Testing nonnested Euler conditions with quadrature-based methods of approximation (Q805126)

For testing the Euler conditions for investigating intertemporal optima, two methods are currently applied: One incorporates information about alternative Euler equation specifications implied by competing economic models as is done in the generalized method of moments (GMM), and the second method constructs a sequence of local alternatives by perturbing the data-generation process of the maintained model in the direction of the alternative model. This paper derives another test statistic based on ``Cox's encompassing principle'' which can be used to discriminate between two nonnested sets of moment conditions. The test depends explicitly on the data generating process of the maintained model and the computation uses quadrature-based methods of approximation.