A predictability test for a small number of nested models

Eleonora Granziera, Kirstin Hubrich, Hyungsik Roger Moon

Research output: Contribution to journalArticle

3 Citations (Scopus)

Abstract

We introduce quasi-likelihood ratio tests for one sided multivariate hypotheses to evaluate the null that a parsimonious model performs equally well as a small number of models which nest the benchmark. The limiting distributions of the test statistics are non-standard. For critical values we consider: (i) bootstrapping and (ii) simulations assuming normality of the mean square prediction error difference. The proposed tests have good size and power properties compared with existing equal and superior predictive ability tests for multiple model comparison. We apply our tests to study the predictive ability of a Phillips curve type for the US core inflation.

Original languageEnglish
Pages (from-to)174-185
Number of pages12
JournalJournal of Econometrics
Volume182
Issue number1
DOIs
Publication statusPublished - 2014 Sep

Fingerprint

Nested Models
Predictability
Phillips Curve
Quasi-likelihood
Multiple Comparisons
Model Comparison
Nest
Bootstrapping
Multiple Models
Likelihood Ratio Test
Prediction Error
Limiting Distribution
Mean square error
Normality
Inflation
Test Statistic
Null
Critical value
Statistics
Benchmark

All Science Journal Classification (ASJC) codes

  • Economics and Econometrics

Cite this

Granziera, Eleonora ; Hubrich, Kirstin ; Moon, Hyungsik Roger. / A predictability test for a small number of nested models. In: Journal of Econometrics. 2014 ; Vol. 182, No. 1. pp. 174-185.
@article{7be3fb8718414049b1585fb67fd4443c,
title = "A predictability test for a small number of nested models",
abstract = "We introduce quasi-likelihood ratio tests for one sided multivariate hypotheses to evaluate the null that a parsimonious model performs equally well as a small number of models which nest the benchmark. The limiting distributions of the test statistics are non-standard. For critical values we consider: (i) bootstrapping and (ii) simulations assuming normality of the mean square prediction error difference. The proposed tests have good size and power properties compared with existing equal and superior predictive ability tests for multiple model comparison. We apply our tests to study the predictive ability of a Phillips curve type for the US core inflation.",
author = "Eleonora Granziera and Kirstin Hubrich and Moon, {Hyungsik Roger}",
year = "2014",
month = "9",
doi = "10.1016/j.jeconom.2014.04.016",
language = "English",
volume = "182",
pages = "174--185",
journal = "Journal of Econometrics",
issn = "0304-4076",
publisher = "Elsevier BV",
number = "1",

}

A predictability test for a small number of nested models. / Granziera, Eleonora; Hubrich, Kirstin; Moon, Hyungsik Roger.

In: Journal of Econometrics, Vol. 182, No. 1, 09.2014, p. 174-185.

Research output: Contribution to journalArticle

TY - JOUR

T1 - A predictability test for a small number of nested models

AU - Granziera, Eleonora

AU - Hubrich, Kirstin

AU - Moon, Hyungsik Roger

PY - 2014/9

Y1 - 2014/9

N2 - We introduce quasi-likelihood ratio tests for one sided multivariate hypotheses to evaluate the null that a parsimonious model performs equally well as a small number of models which nest the benchmark. The limiting distributions of the test statistics are non-standard. For critical values we consider: (i) bootstrapping and (ii) simulations assuming normality of the mean square prediction error difference. The proposed tests have good size and power properties compared with existing equal and superior predictive ability tests for multiple model comparison. We apply our tests to study the predictive ability of a Phillips curve type for the US core inflation.

AB - We introduce quasi-likelihood ratio tests for one sided multivariate hypotheses to evaluate the null that a parsimonious model performs equally well as a small number of models which nest the benchmark. The limiting distributions of the test statistics are non-standard. For critical values we consider: (i) bootstrapping and (ii) simulations assuming normality of the mean square prediction error difference. The proposed tests have good size and power properties compared with existing equal and superior predictive ability tests for multiple model comparison. We apply our tests to study the predictive ability of a Phillips curve type for the US core inflation.

UR - http://www.scopus.com/inward/record.url?scp=84901833571&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84901833571&partnerID=8YFLogxK

U2 - 10.1016/j.jeconom.2014.04.016

DO - 10.1016/j.jeconom.2014.04.016

M3 - Article

AN - SCOPUS:84901833571

VL - 182

SP - 174

EP - 185

JO - Journal of Econometrics

JF - Journal of Econometrics

SN - 0304-4076

IS - 1

ER -