1 Assessment Task – Tutorial Questions Assignment 1 .docx

1
Assessment Task – Tutorial Questions Assignment 1
Unit Code: HA1020
Unit Name: Accounting Principles and Practices
Assignment: Tutorial Questions Assignment 1
Due: 11:30pm 15th May 2020
Weighting: 25%
Total Assignment Marks: 50 Marks
Purpose: This assignment is designed to assess your level of
knowledge of the key topics covered in
this unit
Unit Learning Outcomes Assessed:
1. Understand the logic and assumptions of accounting
procedures;
2. Record business transactions in the journals and ledgers that
make up a business accounting
system;
3. Prepare financial statements; and
4. Analyse and interpret financial statements.

Description: Each week students were provided with three
tutorial questions of varying degrees of
difficulty. These tutorial questions are available in the Tutorial
Folder for each week on Blackboard.
The Interactive Tutorials are designed to assist students with
the process, skills and knowledge to
answer the provided tutorial questions. Your task is to answer a
selection of tutorial questions from
weeks 1 to 5 inclusive and submit these answers in a single
document.
The questions to be answered are:
Week 1 Question (10 marks)
(Note this question is 1.3 in the Pre-recorded Tutorial
Questions)
Compare and contrast Financial Accounting with Management
Accounting. Specify at least three (3)
areas where Financial Accounting and Management Accounting
are different. Support your answer
with examples. (10 marks)

2
Questions)
Select two (2) of the following four (4) financial accounting
assumptions listed below and explain in
your own words the meaning of each one you have selected: (10
marks)
(a.) accounting entity
(b.) accounting period
(c.) monetary
(d.) historical
Questions)
Which of the following events listed below results in an
accounting transaction for Clothing Ltd?
State a reason if it is not an accounting transaction.
1. Clothing Ltd signed a contract to hire a new store manager
for a salary of $150,000 per annum.
The manager will start work next month.

2. The founder of Clothing Ltd., who is also a major
shareholder, purchased additional stock in
another company.
3. Clothing Ltd borrowed $230,000 from a local bank.
4. Clothing Ltd purchased a sewing machine, which it paid for
by signing a note payable.
5. Clothing Ltd issued 10,000 shares to a private investor, who
is also a car business owner, in
return for a new delivery truck.
6. Two investors in Clothing Ltd sold their stock to another
investor.
7. Clothing Ltd ordered some fabric to be delivered next week.
8. Clothing Ltd lent $250,000 to a member of the board of
directors.

3
Questions)
The financial year end for Riverwood Ltd is 30 June.
a. Prepaid insurance as at 1 July 2015 was $4,000. This
represents the cost of one year’s insurance
policy that expires on 30 June 2016.
b. Commissions to sales personnel for the five day working
week ending 2 July 2016, totaling $9,600,
will be paid on 2 July.
c. Sales revenue for the year included $570 of customer deposits
for products that have not yet been
shipped to them.
d. A total of $900 worth of stationery was charged to the office
supplies expense during the year.
On 30 June about $490 worth of stationery is still
considered useful for next year.
e. The company has a bank loan and pays interest annually (in
arrears) on 31 December.
The estimated total interest cost for the calendar year
ended 31 December 2016 is $500.

Required:
(a.) Show the effect of each of the situations above (a. – e.) on
the accounting equation on 30 June
2016. (5 marks)
(b.) Provide the adjusting journal entry for each of the
situations above (a. – e.) on 30 June 2016.
(5 marks)
4

Questions)
SRP Ltd has the following unadjusted trial balance as at 31 Dec
2015.
Account Titles Debit ($) Credit ($)
Cash 19,600
Accounts Receivable 7,000
Supplies 1,300
Prepaid Insurance 900
Equipment 27,000
Accumulated Depreciation 12,000
Other Assets 5,100
Accounts Payable 7,500
Share Capital (3000 shares outstanding all year) 16,000
Retained Profit 10,300
Service Revenue 48,000
COGS 32,900

Total 93,800 93,800
Note: Data not yet recorded as at 31 Dec 2015 includes the
following five (5) transactions:
1) Depreciation expense for 2015 was $3,000.
2) Insurance expired during 2015 was $450.
3) Wages earned by employees but not yet paid on 31 December
2015 was $2,100.
4) The supplies count on 31 December 2015 reflected $800
remaining supplies on hand to be
used in 2016.
5) Income tax expense was $3,150.
Required: (10 marks)
1) Record the 2015 adjusting entries.
2) Prepare an income statement and a classified balance sheet
for 2015 to include the effect of
the five (5) transactions listed above.
3) Prepare closing entries.

5
Submission Directions:
The assignment has to be submitted via Blackboard. Each
student will be permitted one submission
to Blackboard only. Each student needs to ensure that the
document submitted is the correct one.
Academic Integrity
Academic honesty is highly valued at Holmes Institute. Students
must always submit work that
represents their original words or ideas. If any words or ideas
used in a class posting or assignment
submission do not represent the student’s original words or
ideas, the student must cite all relevant
sources and make clear the extent to which such sources were
used. Written assignments that include
material similar to course reading materials or other sources
should include a citation including source,
author, and page number.
In addition, written assignments that are similar or identical to
those of another student in the class is
also a violation of the Holmes Institute’s Academic Conduct
and Integrity Policy. The consequence for

a violation of this policy can incur a range of penalties varying
from a 50% penalty through to
suspension of enrolment. The penalty would be dependent on
the extent of academic misconduct
and the student’s history of academic misconduct issues. All
assessments will be automatically
submitted to Safe-Assign to assess their originality.
Further Information:
For further information and additional learning resources,
students should refer to their Discussion
Board for the unit.
CHAPTER 8 Searching for Mortgage Information Online
131
Chapter 8
IN THIS CHAPTER
» Looking at some safe surfing ideas
» Checking out mortgage sites
Searching for Mortgage
Information Online
Computers, tablets, and smartphones are amazing tools. Used
wisely, they may save you time and money. However, like other
tools (such as a ham-mer), used incorrectly (remember the last
time you whacked your finger
with a hammer?) or for the wrong purpose (tapping a glass

window comes to
mind), today’s technology can cause more harm than good.
Some people have mistaken assumptions about using their
computers and tablet
or phone apps to help them make important financial decisions.
Some believe and
hope that fancy technology can solve their financial problems or
provide unique
insights and vast profits. Often, such erroneous musings
originate from propa-
ganda put forth through “fake news” or social media about how
all your problems
can easily be solved if you just have the right app, spend more
time on particular
websites, and so on.
As computers, technology, and apps continue to proliferate, we
take seriously our
task of explaining how, where, and when to use the Internet to
help you make
important mortgage decisions. In this chapter, we highlight key
concepts and
issues for you to understand as well as list a few of our favorite
websites.
Griswold, R. S., Tyson, E., & Tyson, E. (2017). Mortgage
management for dummies. Retrieved from
http://ebookcentral.proquest.com
Created from apus on 2020-05-04 20:26:39.
C
op
yr
ig

ht
©
2
01
7.
J
oh
n
W
ile
y
&
S
on
s,
In
co
rp
or
at
ed
. A
ll
rig

ht
s
re
se
rv
ed
.
132 PART 3 Landing a Lender
Obeying Our Safe Surfing Tips
Before we get to specific sites that are worthy of your time, in
this section we
provide an overview of how we suggest using (and not being
abused by) your
mortgage-related web surfing or cure-all app. Specific sites, and
especially apps,
will come and go, but these safe surfing tips should assist you
with assessing any
site or app that you may stumble upon.
Shop to find out about rates and programs
The best reason that we can think of to access the Internet when
you’re looking
for a mortgage is to discover more about the going rate for the
various types of
loans you’re considering. Despite all the cautions we raise in
this chapter, shop-
ping for a mortgage online has some attractions:

» No direct sales pressure: Because you don’t speak or meet
with a mortgage
officer (who typically works on commission) when you peruse
mortgage rates
online, you can do so without much pressure. That said, some
sites and apps
are willing to give out specific loan information only after you
reveal a fair
amount of information about yourself, including how to get in
touch with you.
However, on one site where you must register (with all your
contact informa-
tion and more) to list your loan desires, take a look at how the
site pitches
itself to prospective mortgage lenders: “FREE, hot leads! Every
lead is HOT,
HOT, HOT because the borrower has paid us a fee to post their
loan request.”
Although the advantages of online shopping are many, being
savvy and
discrete with who and how you contact prospective lenders is
worthy of a
cautionary reminder. You may think you’re the one shopping,
but on many
sites and apps, you are the one being “sold” to aggressive
marketers of loan
products that may not be what you need. Worse yet, many of
these unscrupu-
lous hucksters don’t even have the loan products and terms they
tease on
their website and their real goal is to lure you in and then turn
around and sell
your information to others. You’ll soon find yourself inundated
with unwanted
emails, texts, and even phone calls.

» Shop when you like: Because most people work weekdays
when lenders and
mortgage brokers are available, squeezing in calls to lenders is
often difficult.
Thus, another advantage of mortgage Internet shopping is that
you can do it
any time of any day when it’s convenient for you. Just be
careful that you don’t
provide personal information to anyone unless you’re sure you
want him to
contact you.
C
op
yr
ig
ht
©
2
01
7.
J
oh
n

W
ile
y
&
S
on
s,
In
co
rp
or
at
ed
. A
ll
rig
ht
s
re
se
rv
ed
.

133
Quality control is often insufficient
Particularly at sites where lenders simply pay an advertising fee
to be part of the
program, you should know that quality control may be
nonexistent or not up to
your standards. “We make your loan request available to every
online lender in the
world,” boasts one online mortgage listing service. We don’t
know too many bor-
rowers willing to work with just any old mortgage company!
Some sites don’t
check to see whether a participating lender provides a high level
of service or
meets promises and commitments made to previous customers.
Again, if you’re going to go loan shopping on the Internet,
examine each site to
see how it claims to review listed lenders. One site we’re
familiar with claims to
demand strict ethics from the companies it lists — no
lowballing or bait-and-
switch tactics — and says it has removed several dozen lenders
from its list for
such violations. That makes us think that the site should do a
better job of screen-
ing lenders upfront!
Beware simplistic affordability calculators
Be highly skeptical of information about the mortgage amount
that you can afford.
Most online mortgage calculators simplistically use overall
income figures and the

current loan interest rate to calculate the mortgage amount a
borrower can
“afford.” These calculators are really spitting out the maximum
a bank will lend
you based on your income. As we discuss in Chapter 1, this
figure has nothing to
do with the amount you can really afford.
Such a simplistic calculation ignores your larger financial
picture: how much (or
little) you have put away for other long-term financial goals
such as retirement or
college educations for your children. Thus, you need to take a
hard look at your
budget and goals before deciding how much you can afford to
spend on a home;
don’t let some slick Java-based calculator make this decision
for you.
Don’t reveal confidential information
unless . . .
Suppose that you follow all our advice in this chapter, and you
find your best
mortgage deal online. You may find yourself solicited to apply
for your mortgage
online as well. However, as you gather your confidential
financial documents, you
may have an unsettling feeling and wonder just how safe and
wise it is to be
entering this type of information into an Internet site.

C
op
yr
ig
ht
©
2
01
7.
J
oh
n
W
ile
y
&
S
on
s,
In
co
rp
or
at

ed
. A
ll
rig
ht
s
re
se
rv
ed
.
We applaud your instincts and concerns! Here’s what you
should do to protect
yourself:
» Do your homework on the business. In Chapter 7, we suggest
a variety of
questions to ask and issues to clarify before deciding to do
business with any
lender — online or offline.
» Review the lender’s security and confidentiality policies. On
reputable
lender websites, you’ll be able to find the lender’s policies
regarding how it
handles the personal and financial information you may share

with it. We
recommend doing business only with sites that don’t sell or
share your
information with any outside organization other than for the
sole purpose of
verifying your creditworthiness needed for loan approval. Be
sure to choose
secure sites that prevent computer hackers from obtaining the
information
you enter.
If you’re simply not comfortable — for whatever reason —
applying for a loan
online, know that most online mortgage brokers and lenders
offer users the abil-
ity to apply for their loan offline (at an office or via loan papers
sent through the
regular mail). They may charge a slightly higher fee for this
service, but if it makes
you feel more comfortable, consider it money well spent.
Be sure to shop offline
You may find your best mortgage deal online. However, you
won’t know it’s the
best unless and until you’ve done sufficient shopping offline as
well. Why shop
offline? You want to be able to see all your options and find the
best one. Online
mortgage options aren’t necessarily the cheapest or the best.
What good is a quote
for a low mortgage rate that a lender doesn’t deliver on or that
you won’t qualify
for because of your specific property, location, or financial
situation? Remember:
Personal service and honoring commitments is highly important.

You may be able to save a small amount of money by taking a
mortgage you find
online. Some online mortgage brokers are willing to take a
somewhat smaller slice
of commission for themselves if they feel they’re saving time
and money process-
ing your loan via an online application. As we discuss in
Chapter 7, mortgage
brokers’ fees do vary and are negotiable. Some online mortgage
brokers are will-
ing to take less than the industry standard cut (1-plus percent).
But just because you’ve been offered a slightly better rate
online, you shouldn’t
necessarily jump on it. Local lender or mortgage brokers may
negotiate with you
to make themselves competitive. However, you have to give
them the opportunity
to do so. Other things being equal, go back to the runner-up on
price and give
C
op
yr
ig
ht
©
2

01
7.
J
oh
n
W
ile
y
&
S
on
s,
In
co
rp
or
at
ed
. A
ll
rig
ht
s
re

se
rv
ed
.
135
them a chance to meet or beat your best offer. You may be
pleasantly surprised
with the results.
Mortgage websites and apps are best used to research the
current marketplace
rather than to actually apply for and secure a mortgage. The
reason: Mortgage
lending is still largely a locally driven and relationship-based
business that varies
based on nuances of a local real estate market.
Beware of paid advertising masquerading
as directories
Some sites on the Internet and apps offer “directories” of
mortgage lenders. Most
sites charge lenders a fee to be listed or to gain a more visible
listing. And, just as
with any business buying a yellow pages listing or Google ad,
higher visibility ads
cost more. Here’s how one online directory lured lenders to
advertise on its site:
Sure, our basic listing is free, but we have thousands of

mortgage companies in
our directory. A free listing is something like a five-second
radio advertisement at
2:00 a.m. on an early Sunday morning. To make your listing
really work for you, you
must upgrade your listing.
Upgrade, here, is a code word for pay for it! For example, a
“gold listing” on this site
costs $600 per year for one state and $360 for each additional
state. What does
that amount of money get the lender?
A Gold Listing sorts your company name to the top of all
listings. In addition, the
Gold Listings receive a higher typeface font and a Gold Listing
icon next to their
name.
Then there is the “diamond listing,” the “platinum listing,” the
“titanium list-
ing,” and you get the idea.
On another directory site, you can find a “directory
enhancement program,”
which for $125 per year enabled a lender to buy a boldface
listing and for $225 per
line per year place descriptive text under the listing. Thus,
prospective borrowers
visiting these sites are looking at the mortgage equivalent of an
online Yellow
Pages advertising directory rather than a comprehensive or low-
cost lender
directory.
If you’re considering using an Internet site or app to shop for a

mortgage, first
investigate the way the site derived the list of lenders. If the
site isn’t upfront
about disclosing this information, be suspicious. Do some
sleuthing like we did;
click on the buttons at the site that solicit lenders to join the
fray. Here you can
C
op
yr
ig
ht
©
2
01
7.
J
oh
n
W
ile
y
&

S
on
s,
In
co
rp
or
at
ed
. A
ll
rig
ht
s
re
se
rv
ed
.
find out how the site attracts lenders and you may also find the
amount lenders

are paying to be listed.
Perusing Our Recommended
Mortgage Websites
In addition to seeking only the highest-quality sources for you,
dear reader, we
don’t want you wasting your time on a wild goose chase for
some unreliable
website or app that’s here today and gone tomorrow. In this
section, we recom-
mend a short list of our favorite mortgage sites. Yes, many more
sites and apps are
out there, but we don’t want to bore you with a huge laundry
list of mortgage-
related sites. And, please remember as we discuss in Chapter 7,
mortgages are
distributed through numerous types of mortgage lenders and
brokers. The Inter-
net and the app craze are just simply another way that these
players can reach
prospective customers.
Useful government sites
Various government agencies provide assistance to low-income
homebuyers as
well as veterans. The U.S. Department of Housing and Urban
Development’s web-
site (see Figure 8-1) at www.hud.gov provides information on
the federal govern-
ment’s FHA loan program as well as links to listings of HUD
and other government
agency–listed homes for sale (foreclosed homes for which the
owners had FHA
loans; see
https://portal.hud.gov/hudportal/HUD?src=/topics/homes_for_

sale). On this site, you can also find links to other useful federal
government
housing–related websites.
Also, if you’re a veteran, check out the VA’s website (see
Figure 8-2; www.
benefits.va.gov/homeloans) operated by the U.S. Department of
Veterans
Affairs. In addition to information on VA loans, veterans and
nonveterans alike are
eligible to buy foreclosed properties on which there was a VA
loan (see the website
http://listings.vrmco.com).
The Federal Citizen Information Center
(www.pueblo.gsa.gov/housing.htm)
offers numerous free and low-cost pamphlets on home financing
topics such as
securing home equity loans, avoiding loan fraud, finding
mortgages and home
improvement loans to make your home more energy efficient,
and qualifying for
a low down payment mortgage. You also want to know the
required lender disclo-
sures so you know what the lender must tell you and what it
means.
C
op
yr

ig
ht
©
2
01
7.
J
oh
n
W
ile
y
&
S
on
s,
In
co
rp
or
at
ed
. A
ll

rig
ht
s
re
se
rv
ed
.
http://www.hud.gov
https://portal.hud.gov/hudportal/HUD?src=/topics/homes_for_sa
le
https://portal.hud.gov/hudportal/HUD?src=/topics/homes_for_sa
le
http://www.benefits.va.gov/homeloans
http://www.benefits.va.gov/homeloans
http://listings.vrmco.com
http://www.pueblo.gsa.gov/housing.htm
137
FIGURE 8-1:
The U.S.
Department of
Housing and
Urban Develop-
ment website

provides
information on
FHA loan
programs and
HUD homes
for sale.
Source: U.S. Department of Housing and Urban Development
FIGURE 8-2:
Visit the
U.S. Department
of Veterans
Affairs website
for information
on VA loans.
Source: U.S. Department of Veterans Affairs
C
op
yr
ig
ht
©

