identifying excessive credit growth and leverage
TRANSCRIPT
Intro Data Predictive model Results Out-of-sample Conclusion
Identifying excessive credit growthand leverage
Lucia Alessi 1 Carsten Detken2
1European Commission - Joint Research Centre
2European Central Bank
This work was carried out while Lucia Alessi was affiliated with the ECB. The views in this presentation arethose of the authors and do not necessarily reflect those of the ECB or the European Commission.
Alessi/Detken - Identifying excessive credit growth and leverage
Intro Data Predictive model Results Out-of-sample Conclusion
Aim of the paperEarly warning indicators for macropru instruments targeting credit
Excessive credit
growth and
leverage
Macropru instruments
(CCBs, SRB, LR, LTV-LTI caps)
TRIGGER
Alessi/Detken - Identifying excessive credit growth and leverage
Intro Data Predictive model Results Out-of-sample Conclusion
Target variableSystemic banking crises and ‘near misses’
Banking crises dataset by Expert Group:
• based on the HoR database compiled by the MaRs
• amended in order to include:
1. only systemic banking crises associated with a domesticcredit/financial cycle
2. periods in which in the absence of policy action or of anexternal event that dampened the credit cycle a crisis as in1. would likely have occurred
Alessi/Detken - Identifying excessive credit growth and leverage
Intro Data Predictive model Results Out-of-sample Conclusion
Target variable
AT eeeeeeeeeeeeeeeeeeee
BE eeeeeeeeeeeeeeeeeeee
CY pppppppppppppppppeeeccccccc
DK pppppppppppppppppeeecccccccccccccccccccccccccccc pppppppppppppppppeeecccccccccccccccccccccc
EE pppppppppppppppppeeeccc eeeeeeeeeeeeeeeeeeee
FI pppppppppppppppppeeecccccccccccccccccc eeeeeeeeeeeeeeeeeeee
FR pppppppppppppppppeeecccccccccc pppppppppppppppppeeecccccccccccccccccccccc
DE pppppppppppppppppeeecccccccccccccccc eeeeeeeeeeeeeeeeeeee
GR pppppppppppppppppeeecccccccccccccccccccccccc
IE pppppppppppppppppeeecccccccccccccccccccccc
IT pppppppppppppppppeeecccccccc eeeeeeeeeeeeeeeeeeee
LV pppppppppppppppppeeecccccccceeeeeeeeeeeee
LU eeeeeeeeeeeeeeeeeeee
MT eeeeeeeeeeeeeeeeeeee
NL pppppppppppppppppeeeccccccpppppppppppppppppeeecccccccccccccccccccccc
PT pppppppppppppppppeeeccccc pppppppppppppppppeeeccccccccccccccccccccc
SK eeeeeeeeeeeeeeeeeeee
SI pppppppppppppppppeeecccccccccccc pppppppppppppppppeeecccccccccccccccccccccccc
ES pppppppppppppppppeeeccccccccccccccccccccccccccccccc pppppppppppppppppeeeccccccccccccccccccc
SE pppppppppppppppppeeecccccccccccccc pppppppppppppppppeeecccccccccceeeeeeeeeeee
GB ppppppppppppeeeccccccccc pppppppppppppppppeeecccccccccccccccc pppppppppppppppppeeecccccccccccccccccccccccccc
precrisis excluded crisis
12 1306 07 08 09 10 1100 01 02 03 04 0594 95 96 97 98 9988 89 90 91 92 9382 83 84 85 86 8776 77 78 79 80 8170 71 72 73 74 75
Alessi/Detken - Identifying excessive credit growth and leverage
Intro Data Predictive model Results Out-of-sample Conclusion
Early warning indicators
• Credit related indicators, based on total credit and bankcredit, credit to households and non-financial corporations,the debt service ratio and public debt
• Real estate indicators based on residential property prices,incl. valuation measures
• Market-based indicators such as the short and long terminterest rates and equity prices
• Macroeconomic variables such as real GDP growth, M3,real effective exchange rate, current account
Alessi/Detken - Identifying excessive credit growth and leverage
Intro Data Predictive model Results Out-of-sample Conclusion
Classification treesIs the minimum systolic blood
pressure over the initial 24 hour period > 91 ?
No Yes
High risk Is age > 62.5?
No Yes
Is sinus tachyardia present?
No Yes
High risk Low risk
Low risk
Alessi/Detken - Identifying excessive credit growth and leverage
Intro Data Predictive model Results Out-of-sample Conclusion
Recursive partitioning
GINI(f ) =∑i,j
Cij fi fj
Alessi/Detken - Identifying excessive credit growth and leverage
Intro Data Predictive model Results Out-of-sample Conclusion
The Random Forest
Bootstrap and aggregation of a multitude of trees, each grownon a randomly selected set of indicators and observations.
