Abstract
The concept of cointegration is widely used in applied non-stationary time series analysis to describe the co-movement of data measured over time. In this paper, we proposed a Bayesian model for cointegration test and analysis, based on the dynamic latent factor framework. Efficient computational algorithms are also developed based on Markov Chain Monte Carlo (MCMC). Performance and efficiency of the the model and approaches are assessed by simulated and real data analysis.
Highlights
Many macroeconomic and financial time series are nonstationary [1], characterized by the existence of stochastic trends or unit roots
The dynamic latent factor model has been proposed for time series analysis for decades, the idea of incorporation of dynamic structures that can accommodate possible non-stationarity flexibly enough via the Bayesian modeling framework and computational strategies, which we proposed as an alternative Bayesian cointegration test, is novel
The matrix state-space form model (2) is primarily used in deriving efficient computational strategies for sampling latent factors as shown. This dynamic latent factor model has been applied for time series analysis for decades, our idea of the incorporation of dynamic structures that can accommodate possible non-stationarity flexibly enough via the Bayesian modeling framework and computational strategies, which we proposed as an alternative Bayesian cointegration test, is novel and shown
Summary
Many macroeconomic and financial time series are nonstationary [1], characterized by the existence of stochastic trends or unit roots. As a more general cointegration test method than the other two, is a complicated model with many degrees of freedom It has to model all the variables at the same time without a clear and straightforward interpretation in terms of exogenous and endogenous variables, which are all less favorable with multivariate cases especially if the relation for some variable is flawed. We propose the framework of Bayesian factorized cointegration analysis as a straightforward and computationally efficient approach for cointegration test and modeling. The dynamic latent factor model has been proposed for time series analysis for decades, the idea of incorporation of dynamic structures that can accommodate possible non-stationarity flexibly enough via the Bayesian modeling framework and computational strategies, which we proposed as an alternative Bayesian cointegration test, is novel.
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
