Intraday Trading Patterns In A Double Auction Agent-Based Stock Market

Financial market microstructure studies the discovery,formation and evolution of asset pricing,and explains how trading mechanism,market structure and behavior of investors effect the process.Market microstructure studies of intraday trading pattern have established that there is a regular pattern of volumes and price near both the open and close of the trading day.But the available literatures fail to explain characteristics of our stock market under continuous double auction trading mechanism.Chinese stock market designs a different trading mechanism with foreign stock markets,therefore the relationship between volume and price in our country not only affected by investor behavior,but also affected by the trading mechanism.Studying the characteristic of volume and price may provide guidance about trading strategy how to collect and use information for investors.Based on available literatures,we develop an agent-based market microstructure simulation with both informed agents and uninformed agents and try to explain the influence on a regular pattern of high volumes near open of the trading day through the aspect of information.Compared to the earlier literature,the innovation of the paper lies in the learning efficiency we set as a proxy of information.For the given market structure we analyze the impact of external information on volumes,prices and trading strategies.Our results demonstrate that information transmitted in the market more effectively with the improvement of learning efficiency,which makes a high volume pattern near the open of the trading day.Fluctuations in stock prices would impact external information,the more information,the greater volatility.When information could not transmitted in the stock market because of the low learning efficiency,there is a time lag effect in short time in the market for information,especially for overnight information.Informed agents prefer to trade at the mid-term and close of a trading day,while uninformed agents prefer to trade at the open.With the improvement of learning efficiency,both informed and uninformed agents make an investment decision at the opening,and the time lag effect disappears.