Cityscape: A Journal of Policy Development and Research - Housing Tenure and Financial Security

by Jake Interrante and Elliot Schmiedl

187

Cityscape: A Journal of Policy Development and Research • Volume 22, Number 1 • 2020

U.S. Department of Housing and Urban Development • Ofﬁce of Policy Development and Research

Cityscape

Building Wealth through

Homeownership:

A Comparative Study of MHP’s ONE

Mortgage Program and FHA

Jake Interrante

Elliot Schmiedl

Massachusetts Housing Partnership

Abstract

This article compares actual wealth building outcomes in the Massachusetts Housing Partnership’s

(MHP) subsidized ONE Mortgage Program to hypothetical outcomes for a borrower who received a

comparable Federal Housing Administration (FHA) insured mortgage with the same loan amount. We

ﬁnd that ONE Mortgage loans had much lower monthly payments than the FHA loans, resulting in

greater overall beneﬁts to borrowers. Although ONE Mortgage loans delivered slightly lower levels of

equity accumulation at time of sale, the net ﬁnancial outcomes still overwhelmingly favored the ONE

Mortgage loans. These ﬁndings are concerning given the large market share of FHA loans among low-

and moderate-income (LMI) and minority homebuyers in Massachusetts. If these borrowers could have

qualiﬁed for the ONE Mortgage program but instead received an FHA loan, our analysis suggests that

they would have lost out on signiﬁcant beneﬁts. For the ﬁrst time, this article quantiﬁes the scale of that

potential loss to Massachusetts’s LMI ﬁrst-time homebuyers.

188 Housing Tenure and Financial Security

Interrante and Schmiedl

Introduction

Home prices in Massachusetts have reached an all-time high (Federal Reserve Economic Data

[FRED], 2019). Accordingly, homeownership has grown more elusive for low- and moderate-

income (LMI) borrowers. These borrowers, often low on savings, are increasingly reliant on

specialized high loan-to-value (LTV) mortgage products that allow them to buy a home with a

smaller downpayment. Borrowers in Massachusetts have a large selection of options for high LTV

mortgages, including products provided by the state’s two Housing Finance Agencies (HFAs),

Massachusetts Housing Partnership (MHPs) and MassHousing Finance Agency (MassHousing).

These products offer special benefits that drastically reduce borrowers’ mortgage costs. Despite the

presence of these affordable alternatives, the largest proportion of LMI first-time homebuyers use

Federal Housing Administration (FHA) loans (Campen, 2018).

FHA loans combine low downpayments with reasonably priced mortgage insurance, making them

more accessible than many other high LTV loans (Fannie Mae, 2019). They have facilitated millions

of home purchases by LMI and minority borrowers. Researchers have increasingly noted, however,

that FHA loans are concentrated in LMI communities and among minority borrowers (Caplin,

Cororaton, and Tracy, 2013; Immergluck, 2011). This concentration could carry significant risk,

both for the FHA program overall and for individual borrowers (Lee and Tracy, 2018).

Lending in Massachusetts has mirrored the national trend, despite the presence of alternative

mortgage options provided by state HFAs. Some worry that lenders in the state are too eager to

offer their clients FHA loans, overselling the loans to borrowers who would be better off with an

HFA product. Fortunately, Massachusetts is particularly well-suited to study the use of FHA loans

because borrowers under 100 percent area median income (AMI) have many alternatives. In this

article, we take the first step in addressing the use of specific mortgage products by examining two

programs that have high degrees of targeting to the state’s LMI homebuyers: (1) FHA-insured loans

and (2) MHP’s ONE Mortgage Program.

This analysis quantifies how ONE Mortgage loans and FHA loans differ in terms of overall

household wealth creation. To do so, we use loan-level simulations to model the amount of wealth

created by the ONE Mortgage and FHA loans and compare them on three dimensions: (1) equity

accumulation realized at the time of property sale, (2) savings on monthly mortgage payments, and

(3) a net financial outcome metric that considers the borrower’s overall wealth building.

Our findings suggest that ONE Mortgage loans delivered borrowers a high degree of savings on

monthly payments. This finding was consistent with our expectations, given the substantial impact

of the ONE Mortgage’s no-private mortgage insurance (PMI) benefit, discounted interest rate,

and interest rate subsidy. Although the ONE Mortgages we examined had slightly lower equity

accumulation at time of sale than FHA loans, every loan we simulated had a higher net financial

outcome as a ONE Mortgage loan than it did as an FHA loan.

This research raises important questions about Massachusetts LMI borrowers’ heavy reliance

on FHA loans. We find that borrowers who would have qualified for an MHP loan but instead

borrowed an FHA loan missed out on significant benefits. Our simulations place the average

overall savings at $19,544 in the ONE Mortgage when compared with an FHA loan. We estimate

189Cityscape

Building Wealth through Homeownership:

A Comparative Study of MHP’s ONE Mortgage Program and FHA

that in 2017 alone, a maximum of about 1,500 FHA borrowers may have qualified for the ONE

Mortgage, about 12.5 percent of all FHA borrowers in the state.

In addition to building our understanding of the potential risks of overreliance on FHA lending

and evaluating the outcomes of the ONE Mortgage program, this research extends the literature

regarding the effect of mortgage choice on household wealth. Our work also offers borrowers and

public agencies a framework for comparing borrower outcomes across loan products, which holds

promise to extend our analysis beyond Massachusetts.

History of LMI Lending in Massachusetts

Although the Civil Rights Act of 1968 and the Equal Credit Opportunity Act of 1974 both

outlawed discriminatory practices in the mortgage market, de facto discrimination remained

common. Prime lenders avoided lending in low-income communities and communities of color

(Munnell et al., 1996), allowing these neighborhoods to become a captive market for subprime

lending at high interest rates. Many of these subprime lenders viewed low-income and minority

borrowers as less financially savvy and therefore targeted them with higher-cost, higher-spread

mortgages (Massey et al., 2016).

