Willy Walker:
In Q1 of this year, there were $131 billion of data center projects in the United States that were turned down by local municipalities. $130 billion of investment, of investment in our future, in jobs for electricians and construction workers that were turned down by local municipalities because they don't want data centers in their municipality. That government pushback on AI on data centers is growing quickly, and where it goes, honestly, is going to have a big impact on how quickly that plane flies and how high it flies.
First of all, welcome to Sun Valley. It is great to have so many of you back here.
It would not be a Sun Valley summer conference presentation on my part if I didn't have some animal photos.
It also wouldn't be a Sun Valley summer conference from a couple years ago, and I've continued this theme forward without photographs from my friend David Yarrow.
We looked this up before today. This is the third most viewed sports image of all time, of all time. And it was taken by David Yarrow. It's the first famous photograph that David Yarrow ever took. And to those of you who were here a number of years ago when David came and presented his magnificent work, this is what launched his career in photojournalism. Let's focus on this photo for two seconds. We've obviously got France-Spain this afternoon, and then we've got England-Argentina tomorrow. The England-Argentina battle tomorrow has a lot behind it to many of you who both follow football as well as global politics. You may recall, A) all of the great matches they've had.
In that match, that's the final in Mexico City when Argentina beat West Germany. But the semifinal was the match against England, where both Maradona scored the divine goal where he basically handed the ball into the net, and then the one where he evaded five English defenders to score the winning goal.
The battle between Argentina and England centers a lot around the 1982 Falklands War. And when I was starting an airline in Argentina back in 1998, my head of flight operations was a gentleman named Marco Benitez. Marco Benitez was a former naval aviator, and Marco Benitez's claim to fame was that he sank the Antelope. The Antelope was the one British destroyer which went down in the Falklands War, and Marco was the person who dropped the bomb on it that sank the Antelope to the bottom of the South Atlantic Sea.
I was watching the Argentina game on Sunday with an Argentine friend of mine, and I said to him, hey Wences, if you think about Messi and Maradona, how do they sort of rank in your view of soccer and soccer royalty and this and that? And he goes, ah, fácil, easy. He goes, Messi, Dios, that's God, Maradona.
And I thought that that was pretty interesting; it is quite emblematic of the way that the Argentines think about Maradona. As well, on Yarrow photographs, we have this picture of Haaland, which was a little bit prescient because when he took that picture of Haaland, Haaland's global stature was not nearly what it is today after the way he played in the World Cup.
We've been waiting for fundamentals in the commercial real estate industry to turn; things sort of hadn't started to get better, and I hope that in both my presentation and other comments today, there is sort of some light at the end of the tunnel as it relates to the fundamentals in commercial real estate, and more specifically to many of you in multifamily. I was out in Chicago at the Bennett-Zell Conference a couple of weeks ago, and there was a huge amount of FOMO in the room. There were 75 of the largest owners of commercial real estate in the world in the room, and everyone seemed to be thinking about SpaceX and Anthropic and Open AI and how they all wish that they were in the tech industry and in that plane flying high and not the cowboy on the horse running along saying, hold on, I want to go on that trip with you.
And there was a lot of talk about SpaceX, and everyone was sort of like, wow, I either invested in SpaceX or I need to invest in SpaceX, trying to capture that movement taking off. Nobody really seemed to be wanting to invest in AvalonBay and Equity Residential and the merger of those two companies. We did have both CEOs of AvalonBay and Equity Residential in the room, and I thought one of the interesting things that they both said was, there were a lot of people in the room who said, why hasn't the stock price of the two companies reacted better? And Ben was very straightforward in saying, this is an efficient play. We're seeing our assets be valued by the market at a level that we don't think is fair value. We are putting these two companies together to drive efficiencies into the two of them and create shareholder value by managing the assets even better.
And as we all know, the combination of those two will likely create a big behemoth. This is Jonathan, the oldest tortoise in the world. He's like 198 years old. But AvalonBay, EQR will create a behemoth in the industry and will likely be like Jonathan, live for a very, very long period of time, and do very, very well.
