Testing a parametric function against a non‐parametric alternative in IV and GMM settings |
| |
Authors: | Tue Gørgens Allan Würtz |
| |
Affiliation: | 1. Research School of Economics, The Australian National University, Canberra ACT 0200, Australia. E‐mail: tue.gorgens@anu.edu.au;2. Department of Economics and Business, Aarhus University, Funglesangs ALLé 4, DK‐8210 Aarhus V, Denmark. E‐mail: awurtz@econ.au.dk |
| |
Abstract: | Summary This paper develops a specification test for functional form for models identified by a conditional moment restriction, including IV and GMM settings. The framework is one where the moment restriction is specified as a function of data, a finite‐dimensional parameter vector and a non‐parametric function (an infinite‐dimensional parameter vector). The null hypothesis is that the moment restriction does not depend on the non‐parametric function. The test is relatively easy to implement and its asymptotic distribution is known. The test performs well in simulation experiments. |
| |
Keywords: | Generalized arc‐sine distribution Generalized method of moments LM statistic Non‐parametric alternative Specification test |
|
|