关键词: 混频数据, 动态因子模型, 马尔科夫区制转换, 经济周期

Abstract: This paper integrates Markov-Switching dynamics into the intercept and variance of the mixed frequency dynamic factor model. A detailed study is conducted to analyze the characteristics of the business cycles during different economic periods in China from both perspectives of the n economic growth rate and economic fluctuation. It shows that Chinese business cycles have distinct phase characteristics. There is an asymmetry in the transition probability between different states. In the context of the new normal economy, there is a high probability to keep the economy running smooth with a narrowing volatility. Besides, there is a basic consistency existing between the coincident index extracted in this paper and the quarterly GDP growth rate, which could reflect the economic conditions in different phases more accurately.

Key words: Mixed Frequency Data, Dynamic Factor Model, Markov Regime Switching, Business Cycle