Leading Indicators of Distress in Danish Banks in the Period 2008-12

Birgitte Vølund Buchholst, Jesper Rangvid

Publikation: Working paperForskning

Resumé

Flere små og mellemstore pengeinstitutter i Danmark er blevet nødlidende, siden den finansielle krise brød ud. I denne analyse anvendes multipel logistisk regression til at identificere, hvilke faktorer der karakteriserer nødlidende pengeinstitutter i Danmark i perioden 2008-12. Faktorerne udvælges fra en bred vifte af forklarende variable, dvs. modellen er ubegrænset. Den estimerede model identificerer de nødlidende pengeinstitutter forholdsvis godt. De variable, der tilsammen bedst beskriver sandsynligheden for, at et pengeinstitut bliver nødlidende, er instituttets kapitaloverdækning i forhold til risikovægtede aktiver, den 3-årige gennemsnitlige udlånsvækst lagget 2 år, ejendomseksponering samt et mål for stabil finansiering (den såkaldte funding-ratio). Variablene er alle fratrukket sektorens gennemsnit for at tage højde for den generelle udvikling i perioden.
Baseret på erfaringer fra denne og tidligere kriser introducerede Finanstilsynet tilsynsdiamanten i 2010 som en del af deres tilsyn med pengeinstitutter. Der estimeres en multipel logistisk regressionsmodel med afvigelser fra tilsynsdiamantens pejlemærker som forklarende variable for at vurdere, om variablene i tilsynsdiamanten adskiller sig fra den ubegrænsede model. Overordnet set understøtter analyserne fastsættelsen af pejlemærker. Resultaterne af denne analyse viser, at afvigelser fra pejlemærkerne vedrørende ejendomseksponering og funding-ratio er statistisk signifikante med forventede fortegn, hvorimod afvigelser fra pejlemærkerne vedrørende udlånsvækst, store engagementer og likviditetsoverdækning er statistisk insignifikante.
OriginalsprogEngelsk
Udgivelses stedKøbenhavn
UdgiverDanmarks Nationalbank
Antal sider37
StatusUdgivet - sep. 2013
NavnDanmarks Nationalbank. Working Papers
Nummer85
ISSN1602-1193

Citer dette

Buchholst, B. V., & Rangvid, J. (2013). Leading Indicators of Distress in Danish Banks in the Period 2008-12. København: Danmarks Nationalbank. Danmarks Nationalbank. Working Papers, Nr. 85
Buchholst, Birgitte Vølund ; Rangvid, Jesper. / Leading Indicators of Distress in Danish Banks in the Period 2008-12. København : Danmarks Nationalbank, 2013. (Danmarks Nationalbank. Working Papers; Nr. 85).
@techreport{2e3cb8159684444d927897b9c9f18f89,
title = "Leading Indicators of Distress in Danish Banks in the Period 2008-12",
abstract = "Several Danish small and medium-sized banks have become distressed during and after the global financial crisis. In this paper, a multiple logistic regression model is used to identify which factors characterize the distressed Danish banks from 2008-12. The factors are chosen from a broad range of variables, i.e. the model is unrestricted. The estimated model identifies the distressed banks fairly well. The variables that altogether best describe the probability of a bank becoming distressed are: a bank's excess capital in per cent of risk weighted assets, the 3 year average lending growth lagged 2 years, property exposure, and a benchmark for stable funding (the so-called funding-ratio). The variables are all adjusted with the sector average to account for the general development during the period. Based on experiences from this and past crises the Danish FSA introduced the so-called Supervisory Diamond as part of its banking supervision in 2010. A multiple logistic regression model is estimated with deviations from limit values set in the supervisory diamond to assess whether the variables in the supervisory diamond differ from the unrestricted model. Overall, the analyses support the establishment of benchmarks. The results of this analysis show that deviations from the benchmarks concerning property exposure and funding-ratio are statistically significant with expected signs. However, deviations from the benchmarks concerning lending growth, large exposures, and excess liquidity cover are statistically insignificant.",
author = "Buchholst, {Birgitte V{\o}lund} and Jesper Rangvid",
year = "2013",
month = "9",
language = "English",
publisher = "Danmarks Nationalbank",
address = "Denmark",
type = "WorkingPaper",
institution = "Danmarks Nationalbank",

}

Buchholst, BV & Rangvid, J 2013 'Leading Indicators of Distress in Danish Banks in the Period 2008-12' Danmarks Nationalbank, København.

Leading Indicators of Distress in Danish Banks in the Period 2008-12. / Buchholst, Birgitte Vølund; Rangvid, Jesper.

København : Danmarks Nationalbank, 2013.

Publikation: Working paperForskning

TY - UNPB

T1 - Leading Indicators of Distress in Danish Banks in the Period 2008-12

AU - Buchholst, Birgitte Vølund

AU - Rangvid, Jesper

PY - 2013/9

Y1 - 2013/9

N2 - Several Danish small and medium-sized banks have become distressed during and after the global financial crisis. In this paper, a multiple logistic regression model is used to identify which factors characterize the distressed Danish banks from 2008-12. The factors are chosen from a broad range of variables, i.e. the model is unrestricted. The estimated model identifies the distressed banks fairly well. The variables that altogether best describe the probability of a bank becoming distressed are: a bank's excess capital in per cent of risk weighted assets, the 3 year average lending growth lagged 2 years, property exposure, and a benchmark for stable funding (the so-called funding-ratio). The variables are all adjusted with the sector average to account for the general development during the period. Based on experiences from this and past crises the Danish FSA introduced the so-called Supervisory Diamond as part of its banking supervision in 2010. A multiple logistic regression model is estimated with deviations from limit values set in the supervisory diamond to assess whether the variables in the supervisory diamond differ from the unrestricted model. Overall, the analyses support the establishment of benchmarks. The results of this analysis show that deviations from the benchmarks concerning property exposure and funding-ratio are statistically significant with expected signs. However, deviations from the benchmarks concerning lending growth, large exposures, and excess liquidity cover are statistically insignificant.

AB - Several Danish small and medium-sized banks have become distressed during and after the global financial crisis. In this paper, a multiple logistic regression model is used to identify which factors characterize the distressed Danish banks from 2008-12. The factors are chosen from a broad range of variables, i.e. the model is unrestricted. The estimated model identifies the distressed banks fairly well. The variables that altogether best describe the probability of a bank becoming distressed are: a bank's excess capital in per cent of risk weighted assets, the 3 year average lending growth lagged 2 years, property exposure, and a benchmark for stable funding (the so-called funding-ratio). The variables are all adjusted with the sector average to account for the general development during the period. Based on experiences from this and past crises the Danish FSA introduced the so-called Supervisory Diamond as part of its banking supervision in 2010. A multiple logistic regression model is estimated with deviations from limit values set in the supervisory diamond to assess whether the variables in the supervisory diamond differ from the unrestricted model. Overall, the analyses support the establishment of benchmarks. The results of this analysis show that deviations from the benchmarks concerning property exposure and funding-ratio are statistically significant with expected signs. However, deviations from the benchmarks concerning lending growth, large exposures, and excess liquidity cover are statistically insignificant.

M3 - Working paper

BT - Leading Indicators of Distress in Danish Banks in the Period 2008-12

PB - Danmarks Nationalbank

CY - København

ER -

Buchholst BV, Rangvid J. Leading Indicators of Distress in Danish Banks in the Period 2008-12. København: Danmarks Nationalbank. 2013 sep.