Mike here. Over the weekend I wrote a post at Wonkblog, "In Defense of the 30 Year Mortgage." Many people have responded to this idea by bringing up the housing market of our neighbors in Canada. In order to keep this conversation running, I have a guest post by David Min, friend of the blog and a University of California, Irvine law professor. Take it away, David:
Does Canada prove the 30-year fixed-rate mortgage is of limited value? Here’s Matt Yglesias from last week:
If you cross the border into Canada it's not like people are living in yurts. It works fine. But since homebuyers have to carry a bit more interest rate risk, they seem to purchase slightly smaller houses. Alternatively if you imagine a jumbo loan scenario where the 30-year fixed rate mortgage lives but with systematically higher interest rates, you'd find that people would have to respond by purchasing slightly smaller houses. And it's not a coincidence that Americans live in the biggest houses in the world.
As I’ve outlined in the past, the dominant mortgage product in Canada is a five-year fixed-rate mortgage, amortized over 25 years, that essentially requires refinancing every five years. This product leaves borrowers open to two important types of mortgage-related risk.
First, there is the risk that interest rates will rise significantly between the time the loan is first originated and the time that it must be refinanced, causing a payment shock that the borrower may not be able to afford. Second, there is the risk that when the loan comes due, there may not be refinancing options available to the borrower, either because the property has declined in value so much that the loan does not meet loan-to-value requirements, or perhaps because banks have reduced their lending due to a credit contraction.
For what it’s worth, Canada has historically had a greater government involvement in its housing finance system, through a combination of government-backed mortgage securitization and mortgage insurance offered by the Canada Mortgage and Housing Corporation (an entity similar in many ways to Fannie and Freddie), as well as governmental reinsurance for all mortgage insurance, which in total accounts for some 70-80 percent of all Canadian home loans. So if you’re looking to Canada as a model of getting the government out of housing finance, look again (and don’t look to Europe, which also has very high levels of government guarantees for housing finance, as I explained recently in congressional testimony).
As to Matt’s broader point about Canadian mortgage finance, there is no question that we can have a housing finance system without the 30-year FRM that drives sufficient capital into housing to meet our needs (both for owner-occupied and rental housing), but that’s not the point of the debate over the 30-year FRM. The key difference between Canada’s five-year FRM and the American 30-year FRM is that the former leaves interest rate risk (and refinancing risk) with consumers, whereas the latter leaves rate risk (and prepayment risk) with financial institutions such as banks, pension funds, and insurance companies.
The key question is whether interest rate risk is better placed with households or with banks and investors. Those of us who favor the 30-year FRM argue that this risk should be placed with the latter, who are better equipped to handle this risk. The available evidence suggests that average mortgage borrowers do not attempt to predict what mortgage rates will be five years down the line. And even if they could do this, they lack access to the financial instruments that might allow them to hedge against this risk. Conversely, banks and MBS investors already spend quite a lot of resources trying to protect against interest rate volatility.
Moreover, when households are unable to deal with interest rate risk, they are unable to make their mortgage payments. This creates a double whammy insofar as higher rate risk for borrowers means higher credit risk for banks and investors. Thus, from a systemic stability standpoint, it seems to make more sense to place rate risk with financial institutions rather than with consumers.
Neither the U.S. nor Canada has experienced significant interest rate increases since the early 1980s, so the difference between the five- and 30-year FRMs has largely been a theoretical debate since that time. But as Karl Case (the economist who helped create the eponymous Case-Shiller home price index) has noted, we have at least one important data point from that last episode of interest rate volatility that suggests the 30-year FRM is preferable from a financial stability standpoint.
Both Vancouver and California had housing booms in the late 1970s, and both of course went through the double-digit interest rate increases of the early 1980s, which led to U.S. mortgage rates settling at about 17-18 percent. Then, as now, the dominant mortgage in the U.S. was the 30-year FRM and the dominant mortgage in Canada was the five-year FRM. Vancouver and California experienced starkly different housing markets in response to this interest rate volatility. Because Canadian mortgages were designed to be refinanced every few years, Canadian borrowers faced enormous payment shocks (with mortgage payments doubling or tripling), which resulted in a huge housing bust, with Vancouver experiencing a 60 percent (!) home price decline in the early 1980s. Conversely, California experienced a few years of a stagnant housing market in which potential sellers simply held onto their existing mortgages, and prices never fell in nominal terms.
This limited historical data suggests that the U.S. 30-year FRM is a more systemically stable product than the shorter duration rollover loan that is popular in Canada. Within the United States, of course, there is ample evidence that the 30-year FRM performs far better than short-term rollover loans. During the Great Depression, the delinquency rates on short-term rollover loans reached 50 percent, as underwater borrowers were unable to find sources of refinancing (sound familiar?). More recently, adjustable-rate mortgages experienced delinquency rates that were two to three times higher than fixed-rate mortgages made to comparable borrowers, as both the Federal Housing Finance Agency and the Mortgage Bankers Association have found.
All of this evidence suggests that critics of the 30-year FRM need to be treading a little more carefully in trashing the benefits of this particular product.