中文摘要：

通过在货币政策非对称性理论基础上，构建包含信贷渠道、产出与价格变量的马尔科夫区制变换向量自回归模型（MS-VAR），检验1998年1月至2016年3月中国货币政策信贷渠道的周期时变效应的结果表明，在样本时间区间内货币政策信贷渠道存在通畅期和阻滞期两区制状态，MS-VAR模型能够有效捕捉信贷渠道周期变换时点；而分区制广义脉冲响应函数的结果显示，信贷渠道通畅期货币政策产出与价格效应明显强于阻滞期。央行应将传导渠道非线性特征纳入货币政策决策考虑范围之内，增强对传导渠道状态转变时点的敏感度，因时制宜确定政策方向和力度。

英文摘要：

Based on the theory of monetary policy asymmetric, this paper constructs a Markov -Switching Vector Auto Regression model （MS-VAR） which concludes credit chan- nel, output and price variables to test periodic time -varying effect of China＇ s monetary pol- icy through credit channel between January 1998 and March 2016. Conclusions can be drawn that, during sample time, the effect of monetary policy through credit channel differs in unobstructed and block system state periodically. MS-AR model can effectively capture the breaking point of periodic transformation of credit channel. State -dependence GIRF re- veals that, monetary policy through credit channel has better output price effect in expedite economic environment than in the obstructed one. According to the results above, central banks should take nonlinearity of transmission channel of monetary policy decisions into consideration, and advance the capability of seizing the turning point of transmission mecha- nism in order to determine the direction and strength of government policies precisely.