Dynamic quality ladder model predictions in nonrandom holdout samples

Linli Xu, Jorge M. Silva-Risso, Kenneth C. Wilbur

Research output: Contribution to journalArticle

1 Citation (Scopus)

Abstract

In light of recent calls for further validation of structural models, this paper evaluates the popular dynamic quality ladder (DQL) model using a nonrandom holdout approach. The model is used to predict data following a regime shift-that is, a change in the environment that produced the estimation data. The prediction performance is evaluated relative to a benchmark vector autoregression (VAR) model across three automotive categories and multiple prediction horizons. Whereas the VAR model performs better in all scenarios in the compact car category, the DQL model tends to perform better on multiple-year horizons in both the midsize car and full-size pickup categories. A supplementary data analysis suggests that DQL model performance in the nonrandom holdout prediction task is better in categories that are more affected by the regime shift, helping to validate the usefulness of the dynamic structural model for making predictions after policy changes.

Original languageEnglish (US)
Pages (from-to)3187-3207
Number of pages21
JournalManagement Science
Volume64
Issue number7
DOIs
StatePublished - Jul 2018

Fingerprint

Prediction model
Quality ladder
Prediction
Regime shift
Vector autoregression model
Car
Policy change
Usefulness
Dynamic structural model
Structural model
Benchmark
Scenarios

Keywords

  • Automobiles
  • Dynamic oligopoly competition
  • Nonrandom holdout validation
  • Product innovation
  • Product quality

Cite this

Dynamic quality ladder model predictions in nonrandom holdout samples. / Xu, Linli; Silva-Risso, Jorge M.; Wilbur, Kenneth C.

In: Management Science, Vol. 64, No. 7, 07.2018, p. 3187-3207.

Research output: Contribution to journalArticle

Xu, Linli ; Silva-Risso, Jorge M. ; Wilbur, Kenneth C. / Dynamic quality ladder model predictions in nonrandom holdout samples. In: Management Science. 2018 ; Vol. 64, No. 7. pp. 3187-3207.
@article{1b88f188d3df4194a23f3d4713cff190,
title = "Dynamic quality ladder model predictions in nonrandom holdout samples",
abstract = "In light of recent calls for further validation of structural models, this paper evaluates the popular dynamic quality ladder (DQL) model using a nonrandom holdout approach. The model is used to predict data following a regime shift-that is, a change in the environment that produced the estimation data. The prediction performance is evaluated relative to a benchmark vector autoregression (VAR) model across three automotive categories and multiple prediction horizons. Whereas the VAR model performs better in all scenarios in the compact car category, the DQL model tends to perform better on multiple-year horizons in both the midsize car and full-size pickup categories. A supplementary data analysis suggests that DQL model performance in the nonrandom holdout prediction task is better in categories that are more affected by the regime shift, helping to validate the usefulness of the dynamic structural model for making predictions after policy changes.",
keywords = "Automobiles, Dynamic oligopoly competition, Nonrandom holdout validation, Product innovation, Product quality",
author = "Linli Xu and Silva-Risso, {Jorge M.} and Wilbur, {Kenneth C.}",
year = "2018",
month = "7",
doi = "10.1287/mnsc.2017.2780",
language = "English (US)",
volume = "64",
pages = "3187--3207",
journal = "Management Science",
issn = "0025-1909",
publisher = "INFORMS Inst.for Operations Res.and the Management Sciences",
number = "7",

}

TY - JOUR

T1 - Dynamic quality ladder model predictions in nonrandom holdout samples

AU - Xu, Linli

AU - Silva-Risso, Jorge M.

AU - Wilbur, Kenneth C.

PY - 2018/7

Y1 - 2018/7

N2 - In light of recent calls for further validation of structural models, this paper evaluates the popular dynamic quality ladder (DQL) model using a nonrandom holdout approach. The model is used to predict data following a regime shift-that is, a change in the environment that produced the estimation data. The prediction performance is evaluated relative to a benchmark vector autoregression (VAR) model across three automotive categories and multiple prediction horizons. Whereas the VAR model performs better in all scenarios in the compact car category, the DQL model tends to perform better on multiple-year horizons in both the midsize car and full-size pickup categories. A supplementary data analysis suggests that DQL model performance in the nonrandom holdout prediction task is better in categories that are more affected by the regime shift, helping to validate the usefulness of the dynamic structural model for making predictions after policy changes.

AB - In light of recent calls for further validation of structural models, this paper evaluates the popular dynamic quality ladder (DQL) model using a nonrandom holdout approach. The model is used to predict data following a regime shift-that is, a change in the environment that produced the estimation data. The prediction performance is evaluated relative to a benchmark vector autoregression (VAR) model across three automotive categories and multiple prediction horizons. Whereas the VAR model performs better in all scenarios in the compact car category, the DQL model tends to perform better on multiple-year horizons in both the midsize car and full-size pickup categories. A supplementary data analysis suggests that DQL model performance in the nonrandom holdout prediction task is better in categories that are more affected by the regime shift, helping to validate the usefulness of the dynamic structural model for making predictions after policy changes.

KW - Automobiles

KW - Dynamic oligopoly competition

KW - Nonrandom holdout validation

KW - Product innovation

KW - Product quality

UR - http://www.scopus.com/inward/record.url?scp=85049393674&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85049393674&partnerID=8YFLogxK

U2 - 10.1287/mnsc.2017.2780

DO - 10.1287/mnsc.2017.2780

M3 - Article

AN - SCOPUS:85049393674

VL - 64

SP - 3187

EP - 3207

JO - Management Science

JF - Management Science

SN - 0025-1909

IS - 7

ER -