A framework which allows for the joint testing ofthe adaptive and rational expectations hypotheses is presented. We assume joint normality of expectations, realizations and variablesin the information set, allowing for parsimonious interpretationof the data; conditional first moments are linear in the conditioningvariables, and we can easily recover regression coefficients fromthem and test simple hypotheses by imposing zero restrictionson these coefficients. The nature of the data, which are responsesto business surveys and are all categorical, requires simulationtechniques to obtain full information maximum likelihood estimates. We use a latent variable model which allows for the constructionof a simple likelihood function. However, this likelihood containsmulti- (four)dimensional integrals, requiring simulators to evaluate. Simulated maximum-likelihood estimation is carried out usingthe Geweke-Hajivassilou-Keane (GHK) method, which is consistentand has low variance. The latter is crucial when maximizing thelog-likelihood directly. Identification of the parameters isachieved by placing restrictions on the response thresholds and/orthe variances. We find that we can reject both hypotheses.
Additional information:
Das Dokument wird vom Publikationsserver der Universitätsbibliothek Mannheim bereitgestellt.
![[img]](https://madoc.bib.uni-mannheim.de/style/images/fileicons/application_pdf.png)
Search Authors in