中文摘要：

关于金融市场建模,经典研究大多是基于布朗运动和跳跃扩散过程,但由于市场中存在着大量的不确定性因素,因此在实际中很难建立精确的统计模型。文章在统计信息不足情形下,以市场“异象”为切入点,充分学习市场“异象”特征,根据动量效应和反转效应构造体现不同特征的基准策略转换概率,设计出具有动量特征或反转特征的新的Switching交易策略。给出并证明了具有动量效应特征的策略的收益下界。通过市场历史数据进行实证分析,并与其他策略进行对比、分析。

英文摘要：

In the area of financial market modeling, classic studies mostly base on Brownian motion and jump diffusion pro- cess, but there are lots of uncertainty factors in the market which increase difficulty to establish accurate statistical model in prac- tice. This paper proposes two new switching portfolio strategies with momentum and reversal characters, respectively. To be specif- ic, in the situation of the lacking of statistical information, this article design different basic strategies＇ switching probability.