Estimating a regime switching pairs trading model
研究了一个离散时间配对交易模型，其中资产价差动态受不可观测的马尔可夫链状态影响，并给出了基于价差观测递归估计市场状态和模型系数的方法。
We consider a discrete time pairs trading model which includes regime changes in the dynamics. The prices of the pair of assets, and so their difference or spread, depend on the state of the market, which in turn is modelled by a finite state Markov chain. Different states of the chain give rise to different parameters in the dynamics of the spread. However, the state of the chain is not observed directly but only through the prices or spread. Based on observations of the spread, this paper provides recursive estimates for both the state of the market and all coefficients in the model.