Abstract
This paper focuses on finding starting-values for the estimation of Vector STAR models. Based on a Monte Carlo study, different procedures are evaluated. Their performance is assessed with respect to model fit and computational effort. I employ (i) grid search algorithms and (ii) heuristic optimization procedures, namely differential evolution, threshold accepting, and simulated annealing. In the equation-by-equation starting-value search approach the procedures achieve equally good results. Unless the errors are cross-correlated, equation-by-equation search followed by a derivative-based algorithm can handle such an optimization problem sufficiently well. This result holds also for higher-dimensional Vector STAR models with a slight edge for heuristic methods. For more complex Vector STAR models which require a multivariate search approach, simulated annealing and differential evolution outperform threshold accepting and the grid search.
Highlights
Whatever the use of an econometric model, estimating its parameters as well as possible given the available information is of crucial importance
The present paper focuses on finding starting-values for the estimation of Vector STAR
I begin by presenting the results of different starting-value search methods of the equation-by-equation approach in Subsection 5.1
Summary
Whatever the use of an econometric model, estimating its parameters as well as possible given the available information is of crucial importance. The present paper focuses on finding starting-values for the estimation of Vector STAR models to solve this complex problem. The inherent stochastics and (controlled) impairments of the objective function of heuristic optimization procedures may deliver advantages in terms of (i) the extent to which the surface area can be explored and (ii) the optimization outcome (higher likelihood) This is important when equation-by-equation Non-linear Least Squares estimation is inefficient or not feasible at all and a system-wide Maximum Likelihood estimation is necessary. Before applying a heuristic algorithm to the whole modeling cycle of a non-linear Vector STAR model, one might initially focus on improving the starting-value search. As soon as the Vector STAR model has cross-correlated error terms, multivariate starting-value search procedures, outperform equation-by-equation approaches.
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.
