In yet another sign that the Great Recession is receding, earnest discussion has begun in the punditocracy on long-term cures for our nation’s home mortgage system, particularly Fannie Mae and Freddie Mac. (Remember them?)
I’m not sure if anything at all will come of this. It’s not clear that Congress has enough gas left in the tank to finish with healthcare, never mind rewiring the mortgage business. If I had to bet money I would say that this window of opportunity, in which there is consensus that something needs fixing, will close without much of anything changing.
But it’s inspired some rare discussion of that great American institution, the thirty year fixed rate mortgage. And it really is a uniquely American institution. Except for Denmark, which has a very peculiar system dating to the Eighteenth Century, only in America do consumers think that borrowing on thirty year fixed rates is normal.
In fact, at least in the view of our elected officials, a thirty year fixed isn’t just normal, it’s a near constitutional right. Like the right to bear arms in National Parks or talk on the phone while driving.
But is it really all that great a deal to begin with? Is fixing the interest rate for thirty years really as valuable as we seem to think?
The great majority of mortgages are for 30 year terms, meaning that if the borrower makes the payments as agreed for thirty years the loan will be fully paid off. (Fifteen year terms exist, but are relatively uncommon, and other terms are rare.) Most, but not all, of those 30 year mortgages have fixed rates for the whole term.
The mortgages that have thirty year terms but not thirty year fixed rates are called adjustable rate mortgages, or ARMs. At the start of the Great Recession they got a lot of bad press, but that seems to have died down. It would be nice to think this is because we all worked out that ARMs were not really the problem, but it also possible that the media got tired of reporting on complicated mathy things and moved on to better narratives like Madoff and bailouts.
ARM is a bit of a misnomer. These mortgages are generally fixed-adjustable hybrids, with a period of fixed interest at the start and then an annually adjusting rate thereafter. So a 5/1 ARM is a mortgage that will have five fixed years followed, at least in principle, by 25 years of annually adjusting rates.
Amongst the many pieces of mortgage information I would like to have but can’t find, is the proportion of ARMs that actually get to the adjusting stage and how long they stay there. Americans move house about once every 14 years on average and refinance their mortgages more frequently than that. Personally, I’m in the process of getting my sixth 5/1 ARM. I only let one of the previous five adjust at all, and then only once.
If you are shopping for a mortgage you are presented with a menu of ARMS (3/1, 5/1, and 7/1 are popular) and the old thirty year fixed. Generally speaking, the interest rate you pay during the fixed period is inversely related to the length of that fixed period, so a 3/1 will have the lowest rate, 5/1 the next lowest, and so on until you get to the thirty year fixed, which will have the highest rate.
So that’s a tradeoff: higher interest rate while fixed in exchange for a longer period of being fixed. How should a person decide which mortgage to choose? Unavoidably, it will involve predicting the future, namely how long it is likely to be that the mortgage will run.
The first reason a mortgage might not actually be in service for thirty years is that the borrower might sell the house before then. That 14 year statistic I cited is real, but somewhat misleading. Homeowners are divided into those for whom it is nearly inconceivable that they will not move within the next ten years or so and those who, if all goes according to plan, are currently in their last house.
The short termers might include those whose careers will likely require them to change cities, those who expect to be able to afford more house in the future, and those who expect to require more or less house in the future because of a growing or shrinking family. Long-termers like me don’t expect ever to need or want to move to another house and look forward to being grumpy old men right where they are now.
The point being that most people can do a decent job of guessing how long it is likely to be before they will sell and move on.
The other reason a mortgage may not live out its full thirty year lifespan is because it gets refinanced. When interest rates are below what a homeowner is currently paying, it makes sense to refinance. How often that is likely to occur is very difficult to predict, but we’ve had several waves of refinancings over the past few decades. (Another statistic I’d like to see: how many of the 30 year mortgages written in 1980 are still with us? I’m betting practically none of them.)
All of this is background to my central question: if your best guess was that you were going to need the mortgage you were taking out today for seven years, wouldn’t you obviously prefer a 7/1 to a thirty year fixed? The thirty year does give you the peace of mind that if you wind up needing the mortgage to live longer, you know what the interest rate will be, but that reassurance comes at a steep price. If the difference between the 7/1 and the thirty year is 0.5% in interest (a reasonable figure I am pulling out of thin air) then over the first seven years the difference in cost of the two mortgages is 3.5% of the principal, and that’s real money.
Turns out, the market share of thirty year fixed mortgages is around 70%. There is no question that some of those 70% of households legitimately are better off with a thirty year, at least based on their predictions of the future. They expect interest rates to never again be as low as they are now and they expect to be in their current house for the duration.
But it is simply inconceivable that homeowners that fit that description are anything like 70% of mortgage borrowers. I wouldn’t be surprised if the number for whom a thirty year fixed is rationally the best choice were a tenth of that.
So why so many thirty year fixed mortgages? Like many such phenomena, a big driver may be ignorance. Consumers just don’t know any better and nobody is explaining this to them. Indeed, an unelaborated preference for the responsible sounding thirty year fixed is a consistent theme from personal finance experts.
The tragedy in all this is that the government spends a lot of effort and truly vast sums of money to build and maintain the infrastructure for a mortgage product that ought not to be so vital or desirable. Much of the exquisitely ornate financial machinery that got gummed up last year, causing the global meltdown, was there because of the need to deal with tens of millions of very long term consumer loans, that is, with thirty year fixed mortgages on American homes.