Transcript

March 30, 2017

Lali Shaffer: Hello, and welcome to the Federal Reserve Bank of Atlanta's ViewPoint Live. We are so glad that you could join us today. I'm Lali Shaffer, senior financial specialist in our Supervision and Regulation Division, and I will be your moderator. I'm joined today by Mike Johnson, our executive vice president, who will share an update on the banking conditions in the United States, and also by Domonic Purviance, senior financial specialist, who will discuss trends and emerging issues in the residential real estate sector. We will be taking your questions today, which you may submit online by pressing the button on your screen.

And as a reminder: the views expressed here are of the presenters, and may not reflect the views of the Federal Reserve Bank of Atlanta, or the Federal Reserve System.

And now, I'll turn it over to Mike.

Michael Johnson: Thank you, Lali, and welcome back. I think it's been a couple of shows ago since you've been here. And also thanks for the disclaimer, I really appreciate that. And welcome, Domonic. I believe this is your first time with us?

Domonic Purviance: First time.

Johnson: Great. So, really looking forward to your presentation. And there's something pretty unique about it: in addition to the content, you will be showing it in Tableau, so it's going to be very dynamic and interactive. I know everyone will enjoy that.

As Lali mentioned, I typically kick this off with a little discussion of banking conditions. The good news through the end of the year, the end of 2016, is, frankly, [that] there's no real story to tell. So we should probably move right on to Domonic, but I'll add a couple of comments along this story—more with an emphasis on what we're watching for in 2017.

If you look at the chart in the upper left-hand corner, which depicts loan growth across community banks, regional banks, and large banks—again, the story is growth is good. The only area of decline was in the large bank segment, mostly driven by the consumer. So I think that's part of what I would like to emphasize here, is a little bit of a slowdown in the consumer segment, overlaid with—at least in newer vintages—maybe some upticks in past-due credits on the consumer side, primarily in the lower FICO bands.

So that's a little something that we're going to keep our eye on, but overall loan growth looks good and in a range that I think is very reasonable. Frankly, I get a bit concerned if we have loan growth that's consistently well above GDP, so in this range I think we're in a good place.

Moving to the panel on the upper right: again, no real story. Noncurrent trends are stable, as you can see, and the only signs of any weakness that you don't see through the aggregate numbers (because of such a strong, low base) is some of the newer vintages, lower FICO band scores, primarily auto, things of that nature—so no real story there. On the earning side, which is on your lower left: earnings are hovering around 1 percent ROA [return on assets], and there's also no real meaningful distinction across community, regional, and large banks—that's a little different from, if we go back a couple years, when the regionals seemed to be outperforming the other segments. But here, everyone's pretty much around the 1 percent bucket, which is not too bad. And it's really reflective, moving over to the lower right, of a very stable, flat, net interest margin. Again, no meaningful trend there, but of course we're keeping our eye on that as interest rates, at least on the short end, rise just a bit.

Another thing that we're keeping our eye on moving into 2017, with that rising-rate environment, is the impact on deposits, on deposit mix. And while most of our banks are actually asset sensitive (and we believe this would be a benefit), we are a bit in uncharted waters, so how fast the asset side reprices is relative to the liability side is something for us to watch.

So, a couple things to watch moving into 2017, but no real story as we leave and move past 2016.

So we shift gears now; we're going to move on to Domonic and talk about residential real estate.

Before we do, in case you missed the big news: we at the Federal Reserve Bank of Atlanta [will in June] have a new president, Raphael Bostic. This is a topic that I know is near and dear to Raphael's heart in that he's a former senior official at HUD, and also a former research economist with the Board—now currently a professor at UCLA, but with a heavy emphasis on real estate and the impact that that has on people's lives.

So, Domonic, no pressure: our new president may be watching, so knock his socks off.

