Research On Carbon Market Risk Defense Method Considering Participants' Behaviors

The carbon market has become the major mechanism for global carbon emission control.However,due to the lack of elasticity of the allowance supply curve,and the fact that allowance demand of regulated companies is greatly affected by the external environment,the drastic fluctuation of carbon prices,especially the extreme carbon prices occure frequently,which is not conducive to capital investment in structural emissions reduction and achieving the emission reduction targets set by policy makers,so the carbon market risks have shown some signs.Therefore,if carbon market regulators want to achive reduction targets at a lower social cost,it is necessary to study the carbon market risk defense method.The research of carbon market risk defense methods mainly uses the carbon market dynamic assessment model,to analyze carbon market risk under different disturbances and to develop its risk defense strategies.However,the current model of market participants are mainly implemented with heuristic and intelligent algorithms,such as regular transactions,and the determined allowance trading volume based on the emission imbalance,which is difficult to reflect the participants,decision-making preferences,and affects the accuracy of market dynamic simulation,furthermore affecting the credibility of the carbon market risk assessment results.The existing risk defense strategy is mainly to carry out corrective control after the crisis occurs,but on the one hand,the deviation of the carbon market risk assessment values is too large to support the optimization of the control plan,and on the other hand,as its control timing is close to the compliance timing,the small emission reduction space of regulated companies will reduce the control effectiveness.Therefore,this paper studies carbon market risk defense methods considering the behavior characteristics of market participants.Specifically,this paper has mainly achieved the following results:1.In order to solve the problem that the participants' decision-making preferences are difficult to obtain,carbon trading experiments are designed to extract the characteristics of transaction behaviors.The mathematical models are not related to participants'behavior characteristics such as allowance allocation methods,allowance auction rules,and allowance exchange models.With the participants' actual decision-making behaviors as an external input to the experiment,the experimental platform is constructed which can support interactive simulation between subjective decision-making behaviors and market rules that reflect objective laws.The phenomena of "ultimatum game" and "loss aversion"are observed in the experiment.Furthermore,experimental results also show that participants'trading behaviors are closely related to price dynamics of carbon market,the distance between the present timing and the compliance timing,and projected emissions imbalances.2.A modeling method combining the merits of experimental economics and agent-based computational methods is proposed,and the behavioral agent model for the participants is constructed,based on the trading samples.Through human-computer interaction simulation based on experimental economics,and extracting knowledge from multi-scene simulation results to qualitatively and quantitatively analyze the influences of key feature variables on participants' decision-making,a behavioral agent model is constructed with the same distribution characteristics of participants' decision-making,and thus a mixed interactive simulation environment is gradually established that is compatible with participants and multi-agents.Finally,the effectiveness of the behavioral agent model is verified by comparing decision-making results of the agents and participants under the verification scenario.The simulation results show that the decision-making deviation between the behavioral agents and the participants is almost equivalent to the"randomness" deviation of the participants themselves in the repeated experiments.This methodology can overcome the scale limitations and meanwhile maintain the merits of experimental economics.3.The carbon market dynamic assessment model considering participants' behaviors is constructed based on the behavioral agent models.Firstly,the amount of power generation and emissions are determined based on the unified clearing price mechanism,and secondly the trading volumes are determined by the behavioral agents,and finally the carbon price is determined through the demand-price model that simulates the impact of supply and demand on prices.According to this process,the dynamic trend of carbon market operation under the entire compliance period is simulated.Taking the carbon price crisis in the European Union Emissions Trading System(EU ETS)as an example,the feasibility of this model is verified by simulating its carbon price trend.The simulation results show that the price dynamics obtained by the simulation are close to the actual overall dynamics of the price,reflecting the impact of the two major disturbances of the"global financial crisis" and the "European debt crisis" on the carbon market price,and furthermore it can describe the mechanism that the influences of economic disturbances are transmitted to the carbon market price gradually through power load,carbon emissions,participants,market position,trading behaviors,and the allowance supply and demand.4.The multi-line risk defense framework of carbon market is proposed to solve the problem that it is difficult to effectively defense risks based on the corrective control strategy.Through the analysis of the actual carbon market crisis defense process,compared with the general law of the crisis defense framework,we discuss its application in carbon market risk defense.Furthermore,combined with the carbon market risk mechanism and space-time characteristics,the multi-line risk defense framework of carbon market is proposed and designed by defining key features such as "disturbance occurrence","parameter violation" and "market malfunction",and the control features and its application scenarios in each line of framework are analyzed,to support the strategy coordination and optimization for carbon market risk defense.5.The control measures of the various defense lines of the framework are coordinated and optimized,based on the carbon market dynamic assessment model that considers participants' behavior characteristics and the multi-line risk defense framework of carbon market.The optimization models and control plans for each defense line are constructed and evaluated,such as preventative control,emergency control,corrective control and coordinated control.Taking the carbon price crisis of the EU ETS in 2011 as an example,and allowance adjustment as the major control measure,the control effects of various control plans are simulated and analyzed under different disturbance scenarios.The simulation results show that in the face of uncertain disturbances,compared with the corrective control,the coordinated preventive and emergency control can reduce the total cost of risk control and enhance the effectiveness of risk defense.