Remarks on U.S. Mortgage Finance
By Arnold Kling
At a conference yesterday on mortgage finance, I managed to turn a discussion of the future of securitization into a discussion of the 30-year fixed-rate mortgage. I think that the 30-year fixed-rate mortgage is an artifact of government intervention, and that without it we would have a simpler, safer mortgage finance system. Below are some remarks, some of which were mine and some of which came from attendees at the conference.
1. The U.S. is the only country with the 30-year fixed-rate mortgage. Other countries get along fine without it.
2. The core of my argument against the thirty-year fixed-rate mortgage is that without government intervention I believe that more borrowers would prefer mortgages where the interest rate is fixed or a shorter period, like five years.
The thirty-year fixed-rate mortgage includes both a default option and a prepayment option. The less money you put down, the more valuable the default option. The lower the cost of originating a new mortgage, the more valuable the prepayment option. Origination costs have fallen considerably over the past twenty years, especially when calculated as a percentage of the loan amount.
Thus, both options have become more valuable in recent years. My guess is that both of these options have been under-priced, due to government intervention. Government subsidizes these options by providing support to FHA, Freddie Mac, and Fannie Mae, as well as by tilting bank capital requirements to subsidize securitized loans that include these options.
I claim that in a completely market-driven mortgage market the interest rate on mortgages with little or no money down would be quite a bit higher than the interest rate on mortgages where the borrower makes a down payment of 10 percent or more. Moreover, my guess is that the interest rate on a mortgage where the interest rate stays fixed for thirty years would be much higher than the rate on a mortgage that stays fixed for five years, because the prepayment option on the latter is less valuable.
If borrowers were confronted with the true cost of the default option, fewer people would buy houses with little money down. Instead, they would choose either to rent or to save up for a large down payment.
If borrowers were confronted with the true cost of the prepayment option, fewer people would elect thirty-year fixed-rate mortgages. Instead, they would elect, say, mortgages with interest rates fixed for five years.
3. It is very inefficient for the government to subsidize home ownership by subsidizing the default and prepayment options through policies that encourage securitization. It could support home purchasers more directly at a much lower cost.
4. Our adjustable-rate mortgages also are peculiar. Rather than being priced at a constant, moderate spread of, say, 1 percentage point over the Treasury rate (or some other market rate), they start out at low “teaser” rates and then adjust upward to well over 2-1/2 percentage points above the index rate. What this creates is an adjustable-rate mortgage with a really valuable prepayment option–a reasonable strategy is to refinance every year, taking full advantage of the low teaser rates.
Why is that? My guess is that this is an instance of Price Discrimination Explains Everything. The bank wants to charge certain people an above-market fixed rate. It needs to find people with little knowledge and financial sophistication. It does so using the teaser product. In the process, it sets up an opportunity for highly-sophisticated borrowers to take advantage of the system by taking teasers and refinancing them on a regular basis. But most people do not want to be bothered with all that refinancing, preferring fixed rates instead.
However, I would not rule out the possibility that lenders are acting against their own self-interest. I wonder whether lenders have ever made money on teaser adjustables, given all the defaults and prepayments.
5. With substantial down payments and interest rates fixed for five years, mortgage finance would not require sophisticated financial engineering. There would be much less risk to transfer, and hence there would be much less value in slicing, dicing, and structuring mortgage securities. Banks could find ways to fund a large share of mortgages.
6. What would borrowers lose? With higher down payments, they would have less opportunity to speculate on house price appreciation. I view this as a feature, not a bug. It would reduce the susceptibility of the housing market to bubbles.
With interest rates fixed for five years, rather than for thirty years, borrowers would face risks from interest-rate cycles. If interest rates go up after five years, your monthly payment could be higher. The severity of this problem depends on how interest rates vary with inflation. If interest rates have moved up because prices have been rising rapidly, then most borrowers should see higher nominal incomes and higher house prices. If so, then on net the borrowers are not in such bad shape.
Historically, a bad time to have taken out a five-year mortgage would have been, say, 1975 or 1976. Five years later, because of inflation, interest rates had really shot up. Many borrowers would have been able to afford the higher payments, but some would not have been able to do so. The result might have been some distressed sales of houses and a temporary decline in house prices. Instead, with thirty-year mortgages what we had was the bankruptcy of many of the lenders–the S&L industry. It is difficult to argue that the latter outcome was better.