Market Behaviors under the Stock Index Circuit Breaker using an Agent-based Approach

Abstract

In this paper, we combine an agent-based model of multi-asset stock market with circuit breaker mechanism and empirical analysis of S&P 500 Index to study market behaviors under the circuit breaker. The artificial stock market model can reproduce the stylized fact that the stock index triggers a circuit breaker. The results show that the smaller the circuit breaker, the more likely the circuit breaker events will occur. And the higher the traders’ index-dependent strength, the more likely the circuit breaker events will occur. From the perspective of market behaviors under the stock index circuit breaker, we find that the market volatility, the correlation of individual stock returns and the convergence of traders’ behavior on the circuit breaker day are higher than those before the circuit breaker day when the circuit breaker in the market is set relatively small or traders refer to the stock index more for decision-making. This is because the smaller circuit breaker mechanism and traders’ more reference to the stock index for decision-making make the behavior of originally heterogeneous traders in the market converge, which aggravates the occurrence of circuit breakers.