Unité mixte de recherche 7235

Reject inference in application scorecards: evidence from France

Ha Thu Nguyen

[en]Credit scoring models are commonly developed using only accepted Known Good/Bad (G/B) applications, called KGB model, because we only know the performance of those accepted in the past. Obviously, the KGB model is not indicative of the entire through-the-door population, and reject inference precisely attempts to address the bias by assigning an inferred G/B status to rejected applications. In this paper, we discuss the pros and cons of various reject inference techniques, and pitfalls to avoid when using them. We consider a real dataset of a major French consumer finance bank to assess the effectiveness of the practice of using reject inference. To do that, we rely on the logistic regression framework to model probabilities to become good/bad, and then validate the model performance with and without sample selection bias correction. Our main results can be summarized as follows. First, we show that the best reject inference technique is not necessarily the most complicated one: reweighting and parceling provide more accurate and relevant results than fuzzy augmentation and Heckman’s two-stage correction. Second, disregarding rejected applications significantly impacts the forecast accuracy of the scorecard. Third, as the sum of standard errors dramatically reduces when the sample size increases, reject inference turns out to produce an improved representation of the population. Finally, reject inference appears to be an effective way to reduce overfitting in model selection.[/en]

AGENDA

lundi 5 décembre 2022

Law, Institutions and Economics in Nanterre (LIEN)

Vincent Lefrère (Institut Mines Telecom)

En salle 614 et en distanciel

Privacy, Data and Competition: The Case of Apps For Young Children

mardi 6 décembre 2022

Recherche et Economie et Socioéconomie Politique, des Institutions et des Régulations (RESPIR)

Eric Monnet (EHESS et PSE)

Le rôle d’amortisseur des banques centrales – une perspective historique

jeudi 8 décembre 2022

Lunch

Georges Prat

12h - 13h, salle 110

Modeling ex-ante risk premiums in the oil market

lundi 12 décembre 2022

Law, Institutions and Economics in Nanterre (LIEN)

Clément Brébion (Copenhagen BS)

En salle 614 et en distanciel

Unemployment Insurance Eligibility and Employment Duration

mardi 13 décembre 2022

Développement Durable Environnement et Energie (DDEE)

Nicolas Astier (Paris School of Economics)

16h-17h

Riding together: eliciting travelers’ preferences for long-distance carpooling

mercredi 14 décembre 2022

Économies du monde musulman

Mohamed Touati Tliba (École Supérieure de Commerce, Alger)

The scientific wealth of nations with special reference to MENA region: a cross-country productivity analysis of academic research

jeudi 15 décembre 2022

Doctorants

Pablo Aguilar Perez

TBA

jeudi 15 décembre 2022

Groupe de travail « Intelligence artificielle »

Matthieu Lapaty (UPMC)

TBA

Inscription aux Newsletters