power sector is essential for meeting climate targets, and many electricity markets now rely on a combination of renewable portfolio standards, per-unit subsidies, and green certificate trading to accelerate renewable deployment. However, because these instruments simultaneously affect firms’ incentives and demand-side acceptance, their joint impacts on pricing behavior and renewable investment decisions remain difficult to characterize. This study investigates the joint effects of renewable energy standards, subsidies, and green certificate prices on pricing behavior and renewable energy investment decisions in the electricity market. By developing a game-theoretic model that incorporates government interventions such as renewable portfolio standards (RPS), per-unit subsidies, and a green certificate trading mechanism, we find that low subsidies typically lead to compliance-driven renewable investments, where generators adjust prices downward to stimulate demand. In contrast, higher fossil fuel costs and subsidies drive up prices and result in increased renewable energy output. Additionally, the price of green certificates plays a pivotal moderating role, influencing the effectiveness of subsidies. To further substantiate these conclusions, we integrate machine learning methods to complement the theoretical framework. Using survey data, we construct a purchase-intention target at the consumer level and apply supervised learning techniques to predict the high-intention consumer group based on variables such as policy cognition, price acceptance, market trust, and environmental support. A Gradient Boosting Classifier (GBC) is used as the primary model, with logistic regression, random forest, and RBF SVM serving as benchmark models. The machine learning results are consistent with and complement the game-theoretic insights. Specifically, they confirm that low subsidy levels are indeed more likely to lead to compliance-driven outcomes, where firms rely on price reductions to boost demand. Furthermore, higher fossil fuel costs and subsidies are shown to push prices upward, stimulating greater renewable energy production. Importantly, the machine learning models also highlight the moderating effect of green certificate prices, confirming their role in enhancing or diminishing the impact of subsidies. These results underscore the importance of coordinated policy design that integrates consumer demand behavior, offering a more comprehensive understanding of market dynamics and effectively balancing compliance pressure, market efficiency, and long-term investment incentives.