The time delay restraining the herd behavior with Bayesian approach

Abstract

We explore the herd behavior of stock prices influenced by the time delay in a finance system with the delayed Heston model. On the basis of a Bayesian approach for delayed Heston model that we proposed and parameter Bayesian estimation, we simulate the absolute deviation between mean residence time of positive and negative returns to characterize the herd behavior. In the plots of absolute deviation against the mean reversion long-run variance of volatility or cross correlation between two Wiener processes of stock price and volatility, the results demonstrate an increasing phenomenon of herd behavior with increasing delay time. In the plot of absolute deviation against the amplitude of volatility fluctuations, the result indicates an optimal delay time matching minimum herd behavior, i.e., the time delay restrains the herd behavior. Also, the optimal mean reversion long-run variance of volatility and cross correlation concerning minimum herd behavior can be observed.