Consistent model-specification tests based on parametric bootstrap

by    A. Leucht

Preprint series: 10-07, Reports on Stochastics and Statistics

A. Leucht

62G10 Hypothesis testing
62E20 Asymptotic distribution theory
62F40 Bootstrap, jackknife and other resampling methods

Abstract: In this paper we establish consistent tests of L2-type for the parametric functional form of the conditional mean of time series with values in $R^d$. A recent result on asymptotic distributions of U-statistics of weakly dependent observations is invoked to obtain the limit distributions of the test statistics. Since the asymptotic distributions depend on unknown parameters in a complicated way, we suggest to apply certain parametric bootstrap methods in order to determine critical values of the tests.

Keywords: Bootstrap; model-specification; U-statistics; time series.

Upload: 2010-08-11

The author(s) agree, that this abstract may be stored as full text and distributed as such by abstracting services.