Everything you always wanted to know
about Basel II in 15 minutes
(a real estate perspective)
Erik Kersten
Senior Policy Advisor
Supervisory Policy – Quantitative Risk Management
Views and opinions expressed in this presentation are those of the author and do not necessarily reflect the position of the Nederlandsche Bank.
Outline
• Overview of Basel II
• Real estate issues under Basel II
Capital requirements under Basel I
Loans
100
(Risk weight 100%)
100
“Worst case”
Performing Loans
Loans losses (8)
92
Owners equity
Deposits
8
92
100
Owners equity
Deposits
0
92
92
92
•Required Capital = (Risk Weight * Exposure value)*8%
Capital requirements under Basel I
Mortgage Loans
100
(Risk weight 50%)
100
“Worst case”
Performing Loans
Loan losses (4)
96
96
Owners equity
Deposits
4
96
100
Owners equity
Deposits
0
96
96
Mortgage collateral under Basel I
• Residential mortgages: if LtV is lower then
threshold: lower risk weight, less capital
• No recognition in capital of commercial real
estate (with some exceptions)
• No realistic assumptions
• Capital treatment fuelled MBS-market
Purposes of the new capital accord
• a comprehensive approach to addressing risks
• Risk sensitive capital requirements
• Promote soundness and safety of the financial system
• Enhance competitive equality
• Accommodate industry best practice
Result:
⇒Widen acceptance of collateral types used in practise
and
⇒widen acceptance of internal models to Credit risk &
Oprisk
Structure of Basel II
Menu of approaches (pillar 1)
• For measuring Credit Risk:
• Standardised Approach
• Foundation Internal Ratings-based Approach
• Advanced Internal Ratings-based Approach
• For measuring Operational Risk:
• Basic Indicator Approach
• Standardised Approach
• Advanced Measurements Approach
• For measuring Market Risk:
• Standardised Approach
• Internal Models Approach
Credit risk (Pillar I)
• The risk of loss due to the fact that an obligor will not
meet its credit obligations in full
• Required Capital = (RW * Exposure value)*8%
(no change!)
• Standardised approach
• RW Based on External ratings (Moody’s, S&P, local
rating agencies)
• Internal ratings based approach
• RW based on Internal ratings (banks own assessment)
Internal Ratings Based Approach
• Capital requirements for an exposure based on
VaR and function of: PD, LGD, EAD and M
• These functions are given by Accord
IRB: the formulae
⎡ look up PD
⎤
8
⎢ 6474
⎥
⎢ Φ − 1 ( PD )
⎥
⎞
⎛
⎞
R
1
⎛
⎞ ⎛⎜
− 1 (0.999 )
⎟ *1,06
⎜
⎟
⎥−
+
×
Φ
PD
)
×
1
+
M
−
2.5
×
b
K = LGD × (Φ ⎢
×
⎜
⎟
{
{⎟
⎜
1
4
4
2
4
4
3
⎟
⎜
1
−
1
.
5
*
b
⎢ 1414−4
⎥
R424414−43
R
⎝
⎠ ⎝
⎝ 4444⎠ 43⎠
1444444
24
look up 99.9%
⎢
⎥
correct for maturity
⎢⎣account for correlatio n in the normal distributi on ⎥⎦ accountfor EL
14
444444444442444444444444
3
calculate PD stressed at 99.9%
RW = K * 12,50
⎛ 1 − e − 50 × PD
Correlation ( R ) = 0.12 × ⎜
⎜
− 50
⎝ 1− e
⎡ ⎛
− 50 × PD
⎞
⎟ + 0.24 × ⎢1 − ⎜ 1 − e
⎟
⎢ ⎜ 1 − e − 50
⎠
⎣ ⎝
For Corporates, Banks & Sovereigns
SME correction : R ranging from 0.08 to 0.20
Retail :
Mortgages : R = 0.15;
credit cards : R = 0.04
and other retail : R ranging from 0.03 to 0.16
⎞⎤
⎟⎥
⎟⎥
⎠⎦
Maturityadjustment(b) = (0.11852− 0.05478× ln(PD))
Onlyfor Corporates, Banks& Sovereigns
2
‘internal’ refers to the inputs
Foundation IRB
Probability of default
Loss Given Default
Exposure at Default
Maturity
Advanced IRB
Own estimates
Supervisory formula
Own estimates
Supervisory formula
Own estimates
Bank’s own estimates or 2.5
yrs
Own estimates
RW vs PD
R W -c u rve s (P D to t 1 0 % )
(L G D = 45 % , M o rtga ges L G D = 1 0 % M =2 ,5 )
250,00 %
200,00 %
150,00 %
RW
C o rp orates
C o rp orates(sm a ll)
m ortg age s
QRE
O the r R eta il
100,00 %
50,00 %
0,00 %
0,00%
2 ,0 0%
4,00%
6 ,0 0%
PD
8,00%
10,00 %
1 2,00%
Conclusion
• IRB is a more risk sensitive way of calculating
capital requirements
• based on statistical properties of portfolio
• and enhances internal management of loans
Real estate in Basel II
Approaches in pillar 1
• Measuring Credit Risk in mortgage lending
• Residential real estate:
• Standardised Approach
• RW down from 50% to 35%
• Monitoring LtV-ratio’s (at least every 3 years)
• Advanced Internal Ratings-based Approach
• banks now have to estimate PD, LGD & EAD for
their retail mortgages (NB collateral
management conditions, e.g. LtV-monitoring
process)
Approaches in pillar 1
•
Measuring Credit Risk in mortgage lending
• Commercial real estate:
• Standardised Approach
• RW based on external rating (non-rated => RW 100%
• Commercial real estate eligible as collateral (NB conditions: e.g. yearly
LtV monitoring)
• Foundation Internal Ratings-based Approach
• RW based on internal estimates of PD
• Commercial real estate eligible as collateral (NB collateral management
conditions, e.g. yearly LtV-monitoring process!)
