Semi-Markov Regime Switching Regression Models
AbstractMarkov switching regression processes belong to the class of Hidden Markov models (HMMs). They provide a higher flexibility than, for example, simple (auto)regression. The main reason for their popularity is the convenient interpretability. For sufficiently long time series, the different regimes can be associated with abrupt macroeconomic events (war, changing governmental policy,etc.). However, it is not always intuitively clear why the regime switching follows a Markov law. Hidden semi-Markov models (HSMMs) are an extension of HMMs. The most appealing property of a HSMM lies in the flexibility of the runlength distributions which are given explicitly instead of implicitly following the geometric distributions of a HMM. We present a generalization of the Markov regime switching framework and introduce the semi-Markov switching (auto)regressive processes. In particular, we focus on the theory for right-censored HSMMs introduced by GuÃ©don in 2003. We present an EM algorithm for auto(regression) models with different state occupancy distributions. Subsequently, we investigate a modified, computational convenient M-step in terms of the One-Step-Late. Finally, the performance of the estimation procedure is analyzed using an economic data set
Download InfoTo our knowledge, this item is not available for download. To find whether it is available, there are three options:
1. Check below under "Related research" whether another version of this item is available online.
2. Check on the provider's web page whether it is in fact available.
3. Perform a search for a similarly titled item that would be available.
Bibliographic InfoPaper provided by Society for Computational Economics in its series Computing in Economics and Finance 2006 with number 438.
Date of creation: 04 Jul 2006
Date of revision:
Hidden semi-Markov model; One-Step-Late algorithm; regime switching; regression; right-censoring;
Find related papers by JEL classification:
- C32 - Mathematical and Quantitative Methods - - Multiple or Simultaneous Equation Models; Multiple Variables - - - Time-Series Models; Dynamic Quantile Regressions; Dynamic Treatment Effect Models; Diffusion Processes
- C34 - Mathematical and Quantitative Methods - - Multiple or Simultaneous Equation Models; Multiple Variables - - - Truncated and Censored Models; Switching Regression Models
- C63 - Mathematical and Quantitative Methods - - Mathematical Methods; Programming Models; Mathematical and Simulation Modeling - - - Computational Techniques
You can help add them by filling out this form.
For technical questions regarding this item, or to correct its authors, title, abstract, bibliographic or download information, contact: (Christopher F. Baum).
If references are entirely missing, you can add them using this form.