Purviance: Thank you. As Mike mentioned, we're going to talk a little bit about housing. As we all are well aware, the housing market has been in a slow steady recovery since the downturn, and at this point we're seeing the housing market start to enter into a transition phase as the recovery strengthens and enters into a new pattern that we really haven't experienced since the recovery began. So, I'm going to talk a little bit about the current conditions we're experiencing. As Mike mentioned, we are entering into a rising rate environment, and there is some concern about how that may impact housing demand moving forward.

This first chart you see is a depiction of home prices in terms of the year-over-year change, as the bars here on the graph. You can see the deep downturn during the recession, and you can see the recovery. Over the past few years, we've been kind of in the single digits; we came out of the recovery at 7 or 8 person, and then we've dropped down to about 5 or 6 percent. At the end of last year, we experienced a steady rise in the rate of appreciation and ended up the year with about 6 percent year-over-year growth.

If you look at the overall index—this is CoreLogic's Home Price Index—if you look at where we were at the peak, somewhere around September of 2006, the index peaked at about 186 (and again, it's indexed to the year 2000 equals 100—so that's kind of the starting reference point). Where we were at the end of last year was about 186, so pretty much we have recovered the loss in value that we experienced during the downturn—overall; there are some markets that are more recovered. If you look at West Coast markets' tighter inventory levels, they've kind of recovered higher than the peak levels, and some markets are still below. But on average, the country is about the same as we were in terms of home values at the peak of the housing market.

So what's been the driving factor for home price appreciation? Well, in this cycle it's been primarily driven by tight inventory levels. So [in] the chart that you see on the screen, the green bar is the total existing home inventory levels for the country, [and] the black line is the months of supply. So you can see where we've been over the past few years, since 2012: right around two million units, and the months of supply has steadily been between four or five months. We actually ended the year—actually the beginning of January, the beginning of this year—with about four months of supply.

The months of supply measures how many months it would take to absorb the current amount of inventory in a market given the current sales philosophy, and typically the equilibrium is between four and six months; so that's considered a balanced market. The closer the months of supply is to four months or below, that means we are entering into a pretty tight inventory market. That's really where we've been over the past few years, and that's the impetus for the strong, upward pressure on home prices.

Moving to the next slide: the dots on the chart are most of the major markets in the country; I have Atlanta highlighted in red. The Y axis is the months of supply of inventory—again, how many months it would take to absorb the current amount of inventory given the current velocity of sales. The X axis here is the year-over-year change in home prices.

The goal of this chart is really to measure the impact of inventory on home prices, so as I animate the chart forward, what you'll see is most of the major markets will begin to move within the equilibrium range—the bar here is between four to six months—and as the markets move, as inventory tightens, you'll see the rate of home price appreciation increase pretty drastically.

So this is kind of through, and you see the years change at the top—pause right here. At the end of 2013 is probably where we've seen the strongest year-over-year increase in home prices. You can see a market like Las Vegas: the months of supply of inventory was about 2.3 months of supply, and they experienced about a 20 percent year-over-year change in home prices. And as I...

Johnson: Domonic, did you invest in Las Vegas?

Purviance: No, no, no; I missed that opportunity. [laughter] There are plenty of investors during this period in Vegas as well as...Phoenix bought a lot of inventory; and really, investors were really the impetus for creating that upward pressure on price during that period.

Johnson: Thanks for that insight.

Purviance: If you keep moving forward, you see [that over] the past couple years most major markets have stayed—and this is where we were November of 2016—most major markets have stayed either within the equilibrium range or just below it, and New York is really the only market above. So, with our markets really sustaining very tight inventory levels, that's created pretty strong upward pressure on prices. So if you were in a market to buy a house today, it's pretty difficult to find a house, and when you do, there's multiple offers and multiple bids. Normally a lot of houses, particularly in good submarkets, have multiple offers in—and offers above the asking price, in many cases.

