A Behavioral Model of Investor Sentiment in Limit Order Markets

By incorporating behavioral sentiment to a model of limit order market, we show that behavioral sentiment not only helps to replicate most of the stylized facts simultaneously in limit order markets, but also plays a unique role in explaining these stylized facts that cannot be explained by noise trading, such as the fat tails in the return distribution, long memory in the trading volume, an increasing and non-linear relationship between trade imbalance and mid-price returns, as well as the diagonal effect or event clustering in order submission types. The results show that behavioral sentiment is an important driving force behind many of the well-documented stylized facts in limit order markets