TIME-VARYING TRANSITION PROBABILITIES FOR MARKOV REGIME SWITCHING MODELS

被引：56

作者：

Bazzi, Marco ^{[1
]}

Blasques, Francisco ^{[2
,3
]}

Koopman, Siem Jan ^{[2
,3
,4
]}

Lucas, Andre ^{[2
,3
]}

机构：

[1] Univ Padua, Padua, Italy

[2] Vrije Univ Amsterdam, Amsterdam, Netherlands

[3] Tinbergen Inst, Amsterdam, Netherlands

[4] Aarhus Univ, CREATES, Aarhus, Denmark

来源：

JOURNAL OF TIME SERIES ANALYSIS | 2017年 / 38卷 / 03期

基金：

新加坡国家研究基金会;

关键词：

Hidden Markov models; observation driven models; time varying parameter; BUSINESS-CYCLE; SERIES MODELS; CONDITIONAL HETEROSKEDASTICITY; STOCK RETURNS; STATIONARITY; ERGODICITY; DRIVEN; RISK;

D O I：

10.1111/jtsa.12211

中图分类号：

O1 [数学];

学科分类号：

0701 ; 070101 ;

摘要：

We propose a new Markov switching model with time-varying transitions probabilities. The novelty of our model is that the transition probabilities evolve over time by means of an observation driven model. The innovation of the time-varying probability is generated by the score of the predictive likelihood function. We show how the model dynamics can be readily interpreted. We investigate the performance of the model in a Monte Carlo study and show that the model is successful in estimating a range of different dynamic patterns for unobserved regime switching probabilities. We also illustrate the new methodology in an empirical setting by studying the dynamic mean and variance behaviour of US industrial production growth.

引用

页码：458 / 478

页数：21

共 42 条

[1] MAXIMUM LIKELIHOOD IDENTIFICATION OF GAUSSIAN AUTOREGRESSIVE MOVING AVERAGE MODELS [J].