In 1989, the Federal Reserve Bank of Boston released a study chronicling a pattern of racial bias

in Boston’s mortgage lending over a 5-year period from 1981–1985. The disparities in lending,

the authors suggested, could not be explained by income, credit history, or other legitimate loan

underwriting factors (Marantz, 1989). The Federal Reserve study laid bare systemic disparate

treatment in the mortgage market and resulted in a massive public outcry. In the wake of that

report, a task force was created comprised of homeownership practitioners from MHP, the

Massachusetts Bankers Association, the Commonwealth of Massachusetts, the City of Boston, and

the Massachusetts Affordable Housing Alliance. This group’s mission was to form a strategy that

would counter the pervasive influence of racial bias in the state’s mortgage market.

The result of this effort was the introduction of two new affordable loan products: the SoftSecond

Program (1991–2013) and then the ONE Mortgage Program (2014–present). These mortgage

products were designed to address traditional barriers to homeownership and close the wealth gap

by providing increased mortgage affordability to LMI and minority first-time homebuyers. The

programs were housed within MHP, a quasi-public state housing finance agency founded in 1985

that works to increase the supply of affordable housing in Massachusetts. Since their introduction

in 1991, more than 21,000 Massachusetts households have used one of these loans to purchase

their first home. Two-thirds of ONE Mortgage loans in Boston and half of the loans statewide

support purchases by households of color. The ONE Mortgage program accounts for about 1

percent of all home purchase loans to all borrowers in the State of Massachusetts and 4 percent

of annual home purchase lending to LMI borrowers (Campen, 2018; Massachusetts Housing

Partnership, 2016).

Meanwhile, national trends over the course of the 1990s and 2000s were characterized by the

growth of mortgage lenders originating high-cost loans in the subprime market. The subprime

For a full history of the ONE Mortgage and SoftSecond programs, see Ziegler, Schmiedl, and Callahan (2017).

190 Housing Tenure and Financial Security

Interrante and Schmiedl

lending industry continued to deliver inferior loans in LMI and minority communities. The

subprime lenders that spent the most money advertising in LMI communities tended to have

higher interest rates than other lenders (Gurun, Matvos, and Seru, 2016). Unsurprisingly, an

analysis of lending in seven metropolitan areas found African-American and Hispanic borrowers to

be 105 and 78 percent more likely to receive high cost mortgages, respectively (Bayer, Ferreira, and

Ross, 2017). Lenders also extracted higher closing costs, resulting in African-American borrowers

spending about $700 more on closing costs than White borrowers (Woodward and Hall, 2010).

When the crisis arrived, the deepest distress fell on LMI and minority homeowners. Even in

the years prior to the crisis, African-American households were 68.2 percent more likely than

their White counterparts to transition back to renting at the conclusion of their homeownership

experience (Sharp and Hall, 2014). The crisis magnified that effect. Foreclosure rates for African-

American borrowers spiked to levels over three times that of White borrowers, while Hispanic

homeowners saw foreclosure rates over four times greater than White households (Garriga,

Ricketts, and Schlagenhauf, 2017).

Because credit had been so easily available in the two decades leading up to the crisis, FHA

volumes were low. That changed rapidly, however, when the subprime mortgage crisis arrived in

2008 (Bhutta, Laufer, and Ringo, 2017). The collapse of the Mortgage Backed Securities (MBS)

market meant that lenders looking to sell their originated loans on the secondary market became

more dependent on selling loans to Fannie Mae and Freddie Mac. To do so, any loan with less

than a 20-percent downpayment would need PMI. PMI companies, however, were raising their

own underwriting standards. These insurers raised their minimum credit score requirements and

lowered their maximum LTV limits, particularly in distressed areas (Avery et al., 2010). As the

alternative products became less attractive, the share of FHA mortgages as a percentage of all first

lien originations for owner-occupied home purchases increased from 5.7 percent in 2006 to 40.8

percent in 2009. As of 2016, 24.4 percent of all first lien originations for owner-occupied home

purchases are FHA loans (Bhutta, Laufer, and Ringo, 2017).

Today, the worst subprime abuses of the crisis are over. Recent research on the mortgage market in

Massachusetts suggests, however, that class and racial disparities may continue to impact the FHA

loan market. FHA loans are most heavily marketed by non-depositories; licensed mortgage lenders

are responsible for 77 percent of FHA loans originated in the state (Campen, 2018). In addition,

African-American and Latino borrowers statewide are much more likely to receive FHA-insured

loans than White borrowers; FHA-insured loans accounted for 35 percent of loans to African-

American households in the greater Boston area, but only 7 percent of loans to White households.

Meanwhile, only 2.4 percent of non-FHA home purchase loans were made to African-American

borrowers. FHA lending accounts for 29 percent of LMI home purchase lending in the state,

compared with about 4 percent for the MHP ONE Mortgage (Campen, 2018).

Affordable Loan Options in Massachusetts and Their Features

This analysis centers around three loan products: the SoftSecond Program; its successor, the ONE

Mortgage Program; and FHA-insured 30-year fixed rate loans. The term sheet comparison in

exhibit 1 summarizes the three programs.

191Cityscape

Building Wealth through Homeownership:

A Comparative Study of MHP’s ONE Mortgage Program and FHA

Exhibit 1

Program Term Sheet Comparison (Single-Family/Condominium Guidelines Shown)

MHP SoftSecond

Loan Program ONE Mortgage Program

FHA-Insured Loans

(30-Year Fixed Rate)

Applicable Borrowers

Borrower must be rst-

time homebuyer (per

FNMA denition) and

primary resident

Borrower must be rst-

time homebuyer (per

FNMA denition) and

primary resident

Borrower must be

primary resident

Max Loan-to-Value 97% 97% 96.5%

Income Limits

100% Area Median

Income

100% Area Median

Income

N/A

Minimum Credit Score 620 640

580 (credit scores below

580 are allowed for 90%

Max loan-to-value loans)

Maximum

Underwriting Ratios

(Housing Ratio/Debt to

Income Ratio)

38% / 43% 36% / 43% 40% / 50%

Maximum Asset Limit $75,000 $75,000 N/A

Mortgage Insurance None None Upfront and Monthly

Loan Structure

First Mortgage: up to

77% of purchase price

Second Mortgage:

20% of purchase price;

interest only for the rst

10 years

Single loan

Interest Rate Subsidy

Subsidy applied years

1-9; repaid at time of

renance or sale

Subsidy applied years

1-7; repaid at time of

renance or sale

N/A

Overall Public Funding

$12,000 (loan loss

reserve + full value

subsidy amount)

$12,000 (loan loss

reserve + full value

subsidy amount)

N/A

Education

Requirements

Prepurchase and

Post-Purchase classes

required

Prepurchase and

Post-Purchase classes

required

N/A

FNMA = Federal National Mortgage Association. N/A = information not available.

