Modeling Frailty Correlated Defaults with Multivariate Latent Factors

Benjamin Christoffersen; Rastin Matin

Modeling Frailty Correlated Defaults with Multivariate Latent Factors

Benjamin Christoffersen, Rastin Matin

Institut for Finansiering

Danmarks Nationalbank

Publikation: Working paper › Forskning

Abstract

Konkursmodeller på virksomhedsniveau er vigtige for bottom-up-modellering af en virksomhedsporteføljes konkursrisiko. Modeller i litteraturen har dog typisk flere strenge antagelser, der kan resultere i bias, særligt en antagelse om en lineær effekt af kovariaterne på log-hazard-skalaen, ingen vekselvirkninger, og en antagelse om en enkel additiv latent faktor på log-hazard-skalaen. Med udgangspunkt i et datasæt bestående af amerikanske virksomheder finder vi evidens for, at disse antagelser er for strenge, at de har en betydning i praksis og, endnu vigtigere, at der er evidens for en tidsvarierende effekt af den relative virksomhedsstørrelse. Vi foreslår en frailty-model til at tage højde for disse effekter, der også kan anvendes til at forecaste for
arbitrære porteføljer. Blandt de undersøgte modeller udviser vores foreslåede frailty-model den bedste out-of-sample-rangering af virksomheder efter deres kreditrisiko og de bedste forecasts af konkursraten for hele økonomien under den seneste globale finansielle krise.

Originalsprog	Engelsk
Udgivelsessted	København
Udgiver	Danmarks Nationalbank
Antal sider	24
Status	Udgivet - 22 jan. 2020

Navn	Danmarks Nationalbank. Working Papers
Nummer	151
ISSN	1602-1193

Emneord

Credit risk
Risk management

Adgang til dokumentet

http://www.nationalbanken.dk/da/publikationer/Documents/2020/01/Working%20paper%20nr.%20151.pdfLicens: Ikke-specificeret

Handle.net

Persistent link

Citationsformater

@techreport{bee24d6cc2a64941b377afc1c5572208,

title = "Modeling Frailty Correlated Defaults with Multivariate Latent Factors",

abstract = "Firm-level default models are important for bottomup modeling of the default risk of corporate debt portfolios. However, models in the literature typically have several strict assumptions which may yield biased results, notably a linear effect of covariates on the log-hazard scale, no interactions, and the assumption of a single additive latent factor on the log-hazard scale. Using a sample of US corporate firms, we provide evidence that these assumptions are too strict and matter in practice and, most importantly, we provide evidence of a time-varying effect of the relative firm size. We propose a frailty model to account for such effects that can provide forecasts for arbitrary portfolios as well. Our proposed model displays superior out-of-sample ranking of firms by their default risk and forecasts of the industry-wide default rate during the recent global financial crisis.",

keywords = "Credit risk, Risk management, Credit risk, Risk management",

author = "Benjamin Christoffersen and Rastin Matin",

year = "2020",

month = jan,

day = "22",

language = "English",

series = "Danmarks Nationalbank. Working Papers",

number = "151",

publisher = "Danmarks Nationalbank",

address = "Denmark",

type = "WorkingPaper",

institution = "Danmarks Nationalbank",

}

TY - UNPB

T1 - Modeling Frailty Correlated Defaults with Multivariate Latent Factors

AU - Christoffersen, Benjamin

AU - Matin, Rastin

PY - 2020/1/22

Y1 - 2020/1/22

N2 - Firm-level default models are important for bottomup modeling of the default risk of corporate debt portfolios. However, models in the literature typically have several strict assumptions which may yield biased results, notably a linear effect of covariates on the log-hazard scale, no interactions, and the assumption of a single additive latent factor on the log-hazard scale. Using a sample of US corporate firms, we provide evidence that these assumptions are too strict and matter in practice and, most importantly, we provide evidence of a time-varying effect of the relative firm size. We propose a frailty model to account for such effects that can provide forecasts for arbitrary portfolios as well. Our proposed model displays superior out-of-sample ranking of firms by their default risk and forecasts of the industry-wide default rate during the recent global financial crisis.

AB - Firm-level default models are important for bottomup modeling of the default risk of corporate debt portfolios. However, models in the literature typically have several strict assumptions which may yield biased results, notably a linear effect of covariates on the log-hazard scale, no interactions, and the assumption of a single additive latent factor on the log-hazard scale. Using a sample of US corporate firms, we provide evidence that these assumptions are too strict and matter in practice and, most importantly, we provide evidence of a time-varying effect of the relative firm size. We propose a frailty model to account for such effects that can provide forecasts for arbitrary portfolios as well. Our proposed model displays superior out-of-sample ranking of firms by their default risk and forecasts of the industry-wide default rate during the recent global financial crisis.

KW - Credit risk

KW - Risk management

KW - Credit risk

KW - Risk management

M3 - Working paper

T3 - Danmarks Nationalbank. Working Papers

BT - Modeling Frailty Correlated Defaults with Multivariate Latent Factors

PB - Danmarks Nationalbank

CY - København

ER -