Wharton's Benjamin Keys discusses his research on mortgages and monetary policy.

The housing collapse and global recession that began in 2008 was a devastating event for so many Americans, with millions losing their homes through mortgages they could not afford. Eight years later, there is still a wealth of information to be mined from that event that could be used to reshape economic policy going forward. Benjamin Keys, a Wharton professor of real estate, recently talked with Knowledge at Wharton about his research on interest rates and the real estate market.

An edited transcript of the conversation follows.

Knowledge at Wharton: Give us a brief overview of your research. What were you trying to find out?

Benjamin Keys: A lot of my recent research has focused on trying to understand the role of monetary policy and how that transmits to households through the mortgage market. Traditionally, we’ve thought of the ways to stimulate the economy as being either through the fiscal side, say through a tax rebate, or through monetary policy, lowering of interest rates. But the challenge is figuring out how those lower interest rates actually hit consumers’ pocketbooks.

In this project, we’re looking at the mortgage market as the key channel through which lower interest rates affect households. When interest rates were lowered in response to the Great Recession, we were able to study this in much more detail.

Knowledge at Wharton: Is this a natural experiment?

Keys: Yes. We had a time period of unusually low interest rates, and that lowers the cost of borrowing for households that are not yet borrowing. But in particular, it benefits mortgage holders, who can either refinance if they have a fixed-rate mortgage or, in the case of an adjustable-rate mortgage, a mortgage that’s indexed to the treasury rate. Their interest rate falls automatically, and their payments fall automatically.

One of the ways that we tried to study the pass-through of interest rates to households was looking at these automatic resets of adjustable-rate mortgages. The way that we tried to tackle this question was to focus on two different types of adjustable-rate mortgage borrowers: Borrowers who had a long fixed period and borrowers who had a short fixed period. One group would be exposed to lower interest rates sooner, rather than later.

What we found was that the group of borrowers who were exposed sooner had interest rate reductions of about 175 basis points. If you translate that into dollars, for the average loan, it reduced their mortgage payments by about $1,500 in the first year and $3,400 over the first two years. To compare that to a traditional tax rebate, you might get a one-time check for $500. This was a much larger effect and something that benefited them throughout the life of the loan.

What we wanted to see was, what was the impact of this reduction? We were able to take advantage of some amazing data that links mortgages to credit records. We could track this reduction in payments through not just mortgage payments, but also to the rest of the credit portfolio. What we found was that mortgage defaults fell 36% for this group that received this reduction in payments, this monetary policy stimulus. So, a very large reduction in default rates. That’s not especially surprising. If you lower people’s mortgage payments, they’re going to be more likely to make those payments.

“If you lower people’s mortgage payments, they’re going to be more likely to make those payments.”

What we thought was more interesting, and more novel, was the fact that we could track where that reduction in payments was landing through the rest of their portfolio. We were able to see a reduction in credit card payments and credit card debt. About 20% of that reduction was going towards paying down credit cards, and that was especially a large factor for those consumers who had a lot of credit card debt to begin with. They’re taking this reduction in mortgage payments and transferring that money to their highest-cost debt. We sort of see the re-balancing of the household portfolio.

What we also see is households going out and spending that money in the form of buying new cars. Consumers who receive this reduction in mortgage payments, relative to the group that was going to receive the reduction later, spent about 10% of that monetary policy stimulus on new car purchases.

Knowledge at Wharton: Are there other key takeaways that you found?

Keys: One of the things that we wanted to highlight in this study was the benefits of automatic transmission of monetary policy through adjustable-rate mortgage contracts. Again, the borrower doesn’t have to do anything. They get a letter in the mail that says, “Congratulations, your payment has fallen.” Of course, we’re coming to a time period where interest rates might rise, and they’ll get the opposite letter, “Sorry, your payment’s going to go up.” But that’s a really stark contrast to the fixed-rate case, where you have to actively refinance your mortgage. You have to pay attention to what the current rates are. You need to reach out to a lender to be re-evaluated and re-underwritten for that loan. Many people who are underwater on their mortgage couldn’t qualify for refinancing, so there’s a really nice benefit to adjustable-rate mortgages through this automatic transmission of monetary policy, which I think was under-appreciated.

Knowledge at Wharton: I can see how this could have a lot of implications, particularly for policymakers. What do you think are some of the practical takeaways from this?

Keys: I think one of the challenges for policymakers going forward is to rethink the predominant mortgage contract, which is the fixed-rate, 30-year mortgage. That’s a contract that’s been popularized by Fannie Mae and Freddie Mac. … Those contracts are really nice for borrowers in certain ways and really challenging for other types of borrowers.

On the good side, it’s a fixed payment every month for 30 years. You know exactly how much you’re going to owe every month for 30 years, and that type of stability allows you to make very long-term plans. It’s very easy to assess whether you’re going to be able to make that payment or not. The challenge of the fixed-rate contract is that you have to actively refinance that contract, and you have to be approved for that contract. So when credit dries up, as we saw it dry up in the wake of the Great Recession, and standards are tightened, there’s going to be a large group of people who are going to be unable to refinance.

“One of the challenges for policymakers going forward is to rethink the predominant mortgage contract, which is the fixed-rate, 30-year mortgage.”

This is what the Home Affordable Refinance Program (HARP) was designed to counteract. We wouldn’t need such an extensive HARP program if more consumers were on adjustable-rate mortgages. The monetary policy impact would pass directly through this automatic refinancing. The risk there is that consumers are going to be bearing that upside risk as well. When rates rise, their payments are going to go up. We might not think that a lot of low-income households should be facing that kind of risk. They should be insulated from future interest rate changes.

I think one of the key policy takeaways is that in downturns, adjustable-rate mortgages look like a really convenient mechanism through which households can benefit from policy. The challenge going forward is to think about what’s going to happen on the upside.

Knowledge at Wharton:  What’s next for this research?

Keys: Part of the research going forward is going to focus on the impacts of rate increases as we begin to see those. This is a project where we’re waiting in the wings for the Fed to raise rates. I’m not going to speculate on when that’s going to happen, but I think as we see rates rise, we’ll see these types of adjustable-rate mortgage contracts become costlier for borrowers. They’re either going to want to refinance out of those, or we’re going to see the impact on their pocketbooks.

The other thing we’d like to study in more detail is the regional impacts. There are parts of the country where adjustable-rate mortgages are more prevalent or less prevalent. In our existing work, we’ve done some analysis looking at the regional impacts, and we’ve shown that the regions of the country that had more of these adjustable-rate mortgages, where rates fell faster and payments fell faster, recovered faster during the Great Recession. We saw increased house prices, we saw more auto sales, and we actually saw increased local employment in those parts of the country. We want to do more to study the regional impacts and get a better understanding about how a national policy like setting interest rates can have really divergent impacts across different parts of the country.