2
01
7.
J
oh
n
W
ile
y
&
S
on
s,
In
co
rp
or
at
ed
. A
ll
rig
ht
s

re
se
rv
ed
.
Fannie Mae (www.fanniemae.com) has many resources for
mortgage borrowers
and homebuyers. In addition to helping you find mortgage
lenders for home pur-
chases, improvements, or refinances, the site can also turn you
onto helpful
worksheets and counseling agencies. Freddie Mac
(www.freddiemac.com) offers
similar (although not as extensive) resources.
Finally, if you’re trying to fix your problematic credit report,
don’t waste your
money on so-called credit-repair firms, which often
overpromise — and charge
big fees for doing things that you can do yourself. In addition to
following our
credit-fixing advice in Chapters 2 and 3, also check out the
Federal Trade Com-
mission’s website (www.ftc.gov) for helpful credit-repair and
other relevant
advice regarding borrowing.
Mortgage information and shopping sites

HSH Associates (www.hsh.com) is the nation’s largest collector
and publisher of
mortgage information. If you’re a data junkie, you’ll enjoy
perusing the HSH site,
which includes up-to-date mortgage rates and graphs showing
recent trends (see
Figure 8-3).
Some lenders do choose to advertise online at HSH’s website
and you can obtain
their rates through the website’s ad links.
FIGURE 8-3:
The website of
HSH Associates,
publisher of
mortgage
information.
Source: HSH Associates
C
op
yr
ig
ht
©

re
se
rv
ed
.
http://www.fanniemae.com
http://www.freddiemac.com
http://www.ftc.gov
http://www.hsh.com
139
Many online mortgage brokers and lenders provide rate quotes
and assist with
your loan shopping. The interactive features of some sites even
allow prospective
borrowers to compare the total cost of loans (including points
and fees) under dif-
ferent scenarios (how long you keep the loan and what happens
to the interest
rate on adjustable-rate mortgages). Interpreting these
comparisons, however,
requires a solid understanding of mortgage lingo and pricing.
Two other sites that we like are www.bankrate.com and
www.realtor.com. Bank
Rate’s site offers lots of information and perspectives on many
types of consumer
loans including mortgages. Realtor.com’s Mortgage section is
more focused on

mortgages. On both sites, you can shop for specific mortgages.
C
op
yr
ig
ht
©
2
01
7.
J
oh
n
W
ile
y
&
S
on
s,
In

co
rp
or
at
ed
. A
ll
rig
ht
s
re
se
rv
ed
.
http://www.bankrate.com
http://www.realtor.com
C
op

yr
ig
ht
©
2
01
7.
J
oh
n
W
ile
y
&
S
on
s,
In
co
rp
or
at
ed
. A

ll
rig
ht
s
re
se
rv
ed
.
2015 V43 4: pp. 993–1034
DOI: 10.1111/1540-6229.12105
REAL ESTATE
ECONOMICS
A Tale of Two Tensions: Balancing Access
to Credit and Credit Risk in Mortgage
Underwriting
Marsha J. Courchane,* Leonard C. Kiefer** and Peter M.
Zorn***
Over the years 2000–2007, mortgage market underwriting
conditions eased in
response to public policy demands for increased
homeownership. This eas-
ing of acceptable credit risk in order to accommodate increased
access to

credit, when coupled with the unanticipated house price
declines during the
Great Recession, resulted in substantial increases in
delinquencies and fore-
closures. The response to this mortgage market crisis led to
myriad changes in
the industry, including tightened underwriting standards and
new market regu-
lations. The result is a growing concern that credit standards are
now too tight,
restricting the recovery of the housing market. Faced with this
history, policy an-
alysts, regulators and industry participants have been forced to
consider how
best to balance the tension inherent in managing mortgage
credit risk without
unduly restricting access to credit. Our research is unique in
providing explicit
consideration of this trade-off in the context of mortgage
underwriting. Using
recent mortgage market data, we explore whether modern
automated under-
writing systems (AUS) can be used to extend credit to
borrowers responsibly,
with a particular focus on target populations that include
minorities and those
with low and moderate incomes. We find that modern AUS do
offer a potentially
valuable tool for balancing the tensions of extending credit at
acceptable risks,
either by using scorecards that mix through-the-cycle and stress
scorecard ap-
proaches or by adjusting the cutpoint—more relaxed cutpoints
allow for higher
levels of default while providing more access, tighter cutpoints
accept fewer

borrowers while allowing less credit risk.
Introduction
U.S. residential mortgage markets changed dramatically during
the past sev-
eral years. In the early 2000s, public policy focused on
expanding credit
access and homeownership and specifically targeted a reduction
in the home-
ownership gap between minority and non-minority households
and between
*Charles River Associates or [email protected]
**Freddie Mac or [email protected]
***Freddie Mac or [email protected]
C© 2015 American Real Estate and Urban Economics
Association
994 Courchane, Kiefer and Zorn
higher and lower income families.1 Relaxation of underwriting
standards, ac-
companied by a surge in subprime lending and an attendant
proliferation of
new products, resulted in many borrowers who could not meet
traditional un-
derwriting standards being able to obtain home mortgages and
achieve home
ownership.
However, the environment changed with the mortgage market
crisis of 2007
and 2008 when the subprime sector collapsed nearly entirely

and delinquency
and foreclosure rates increased throughout the country. In
response, underwrit-
ing standards tightened and legislation was passed imposing
more stringent
regulations on the mortgage industry, particularly the Dodd-
Frank Act reg-
ulations, which introduced both Qualified Mortgage (“QM”) and
Qualified
Residential Mortgage (“QRM”) standards. While providing
assurance that
the performance of recent mortgage originations will reduce the
likelihood of
another housing crisis, this tightening of standards comes at a
significant cost
in terms of access to credit. Balancing the tension between
access to credit
and the management of credit risk remains an ongoing concern.
The rich history of mortgage performance data over this period
offers an
opportunity to better distinguish mortgage programs and
combinations of
borrower and loan characteristics that perform well in stressful
economic en-
vironments from those that do not. The relaxed underwriting
standards of the
2000s provide plentiful performance information on borrowers
who stretched
for credit, but then experienced the stressful post-origination
environment of
declining house prices and rising unemployment. While many of
these loans
performed poorly, a large number performed well. Our goal is to
identify the
characteristics that distinguish between these two groups.

We specifically examine whether the recent data can be used to
create a mod-
ern automated underwriting scorecard that effectively and
responsibly extends
mortgage credit to the general population, and to underserved or
targeted bor-
rowers who reside in low-income communities, make low down
payments
and have poorer credit histories. Our analysis focuses on
mortgage under-
writing, rather than mortgage pricing. This reflects the two-
stage approach
to mortgage lending broadly practiced in the United States—
originators first
underwrite applications to determine whether they qualify for
origination, and
then price the loans that are originated successfully.
1For example, former United States Department of Housing and
Urban Development
(HUD) Secretary Mel Martinez states in 2002 that “The Bush
Administration is
committed to increasing the number of Americans, particularly
minorities, who own
their own homes.”
A Tale of Two Tensions 995
There are four steps necessary to complete this exercise. First,
we empirically
estimate a mortgage delinquency model. Second, we convert the
estimated
delinquency model to an underwriting scorecard for assessing

risk, where
higher scores signify higher risk. Third, we determine a
scorecard value (a
“cutpoint” or risk threshold) that demarcates the marginal risk
tolerance—
score values equal to or below the cutpoint are viewed as
acceptable risk;
score values above the cutpoint are not. Fourth, we process
borrowers through
this prototype of an automated underwriting system. We then
determine the
proportion of the population of mortgage applicants that is
within acceptable
risk tolerances, and the historic performance of these
“acceptable” loans.
The main data we use for this analysis are loan-level
observations from
CoreLogic on mortgages originated in the conventional (prime
and subprime)
and government sectors from 1999 through 2009. For each of
the three market
sectors, we separately estimate the probability that borrowers
will become
90-days or more delinquent on their loans within the first three
years after
origination. Included in the model are standard controls for
borrower and loan
characteristics, as well as for key macroeconomic factors
affecting mortgage
performance post-origination (specifically, changes in house
prices, interest
rates and unemployment rates).
Underwriting scorecards provide ex ante assessments of
mortgage risk at

origination, so creating scorecards requires appropriate
treatment of the
post-origination variables in our estimated models. Two broad
approaches are
possible. One approach attempts to forecast post-origination
variables across
borrower locations and over time. The other approach sets post-
origination
variables to constant values for all borrowers and all time
periods. We use
the latter approach. Specifically, we create two separate
scorecards. The first
scorecard sets post-origination values of house prices, interest
rates and un-
employment rates to their constant long run average levels (a
“through-the-
cycle” scorecard). The through-the-cycle scorecard is inherently
“optimistic”
with respect to credit risk, and therefore reflects a focus on
access to credit.
The second scorecard sets post-origination values of house
prices, interest
rates and unemployment rates to the varying ex post values
defined by the
Federal Reserve in an adverse scenario (a “stress” scorecard) as
defined in the
2014 supervisory stress test for very large banking
organizations.2 The stress
scorecard focuses on “tail” events that are unlikely to occur and
is meant to
prevent crisis outcomes such as those observed during the Great
Recession.
This scorecard therefore represents a focus on credit risk
management.
2See http://www.federalreserve.gov/bankinforeg/stress-

tests/2014-appendix-a.htm.
The next challenge requires choosing appropriate scorecard
cutpoints for
delimiting loans within acceptable risk tolerances. This, in
combination with
the choice of scorecard, is where much of the tension between
credit access
and credit risk resides. Higher cutpoints provide greater access
at the cost of
increasing credit risk; lower cutpoints limit credit risk but
restrict access.
As the choice of a cutpoint is a complicated policy/business
decision, we
provide results for a variety of possible cutpoints, ranging from
a low of a
5% delinquency rate to a high of a 20% delinquency rate. In an
effort to put
forward a possible compromise between access and credit risk,
we explore in
more detail results for alternative cutpoints that are market-
segment-specific;
5% for prime loans, 15% for subprime loans and 10% for
government loans.
We argue that these values represent reasonable risk tolerances
by approxi-
mating the observed delinquency rates in these segments
between 1999 and
2001.
The combination of scorecards and cutpoints creates working

facsimiles of
modern AUS, and we apply these systems to both the full and
target pop-
ulations.3 For this exercise, our “target” population is defined
as borrowers
with loan-to-value (“LTV”) ratios of 90% or above, with FICO
scores of 720
or below or missing, and who are located in census tracts with
median in-
comes below 80% of area median income. This group is
generally reflective
of “underserved” borrowers for whom there is particular policy
concern.
We find that automated underwriting, with a judicious
combination of score-
card and cutpoint choice, offers a potentially valuable tool for
balancing the
tensions of extending credit at acceptable risks. One approach
entails using
scorecards that mix the through-the-cycle and stress scorecard
approaches
to post-origination values of key economic variables. Moving
closer to a
through-the-cycle scorecard provides more focus on access to
credit. Moving
closer to a stress scorecard provides more focus on the control
of risk. The
second approach is to adjust the cutpoint—more relaxed
cutpoints allow for
higher levels of default while providing more access, tighter
cutpoints have
accept fewer borrowers while allowing less credit risk.
Previous Literature

A considerable body of research has examined outcomes from
the mortgage
market crisis during the past decade. Of particular relevance for
this research
3We weight the data using weights based on the proportion of
the target population in
the Home Mortgage Disclosure data (“HMDA”) to ensure that
the target population
in our data is representative of the target population in HMDA.
This allows us to draw
inferences to the full population.
are studies that examine specific underwriting standards and
products that
may be intended for different segments of the population, or
that address the
balancing of access to credit and credit risk.
A recent paper by Quercia, Ding, and Reid (2012) specifically
addresses the
balancing of credit risk and mortgage access for borrowers—the
two tensions
on which we focus. Their paper narrowly focuses on the
marginal impacts of
setting QRM product standards more stringently than those for
QM.4 They
find that the benefits of reduced foreclosures resulting from the
more stringent
product restrictions on “LTV” ratios, debt-to-income ratios
(“DTI”) and credit
scores do not necessarily outweigh the costs of reducing

borrowers’ access
to mortgages, as borrowers are excluded from the market.
Pennington-Cross and Ho (2010) examine the performance of
hybrid and ad-
justable rate mortgages (ARMs). After controlling for borrower
and location
characteristics, they find that high default risk borrowers do
self-select into
adjustable rate loans and that the type of loan product can have
dramatic im-
pacts on the performance of mortgages. They find that interest
rate increases
over 2005–2006 led to large payment shocks and with house
prices declin-
ing rapidly by 2008, only borrowers with excellent credit
history and large
amounts of equity and wealth could refinance.
Amromin and Paulson (2009) find that while characteristics
such as LTV,
FICO score and interest rate at origination are important
predictors of defaults
for both prime and subprime loans, defaults are principally
explained by house
price declines, and more pessimistic contemporaneous
assumptions about
house prices would not have significantly improved forecasts of
defaults.
Courchane and Zorn (2012) look at changing supply-side
underwriting stan-
dards over time, and their impact on access to credit for target
populations
of borrowers.5 They use data from 2004 through 2009,
specifically focusing

on the access to and pricing of mortgages originated for
African-American
and Hispanic borrowers, and for borrowers living in low-income
and minor-
ity communities. They find that access to mortgage credit
increased between
2004 and 2006 for targeted borrowers, and declined
dramatically thereafter.
The decline in access to credit was driven primarily by the
improving credit
mix of mortgage applicants and secondarily by tighter
underwriting standards
4For details of the QRM, see Federal Housing Finance Agency,
Mortgage Market Note
11-02. For details of the QM, see
http://files.consumerfinance.gov/f/201310_cfpb_qm-
guide-for-lenders.pdf.
5See also Courchane and Zorn (2011, 2014) and Courchane,
Dorolia and Zorn (2014).
associated with the replacement of subprime by FHA as the
dominant mode
of subprime originations.
These studies all highlight the inherent tension between access
to mortgage
credit and credit risk. They also stress the difficulty in finding
the “cor-
rect” balance between the two, and suggest the critical
importance of treat-
ing separately the three mortgage market segments—prime,

subprime and
government-insured (FHA)—because of the different borrowers
they serve
and their differing market interactions. The research also
provides some op-
timism that a careful examination of recent lending patterns will
reveal op-
portunities for responsibly extending credit while balancing
attendant credit
risks.
Data
Our analysis uses CoreLogic data for mortgages originated
between 1999
and 2009. The CoreLogic data identify prime (including Alt-A),
subprime
and government loans serviced by many of the large, national
mortgage
servicers. These loan-level data include information on
borrower and loan
product characteristics at the time of origination, as well as
monthly updates
on loan performance through 2012:Q3. Merged to these data are
annual
house price appreciation rates at a ZIP code level from the
Freddie Mac
Weighted Repeat Sales House Price Index, which allow us to
update borrower
home equity over time.6 We prefer this house price index to the
FHFA’s,
as the latter are not available at the ZIP code level. The
CoreLogic data
do not provide Census tract information, so we use a crosswalk
from ZIP
codes to 2000 Census tracts.7 We also merge in county-level

unemployment
rates from the Bureau of Labor Statistics, which are seasonally
adjusted by
Moody’s Analytics.8 Finally, we include changes in the
conventional mortgage
market’s average 30-year fixed mortgage (“FRM”) rate reported
in Freddie
Mac’s Primary Mortgage Market Survey.9
The CoreLogic data are not created through a random sampling
process and
so are not necessarily representative of the overall population,
or our target
6While these data are not publicly available, the metro/state
indices can be found
which are available at:
http://www.freddiemac.com/finance/fmhpi/.
7Missouri Census Data Center, available at:
http://mcdc.missouri.edu/
websas/geocorr12.html.
8The unemployment rate is from the BLS Local Area
Unemployment Statistics
(http://www.bls.gov/lau/).
9These data are available publicly at:
http://www.freddiemac.com/pmms/pmms30.htm.
population. This is not a problem for estimating our
delinquency model, but it
does create concern for drawing inference with our scorecards.
To address this
potential concern, we apply appropriate postsample weights

based on HMDA
data to enhance the representativeness of our sample. We
develop weights by
dividing both the HMDA and the CoreLogic data into
categories, and then
weight so that the distribution of CoreLogic loans across the
categories is the
same as that for HMDA loans. The categories used for the
weighting are a
function of loan purpose (purchase or refinance), state, year of
origination and
loan amount. Because we rely on a postsample approach and
cannot create
categories that precisely define our target population, our
weighting does not
ensure representativeness of the CoreLogic data for this group.
Nevertheless,
it likely offers a significant improvement over not weighting.
We also construct a holdout sample from our data to use for
inference. This
ensures that our estimated models are not overfitted. The
holdout sample was
constructed by taking a random (unweighted) sample of 20% of
all loans in
our database. All summary statistics and estimation results
(Tables 1 and 2
and Appendix) are reported based on the unweighted 80%
estimation sample.
Consistent with our focus on identifying responsible credit
opportunities, we
restrict our analysis to first lien, purchase money mortgage
loans. Summary
statistics for the continuous variables used in our delinquency
estimation are
found in Table 1. Table 2 contains summary statistics for the

categorical
variables.
As shown in Table 1, the average LTV at origination is 97% for
government
loans. This is considerably higher than for the prime market,
where first
lien loans have LTVs less than 80%, on average.10 We also
observe the
expected differences in FICO scores, with an average FICO
score in the prime
sector of 730, 635 for subprime and 674 for government loans.
The prime
market loan amount (i.e., unpaid principal balance, or UPB, at
origination)
averages $209,000 with the government loan amount the lowest
at a mean of
$152,000. The mean value in the subprime population is below
that for prime
at $180,000. DTI ratios do not differ much between prime and
government
loans, and the DTI for subprime is unavailable in the data. As
DTI is a key
focus in the efforts of legislators to tighten underwriting
standards, we use it
when available for estimation. The equity measures post-
origination reflects
the LTV on the property as house prices change in the area.
All three markets faced significant house price declines, as
captured by the
change in home equity one, two or three years after origination.
For all three
10The mean LTV for subprime mortgages is surprisingly low at
83%, although this

likely reflects the absence of recording second lien loans, which
would lead to a higher
combined LTV.
T
ab
le
1
�
S
um
m
ar
y
st
at
is
ti
cs
fo
r
co
nt

in
uo
us
va
ri
ab
le
s
(8
0%
es
ti
m
at
io
n
sa
m
pl
e)
—
st
at
is

ti
cs
no
t
w
ei
gh
te
d.
A
ll
P
ri
m
e
S
ub
pr
im
e
G
ov
er
nm

en
t
L
oa
ns
(u
nw
ei
gh
te
d)
21
,7
25
,5
28
13
,8
94
,1
40
2,
50
3,

16
4
5,
32
8,
22
4
C
hg
in
M
or
tg
ag
e
R
at
e,
O
ne
Y
ea
r
P

os
t-
O
ri
gi
na
ti
on
−0
.1
4%
−0
.1
2%
0.
04
%
−0
.3
0%
%
M
is
si

ng
0.
00
%
0.
00
%
0.
00
%
0.
00
%
S
t.
D
ev
.
0.
61
%
0.
60
%

0.
58
%
0.
64
%
C
hg
.
in
M
or
tg
ag
e
R
at
e,
T
w
o
Y
ea
rs

P
os
t-
O
ri
gi
na
ti
on
−0
.2
1%
−0
.2
2%
−0
.1
0%
−0
.2
5%
%
M
is

si
ng
0.
00
%
0.
00
%
0.
00
%
0.
00
%
S
t.
D
ev
.
0.
56
%
0.
57

%
0.
54
%
0.
53
%
C
hg
.
in
M
or
tg
ag
e
R
at
e,
T
hr
ee
Y
ea

rs
P
os
t-
O
ri
gi
na
ti
on
−0
.2
4%
−0
.2
3%
−0
.2
7%
−0
.2
9%
%
M

is
si
ng
0.
00
%
0.
00
%
0.
00
%
0.
00
%
S
t.
D
ev
.
0.
55
%
0.

