Dynamic order dispersion and volatility persistence in a simple limit order book model

We extend the dynamics of a basic stylized limit order book model introduced in Chiarella and Iori, (2002). The original model is capable of generating some key market microstructure features, but it cannot recreate longer range persistence in volatility. We explore a very simple and intuitive addition to the stylized, near zero intelligence behavior of traders that is capable of delivering persistent volatility and fat tailed return distributions. Specifically, we introduce a feedback mechanism where agents adjust their order placement in response to recent volatility: when short-term volatility rises relative to long-term volatility, traders place their limit orders farther from the current market price. This volatility-responsive order shading creates a self-reinforcing cycle where higher volatility leads to sparser order books, which in turn amplifies subsequent price movements and volatility. We also extend our results to some well known properties of realized volatility measures. Our model reveals that all these properties are generated by order book patterns which are consistent with empirical liquidity patterns in market microstructure studies. We feel our parsimonious agent-based model may be used as a test bed for developing agents with more complex behavior.