So what's been the impact of that? I have here a map of the market of Denver, and it's locked in at August of 2005. Every ZIP code here is color-coded by the median home price. So everything in blue represents a ZIP code where the median home price is above $200,000, and everything in orange is a ZIP code where the median home price is below $200,000—and some of these darker areas are areas where there's no data, so this is mostly where the housing activity takes place in the market of Denver. The chart on the right is the median home price for the overall area. So I'm going to animate the chart forward and you'll see the bars increase over time with the increase of home prices, and you'll also see a lot more blue fill in in this map.

So, for example, this particular ZIP Code 80219: in August of 2005, the median home price was $168,000. We'll see what's happened as—over time—as home prices have increased. So you'll see the changes; you'll see bars kind of increasing and becoming blues. During the last housing boom, the market of Denver median home prices increased above $200,000. Then, during a correction, they declined again. But since around 2013, you'll see a sharp increase in the median price for the entire region. That's 2016, and that's the end of the year in November. You can see median home prices went from $200,000–230,000 and increased to about $330,000, just within the span of those years. And if you look at the map here, similar ZIP codes that we were looking at before—I think we looked at that one—now the median home price is $354,000.

So you can see over the past few years, by the inventory levels being so tight, and demand being strong as we've come out of the recovery, we're seeing a sharp increase in median home prices within...and Denver is just an example; we're seeing the same in many other regions.

Johnson: So just to be clear: you picked Denver because it's essentially a median example, and not an extreme example?

Purviance: I would say it's a good...Denver's a market that's experienced pretty strong growth in their economy, and so it's pretty emblematic of markets that have experienced stronger growth. If you look at a market like Dallas or Atlanta, you'll see similar trends. Markets that have not experienced as much growth—like Chicago, or some other markets, or Detroit—you may not see the same trend. So it's an example of a market that's had pretty positive job growth.

Johnson: Right. Thank you.

Purviance: This chart moves into the new housing market. This shows the distribution of new home starts by price point, and you can see in March of 2005, the largest share of new home starts were below $200,000. Now animate it forward, and you can see where we ended up today, since the recovery—so this is '15, and this is '16. So you can see, we went from a large share of our new home starts below $200,000 to really the largest single price range is between $300,000 and $400,000—and that's pretty similar in most markets.

The impact of that overall has been a big shift in the difference between existing home prices and new home prices. In this chart, the green bar is the median new home price, and the black line is the existing home price. You can see historically—we're comparing apples to oranges, because the new home[s] might be different square footages from an existing home—but just to show the trend overall, the median new home price and existing home price really didn't have that much of a difference historically. It's been maybe a $15,000 difference, on average.

But since the recovery, there's been a big separation between what the median existing home price is and what the median new home price is. This has to do with the fact that a lot of builders concentrated their new developments in your "A" submarkets—"A" submarkets meaning they're closer to employment centers, in good school districts—they also built larger houses, and they tend to be more expensive. So, since the recovery—although, if you look at overall starts, they are not as high as they've been historically, but the starts that we have seen have been for larger and more expensive units.

So what's the impact of that? I have here on the left a map of the city of Atlanta, and on the right is a bar chart that shows the distribution of new home starts by price point. This is in September of 2004, and if you were in the Atlanta region you can see all the blue ZIP codes—the ZIP codes that have at least 175–200 starts below $250,000. So if you were looking for new home in Atlanta below $250,000, you pretty much could find something anywhere in the region. And if you look at price points, [it's] the same thing we saw nationally: a large share of new home starts were below $200,000 and below $250,000.

So, as you fast forward to where we are today, you can see the shift in the market. We weren't doing too many starts at all during the downturn, but this is where we ended up today. So a big drop in units under $200,000. The largest share is between $300,000 and $400,000, like we saw nationally. And if you look at the market, there are very few ZIP codes today within the Atlanta region where you can find a unit below $250,000. This is pretty similar if you look at Houston or Dallas or Charlotte or Nashville—you'll see pretty much the same trend. Part of this is because new home builders have concentrated in the higher-end price points in your "A" submarkets, where it's little bit more expensive.