55
%
0.
53
%
0.
55
%
D
eb
t-
to
-I
nc
om
e
R
at
io
(D
T
I)
36
.0

8%
35
.8
6%
0.
00
%
36
.8
4%
%
M
is
si
ng
62
.0
3%
58
.7
7%
99
.9
8%

52
.7
2%
S
t.
D
ev
.
15
.1
0%
15
.2
1%
0.
00
%
14
.6
9%
E
qu
it
y

M
ea
su
re
O
ne
-Y
ea
r
P
os
t-
O
ri
gi
na
ti
on
95
.7
2%
95
.5
3%

93
.8
6%
97
.3
8%
%
M
is
si
ng
3.
04
%
2.
06
%
8.
52
%
3.
02
%
S

t.
D
ev
.
10
.3
5%
10
.8
0%
9.
33
%
9.
15
%
E
qu
it
y
M
ea
su
re

T
w
o-
Y
ea
rs
P
os
t-
O
ri
gi
na
ti
on
94
.7
9%
94
.5
3%
94
.5
3%

95
.7
4%
%
M
is
si
ng
3.
04
%
2.
06
%
8.
52
%
3.
02
%
S
t.
D
ev

.
20
.2
0%
21
.0
2%
22
.1
2%
15
.9
5%
E
qu
it
y
M
ea
su
re
T
hr
ee

-Y
ea
rs
P
os
t-
O
ri
gi
na
ti
on
96
.0
4%
95
.7
3%
10
0.
67
%
94
.3

9%
%
M
is
si
ng
3.
04
%
2.
06
%
8.
52
%
3.
02
%
S
t.
D
ev
.
28

.5
0%
29
.0
5%
34
.6
4%
21
.8
4%
T
ab
le
1
�
C
on
ti
nu
ed

.
A
ll
P
ri
m
e
S
ub
pr
im
e
G
ov
er
nm
en
t
F
IC
O
70
1.

9
72
7.
9
63
5.
1
66
5.
1
%
M
is
si
ng
14
.2
4%
15
.7
7%
9.
08
%

12
.7
0%
S
t.
D
ev
.
69
.8
55
.1
63
.6
68
.8
L
oa
n-
to
-V
al
ue
R

at
io
(L
T
V
),
at
O
ri
g.
82
.9
1%
78
.3
4%
83
.2
2%
96
.9
4%
%
M

is
si
ng
1.
62
%
1.
54
%
2.
43
%
1.
44
%
S
t.
D
ev
.
14
.9
5%
14

.6
4%
11
.0
9%
7.
07
%
L
oa
n
A
m
ou
nt
(O
ri
g.
U
P
B
($
))
$1

90
,5
65
$2
07
,3
70
$1
80
,3
74
$1
44
,5
34
%
M
is
si
ng
0.
00
%
0.

00
%
0.
00
%
0.
00
%
S
t.
D
ev
.
$3
28
,1
41
$3
62
,9
15
$2
67
,0

06
$2
27
,5
22
U
ne
m
pl
oy
m
en
t
R
at
e
O
ne
-Y
ea
r
P
os
t-

O
ri
gi
na
ti
on
5.
58
%
5.
40
%
5.
03
%
6.
47
%
%
M
is
si
ng
0.

11
%
0.
10
%
0.
22
%
0.
08
%
S
t.
D
ev
.
2.
27
%
2.
09
%
1.
53

%
2.
88
%
U
ne
m
pl
oy
m
en
t
R
at
e
T
w
o-
Y
ea
rs
P
os
t-

O
ri
gi
na
ti
on
6.
02
%
5.
94
%
5.
40
%
6.
64
%
%
M
is
si
ng
0.

09
%
0.
09
%
0.
12
%
0.
09
%
S
t.
D
ev
.
2.
56
%
2.
53
%
1.
93

%
2.
85
%
U
ne
m
pl
oy
m
en
t
R
at
e
T
hr
ee
-Y
ea
rs
P
os
t-

O
ri
gi
na
ti
on
6.
57
%
6.
55
%
6.
45
%
6.
69
%
%
M
is
si
ng
0.

08
%
0.
08
%
0.
05
%
0.
09
%
S
t.
D
ev
.
2.
78
%
2.
81
%
2.
72

%
2.
72
%
Table 2 � Summary statistics for categorical (class) variables
(80% estimation
sample)—statistics not weighted.
All Prime Subprime Government
ARM 12.60% 48.50% 4.72% 14.91%
Balloon 0.39% 4.91% 0.05% 0.82%
FRM-15 7.68% 1.61% 1.16% 5.63%
FRM-30 68.25% 22.31% 90.14% 67.79%
FRM-Other 4.48% 1.71% 2.73% 3.81%
Hybrid 6.59% 20.97% 1.20% 7.04%
Other 41.05% 33.13% 43.57% 40.71%
Retail 33.70% 21.20% 22.12% 29.87%
Wholesale 25.25% 45.67% 34.31% 29.42%
Full Documentation 29.83% 49.38% 41.80% 34.52%
Missing 38.89% 18.44% 42.30% 37.35%
Not Full Documentation 31.27% 32.19% 15.90% 28.13%
Owner Occupied 83.43% 85.88% 91.84% 85.48%
Not Owner Occupied 16.57% 14.12% 8.16% 14.52%
Condo 13.82% 7.75% 6.97% 11.70%
Single Family 86.18% 92.25% 93.03% 88.30%
mortgage market segments, post-origination equity measures
(post-origination

estimated LTV) averaged over 90%. Post-Origination
unemployment rates are
highest, on average, in the geographies with government loans,
although the
differentials among market segments fell after three-year post-
origination.
Table 2 presents the summary statistics for the categorical
(class) variables
in our sample. Some expected results emerge. The subprime
segment has the
largest share of loans originated through the wholesale channel
at 45.7%,
while the wholesale share for the prime segment was only
25.2%. Nearly half
(48.5%) of subprime loans were “ARM” loans, while only
22.3% of subprime
loans were the standard 30-year FRM product. In contrast,
69.1% of prime
loans were 30-year FRMs and an additional 7.8% were 15-year
FRMs. Nearly
all of the government loans (91.2%) were 30-year FRMs. The
documentation
figures are somewhat surprising, with nearly half (49.4%) of
subprime loans
indicating full documentation. The low share of full
documentation loans in
the prime sector (about 30%) likely reflects the inclusion of
Alt-A loans,
which are defined to be prime loans in the CoreLogic data.11
In our analyses, we focus on access to credit and credit risk
outcomes for all
borrowers. However, many homeownership and affordable
lending programs

11Historically, Alt-A loans were originated through prime
lenders, offering their more
credit worthy customers a simpler origination process.
focus more narrowly on assessing opportunities for responsibly
extending
mortgage credit to borrowers with low down payments and poor
credit his-
tories, or who are otherwise underserved by the prime market
(“target pop-
ulation”). As a result of long standing public policy objectives
focused on
the value of homeownership, both government insured mortgage
programs
(such as FHA) and the GSEs have long held missions to meet
the needs of
underserved borrowers, including low income, minority and
first-time home-
buyers.12 Programs meeting this mission are tasked with
balancing access to
credit for borrowers with any attendant increases in credit risk.
Therefore, aside from our focus on the opportunities provided to
the full pop-
ulation of borrowers, we also provide an analysis of scorecard
outcomes for a
specific target population. We define this target population as
borrowers who
receive first lien, purchase money mortgages on owner-occupied
properties
located in census tracts with median incomes below 80% of the
area median

income, with FICO scores less than or equal to 720 and with
LTV ratios
greater than or equal to 90%.
Limiting our analysis to borrowers who live in lower income
census tracts is
especially constraining, as many borrowers with high LTVs and
lower FICO
scores live elsewhere. However, our data lack accurate income
measures, and
public policy considerations encourage us to include an income
constraint in
our definition of the target population.13 As a consequence,
loans to target
borrowers account for a small percentage of the total loans
made during our
period of study (roughly 4%). We can be assured, however, that
our target
population is composed of borrowers who are an explicit focus
of public
policy.
Figure 1 provides a graphical illustration of the HMDA-
weighted distribution
of target population loans in our sample across the three market
segments.
The dramatic shift over time in the share going to the
government sector is
obvious, as is the reduction in the number of loans originated to
the target
population by all three segments, combined, post-crisis.
12Both the Community Reinvestment Act (CRA) and the
Federal Housing Enterprises
Financial Safety and Soundness Act of 1992 (the 1992 GSE Act)
encouraged mortgage

market participants to serve the credit needs of low- and
moderate-income borrowers
and areas.
13For example, GSE affordable goals are stated with respect to
low- and moderate-
income borrowers and neighborhoods.
Figure 1 � Target population by year and market segment
(weighted).
Analysis
Our analysis begins with the estimation of mortgage
performance models over
the crisis period. We use loan-level origination data from 1999
through 2009
to estimate models of loans becoming 90-days or more
delinquent in the first
three years after origination. These models include standard
borrower and loan
characteristics at origination, as well as control variables
measuring changes
in house prices, unemployment rates and interest rates post-
origination. They
also include several interaction terms for the borrower, loan and
control
variables.
We then use our estimated delinquency models to specify two
underwrit-
ing scorecards—a through-the-cycle scorecard and a stress
scorecard.14 We

next apply various cutpoints (risk thresholds) to our scorecards
to define
levels of acceptable risk. By definition, loans with risk scores
(delinquency
probabilities) at or below the cutpoint are assumed to be within
appropriate
(acceptable) risk tolerances.
The scorecard and cutpoint combinations provide working
prototypes of an
AUS. Our final step applies these prototypes to the full and
target populations
and assesses the results.
Estimating the Models
We estimate three separate delinquency models based on an
80% sample
of first-lien, purchase money mortgage loans in our data.
Separate models
14Additional scorecards constructed using “perfect foresight”
and macroeconomic
forecasts are available from the authors upon request.
were estimated for prime loans (including Alt-A loans),
subprime loans and
government loans, using an indicator provided in the CoreLogic
data to as-
sign each loan to its appropriate segment.15 We estimate
separate models for
each market sector because we believe that there is clear market

segmenta-
tion in mortgage lending. In the conventional market the
lenders, industry
practices, market dynamics and regulatory oversight have
differed between
the prime and subprime segments.16 A similar distinction exists
between the
conventional and government segments—the latter focuses on
first-time bor-
rowers and lower income households. Moreover, acceptable risk
tolerances
will necessarily vary across segments, as may concerns
regarding access to
credit.
Our process differs from the typical construction of
underwriting systems in
two important ways. First, while the CoreLogic data are
reasonably rich in
variables, they do not contain the detailed credit variables, such
as tradeline
balance to limits, number of open tradelines and presence of
mortgage late
payments, which are a key component of most underwriting
models. As
a result, our model assesses risk less accurately than production
versions.
Second, typical models are estimated on historical data, but the
resulting
scorecards are applied to future applications (i.e., out of
sample). However,
lacking knowledge and data on future states of the world, we
use historical
assessments and apply our scorecard to the 20% holdout sample.
Thus, our
scorecard may assess risk more accurately than production

versions, given that
the data were contemporaneously generated. However, we
believe that this is
not a critical concern because we seek to illustrate how certain
scorecard and
cutpoint combinations might affect outcomes under future
stressful market
conditions.
The dependent variable in our estimation is a loan becoming 90
days or
more delinquent in the first three years after origination.
Continuous ex-
planatory variables include borrower FICO scores, mortgage
market interest
rates (Freddie Mac Primary Mortgage Market Survey rates),
updated LTVs
(derived using the Freddie Mac House Price Index) and local
unemploy-
ment rates.17 The models also include categorical explanatory
variables for
15Because this field is determined at CoreLogic, we are unable
to define the specific
parameters around the determination of subprime.
16This structural segmentation also loosely translates into
separation on the basis of
risk—the prime segment generally caters to lower risk
borrowers, while the subprime
segment generally caters to higher risk borrowers—however, the
distinction along this
dimension is far from perfect. We are segmenting by market
structure, not simply by
market risk.
17We include variables estimated to model post-origination
home equity one-

year, two-year and three-year post-origination. If equity one-
year post-origination
loan amount ($50,000–$150,000, $150,000–$250,000,
$250,000–$350,000,
$350,000–$450,000 and greater than $450,000); documentation
type (full
documentation, low documentation and missing documentation);
origination
channel (retail, wholesale and other); original LTV (less than
40%, 40–60%,
60–75%, 75–80%, 80–85%, 85–90%, 90–95%, 95–105%, 105
to–115% and
greater than 115%); product type (ARM, balloon, 15-year FRM
[“FRM-15”],
30-year FRM [“FRM-30”] and other FRM and hybrids [“FRM-
other”]) and
condo and owner occupancy indicators. Finally, interactions
were included
between FICO score of borrower and whether FICO was
missing, FICO
score and loan amount, loan amount and LTV and FICO score
and LTV.18
The estimation results, based on the 80% sample, are presented
in Appendix
Tables A.1.a (prime), A.2.a (subprime) and A.3.a (government).
Goodness-
of-fit plots, applied to the holdout sample, are found in Figures
A.1.b (prime),
A.2.b (subprime) and A.3.b (government). Most of the variables
in the prime

delinquency model (Figure A.1) had the expected signs. Loans
with LTV
ratios less than 80%, full documentation loans, retail channel
loans, loans
under the conforming limits and FRM loans are all less likely to
become
delinquent. As FICO score increases, the delinquency
probability falls. Loans
with higher LTV values have higher delinquency rates, with the
loans in the
over 100 LTV categories most likely to go delinquent. Owner
occupants are
less likely to become delinquent.
Most of the subprime results (Figure A.2.a) are similar to those
in the prime
model with a few exceptions. As in the prime segment, owner-
occupied and
FRM loans are less likely to become 90 days delinquent. LTV
also has a
similar relationship with delinquency in both the prime and
subprime mod-
els; however, the parameter estimates on the high LTV prime
loans exceed
those for subprime, perhaps reflecting missing second lien loans
in the sub-
prime population. There are some differences in the two models,
however.
For example, full documentation subprime loans are more likely
to become
is defined as ltv_1yr, then ltv_1yr = ltv*upb_1yr/[upb /( 1 +
hpa_1yr_orig)]
where ltv = LTV at origination, upb = upb at origination and
hpa_1yr_orig =
house price growth (in %) from origination to one year after

origination and
upb_1yr = upb 1 year after origination, assuming no
delinquency/curtailment,
and a fully amortizing loan. Then, upb_1yr= upb +
FINANCE(’CUMPRINC’,
initial_interest_rate/1200, original_term, upb, 1, 12); where
FINANCE is a SAS
function used to compute the cumulative principal. See:
http://support.sas.com/
documentation/cdl/en/lrdict/64316/HTML/default/viewer.htm#a
003180371.htm
18We use an indicator variable for observations with missing
DTI in the prime and
government segments (this is the omitted category.) DTI is
missing for all subprime
loans. For missing FICO, we create a variable denoted as FICO2
and set scores when
missing to 700 and to actual values otherwise. We interact
FICO2 with a dummy for
missing status.
delinquent, as are loans from the retail channel. Higher FICO
scores do not
reduce the likelihood of subprime mortgage delinquency.
For the government segment, retail channel has the negative
sign we observed
in prime. Nearly all government loans are full documentation,
so the result
carries little meaning. Finally, higher LTV and lower FICO
government loans
scores have an increased probability of delinquency. Owner

occupancy and
the retail channel reduce the probability of delinquency, as do
FRM loans.
We assess model fit by comparing model predictions to actual
outcomes.
The results of these comparisons are provided in the Appendix
as Figures
A.1.b, A.2.b and A.3.b for the prime, subprime and government
estimations,
respectively.19 In general, we see that the models fit well.
Specifically, the
scatter plots remain relatively close to the 45-degree reference
line. To the
extent that there is any systematic error in the model, it occurs
for lower
risk loans (toward the bottom left of the chart). This causes
relatively little
concern for our analysis because it is most important that the
model is well-fit
in the area around likely cutpoints, which is located in the well-
fitting higher
risk (upper right-hand) section of the charts.
Finally, in Figures 2.a (prime), 2.b (subprime) and 2.c
(government), we use
predictions from our estimated model on the weighted holdout
sample to
provide a distributional sense of loans originated throughout the
years in our
sample. Specifically, using our predictions, we rank order loans
within each
market segment and origination year into approximately 200
buckets. For
each bucket of loans, we compute the realized default rate. We
then generate

a box plot illustrating the distribution of average default rates
over our 200
buckets by market segment and year.20
The three charts below immediately highlight the dramatic
increase in delin-
quency rates that occurred during the crisis years of 2005–2008.
Clearly,
19Loans in each segment are first grouped by model prediction,
and then divided into
200 equally sized buckets of loans with similar model
predictions. The mean model
prediction and actual delinquency rates are calculated for each
bucket, and then plotted
in log-log scale. The model prediction is measured on the
horizontal axis, and the
actual delinquency rate is measured on the vertical axis. A 45-
degree reference line
is drawn in each chart, reflecting the combination of points
where the models are
perfectly predicting.
20The “box” in the box plot shows the interquartile range
(“IQR”)—the scores between
the 25th and the 75th percentiles. The “whiskers” go down to
the 5th percentile, and
up to the 95th percentile of scores. The 50th percentile (the
median) falls within
the box. The data are weighted via HMDA to more accurately
reflect the underlying
population.

