Discussion of
\House Prices, Foreclosures, and Bail-Outs"
by Garriga and Schlagenhauf
Dirk Krueger
University of Pennsylvania, CEPR, and NBER
Conference at the Sveriges Riksbank
September 19, 2008
Change in Foreclosure Rate -2006Q4 to 2007Q4
-.5
0
.5
1
1.5
2
2.5
Figure 1: Annual Change in OFHEO HPI
vs. Change in Foreclosure Rate
FL
NV
CA
RI
AZ
MI
-10
MA
IL ME
DE
NY WI
CT HI
NH MD
VA
INDCVT
OH
GA
CO IAKY ID
WY
MOWV OR
SC OK
WA MT
SDNM
NE KS
AK
PA AR AL
LA TX
ND
TN
NC
MS
MN
NJ
-7.5
-5
-2.5
0
2.5
5
7.5
Percentage Change in OFHEO HPI - 2006Q4 to 2007Q4
UT
10
Note: Slope coefficient = −0.126 with an R-square of 0.72.
Source: Calomiris, Longhofer and Miles (2008)
35
1
Introduction
The boom in ownership in the United States that was initiated in 1994 started to have
signi…cant impact in house prices around 2002. Figure 1 describes the real appreciation rate
measured by di¤erent house price indices.
Figure 1: Evolution of House Prices
1.800000
1.600000
OFHEO Index
Conventional Mtg Index
S&P/ Case Shiller Index
1.400000
1.200000
1.000000
M
ar
-9
8
Ju
l-9
N 8
ov
-9
M 8
ar
-9
9
Ju
l-9
N 9
ov
-9
M 9
ar
-0
0
Ju
l-0
N 0
ov
-0
M 0
ar
-0
1
Ju
l-0
N 1
ov
-0
M 1
ar
-0
2
Ju
l-0
N 2
ov
-0
M 2
ar
-0
3
Ju
l-0
N 3
ov
-0
M 3
ar
-0
4
Ju
l-0
N 4
ov
-0
M 4
ar
-0
5
Ju
l-0
N 5
ov
-0
M 5
ar
-0
6
Ju
l-0
N 6
ov
-0
M 6
ar
-0
7
Ju
l-0
7
0.800000
The …gure combines a relatively steady appreciation of house prices between 1998 and 2002,
a much rapid increase in house prices between 2003 and 2005, with a …nal decline after
the summer 2005. Changes in house prices have a signi…cant impact in the homeowners
portfolios since they change the level of equity accrued in the dwelling. in periods with high
appreciation homeowners can borrow against a collateral with an increased value to increase
consumption or can opt to sell the property and purchase a bigger one, but in periods with
falling prices the outstanding debt can be larger than the market value of the property making
default a viable option for some homeowners. The evidence seems to suggest a connection
between house prices and housing foreclosures. Next …gure, summarizes the evolution of
seriously delinquent mortgages between 1990 and 2007.1
Figure 2: Evolution of Seriously Delinquent Mortgages
Delinquent
3.5
Mortgages:
US
Percent
3
2.5
2
1.5
119 9 0
1992
1994 1996
(Source:
1998 2000 2002 2004 2006 2008
Time
Mortgage Bankers Association)
1
The concept of "seriously delinquent mortgages" is calculated by adding the percentages of mortgage
payments 90 days past due and the percentages of inventory of mortgages in foreclosure. "Inventory of
Mortgages in foreclosure" refers to the total number of loans in the legal process of foreclosure as a percentage
of the total number of mortgages in the pool during a quarter. The number of loans in the process of
foreclosure during a quarter means that some foreclosures may have started in other quarters but have yet
to be resolved.
2
1
Introduction
The boom in ownership in the United States that was initiated in 1994 started to have
signi…cant impact in house prices around 2002. Figure 1 describes the real appreciation rate
measured by di¤erent house price indices.
Figure 1: Evolution of House Prices
1.800000
1.600000
OFHEO Index
Conventional Mtg Index
S&P/ Case Shiller Index
1.400000
1.200000
1.000000
M
ar
-9
8
Ju
l-9
N 8
ov
-9
M 8
ar
-9
9
Ju
l-9
N 9
ov
-9
M 9
ar
-0
0
Ju
l-0
N 0
ov
-0
M 0
ar
-0
1
Ju
l-0
N 1
ov
-0
M 1
ar
-0
2
Ju
l-0
N 2
ov
-0
M 2
ar
-0
3
Ju
l-0
N 3
ov
-0
M 3
ar
-0
4
Ju
l-0
N 4
ov
-0
M 4
ar
-0
5
Ju
l-0
N 5
ov
-0
M 5
ar
-0
6
Ju
l-0
N 6
ov
-0
M 6
ar
-0
7
Ju
l-0
7
0.800000
The …gure combines a relatively steady appreciation of house prices between 1998 and 2002,
a much rapid increase in house prices between 2003 and 2005, with a …nal decline after
the summer 2005. Changes in house prices have a signi…cant impact in the homeowners
portfolios since they change the level of equity accrued in the dwelling. in periods with high
appreciation homeowners can borrow against a collateral with an increased value to increase
consumption or can opt to sell the property and purchase a bigger one, but in periods with
falling prices the outstanding debt can be larger than the market value of the property making
default a viable option for some homeowners. The evidence seems to suggest a connection
between house prices and housing foreclosures. Next …gure, summarizes the evolution of
seriously delinquent mortgages between 1990 and 2007.1
Figure 2: Evolution of Seriously Delinquent Mortgages
Delinquent
3.5
Mortgages:
US
Percent
3
2.5
2
1.5
119 9 0
1992
1994 1996
(Source:
1998 2000 2002 2004 2006 2008
Time
Mortgage Bankers Association)
1
The concept of "seriously delinquent mortgages" is calculated by adding the percentages of mortgage
payments 90 days past due and the percentages of inventory of mortgages in foreclosure. "Inventory of
Mortgages in foreclosure" refers to the total number of loans in the legal process of foreclosure as a percentage
of the total number of mortgages in the pool during a quarter. The number of loans in the process of
foreclosure during a quarter means that some foreclosures may have started in other quarters but have yet
to be resolved.