Source: HUD FHA Single Family Housing Policy Handbook

MHP’s Subsidized Mortgage Programs

SoftSecond Loan Program (1991–2013)

The SoftSecond Loan Program was targeted at first-time homebuyers under 100-percent AMI. It

offered a 30-year fixed interest rate with a minimum 3-percent downpayment. Although the loans

could be offered up to 97-percent LTV, they were originated without mortgage insurance (MI).

The loans avoided MI using a unique loan structure that combined public reserves with private

financing. SoftSecond loans had a dual mortgage structure consisting of a 77-percent first mortgage

and a 20-percent second mortgage, both offered by the participating lender. While the first loan

was a conventional loan with a 30-year amortization schedule, the second loan was interest

192 Housing Tenure and Financial Security

Interrante and Schmiedl

only for the first 11 years. MHP covered lenders’ risk on the second mortgage by setting aside a

percentage of the loan in a dedicated loan loss reserve fund. The average loan loss reserve (LLR)

amount in the SoftSecond Program was $1,979 per loan.

Participating lenders also provided affordability assistance by agreeing to offer these loans at a

discounted interest rate without charging any points. The SoftSecond Program was offered to

borrowers at a discounted interest rate, capped at the Freddie Mac weekly rate minus 30 basis

points.

As a further benefit, eligible homebuyers below 80 percent AMI received a state-funded interest

subsidy. The intent of the interest subsidy is to lower a borrower’s monthly payments early in the

loan in order to transition first-time homebuyers into the higher costs and maintenance demands

of their new home. In the SoftSecond Program, the subsidy was paid out over the course of the

first 10 years of the loan. Borrowers repaid the subsidy when they sold their property or refinanced

out of the program. We provide an example subsidy schedule in exhibit 2. The amount of public

assistance per loan (loan loss reserve plus the full interest subsidy) was capped at $12,000. The

average full value subsidy amount (for loans receiving subsidy) was $7,301. Over the course of the

SoftSecond Program’s history, 7,033 borrowers received interest subsidy, which is about 40 percent

of the 17,410 overall borrowers.

Exhibit 2

Sample SoftSecond Loan Program Subsidy Schedule

Purchase Price: $350,000

Downpayment: $10,500 (3%)

Mortgage Amount: $339,500 (97%)

Interest Rate: 4.00%

Full Value Subsidy Amount: $10,415

Year

1st Mortgage

Payment

(a)

2nd Mortgage

Payment

(b)

2nd Mortgage

Subsidy Amount

(c)

Total Payment

(a + b + c)

1–5 $1,286.63 $233.33 -$123.94 $1,396.03

6 $1,286.63 $233.33 -$99.15 $1,420.82

7 $1,286.63 $233.33 -$74.36 $1,445.61

8 $1,286.63 $233.33 -$49.82 $1,470.15

9 $1,286.63 $233.33 -$24.91 $1,495.06

10 $1,286.63 $233.33 $0.00 $1,519.97

11–30 $1,286.63 $424.19

$0.00 $1,710.82

In year 11, the second mortgage payment increases as the loan transitions from interest only to fully amortizing.

Borrowers are able to receive subsidy forgiveness in certain hardship scenarios when they sell their

property with a relatively small amount of price appreciation. In the first 5 years of their mortgage,

they must repay the lesser of total subsidy received or the total amount of appreciation. After the

first 5 years, they would repay the lesser of the total subsidy received or 20 percent of the total

amount of appreciation.

193Cityscape

Building Wealth through Homeownership:

A Comparative Study of MHP’s ONE Mortgage Program and FHA

ONE Mortgage Loan Program (2013–Present)

Until 2008, the SoftSecond Loan Program’s interest-only 20-percent second mortgage was not

controversial. It provided the benefit of lower monthly payments and eliminated the need for costly

PMI. As the mortgage crisis unfolded nationwide and the amount of delinquencies and defaults

increased, however, MHP decided to investigate alternative program models. Interest-only second

mortgages had gained negative notoriety during the crisis as risky products that were created to

help get people into homes without guarding against rising payments. It was important to not only

find a way to maintain the low monthly payments of the SoftSecond Program, but also to increase

the pace of equity building.

Thus, in 2013 MHP redesigned and rebranded the SoftSecond Loan Program as the ONE

Mortgage Program. Many aspects of the ONE Mortgage were left unchanged from the earlier

SoftSecond model. ONE Mortgages are a 30-year fixed rate mortgage with a minimum 3-percent

downpayment. Like SoftSecond, ONE uses a publicly funded loan loss reserve to offer borrowers

a no-PMI benefit and is offered to the borrower at the same 30-basis-point discount as the

SoftSecond program. It adopted a new structure that incorporated a 97-percent fully amortizing

first mortgage.

Other aspects of the program were altered to comply with post-crisis mortgage lending standards,

however. The ONE Mortgage program abandoned the SoftSecond Program’s two-mortgage

structure in favor of a single 97-percent note. The ONE Mortgage Program also restructured the

loan-loss reserve, accepting a higher share of potential losses. The average ONE Mortgage LLR

since the program’s inception has been $2,364, a 19-percent increase over the average SoftSecond

LLR. Because borrowers build their equity cushion faster with the one-mortgage structure,

however, the likelihood of a loan-loss event is reduced. The ONE Mortgage subsidy also has a

shorter subsidization schedule than the SoftSecond Program, lasting just 7 years compared with

the SoftSecond’s 10 (exhibit 3). The average full-value subsidy for a ONE Mortgage borrower has

been $7,672, 5 percent higher than the average subsidy in the SoftSecond Program. Other changes

were made to MHP’s eligibility and compliance requirements as the program rolled out, including

debt-to-income (DTI) limits, credit score minimums, and liquid asset maximums.