Figure 2 � a. Prime, b. Subprime and c. Government.
there is justification for a general concern over credit risk and
the perfor-
mance of loans under stressful conditions in particular. It is also
interesting
to compare the relative performance of loans across the market
segments. In
the years 1999–2004, prime loans were clearly the best
performing, followed
by government loans, and then subprime. However, the
government segment
performed roughly as well as the prime segment during the
crisis years. This
is because Alt-A mortgages are primarily allocated to the prime
segment (as
they were primarily originated by prime lenders) and these
loans performed
very poorly. Regardless, both the prime and government
segments look much
better than subprime during the crisis years—as subprime
delinquency rates
reached an average of 40% for 2006 and 2007 originations.
Deriving the Scorecards
The second step of our analysis is to derive prime, subprime and
government
scorecards from the estimated models. Scorecards are an ex ante
assessment
of the credit risk at origination associated with a particular

borrower/loan
combination. Our estimated delinquency models provide the
basis for this
assessment; although these models include both ex ante and ex
post (post-
origination) explanatory variables. The appropriate treatment of
the post-
origination explanatory variables is the key challenge for
scorecard creation.
One approach, arguably the most typical, is to treat post-
origination explana-
tory variables as controls in the scorecard.21 That is, to keep
the value of these
variables constant across borrowers and over time. This is the
approach we
use here, although we create two variants. The first version we
call a “through-
the-cycle” scorecard. For this scorecard, we set post-origination
variables to
approximately their long run averages (house prices are set at a
2% annual
increase, interest rates are assumed to remain unchanged after
origination
and unemployment rates are set at 6%). This provides a
generally “friendly”
view of credit risk, and so is reflective of a concern for access
to credit is-
sues. The through-the-cycle scorecard also has the policy
advantage of being
countercyclical.
We also create a second version that we call our “stress”
scorecard. For this
scorecard, we incorporate the values of the ex post explanatory
variables used

by the Federal Reserve Board in its 2014 severely adverse stress
test scenario.
The Federal Reserve’s provides paths under the severely
adverse scenario for
several macroeconomic variables, including unemployment,
house prices and
mortgage rates. This represents a hypothetical scenario
containing both reces-
sion and financial market stress aimed to assess the resiliency
of U.S. financial
institutions. In this regard, our scorecard represents an outer-
bound possibility
of risk, and is clearly reflective of a concern for credit risk. For
each loan
scored under the stress scorecard, we use cumulative house
price declines of
12.4%, 24.2% and 24.7 for one-year, two-year and three-year
post-origination,
respectively, to update our equity variable in the stress
scorecard. We expect
the stress scorecard to be very tight with respect to access,
which should limit
its risk exposure during downturns. Separate scorecards are
created for each
21An alternative approach is to forecast at origination the future
values of the
ex post explanatory variables. This is a challenging task in both
theory and prac-
tice. A prototype version of such a scorecard is available from
the authors on request.

of the models/markets: prime, subprime and government. We
believe that it
is enlightening to compare and contrast the results of the
through-the-cycle
and stress scorecards for each market. While we expect the
through-the-cycle
scorecard to lead to increased access to credit relative to the
stress scorecard,
it may achieve this at the cost of higher credit risk. We expect,
for example,
that it will perform significantly worse than the stress scorecard
during down
cycles.
Choice of Cutpoints
The third step in our analysis is to choose scorecard cutpoints.
The cutpoints
set the marginal risk tolerance for the scorecards, and so
determine the lev-
els at which loans switch from “acceptable” to “unacceptable”
risks. The
cutpoints therefore set the extreme bounds of within-tolerance
risk for the
scorecards, and are critical in setting the balance between
access to credit and
credit risk concerns.
Both policy and business considerations influence the
determination of cut-
points. For example, a 10% delinquency rate might be viewed as
an acceptable
prime cutpoint during boom years when the market is optimistic
and public
policy is focused on expanding access to credit. However, that
same 10%

delinquency rate might be viewed as too high for a prime
cutpoint during a
recession, when the market is trying to limit credit exposure and
public policy
has shifted its focus to managing systemic risks and taxpayer
losses. Lenders
with more tolerance for risk might choose to operate in the
subprime market
segment, and will accept higher risk thresholds than lenders
who want to
operate in the prime segment. Government-insured risk
tolerance levels may
vary with the health of the mortgage insurance fund, as well as
other policy
considerations.
It is not our intention to propose “correct” cutpoints for our
scorecards. Rather,
our goal is to illustrate how the interactions between scorecards
and possible
cutpoints affect access to credit and the management of credit
risk, and to
illustrate the potential for possible compromises. Toward this
end, we provide
a set of potential cutpoints for each scorecard. Specifically, we
provide results
for cutpoints of 5%, 10%, 15% and 20% delinquency rates for
each of our
scorecards. This allows us to provide a range of alternative
impacts on both
the full and target populations.
To simplify our presentation and focus our analysis, we also
concentrate on
a select cutpoint for each market that offers a possible
compromise between

managing access and credit risk. This is determined by choosing
among our
four cutpoints for each market the one that most closely
approximates the
actual delinquency rate of marginal loans originated in the years
1999–2001.
These years provide origination cohorts that experienced a
relatively benign
economic environment for the first three years after origination
(neither ex-
pansive nor depressed), and their realized performance is not
unduly affected
by factors outside the control of underwriting.
Underwriting in the prime market during the 1999 –2001 period
was rela-
tively standardized (arguably, neither too loose nor too tight),
so we set the
select cutpoint at the realized performance of borrowers around
the 90th risk
percentile from the full model prediction. This performance is
most closely
approximated by a cutpoint of 5% delinquency rates for the
prime market
(see Figure 2.a), and by construction this results in about 90%
of the prime
loans originated in 1999–2001 being viewed as acceptable
risk.22
The subprime performance distribution (see Figure 2.b) displays
a markedly

different time trend than that observed in the prime market.
Realized per-
formance in the years 1999–2001 was worse than the
performance of the
2002–2004 cohorts. This suggests that subprime underwriting in
the earlier
period was not as standardized as it was in the prime market
during those
years. Moreover, the differential in risk between prime and
subprime lending
appears greater in the earlier years, suggesting that subprime
lending was
relatively less conservative than prime lending in 1999–2001.
Finally, the
overall tolerance for accepting risk in mortgage lending has
clearly declined
in the recent environment. These factors persuade us to use a
more restrictive
standard for determining marginal borrows in the subprime
segment than we
do in the prime segment. For the subprime market, we choose a
cutpoint of a
15% delinquency rate, which results in only about one-half of
the subprime
loans originated in in 1999–2001 being viewed as acceptable
risk in terms of
their realized outcomes.
The performance distribution of government-insured loans is
shown in
Figure 2.c. As with the subprime market, the box plots for
government mort-
gage lending suggest that underwriting was not as standardized
or (relatively)
conservative as in the prime market from 1999 to 2002.
Particularly, striking

is the more limited relative increase in the risk distributions of
the 2006–2008
originations than the increase experienced by these cohorts in
the subprime
market. We therefore again impose a more restrictive standard
for determining
the marginal borrowers in the government market, but mitigate
this somewhat
because of the government sector’s explicit goal of providing
credit to first
22The 90th risk percentile is the scorecard prediction level that
separates the 10% of
borrowers with the highest predicted risks from the remaining
90% of borrowers with
lower predicted risks.
time and traditionally underserved home buyers. This yields a
select cutpoint
for the government sector of a 10% delinquency rate, which
results in about
60% of the 1999–2001 cohort being viewed as acceptable risk in
terms of
their realized outcomes.
Applying Scorecards to the Full and Target Populations
Our last step applies our automated underwriting scorecards to
the full and tar-
get populations. As noted earlier, the target population
represents only about
4% of overall originations during our period of study. Although

a restrictive
definition, we believe that our resulting target population is
highly reflective
of the population focused on by most affordable housing
initiatives. We use
our through-the-cycle and stress scorecards to separately score
borrowers, and
then determine the percent of the population assessed as
acceptable risks by
each scorecard using the alternative cutpoints (expressing risk
thresholds of
5%, 10%, 15% and 20% delinquency rates). Table 3.a presents
the results for
the full population, while Table A.4a, in the Appendix, presents
the results
for the target population. Table 3.b presents the share of
defaults for the full
population for each scorecard in each risk threshold range.
Table A.4b in the
Appendix provides similar values for the target population.
As indicated in Table 3.a, using a cutpoint of 5%, we find that
85.1% of
all prime borrowers are viewed as acceptable risks by the
through-the-cycle
scorecard over 1999–2009, while the stress scorecard yields
only 60.8 accept-
able risks among prime borrowers. If the risk threshold is
relaxed to a level
of 10%, 97.6% of the prime market borrowers are acceptable
risk. Over time,
at a 5% prime cutpoint, the percent of acceptable risks falls as
the origina-
tion population reflects the historic relaxation of underwriting
standards—the
market includes more high risk borrowers, so a lower percent

are accepted
with the 5% cutpoint.
For the subprime market segment, the through-the-cycle
scorecard assesses
39.7% as acceptable risk with a default risk threshold set at
15%, while only
4.7% are acceptable using the stress scorecard. If the risk
threshold is held to
5%, only 2.5% are accepts.
Finally, in the government segment, using a 10% threshold,
61.1% of bor-
rowers are acceptable risks with the through-the-cycle
scorecard. At a 5%
threshold, only 4.7% would have been able to receive
mortgages.
As expected, the through-the-cycle scorecard accepts more
borrowers than
does the stress scorecard during a stressful environment, with
the differ-
ential impacts of the two scorecards varying by market segment.
The prime
T
ab
le
3.
a

�
P
er
ce
nt
of
ac
ce
pt
ab
le
ri
sk
bo
rr
ow
er
s
by
ri
sk
th
re
sh

ol
d
ra
ng
e,
fu
ll
po
pu
la
ti
on
.
S
co
re
ca
rd
:
T
hr
ou
gh
-t

he
-C
yc
le
P
ri
m
e
19
99
20
00
20
01
20
02
20
03
20
04
20
05
20
06

20
07
20
08
20
09
A
L
L
L
es
s
th
an
5%
89
.1
85
.6
91
.0
89
.8

87
.5
85
.5
83
.6
78
.5
71
.2
83
.1
92
.1
85
.1
5–
10
%
9.
3
12
.0
7.

7
9.
1
10
.9
12
.5
13
.9
17
.4
22
.2
13
.9
7.
2
12
.5
10
–1
5%
1.
0

1.
5
0.
8
0.
7
1.
1
1.
3
1.
7
2.
5
3.
6
1.
9
0.
5
1.
5
15
–2

0%
0.
4
0.
6
0.
3
0.
2
0.
3
0.
4
0.
4
0.
8
1.
4
0.
6
0.
1
0.

5
G
re
at
er
th
an
20
%
0.
2
0.
3
0.
2
0.
2
0.
2
0.
3
0.
3
0.

7
1.
6
0.
5
0.
1
0.
4
S
ub
pr
im
e
19
99
20
00
20
01
20
02
20
03

20
04
20
05
20
06
20
07
20
08
20
09
A
ll
L
es
s
th
an
5%
0.
9
0.
8

0.
7
1.
6
4.
1
3.
9
3.
0
2.
5
2.
6
5.
7
2.
2
2.
5
5–
10
%
8.

4
8.
0
7.
6
13
.7
19
.7
14
.8
13
.9
12
.0
9.
9
8.
3
8.
4
12
.9
10

–1
5%
21
.0
19
.8
19
.1
23
.5
25
.7
28
.5
27
.3
23
.6
19
.8
33
.4
74
.3

24
.3
15
–2
0%
30
.9
35
.7
35
.8
35
.2
28
.9
33
.6
31
.6
29
.4
29
.1
39

.4
8.
1
32
.2
G
re
at
er
th
an
20
%
38
.7
35
.8
36
.7
26
.0
21
.6
19

.3
24
.2
32
.4
38
.6
13
.2
7.
1
28
.1
G
ov
er
nm
en
t
19
99
20
00

20
01
20
02
20
03
20
04
20
05
20
06
20
07
20
08
20
09
A
ll
L
es
s
th

an
5%
3.
4
2.
5
3.
6
4.
7
5.
3
6.
5
5.
3
4.
7
4.
3
4.
5
5.
1

4.
7
5–
10
%
45
.6
31
.8
52
.7
51
.7
56
.8
55
.8
49
.4
44
.2
48
.1
57

.8
62
.1
56
.4
10
–1
5%
44
.9
59
.3
36
.8
38
.3
33
.6
32
.2
39
.2
41
.6

42
.4
34
.9
30
.1
34
.8
15
–2
0%
1.
5
3.
1
3.
3
2.
0
1.
2
1.
2
2.

3
4.
8
1.
3
0.
5
0.
3
1.
2
G
re
at
er
th
an
20
%
4.
6
3.
4
3.

7
3.
3
3.
2
4.
3
3.
8
4.
7
4.
0
2.
3
2.
5
2.
9
T
ab

le
3.
a
�
C
on
ti
nu
ed
.
S
co
re
ca
rd
:
S
tr
es
s
P
ri
m
e

19
99
20
00
20
01
20
02
20
03
20
04
20
05
20
06
20
07
20
08
20
09
A
L

L
L
es
s
th
an
5%
65
.5
55
.6
66
.8
66
.6
63
.5
60
.1
56
.8
51
.9
46

.9
57
.3
75
.1
60
.8
5–
10
%
26
.0
32
.7
26
.6
25
.7
26
.8
28
.4
30
.1

30
.2
28
.8
29
.3
18
.6
27
.2
10
–1
5%
6.
1
8.
3
4.
7
5.
8
7.
1
8.

3
9.
0
11
.6
14
.7
8.
8
4.
8
8.
2
15
–2
0%
1.
3
1.
9
1.
1
1.
1

1.
6
2.
0
2.
4
3.
3
4.
6
2.
4
1.
0
2.
1
G
re
at
er
th
an
20
%

1.
1
1.
6
0.
8
0.
8
1.
1
1.
3
1.
7
2.
9
5.
0
2.
2
0.
5
1.
7

S
ub
pr
im
e
19
99
20
00
20
01
20
02
20
03
20
04
20
05
20
06
20
07
20

08
20
09
A
ll
L
es
s
th
an
5%
0.
0
0.
0
0.
0
0.
0
0.
2
0.
6

0.
3
0.
2
0.
2
0.
0
0.
0
0.
2
5–
10
%
0.
5
0.
3
0.
3
0.
8
2.

6
2.
3
2.
1
1.
7
1.
8
4.
7
4.
8
1.
5
10
–1
5%
2.
5
1.
9
1.
8

2.
4
4.
7
3.
6
3.
2
3.
0
3.
0
3.
1
1.
7
3.
0
15
–2
0%
4.
5
4.

6
4.
2
6.
9
8.
5
5.
9
6.
0
5.
5
5.
2
4.
9
4.
1
5.
9
G
re
at

er
th
an
20
%
92
.5
93
.3
93
.6
89
.9
83
.9
87
.7
88
.4
89
.6
89
.9
87

.3
89
.5
89
.3
G
ov
er
nm
en
t
19
99
20
00
20
01
20
02
20
03
20
04
20

05
20
06
20
07
20
08
20
09
A
ll
L
es
s
th
an
5%
0.
7
0.
5
0.
1
0.

2
0.
6
0.
9
0.
7
0.
9
0.
7
0.
6
0.
4
0.
5
5–
10
%
5.
0
3.
8

6.
9
9.
0
10
.4
11
.9
9.
8
8.
8
8.
1
10
.3
11
.5
9.
8
10
–1
5%
22

.5
15
.2
26
.0
29
.9
30
.8
31
.3
28
.1
25
.9
29
.2
35
.0
38
.5
32
.4
15

–2
0%
50
.1
35
.9
42
.5
40
.3
39
.1
36
.3
34
.7
32
.7
36
.9
37
.2
34
.8

37
.6
G
re
at
er
th
an
20
%
21
.6
44
.5
24
.6
20
.6
19
.1
19
.6
26
.7

31
.8
25
.0
17
.0
14
.8
19
.8
T
ab
le
3.
b
�
S
ha
re
of
to

ta
l
de
fa
ul
ts
by
ri
sk
th
re
sh
ol
d
ra
ng
e,
fu
ll
po
pu
la
ti
on

.
S
co
re
ca
rd
:
T
hr
ou
gh
-t
he
-C
yc
le
P
ri
m
e
19
99
20
00

an
5%
71
.9
71
.7
74
.4
66
.4
61
.9
60
.6
61
.9
58
.2
41
.3
50
.5
55
.1

55
.9
5–
10
%
19
.7
19
.7
18
.1
24
.0
28
.3
29
.3
28
.7
30
.6
37
.8
31

.3
36
.9
30
.7
10
–1
5%
4.
4
4.
4
4.
0
4.
4
4.
2
5.
0
5.
6
5.
9

9.
0
8.
4
3.
1
6.
5
15
–2
0%
2.
2
2.
3
1.
6
2.
2
2.
5
2.
2
1.

9
2.
3
5.
1
4.
6
2.
2
3.
1
G
re
at
er
th
an
20
%
1.
9
1.
9
1.

8
3.
1
3.
2
3.
0
1.
9
2.
9
6.
7
5.
1
2.
7
3.
9
S
ub
pr
im

e
19
99
20
00
20
01
20
02
20
03
20
04
20
05
20
06
20
07
20
08
20
09
A

ll
L
es
s
th
an
5%
0.
6
0.
3
0.
6
0.
5
1.
2
1.
0
0.
6
0.
5
0.

4
0.
6
0.
0
0.
6
5–
10
%
7.
4
5.
7
10
.0
7.
1
11
.7
8.
0
5.
5

4.
4
4.
5
5.
5
4.
2
6.
5
10
–1
5%
21
.1
18
.4
18
.3
22
.4
25
.0
20

.6
18
.8
17
.0
13
.4
26
.6
57
.3
19
.2
15
–2
0%
31
.9
34
.1
31
.5
38
.2

34
.9
36
.7
35
.0
33
.2
32
.0
45
.3
15
.5
34
.3
G
re
at
er
th
an
20
%

39
.1
41
.6
39
.6
31
.7
27
.3
33
.7
40
.1
44
.9
49
.7
22
.0
23
.1
39
.5

G
ov
er
nm
en
t
19
99
20
00
20
01
20
02
20
03
20
04
20
05
20
06
20
07

20
08
20
09
A
ll
L
es
s
th
an
5%
0.
7
0.
5
1.
0
1.
7
1.
7
2.
0

1.
7
1.
6
2.
1
1.
7
2.
1
1.
6
5–
10
%
36
.9
28
.2
34
.2
39
.9
44

.7
42
.2
37
.9
35
.2
38
.2
47
.4
52
.9
43
.0
10
–1
5%
49
.2
54
.8
45
.7

41
.9
38
.5
39
.0
44
.7
45
.7
46
.1
38
.9
36
.7
42
.2
15
–2
0%
1.
7
2.

