Indirect inference with a non-smooth criterion function

Research output: Contribution to journalArticleResearchpeer-review

Abstract

Indirect inference requires simulating realizations of endogenous variables from the model under study. When the endogenous variables are discontinuous functions of the model parameters, the resulting indirect inference criterion function is discontinuous and does not permit the use of derivative-based optimization routines. Using a change of variables technique, we propose a novel simulation algorithm that alleviates the discontinuities inherent in such indirect inference criterion functions, and permits the application of derivative-based optimization routines to estimate the unknown model parameters. Unlike competing approaches, this approach does not rely on kernel smoothing or bandwidth parameters. Several Monte Carlo examples that have featured in the literature on indirect inference with discontinuous outcomes illustrate the approach, and demonstrate the superior performance of this approach over existing alternatives.

Original languageEnglish
Number of pages23
JournalJournal of Econometrics
DOIs
Publication statusAccepted/In press - 2019

Keywords

  • Discontinuous objective functions
  • Dynamic discrete choice models
  • Indirect inference
  • Simulation estimators

Cite this

@article{646fbaed9f9c41958e204b64cff7377f,
title = "Indirect inference with a non-smooth criterion function",
abstract = "Indirect inference requires simulating realizations of endogenous variables from the model under study. When the endogenous variables are discontinuous functions of the model parameters, the resulting indirect inference criterion function is discontinuous and does not permit the use of derivative-based optimization routines. Using a change of variables technique, we propose a novel simulation algorithm that alleviates the discontinuities inherent in such indirect inference criterion functions, and permits the application of derivative-based optimization routines to estimate the unknown model parameters. Unlike competing approaches, this approach does not rely on kernel smoothing or bandwidth parameters. Several Monte Carlo examples that have featured in the literature on indirect inference with discontinuous outcomes illustrate the approach, and demonstrate the superior performance of this approach over existing alternatives.",
keywords = "Discontinuous objective functions, Dynamic discrete choice models, Indirect inference, Simulation estimators",
author = "Frazier, {David T.} and Tatsushi Oka and Dan Zhu",
year = "2019",
doi = "10.1016/j.jeconom.2019.06.003",
language = "English",
journal = "Journal of Econometrics",
issn = "0304-4076",
publisher = "Elsevier",

}

Indirect inference with a non-smooth criterion function. / Frazier, David T.; Oka, Tatsushi; Zhu, Dan.

In: Journal of Econometrics, 2019.

Research output: Contribution to journalArticleResearchpeer-review

TY - JOUR

T1 - Indirect inference with a non-smooth criterion function

AU - Frazier, David T.

AU - Oka, Tatsushi

AU - Zhu, Dan

PY - 2019

Y1 - 2019

N2 - Indirect inference requires simulating realizations of endogenous variables from the model under study. When the endogenous variables are discontinuous functions of the model parameters, the resulting indirect inference criterion function is discontinuous and does not permit the use of derivative-based optimization routines. Using a change of variables technique, we propose a novel simulation algorithm that alleviates the discontinuities inherent in such indirect inference criterion functions, and permits the application of derivative-based optimization routines to estimate the unknown model parameters. Unlike competing approaches, this approach does not rely on kernel smoothing or bandwidth parameters. Several Monte Carlo examples that have featured in the literature on indirect inference with discontinuous outcomes illustrate the approach, and demonstrate the superior performance of this approach over existing alternatives.

AB - Indirect inference requires simulating realizations of endogenous variables from the model under study. When the endogenous variables are discontinuous functions of the model parameters, the resulting indirect inference criterion function is discontinuous and does not permit the use of derivative-based optimization routines. Using a change of variables technique, we propose a novel simulation algorithm that alleviates the discontinuities inherent in such indirect inference criterion functions, and permits the application of derivative-based optimization routines to estimate the unknown model parameters. Unlike competing approaches, this approach does not rely on kernel smoothing or bandwidth parameters. Several Monte Carlo examples that have featured in the literature on indirect inference with discontinuous outcomes illustrate the approach, and demonstrate the superior performance of this approach over existing alternatives.

KW - Discontinuous objective functions

KW - Dynamic discrete choice models

KW - Indirect inference

KW - Simulation estimators

UR - http://www.scopus.com/inward/record.url?scp=85068505056&partnerID=8YFLogxK

U2 - 10.1016/j.jeconom.2019.06.003

DO - 10.1016/j.jeconom.2019.06.003

M3 - Article

JO - Journal of Econometrics

JF - Journal of Econometrics

SN - 0304-4076

ER -