Rise of the Assumable Mortgage

In the past, the mortgage insurance on a 30-year fixed rate mortgage backed by the FHA would phase out after the mortgagee had reduced their loan-to-value ratio to 78% through principle reduction.  For a mortgagee with a 3.5% down payment [1] and a $200,000[2] 30-year fixed rate mortgage at a rate of 4%, the LTV would fall to 78% after a little more than 9 years.  In addition, there would be an annual mortgage insurance premium rate of 1.35%, which is $2,700 extra paid annually or $225 per month.

Over the last three decades, the average 30-year fixed mortgage rate fell steadily from a peak of 16.63% in 1981 to 3.66% in 2012.  As a result, fewer mortgagees waited to pay down their mortgage to release the mortgage insurance.  Rather, as prices appreciated, they took advantage and refinanced into a lower rate mortgage without the mortgage insurance.  The falling rates led to faster pre-payment of loans.  Furthermore, the average tenure on a home hovered at just six years for nearly two decades, rising in recent years in part due to the large number of underwater owners, weak job prospects, tight credit standards, and price uncertainty.

However, mortgage rates are expected to rise over the next decade thanks to economic pressures, with the cessation of MBS and Treasury purchases by the Federal Reserve, and due to the large amount of government debt outstanding.  As long term mortgage rates rise, holders of FHA mortgages will find that they hold a unique and valuable asset: an assumable mortgage.   Mortgages financed by the FHA are assumable by another homebuyer, so long as the second person can qualify for FHA financing.  Thus, a buyer in the future who is facing a market rate of 5.0% on a mortgage could assume an FHA mortgage of 3.5% for the bulk of the financing and pay cash for the rest or use a second mortgage.[3] The number of mortgages issued by the FHA ballooned in recent years and the FHA estimates that a higher than average share its 30-year fixed rate mortgages will be held for the full 30-year term.  What’s more, this trend is forecast to rise again with originations after 2013, precisely when rates are forecast to rise.

What Does this Mean to the Buyer?

Assuming a lower rate loan reduces the cost of borrowing in an environment of rising interest rates.  Assuming that home values rise at the historical 3% average, that $207,250 home would be worth roughly $247,500 after six years when the owner decides to sell.  The next buyer could assume the remaining balance of roughly $172,000 and a second mortgage for the $83,000 balance at a hypothetical retail rate of 5.0%.  The difference in the monthly payment for the new homeowner is nearly $72 dollars with the assumed mortgage versus if the entire sum were financed at the retail rate.  However, if rates were to rise further, the difference is even larger reaching more than $900 a month if rates rise to an unlikely 12%.

How Does the Original Owner Benefit from this Deal?

The homeowner who sells their home has an asset in the assumable loan.  In today’s market, you can pay your mortgage broker roughly 1% of the value of the loan to buy down the rate by an eighth of a percentage point (e.g. from 4% to 3.875%)[4].  If that pricing held in the future, the original mortgagee in the above example would have an asset worth $13,747 ([5% - 4%]/0.125% X 1% X $172,000) and the value would rise if the difference between the rate on their mortgage and the market rate increased.  Furthermore, if the FHA did not continue to collect mortgage insurance, the benefit would be even greater.

For the future home seller, these extra monies could be used as equity for trade-up or to buy an assumable mortgage on another home.  It could also be used to entice reticent buyers in a slow market or to offset costs of selling.

But How Does the Assumability of the Mortgage Help the FHA?

It doesn’t.  In fact, it actually creates a problem for the FHA as it raises the risk to the agency when the second mortgagee takes over the loan and the mortgage insurance is released when the LTV hits 78%; just four years after the second transaction.  The FHA would take on new credit risks without offsetting compensation.  The second owner could then resell the loan to a third person who pays no mortgage insurance at all to the FHA, but still poses a risk to the FHA.  Thus, the FHA’s new permanent mortgage insurance policy makes sense in this respect.  Given the historic tenure term of six to seven years, an FHA 30-year mortgage could be used by four or even five different homeowners, each with different risk characteristics and propensities to default and each facing different economic stresses.

The FHA also benefits from the permanent mortgage insurance policy as it helps to boost the FHA’s earning on each loan.  This would support the company’s current book of business in the present, lifting it closer to its statutory reserve requirements and back into the black.  However, the fund could be hurt by the large volume of mortgages issued in recent years that may be assumed by new borrowers in the future without compensating mortgage insurance.  The FHA benefited from tighter underwriting and better performance from 2007 to present, but mortgagees who assume these loans in the future might not have the same low risk characteristics.

Selection bias might also benefit the FHA.  A buyer who is aware of the benefits of an assumable loan may be more financially savvy.  Furthermore, the lower monthly payment would soften the blow of a loss of employment or other financial hardship that could trigger a default.

Finally, the overall housing market will benefit from assumability, though the mortgage industry might lose on compensation.  Borrowers will have a means to mitigate increases in mortgage rates that are anticipated in the coming years, softening any erosion of affordability or flat income growth.  Furthermore, borrowers served by the FHA, the credit and savings impaired, are more likely to be impacted by marginal changes in affordability.