7
6.
3
3.
4
2.
3
2.
0
2.
8
3.
3
1.
3
0.
6
0.
3
1.
9
G
re
at

er
th
an
20
%
11
.5
13
.8
12
.8
13
.2
12
.9
14
.8
12
.9
14
.2
12
.4
11

.5
8.
0
11
.3
T
ab
le
3.
b
�
C
on
ti
nu
ed
.
S
co
re
ca

rd
:
S
tr
es
s
P
ri
m
e
19
99
20
00
20
01
20
02
20
03
20
04
20
05

20
06
20
07
20
08
20
09
A
ll
L
es
s
th
an
5%
39
.4
34
.5
35
.0
31
.6

30
.5
28
.9
27
.4
26
.0
17
.0
23
.1
33
.1
25
.5
5–
10
%
36
.5
40
.2
44

.6
39
.7
36
.5
36
.8
40
.3
37
.7
29
.7
33
.2
26
.7
35
.7
10
–1
5%
12
.4

14
.2
10
.6
16
.3
19
.9
20
.5
19
.1
20
.7
26
.0
20
.5
25
.8
20
.8
15
–2

0%
5.
6
4.
7
4.
6
4.
4
4.
8
5.
4
6.
2
7.
0
10
.0
8.
2
7.
1
7.

3
G
re
at
er
th
an
20
%
6.
0
6.
4
5.
3
8.
0
8.
3
8.
3
6.
9
8.

6
17
.2
15
.0
7.
4
10
.7
S
ub
pr
im
e
19
99
20
00
20
01
20
02
20

03
20
04
20
05
20
06
20
07
20
08
20
09
A
ll
L
es
s
th
an
5%
0.
0
0.

0
0.
0
0.
0
0.
0
0.
0
0.
0
0.
0
0.
0
0.
0
0.
0
0.
0
5–
10
%

0.
4
0.
1
0.
3
0.
1
0.
9
0.
6
0.
4
0.
3
0.
2
0.
0
0.
0
0.
4

10
–1
5%
1.
8
1.
1
1.
6
1.
1
1.
5
1.
2
1.
0
0.
7
1.
0
1.
6
0.

0
1.
1
15
–2
0%
4.
3
3.
3
6.
2
3.
1
4.
9
3.
6
2.
2
2.
1
2.
3

2.
6
4.
2
3.
1
G
re
at
er
th
an
20
%
93
.6
95
.4
91
.7
95
.6
92
.7

94
.5
96
.3
96
.9
96
.5
95
.8
95
.8
95
.4
G
ov
er
nm
en
t
19
99
20
00

an
5%
0.
0
0.
0
0.
1
0.
0
0.
1
0.
1
0.
1
0.
1
0.
2
0.
1
0.
0

0.
1
5–
10
%
1.
8
1.
3
2.
0
3.
7
4.
5
4.
5
4.
7
3.
7
3.
9
5.

0
5.
9
4.
2
10
–1
5%
19
.3
13
.8
16
.9
20
.9
22
.7
22
.7
21
.6
20
.6

22
.8
27
.6
30
.9
23
.9
15
–2
0%
49
.3
43
.7
39
.7
38
.4
37
.6
36
.5
31

.6
32
.0
36
.1
38
.4
38
.4
38
.7
G
re
at
er
th
an
20
%
29
.6
41
.2
41

.3
36
.9
35
.1
36
.2
42
.1
43
.5
37
.0
28
.8
24
.8
33
.2
market, even with a lower risk threshold (5% cutpoint), accepts
a significantly

higher percentage of borrowers using either scorecard. The
stress scorecard,
because of its very pessimistic view of post-origination
outcomes, completely
eliminates access to credit in the subprime segment, and nearly
eliminates the
possibility of acceptable credit risks in the government segment.
Table A.4.a, as shown in the Appendix, provides results for
acceptable risks
among target borrowers. Using our select set of cutpoints, we
find that 55.1%
of the prime targeted borrowers are viewed as acceptable risks
by the through-
the-cycle scorecard. The stress scorecard yields 12.8%. For the
subprime
market, these values are 24.9% and 0.2%, respectively, and for
the government
market, they are 54.3% and 5.2%, respectively.
All of the borrowers in the full population received loans under
the standards
present at the time of origination. Applying a modern version of
a through-the-
cycle or stress scorecard, many of those borrowers would have
failed to qualify
for a loan. This suggests that AUS also offer some potential for
responsibly
extending credit to the target population. However, the through-
the-cycle and
stress scorecards offer competing policy trade-offs. The
through-the-cycle
scorecard extends credit to a larger percentage of the target
population by
providing greater access during expansionary cycles. The stress
scorecard

severely restricts access during periods of financial stress, as
designed.
Table 3.b provides the share of defaults by risk threshold. In the
prime mar-
ket, at a 5% cutpoint, 85.1% of the borrowers are accepted but
this group is
responsible for only 55.9% of the defaults. In contrast, at a 15%
subprime
cutpoint, the through-the-cycle scorecard accepted 39.7% of
borrowers rep-
resenting 26.3 of defaults. Finally, in the government segment,
while 61.1%
were acceptable at a 10% cutpoint, that group’s share of
defaults was 44.6%.
Only the prime market is of interest when assessing the default
share by
cutpoint using the stress scorecard for the full population (Table
3.b). The
stress scorecard accepted 60.8% using the 5% cutpoint and this
resulted in
loans that comprised 25.5% of the total defaults in the
population. For the
target population (Table A.4.b), the share of defaults for target
borrowers in
the prime segment is 35.1% with a 5% cutpoint; while it is
17.5% for subprime
with a 15% cutpoint and 45.7% for government with a 10%
cutpoint. In every
case, the percent of acceptable risks outweighs the default
share.
In summary, acceptable levels of risk can be achieved in two
ways. Either the
scorecard can reflect a more stressful environment post-

origination, which
means that the post-origination values of the macroeconomic
variables are
more pessimistic relative to the through-the-cycle scorecard, or
the cutpoints
can be adjusted. For example, in examining the results for the
prime scorecard,
85.1% of the full population posed acceptable risks using the
through-the-
cycle scorecard using a 5% cutpoint. That dropped to 60.8%
using the stress
scorecard. Approximately the same percentage of prime
mortgages (88.0%)
could be accepted using the stress scorecard, but that requires
relaxing the
cutpoint to 10%. It is similarly likely that using the through-
the-cycle score-
card and tightening the cutpoints would lower the percent of
acceptable risks
in a manner similar to the application of the stress scorecard.
It is clear that using both levers to manage credit risk (a stress
scorecard
and tight cutpoints) virtually eliminates credit access in the
subprime and
the government market segments. For example, a 15% cutpoint
and a stress
scorecard in the subprime segment results in only 4.7% of the
full popula-
tion being approved for loans. Using the stress scorecard in the
government

segment, with a cutpoint 10%, means that only 10.3% of
borrowers are ac-
ceptable credit risks. For the target population borrowers, using
the stress
scorecard and tighter cutpoints means that no borrowers are
viewed as posing
acceptable risks.
In Figures 3.a.1, 3.a.2, 3.b.1, 3.b.2, 3.c.1 and 3.c.2, we provide
accept rates
(1), the share of defaults (2) and realized default rates (3) for
the prime (a),
subprime (b) and government (c) segments, respectively, for the
through-the-
cycle (1) and stress (2) scorecards. These figures provide
information for
the full and target populations. All of these figures are based on
the select
cutpoints using the holdout sample for the given markets.
The accept rate trend lines provide information similar to that
found in Tables
3.a and A.4.a, but the results across segments and across
scorecards for the
full and target populations can be more readily compared in the
figures. It is
clear that the stress scorecard has reduced access to credit with
lower accept
rates (e.g., Figure 3.a.1.1 compared to Figure 3.a.2.1), but also
results in a
lower share of defaults (e.g., Figure 3.a.1.2 compared to Figure
3.a.2.2).
The figures also provide the average realized default rates for
those loans that
were judged to be acceptable credit risk for each scorecard. In

viewing the
realized default rate for the through-the-cycle scorecard, it is
interesting to
note that even though the scorecard and risk threshold are the
same for the full
and target populations (Figure 3.a.1.3), the target population
performs worse
in every year because it includes the riskier borrowers that are
meeting the
uniformly applied cutpoints. Further, even though the
scorecard’s through-the-
cycle values and the cutpoint are identical over time, the
realized performance
of the acceptable loans meeting the cutpoint is much worse in
the years when
the actual macroeconomic post-origination variables were at
higher “stress”
Figure 3 � Through-the-cycle: prime accept rate.
levels. This same effect can be observed for the stress scorecard
(Figure
3.a.2.3), although it is attenuated because that scorecard already
uses higher
“stress” levels of the macroeconomic variables.
Figures 3.b.2.1 (subprime) and 3.c.2.1. (government) clearly
demonstrate that
using a stress scorecard, with cutpoints that are reasonable for
the through-
the-cycle scorecard, would nearly completely eliminate any
subprime or

government lending.
We provide Table 4 as a final part of our analysis, which gives
summary
statistics on key variables by scorecard and market segment.
This provides
some detail on the types of loans, loan characteristics and
performance for the
loans that are in the acceptable risk class. The loans categorized
as requiring
Figure 3 � Continued.
more “caution” are those that would not be approved as their
level of risk is
higher than the risk threshold. If judgmental overrides are
applied to waive
the cutpoint, this requires the use of “caution.”
The acceptable risk loans are characterized by having higher
mean FICO
scores and lower LTV values than the caution loans. For both
the prime
and subprime segments, acceptable risk loans are more likely to
carry fixed
rates, although the difference is not very great in subprime. In
the government
segment, as nearly 90% of the market is FRM, there is less
distinction. Results
for the stress scorecard are marginally different. For the full
population, mean
FICO scores of the acceptable risk loans are higher and LTV

values lower for
Figure 3 � Continued.
the prime market segment, and even a higher percent of the
loans is FRM.
Mean FICOs are also higher in the government market.
Conclusions and Implications
Using residential mortgage default models for each of the three
segments
of the U.S. market (prime, subprime and government), we
construct two
scorecards (through-the-cycle and stress), and then apply each
scorecard to
the full population of borrowers and to a specific group of
targeted borrowers
to determine expected performance. Such scorecards can be
used to balance
the competing tensions between access to credit and credit risk.
T
ab
le
4
�

S
um
m
ar
y
st
at
is
ti
cs
by
sc
or
ec
ar
d
an
d
m
ar
ke
t
se
gm

en
t.
S
co
re
ca
rd
:
T
hr
ou
gh
-t
he
-C
yc
le
P
ri
m
e
S
ub

pr
im
e
G
ov
er
nm
en
t
R
is
k
T
ar
ge
t
T
ot
al
T
ar
ge
t
T

ot
al
T
ar
ge
t
T
ot
al
V
ar
ia
bl
e
C
la
ss
P
op
ul
at
io
n
P

op
ul
at
io
n
P
op
ul
at
io
n
P
op
ul
at
io
n
P
op
ul
at
io
n
P

op
ul
at
io
n
D
90
in
T
hr
ee
Y
ea
rs
A
cc
ep
t
8%
3%
13
%
11
%

11
%
7%
D
90
in
T
hr
ee
Y
ea
rs
C
au
ti
on
18
%
14
%
24
%
22
%

16
%
13
%
M
ea
n
F
IC
O
A
cc
ep
t
65
0
73
5
61
1
62
9
62
3

66
2
M
ea
n
F
IC
O
C
au
ti
on
64
5
70
1
59
8
62
0
68
98
L
T

V
A
cc
ep
t
93
%
75
%
92
%
78
%
98
%
96
%
L
T
V
C
au
ti
on

97
%
92
%
92
%
85
%
99
%
99
%
C
on
do
A
cc
ep
t
14
%
14
%
9%

9%
10
%
10
%
C
on
do
C
au
ti
on
9%
10
%
5%
7%
1%
1%
F
R
M
A
cc

ep
t
95
%
85
%
28
%
30
%
91
%
93
%
F
R
M
C
au
ti
on
86
%
74

%
27
%
21
%
95
%
96
%
S
co
re
ca
rd
:
S
tr
es
s
D
90
in
T
hr

ee
Y
ea
rs
A
cc
ep
t
6%
2%
9%
8%
10
%
5%
D
90
in
T
hr
ee
Y
ea
rs

C
au
ti
on
13
%
9%
22
%
19
%
14
%
10
%
M
ea
n
F
IC
O
A
cc
ep

t
64
9
74
0
61
0
63
1
63
5
68
9
M
ea
n
F
IC
O
C
au
ti
on
64

8
71
4
60
1
62
3
63
2
67
3
L
T
V
A
cc
ep
t
91
%
71
%
93
%

77
%
97
%
93
%
L
T
V
C
au
ti
on
95
%
88
%
92
%
83
%
98
%
98

%
C
on
do
A
cc
ep
t
13
%
14
%
7%
9%
17
%
17
%
C
on
do
C
au
ti

on
11
%
12
%
6%
8%
5%
5%
F
R
M
A
cc
ep
t
98
%
88
%
2%
17
%
71

%
83
%
F
R
M
C
au
ti
on
90
%
75
%
32
%
29
%
95
%
96
%
N
o

te
:
S
ta
ti
st
ic
s
ca
lc
ul
at
ed
us
in
g
H
M
D
A
de
ri
ve
d

w
ei
gh
ts
.
F
IC
O
sc
or
e
on
ly
ca
lc
ul
at
ed
us
in
g
ob
se
rv

at
io
ns
w
it
h
no
nm
is
si
ng
F
IC
O
.
Responsible lending requires the choice of an appropriate
scorecard and/or
the appropriate setting of risk tolerances, and in AUS, this is
operationalized
by setting post-origination macroeconomic values (choice of
scorecard) or by
choosing the scorecard cutpoints that determine the maximum
level of ac-

ceptable risk. This process requires judgment that balances
policy, regulatory
and business considerations, all of which may and do change
over time. This
simple observation highlights the temporal nature of responsible
lending—
risks that are viewed as acceptable in one period may be viewed
as excessive
in another.
As expected, for a given cutpoint, a through-the-cycle scorecard
accepts more
borrowers than the stress scorecard during a stressful time
period, with the
differentials varying by market segment. The prime market,
with a lower
risk threshold (5% cutpoint), accepts a significantly higher
percentage of
borrowers using either scorecard, while the subprime market
yields the lowest
percentage of borrowers using the through-the-cycle scorecard
(but slightly
higher acceptable rate than the government market when using
the stress
scorecard).
We find that the scorecards offer potential for responsibly
extending credit to
the target population. However, our results again indicate that
the through-the-
cycle and stress scorecards offer competing policy trade-offs.
The through-the-
cycle scorecard extends credit to a larger percentage of the
target population
by providing greater access during expansionary cycles. The
stress scorecard

severely restricts access during periods of financial stress, as
designed.
Our analysis is suggestive rather than definitive. It is sensitive
to the scorecard
used and to the choice of cutpoint (risk tolerance). It is
encouraging, however,
because it suggests that AUS offer potential for responsibly
extending credit
to all borrowers, including those in a target population. The size
of this
impact depends critically on the risk tolerances incorporated
into the AUS.
Regardless of the chosen level of risk, our analysis identifies a
portion of
the population to whom lenders can responsibly extend credit,
even during
stressful environments. This requires either adopting a
scorecard that reflects a
stress scenario or tightening cutpoints. Doing both, however,
results in almost
no loans being offered to the target population by either the
subprime market
or government market, and a significant reduction in the percent
of the target
population getting loans in the prime market.
To conclude, we find that automating underwriting scorecards
can provide a
valuable tool for managing risks for mortgage originations.
However, modern
AUS are unlikely by themselves to be a panacea for providing
access to credit

to targeted populations.23 With tight standards, target
borrowers face signif-
icant challenges in obtaining mortgage credit. Successful
homeownership
outreach programs may well require additional enhancements to
the under-
writing process such as prepurchase counseling and high-touch
origination
and servicing. These programs also often consider non-
traditional sources of
data, such as rental payment history, when assessing borrower
risk. Enhancing
modern automated underwriting along these dimensions is not a
simple mat-
ter, but doing so offers the potential of further expanding credit
access, while
managing the attendant risks and increasing accuracy, albeit at
potentially
higher cost.24
Our analysis also highlights how the treatment of the post-
origination environ-
ment has significant policy implications. From a
macroeconomic perspective,
the through-the-cycle scorecard has the desirable characteristic
of extend-
ing access to credit—it is more countercyclical during
recessions. The stress
scorecard, in contrast, is more procyclical in design. It also
reduces total
credit losses because it views the post-origination environment
as more risky.
This presents a challenging policy conundrum, requiring a
balancing of the

two tensions.
We thank seminar participants at Government Intervention in
Residential
Mortgage Markets, held December 3 and 4, 2013, Charlotte NC
and at the
IBEFA/WEA conference in Denver, June 30, 2014. In
particular, we thank Paul
Willen for his helpful suggestions. All views and opinions are
those of the
authors and do not reflect the views or opinions of Charles
River Associates
or its Board of Directors, of Freddie Mac or its Board of
Directors or of
the Federal Housing Finance Agency or any of Freddie Mac’s
regulators.
An earlier version of this article appears in Homeownership
Built to Last:
Lessons from the Housing Crisis on Sustaining Homeownership
for Low-
Income and Minority Families, 2014.
References
Amromin, G. and A.L. Paulson. 2009. Comparing Patterns of
Delinquency among
Prime and Subprime Mortgages. Federal Reserve Bank of
Chicago, Economic Perspec-
tives 2Q 18–37. Available at
http://www.chicagofed.org/digital_assets/publications/
economic_perspectives/2009/ep_2qtr2009_part2_amromin_pauls
on.pdf, last accessed
November 5, 2014.
23It is worth pointing out that most actual automated
underwriting systems include

many more detailed credit variables than are available in the
CoreLogic data or
included in our scorecard. The addition of these variables would
certainly improve
the accuracy of our delinquency model, but we expect that
would have relatively little
impact on extending credit to the target population.
24See for example Avila, Nguyen, and Zorn (2013) on the value
of counseling and
Moulton and Quercia (2013) on the use of high-touch servicing.
Avila, G., H. Nguyen and P. Zorn. 2013. The Benefits of Pre-
Purchase Homeownership
Counseling. Working Paper, Freddie Mac.
Bureau of Labor Statistics, Local Area Unemployment
Statistics. Avail-
able at: http://www.bls.gov/lau/, seasonally adjusted by
Moody’s Analytics,
last accessed November 5, 2014. Consumer Financial Protection
Bureau.
2013. Basic Guide for Lenders: What is a Qualified Mortgage.
Available
at http://files.consumerfinance.gov/f/201310_cfpb_qm-guide-
for-lenders.pdf, last ac-
cessed January 3, 2015.
Consumer Financial Protection Bureau. 2013. Basic Guide for
Lenders: What Is a
Qualified Mortgage? Available at
http://files.consumerfinance.gov/f/201310_cfpb_qm-
guide-for-lenders.pdf
Courchane, M.J., R. Darolia and P.M. Zorn. 2014. The Downs
and Ups of FHA Lend-

ing: The Government Mortgage Roller Coaster Ride. Journal of
Housing Economics
24: 39–56.
Courchane, M.J. and P.M. Zorn. 2011. A Changing Credit
Environment and Its Impact
on Low-Income and Minority Borrowers and Communities. In
Moving Forward: The
Future of Consumer Credit and Mortgage Finance. N. Retsinas
and E. Belsky, eds.
Washington, DC: Brookings Institution Press. 86–117.
—— and ——. 2012. The Differential Access and Pricing of
Home Mortgages: 2004–
2009. Real Estate Economics 40(S1): S115–S158.
—— and ——. 2014. Homeownership Built to Last: Lessons
from the Housing Crisis
on Sustaining Homeownership for Low-Income and Minority
Families. E.S. Belsky,
C.E. Herbert and J.H. Molinsky, eds. Washington, DC:Joint
Center for Housing Stud-
ies, Harvard University and Brookings Institution Press.
Federal Housing Finance Agency, Mortgage Market Note 11-02:
Qual-
ified Residential Mortgages. April 11. Available at
http://www.fhfa.gov/
PolicyProgramsResearch/Research/Pages/Mortgage-Market-
Note-11-2.aspx, last ac-
cessed November 5, 2014.
Freddie Mac House Price Index, Weighted Repeat Sales Index
(WRSI) at the zip code
level. The metro/state index. Available at
http://www.freddiemac.com/finance/fmhpi/,
last accessed November 5, 2014.
Moulton, S. and R.G. Quercia. 2013. Access and Sustainability
for First-Time
Homebuyers: The Evolving Role of State Housing Finance
Agencies. Available

at http://www.jchs.harvard.edu/sites/jchs.harvard.edu/files/hbtl-
10.pdf, last accessed
November 5, 2014, Presented at Joint Center for Housing
Studies, Harvard University.
Pennington-Cross, A. and G. Ho. 2010. The Termination of
Subprime Hybrid and
Fixed Rate Mortgages. Real Estate Economics 38(3): 399–426.
Quercia, R.G., L. Ding, and R. Carolina. 2012. Balancing Risk
and Access:
Underwriting Standards for Qualified Residential Mortgages.
Working paper
of the Center for Community Capital and the Center for
Responsible Lend-
ing. March 2012. Available at
http://www.responsiblelending.org/mortgage-
lending/research-analysis/Underwriting-Standards-for-
Qualified-Residential-
Mortgages.pdf, last accessed November 5, 2014. SAS Institute,
http://support.sas.com/
documentation/cdl/en/lrdict/64316/HTML/default/viewer.htm#a
003180371.htm
United States Department of Housing and Urban Development.
2002. Bush Admin-
istration Unveils New Homeownership Initiative: Martinez
Announces $1000 Home-
buyer Cash Back Incentive. HUD Archives 02–075.
Table A.1.a � Prime market estimation results (estimated on the
80% sample).
Variable Value Estimate Std Error Prob Chi Sq

