The Trump administration has tried to seize the affordability mantle by proposing a move to 50-year mortgages. Unfortunately, the math doesn’t add up: a 50-year mortgage is a pretty bad idea.
The United States is unique globally in that our dominant mortgage product is the 30-year, fixed-rate, fully-prepayable, fully-amortized mortgage. The 30-year fixed is the American mortgage, and it is a wonderful financial product. It’s also one that only exists because of substantial government involvement in the market. But shifting it out to a 50-undermines the benefits of the product.
A brief history: prior to the New Deal, the standard US mortgage product was a short-term, fixed-rate, non-amortized mortgage—a bullet loan. The pre-New Deal bullet loan had to be refinanced frequently, which was disastrous if markets were frozen (as in 1929) or if property values dropped and the homeowner had no equity or if the homeowner’s financial situation deteriorated. The problems with this sort of loan were apparent even to the Hoover Administration, and after a whole bunch of government interventions over the 1930s and 40s (creation of Federal Home Loan Banks, Federal Housing Administration, Fannie Mae, and the Veterans Administration), the United States’ mortgage market shifted to a 20-year, amortized fixed-rate mortgage and ultimately the 30-year fixed. (If you’re looking for a good book on the topic, I can recommend one….)
The 30-year fixed is one of the most consumer-friendly financial products around (the other contender being federal student loans!). Here’s what makes it great:
- It is fully-amortized. That means that the borrower is constantly building up equity in the property. That’s important both because it means that the house is an investment (and one of the few highly leveraged investments available to a consumer) and because it protects the homeowner in case the homeowner needs to move and property values have declined.
- It is fixed-rate. That means that it provides a reliable centerpiece for a family budget. The mortgage is likely the family’s largest regular expense and they can plan around it. It also means that the homeowner is shielded from interest rate risk. There’s no good hedge for consumers to deal with rate risk, so this is really important. Adjustable-rate mortgages might look cheaper, but on an option-adjusted basis they are not. They really are not a good product for consumers.
- It is fully prepayable. This means that the borrower benefits from huge optionality. If rates go down, the homeowner can refinance (having build up some equity because of the amortization). Likewise, if the homeowner chooses to sell the house, the mortgage can always be paid off (and contractual due-on-sale clauses mean that it must be).
- It is for 30 years. There’s nothing inherently magical about 30-years, but it does more or less match with the time period from first home purchase to retirement (at least traditionally): buy your home at ago 30 or 35 and pay off the mortgage by the time you’ve retired, at which point you’ve built up a nice next egg from the property appreciation. The long term helps lower monthly payments, which enhances affordability, but it is still bounded by having mortgage payments conclude around the time that the borrower’s working income ends. Notice that this doesn’t work so well today, however, when the average age of first home purchase is 40.
A. The Rate on the 50-Year Mortgage Will be Higher than on the 30-Year Mortgage
The thinking (if you can call it that) behind the 50-year mortgage proposal is that by extending the amortization period of the loan, the monthly payment is reduced. That’s true if all else is equal, meaning that there is no change in the principal or interest rate. For illustrative purposes, let use the latest National Mortgage Database numbers.
The average size of a new mortgage origination these days is $366,000, and the average interest rate is 6.4%. On a 30-year mortgage, the monthly payment would be $2,289.35, whereas the same mortgage over 50-years would have a monthly payment of $2,035.69. Saving families $253/month would be a nifty trick, right?
The problem is it doesn’t work. The interest rate would never be the same on otherwise identical 30-year and 50-year mortgages. That’s because lenders have to price for the greater optionality on the 50-year mortgage. How much more would they charge?We can get a ballpark of what the difference might be by comparing the rates on 15- and 30-year mortgages from the same lender. Here’s what Bank of America offers. It lists a 15 year at 5.788% APR and a 30 year at 6.467% APR.1 That’s 67.7 basis points for 15 years of optionality. Pro-rated to 20 years that’s 90.3 basis points for the extra optionality on a 50-year mortgage over the 30-year.Similarly, PennFed Credit Union lists 5.409% for a 15-year mortgage and 6.174% for a 30-year loan. So that’s an extra 76.5 basis points for 15 years of additional optionality. Pro-rated to 20 years, that 102 basis points for the extra optionality on a 50-year mortgage over the 30-year.
B. The 50-Year Mortgage Might Not Even Lower Monthly Payments
If we compare a 30-year mortgage of $366,000 at 6.4% with a 50-year mortgage of $366,000 at either 7.33% or 7.42%, we see that the 50-year mortgage is less affordable even in terms of monthly payment. The monthly payment on the 30-year would be $2,289.35, while the monthly payment on the 50-year would be either $2,295.07 or $2,315.45.
Now you might quibble with the optionality premium for the 50-year mortgage. But to even get the monthly payments the same, the optionality premium would have to be a mere 90 basis points. And even if it were as small at 75 basis points (which would be hard to imagine), it really doesn’t make much difference, as the monthly payment would only drop to $2,224.30.
C. The 50-Year Mortgage Results in Much Slower Equity Build and Massively Higher Interest Expense
You might say that $65/month is nothing to sniff at, but let’s recall that the change in monthly payment isn’t free. It comes at the expense of slower equity buildup and much higher total interest expense to maturity. How much?
