Duration Dependence Of The Business Cycle In China

In 2008, the global financial crisis triggered by US subprime mortgage let us realize again that the economy of a country or region will never keep booming forever. Look at the history of economic development of countries in the world, each country's economic growth has always presented alternating volatility characteristics of expansion and contraction.Scholars in China take little account of the duration dependence characteristics in the economic cycle, whereas, foreign scholars have confirmed that the duration dependence characteristics is one of the important experience features in business cycle. In short, duration dependence is defined as follows; the termination probability staying one state of an event depends its elapsed time. As for economic activity, the termination probability of an expansion regime or contraction regime depends the age of the current state. As we all known, an economic cycle fluctuation never and will never last forever. So an obvious question is that whether the termination probability of a regime or the whole economic cycle will change as the time goes by. Specifically speaking, positive duration dependence means that the longer the economy remained in one state, the more likely it was to change to the other; negative duration dependence represents that an economic cycle state will tend to last forever.In this paper, we extend Markov-switching model by Hamilton and put the duration dependence characteristic of the economic cycle fluctuations into the model, thus forming Duration dependence Markov-switching VAR model (from now on DDMS-VAR). By using the macro-monthly data of the four coincident indicators from January 1980 to May 2010 and Bayesian Gibbs sampling algorithm to estimate the model parameters, we confirm whether China business cycle existing duration dependence through comparative analysis. Since the reform and opening up, the empirical evidence shows that, the results support the hypothesis of two regimes in Chinese economy:i.e. "low growth regime" and "high growth regime". There are 10 low growth regime intervals and the average duration was 14.88 months. The economic cycle fluctuation shows unobvious positive duration dependence characteristic. Otherwise, there are 11 high growth regime intervals and the average duration was 20.08 months. Evidence is presented suggesting that high growth regime exists significant positive duration dependence and the transition probability close to 0.75 when the economy remained in high growth regime for 60 months.It is significant to find this experience feature for our government in issuing some macro-economic control policies and maintaining economy development with sustained, stable and high-speed. It is mainly reflected in the counter-cycle policies and the authority could make more accurate judgments in issuing policies time. It is very important to decide the time of policy implementation considering the time delay effect of policy. Given that it has different duration dependence characteristic in different regimes, the government issued the policies in force should be distinguished. We need take stronger policy force in low-speed growth regime and weaker policy force respectively in high-speed growth regime. So we can balance economic development and maintain social progress harmoniously.This paper has two main innovations:(1) we extend the popular Markov-switching VAR model to research the duration dependence characteristics in china business cycle fluctuation and estimate the model parameters with Bayesian Gibbs sampling algorithm. To some extend, this method can help improve the shortage of lacking macro-monthly data in China. What's more, we can make full use of sample information and prior information of model parameters to assure less variance of estimators and get more precise predictions. (2) We not only give an Empirical Investigation to the duration dependence characteristics in China economic cycle but also find the changes of parameters and regime characteristics by comparative analysis when considering or not considering the duration dependence characteristics, thus we can find out accurately the influence of the duration dependence characteristics to the business cycle fluctuation in China.