s 2
g a
niver
ersity,
explo
2007
s in
ation
c con
om th
of fin
to m
f the
gulat
ng fro
that attempt to decompose the causes of the great surge in
defaults from 2006 to 2009 using a large data set with loan
level information. One outgrowth of this research is a set of
metrics that can be used to track the quality of mortgage
writing characteristics that were known to investors in the
pools. With these data we can decompose defaults into
three parts: those due to worsening of economic condi-
tions, those due to observable changes in underwriting
standards, and a set of time varying fixed effects that we
attribute to moral hazard.
We identify moral hazard by looking for discontinuities
or ‘‘notches’’ in default behavior. The underlying economics
1051-1377/$ - see front matter � 2011 Published by Elsevier Inc.
⇑ Corresponding author.
E-mail address: rvo@gwu.edu (R.V. Order).
Journal of Housing Economics 20 (2011) 141–151
Contents lists available at ScienceDirect
Journal of Housi
w.el
doi:10.1016/j.jhe.2011.04.005
and housing gyrations of the last few years. For example, in
October 2005, near the peak of the housing ‘‘bubble’’
Federal Reserve Chairman Ben Bernanke was still arguing
that housing prices ‘‘largely reflect strong economic funda-
mentals’’ and were no cause for concern. Given the fire-
storm that erupted shortly thereafter and that spread
from the mortgage and housing markets to the national
and world economy, it is undoubtedly meiosis to say that
better methods and metrics for monitoring mortgage mar-
kets are urgently needed.
Toward this end, this research presents initial estimates
which they can only measure indirectly, using fixed effects
for observation year. Their results suggest that there was a
trend toward lower underwriting standards characterized
by two major periods of deterioration, one in the middle
and late 1990s and one after 2002. After 2002 the favorable
economic conditions that had masked underwriting deteri-
oration changed, and defaults increased sharply.
The data in this research contains loan level information
on mortgages in non-agency (not Fannie Mae, Freddie Mac,
or Ginnie Mae) mortgage-backed securities that were orig-
inated from 2000 to 2008. This data set includes the under-
1. Introduction and overview
Perhaps the most important task
since the great depression has been
only of individual banks, but also o
i.e., systemic risk. Unfortunately, re
to anticipate the systemic risks arisi
ancial regulators
onitor risks not
banking system,
ors were unable
m the mortgage
portfolios and separate the sources of defaults into indices
for underwriting, moral hazard and economic conditions.
Our analysis is an extension of the earlier work in
Anderson et al. (2011), ACV henceforth. ACV use aggre-
gated data (foreclosure rates by state) that extend back
to the 1990s to decompose defaults into those caused by
economic conditions and those caused by underwriting,
conditions.
� 2011 Published by Elsevier Inc.
The great surge in mortgage default
of economic conditions, underwritin
Dennis R. Capozza a, Robert Van Order b,⇑
aDale Dykema Professor of Business Administration, Ross School of Business, U
bOliver T. Carr Professor of Finance and Real Estate, George Washington Univ
a r t i c l e i n f o
Article history:
Available online 12 May 2011
Keywords:
Credit risk
Credit crunch
Moral hazard
a b s t r a c t
In this research we
nated from 2000 to
economic condition
observe the inform
as well as economi
hazard. Estimates fr
journal homepage: ww
006–2009: The comparative roles
nd moral hazard
sity of Michigan, USA
USA
it the power of a large and rich sample of individual loans origi-
to study the relative roles of underwriting, moral hazard and local
the Great Surge in mortgage defaults. With these data we can
available to investors and control for observable underwriting
ditions. We can also use the data to infer the share due to moral
ese data suggest that much of the variation was due to economic
ng Economics
sevier .com/locate / jhec
through 2010. There is a rising trend with occasional level-
uarter
142 D.R. Capozza, R.V. Order / Journal of Housing Economics 20 (2011) 141–151
suggest that default should be a continuous function of
underlying variables like loan to value ratio (LTV) and credit
score, as well other, harder-to-observe variables. However,
pricing and screening tend to be done over discrete inter-
vals in pricing and underwriting matrices. The conditions
for moral hazard occur when loan sellers and securitizers
have access to better information than that available to
investors. The question is: what sorts of mortgages will be
delivered by traders with asymmetric information?
An 80% LTV and a 620 credit (FICO) score appear to be
critical minimum standards of quality because an LTV
above 80 requires insurance, and at least a 620 FICO is gen-
erally required for agency purchase. Our hypothesis is that
loan sellers who possess superior information relative to
investors will tend to deliver loans that just meet these
minimum standards. Because of these cutoff points in the
standards for loan quality, we expect to see notches in
the default function at or around these points.