⇓Robust technique
Alessi/Detken - Identifying excessive credit growth and leverage
Intro Data Predictive model Results Out-of-sample Conclusion
Random forest performanceAUROC=0.94, out-of-sample missclassification=7%
0 0.2 0.4 0.6 0.8 10
0.2
0.4
0.6
0.8
1
False Positive Rate
Tru
e P
ositi
ve R
ate
Alessi/Detken - Identifying excessive credit growth and leverage
Intro Data Predictive model Results Out-of-sample Conclusion
Random Forest rankingKey indicators
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
Indi
cato
r Im
port
ance
Ban
k cr
edit/
GD
P
Glo
bal c
redi
t/GD
P g
ap (
400K
)
HH
cre
dit/G
DP
Deb
t ser
vice
rat
io
Glo
bal c
redi
t gro
wth
Ban
k cr
edit/
GD
P g
ap (
400K
)
Hou
se p
rice/
inco
me
Bro
ad c
redi
t/GD
P
Hou
se p
rice
gap
(400
K)
Hou
se p
rice
grow
th
ST
rat
e
HH
Deb
t ser
vice
rat
io
Glo
bal c
redi
t/GD
P g
ap (
26K
)
Ban
k cr
edit
grow
th
NF
C c
redi
t/GD
P
Equ
ity p
rice
grow
th
M3
gap
(26K
)
Bas
el g
ap
M3
gap
(400
K)
Hou
se p
rice
gap
(26K
)
NF
C D
ebt s
ervi
ce r
atio
NF
C c
redi
t/GD
P g
ap (
400K
)
Hou
se p
rice/
rent
Gov
’t de
bt
Bro
ad c
redi
t gro
wth
HH
cre
dit/G
DP
gap
(40
0K)
NF
C c
redi
t/GD
P g
ap (
26K
)
LT g
ov’t
bond
yld
HH
cre
dit g
row
th
Hou
sing
loan
gro
wth
Cur
rent
acc
ount
Bro
ad c
redi
t/GD
P g
ap (
26K
)
Ban
k cr
edit/
GD
P g
ap (
26K
)
HH
cre
dit/G
DP
gap
(26
K)
Effe
ctiv
e ex
ch. r
ate
GD
P g
row
th
M3
grow
th
NF
C c
redi
t gro
wth
Alessi/Detken - Identifying excessive credit growth and leverage
Intro Data Predictive model Results Out-of-sample Conclusion
Early warning tree
Warningcrisis pr.= 1
7 obs.
HH credit/GDP
Bank credit/GDP
< 92
>54.5
> 92
<3.6 p.p.
Warningcrisis pr.= 0.62
63 obs.
No warningcrisis pr.= 0227 obs.
No warning
crisis pr.= 0.07375 obs.
Debt service ratio
> 10.6%
Debt service ratio
> 18%< 18%
No warning
crisis pr.= 018 obs.
Warningcrisis pr.= 1
5 obs.
M3 gap
< ‐0.2 p.p. > ‐0.2 p.p.
ST rate
House price gap
No warningcrisis pr.= 0226 obs.
< 10.6%
< ‐0.5%
Bank credit growth
> 7.3%
> ‐0.5%
< 0.2 p.p. > 0.2 p.p.< 7.3%
No warningcrisis pr.= 0.25
4 obs.
Bank credit gap
>3.6 p.p. <54.5
No warningcrisis pr.= 012 obs.
Warningcrisis pr.= 0.9139 obs.
< 8%
House price/income
> 27 p.
> 8%
< 2 p.p. > 2 p.p.
< 27 p.
House price growth
Basel gap
House price/income
> ‐3 p.< ‐3 p.
No warningcrisis pr.= 0189 obs.
HH credit/GDP
Equity price growth
> 36 %< 36 %
WarningNo warningcrisis pr.= 0.75
8 obs.crisis pr.= 0.08
83 obs.
Bank credit/GDP
Bank credit/GDP
House price/income
crisis pr.= 0.131 obs.
No warningcrisis pr.= 0
5 obs.
Warning Warning
WarningNo warning
< 46.7 > 46.7
> 42.4< 42.4 > 98.5< 98.5
> ‐2 p.< ‐2 p.
No warning
crisis pr.= 0.9738 obs.
crisis pr.= 0.3312 obs.
crisis pr.= 0.754 obs.
crisis pr.= 07 obs.
Alessi/Detken - Identifying excessive credit growth and leverage
Intro Data Predictive model Results Out-of-sample Conclusion
Evaluation metricsCrisis No Crisis
Signal A BNo signal C D
θ = 2/3
TPR AA+C 85%
FPR (Type II error) BB+D 4%
Type I error CA+C 15%
N2S BB+D)/
AA+C 5%
Loss θ CA+C + (1− θ) B
B+D 0.12
Usefulness min[θ;1− θ]− Loss 0.22
Rel. Usefulness Usefulnessmin[θ;1−θ] 0.65
Alessi/Detken - Identifying excessive credit growth and leverage
Intro Data Predictive model Results Out-of-sample Conclusion
Out-of-sample exerciseImagine you were in mid-2006
Crisis No crisisWarning FR, IE, ES, FI, IT
SE, DK, UKNo warning GR, PT, LV, AU, BE, LU, DE,
SI, NL EE, SK, MT, CY**Crisis started beyond prediction horizon.
Not classified in terminal nodes owing to lack of data.
AT BE CY DK EE FI FR DE GR IE IT LV LU MT NL PT SK SI ES SE GB2006Q3 x x x x x x x x2006Q4 x x x x x x x x2007Q1 x x x x x x x x2007Q2 x x x x x x x x2007Q3 x x x x x x x x2007Q4 x x x x x x x x x2008Q1 x x x x x x x x x2008Q2 x x x x x x x x x2008Q3 x x x x x
Alessi/Detken - Identifying excessive credit growth and leverage
Intro Data Predictive model Results Out-of-sample Conclusion
Conclusion
• The Random Forest/Early Warning methodology canbecome a useful quantitative tool to:
• spur discussion on country risks• provide information on the most appropriate policy
instrument to address identified vulnerabilities
• Additional relevant (potentially country specific) informationcan be included
Alessi/Detken - Identifying excessive credit growth and leverage