194 Housing Tenure and Financial Security

Interrante and Schmiedl

Exhibit 3

Sample ONE Mortgage Program Subsidy Schedule

Purchase Price: $350,000

Downpayment: $10,500 (3%)

Mortgage Amount: $339,500 (97%)

Interest Rate: 4.00%

Full Value Subsidy Amount: $6,369

Year 1st Mortgage Payment MHP Subsidy

Total Monthly Mortgage

Payment

1–4 $1,621 -$96 $1,525

5 $1,621 -$72 $1,549

6 $1,621 -$49 $1,572

7 $1,621 -$23 $1,598

8–30 $1,621 $0 $1,621

MHP = Massachusetts Housing Partnership.

FHA-Insured Loans

Since 1934, FHA-insured mortgage loans have been one of the Federal government’s premier

products for increasing access to homeownership. The FHA uses federally backed insurance to

facilitate private-sector mortgage lending. FHA has the broadest borrower eligibility guidelines of

the various government-backed loan options. Unlike the SoftSecond and ONE Mortgage Programs,

FHA loans do not have a set maximum income, maximum asset limit, or minimum credit score.

This makes FHA loans an attractive option for borrowers from a range of income classes and levels

of financial health. Also, unlike the SoftSecond and ONE Mortgage Program, FHA requires a 3.5

percent-downpayment and mortgage insurance. FHA insurance premiums are applied in the form

of an initial payment, the upfront mortgage insurance premium (upfront MIP), and an ongoing

“annual” payment. FHA insurance can be used to cover a variety of loan terms.

MassHousing Mortgages

Unlike most states, Massachusetts has two HFAs, both offering income-restricted mortgage options.

MHP offers the ONE Mortgage, whereas its larger counterpart, MassHousing, offers a suite of

income-restricted mortgage options. The two organizations take very different approaches to

their loan products, reflecting two fundamentally different approaches to the mortgage market.

MassHousing is a wholesale lender and its lending is integrated into the secondary market.

Their loans are originated by partner lenders, which include banks, credit unions, and licensed

mortgage lenders. The loans are then serviced by MassHousing and sold to government-sponsored

enterprises on the secondary market.

In contrast, SoftSecond and ONE Mortgage loans are held in

portfolio by the originating partner lender.

In this paper, any reference to FHA loans is to the 30-year ﬁxed rate variety.

See MassHousing Agency Backgrounder for more detail about MassHousing’s wholesale lending model.

www.masshousing.com

195Cityscape

Building Wealth through Homeownership:

A Comparative Study of MHP’s ONE Mortgage Program and FHA

MassHousing products are limited to borrowers below 135-percent AMI, with some products

(such as the MassHousing 100 product) targeted at borrowers under 100-percent AMI.

MassHousing’s income guidelines are calculated in accordance with the Fannie Mae Selling Guide.

In contrast, MHP compliance is based on an estimate of the entire household’s income, including

nonborrowers, which means that a borrower who is over 100-percent AMI by MHP’s definition

might still qualify for a MassHousing product limited to 100-percent AMI.

Massachusetts’s large number of variously overlapping Housing Finance Administration-mortgage

products raises interesting questions about how borrowers decide among those products.

MassHousing offers a suite of purchase loan products (five as of the writing of this paper),

including a Federal Housing Administration option. These products are designed for various

income brackets and financing scenarios. Because the large number of variously overlapping HFA

products would introduce a large degree of complexity, we will limit our analysis to a comparison

between ONE Mortgage Loans and FHA loans. Of note, it is important to clarify that one cannot

distinguish if an FHA loan was originated through an HFA or by another private institution.

Prior Evaluations of Mortgage Subsidization

The ONE Mortgage is a combination of two types of public subsidy: (1) monthly payment

reduction in the form of an interest rate discount and interest rate subsidy, and (2) no mortgage

insurance (despite down payments as low as 3-percent) due to the presence of a publicly funded

loan loss reserve. Because borrowers receive all those benefits together, it is difficult to tease out the

effects of the individual subsidies.

Several studies have evaluated the impact of financing options on borrowers’ wealth creation,

although most focus on the relative benefits of participation in these programs in comparison

to renting. The most direct parallels to our study are the several studies carried out on the Self-

Help Ventures Community Advantage Program (Grinstein-Weiss et al., 2011; Stegman, Freeman,

and Paik, 2007). The Community Advantage Program portfolio contains a variety of Community

Reinvestment Act (CRA)-eligible mortgage products. Most loans in the portfolio are like the ONE

Mortgage in that they are 3-percent downpayment loans with no PMI. Program participants who

became homeowners earned $10,196 more in net assets than renters over a 3-year period.

International subsidy experiments also offer insight into the potential of mortgage subsidies to

increase homeownership rates. Like the Massachusetts Housing Partnership’s ONE Mortgage

Program, Portugal’s Credito Bonificado program from 1986 provided interest subsidies specifically

targeted at a low-income population. As in the Massachusetts Housing Partnership loans, Credito

Bonificado subsidies were provided through the terms of the loan itself rather than through the

tax code. The program increased borrowers’ probability of purchasing a home by between 2 and

4 percentage points. A 1-percent increase in interest rate corresponded to a decrease in borrowing

between 1.3 and 2.8 percent (Martins and Villanueva, 2005).

See MassHousing General Underwriting Guide for a full overview of MassHousing’s underwriting parameters.

196 Housing Tenure and Financial Security

Interrante and Schmiedl

The Effect of Interest Rate Discounts and Subsidies

High interest rates raise a mortgage’s monthly payments, increasing the likelihood of default.

Higher debt-to-income ratios are strongly associated with increased loan delinquency (Campbell

and Cocco, 2011). Evidence from the U.S. Home Affordable Modification Program demonstrates

the role interest rates play in mortgage sustainability. The program, a federal initiative introduced in

2009 to aid homeowners at risk of foreclosure, provided incentives to lenders to provide borrowers

with loan modifications that made the mortgages more affordable. Lenders were able to make

several adjustments to make the loan more affordable, including adjustments to the interest rate.