We do indeed find notches, particularly at 80% LTV.
However, we find that they are not especially important
in explaining the surge in defaults, which appears to be
due primarily to deteriorating economic conditions, partic-
ularly house price declines, which had previously been
Fig. 1. All foreclosures started: US 1979–2010 quarterly data. Four q
Delinquency Survey.
very favorable. Because economic conditions were so
favorable, they masked, or perhaps even contributed to, a
trend of deteriorating underwriting conditions.
In the next section we document the Great Surge, its
implications for both prime and subprime loans, and the
concurrent economic conditions, especially house prices.
The third section develops and estimates models of default
using the loan level data set. The model is then applied to
extract two types of estimates of the importance of eco-
nomic conditions, underwriting and moral hazard in the
great surge.
2. Background and summary data
2.1. The long run trend in foreclosures
Fig. 1 graphs the time series of foreclosures started as a
percent of the outstanding number of loans from 1979
ing off. Between 1979 and 2002 foreclosure rates quadru-
pled from .48% to 1.96%. In the impressive Great Surge
from 2006 to 2009, foreclosure rates almost quadrupled
again from 1.60% to 5.68% in just three years. Our purpose
with this research is to analyze the available data to enable
a deeper understanding of both the trend and the surge.
The deterioration in mortgage performance has varied
by loan type. Fig. 2 presents data for 1998–2010 on foreclo-
sures started by major product type. The vertical axis is the
annualized percent of loans that enter the foreclosure pro-
cess over each four quarters.1 Note in particular the history
of subprime. Foreclosures fell after the 2001 recession but
then increased sharply after 2005. A similar pattern, but
on a smaller scale and with about a one year lag, occurred
in the prime mortgage data, suggesting that there is a com-
mon factor affecting both prime and subprime and that the
surge in foreclosures is not just a subprime issue. The lag be-
tween prime and subprime makes subprime the canary in
the coal mine. Subprime borrowers respond more quickly
to financial stress than prime borrowers.2
averages annualized. Source: Mortgage Bankers Association National
3. The trends in house prices
The data in the previous section suggest that there may
be an important role that economic conditions play in the
pattern of defaults. Modern contingent-claims based theo-
ries of mortgage valuation treat the borrower’s position as
long a put on the collateral.3 One implication of this ap-
proach is that the put option should be sensitive to the value
of the collateral. When collateral prices are rising, a finan-
cially stressed borrower with equity will rationally choose
to sell the collateral rather than default. Correspondingly,
when collateral prices are falling and equity becomes nega-
tive, the stressed borrower is more likely to choose default.
1 The MBA data do not track how many actually went through
foreclosure to REO, real estate owned by lenders.
2 Subprime responds earlier; but, the eventual total response is smaller
in percentage terms than for prime loans, albeit much larger in absolute
terms.
3 The early example is Findlay and Capozza (1977).
D.R. Capozza, R.V. Order / Journal of Housing Economics 20 (2011) 141–151 143
Therefore, it is important to understand both collateral
prices and financial stress as they relate to foreclosures.
Fig. 3 plots real and nominal house prices since 1975.
There is a cyclical pattern to real house prices from 1975
to 1997 within a narrow range. However, during the boom
years from 1997 to 2007, real house prices rose about 40%
above what had been the long run level until that time.
Such steep increases should be expected to greatly reduce
the need for stressed borrowers to default. Since many
lenders develop underwriting models by evaluating recent
loan performance data, any underwriting models created
using data from this boom period would underestimate
the risks to lenders in more average times.
Fig. 3. Real and nominal house price indices,
Fig. 2. Rate of foreclosures started by loan type, 1998–2010 (%). Sourc
Fig. 4 provides another perspective on the recent period
by plotting the real appreciation rate of house prices. The
long run evidence (e.g., Eichholz, 1997) is that real house
prices appreciate at rates close to zero over decades and
centuries. Thus the 4–6% annual appreciation rates of the
last decade are extraordinary.
Fig. 5 illustrates the variation in house prices for se-
lected metro areas. It shows the extreme ups and down
of some the metro areas like San Diego and Miami that
had price ‘‘bubbles.’’ These areas currently have high fore-
closure rates following price declines. Other metro areas
had smaller increases (Boston and Detroit) with slightly
displaced peaks and troughs. Detroit did not experience
1975–2008 (2008 = 1.0). Source: FHFA.
e: Mortgage Bankers Association National Delinquency Survey.
the ‘‘bubble’’ level of price increases, but nevertheless has
been experiencing elevated default rates. High and persis-
These non-Agency data extend only to 2000 but contain a
large amount of data at the loan level on borrower and loan
characteristics.
