Forecasting recovery rates on non-performing loans with machine learning

Anthony Bellotti, Damiano Brigo, Paolo Gambetti, Frédéric Vrins

Research output: Journal PublicationArticlepeer-review

32 Citations (Scopus)

Abstract

We compare the performance of a wide set of regression techniques and machine-learning algorithms for predicting recovery rates on non-performing loans, using a private database from a European debt collection agency. We find that rule-based algorithms such as Cubist, boosted trees, and random forests perform significantly better than other approaches. In addition to loan contract specificities, predictors that refer to the bank recovery process — prior to the portfolio's sale to a debt collector — are also shown to enhance forecasting performance. These variables, derived from the time series of contacts to defaulted clients and client reimbursements to the bank, help all algorithms better identify debtors with different repayment ability and/or commitment, and in general those with different recovery potential.

Original languageEnglish
Pages (from-to)428-444
Number of pages17
JournalInternational Journal of Forecasting
Volume37
Issue number1
DOIs
Publication statusPublished - 1 Jan 2021

Keywords

  • Credit risk
  • Debt collection
  • Defaulted loans
  • Loss given default
  • Superior set of models

ASJC Scopus subject areas

  • Business and International Management

Fingerprint

Dive into the research topics of 'Forecasting recovery rates on non-performing loans with machine learning'. Together they form a unique fingerprint.

Cite this