A 1-percent reduction in monthly payments was found to reduce the probability of redefaulting

by 0.23 percentage point, and a 1-percent reduction in interest rate reduced the probability of

redefault by 0.17 percentage point (Schmeiser and Gross, 2015).

Interest rates also impact demand for mortgage financing. Based on an analysis of high credit score

borrowers, Lo (2017) argues that a 25-basis-point decrease in mortgage rates for people with high-

FICO scores made those individuals 50 percent more likely to apply for a loan and also increased

the loan size by an average of $15,000.

The Effect of Low Downpayment Requirements

The size of a borrower’s required downpayment is perhaps the largest determinant of whether a

borrower will be able to purchase a home. Quercia, McCarthy, and Wachter (2002) find that loan-

to-value constraints—and therefore the difficulty of saving up a sufficient downpayment—are the

most important financial factors affecting a borrower’s likelihood of achieving homeownership.

The effect of higher downpayment requirements is an even larger effect than increases in monthly

payment due to higher interest rates. In their model, increasing the maximum LTV from 80 percent

to 97 percent led to a 3.25-percent increase in the probability of becoming a homeowner.

Although higher LTV loans have the advantage of increasing access to homeownership, they

perform worse than lower LTV counterparts. FHA borrowers are 2.5 times as likely to experience

a foreclosure event if the loan is originated at 95-percent LTV, compared with 80 percent (Lam,

Dunsky, and Kelly, 2013). LTV has effects on duration of tenancy as well; high-LTV borrowers take

longer to sell their properties as they hope to build up more equity prior to their sale (Genesove

and Mayer, 1994).

The Effect of Subsidizing MI

The advent of MI has had a major positive effect on lenders’ increasing willingness to extend

credit to high LTV borrowers (Goodman and Kaul, 2017). On the other hand, MI is a significant

cost to borrowers, potentially resulting in an effective cost increase of several hundred dollars per

month. The effect of MI payments on borrower outcomes is not widely isolated in the literature.

Because MI is generally applied as either an upfront cost that is factored into the loan amount or

as a monthly payment, its effects can be predicted to be similar to the effects of an interest rate

increase, however. The 2015 reduction of the FHA monthly insurance premium created a natural

experiment for researchers to observe the effect of mortgage insurance on loan demand. A 50-basis-

point reduction in mortgage insurance payments resulted in a 14-percent increase in home

purchase borrowing by otherwise qualified borrowers. (Bhutta, Laufer, and Ringo, 2017).

197Cityscape

Building Wealth through Homeownership:

A Comparative Study of MHP’s ONE Mortgage Program and FHA

Data and Methodology

Our analysis is based on a subset of 349 loans from MHP’s SoftSecond database. This database

contains over 21,000 loans and extends back to the creation of the SoftSecond Loan Program in

1991. The intent of our analysis is to quantify the difference between a fully subsidized MHP loan

and a comparable FHA loan over the lifetime of that loan.

Filtering the Data Set

The most relevant data on home price appreciation in MHP’s mortgage data set comes from

borrowers who received an MHP subsidy and subsequently exited the program via a home sale

event. When a borrower who received an MHP subsidy sells their property, they are required to

report their sales price to MHP. MHP uses that information to determine whether the borrower’s

home price appreciation was modest enough to trigger the program’s subsidy forgiveness

provisions. MHP does not collect home appreciation data for non-subsidized loans,

nor does MHP collect price appreciation data during a refinance, as these do not trigger any

subsidy forgiveness scenarios.

In order to take advantage of the extra data

associated with MHP subsidized property sales,

we first limited the data set to subsidized loans

that had already experienced a sale event.

Our database contained a sales price for each

of these loans, which we used to determine

the borrower’s actual amount of equity

accumulation. Note that selecting only loans

that received subsidy means that all loans in

our subset also meet the subsidy criteria, which

require the borrower’s household income to be

below 80-percent AMI and their Housing-to-

Income ratio to be above 28 percent. Because

only a few ONE Mortgage borrowers meeting

our filters have sold their properties to date, this

data set exclusively contains SoftSecond loans.

Next, we filtered out any multifamily (two- or three-unit) properties, which are not directly

comparable with the single-family and condominium units that composed the bulk of our subset.

Finally, we excluded properties with any form of affordability deed restriction. Affordability deed

restrictions in Massachusetts allow low- and moderate-income borrowers to purchase properties

at below-market prices but require that the borrower also sell their property at a below-market

price. This creates an artificial limit on the amount of appreciation the homebuyer can experience,

meaning they are not comparable with unrestricted market units. Our database did not capture

deed restrictions and certain other loan characteristics until 2004, so all loans prior to this cutoff

have been excluded.

Exhibit 4

Loan Subset Criteria

• Loan closed between 1/1/2004 and 5/31/2013.

• The subject property has been sold as of

August 2018.

• The subject property is a condominium or

single-family home.

• The subject property does not have any deed

restriction limiting the price appreciation of the

property.

• Loan received MHP interest subsidy and meets

the following subsidy award criteria:

o Household Income below 80 percent AMI.

o Unsubsidized Housing to Income ratio is

greater than 28 percent.

AMI = area median income.

MHP = Massachusetts Housing Partnership.

198 Housing Tenure and Financial Security

Interrante and Schmiedl

Constructing Our Comparisons

Our analysis compares actual borrower outcomes in the SoftSecond Loan Program to hypothetical

outcomes for a comparable FHA mortgage. To compare these programs, we created three data sets

each containing 349 loans:

1. A set of real SoftSecond loans drawn from MHP’s loan database.

2. A set of simulated ONE Mortgage loans. Each loan in this set is based on a loan in the

SoftSecond data set. For each loan, we hold constant the total loan amount, the interest

rate, and the full value subsidy amount. The monthly payments are recalculated to reflect

the differences in amortization between the two-mortgage structure of the SoftSecond

Program and the ONE Mortgage Program. We also alter the subsidy payment schedule to

reflect the new shorter schedule of the ONE Mortgage Program.