Intercept −5.2079 0.0164 <0.0001
LTV Ratio 40–60 −1.5594 0.0126 <0.0001
60–75 −0.5847 0.00767 <0.0001
75–80 −0.3424 0.00678 <0.0001
80–85 0.1571 0.0134 <0.0001
85–90 0.1765 0.00864 <0.0001
90–95 0.4264 0.00913 <0.0001
95–105 0.5239 0.00803 <0.0001
105–115 1.7391 0.0183 <0.0001
>115 1.8144 0.0218 <0.0001
DTI ratio <20 −0.364 0.00678 <0.0001
20–30 −0.2646 0.00479 <0.0001
30–40 0.0802 0.00329 <0.0001
40–45 0.2991 0.00382 <0.0001
45–50 0.285 0.00453 <0.0001
>50 0.33 0.00427 <0.0001
FICO −0.00009 6.15E-06 <0.0001
Loan amount 50–150k 0.1187 0.00711 <0.0001
150–250k −0.1391 0.00848 <0.0001
250–350k −0.1926 0.0122 <0.0001
350–450k −0.1026 0.014 <0.0001
>450k −0.1327 0.0163 <0.0001
Documentation type Full −0.3297 0.00222 <0.0001
Missing 0.0917 0.00247 <0.0001
Origination channel Other 0.176 0.00189 <0.0001
Retail −0.1676 0.00204 <0.0001
Owner-occupied yes −0.00108 0.00191 0.57
Product ARM −0.0317 0.00483 <0.0001

Balloon 0.4432 0.0153 <0.0001
FRM-15 −1.0696 0.00981 <0.0001
FRM-30 −0.416 0.00402 <0.0001
FRM-Other 0.73 0.00523 <0.0001
Condo No 0.0595 0.00189 <0.0001
Mortgage rate One year after −0.0986 0.00301 <0.0001
Two years after −0.1032 0.00339 <0.0001
Three years after −0.0761 0.0031 <0.0001
Unemployment rate One year after −0.1493 0.00127 <0.0001
Two years after 0.055 0.00145 <0.0001
Three years after 0.0761 0.00101 <0.0001
House price appreciation One year after 1.2707 0.0257 <0.0001
Two years after −0.1657 0.0195 <0.0001
Note: Also included in the estimation are interactions between:
(1) FICO score and
loan amount, (2) loan amount and LTV ratio and (3) FICO score
and LTV ratio.
Table A.2.a � Subprime market estimation results (estimated on
the 80% sample).
Intercept −6.6489 0.0379 <0.0001
LTV 40–60 −1.0644 0.0384 <0.0001

60–75 −0.4261 0.0206 <0.0001
75–80 0.0908 0.017 <0.0001
80–85 0.2529 0.021 <0.0001
85–90 0.1539 0.0187 <0.0001
90–95 0.2071 0.0203 <0.0001
95–105 0.1302 0.0193 <0.0001
105–115 1.1918 0.0413 <0.0001
>115 0.9405 0.0568 <0.0001
FICO 0.00056 1.7E-05 <0.0001
Loan amount 50–150k 0.273 0.0156 <0.0001
150–250k −0.02 0.0187 0.2845
250–350k −0.1659 0.0313 <0.0001
350–450k −0.1991 0.0448 <0.0001
>450k −0.3163 0.0419 <0.0001
Documentation type Full 0.2814 0.00306 <0.0001
Missing −0.4448 0.00431 <0.0001
Retail 0.0811 0.00363 <0.0001
Owner-occupied Yes −0.0281 0.00287 <0.0001
Product ARM −0.1136 0.00489 <0.0001
Balloon 0.294 0.00735 <0.0001
FRM-15 −0.4165 0.0168 <0.0001
FRM-30 −0.3156 0.00546 <0.0001
FRM-Other 0.3038 0.0111 <0.0001
Condo No 0.0776 0.00364 <0.0001
Mortgage rate One year after 0.0854 0.00387 <0.0001
Two years after −0.0917 0.00475 <0.0001
Three years after −0.2461 0.00496 <0.0001

Two years after −0.0617 0.00257 <0.0001
Equity One year after 5.0494 0.056 <0.0001
Two years after −1.3697 0.0358 <0.0001
(1) FICO score and
and LTV ratio.
Table A.3.a � Government market estimation results (estimated
on the 80% sample).
Intercept −9.5488 0.066 <0.0001
LTV ratio <40 n/a n/a n/a
40–60 3.9636 0.1215 <0.0001
60–75 4.8392 0.0691 <0.0001
75–80 5.1345 0.064 <0.0001
80–85 5.2071 0.0693 <0.0001
85–90 5.2011 0.0644 <0.0001
90–95 5.3195 0.0625 <0.0001
95–105 5.5381 0.0608 <0.0001
105–115 6.9456 0.0775 <0.0001
>115 7.3719 0.0553 <0.0001

DTI ratio <20 −0.5595 0.00982 <0.0001
20–30 −0.0545 0.00668 <0.0001
30–40 0.0912 0.0048 <0.0001
40–45 0.2243 0.00564 <0.0001
45–50 0.2035 0.00622 <0.0001
>50 0.131 0.00653 <0.0001
FICO −0.0002 0.00002 <0.0001
Loan amount 50–150k 5.7136 0.0612 <0.0001
150–250k 5.3771 0.0667 <0.0001
250–350k 5.2903 0.0791 <0.0001
350–450k 4.9765 0.109 <0.0001
>450k −26.984 0.272 <0.0001
Documentation type Full 0.0563 0.00263 <0.0001
Missing −0.1174 0.00303 <0.0001
Retail −0.1423 0.00287 <0.0001
Owner-occupied yes −0.0967 0.00327 <0.0001
Product ARM −0.1674 0.0138 <0.0001
Balloon 0.1649 0.0549 0.0026
FRM-15 −0.5724 0.02 <0.0001
FRM-30 −0.263 0.0123 <0.0001
FRM-Other 1.3566 0.0133 <0.0001
Condo no 0.1757 0.00378 <0.0001
Mortgage rate One year after −0.0033 0.00305 0.2872
Two years after −0.0391 0.00361 <0.0001

Two years after 0.0934 0.00204 <0.0001
Equity One year after 1.4957 0.0418 <0.0001
Two years after −1.1064 0.0422 <0.0001
(1) FICO score and
and LTV ratio.
T
ab
le
A
.4
.a
�
P
er
ce
nt
of
ac
ce

pt
ab
le
ri
sk
bo
rr
ow
er
s,
by
ri
sk
th
re
sh
ol
d
ra
ng
e,
ta
rg
et

po
pu
la
ti
on
.
S
co
re
ca
rd
:
T
hr
ou
gh
-t
he
-C
yc
le
P
ri

m
e
19
99
20
00
20
01
20
02
20
03
20
04
20
05
20
06
20
07
20
08
20
09

A
L
L
L
es
s
th
an
5%
70
.7
69
.2
74
.7
65
.7
58
.2
51
.1
45
.1
33

.0
29
.9
49
.5
42
.2
55
.1
5–
10
%
24
.3
24
.9
21
.0
30
.3
36
.9
42
.3

43
.9
50
.3
49
.1
39
.1
51
.9
36
.0
10
–1
5%
2.
8
2.
8
2.
8
2.
2
2.

7
3.
9
6.
6
9.
0
10
.4
6.
1
2.
7
4.
7
15
–2
0%
1.
7
1.
9
1.
0

1.
0
1.
2
1.
2
2.
1
3.
2
4.
1
2.
2
0.
6
1.
9
G
re
at
er
th
an

20
%
0.
6
1.
2
0.
4
0.
8
1.
0
1.
5
2.
3
4.
6
6.
5
3.
1
2.
7

2.
2
S
ub
pr
im
e
19
99
20
00
20
01
20
02
20
03
20
04
20
05
20
06
20

07
20
08
20
09
A
ll
L
es
s
th
an
5%
0.
0
0.
0
0.
0
0.
0
0.
0

0.
0
0.
0
0.
0
0.
0
0.
7
1.
6
0.
0
5–
10
%
2.
0
2.
9
1.
5
1.

8
5.
5
5.
1
7.
2
7.
2
3.
9
5.
7
8.
3
4.
4
10
–1
5%
16
.0
11
.2

12
.7
16
.3
24
.8
28
.6
25
.3
21
.9
15
.5
23
.8
43
.2
20
.5
15
–2
0%
27

.7
26
.4
25
.3
36
.1
32
.6
35
.2
30
.1
23
.1
26
.8
33
.8
28
.6
29
.7
G

re
at
er
th
an
20
%
54
.3
59
.5
60
.4
45
.8
37
.0
31
.1
37
.5
47
.8
53

.8
36
.1
18
.3
45
.4
G
ov
er
nm
en
t
19
99
20
00
20
01
20
02
20
03
20

04
20
05
20
06
20
07
20
08
20
09
A
ll
L
es
s
th
an
5%
0.
3
0.
2
0.

9
1.
4
1.
5
1.
6
1.
6
1.
7
1.
5
1.
8
2.
8
1.
6
5–
10
%
39
.8

23
.8
45
.3
48
.1
56
.6
57
.6
51
.4
44
.9
48
.9
58
.4
65
.3
52
.7
10
–1

5%
52
.9
68
.4
45
.0
43
.5
37
.3
35
.1
39
.7
38
.4
44
.0
36
.7
29
.4
40

.2
15
–2
0%
1.
8
3.
8
4.
9
3.
3
1.
4
1.
2
2.
9
9.
2
1.
3
0.
6

0.
3
2.
2
G
re
at
er
th
an
20
%
5.
2
3.
8
4.
0
3.
6
3.
3
4.
4

4.
4
5.
8
4.
4
2.
5
2.
2
3.
4
T
ab
le
A
.4
.a
�
C
on

ti
nu
ed
.
S
co
re
ca
rd
:
S
tr
es
s
P
ri
m
e
19
99
20
00
20
01

20
02
20
03
20
04
20
05
20
06
20
07
20
08
20
09
A
ll
L
es
s
th
an
5%

27
.0
20
.4
20
.3
15
.2
12
.1
8.
8
7.
3
3.
8
3.
3
8.
4
6.
0
12
.8

5–
10
%
49
.9
54
.5
61
.5
59
.6
54
.8
51
.0
46
.9
38
.7
37
.2
51
.2
48

.1
50
.6
10
–1
5%
16
.1
17
.2
12
.2
19
.0
25
.4
29
.4
28
.6
33
.9
31
.5

23
.3
31
.0
23
.7
15
–2
0%
3.
7
3.
7
3.
1
3.
4
4.
2
6.
3
9.
2
10

.3
10
.8
8.
5
10
.6
6.
2
G
re
at
er
th
an
20
%
3.
2
4.
3
2.
9
2.

8
3.
6
4.
5
8.
0
13
.4
17
.1
8.
6
4.
2
6.
7
S
ub
pr
im
e
19

99
20
00
20
01
20
02
20
03
20
04
20
05
20
06
20
07
20
08
20
09
A
ll
L

es
s
th
an
5%
0.
0
0.
0
0.
0
0.
0
0.
0
0.
0
0.
0
0.
0
0.
0
0.

0
0.
0
0.
0
5–
10
%
0.
0
0.
0
0.
0
0.
0
0.
0
0.
0
0.
0
0.
0

0.
0
0.
7
1.
6
0.
0
10
–1
5%
0.
0
0.
5
0.
0
0.
0
0.
1
0.
3
0.

2
0.
3
0.
2
0.
0
0.
0
0.
2
15
–2
0%
1.
2
1.
9
0.
9
1.
0
2.
1

2.
1
3.
4
3.
8
2.
5
3.
2
4.
7
2.
1
G
re
at
er
th
an
20
%
98
.8

97
.6
99
.1
99
.0
97
.8
97
.5
96
.4
95
.9
97
.3
96
.1
93
.8
97
.7
G
ov

er
nm
en
t
19
99
20
00
20
01
20
02
20
03
20
04
20
05
20
06
20
07
20
08

20
09
A
ll
L
es
s
th
an
5%
0.
0
0.
0
0.
0
0.
0
0.
0
0.
0
0.
0

0.
0
0.
0
0.
0
0.
1
0.
0
5–
10
%
1.
0
0.
6
2.
6
4.
0
5.
5
6.

6
6.
1
6.
5
5.
4
7.
8
9.
0
5.
2
10
–1
5%
18
.4
10
.6
21
.8
25
.9

28
.8
29
.9
27
.8
25
.8
30
.6
33
.9
35
.3
27
.3
15
–2
0%
53
.1
35
.4
44

.1
44
.5
44
.4
41
.8
36
.4
31
.5
38
.2
40
.6
40
.8
42
.1
G
re
at
er
th

an
20
%
27
.5
53
.3
31
.5
25
.6
21
.2
21
.6
29
.7
36
.1
25
.8
17
.7
14

.8
25
.4
T
ab
le
A
.4
.b
�
S
ha
re
of
to
ta
l
de
fa
ul
ts

by
ri
sk
th
re
sh
ol
d
ra
ng
e,
ta
rg
et
po
pu
la
ti
on
.
S
co
re
ca

rd
:
T
hr
ou
gh
-t
he
-C
yc
le
P
ri
m
e
19
99
20
00
20
01
20
02

20
03
20
04
20
05
20
06
20
07
20
08
20
09
A
ll
L
es
s
th
an
5%
62
.8

61
.4
71
.2
51
.5
45
.9
38
.0
31
.0
23
.8
15
.3
31
.5
13
.9
35
.1
5–
10

%
26
.7
25
.2
20
.8
34
.4
39
.7
46
.2
44
.8
45
.3
47
.1
41
.0
66
.7
40

.3
10
–1
5%
5.
1
7.
8
3.
6
7.
1
5.
4
5.
3
9.
8
13
.8
15
.0
11
.5

9.
4
10
.4
15
–2
0%
3.
8
3.
1
1.
3
2.
9
3.
8
4.
5
6.
4
6.
5
7.

8
5.
7
0.
0
5.
4
G
re
at
er
th
an
20
%
1.
6
2.
6
3.
0
4.
0
5.

2
5.
9
7.
9
10
.6
14
.7
10
.3
10
.0
8.
8
S
ub
pr
im
e
19
99
20
00

an
5%
0.
0
0.
0
0.
0
0.
0
0.
0
0.
0
0.
0
0.
0
0.
0
0.
0
n/
a

0.
0
5–
10
%
2.
4
0.
8
2.
4
0.
9
2.
6
2.
7
2.
5
3.
6
1.
1
0.

0
n/
a
2.
4
10
–1
5%
12
.1
9.
7
9.
8
16
.7
23
.0
18
.1
16
.3
15
.1

9.
8
11
.3
n/
a
15
.1
15
–2
0%
31
.6
28
.7
23
.5
34
.9
35
.3
30
.0
26

.4
21
.3
28
.4
48
.0
n/
a
27
.5
G
re
at
er
th
an
20
%
53
.9
60
.9
64

.3
47
.4
39
.0
49
.1
54
.7
59
.9
60
.7
40
.7
n/
a
55
.0
G
ov
er
nm

en
t
19
99
20
00
20
01
20
02
20
03
20
04
20
05
20
06
20
07
20
08
20
09

A
ll
L
es
s
th
an
5%
0.
2
0.
1
0.
6
1.
0
0.
7
0.
9
1.
7
0.
8

0.
5
1.
1
2.
1
1.
0
5–
10
%
31
.6
26
.6
31
.7
39
.9
47
.4
46
.3
48

.2
46
.0
43
.5
53
.4
59
.5
44
.7
10
–1
5%
58
.7
57
.8
49
.4
41
.4
38
.6

40
.1
39
.0
36
.5
45
.7
37
.1
31
.2
42
.5
15
–2
0%
1.
7
2.
4
7.
1
4.

7
2.
1
2.
1
2.
8
3.
3
0.
7
0.
1
0.
3
2.
3
G
re
at
er
th
an
20

%
7.
9
13
.0
11
.3
13
.1
11
.1
10
.7
8.
3
13
.4
9.
6
8.
3
6.
8
9.

6
T
ab
le
A
.4
.b
�
C
on
ti
nu
ed
.
S
co
re
ca
rd
:
S

tr
es
s
P
ri
m
e
19
99
20
00
20
01
20
02
20
03
20
04
20
05
20
06
20

07
20
08
20
09
A
ll
L
es
s
th
an
5%
25
.8
17
.3
13
.0
10
.2
6.
9
6.

