1. A Comparison between MS-VECM and MS-VECMX on Economic Time Series Data.
- Author
-
Seuk-Wai Phoong, Ismail, Mohd Tahir, and Siok-Kun Sek
- Subjects
- *
MARKOV processes , *VECTOR error-correction models , *TIME series analysis , *MULTIVARIATE analysis , *SWITCHING theory , *STOCK exchanges , *PETROLEUM product sales & prices - Abstract
Multivariate Markov switching models able to provide useful information on the study of structural change data since the regime switching model can analyze the time varying data and capture the mean and variance in the series of dependence structure. This paper will investigates the oil price and gold price effects on Malaysia, Singapore, Thailand and Indonesia stock market returns. Two forms of Multivariate Markov switching models are used namely the mean adjusted heteroskedasticity Markov Switching Vector Error Correction Model (MSMH-VECM) and the mean adjusted heteroskedasticity Markov Switching Vector Error Correction Model with exogenous variable (MSMH-VECMX). The reason for using these two models are to capture the transition probabilities of the data since real financial time series data always exhibit nonlinear properties such as regime switching, cointegrating relations, jumps or breaks passing the time. A comparison between these two models indicates that MSMH-VECM model able to fit the time series data better than the MSMH-VECMX model. In addition, it was found that oil price and gold price affected the stock market changes in the four selected countries. [ABSTRACT FROM AUTHOR]
- Published
- 2014
- Full Text
- View/download PDF