The title of this post may sound like a rhetorical question, but I assure you I don’t mean it that way. It’s a question I’ve been pondering for a while as I come across weird datapoints from the history of home financing around the world. A 30-year fixed-rate mortgage is maybe best defined by what it is not: it’s not a one-year floating-rate mortgage.
What would a one-year floating-rate mortgage look like?
There’s no mechanical reason a one-year floating-rate mortgage couldn’t exist. At the beginning of the first year, a home-buyer would take out a loan for the purchase price of a house, secured by the house as collateral. They’d make payments throughout the year at whatever the prevailing interest rate was (the borrower and lender could agree on whether payments would fluctuate or whether payments would be fixed and the amount of interest and principal paid would fluctuate), and then at the end of the first year, the home-buyer could either repay the remaining principal or take out another one-year floating rate mortgage to repay the balance of the original one.
While there’s no mechanical reason such a system couldn’t exist, it has some obvious drawbacks:
- Vulnerability to changes in house value. In a rising housing market, such a system would be the equivalent of annual cash-out refinancing. In a falling housing market, it would be an annual cash call, since the borrower would have to pay the previous year’s lender the amount borrowed, even if the next year’s lender was unwilling to lend the full amount given the lower value of the house in year 2.
- Uncertain total repayment term or uncertain payments. In a period of rising interest rates, each year the borrower would repay less and less of their loan’s principal (or face rising monthly payments). That means it might take many more than 30 years to fully own a home, even if at the beginning of each year you intend to pay off one thirtieth of your remaining principal. Uncertain monthly payments also would make it difficult to plan your expenses in advance.
- Borrower credit risk. Even if economy-wide interest rates remained steady, the interest rate an individual borrower might be forced to pay could vary from year to year as their personal credit profile changes. A higher income might lower their monthly payment in one year, while a lost job might make it impossible to roll over their loan at all — even if they have the next year’s worth of payments saved up in cash!
- Lender risk. In case of an economy-wide shock to the credit market, it might not be possible to refinance a loan at all, even for the most credit-worthy borrowers. Mass foreclosures and evictions based on banking sector malfeasance would be a constant risk in an economy of one-year mortgages.
- Huge transaction costs. Presumably such a system would create streamlined refinancing offers (perhaps even interest rate discounts after a certain number of years) but verifying the credit and income of a borrower and the value of a home each year would still require an enormous amount of work, and create enormous scope for human error.
That said, there are some obvious advantages as well, for both borrowers and lenders:
- Lower interest rates. If lenders didn’t have to squint at the horizon and forecast the future trajectory of interest rates, they would be able to offer much lower interest rates today. You can see this in the interest rate differential between 30-year and 15-year fixed-rates mortgages. It would be reasonable to expect the rate to be even lower for one-year floating-rate mortgages.
- Less prepayment risk. When a lender makes a mortgage loan today, they’re forced to take into account the risk that if interest rates fall their customers are likely to repay their loans early and refinance into a lower-rate mortgage. That creates an unpredictable, one-way risk for the lender and for the buyers of mortgage securities. Rolling one-year loans would drastically reduce that risk, since few people would be willing to go through the hassle of refinancing a loan mid-year.
- Reduce borrower debt. Since each year borrowers would only borrow the amount they needed to repay the balance of the previous year’s loan, they could use annual savings to reduce the amount borrowed each year. This is the same logical process as accelerating payments on a 30-year mortgage, but the annual refinancing process might provide a sense of immediacy or urgency to the process, or simply given borrowers a chance to look at their household assets and expenses.
- Increase home equity. Lenders would presumably insist on large down payments in order to shield themselves from the risk of a decline in the house’s value between the time they make a loan and the time the borrower seeks another. Much higher home equity might contribute to fewer foreclosures and evictions in general.
