A Study on Price Volatility in the Carbon Emissions Trading Market Based on the GARCH-ANN Model

Abstract

With the advancement of the global "dual-carbon" strategy, emissions trading systems (ETS) have gradually become a central policy tool in driving green economic transformation. However, sharp fluctuations in carbon prices not only threaten market stability but also directly impact corporate emission reduction decisions and long-term investments. This study innovatively combines the GARCH model's ability to capture volatility clustering with the artificial neural network (ANN)'s strong capability in modeling nonlinear relationships to construct a GARCH-ANN hybrid model. The model systematically analyzes the volatility characteristics of carbon allowance prices and their dynamic response mechanisms to exogenous shocks. Taking the EU Emissions Trading System (EU ETS) and China’s pilot markets as research objects, the study incorporates high-frequency trading data and policy dummy variables, and empirically tests the predictive performance of the hybrid model. The results show that the hybrid model significantly outperforms single models, and policy shocks exhibit time-varying effects. This study provides both theoretical and empirical support for the design of carbon market regulatory policies, corporate green investment decisions, and the interconnection of international carbon markets.