0
3.
7
1.
9
0.
8
3.
8
3.
1
6.
4
5–
10
%
43
.4
48
.6
62
.1
47
.3

45
.0
38
.7
34
.6
27
.6
21
.4
35
.0
18
.6
34
.9
10
–1
5%
17
.4
17
.9
13

.7
24
.6
31
.3
34
.7
31
.1
32
.0
32
.9
27
.1
38
.9
28
.3
15
–2
0%
6.
6

7.
6
4.
7
6.
1
3.
6
6.
4
9.
2
11
.7
12
.4
11
.1
22
.1
9.
4
G
re

at
er
th
an
20
%
6.
9
8.
6
6.
5
11
.8
13
.2
14
.2
21
.4
26
.9
32
.5

23
.1
17
.3
21
.0
S
ub
pr
im
e
19
99
20
00
20
01
20
02
20
03
20
04
20

05
20
06
20
07
20
08
20
09
A
ll
L
es
s
th
an
5%
0.
0
0.
0
0.
0

0.
0
0.
0
0.
0
0.
0
0.
0
0.
0
0.
0
n/
a
0.
0
5–
10
%
0.
0
0.

0
0.
0
0.
0
0.
0
0.
0
0.
0
0.
0
0.
0
0.
0
n/
a
0.
0
10
–1
5%

0.
0
0.
0
0.
0
0.
0
0.
0
0.
1
0.
1
0.
1
0.
0
0.
0
n/
a
0.
1

15
–2
0%
0.
8
0.
8
1.
3
0.
3
0.
8
1.
1
0.
7
1.
7
0.
6
0.
0
n/

a
1.
0
G
re
at
er
th
an
20
%
99
.2
99
.2
98
.7
99
.7
99
.2
98
.8
99

.1
98
.2
99
.4
10
0.
0
n/
a
98
.9
G
ov
er
nm
en
t
19
99
20
00
20
01

0.
0
0.
0
0.
0
0.
0
0.
0
0.
0
0.
0
0.
0
0.
0
0.
0
0.
0
0.
0

5–
10
%
0.
3
0.
7
1.
3
2.
9
2.
7
3.
4
6.
1
4.
0
2.
6
5.
3
6.

7
3.
4
10
–1
5%
12
.9
12
.1
15
.4
19
.7
24
.1
22
.9
26
.2
27
.4
26
.5

31
.1
34
.1
23
.7
15
–2
0%
59
.0
47
.8
41
.6
39
.9
41
.1
42
.6
32
.6
33

.0
40
.7
41
.8
39
.2
42
.9
G
re
at
er
th
an
20
%
27
.8
39
.4
41
.6
37

.5
32
.1
31
.1
35
.1
35
.7
30
.2
21
.8
20
.0
30
.0
Figure A.1.b � Conventional Prime Estimation Goodness of Fit.
Figure A.2.b � Conventional SubPrime Estimation Goodness of
Fit.

Figure A.3.b � Conventional Estimation Goodness of Fit.
Copyright of Real Estate Economics is the property of Wiley-
Blackwell and its content may
not be copied or emailed to multiple sites or posted to a listserv
without the copyright holder's
express written permission. However, users may print,
download, or email articles for
individual use.
CHAPTER 4 Fathoming the Fundamentals 65
Chapter 4
IN THIS CHAPTER
» Understanding the basic building
blocks of mortgages
» Looking at mortgage terminology
» Finding out about prepayment
penalties and private mortgage
insurance
Fathoming the
Fundamentals

Like brain surgeons, nuclear physicists, pizza makers, and all
other highly skilled professionals, financial wizards have
developed their own weird customs, practices, and terminology
over the centuries. If you want to do
business with financiers, knowing how to speak their language
helps, because
they rarely bother to speak yours. A steady diet of jumbo loan
with points au gratin
on the side and the infamous house specialty, prepayment
penalty flambé, for des-
sert leaves even the hardiest borrower intellectually constipated.
Worse, some unscrupulous lenders may use your fiscal
ignorance to maneuver you
into getting a loan that’s good for them but bad for you. Even
though an assort-
ment of loans may outwardly appear to be equally attractive,
they’re usually
not — not by a long shot.
The good news is that lending ain’t rocket science. This chapter
explains what
makes a loan tick and helps you speak the language of lending
like a pro. (Chapter 5
takes you through the particulars of choosing the best loan for
you.)
C
op
yr

rig
ht
s
re
se
rv
ed
.
66 PART 2 Locating a Loan
Grasping Loan Basics: Principal, Interest,
Term, and Amortization
Money isn’t magical. It’s a commodity or consumer product like
HDTVs and toast-
ers. Lending institutions such as banks, savings and loan
associations (S & Ls),
and credit unions get their raw material (money) in the form of
deposits from
millions of people just like you. Then they bundle your cash
into neat little pack-
ages called loans, which they sell to other folks who use the
money to buy cars,
college educations, and cottages. Lenders make their profit on
the spread (differ-
ential) between what they pay depositors to get money and what
they charge bor-
rowers to use the money until the lender is fully repaid.

All loans have the following four basic components:
» Principal: Even though both words are spelled and
pronounced the same
way, the principal we’re referring to isn’t that humorless old
coot who ruled
your high school with an iron fist. We’re talking about a sum of
money owed
as a debt: the dollar amount of the loot you borrow to acquire
whatever it is
that your heart desires.
» Interest: No linguistic confusion here — interest is what
lenders charge you to
use their product: money. It accumulates over time on the
unpaid balance of
money you borrowed (the outstanding principal) and is
expressed as a
percentage called the interest rate. For instance, you may be
paying an interest
rate of 19.8 percent or more on the unpaid balance of your
credit card debt.
(We recommend that you pay off credit card balances as soon as
possible!)
Consumer interest for outstanding balances such as credit card
debt and a car
loan is not deductible on your federal or state income tax return.
Interest paid
on a home loan, conversely, can be used to reduce your state
and federal
income tax burdens. There’s a major difference in how you
borrow money.
Understanding these income tax write-off rules can save you big
bucks.

» Term: All good things come to an end sooner or later. A
loan’s term is the
amount of time you’re given by a lender to repay money you
borrow.
Generally speaking, small loans have shorter terms than large
loans. For
instance, your friendly neighborhood credit union may give you
only four
years to pay back a $20,000 car loan. That very same lender
will graciously
fund a loan with a 30-year term so you have plenty of time to
repay the
$200,000 you borrow to buy your dream home.
Lenders allow more time to pay back large loans to make the
monthly
payments more affordable. For example, you’d spend $568 a
month to repay
a $100,000 loan with a 5.5 percent interest rate and a 30-year
term. The same
loan costs $818 a month with a 15-year term. Even though the
15-year loan’s
C
op
yr
ig
ht
©

re
se
rv
ed
.
payment is $250 per month higher, you’d pay far less interest
on it over the
life of the loan:
$818/month × 180 months for a $100,000 loan repayment =
$47,240 in
interest over 15 years
versus
$568/month × 360 months for a $100,000 loan repayment =
$104,480
interest over 30 years
Don’t let a seemingly low monthly payment (with a longer-term
loan) fool you
into paying a lot more interest over the long haul.
» Amortization: Amortization is an ominous word lenders use
to describe the
tedious process of liquidating a debt by making periodic
installment payments
throughout the loan’s term. Loans are amortized (repaid) with

monthly
payments consisting primarily of interest during the early years
of the loan
term and principal, which the lender uses to reduce the loan’s
balance. If your
loan is fully amortized, it will be repaid in full by the time
you’ve made your
final loan payment. You’ll gasp in astonishment and sadness
when you read
Appendix B and see with your own eyes how long it takes to
repay half of the
original loan amount.
Deciphering Mortgage Lingo
Just for the heck of it, ask the next thousand people you meet
what a mortgage is.
Approximately 999 of them will tell you that it’s a loan used to
buy a home.
Amazingly, every one of them is wrong. Common usage aside, a
mortgage is not
simply a loan. This section clarifies what a mortgage is and
isn’t.
So . . . what’s a mortgage?
Mortgage is a word lenders use to describe a formidable pile of
legal documents
you have to sign to get the money you need to buy or refinance
real property. What’s
real property? It’s dirt — plain old terra firma and any
improvements (homes,
garages, cabanas, swimming pools, tool sheds, barns, or other
buildings) perma-
nently attached to the land.
Mortgages aren’t used only to facilitate home purchases.

They’re used whenever
people acquire any kind of real property, from vacant lots to
commercial real
estate such as shopping centers and the Empire State Building.
C
op
yr
ig
ht
©
2
01
7.
J
oh
n
W
ile
y
&
S
on

s,
In
co
rp
or
at
ed
. A
ll
rig
ht
s
re
se
rv
ed
.
In case you’re curious, anything that isn’t real property is
classified as personal
property. Moveable or impermanent possessions such as stoves,
refrigerators,
dishwashers, washers and dryers, window treatments, flooring,

chandeliers, and
fireplace screens are examples of personal property items that
are frequently
included in the sale of real property.
Mortgages encumber (burden) real property by making it
security for the repay-
ment of a debt. A first mortgage ever so logically describes the
very first loan
secured by a particular piece of property. The second loan
secured by the same
property is called a second mortgage, the third loan is a third
mortgage, and so on.
You may also hear lenders refer to a first mortgage as the senior
mortgage. Any
subsequent loans are called junior mortgages. Money imitates
life.
This type of financial claim on real property is called a lien.
Proper liens invariably
have two integral parts:
» Promissory note: This note is the evidence of your debt, an
IOU that specifies
exactly how much money you borrowed as well as the terms and
conditions
under which you promise to repay it.
» Security instrument: If you don’t keep your promise, the
security instrument
gives your lender the right to take steps necessary to have your
property sold
to satisfy the outstanding balance of the debt. The legal process
triggered by
the security device is called foreclosure. We sincerely and
fervently hope that

the closest you ever get to foreclosure is reading about it in this
book (see
Chapter 14 for details).
From a lender’s perspective, each junior mortgage (subsequent
mortgage after the
first loan on the property) is increasingly risky, because in the
event of a foreclo-
sure, mortgages are paid off in order of their numerical priority
(seniority). In
plain English, the second mortgage lender doesn’t get one cent
until the first
mortgage lender has been paid in full. If a foreclosure sale
doesn’t generate
enough money to pay off the first mortgage, that’s tough luck
for the second
lender. Due to the added risk, lenders charge higher interest
rates for junior
mortgages.
How to scrutinize security instruments
The security instrument used in your transaction can vary from
one state to the
next depending on where the property you’re financing is
located. Mortgages and
deeds of trust are the most common types of security
instruments. Without fur-
ther ado, we give you some important information about them.
C
op

ll
rig
ht
s
re
se
rv
ed
.
Mortgages as security instruments
As a legal concept, mortgages have been around centuries
longer than deeds of
trust, their relatively newfangled siblings. That’s why folks
nearly always refer to
real property loans as mortgages even if they live in one of the
many states where
a deed of trust is the dominant security instrument. The other
states use mort-
gages as security instruments.
The seniority of mortgages explains why they’re the prevalent
security instru-
ment in many states east of the Mississippi River, the first part
of the country to
be settled. Check with your real estate agent or lender to find
out which kind of

security instrument is used where your property is located.
Here’s how mortgages operate:
» Type of instrument: A mortgage is a written contract that
specifies how your
real property will be used as security for a loan without actually
delivering
possession of the property to your lender.
» Parties: A mortgage has two parties — the mortgagor (that’s
you, the
borrower) and the mortgagee (the lending institution). You
don’t get a
mortgage from the lender. On the contrary, you give the lender
a mortgage on
your property. In return, the mortgage holder (lender) loans you
the money
you need to purchase the property.
» Title: Title refers to the rights of ownership you have in the
property. A
mortgage requires no transfer of title. You keep full title to your
property.
» Effect on title: The mortgage creates a lien against your
property in favor of
the lending institution. If you don’t repay your loan, the lender
usually has to
go to court to force payment of your debt by instituting a
foreclosure lawsuit.
If the judge approves the lender’s case against you, the lender is
given
permission to hold a foreclosure sale and sell your property to
the highest
bidder.

Deeds of trust as security instruments
Mortgages and deeds of trust are both used for exactly the same
purpose: They
make real property security for money you borrow. However,
mortgages and deeds
of trust use significantly different methods to accomplish that
same purpose. The
following list highlights the features of a deed of trust:
» Type of instrument: The security given isn’t a written
contract. It’s a special
kind of deed called a trust deed.
C
op
yr
ig
ht
©
2
01
7.
J
oh
n

» Parties: The trust deed involves three parties: a trustor (you,
the borrower), a
beneficiary (the lender), and a trustee (a neutral third party such
as a title
insurance company or lawyer who won’t show any favoritism to
you or the
lender).
» Title: The trust deed conveys your property’s naked legal
title to the trustee,
who holds it in trust until you repay your loan. Don’t worry,
dear reader; you
retain possession of the property. Your lender holds the actual
trust deed and
note as evidence of the debt.
» Effect on title: Like a mortgage, a trust deed creates a lien
against your
property. Unlike a mortgage, however, the lender doesn’t have
to go to court
to foreclose on your property. In most states, the trustee has
power of sale,
which can be exercised if you don’t satisfy the terms and
conditions of your
loan. The lender simply gives the trustee written notice of your
default and
then asks the trustee to follow the foreclosure procedure
specified by the
deed of trust and state law. Most lenders prefer having their
loans secured by
a deed of trust. Why? Compared to a mortgage, the foreclosure
process is

much faster and less expensive.
For simplicity’s sake in this book, we use mortgage, deed of
trust, and the loan you
get to buy a home as interchangeable terms. You, however, must
promise us that
you’ll always remember the difference and who explained it to
you!
Eyeing Classic Mortgage Jargon Duets
Just because you can speak mortgage fluently doesn’t mean
you’ll be able to com-
municate with lenders. The following sections offer more
essential loan jargon.
Consider these dynamic duos: mortgage loan options such as
fixed or adjustable
rate, government or conventional, primary or secondary,
conforming or jumbo,
and long- or short-term.
Fixed or adjustable loans
FRM, ARM, or whatever — don’t let the alphabet soup of
mortgages available
today confuse you. No matter how complicated the names
sound, all loans fall into
one of the following basic classifications:
» Fixed: This type of loan either has an interest rate or a
monthly payment that
never changes. A fixed-rate mortgage (FRM) is just what it
claims to be — a
mortgage that keeps the same interest rate throughout the life of
the loan.

or
at
ed
. A
ll
rig
ht
s
re
se
rv
ed
.
Even though you have a fixed-rate mortgage, your monthly
payment may vary
if you have an impound account (for folks who put less than 20
percent cash
down when purchasing their homes). In addition to the monthly
loan pay-
ment, some lenders collect additional money each month for the
prorated
monthly cost of property taxes and homeowners insurance. The
extra money
is put into an impound account by the lender, who uses it to pay
the borrow-

er’s property taxes and homeowners insurance premiums when
they’re due.
If either the property tax or the insurance premium happens to
change (and
they do typically increase annually), the borrower’s monthly
payment is
adjusted accordingly.
» Adjustable: Either the interest rate or the monthly payment or
both interest
rate and monthly payment change (adjust) with this kind of
loan. The follow-
ing are examples of adjustable mortgages:
• An adjustable-rate mortgage (ARM) is a loan whose interest
rate can vary
during the loan’s term.
• A hybrid loan merges an FRM and an ARM. The hybrid loan’s
interest rate
and monthly payment are fixed for a specific period of time,
such as five
years, and then the mortgage converts into an ARM for the
remainder of
the loan term.
Just because a mortgage’s monthly payment is fixed doesn’t
mean the loan is a
good one. For instance, some ARMs have monthly payments
that don’t always
change, even though the loan’s interest rate can change and
increase. This can
lead to negative amortization, an unpleasant situation where the
loan balance
increases every month, even though you faithfully make the
monthly loan pay-

ments. After the subprime crisis, few lenders offer negative
amortization loans.
You can find an in-depth analysis of ARMs and negative
amortization in Chapter 5.
For now, be advised that we strongly urge you to avoid loans
that have the poten-
tial for negative amortization.
Government or conventional loans
Through either insuring or guaranteeing home loans by an
agency of the federal
government, Uncle Sam is a major player in the residential
mortgage market. Such
mortgages are called, you guessed it, government loans. The
remaining residential
mortgages originated in the United States are referred to as
conventional loans.
Here’s a quick recap of government loans:
» Federal Housing Administration (FHA): The FHA was
established in 1934
during the depths of the Great Depression to stimulate the U.S.
housing
market. It primarily helps low-to-moderate income folks get
mortgages by
C
op
yr

rig
ht
s
re
se
rv
ed
.
issuing federal insurance against losses to lenders who make
FHA loans. The
FHA is not a moneylender. Borrowers must find an FHA-
approved lender such
as a credit union, bank, or other conventional lending institution
willing to
grant a mortgage that the FHA then insures. Not all commercial
lenders
choose to participate in FHA loan programs due to their
complexities.
Depending on which county within the United States the home
you want to
buy is located, you may be able to get an FHA-insured loan of
up to $636,150.
The minimum loan amount under this program is $275,665 with
a $636,150
maximum as of 2017. The loan limit varies based on the cost of
housing in each

area. (For current, up-to-date lending limits by area, visit the
FHA Mortgage
Limits web page at
https://entp.hud.gov/idapp/html/hicostlook.cfm.)
» Department of Veterans Affairs (VA): Congress passed the
Serviceman’s
Readjustment Act, commonly known as the GI Bill of Rights, in
1944. One of its
provisions enables the VA to help eligible people on active duty
and veterans
buy primary residences. Like the FHA, the VA has no money of
its own. It
guarantees loans granted by conventional lending institutions
that participate
in VA mortgage programs. This can be an excellent program if
you qualify.
» U.S. Department of Agriculture (USDA): The USDA oversees
the Rural
Housing program. This is a popular program for owner-occupied
homes
outside metropolitan areas. The loans offer $0 down and
affordable mortgage
insurance. However, there are restrictions on location, income,
and assets. If
you qualify, this is usually your best $0 down option, besides a
VA loan.
» Farmers Home Administration (FmHA): Like the FHA, VA,
and USDA, the
FmHA isn’t a direct lender. Despite its name, you don’t have to
be a farmer to
get a Farmers Home Administration loan. You do, however,
have to buy a
home in the sticks. The FmHA insures mortgages granted by

participating
lenders to qualified buyers who live in rural areas.
FHA, VA, and FmHA mortgages have more attractive
features — little or no cash-
down payments, long loan terms, no penalties if you repay your
loan early, and
lower interest rates — than conventional mortgages. However,
these loans aren’t
for everyone. Government loans are targeted for specific types
of homebuyers,
have maximum mortgage amounts established by Congress, and
may require an
inordinately long time to obtain loan approval and funding. In a
desirable urban or
hot market where homes generate multiple offers, buyers using
government loans
often lose out to people using conventional mortgages that can
be funded quicker.
Primary or secondary mortgage market
Lenders make loans directly to folks like you in what’s called
the primary mortgage
market. Few lending institutions keep mortgages they originate
in vaults sur-
rounded by heavily armed guards. Lenders sell most of their
mortgages to pension
C
op

ll
rig
ht
s
re
se
rv
ed
.
https://entp.hud.gov/idapp/html/hicostlook.cfm
funds, insurance companies, and other private investors as well
as certain gov-
ernment agencies in the secondary mortgage market. Why do
mortgage lenders sell
mortgages they originate? They want to make a profit and to
obtain more funds
to lend.
Uncle Sam is an extremely important force in the secondary
mortgage market
through two federally chartered government organizations —
the Federal National
Mortgage Association (FNMA, or Fannie Mae) and the Federal
Home Loan Mortgage
Corporation (FHLMC, endearingly known as Freddie Mac). One
of the primary
missions of Fannie Mae and Freddie Mac is to stimulate

residential housing con-
struction and home purchases by pumping money into the
secondary mortgage
market.
Fannie Mae and Freddie Mac boost home purchases and
construction by purchas-
ing loans from conventional lenders and reselling them to
private investors.
These government programs are far and away the two largest
investors in U.S.
mortgages.
These programs aren’t meant to subsidize rich folks. To that
end, Congress estab-
lishes upper limits on mortgages Fannie Mae and Freddie Mac
are authorized to
purchase. Table 4-1 shows the 2017 maximum mortgage
amounts for one- to
four-unit properties. Note: These are the general loan limits for
most areas, but if
you’re buying a property in a so-called “high-cost” area, the
maximum mortgage
amounts are 50 percent higher than those in Table 4-1.
Congress periodically readjusts these maximum mortgage
amounts to reflect
changes in the prevailing average price of property. Any good
lender can fill you
in on Fannie Mae’s and Freddie Mac’s current loan limits.
TABLE 4-1 2017 Fannie Mae and Freddie Mac Maximum
Mortgage
Amounts for One- to Four-Unit Properties
# of Units Continental U.S.

