An behavior finance approach to asset price correlations and applications to financial engineering

在全球金融危机之后，一些关于美国股票市场的研究报告显示，行业之间的相关性更高。这种现象可能无法通过传统的风险收益分析来解释。我们应用一个包含可逆链和不可逆链的马尔可夫转换模型，发现市场的结构变化似乎发生在2003年，早在危机之前，我们还分析了买卖价差，流动性的措施之一，在信用违约互换（CDS）市场。这些价差是由市场参与者的行为决定的。我们成功地拟合了一个与时变布朗运动相结合的自激过程。 此外，我们研究了基金经理的最优停止问题的基金杠杆。

After the global financial crisis, a number of research papers on the U.S. stock market report a higher correlation among indusries. This phenomenon may not be explained through the conventional risk-return analysis. We apply a Markov switching model that contains both reversible and irreversible chains and find that the structural change of the market seemed to occur in 2003, well before the crisis.We also analyse bid-ask spreads, one of the liquidity measures, in the credit default swap (CDS) markets. These spreads are determined by the behaviors of the market participants. We successfully fit a self-exciting process combinded with time-changed Brownian motion. Moreover, we investigate the fund manager's optimal stopping problem where the fund is leveraged.