Someone Must Get Hurt

The one group that loses from permanent mortgage insurance is those borrowers who put down less than 10% and hold the loan for 30 years or those who hold for more than 10 years and sell but don’t allow their loan to be assumed.  At the extreme, those borrowers would pay an extra $56,700 in mortgage insurance over the life of the 30-year loan; monies that could be used to pay down principle, student loans or some other more productive use.  The permanent mortgage insurance is an onerous change for this group.  For this reason, the FHA might re-think the mechanics of its mortgage insurance policy so that it releases after 78%, but kicks in again if the loan is assumed.  In this way the program is protected from the risks of assumption without overcharging the consumer.

The FHA faces many challenges going forward.  Rising mortgage rates and a changing regulatory landscape may alter the typical borrower’s profile as well as the agency’s market share.  Mortgage assumability will also impact the type of borrower the FHA receives, so it is with keen foresight that the agency changed its policy to shore up its books against future risks.  Furthermore, the consumer gains an asset, which will help keep the market fluid in an environment of rising mortgage rates.  However, many households will over pay with permanent mortgage insurance.  More should be done to protect consumers who are likely to hold these loans to term.

[1] Under the new FHA policy, the FHA releases the MI requirement after 11 years if the borrower puts down 10%.  Why is the mortgage insurance not for the life of these loans? After, 11 years, the borrower’s LTV falls to roughly 68%.  However, a borrower who puts down 10% in the FHA program is likely credit impaired, unlikely to shift to private financing, and more likely a trade-up buyer who is less likely to re-sell (70% of first time buyers put down less than 10%).
[2] The initial home value is $207,250.  I rounded for simplicity in this example.
[3] Second mortgage may be more difficult to use in the future under the final Basel III and Qualified Mortgage Rule, but these regulations are as yet unfinished and financial innovation could step in to fill the gap.
[4] The current rate charged by brokers for the buy-down changes over time, but the example is just an illustration of how this benefit can and is quantified on a regular basis.

• The New York Times reviewed a new research paper by authors from Dartmouth and the University of Warwick in England that implies a rise in the homeownership rate is bad for the economy because fast rising homeownership was associated with much higher unemployment rate. The paper gave examples of several southern states with high unemployment rates as evidence.
• Economic vitality should be assessed not by the unemployment rate but by job growth. A state may have a very low unemployment rate yet have no job growth. What is needed for economic and income growth, in short an improving standard of living, is job growth – the variable this critique focuses on.
• The paper cites California and Wisconsin as good states with no increase in homeownership and a low state unemployment rate. Georgia and South Carolina were examples used as bad states with rising homeownership and a high state unemployment rate.
• The paper used 2010 unemployment rates to compare. But 2010 may not be the best choice of year, as it marks the deep depths of the Great Recession. A dynamic economy like that of the United States constantly undergoes a “creative destruction” that is critical to long-run faster economic growth and states that tend to grow fast could also be vulnerable to bigger short-term setbacks.
• If using the year 2007, before the economic downturn, as the end point, and using payroll growth (not the unemployment rate) for comparisons, then Georgia and South Carolina are shown to easily outperform California and Wisconsin (see table below).
• The Civil Rights Act passed in 1964. The southern states’ job market conditions must be taken into consideration with this monumental legislation. Therefore, the start date for our job growth is 1964 (see table below).
• Our quick analysis shows a clear superior economic performance by Georgia and South Carolina over California and Wisconsin. A choice of different start and end years may have yielded different conclusions. At the same time, we should not automatically assume other time periods used, such as the ones implied in the NYT article, as infallible.
• Interest rates and mortgages rates vary with the business cycle. Historically, housing has led the United States out of recessions because of the low borrowing costs at the end of a recession. As borrowing costs rise, business activity is choked off as is housing consumption, the former having a more direct link to employment. Thus, housing consumption could be acting as an instrument for borrowing costs and the fluctuations in employment could be ascribed to it.
• The authors do not explore variation within the labor markets across states and time. For instance, certain regions experienced change in the industries driving their economies as the South experienced a sharp increase in skilled manufacturing and financial services since 1980. The regional observation might point to differences in patterns of economic development.
• The authors note that their working paper “makes a simple statistical contribution and discusses possible mechanisms. The detailed nature of any housing-labor externality remains poorly understood.” Furthermore, the structure of their analysis, “is potentially a weakness and means that some underlying omitted variable, or causal force, might be responsible for the link between,” the homeownership and subsequent changes in unemployment.
• Finally, it is almost certain that by the time of next Congressional Reapportionment Georgia and South Carolina will have gained a seat or two while California and Wisconsin will have lost a seat or two. That is a direct consequence of the state’s ability to draw new people because of better economic conditions. Rising homeownership rates in southern states look to have been a great benefit to the states’ economies, exactly contrary to the assertion made by New York Times.

• In spite of reduced affordability from last month and last year’s near-record levels, the median income U.S. family earns almost double what is needed to purchase the median priced home, so affordability remains high.
• Housing affordability is down for the month of March in the U.S. as rising incomes were not enough to completely offset higher mortgage rates and home prices from February to March.
• From one year ago, affordability is down as lower mortgage rates and higher incomes have not completely offset home price gains.
• By region, affordability is down from one month ago in all regions except the Midwest, where there was no change. From one year ago, affordability is higher in the Northeast and Midwest and lower in the South and West as huge home price gains overwhelmed slightly lower mortgage rates.
• Check out the full data release here.
• The Housing Affordability Index calculation assumes a 20 percent down payment and a 25 percent qualifying ratio (principle and interest payment to income). See further details on the methodology and assumptions behind the calculation here.