nach oben

Journal of Financial Services Research

Erschienen in:

01.10.2014

Ensemble Predictions of Recovery Rates

verfasst von: João A. Bastos

Erschienen in: Journal of Financial Services Research | Ausgabe 2/2014

Einloggen

Aktivieren Sie unsere intelligente Suche, um passende Fachinhalte oder Patente zu finden.

search-config

KI-gestützte Suche

Aus

Abstract

In many domains, the combined opinion of a committee of experts provides better decisions than the judgment of a single expert. This paper shows how to implement a successful ensemble strategy for predicting recovery rates on defaulted debts. Using data from Moody’s Ultimate Recovery Database, it is shown that committees of models derived from the same regression method present better forecasts of recovery rates than a single model. More accurate predictions are observed whether we forecast bond or loan recoveries, and across the entire range of actual recovery values.

Vorheriger Artikel Why Do Banks Provide Leasing?

Nächster Artikel The Effect of Bank-firm Relationships on Sell-side Research

Sie haben noch keine Lizenz? Dann Informieren Sie sich jetzt über unsere Produkte:

Springer Professional "Wirtschaft"

Online-Abonnement

Mit Springer Professional "Wirtschaft" erhalten Sie Zugriff auf:

über 67.000 Bücher
über 340 Zeitschriften

aus folgenden Fachgebieten:

Bauwesen + Immobilien
Business IT + Informatik
Finance + Banking
Management + Führung
Marketing + Vertrieb
Versicherung + Risiko

Jetzt Wissensvorsprung sichern!

Jetzt informieren

Springer Professional "Wirtschaft+Technik"

Online-Abonnement

Mit Springer Professional "Wirtschaft+Technik" erhalten Sie Zugriff auf:

über 102.000 Bücher
über 537 Zeitschriften

aus folgenden Fachgebieten:

Automobil + Motoren
Bauwesen + Immobilien
Business IT + Informatik
Elektrotechnik + Elektronik
Energie + Nachhaltigkeit
Finance + Banking
Management + Führung
Marketing + Vertrieb
Maschinenbau + Werkstoffe
Versicherung + Risiko

Jetzt Wissensvorsprung sichern!

Jetzt informieren

For a discussion on the importance of the negative correlation between default probabilities and recovery rates see, e.g. Das (2007).

The optimal lower limit on the number of observations in the nodes may be very low, even when the model precision is estimated out-of-sample to prevent over-fitting the data. This generates highly complex trees when the sample size is large. For Moody’s URD, I found an optimal lower limit of just 2 observations per node, which segments the data into 170 distinct regions.

Out-of-sample estimates of errors on unseen data are obtained through 10-fold cross-validation. In this procedure, the data are divided into 10 groups of approximately the same size. Nine groups are used for estimation and one group is used for evaluating an out-of-sample error. Each of the 10 groups is in turn set aside to serve temporarily as an independent test sample. Then, the out-of-sample errors of the 10 samples are combined to obtain an estimate of the error on unseen data using all available observations.

The bootstrap samples leave many observations out. For reasonably large data sets such as Moody’s URD, on average 36.8 % of the observations in the bootstrap samples are duplicates.

Acharya VA, Bharath ST, Srinivasan A (2007) Does industry-wide distress affect defaulted firms? Evidence from creditor recoveries. J Financ Econ 85:787–821CrossRef

Basel Committee on Banking Supervision (2006) International convergence of capital measurement and capital standards. Bank for International Settlements

Bastos JA (2010) Forecasting bank loans loss-given-default. J Bank Financ 34:2510–2517CrossRef

Breiman L, Friedman JH, Olshen RA, Stone CJ (1984) Classification and regression trees. Wadworth International Group, Belmont, California

Breiman L (1996) Bagging predictors. Mach Learn 24:123–140

Caselli S, Gatti S, Querci F (2008) The sensitivity of the loss given default rate to systematic risk: new empirical evidence on bank loans. J Financ Serv Res 34:1–34CrossRef

Das SR (2007) Basel II: correlation related issues. J Financ Serv Res 32:17–38CrossRef

Dermine J, Neto de Carvalho C (2006) Bank loan losses-given-default: a case study. J Bank Financ 30:1291–1243CrossRef

Emery K, Cantor R, Keisman D, Ou S (2007) Moody’s Ultimate Recovery Database. Moody’s Investors Service

Grunert J, Weber M (2009) Recovery rates of commercial lending: empirical evidence for German companies. J Bank Financ 33:505–513CrossRef

Gupton GM, Stein RM (2005) LossCalc V2: dynamic prediction of LGD. Moody’s Investors Service

Loterman G, Brown I, Martens D, Mues C, Baesens B (2012) Benchmarking regression algorithms for loss given default modeling. Int J Forecast 28:161–170CrossRef

Papke LE, Wooldridge JM (1996) Econometric methods for fractional response variables with an application to 401(K) plan participation rates. J Appl Econ 11:619–632CrossRef

Varma P, Cantor R (2005) Determinants of recovery rates on defaulted bonds and loans for North American corporate issuers: 1983–2003. J Fixed Income 14:29–44CrossRef

Verikas A, Kalsyte Z, Bacauskiene M, Gelzinis A (2010) Hybrid and ensemble-based soft computing techniques in bankruptcy prediction: a survey. Soft Comput 14:995–1010CrossRef

Witten IH, Frank E (2005) Data mining: practical machine learning tools and techniques. Morgan Kaufmann Publishers

Titel: Ensemble Predictions of Recovery Rates
verfasst von: João A. Bastos
Publikationsdatum: 01.10.2014
Verlag: Springer US
Erschienen in: Journal of Financial Services Research / Ausgabe 2/2014
Print ISSN: 0920-8550
Elektronische ISSN: 1573-0735
DOI: https://doi.org/10.1007/s10693-013-0165-3