3. A set of simulated FHA loans. Like the simulated ONE Mortgage loans, each simulated

FHA loan is based on a loan from the SoftSecond set. For the FHA loans, we keep only

the loan amounts constant. We modify the interest rate to match FHA’s historical average at

the time of origination using historical interest rates gathered from FHA’s Announcements

Archives (FHA 2019). Upfront private-mortgage-insurance payments are included in

the loan amount, which is a common practice in FHA loans. In addition, these loans are

assumed to have mortgage insurance until reaching 78-percent LTV, a common feature of

FHA loans prior to June 2013.

Comparison Metrics

Our evaluation of the differences between these programs was based on three dimensions of wealth

accumulation: (1) total monthly payment amount, (2) equity accumulation, and (3) net financial

outcomes.

To give a baseline reflection of the time-value of money, all savings have been inflation adjusted

to 2018 dollars. This makes our analysis sensitive not only to the differences in monthly payment

amount, but also the timing of the monthly payments. This adjustment is particularly important

when considering the benefits of the interest subsidy. MHP structures its interest subsidy on a

declining schedule to deliver the largest impact early in the loan’s amortization. (See Exhibit 3.)

Besides adjusting for inflation, this analysis does not make additional assumptions about how

borrowers might use the savings derived from lower monthly payments (for example, by paying

down credit cards or investing in a savings instrument), although doing so would give additional

weight to savings rendered early in the life of the loan.

ONE Mortgage subsidy schedule calculation (where “Full Value Subsidy” is the total amount of subsidy funds

to be disbursed):

Year Annual Subsidy Amount

Years 1–4 Full Value Subsidy/5.5

Year 5 Full Value Subsidy/7.33

Year 6 Full Value Subsidy/11

Year 7 Full Value Subsidy/22

199Cityscape

Building Wealth through Homeownership:

A Comparative Study of MHP’s ONE Mortgage Program and FHA

Equity accumulation in this article is a measure of the total proceeds to the borrower when they

sell their property. It is measured by subtracting the principal balance remaining on the loan at the

time of sale from the sale price of the home. In the case of ONE Mortgage and SoftSecond loans,

the subsidy repayment is also subtracted

Equity Accumulation = Sale Price - Principal Balance - Subsidy Repayment

Total monthly payments were calculated by adjusting the monthly payments for inflation and then

summing the borrower’s monthly payments, from the time they closed on their loan until sale:

Total Monthly Payments =

∑

Monthly Payments

Finally, we created a Net Financial Outcome measure that captures the overall financial benefit to

borrowers taking into consideration both equity accumulation and total monthly payments:

Net Financial Outcome = Equity Accumulation - Total Monthly Payments

Results

Equity Accumulation

In general, FHA loans offered borrowers slightly higher equity accumulation than ONE Mortgage.

As seen in exhibit 5, on average, ONE Mortgage borrowers accumulated $67,534 in equity

accumulation compared with $70,806 for FHA borrowers. Exhibit 6 shows loan level differences.

The median loan built 4.4-percent less equity as a ONE loan than it did as an FHA. Approximately

84 percent of the loans modeled would have had higher equity accumulation under an FHA loan

than a ONE Mortgage loan. Based on a two-tailed, two-sample t-test, using a 95 percent confidence

level as the threshold, the difference in the means of the two groups is not statistically significant.

In practice, MHP does grant borrowers partial subsidy forgiveness if their property has had little appreciation. The

MHP Subsidy Note allows borrowers to repay the lesser of either: (a) the amount of subsidy they received or (b)

20 percent of net appreciation. For simplicity, we assume that all borrowers repay the amount of subsidy they have

received.

200 Housing Tenure and Financial Security

Interrante and Schmiedl

Exhibit 5

Average Equity Built

$80,000

$10,000

$20,000

$30,000

$40,000

$50,000

$60,000

$70,000

All

ONE

FHA

(N=349)

$67,533.73

$70,806.25

Exhibit 6

Total Equity at Loan Repayment - ONE Versus FHA (all equity in 2018 dollars)

FHA = Federal Housing Administration.

Total Equity at Loan Repayment (2018 Dollars)

Loans Ordered by Total Equity Under FHA

201Cityscape

Building Wealth through Homeownership:

A Comparative Study of MHP’s ONE Mortgage Program and FHA

Although not statistically significant, the fact that FHA builds equity slightly faster than ONE is

surprising, given that ONE Mortgages feature a discounted interest rate (ONE Mortgage interest

rates are capped at 30 basis points below the Freddie Mac Primary Mortgage Market Survey).

The difference in equity building is largely due to the required repayment of the MHP Subsidy

Mortgage, which every borrower in this set received. Borrowers in this set repaid an average of

$7,698 in subsidy. In many cases, this repayment was enough to outweigh the equity-building

benefits associated with the ONE Mortgage’s lower interest rate. The effect of the subsidy repayment

was greatest for borrowers with relatively small first mortgages. In the program as a whole, not all

borrowers receive the subsidy and some that do may receive a partial subsidy forgiveness, so we

expect that equity accumulation was somewhat more favorable towards borrowers in the program

overall.

Monthly Payments

While FHA might result in higher equity realized at repayment, the ONE Program compared very

favorably to an equivalent FHA loan in terms of the borrower’s total monthly payments over the life

of the loan. Both the mean and median life of loans within the data set are 6.43 years. As exhibit

7 shows, loans modeled as FHA mortgages had payments about 33 percent higher than when

modeled as ONE mortgages. Not only did the overall averages favor the ONE Mortgage Program

over FHA, but every single loan in the data set would have lower total payments under ONE than

under a comparable FHA mortgage.

The distribution of total payments (exhibit 7) shows the stark difference between the products.

Exhibit 7

Average of Total Payments Over the Life of the Loan

$80,000

$90,000

$100,000

$10,000

$20,000

$30,000

$40,000

$50,000

$60,000

$70,000

All

ONE

FHA

(N=349)

$69,596.35

$92,413.66

FHA = Federal Housing Administration.

202 Housing Tenure and Financial Security

Interrante and Schmiedl

Exhibit 8

Total Payments Over Life of Loan - ONE v. FHA

FHA = Federal Housing Administration.

Net Financial Outcome

In terms of net financial outcome, every loan in our subset performed better as a hypothetical ONE

Mortgage compared with a hypothetical FHA loan. Exhibit 9 shows the relative net cost or gain as a

percentage of purchase price over the life of the loan (amount of equity gained less the cumulative

monthly payment). On average, this estimated net financial difference was the equivalent of 10.7

percent of the original purchase price.