A 30-year fixed-rate mortgage is a one-year floating-rate mortgage with a bunch of riders
That brings me back to the question, what is a 30-year fixed-rate mortgage? In the same way that a delayed variable annuity is an immediate fixed annuity with a bunch of riders on the contract, I think of 30-year fixed-rate mortgages as one-year floating-rate mortgages with a bunch of riders:
- instead of a floating interest rate, a fixed interest rate;
- instead of variable payments, fixed payments;
- instead of annual credit assessment, assessment just once at the beginning of 30 years;
- instead of having mandatory cash calls for any change in your home equity at the end of each year, never having a cash call for declines in your home equity;
- instead of being vulnerable to changes in the supply of credit, being guaranteed the same terms as long as you stay current on your payments.
Of course, just as with a delayed variable annuity, each of these riders comes at a cost, generally in the form of higher interest rates.
Once you know you’re paying for each rider, you could decide which ones are worth paying for
What got me thinking about this most recently was seeing the statistic that the median length of tenure in single-family owner-occupied housing is 15 years (6 years for multi-family condo owners). In other words, 50% of single-family homeowners move before spending 15 years in their home. But between 85% and 90% of mortgages are for 30 years. That means 35-50% of home buyers are paying for riders they don’t use: fixed interest rates and payments and credit insurance for the second 15 years of their mortgage.
That begs the question: do the 50% of homeowners who move within 15 years know in advance they’ll move within 15 years? If so, they’re overpaying for those 15 years; they’d be strictly better off with a 15 year mortgage (even if that meant a balloon payment in year 15, by which point they plan to sell the house anyway). If not, what they’re paying for is an insurance policy so that if they end up belonging to the 50% that stays longer than 15 years, they won’t have to renegotiate new terms and be susceptible to intervening changes in the credit market.
That insurance is worth something, without a doubt. But the price is higher interest payments and lower principal payments during each of the first 15 years. By saving that money instead, home buyers could “self-insure” against the same risk, with the potential upside of being able to keep the savings if they do, in fact, move before then.
The problem with the federal housing finance agencies is that they subsidize one particular set of riders
On the one hand, the idea of securitizing standardized mortgages and making it possible for investors to select the particular borrower and loan characteristics they are looking for while freeing up bank capital to underwrite and issue new loans is a fairly reasonable idea. On the other hand, the private sector has demonstrated absolutely no interest in doing so, and it was left up to the federal government to create the so-called “government-sponsored enterprises,” or GSE’s (the failure of the GSE’s in 2008-2009 suggests one possible explanation of the reluctance of the private sector to undertake this enterprise).
The problem is not that the government has decided to subsidize homeownership. The problem is that doing so has made offering unsubsidized mortgages uneconomical, since mortgages that don’t qualify for sale and securitization have to be held on the lender’s books or privately securitized (i.e. without an implicit government guarantee), tying up that capital until the loan is repaid, while qualified mortgages can be immediately sold and the same money lent again to another borrower, with a new set of origination fees piled on top.
Freddie Mac has this fairly hilarious document signed by their “VP Chief Economist” Sean Becketti explaining that:
“The considerable benefits of the 30-year fixed rate mortgage to consumers are beyond question. However, this type of mortgage isn’t a natural fit for lenders. All the features that benefit the consumer—long term, fixed interest rate, and the option to prepay the loan without penalty—create serious headaches for lenders. As a result, the federal government created Freddie Mac and other institutions that allow lenders to hand these headaches over to the capital markets, where sophisticated portfolio managers have the tools and expertise to manage the investment risks of the 30-year mortgage.”
Now, 15-year (and shorter) fixed-rate and floating-rate mortgages are still available, and under the right circumstances may even be qualified mortgages for securitization purposes. But balloon loans, which require repayment or refinancing in the final year of the mortgage, are decidedly verboten. In other words, for a mortgage to be qualified for GSE securitization, it has to be repayable in equal payments over the term of the mortgage, which in practice forces people who have no intention of staying in their home for 30 years into 30-year mortgages — and paying the correspondingly higher interest rates for each of the years they actually remain in the home.
I own 50 common stock shares of Freddie Mac and of Fannie Mae. I’ve lost about $150 on them since I bought them, but remain hopeful that the current administration will sell out the American public by allowing the GSE’s to recapitalize and start paying dividends again, in which case I’ll make a fortune.