The cross section of stock returns in an artificial stock market
开发了一个基于智能体的股票市场模拟模型，能复现真实股市的时间序列和横截面特征，包括随机游走、过度波动、因子溢价等，可作为金融研究的补充实验工具。
In this paper, we develop an Artificial Stock Market (ASM) - an agent-based simulation model of the stock market. We present the model and demonstrate its ability to replicate the main empirical features of equity markets. Our model can generate the stylized facts in the time-series domain such as the random walk property, long memory, excess volatility, negative skewness, and excess kurtosis. In the cross-section of stocks, the model generates excess co-movement as well as factor premia as in real-life equity markets. A well-defined ASM can generate large amounts of data, and function as a laboratory. Hence, it can be a complementary methodological approach to research in financial markets.