144 D.R. Capozza, R.V. Order / Journal of Housing Economics 20 (2011) 141–151
5. Default models
ACV use the data set from the Mortgage Bankers Asso-
ciation which is longer, but aggregated by state. Here we
estimate using the loan level data from non-Agency pools.
4 The UFA indices are available to academic researchers at
www.ufanet.com.
4. A summary statistic for economic conditions
The Fig.s above highlight the possible effects of eco-
nomic conditions on foreclosures. In the analysis that fol-
lows we split foreclosure rates into components arising
from underwriting policy and changes in economic condi-
tions. As a summary measure of economic conditions we
use the quarterly ‘‘ForeScore’’ Default Risk indices by state
compiled by University Financial Associates (UFA), which
track the effect of local and national economic conditions
on the probability of a constant quality loan ever default-
ing.4 The UFA indices are derived from a model of default,
using both local conditions and loan characteristics to ex-
plain default. The indices holds the loan characteristics con-
stant and project the impact of economic conditions on
default. House price changes are the most important driver
of the indices with other economic, demographic, political
and topographic variables explaining the balance. The indi-
ces enable parsimonious estimation of the equations that
follow. A more detailed explanation is contained in ACV
(2011).
Fig. 6 illustrates one use of the index to track nation-
wide default risks over time. In this case the constant qual-
ity loan is moved through time and space to create a na-
tional index for each vintage by averaging across
locations each year. The Index has varied between 60 and
290, i.e., the variation in economic conditions has been suf-
ficient to cause a quadrupling of default rates from trough
to peak during the last decade on a constant quality loan.
There have been two trends in the Default Risk Index:
improvement from 1990 until around 2002 and then a
sharp deterioration. It should be noted that the Index is a
forward looking life-of-loan prediction for loans of the indi-
cated vintage. The projections in the figure use actual data
to the extent they are available, and then (for recent indi-
ces) use forecasts, e.g., of house prices, over the life of each
loan vintage. When the index begins to increase from 2003
on, it is not necessarily because the model ‘‘expects’’ the
indicated vintage to default at high rates immediately.
Any increase during the life of the loan will affect the
life-of-loan index value for that vintage. The figure sug-
gests that indeed economic conditions could be a major
factor in explaining recent history.
tent levels of unemployment in Detroit create high levels
of financial stress for borrowers that interact with the
declining collateral prices. ACV exploit this time and spa-
tial variation in foreclosures in their analysis of the rate
of foreclosures in the MBA serviced portfolio data.
The default model is as follows: let the conditional
probability of default for a loan to borrower i, originated
at time v in region r, observed at time t be:
dvitr ¼ aðt � vÞebXðr;tÞþcY
iðr;vÞþdGðrÞ ð1Þ
where X(r,t) is a vector of time varying covariates that de-
scribe the economy in region r at time t; Yi(r,v) is a vector
of characteristics of loans in region r at time of origination,
v; G(r) is a vector of variables that are not time varying and
describe region r; a(t-v) is the baseline hazard for loan age
t-v;b, c and are vectors of coefficients.
6. The ACV Model
ACV estimate Eq. (1) from the MBA data but do not ob-
serve individual loans, nor do they know origination year,
so only dtr, the share of loans in region (state) r that go into
foreclosure at time t, is observed. The ACV model aggre-
gates across individuals and origination years.
dtr ¼ ebXþdGðrÞ
X
v
X
i
aðt � vÞecYiðr;vÞ=nrt ð2Þ
where nrt are is the number of loans originated prior to
time t in region r that are still alive at time t. This is what
we estimate first.
Taking logarithms of both sides of (2):
logðdtrÞ ¼ bXðr; tÞ þ dGðrÞ
þ log
X
v
X
i
aðt � vÞecYiðr;vÞ=nrt
!
ð3Þ
which can be simplified to
logðdtrÞ ¼ bXðr; tÞ þ dGðrÞ þ et ¼ mtr þ drfr þ et ð4Þ
where fr is a fixed effect for region r and mtr is the Fore-
Score index that applies to loans originated in state r at
time t, and et is an error term.