Loans Ordered by Total FHA Payments

Total Payments Over Life of Loan (2018 Dollars)

203Cityscape

Building Wealth through Homeownership:

A Comparative Study of MHP’s ONE Mortgage Program and FHA

Exhibit 9

Net Difference in Financial Outcome (as a Percentage of Original Purchase Price) When ONE

Mortgage is Modeled Against FHA

FHA = Federal Housing Administration.

It is important to note that although all loans would have experienced better financial outcomes

under ONE than FHA, not all net outcomes are net positive. Equity gained upon the sale of a

home is naturally offset by payments over the life of the loan. In most cases, the net payments on

a mortgage exceed price appreciation and amortization. In our data set, 56 percent of ONE loans

would have had a net cost over the life of the loan, and 70 percent of FHA loans would have a net

cost. Descriptive statistics for our analysis are shown in exhibit 10.

Exhibit 10

Descriptive Statistics (N=349)

Mean Median Minimum Maximum

Monthly

Payments

(Cumulative)

SoftSecond $66,531.33 $61,742.94 $2,625.77 $185,836.12

ONE Mortgage $69,596.35 $64,386.43 $2,738.41 $193,360.58

FHA $92,413.66 $88,766.47 $3,942.46 $235,122.52

Equity

Accumulation

SoftSecond $64,448.00 $53,314.04 $(35,546.51) $301,438.05

ONE Mortgage $67,533.73 $56,527.68 $(32,525.56) $305,119.12

FHA $70,806.25 $60,376.19 $(27,828.98) $307,749.09

Net Financial

Outcome

SoftSecond $(2,083.33) $(9,694.75) $(132,823.32) $239,265.21

ONE Mortgage $(2,062.62) $(9,138.39) $(131,369.54) $239,683.30

FHA $(21,607.40) $(27,188.15) $(163,504.38) $217,274.47

FHA = Federal Housing Administration.

Net Financial Difference Between ONE and FHA

as a % of Purchase Price

Loans Ordered by Net Difference as a Percentage of Purchase Price

204 Housing Tenure and Financial Security

Interrante and Schmiedl

Delinquency

Massachusetts Housing Partnership Loans have lower delinquency and foreclosure rates than both

FHA loans and the average loan originated in the state (exhibit 11).

Exhibit 11

Massachusetts Delinquency Rates Massachusetts Foreclosure Rates

FHA = Federal Housing Administration.

MHP = Massachusetts Housing Partnership.

Source: MBS National Delinquency Surveys, 2009-2019; MHP eS2 database

The ONE Mortgage program was introduced in 2014 and is still relatively new. It has only existed

during times of economic expansion and its performance has not been tested in a crisis. Exhibit 12

shows MHP delinquency rates over time for three selected vintages (2005, 2010, and 2015). The

2005 and 2010 vintages include only SoftSecond loans, whereas the 2015 vintage contains only

ONE Mortgage loans. The stark contrast between the 2005 and 2010 delinquency data shows the

unpredictable effect a recession can have on a seemingly low-delinquency Housing Finance Agency

product. Given the ONE Mortgage program’s general similarity to the SoftSecond Program, we

expect that performance would be generally comparable with the SoftSecond Program’s.

MHP’s delinquency rates are a measure of all currently delinquent loans (>30 days delinquent, but not reported as

in foreclosure) divided by the number of active loans. Our foreclosure rate reports the percentage of loans that MHP

lenders, who service the loans, report as being in the process of foreclosure.

205Cityscape

Building Wealth through Homeownership:

A Comparative Study of MHP’s ONE Mortgage Program and FHA

Exhibit 12

MHP Delinquency Rates by Vintage MHP Foreclosure Rates by Vintage

MHP = Massachusetts Housing Partnership.

Source: MBS National Delinquency Surveys, 2009-2019; MHP eS2 database

The ONE Mortgage Program and SoftSecond Program both feature built-in mechanisms that

address delinquency and foreclosure. In addition to a lower monthly payment, ONE Mortgage

borrowers receive ongoing delinquency counseling. Any time a borrower becomes 30 days

delinquent, a counselor working for one of MHP’s partner counseling agencies will offer free

services. These independent nonprofit agencies provide borrowers a trusted third party who

will help them resolve their delinquency. Depending on the borrower’s desire to engage with the

counselor’s outreach efforts, these counseling sessions can be extensive. They may span dozens

of interactions with borrowers over the course of months or even years. The assistance provided

ranges from simple financial advice to more involved interactions, such as loan modification

mediation between the borrower and lender.

How Many Federal Housing Administration Borrowers Could Have Qualiﬁed for the

ONE Mortgage?

Due to the limitations of publicly available FHA data, it is difficult to create a satisfying estimate of

the number of FHA borrowers who might have qualified for the ONE Mortgage Program. Home

Mortgage Disclosure Act (HMDA) data lacks critical pieces of ONE Mortgage compliance data

including FICO score, first-time homebuyer status, and borrower debt. Ginnie Mae mortgage-

backed securities data provides another potential route, but it also lacks critical information.

Although it does record borrower credit scores and debt to incomes, it lacks both borrower and

household incomes and debt amount. Exhibit 13 illustrates the shortcomings of each data set.

206 Housing Tenure and Financial Security

Interrante and Schmiedl

Exhibit 13

Availability of MHP Borrower Qualications in Publicly Available Data Sets

MHP Borrower Qualiﬁcations

First-Time Homebuyer?

Household Income

(Including Nonborrowers)

Qualifying Income

Credit Score

Borrower Assets

Loan Size

HTI

DTI

Upper Bound: FHA

Loans That Could Have

Been ONE (Annual)

Upper Bound: FHA

Loans That Could Have

Been ONE (Annual)

Ginnie

Mae MBS

  

1,516 $382M

HMDA

  

1,555 $382M

DTI = Debt to Income. FHA = Federal Housing Administration. HMDA = Home Mortgage Disclosure Act. HTI = Housing to Income.

