Stochastic-Simulation Tests of Nonlinear Econometric Models
Stochastic simulations of nonlinear dynamic econometric models have been used in various ways in the past. This chapter discusses how stochastic simulations can be exploited to develop appropriate system-specification tests for nonlinear systems. The approach is through auxiliary regressions of stochastic simulation errors to develop asymptotically valid significance tests of the predictive performance of the model. The first section discusses Adrian Pagan's critique of the use of simulations in testing nonlinear models for misspecification. The related issue of the informational content of multi-period-ahead predictions is also analyzed in this section. The stochastic simulations that it uses to form the prediction-based tests and their basic asymptotic properties are reviewed in the second section. The last section then develops the auxiliary regressions leading to our prediction-based tests.
Oxford Scholarship Online requires a subscription or purchase to access the full text of books within the service. Public users can however freely search the site and view the abstracts and keywords for each book and chapter.
If you think you should have access to this title, please contact your librarian.