Font Size: a A A

Markov Regime-switching Model And Its Application To Volatility Analysis Of Chinese Inflation

Posted on:2009-02-10Degree:MasterType:Thesis
Country:ChinaCandidate:L F XingFull Text:PDF
GTID:2189360278963717Subject:Quantitative Economics
Abstract/Summary:PDF Full Text Request
If we observe time series for a long time, we can find that sometimes there are some time periods, which act differently from the prophases. And the remarkable change in the time series is sometimes regarded as a switch in the regression equation from one regime to another. Markov-switching model takes the regime-switch as an endogenous variable and a random process, which enable it to describe all the remarkable structural change in one united model and help to forecast. After explaining the meaning and dynamics of contemporary research on Markov regime-switching, this paper introduces the specification, estimation and forecasting of the model. We use Markov regime-switching model with two regimes and two lags to investigate the dynamic path of China inflation rate since 1983. According to the result of estimation and the analysis of filter probabilities and smooth probabilities, we find that there exist two regimes, which are high inflation regime and low inflation regime, and there are different transition probabilities between the two regimes. The duration in each regime is different, also is the degree of fluctuation. At the same time, we also estimate with AR model, and compared to traditional AR model, the LR-test shows Markov regime-switching model describes the data generating process better. At the end, we get one-step ahead forecast probability of low inflation regime and one-step ahead expected value of inflation rate, by which we can get in which regime the next period is, and we also can make corresponding policy on the basis of the forecasting.
Keywords/Search Tags:Markov regime-switching model, filter probability, smooth probability, inflation rate, model forecast
PDF Full Text Request
Related items