针对近几年在研究金融市场超高频序列时出现的ACD模型和SCD模型,运用MCMC方法采用沪市分笔交易数据得到两类模型的参数进行估计值及DIC值,分析了模型的收敛性,稳健性、拟合效果及复杂性,结果表明,两类模型均趋于收敛,在稳健性及收敛速度方面,ACD模型有优势,而在数据拟合效果及复杂性方面,SCD模型具有优势. Aiming at the ACD model and SCD model which appeared in the research of UHF sequence in financial market in recent years, the MCMC method was used to get the estimated values ​​and DIC values ​​of the two models’ parameters by using Shanghai stock transaction data. The convergence of the model , The robustness, the fitting effect and the complexity. The results show that both models tend to converge, and the ACD model has advantages in terms of robustness and convergence speed, while the SCD model has advantages in data fitting effect and complexity .