Fannie Mae's chief economist unpacks divergent views on housing

There’s a limited return to the office now. Do the migration numbers reflect that?

I think they will, but if you only come into the office two days a week, you still may relocate your house further out, where land is less expensive. It's still relatively cheaper, because you have to absorb less in the way of a commute than before the pandemic.

You're also seeing the impact of remote or hybrid work on office space valuation. There'll be negotiating between property owners and tenants. Part of the decision companies are making will be based on trying to understand total productivity change from changes in staffing. It won't be the same for all job categories.

My personal view is we're still probably two or three years away from stability on this front. I think companies are still sorting it out.

A related question is what's going to happen with loans on office space. Many of them were made at very low rates, but their financing costs are going to reset. At the moment, unless leases are being broken and that kind of thing, there's probably not an immediate threat. But certainly you see pressure in the regional banks because of it. They originated many loans at very low rates that will reset at higher ones. Investors are still significantly concerned about what's going to happen in those regional banks in general, and if you look at the bank term-lending facility that was set up when Silicon Valley Bank went down, their usage is still very high, which also suggests continued stress.

I mention commercial trends to get people to think outside of the class of real estate that they might be involved in. A lot of the residential lenders focus on people making or refinancing mortgages and aren't necessarily thinking about how the people for whom they are making those mortgages might be changing their life circumstances.

What does the current loan performance outlook look like?

If the Fed achieves their desired success in loosening labor markets, it means increasing unemployment. I was asked recently if there is a rule of thumb in regards to how much of an increase in unemployment creates an increase in serious delinquencies. That's an interesting question now.

In the last two serious downturns, there has been huge intervention into the market from the government. So you can't just look at the data from the past. We're now less certain about what that relationship would be. It's reasonable to expect another policy intervention.

I've been asking lenders, do you believe that mortgage lending is still secured lending? Have these interventions severed the security from the loan? They say, "I don't know." The reason I ask is there might be questions about whether risk premiums are wide enough if there's acknowledgement that you no longer have access to the asset. If you think about capital markets and functioning and how pricing takes place, if you've lost the security, you would think risk premiums would rise. So maybe that's a piece of why yield spreads are wider.

There are other things like the Fed exiting mortgage-backed securities. They're the single biggest holder in the world. Who replaces them? It's going to be somebody that's going to care about yield. We've got time to figure it out because the Fed's portfolio is running off very slowly.

You also have to ask, is there a reputational risk for mortgages in the banking sector now given Silicon Valley Bank's problems? What's going to happen with bank rules? Maybe that affects spreads. Nobody believes that a 7% mortgage is going to be around for long, so prepayment risk could grow. There are a bunch of things that address the question of why spreads are as wide as they are.

Our forecast has no rate changes until May of '24, when we think there will be a mild recession underway, and the Fed will start to cut slowly.

What’s the Fed looking for in housing when it comes to rate policy?

There, they're looking at the rental component of the services part of the Consumer Price Index and Personal Consumption Expenditures. It's easier to see in CPI. It includes actual rent changes, plus equivalents for owners measured against the home purchase market. They're looking for a slowdown in that in the rent component and that has started to happen. Our thought is that rents this year in multifamily will increase. Our formal forecast of house prices, which we do on a quarterly basis, right at the moment is about 4% for the full year of 2023. But prices have been rising more strongly since we did that estimate, so we have to rerun all the models.

Prices have risen. That will work against what the Fed is trying to achieve in housing. They've gotten some benefit from policy but maybe not as much as they had hoped.

We have recession-level home sales volumes. How does that play into the current outlook?

If you talk to the average person on the street and if they're not in the market to buy, you might get from them the "housing is OK" feeling. But if you're a loan officer, this is a recession.

Our National Housing Survey suggests people think of this as a bad time to buy a house, because prices are rising and interest rates rose again. So there's this sort of conflicting perspective in the marketplace.