MBS = Mortgage Backed Securities. MHP = Massachusetts Housing Partnership.

Sources: HMDA LAR 2017; Ginnie Mae MBS Monthly Loan Level data

Filtering for FHA loans that met all HMDA-provided ONE Mortgage qualifying criteria in 2017

yields a subset of 1,555 loans in the total amount of $382 million. A similar filtering of this data set

for the Ginnie Mae MBS data set indicates 1,516 eligible loans, a total of $382 million in lending.

These estimates, neither of which applies the full panel of ONE Mortgage qualifications, should be

taken as reasonable upper bounds of the number of FHA loans originated annually that might have

qualified as ONE Mortgages.

Although they come from two different data sets filtered on different qualifying variables, the

estimates arrived at a very similar numbers of loans and nearly identical gross dollar amounts.

It is certainly tempting to read more into that match than it merits. Note that this estimate is not

robust enough to accurately predict the actual size of the overlapping group of borrowers. Rather,

we intend it simply to provide a sense of scale. Future research could refine the accuracy of our

estimates using data from proprietary FHA loan databases.

Discussion

We live in an era in which LMI households are squeezed for every last dollar. In 2017, 59 percent

of households could not cover a $400 expense using cash or its equivalent (Board of Governors

of the Federal Reserve Board, 2018). Over half of young adults who went to college in 2017 took

on some personal debt, while one-fifth of them were behind on their payments (Federal Reserve

Board, 2018). This student loan debt, increasing along with rising home prices in urban markets,

makes it more difficult to afford the monthly costs of homeownership. To combat these statistics

and still encourage homeownership, it is increasingly important to focus on the development of

financial products that lower costs and increase housing stability.

Historically, homeownership has been the single biggest driver of household wealth in the United

States. As Goodman and Mayer (2018) note, although homeownership generally offers households

207Cityscape

Building Wealth through Homeownership:

A Comparative Study of MHP’s ONE Mortgage Program and FHA

superior wealth building when compared with renting, the advantage of owning a home is highly

dependent on assumptions about home price appreciation and the relative costs of homeownership

and renting. The terms of the mortgage and the types of subsidization used play a major role in

evaluating the overall benefit of homeownership.

Since each new homebuyer’s situation is different, it is difficult to broadly claim that one approach

is better than another. Borrower preferences can vary on several key dimensions. While one person

might place a higher value on monthly savings, another might place a higher value on equity

accumulation. Some might be looking at the home as a long-term family asset, while others are

looking at the home as a short-term investment.

The difficult tradeoffs involved in borrower preferences are reflected in our analysis: the ONE

Mortgage program slightly underperformed FHA loans on measures of equity building, largely

due to the sample selection, which limited our analysis to subsidized ONE Mortgage loans.

The mechanics of the subsidy are to diminish equity accumulation in exchange for significantly

lower monthly payments. Therefore, the subsidy mortgage is a major part of the ONE Mortgage’s

advantage over FHA in monthly payments but causes the borrower to lose out on overall equity.

How should borrowers think about this tradeoff?

The MHP subsidy mortgage can be seen as a deferred amortization mechanism that transfers funds

from the proceeds of the borrower’s eventual home sale to a buydown of their monthly payments.

It is disbursed to the borrower on a monthly basis over the course of the first 7 years until the full

value of the loan has been paid out. Because it is a zero-interest loan, the subsidy is repaid at time

of property sale. Although this design means the subsidy is a net-zero prospect to the borrower (in

nominal terms), it subtly alters the economics of a mortgage loan from the borrower’s perspective.

Normally, the proceeds of equity accumulation can only be accessed during a refinance or at the

time of sale. This means that for the most part, these funds remain inaccessible to the borrower,

even though their preference may be to access them earlier in the life of the loan. The MHP subsidy

allows the borrower to do so with no fees or penalties.

Policymakers should be sensitive to the fact that there is more to wealth building than equity

accumulation alone; homeowners are interested in monthly savings as well as overall equity

accumulation. Subsidization methods that allow for this kind of liquidity earlier in the loan

address these borrower preferences. More affordable payments keep more money in the pockets

of LMI homeowners. Affordability has the added benefit of stability, making it easier for low-

income owners to maintain payments and avoid default and foreclosure. The FHA delinquency

rate is consistently two to three times higher than MHP, whereas its foreclosure rate is about

twice that of MHP. Contrary to expectations about high loan-to-value loans, MHP actually has

lower delinquency and foreclosure rates than other overall rates for mortgage originations in

Massachusetts, despite targeting LMI households.

Throughout the programs’ histories, both SoftSecond and ONE Mortgage have had lower

delinquency and foreclosure rates than the average Massachusetts mortgage loan. These loan

performance figures benefit both borrowers and originating lenders. While borrowers have

safer, more sustainable loans, lenders can produce more loans because of their relatively

208 Housing Tenure and Financial Security

Interrante and Schmiedl

strong performance. Most of our focus has been on affordability and wealth building, but the

sustainability of homeownership is another important factor to consider when crafting housing

policy at all levels. In addition to the losses foreclosures entail for the borrower, a single foreclosure

is associated with an average loss to the loan holder of over $58,000. Foreclosures also cost cities

and neighborhoods, to the tune of $27,000 and $10,000 respectively (Immergluck and Smith,

2006).

Regardless of the tradeoffs between monthly payments and equity built into the ONE Mortgage,

the overall takeaway of our study is clear: borrowers have better financial outcomes using a

ONE Mortgage compared with an FHA loan. The average net financial difference between ONE

Mortgages and FHA loans in our data set was the equivalent of 10.7 percent of the original

purchase price, and every borrower was better off in terms of net financial outcome when modeled

as a ONE Mortgage rather than an FHA loan.

HMDA data reveals, however, that FHA lending accounts for 29 percent of low- and moderate-

income home purchase lending in the state compared with the ONE Mortgage’s 4 percent

(Campen, 2018; Massachusetts Housing Partnership, 2016). In the post-crisis era, FHA

emerged as the next-best option for lenders who could no longer offer high cost loans. This was

consistent with its original intent as the loan program of last resort (Immergluck, 2011). Why are

Massachusetts borrowers using a more expensive last resort option when a more affordable State-