• Advanced Internal Ratings-based Approach
• RW based on internal estimates of PD, LGD, EAD (&M)
• Commercial real estate eligible as collateral (NB same conditions, but
more freedom in way of meeting those conditions)
Approach in Pillar 2
• Banks are free to develop their own models
• Ensuring sound internal processes to assess
risks and capital adequacy
• Active dialogue between banks and their
supervisors,
• Identify deficiencies
• Take prompt and decisive action
Focus on IRB
• Most mortgages will probably be subject to IRBregime
• (Bigger banks use IRB)
What information do we need?
• How good is the obligor
• Probability of default; PD
• What determines a PD?
• Relation between PD and LtV (low LtV tend to have lower PD?)
What information do we need?
• How much will we recover after default
• Loss given default; LGD
• What determines a LGD?
• Relation between LGD and LtV
• LGD is more then Loans – Current Market Value!
• economic loss: recovery value, time & costs,
• down turn effect (remember, we’re talking UL)
What information do we need?
• How much money is the obligor likely to owe us
when a default occurs
• Exposure at default; EAD (Credit Conversion
Factor)
• What determines an EAD?
• All kind of ‘options’ with mortgage-lending
LGD: strong effect on RW
• LGD↑ =► RW↑
Mortgage RW vs PD, at different LGD levels
250,00%
200,00%
150,00%
RW
LGD = 45%
LGD = 25%
LGD = 10%
100,00%
50,00%
0,00%
0,00%
1,00%
2,00%
3,00%
4,00%
5,00%
PD
6,00%
7,00%
8,00%
9,00%
Importance?
N
et
he
r
D lan
Sw en d
itz ma s
er rk
Ic lan
el d
an
d
U
K
Ire US
Sw lan
Po ed d
rt en
N ug
or a l
w
a
G Sp y
er ai
m n
a
EUny
E 1
Lu F U 5
i
xe n 25
m lan
bo d
ur
B Ma g
el lta
g
F r ium
a
G nc
re e
Es ec
t e
A oni
us a
La tria
tv
ia
I
C tal
y y
C pru
r
Li o s
th at
ua ia
H
n
C
u
ze S ng ia
ch lo a
r
R va y
ep ki
u a
Po bli
Sl la c
o n
B ven d
ul ia
g
T u a ri
R rk a
om e
a y
Se nia
R rbi
u a
U ss
kr ia
ai
ne
Importance?
Overview of EU residential mortgage markets 2005
100%
70%
50.000
90%
Mortgage Debt to GDP ratio
80%
Mortgage debt per capita right hand scale
0%
45.000
40.000
35.000
60%
30.000
50%
25.000
40%
20.000
30%
15.000
20%
10.000
10%
5.000
0
Importance?
in € 1000
Gemiddelde koopsom en hypotheeksom (per kw)
290
in %
14
116
ratio
Gemiddelde LTV (per kwartaal)
Koopsom
270
Hypotheeksom
12
114
250
prijsmutatie j/j rechter as
10
112
230
8
210
6
190
4
170
2
102
0
100
150
2000
2001
Bron: Kadaster
2002
2003
2004
2005
2006
4e kw
2006
110
108
106
104
2000
2001
Bron: Kadaster
2002
2003
2004
2005
2006
Mortgage lending important?
Yes