When interest rates rose past 7% the first time this past year, what you saw was the securitized market froze up briefly, and rapidly builders started offering buydowns. Why would they do that? Historically, builders in order to cut prices would offer additional attributes to the house. Why didn't they do that? There are two data points that speak to that. One is right now the share of new home sales as a percent of the total is at the highest level it has ever been at. Second, within those new home sales, the share of new home sales going to first time homebuyers is high.

A first-time buyer is not interested in attributes, they're looking for a basic house. Their problem is getting a down payment and affording the financing.

That's a really important nuance. Warren Buffett just spent several hundred of millions of dollars, buying a piece of D.R. Horton, a builder that serves first-time buyers. That suggests the supply issue is going to stick with us for a while. That's where all the pressure is from the demand side in the new home market, because there just aren't existing properties available for sale.

The builders are working through the backlog they had because of supply issues. Those are largely solved. As sales proceed, they're likely to enter a period where margins will widen again.

At what point does the resilience seen in home prices give out?

I think it is hard to say, but I have to forecast because it's my job. I would watch for the point in time when the peak first-time homebuyer age of the millennials passes. It depends on what numbers people use, but it looks to us like it's maybe three years from now.

There have been famous papers by well-known economists that have predicted the collapse of the housing market, to their detriment, because it never happened, so I'm not going to predict a collapse. People thought 10 years ago that millennials would not want to own homes because they saw the damage from foreclosures during the Great Recession. But when we surveyed in June of 2010, 90% of them eventually wanted to own a home. This is in my view, a long-standing permanent impulse of people in the United States. So I don't foresee any collapse of the housing market.

But demographics do suggest a change in the balance between supply and demand, which would slow the pace of price appreciation. Also there are periodic price declines on a regional basis. Right now you're seeing some big declines on the West Coast in some of what were previously the hottest markets, but on average across the country prices have risen. Migration has definitely shifted from the West Coast to the Southeast as people have been moving.

So I would say it's more likely to be a rebalancing of supply and demand that would kind of flatten the market out or return it to a more normal pace of appreciation.

Supply also is a case of the baby boomers aging. While their aspiration is to age in place, at some point infirmity sets in, and some of them have to start exiting their existing homes. So the question is, does the supply on the market become greater as a result of that?

We have surveyed people 60 years of age and older that own homes, to ask them what is your life gameplan. People are always talking about reverse mortgages, but they hate reverse mortgages.

Why are people saying they hate reverse mortgages?

There's probably a couple of things, one of them psychological. It suggests a financial inefficiency in their past behavior and recognition of that in some way. It largely removes the option of giving the house to the next generation, not that the next generation always wants the house.

Most people that take reverse mortgages are forced to take them by health-related issues, which usually means your time in the house could be limited, and the loan can be financially inefficient. You have to make assumptions about the future value of the house and you and the lender may have wildly different views of that.

There are exceptions. Mutual of Omaha has reverse mortgages as a component of a sophisticated financial plan for many households but that tends to be properties that have a value of $1 million or more.

Lenders are more focused on cost savings. How has that impacted technology?

My thinking has been that it's a very inefficient business and technology should be brought to bear to streamline processes, changing fixed costs to variable expenses. Lenders have had a different idea, particularly independent mortgage banks. They needed to get to the consumer, so they spent all their money on borrower-facing technologies.

Now, what you're seeing is that they feel pretty good about consumer technologies and they recognize we should have been spending time becoming more financially efficient. Today in our Lender Sentiment Survey, it's clearly cost management that's the priority.

What's an example of cost-efficient technology? Artificial-intelligence driven workflow?

I'm sure it will be but there are lots of fair lending considerations and things. You have to be very careful about the application of it. Things like facial recognition technologies and things like that have been shown to have some biases built into them. That's going to be one of the things that will have to be satisfied before deployment.

I think there's been a lot of experimentation and development of it but don't know that there's anybody that has had a high degree of success. It's of interest to me that Intercontinental Exchange has acquired technology in so many parts of the mortgage process. They may have the tools to do some work in that space that others might not but it remains to be seen.

Current click-through charges extract value. Smaller companies are creating their own loan origination systems primarily because the big LOS vendor charges absorb a lot of what would be the cost savings.