2
Motivation
Strong negative correlation between changes in house prices and changes
in foreclosure rates in the U.S.
Conventional wisdom I: price declines lower owners' equity in the home
and thus cause changes in default rates on mortgages (foreclosures).
Conventional wisdom II: rising foreclosures increase supply of homes
on the market, cause price declines.
Motivation
(Quantitative) Theory: home prices and foreclosure rates are jointly
determined in general equilibrium. Garriga & Schlagenhauf's research
agenda: provide us with this quantitative theory
This paper: causality runs from price changes to foreclosure rates.
{ Construct model that matches homeownership rates, foreclosure
rates
{ Engineer a change in the house price and document what happens
to foreclosure rates in the model.
Model: Key Ingredients
Chambers, Garriga & Schlagenhauf's (2007) model of housing market
{ OLG model with uninsurable idiosyncratic income risk
{ Relative price of real estate, p; and risk free rate r exogenous.
{ Endogenous housing tenure and mortgage decisions
{ Housing investment lumpy and subject to transaction costs and
idiosyncratic price shocks .
Introduction of mortgage default as in Krueger and Jeske (2007).
Time line for households that
owned in previous period
Stay put
c,d,a’,Ir
Sell, rent
Sell, buy
c,d,a’
ξ~πξ
Foreclose?
State s=(a,h,z,n,ε,j)
Today
c,d,a’,h’,z’,Ir
State s’=(a’,h’,z’,n’,ε’,j+1)
Tomorrow
Mortgage Contracts and Default
Pre-commitment to \sell". House price shock
(1
s ) ph
realized. Default i
D(z; n)h < 0
Threshold (z; n) such that default i
< (z; n): Corresponding
default probability (z; n) and mortgage interest rate %(z ):
(z; n); (z; n); %(z ) depend on mortgage type z and length
n; but not on characteristics of borrower, s; or size of the house h:
Note: for default only di erence in contracts is D(z; n):
Prices and Default Decision
πξ
Default Region
ξ*(z,n;p)
ξ
Quantitative Prediction of Benchmark Model
Model matches homeownership rates over the life cycle and aggregate
statistics remarkably well.
Focus on foreclosure rates d = dF R sF R + dGP sGP
{ zF R : Standard 30 year xed rate contract with 20% down
{ zGP : 30 year contact, low downpayment, growing payments
Stat.
Data (98)
Model
d
1.0%
1.8%
sF R
85%
61%
dF R
0.8%
1.7%
sGP
15%
39%
dGP
2.0%
2.0%
Thought Experiment: Unexpected Exogeneous Fall
in House Price p
Default if
(1
s ) ph
D(z; n)h < 0:
For contract z a decline from p to p0 leads to an increase in
Is the model \continuous" in p if
2
; and
(z; n):
is a very nite set?
Note that for given contract z heterogeneity in s does not help since
(z; n) not a function of s: Mortgage contract is a choice, z = Z (s):
Prices and Default Decision
p’<p
πξ
Default Region
ξ*(z,n;p)
ξ*(z,n;p’)
ξ
Prices and Default Decision
p’<p
πξ
Default Region
ξ*(z,n;p)
ξ*(z,n;p’)
ξ
Prices and Default Decision
p’<p
Default Region p'
πξ
Default Region p
ξ*(z,n;p)
ξ*(z,n;p’)
ξ
Fall in House Price p by 15 Percent: Results
Ownership rate not a ected much (despite the fact that R=p changes
signi cantly).
E ects on mortgage default:
Stat.
Data (98)
Data (07)
Model (98)
Model (07)
d
1.0%
2.8%
1.8%
2.7%
sF R
85%
77%
61%
72%
dF R
0.8%
1.2%
1.7%
2.2%
sGP
15%
23%
39%
28%
dGP
2.0%
7.4%
2.0%
4.0%
Conclusions
Ambitious paper. Introduces endogenous default into elaborate model
of housing.
Model ts homeownership rates well.
Di erential default rates by mortgage type? Market shares? Reasonably well.
{ Very coarse set of mortgage contracts
{ Model has signi cant heterogeneity, but maybe not enough (unobserved di erences in types?)
Future: Even More Ambitious Paper
What fundamental factors jointly determine
(recent) trends in house prices and mortgage
default rates?