Working Paper: NBER ID: w11758
Authors: Bruce N. Lehmann
Abstract: This paper discusses inference for rational expectations models estimated via minimum distance methods by characterizing the probability beliefs regarding the data generating process (DGP) that are compatible with given moment conditions. The null hypothesis is taken to be rational expectations and the alternative hypothesis to be distorted beliefs. This distorted beliefs alternative is analyzed from the perspective of a hypothetical semiparametric Bayesian who believes the model and uses it to learn about the DGP. This interpretation provides a different perspective on estimates, test statistics, and confidence regions in large samples, particularly regarding the economic significance of rejections in rational expectations models.
Keywords: No keywords provided
JEL Codes: C1; C2; C3; C4; C5
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| rational expectations (D84) | distorted beliefs (D91) |
| econometrician's estimates (C51) | economic agents' beliefs (D84) |
| posterior beliefs (D80) | GMM estimates (C51) |