So, the big question now is: what is the impact [on] housing affordability? Because that's been the big question that has been asked [by] many people in the housing sector. So I have here in this chart—this shows the affordability index by market. So you can see at the top the most affordable market would be Minneapolis, and the second most affordable is Atlanta. Again, the affordability index looks at what the median income household can afford given the median-priced house and the current interest rate. So one of the reasons why Minneapolis is number one—on our list we don't have all markets, just kind of the major markets included here—but Minneapolis is one of the most affordable markets, not because houses are necessarily cheaper but because incomes are pretty high in Minneapolis.

And so you can see the affordability by each market. Now, what I've done here is I've set the affordability index for each market at the interest rate of 3.95—and this is a 30-year fixed rate—and what we're able to do is see what happens to affordability as you increase interest rates over time. So that's 3.95, that's 4.5 percent, that's 5 percent, 5.5, 6, 6.5, and 7. So, theoretically—and this assumes that the median home price doesn't change—so if you keep the median home price the same and you keep incomes the same, if the 30-year fixed interest rate goes up to 7 percent, you'll see a lot more markets becoming unaffordable. So anything in orange is a market where the affordability index is below 100—so if it's 100, that's affordable; above 100 is more affordable; below 100 it's unaffordable. So you can see the impact over time.

Another way to look at it is the share of income that goes to housing. So for example, San Francisco today, just at a 3.95 percent interest rate: 50 percent of your income would be consumed by housing in the San Francisco market. There's some nuances to each market; a lot of the median income in San Francisco doesn't necessarily measure everyone's income, but just as a common measure to look at, you can see what happens over time as I raise the interest rate—so that's 4.5, 5, 5.5, 6, 6.5, and 7.

Johnson: You overlay this with the qualified mortgage rules, and we've got a problem.

Purviance: Right, right. And so, all this to say, one of the things that we at the Atlanta Fed have been focused on is the real impact of declining levels of affordability, and how the market resolves this issue and what impact it could potentially have long term. Most of the decline in affordability is driven by that upward pressure in home prices, which has been driven by the lack of inventory in our market. And, moving forward, we have to see what happens and how the market responds to a growing interest-rate environment.

So with that I'll turn it back over to Lali, and we'll take your questions.

Shaffer: All right. Thanks, Domonic. Again, you can submit your questions online by pressing the button at the bottom of your screen. We do have one here, Domonic: in your data, which real estate markets do you see being the weakest?

Purviance: "Weakest" is a difficult word to quantify. It depends on what you mean.

Shaffer: Price decline?

Purviance: If you're talking about price declines, most markets have experienced positive increases in home prices, and that's because inventory has been low. There are some markets that—like Dallas, for example, that's experiencing strong price appreciation, and a market like Chicago that's experiencing weaker price appreciation. But very few markets are experiencing a decline in appreciation because of the tight inventory levels overall.

Shaffer: Okay. Well, you mentioned that inventory levels overall have remained low over the past few years. What's been driving that?

Purviance: The primary driver—well, there are several. One is the fact that so many people refinanced their mortgages as interest rates declined, and so they've locked in their mortgage payments at a very low rate because interest rates were so low. So there's a disincentive for them to sell. If they sold their home, they would have to buy another home, and it would drastically increase their cost, their housing cost, per month.

The other reason is, during the housing boom, we drastically overbuilt housing. If you look at the period between 2000 and 2010, that whole decade was a decade of pretty large expansion in construction. Since then—since 2011 to now—we've pretty much underbuilt housing, particularly in the affordable price ranges. If you look before, builders have not really produced a whole lot of product in the entry-level price points. Most of that is because [of] higher land cost, higher labor cost—it's very difficult for them to introduce product that's cheaper. So the result is that we're not building enough housing, and the housing that we are building typically is more expensive.

If you add those two together, it creates a situation where we're pretty undersupplied in terms of inventory.

Johnson: Can I exercise a little privilege and ask Domonic a question on the fly as well? Given what you just mentioned, how do you think about the big growth in multifamily building at the same time? So, underbuilding of housing—perhaps in some markets, anyway—overbuilding of multifamily—how do you think about that dynamic, maybe from the affordability standpoint?