Alaska, Hawaii, Guam &
U.S. Virgin Islands
1 $424,100 $636,150
2 $543,000 $814,500
3 $656,350 $984,525
4 $815,650 $1,223,475
C
op
yr
ig
ht
©
2
01
7.
J
oh
n
W
ile

y
&
S
on
s,
In
co
rp
or
at
ed
. A
ll
rig
ht
s
re
se
rv
ed
.

Conforming or jumbo loans
This delicious tidbit of information can save you big bucks:
Conventional mort-
gages that fall within Fannie Mae’s and Freddie Mac’s loan
limits are referred
to as conforming loans. Mortgages that exceed the maximum
permissible loan
amounts are called jumbo loans or nonconforming loans.
When Congress passed the Economic Stimulus Act of 2008 (The
Act), it also cre-
ated a brand-new type of mortgage neatly notched between a
conforming loan
and a jumbo loan. We now have three tiers of mortgages:
» True conforming loans include loan amounts up to $424,100.
These loans,
also called traditional conforming loans, have the lowest
interest rates.
» Jumbo conforming loans encompass loan amounts from
$424,100 up to a
maximum of $636,150 and are designed for high-cost areas (the
precise
amount varies by area). Some lenders call these conforming
jumbos, super
conforming, or jumbo light loans. Whatever. Loans of this size
generally have
interest rates anywhere from half a percent to a full percent (or
more) higher
than the true conforming loan.
» True jumbos are loans that exceed $636,150. As you’d
expect, the largest
loans are also the most expensive. Their interest rates usually

run a full
percent point or more above jumbo conforming loans.
Fannie Mae and Freddie Mac both imposed tougher qualifying
standards on jumbo
conforming loans than they have for true conforming loans.
Some examples of
these tougher standards: Jumbo conforming loans are limited to
single-family
dwellings, require that you have at least a 700 FICO score if
your loan-to-value
(LTV) ratio exceeds 75 percent (for Freddie Mac) or 80 percent
(for Fannie Mae),
and specify that monthly payments on your combined total debt
can’t exceed
45 percent of your income.
Fannie’s and Freddie’s jumbo conforming loan programs were
originally sched-
uled to expire December 31, 2008, but Congress keeps
extending them, and these
programs are still in place as of 2017. Be sure to check with
your lender regarding
the current status of these loans.
You pay dearly for nonconformity. The higher the loan amount,
the bigger the
thud if your loan goes belly up. Reducing the loan-to-value ratio
is one way lend-
ers cut their risk. To that end, conventional lenders generally
insist on more than
the usual 20 percent down on jumbo loans. You’ll probably be
required to make at
least a 25 percent cash down payment. Interest rates on
nonconforming fixed-
rate mortgages generally run from 3⁄8 to ½ a percentage point

higher than con-
forming FRMs. When mortgage money is tight, the interest rate
spread between
C
op
yr
ig
ht
©
2
01
7.
J
oh
n
W
ile
y
&
S
on

s,
In
co
rp
or
at
ed
. A
ll
rig
ht
s
re
se
rv
ed
.
conforming and jumbo FRMs is higher; when mortgage money
is plentiful, the
spread decreases.
If you find yourself slightly over Fannie Mae’s and Freddie
Mac’s limit for either

true conforming loans or the jumbo conforming loans, don’t
despair. You can
either buy a slightly less expensive home or increase your cash
down payment
just enough to bring your mortgage amount under their loan
limits or possibly use
a small second mortgage. In Chapter 2, we include a lengthy
list of financial
resources you may be able to tap for additional cash.
Long-term or short-term mortgages
Any loan that’s amortized more than 30 years is considered to
be a long-term
mortgage. Reversing that guideline, short-term mortgages are
loans that must be
repaid in less than 30 years. Wow. Definitions that actually
make sense.
These standards harken back to less complicated times before
the late 1970s when
people could get any kind of mortgage they wanted as long as it
was a 30-year,
fixed-rate loan. Back then, choices for a short-term mortgage
were nearly as
limited. Homebuyers could have an FRM with either a 10- or
15-year term or a
balloon loan with, for example, a 30-year amortization schedule
and a 10-year due
date. They made the same monthly principal and interest
payments for ten years
and then got hammered with a massive balloon payment to pay
off the entire
remaining loan balance. (The reality was that homeowners
simply had to refi-
nance the remaining loan balance through a new loan either with
their current

lender or another lender.)
The total interest charges on short-term mortgages are less than
total interest
paid for equally large long-term loans at the same interest rate
because you’re
borrowing the money for less time. Because a lender has less
risk with a short-
term loan, such loans usually have lower interest rates than
comparable long-term
mortgages. For instance, the interest rate on a conforming 15-
year, fixed-rate
mortgage is generally about ½ a percentage point lower than a
comparable
30-year FRM.
In our prior example (see the section “Grasping Loan Basics:
Principal, Interest,
Term, and Amortization”), we say that you’d spend $568 a
month to repay a
$100,000 FRM with a 5.5 percent interest rate and a 30-year
term. The same FRM
with a 15-year term and 5.5 percent interest rate costs $818 a
month. If that loan
has a 5 percent interest rate, its payment would drop to $791 per
month. The half-
point interest rate cut saves you an additional $4,860 over the
life of the loan
($818 – $791 = $27 per month × 180 months). Not too shabby!
C

. A
ll
rig
ht
s
re
se
rv
ed
.
Even though short-term loans have lower interest rates than
their long-term
cousins, qualifying for a short-term loan is more difficult due to
the higher
monthly loan payments. Lenders generally don’t want you
spending much more
than 30 to 35 percent of your gross monthly income on
mortgage payments. Even
if you can qualify for a short-term loan, it may not be in your
best interests (pun
intended) to irrevocably lock yourself into the higher monthly
payments. Will
higher loan payments deplete the cash reserves you ought to
maintain for emer-
gencies? Can you afford higher loan payments and still
accomplish all the other

financial goals we cover in Chapter 1? We devote Chapter 12 to
a stimulating anal-
ysis of the pros and cons of paying off a mortgage more rapidly
than is required
by the lender.
Introducing the Punitive Ps
Certain warnings are drilled into people until they become as
reflexive as the way
your leg convulsively jerks when a doctor hits your knee with
that little pointy
rubber hammer. Don’t stuff yourself on sweets just before
sitting down to a good,
healthy meal. Don’t forget to floss and brush your teeth. Don’t
drink and drive.
Think before you post something on social media! Other
injurious hazards are
more insidious. The following sections offer words to the wise
about two of them
related to mortgages.
Prepayment penalties
Some lenders punish borrowers severely for repaying all or part
of their conven-
tional loan’s remaining principal balance before its due date. As
punishment, they
impose a charge known as a prepayment penalty. Prepayment
penalties aren’t
permitted on FHA, VA, USDA, and FmHA mortgages (see the
earlier section, “Gov-
ernment or convention loans,” for more information on these
kinds of mortgages).
How much money are we talking about? That depends.
Maximum permissible
prepayment penalties vary widely from state to state, from one

lender to the
next — and even from one loan to the next on mortgages
offered by the same
lending institution. Some lenders will waive the prepayment
penalty if you get a
new loan from them when you refinance your mortgage or if
you’re forced to pay
off the loan because you sell your house.
Less sympathetic souls force you to pay upward of 3 percent on
your unpaid loan
balance, which equals $3,000 on every $100,000 you prepay.
Even less humane
lenders may insist on a penalty equal to six month’s interest on
your outstanding
loan balance. If, for example, your mortgage’s interest rate is 8
percent per annum,
you’d have to pay $4,000 per $100,000 of principal you repay
early.
C
op
yr
ig
ht
©
2
01

7.
J
oh
n
W
ile
y
&
S
on
s,
In
co
rp
or
at
ed
. A
ll
rig
ht
s
re
se

rv
ed
.
Now that we have your attention, here’s how to determine
whether the lender can
impose a prepayment penalty:
» Ask: Now that you know what to ask, don’t be shy. Look
your loan officer right
in the eye and specifically inquire whether the loan you’re
considering has a
prepayment penalty. If it does, we strongly urge you to keep
looking until you
find another equally wonderful mortgage without a prepayment
penalty.
Some lenders will be willing to negotiate and reduce or even
eliminate the
prepayment penalty — all you have to do is ask!
» Read: Even if the lender says the loan doesn’t have a
prepayment penalty, don’t
take chances. Verify that the mortgage doesn’t have a
prepayment penalty
clause by carefully reading the federal truth-in-lending
disclosure you’ll receive
from the lender soon after submitting your loan application.
Even good lenders
frequently don’t know the nuances of every single loan they
offer.

» Read again: Check, double-check, and check again. You must
scrutinize one
last document to be sure that your loan doesn’t have a
prepayment penalty —
the promissory note. Read it with care. Make sure a prepayment
penalty
clause doesn’t somehow manage to mysteriously creep into your
mortgage
before you sign the final loan documents.
Some mortgages have soft prepayment penalties, which may be
waived at the lend-
er’s discretion if you sell an owner-occupied one- to four-unit
property after
you’ve owned the property at least one year. Soft prepayment
penalties are infi-
nitely preferable to hard prepayment penalties, which are
always enforced without
exception.
You may be tempted to get a loan with a prepayment penalty,
because you’re abso-
lutely certain that there’s no way you’ll ever pay it off early.
Trust us when we say
that circumstances have a way of changing when you least
expect them to. Utterly
unforeseen life changes force folks to sell property whether they
want to or not.
Divorce happens. People find their employer has transferred
them to another state
or worse — fired them! Folks pass away prematurely. Life
happens.
You may decide, in your infinite wisdom, to get a mortgage that
has a prepayment

penalty. Fine. If your mom couldn’t make you eat your
vegetables, how can we
make you follow our sage advice? At least make sure that you
completely under-
stand the terms and conditions of your mortgage contract’s
prepayment penalty
clause regarding the following:
» The amount you can prepay without penalty: For instance,
some lenders
permit you to prepay up to 20 percent of your original loan
amount or current
loan balance without penalty each calendar year. Others impose
a penalty
C
op
yr
ig
ht
©
2
01
7.
J
oh

n
W
ile
y
&
S
on
s,
In
co
rp
or
at
ed
. A
ll
rig
ht
s
re
se
rv
ed
.

from the very first dollar of any prepayment. The more you can
prepay
without penalty, the better.
» When you can prepay without penalty: You may be allowed
to prepay a
specific amount of money or percentage of your original loan
balance
quarterly without penalty. Other lenders let you prepay funds
without penalty
only once a year. The faster you can prepay without penalty, the
better.
» The duration of prepayment penalty: Mortgages on owner-
occupied
residential property often specify that the prepayment penalty
expires three
to five years after loan origination. Other home mortgages have
prepayment
penalties over the full term of the loan. The faster the
prepayment penalty
vanishes, the better.
» The severity of prepayment penalty: Some prepayment
penalties diminish
in severity as the mortgage matures. You could, for example, be
penalized
5 percent on any funds prepaid within one year of loan
origination, 4 percent
in the second year, 3 percent for the third year, and so on. Other
mortgages

impose the same vicious penalty as long as the prepayment
clause is in effect.
Declining penalties are better.
Can you tell we’re not big fans of prepayment penalties?
Private mortgage insurance (PMI)
Mortgage insurance protects lenders from losses they may incur
due to the dreaded
double whammy of default and foreclosure. Uncle Sam provides
the mortgage
insurance on government loans (FHA, VA, USDA, and FmHA).
Private insurance
companies provide private mortgage insurance (PMI) on all
other loans.
Who pays for this insurance? You, of course — if you want a
conventional loan and
can’t make at least a 20 percent cash down payment on the
property you’re buying
or refinancing. (If that doesn’t apply to you, school’s out. You
have our permission
to skip the rest of this chapter.)
“Wait a second,” you say. “That seems incredibly inequitable,
even for lenders.
I pay for the insurance, but my lender gets the proceeds? What’s
in it for me?”
A loan. It’s the only way to get conventional financing with a
low cash down
payment. That’s the deal. Take it or leave.
Twenty percent is a magic number to institutional lenders. They
made a fascinat-
ing empirical discovery after suffering through years of
expensive, unpleasant

experiences with belly-flopped loans. At least a 20 percent
down payment is nec-
essary to protect their investment (the mortgage) if you cut and
run on your loan.
We know you’re wonderful and would never default on your
mortgage. Unfortu-
nately, lenders don’t know you nearly as well as we do.
C
op
yr
ig
ht
©
2
01
7.
J
oh
n
W
ile
y
&

S
on
s,
In
co
rp
or
at
ed
. A
ll
rig
ht
s
re
se
rv
ed
.
Look at things from their perspective. Suppose that you put only
10 percent cash

down. A severe recession occurs, and property values drop 15
percent. You lose
your job because your business fails, and you can’t make your
monthly loan pay-
ments. The lender is forced to take your house away from you in
a foreclosure
action and sell it to satisfy your debt. Farfetched? Hardly. Read
your local paper.
Stranger things happen every day. Witness the jump in
foreclosures in most areas
in the years just before and after the 2008 financial crisis and
recession.
After the poor, misunderstood lender involuntarily takes back
your now vacant
home, fixes it up to make it marketable, and pays the real estate
commission,
property transfer tax, and other customary expenses associated
with the sale of
your house, there won’t be nearly enough money left to pay off
your loan. Your
lender will lose his corporate shirt. If that scenario happens too
often, the lender
goes belly up.
You may be able to deduct your PMI premiums on your federal
tax return. For
loans that commenced after 2006, borrowers with an adjusted
gross income (AGI)
of up to $100,000 may deduct their PMI premiums as they do
mortgage interest
on IRS Form 1040, Schedule A. The deduction is phased out in
10 percent incre-
ments for each $1,000 in increased income above $100,000.
Above $109,000, PMI
isn’t tax deductible.

What you’ll end up spending for PMI depends on the following
factors:
» Type of loan: For example, ARMs generally have higher PMI
premiums than
FRMs. (The previous sentence would have been utterly
unintelligible gibberish
before you read this chapter. See how well you’ve mastered the
lingo? We’re
so proud of you.) If you don’t understand this sentence, check
out the section
“Eyeing Classic Mortgage Jargon Duets,” earlier in this chapter.
» Loan amount: Your PMI premium is partially based on a
percentage of the
loan amount — the more you borrow, the more you’ll pay for
PMI.
» Loan-to-value (LTV) ratio: LTV ratio is the loan amount
divided by the
appraised value of the property you’re buying or refinancing.
The higher the
LTV ratio, the greater the risk of default to the lender and,
hence, the higher
your PMI premium.
» Credit Score: Your PMI premium is also partially based on
your credit score.
» The insurance company issuing your PMI: This is the least
important factor
because PMI charges usually vary relatively little from one
insurance provider
to the next. It can’t hurt, however, to instruct your lender to
shop around for

co
rp
or
at
ed
. A
ll
rig
ht
s
re
se
rv
ed
.
Even though PMI charges don’t usually vary much from one
insurer to the next,
the type of loan they insure and geographical areas of coverage
can vary wildly.
The late 2000s mortgage market problems made lenders more
cautious. Ditto PMI
insurers. MGIC (Mortgage Guaranty Insurance Corporation, the
largest private

mortgage insurer), Radian Group, and Genworth Financial (two
other large insur-
ers) are now much more selective about loans they’ll insure.
Insurers are skittish
now about property in distressed markets where values are
declining and loans
with less than 5 percent cash down. Your lender may have to
shop around to find
a PMI provider who’ll issue your policy.
PMI origination fees and monthly premiums change frequently.
Check with your
lender for specifics on PMI expenses for your loan.
PMI isn’t a permanent condition. You can discontinue it by
proving you have at
least 20 percent equity in your property. Equity is the difference
between your
home’s current market value and what you owe on it. The magic
20 percent can
come from a variety of sources: an increase in property values;
paying down your
loan; improving the property by, for example, modernizing the
kitchen or adding
a second bathroom; or any combination of these factors. To
remove PMI, your
lender will no doubt insist that you have the property appraised
(at your expense,
of course) to establish its current market value. Spending a few
hundred dollars
for an appraisal that’ll save you hundreds or more a year in PMI
expenses is a wise
investment. (We also thoughtfully include a section in
Chapter 6 about how you
may be able to use 80-10-10 financing to avoid paying PMI.)

rp
or
at
ed
. A
ll
rig
ht
s
re
se
rv
ed
.

1 Assessment Task – Tutorial Questions Assignment 1 .docx

More Related Content

Similar to 1 Assessment Task – Tutorial Questions Assignment 1 .docx (20)

More from mercysuttle (20)

Recently uploaded (20)

1 Assessment Task – Tutorial Questions Assignment 1 .docx