Abstract
It is frequent to deal with parametric models that are difficult to analyze, due to the large number of data and/or parameters, complicated nonlinearities, or unobservable variables. The aim is to explain how to analyze such models by means of a set of simplified models, called instrumental models, and how to combine these instrumental models in an optimal way. In this respect a bridge between the econometric literature on indirect inference and the statistical literature on composite likelihood is provided. The composite indirect inference principle is illustrated by an application to the analysis of corporate risks.
| Original language | English |
|---|---|
| Pages (from-to) | 30-45 |
| Number of pages | 16 |
| Journal | Econometrics and Statistics |
| Volume | 7 |
| DOIs | |
| Publication status | Published - 1 Jul 2018 |
| Externally published | Yes |
Keywords
- Asymptotic single risk factor
- Composite likelihood
- Corporate risk
- Indirect inference
- Instrumental model
- Pseudo maximum likelihood