In complex financial markets, investors’ behavior is often influenced by those around them, which in turn generates a herd effect. This paper explores the emergence process of flocking behavior in the tradable green certificate market using an Agent-based nonlinear dynamics model. By constructing a simulation environment, the impact of investors’ decision-making based on imitation behavior is analyzed, and the evolution mechanism of herd behavior in the market is revealed. The experimental results show that the intensity of flocking behavior is affected by the quality of information, the initial state of the market and individual imitation tendency. In the baseline scenario, when the imitation probability of low-quality information subjects is high, the flocking behavior shows a significant increase; the flocking behavior of high-quality information subjects fails to show up when the initial imitation probability is low. Specifically, when the initial imitation scale of high-quality information subjects is less than 0.439, the flocking behavior does not appear, while the imitation behavior of low-quality information subjects has a significant impact on the market price. The experiment verifies the validity of the model, and the deviation of the simulation data from the actual market data is less than 0.05, indicating the feasibility of the model in practical application. The study provides a theoretical basis for understanding the volatility of the green certificate market and predicting herd behavior.
