Meta-Learning Approaches for Recovery Rate Prediction

33 Pages Posted: 26 Mar 2022

See all articles by Francesco Roccazzella

Francesco Roccazzella

IESEG School of Management

Paolo Gambetti

Louvain Finance Center (LFIN), UCLouvain

Frédéric D. Vrins

LFIN/LIDAM, UCLouvain

Abstract

While previous academic research highlights the potential of machine learning and big data for predicting corporate bond recovery rates, the operations management challenge is to identify the relevant predictive variables and the appropriate model. In this paper, we use meta-learning to combine the predictions from 20 candidate linear, nonlinear and rule-based algorithms and we exploit a data set of predictors including security-specific factors, macro-financial indicators and measures of economic uncertainty. We find that the most promising approach consists of models combinations trained on security-specific characteristics and a limited number of well-identified theoretically sound recovery rate determinants, including uncertainty measures. Our research provides useful indications for practitioners and regulators targeting more reliable risk measures in designing micro and macro-prudential policies.

Keywords: Finance, Forecasting, Credit Risk, Machine Learning, Recovery rate

Suggested Citation

Roccazzella, Francesco and Gambetti, Paolo and Vrins, Frederic Daniel, Meta-Learning Approaches for Recovery Rate Prediction. Available at SSRN: https://ssrn.com/abstract=4067066 or http://dx.doi.org/10.2139/ssrn.4067066

Francesco Roccazzella (Contact Author)

IESEG School of Management ( email )

Paolo Gambetti

Louvain Finance Center (LFIN), UCLouvain ( email )

34 Voie du Roman Pays - L1.03.01
Louvain-la-Neuve, 1348
Belgium

Do you have negative results from your research you’d like to share?

Paper statistics

Downloads
38
Abstract Views
188
PlumX Metrics