On the 30-year mortgage of $366,000 at 6.4%, the homeowner builds up about $57,000 in equity over 10 years, assuming no appreciation. Total life-of-mortgage interest would be $458,166.59.
In contrast on the 50-year mortgage of $366,000 at 7.42%, the homeowner builds up only about $11,000 in equity over 10 year, assuming no appreciation. Total life-of-mortgage interest would be $1,023,271.66.
What’s worse, is that a 50-year mortgage will last well past the earning lifetime of most individuals. If you get a mortgage at age 40, you’d be in it until age 90, which is beyond the average American’s life expectancy. That’s like getting a 10-year car loan on a car with a remaining useful life of 3 years. Come to think of it, the that’s exactly what’s going on here. The 50-year mortgage proposal is an attempt to bring used car salesmanship to the home mortgage market. It’s all about “how much can you afford per month,” never minding that it’s a terrible deal overall, even for affordability.
D. If the 50-Year Mortgage Lowers the Cost of Credit, It Will Raise Home Prices, Undercutting Any Affordability Gains
Suppose that the 50-year mortgage does in fact lower monthly payments for borrowers and that they opt for it because of that. What’s going to happen? Home prices will be bid up. A borrower who could afford a $400,000 loan with the 30-year fixed’s monthly payment can now afford, say, a $450,000 loan with the smaller monthly payment of a 50-year fixed. The result? The borrower will borrow $450,000, as will all the competing borrowers, and they will bid up home prices. The higher home prices will in turn undercut any affordability gains from the 50-year fixed. Someone just didn’t think this stuff through.
There is really nothing to like about a 50-year mortgage relative to the 30-year fixed.
Use the APR, rather than the interest rates here because there’s an unequal amount of buyer’s points for the two loans. Those are finance charges, so they get included in the APR.
↩︎
Comments
3 responses to “50-Year Mortgages? The Numbers Don’t Add Up”
You beat me to this, Adam. My days are taken up teaching Consumer Finance out of your textbook. You are absolutely right, but I will take up your opportunity to quibble with the math just a little bit. The cost of the additional “optionality” is not a linear function. The difference between the interest rate on a 30-year and a 50-year is not likely to be huge. If I was not running to class, I would look to see if there were government or corporate securities that might suggest what the spread is. As a thought experiment, the difference between the interest rate on a 1,000-year and a 10,000-year loan must approach zero. To channel Spinal Tap, how much more uncertainty can it be — none more uncertain.
Fwiw, I think the larger point about debt peonage is the biggest problem. These are not instruments designed to promote home ownership but housing. Rather this solution, which will not promote equity build-up as you point out, it would be better to promote more affordable rental housing.
One thing that I have not seen discussed is that any mortgage greater than 30 years cannot be a qualified mortgage (QM). 12 CFR § 1026.43(e)(2)(ii). These mortgages would be subject to the full panoply of ability-to-repay requirements. The CFPB could amend that rule, but isn’t the administration’s position that the CFPB is going to run out of funding (per your previous post) — tongue firmly planted in cheek.
Bob is of course right about it all. There’s a SF Fed study that tries to extrapolate the extra cost of issuing 50-year rather than 30-year Treasury bonds. (https://www.frbsf.org/research-and-insights/publications/economic-letter/2021/11/what-would-it-cost-to-issue-50-year-treasury-bonds/) It concludes that it’s probably around 20 basis points. I don’t know how well that translates to mortgages, but it suggests a floor at the very least.
I would expect something higher for a few reasons.
1. The liquidity of 50-year mortgages will be lower than 30-year mortgages–they’ll likely be securitized in separate pools.
2. The SF Fed study is based on an extrapolation of government bond rates in Austria, France, Switzerland, UK. Unlike a home mortgage, those bonds are not callable, so the investor is only exposed to the risk of rates rising, not falling.
3. The credit risk on the European bonds is minimal. That’s not the case with a home mortgage, and an extra 20 years of credit risk exposure isn’t nothing–that is the difference between a 30-year mortgage and a 50-year mortgage is not solely interest rate optionality.
So maybe we aren’t looking at 102 basis points, but something more like 50 basis points. If so, there’s some savings on the monthly payment, but nothing that’s a game-changer for affordability. What’s more, this ignores the recursive effect of cheaper credit: higher housing prices. If credit is cheaper, people bid up home prices because the housing supply is limited. Higher housing prices will mean larger mortgages, which means larger monthly payments, even with a longer amortization period. The result is to cancel out most or all of the monthly payment savings.
I will note, however, that even a long-term mortgage that fails to build up equity is better than renting. First, there might be some tax benefits to the mortgage. But more importantly, homeownership is a protection against gentrification or against changes with the landlord. It guaranties continued housing tenure. A renter faces an increase in rents if the neighborhood improves; a homeowner sees an appreciation in property value. That’s a huge reason to favor homeownership–it’s an economic hedge.
Regarding QM, don’t forget that the 30-year requirement is in the statute, 15 USC 1639c(b)(2)(A)(viii). It does allow for the requirement to be changed by regulation by the CFPB. The CFPB can only change the requirement, however, “upon a finding that such regulations are necessary or proper to ensure that responsible, affordable mortgage credit remains available to consumers in a manner consistent with the purposes of this section, necessary and appropriate to effectuate the purposes of this section”. I’m unsure if the Bureau–even if it were operating–could make this showing, precisely because of the recursive effect of cheaper credit.