|
||
Chapter
6
Using
Credit-Bureau
Scores
A
transformation
in
the
way
mortgage
applications
are
processed
is
beginning
to
reshape
the
residential
finance
industry.
Automated
underwriting’s
ability
to
evaluate
a
multitude
of
risk
factors
accurately
and
objectively
for
each
loan
application
makes
it
a
superior
underwriting
tool.
However,
the
move
to
automated
underwriting
will
not
occur
overnight.
Until
the
conversion
is
complete,
credit-bureau
scores
will
serve
as
an
important
bridge
between
traditional
loan-
assessment
methods,
which
rely
on
human
underwriters
to
weigh
myriad
pieces
of
information,
and
automated
underwriting.
Credit-bureau
scores
capture
only
one
dimension
of
the
lending
decision
and,
therefore,
lack
automated
underwriting’s
comprehensiveness
and
ability
to
account
for
the
layering
of
risk.
Nonetheless,
credit-bureau
scores
are
an
extremely
useful
tool
in
the
hands
of
human
underwriters.
Credit-Bureau
Scores
Improve
Manual
Underwriting
In
July
1995,
after
conducting
statistical
tests
that
demonstrated
the
predictive
power
of
credit-bureau
scores,
Freddie
Mac
recommended
that
lenders
use
these
scores
to
supplement
their
manual
underwriting
practices.
Three
months
later,
Fannie
Mae
similarly
endorsed
the
supplemental
use
of
credit-bureau
scores.
The
use
of
credit-bureau
scores
enables
lenders
to
expedite
their
reviews
of
borrowers’
credit
profiles.
Freddie
Mac
advises
lenders
that
applicants
with
FICO
scores
above
660
are
likely
to
have
acceptable
credit
reputations.
In
these
cases,
lenders
can
focus
on
verifying
the
consistency
of
the
information
provided
and
establishing
whether
additional
risk
exists
that
is
not
accounted
for
in
the
credit-bureau
score.
For
applicants
with
FICO
scores
between
620
and
660,
the
credit
profile
is
uncertain,
and
lenders
perform
the
same
review
as
they
would
when
underwriting
without
credit-bureau
scores.
FICO
scores
below
620
indicate
high
risk,
and
a
particularly
thorough
review
is
needed.1
Using
credit-bureau
scores
in
this
way
increases
both
the
accuracy
and
efficiency
of
manual
underwriting.
What
Creates
a
High-Risk
Score?
While
a
borrower’s
credit-bureau
score
depends
on
a
number
of
factors,
the
most
important
is
the
borrower’s
proven
willingness
and
ability
to
repay
debts.
High-risk
scores
go
to
those
applicants
with
a
history
of
repayment
problems,such
as
bankruptcies
or
chronic
late
payments.
A
Freddie
Mac
analysis
of
a
sample
of
25,000
loans
insured
by
FHA
illustrates
the
importance
of
these
risk
factors.
For
this
purpose,
Freddie
Mac
defined
high-risk
borrowers
as
those
with
FICO
scores
below
620
and
lower-risk
borrowers
as
those
with
FICO
scores
above
660.
Exhibit
10
reveals
key
differences
between
these
two
groups.
For
example,
88
percent
of
high-risk
borrowers
showed
a
previous
60-day
delinquency,
while
35
percent
were
currently
60
days
past
due
on
a
loan
payment.
In
contrast,
17
percent
of
lower-risk
borrowers
had
fallen
60
days
behind
in
the
past,
and
just
1
percent
were
currently
60
days
delinquent.
Along
with
a
history
of
missed
payments,
a
jump
in
recent
credit
activity
bears
close
scrutiny.
The
addition
of
several
new
credit
accounts
or
the
maximum
use
of
existing
credit
lines
may
jeopardize
a
borrower’s
ability
to
handle
a
new
mortgage
obligation.
Borrowers
with
high-risk
scores
are
much
more
likely
to
exhibit
signs
of
overextension.
For
the
sample
of
borrowers
illustrated
in
Exhibit
10,
23
percent
had
used
80
percent
or
more
of
their
available
credit,
and
20
percent
showed
more
than
four
inquiries
for
new
credit
in
the
past
year.
In
contrast,
lower-risk
borrowers
were
far
less
likely
to
show
signs
of
expanding
credit
use:
only
6
percent
had
tapped
into
80
percent
or
more
of
their
available
credit,
and
only
5
percent
had
more
than
four
inquiries
for
new
credit.2
Who
Has
Riskier
Scores?
By
and
large,
the
majority
of
applicants
have
established
adequate
credit
histories
to
qualify
for
mortgage
loans.
Nonetheless,
policymakers
and
the
housing
finance
industry
are
raising
concerns
about
borrowers
with
high-risk
scores,
particularly
in
light
of
various
studies
indicating
that
minority
families
tend
to
experience
greater
credit
difficulties.3
A
Freddie
Mac
analysis
of
the
distribution
of
credit-bureau
scores
confirms
these
findings.
African-American
borrowers,
for
example,
were
about
three
times
as
likely
to
have
high-risk
credit-bureau
scores—defined
as
FICO
scores
below
620—as
were
White
borrowers,
based
on
Freddie
Mac’s
1994
mortgage
purchases.
Hispanic
borrowers
were
about
twice
as
likely
as
White
borrowers
to
have
high-risk
scores.4
While
credit-bureau
scores
may
vary
across
racial
and
ethnic
groups,
their
predictive
power
does
not.