The error term is quite complicated. It is a weighted
average of underwriting characteristics of the pool of loans
across the different vintages. ACV decompose the error
term in (4) into time fixed effects and everything else to
get:
logðdtrÞ ¼ mtr þ drfr þ dtft þ uðr; tÞ ð5Þ
where ft is a set of fixed effects for time and u is again com-
plicated. Use of the time effect, ft, as a proxy for credit stan-
dards means they cannot distinguish changes in loan
quality that are deliberate changes in the Y vector from
other unobserved changes in loan characteristics. A short-
coming of this aggregation across vintages is that it risks
confusing changes in standards with changes in the his-
toric distribution of loans by vintage and their survival
rates.5
5 We also do not consider the possibility that the time fixed effects might
be due to changes in borrower behavior, such as an increased willingness to
default.
and include state fixed effects.
purch
D.R. Capozza, R.V. Order / Journal of Housing Economics 20 (2011) 141–151 145
ACV estimate equations of the form.
logðdtrÞ ¼
X�l
t¼�1
atmðr; tÞ þ drfr þ dtft þ
X�q
t¼0
ctut þ et ð6Þ
where bars over variables indicate a four-quarter moving
average of the variable, and l and q are lag lengths. From
(5) we should expect the sum of the coefficients of mðr; tÞ
in (6) to be close to one.
7. Results for MBA data: the relative roles of economic
conditions and underwriting
7.1. Simulations
ACV use their estimated equations along with the fixed
Because the mtr are the probability of ever defaulting
they do not apply to the same time period as dtr; and be-
cause there are lags in adjustment of dtr to changes in
mtr, ACV estimate versions of (5) where both dtr and mtr
are four quarter moving averages and the right hand side
has lags. ACV allow u to be an autoregressive process,
Fig. 4. US real house price appreciation, 1992–2010 seasonally adjusted,
effects to decompose foreclosure rates into a part due to
the economic multipliers and a part due to the year fixed
effects. The year fixed effects conditional on the multipliers
are the estimates of the underwriting component, i.e., of
default rates after controlling for economic conditions.
When normalized, the fitted values from the regression,
i.e., difference between the unconditional year indicators
(i.e., the actual yearly foreclosure rates) and the year indi-
cators conditional on economic conditions is an estimate of
the economic component. By construction the two add up
to the actual level of foreclosures. Fig. 7 presents results
using the ACV Model for all loans.
The yellow line gives the part due to economic condi-
tions (how foreclosures would have moved had underwrit-
ing not changed), which promoted declining foreclosures
until 2004. The pink line shows the contribution of under-
writing (how foreclosures would have changed had eco-
nomic conditions not changed), which was positive early
in the period, negative later and sharply positive in 2006
to 2007. For example, in Fig. 7 the red curve for underwrit-
ing in 2004 is 1.0 while actual defaults are 0.25 and the
economic conditions index is �0.75. The interpretation is
that while actual default rates rose 25% from 1990 to
2004, if economic conditions had not been so favorable,
foreclosures started would have risen by 100% instead of
25%. Stated differently, underwriting quality eroded en-
ough to double the level of foreclosures started by 2004;
but only a 25% increase was realized because favorable
economic conditions offset 3=4 of the potential increase.
During this period house prices appreciated steadily in
most of the country.
Note that the underwriting effects refer to the year in
which the loans are observed, not the year in which they
were originated. The poor underwriting results in 2006
and 2007 are for loans that were originated earlier. The fig-
ure suggests that the post 2005 increase in foreclosures can
be apportioned about equally between the underwriting
and economic conditions.
The spectacular increase in foreclosures after 2005 is
unprecedented in the data. Economic conditions and
underwriting quality typically moved in opposite direc-
tions in the 1990s. This negative correlation is consistent
ase only index, quarterly data, four quarter price changes. Source: FHFA.
with lenders becoming more conservative when economic
conditions are weak. However, after 2002–2005, economic
conditions and quality both deteriorated, breaking the ear-
lier pattern and suggesting a possible structural break or
regime shift in this market that is consistent with a moral
hazard story. The data suggest that the post 2005 increase
in foreclosures can be apportioned about equally between
the underwriting and economic conditions explanations.
7.2. Loan level model
In this section we present results for the loan level mod-
el, which uses the non-Agency data to estimate a full ver-
sion of Eq. (1). The data include prime, subprime and Alt-
A loans.6 While that data set is rich
本文档为【The great surge in mortgage defaults 2006–2009】,请使用软件OFFICE或WPS软件打开。作品中的文字与图均可以修改和编辑,
图片更改请在作品中右键图片并更换,文字修改请直接点击文字进行修改,也可以新增和删除文档中的内容。
该文档来自用户分享,如有侵权行为请发邮件ishare@vip.sina.com联系网站客服,我们会及时删除。
[版权声明] 本站所有资料为用户分享产生,若发现您的权利被侵害,请联系客服邮件isharekefu@iask.cn,我们尽快处理。
本作品所展示的图片、画像、字体、音乐的版权可能需版权方额外授权,请谨慎使用。
网站提供的党政主题相关内容(国旗、国徽、党徽..)目的在于配合国家政策宣传,仅限个人学习分享使用,禁止用于任何广告和商用目的。