Purviance: Well, multifamily is interesting. The large increase in multifamily construction is primarily driven by millennials, who are coming out of school or moving out of their parents' basements (or however you put that), and they're very interested in living in town. So that's driven this demand for a lot of high-end, class-A apartment space. If you look at it overall—if you're in Midtown Atlanta, you see all these buildings going up—but overall, if you look at household formation, it is much higher than new home construction, whether that's multifamily or new.

So we're still not building enough houses, even though we've kind of seen an increase in high-end multifamily. There is a possibility that we are overbuilding at the high end of the market, but overall, we're still underbuilding in terms of household formation that we're seeing.

Shaffer: Thanks, Domonic. What about the potential risk to the economy of declining housing affordability?

Purviance: I think if housing affordability continues to decline, the most likely outcome is there will be some kind of correction in terms of pricing, whether demand declines because people just can't afford it, or we'll see some kind of alternative mortgage instruments to maybe get people in housing, or some kind of decline in overlays—there'll be some kind of reaction in the market as people lose the ability to afford to buy.

The one caveat that I would give to that is: the last time we saw a correction in home prices, we were in an overbuilt market. Today, if we see a correction in home prices, we're in an underbuilt market. So the type of correction that we'll see moving forward will be much different than what we saw a few years ago, and we'll have to figure out how we deal with it and how we adjust to it.

Shaffer: There is a question here regarding the decline in inventory: is any of that related to the lack of skilled workers in the housing industry?

Purviance: I think the lack of skilled workers has a direct impact on the cost of construction for builders. In some cases, it does impact their ability to complete homes at a rapid pace, but for the most part, it's really [that] the lack of skilled labor has caused the cost of construction to move up. So, builders have gotten the message that they may have—we showed that chart earlier where the median new home prices increased pretty drastically over the past few years. Builders are looking to try to find ways to introduce cheaper product, and normally that means build higher density or build further out. The issue is, if you drop prices and your costs remain the same, that means that the logical area where you'll cut is in your margins.

So there's a question [as] to how builders are actually going to respond to having to move down in price point, being [that] their labor costs are much higher than they've been over the past few years.

Shaffer: I've got another one here for you—they're rolling in! This person, they're saying, "My assumption is that housing prices have risen faster than wages over the past decade. What are your thoughts?"

Purviance: Yes, I agree. [laughter] In fact... [laughter]

Johnson: And hence the implications for affordability.

Purviance: Yes. As home prices have increased, and we're starting to see an increase in interest rates, it [could] be offset if we saw some increase in wages. We are starting to see wages go up, but not at the rate that's in line with the increase in home prices. That's really what's driving the decline in housing affordability.

Shaffer: Let's see. We might have time for one more question. Mike hasn't had an opportunity to speak yet, so I'm going to throw one [to] you.

Johnson: Oh, thank you. I have plenty of opportunities to speak, but...

Shaffer: [laughter] Any more issues or challenges that you think banks will be facing in 2017?

Johnson: That's a great question, and I tried to highlight some of that as we walked through the banking conditions discussion earlier. So I'll just mention [that at the] top of our radar screen—risk-wise, anyway—is: cyber, cyber, cyber. So, just to be clear: cyber. For our district, commercial real estate concentrations are always an issue, and I think that's a specific product/specific market conversation—it's not coincidental that I asked the multifamily question as well—and with a rising rate environment, deposit mix, and how that's going to impact net interest margin. But equally important, if not more so: the implications for liquidity.

Those are the unknowns, I guess, that I would say are at the top of our radar screen. That's a good question.

Shaffer: Thank you so much; I just want to say thank you to the audience for your time today. We will make slides available to all registrants and, as always, we have banking information on our Atlanta Fed website. We look forward to seeing you at the next ViewPoint Live, and this concludes today's webcast.