Based
on
FreddieMac’s
1994
loan
purchases,
for
example,
Exhibit
11
shows
that—whether
the
borrower
was
African-
American,
Hispanic
or
White—loans
for
borrowers
with
FICO
scores
greater
than
660
performed
better
than
loans
for
borrowers
with
scores
between
620
and
660,
which
in
turn
performed
better
than
loans
with
FICO
scores
below
620.
FICO
scores
also
are
highly
predictive
across
income
groups.
For
the
same
sample,
Exhibit
12
shows
that
loans
to
borrowers
with
FICO
scores
above
660
and
between
620
and
660
outperformed
loans
to
borrowers
with
scores
below
620
for
all
income
levels.
Addressing
Key
Questions
Inaccuracies.
How
can
borrowers
be
sure
that
their
credit-bureau
scores
are
based
on
accurate
information?
Currently,
the
nation’s
three
main
credit
repositories
maintain
a
total
of
150
million
to
200
million
individual
credit
files.
Keeping
the
files
accurate
and
up-to-date
is
vital
to
the
reliability
of
the
statistical
scores
used
to
rate
borrowers
with
this
information.
Fortunately,
credit
grantors
and
credit
information
providers
share
the
same
need
as
consumers
for
complete
and
accurate
credit
files.5
For
credit
grantors,
accurate
credit
files
are
critical
to
sound
lending
decisions.
For
credit
information
providers,
reliable
information
is
what
keeps
them
in
business.
Laws
passed
by
Congress
more
than
a
generation
ago
provide
further
assurances
that
credit
files
will
be
well
maintained.
These
laws
protect
the
privacy
of
credit
information
and
give
consumers
the
tools
to
require
credit
repositories
to
correct
inaccuracies.
(See
Consumer
Protection
and
Credit
Records)
In
the
event
that
some
consumer
credit
records
still
contain
errors,
Freddie
Mac
advises
lenders
to
review
evidence
that
the
loan
applicant
provides
about
inaccuracies.
Lenders
should
incorporate
this
information,
as
appropriate,
when
underwriting
the
loan.
Representativeness.
Are
minority
households
sufficiently
represented
in
the
samples
of
loans
used
to
compile
credit-bureau
scores?
In
a
recent
report,
FICO-score
developer
Fair,
Isaac
and
Company,
Inc.
assessed
the
degree
to
which
the
loan
repayment
experience
of
minority
borrowers
is
adequately
represented
in
credit
repository
data.
The
study
found
that
residents
of
“high-minority
areas”
account
for
7.8
percent
of
adults
18
years
and
older
and
6.7
percent
of
consumers
with
credit
reports
maintained
by
credit
repositories.
Fair,
Isaac
concluded
that,
while
the
figures
indicate
a
slight
underrepresentation
of
high-minority-area
residents
in
the
records
of
credit
repositories,
the
data
clearly
cover
a
significant
number
of
minority
households.6
The
fact
that
credit-bureau
scores
are
powerful
predictors
for
minority
borrowers
confirms
that
they
are
sufficiently
represented.
Finance
company
use.
Does
prior
use
of
finance
companies
by
minority
households
push
their
credit-bureau
scores
into
a
high-risk
category? Fair, Isaac’s recent report also examined the use of finance companies. It detected little variation in the number of finance company accounts used by consumers regardless of the racial composition of their neighborhoods.
(See
Use
of
Finance
Companies)
Freddie
Mac
reached
a
similar
conclusion
based
on
an
analysis
of
25,000
FHA
loans.
The
use
of
finance
companies
was
similar
among
African-American,
Hispanic
and
White
borrowers,
as
well
as
across
income
groups.
Nontraditional
credit.
What
happens
when
families
without
established
credit
histories
apply
for
a
loan?
The
mortgage
industry
has
made
strides
in
recent
years
to
find
alternative
ways
to
evaluate
and
consider
the
small
number
of
households
who
have
yet
to
establish
credit.
Some
lenders,
for
example,
have
begun
to
rely
on
payment
histories
for
rent,
utilities
and
other
recurring
obligations.
However,
credit
repositories
do
not
routinely
collect
payment
information
from
landlords
or
utility
providers,
making
this
information
difficult
to
verify
efficiently.
In
cases
where
sufficient
traditional
information
is
unavailable,
Freddie
Mac
encourages
lenders
to
use
alternate
ways
of
evaluating
credit
to
accommodate
the
needs
of
these
borrowers.
Fortunately,
consumers
are
able
to
go
from
no
credit
history
to
an
acceptable
one
relatively
quickly,
in
perhaps
one
or
two
years.
To
do
this,
they
need
to
open
and
use
several
credit
accounts,
and
make
timely
payments
without
running
up
large
balances.
As
an
interim
step
to
the
widespread
adoption
of
automated
underwriting,
credit-bureau
scores
can
help
simplify
and
improve
the
mortgage-approval
process.
Automated
underwriting
builds
on
this
accuracy
and
efficiency.    
Footnotes:
|
||
|
In
contrast,
credit-bureau
scores
appear
to
vary
less
across
income
groups.
Based
on
the
same
data,
for
example,
borrowers
earning
less
than
80
percent
of
area
median
income
were
only
slightly
more
likely
to
have
high-risk
scores
than
borrowers
with
incomes
exceeding
120
percent
of
area
median.
The
growing
reliance
on
credit-bureau
scores
for
all
types
of
lending
has
raised
a
number
of
concerns.
Addressing
them
confirms
the
value
of
this
tool
to
supplement
manual
underwriting.
1.