Over the last year, there have been a lot of things that have had all of us a little bit worried, that there are things lurking in the bush. Tariffs a year ago were very much on all of our minds. And how would tariffs play into the economy? As we all know, surprisingly, tariffs haven't had the impact that many of us thought tariffs would have. We all also thought that immigration policy was going to have a massive impact on both the cost of labor as well as construction starts, and whether people would show up at job sites and be able to build the buildings that all of you either are building or are owning.
As I have learned from talking to many of you across the country, project size to project size has had almost no impact on either the cost of labor, which has actually gone down in many instances, or the supply of labor. Very, very interesting. And so those two issues, as it relates to the tariff impact on the economy as well as labor and access to labor, those two things did not materialize.
And yet at the same time, we've all sat around and thought about surviving to '25. A lot of people in this room were sitting there in '23 and '24 just saying, economy ain't great, rates have gone up, not a lot of money to be made, but if I can just survive until '25, hang out in that tree for a little bit longer, things will start to change. And not telling anyone in this room anything you all don't know, unfortunately, we all survived to '25, and then nothing had happened.
And now we're in '26, and things seem to start to be coming back, but it's not anything that people had sort of projected as it relates to either rent growth, rate growth, anything in the commercial real estate space. The saying now seems to be there will be a slice of heaven in '27, and if we can get to '27 and be those monkeys with our bananas sitting around, maybe there will be a slice of heaven for all of us in '27. But I think one of the big important things for all of us to do is to try and break out the noise on the right from the signal on the left.
What is it that the data is actually telling us? Because there is so much noise in the economy today. It feels as if, and Arthur Brooks is going to talk about this in a little bit, every one of us has an Instagram feed that is filled with noise.
It's clickbait; it's fun to look at, but it doesn't really tell you anything. And, oh, by the way, it's also manufactured to tell you something that it actually wants you to think, if you will. And so it's gotten quite hard to differentiate between what I am actually seeing in the data and what I am actually hearing in the marketplace.
And so I want to put forth a little bit of data in this presentation that hopefully helps break out between the signal and the noise. One of the things that has clearly happened is that the job front has actually stayed very strong. Unemployment is staying down in the sort of 4% to 4.5% range. It has held up a lot stronger than many people had expected. With that said, the cutting off of the border, the solidification of the southern border, has, as everyone in this room knows, dropped down both illegal immigration dramatically, and then the Trump administration has held a very strong stance on legal immigration, has not increased the amount of legal immigration into the United States. And that lack of new entrance to our economy has put a lot of pressure on multifamily occupancy levels.
And I think that as I talk to people about the fundamentals, and in a moment I'll show you the supply-demand curve on multifamily, but the missing component to the story right now has been everything was set up for us surviving to '25 and then things kind of taking off as the excess supply had been absorbed, but you didn't get rent growth. And the one piece to all that, I think, is immigration policy and how immigration policy has just impacted the number of people seeking shelter in the United States, household formation in the United States, and what that has done to rent. We very clearly, in a moment I will show you, some of the numbers as it relates to green shoots that we're seeing across the country as it relates to that story changing.
One of the things to think about as it relates to AI, some of you may listen to the All-In Podcast, which I try to listen to on a weekly basis. When those four gentlemen stay inside of their lanes on technology, I think they're about as good as they get. When those four gentlemen go outside of their four lanes into foreign policy and all sorts of other things, I think they get about as bad as they get. But when they stay on technology, boy, oh boy, are they insightful. And last week they had a big debate about employment, and they sat there and basically were showing stats that say that the need for computer engineers has actually gone up dramatically, that companies that are implementing AI right now are growing faster than companies that are not, and that the overall employment landscape is holding much stronger than many people thought the Armageddon of displacement of jobs due to AI. And they sat there, and they were giving one of the panelists a hard time, J-Cal, because he was basically saying that jobs are going to be taken away and certain pieces of the industry are going to go away. And they basically all threw all this data at him, and he couldn't come up with one use case or one example where AI has completely taken out either an industry or a component part of an industry.
I put this slide up, which is Teletech. Teletech is a call center operating business. At the upper left-hand point where it said peak 112, Teletech's market cap in 2022 was $5 billion. Teletech's market cap today on the right-hand side is $175 million. That is AI right there, completely disintermediating a company. There is no need for companies to go and hire third-party customer service agents sitting in a traditional call center and answering your telephone calls when a chatbot can do it better, more accurately, quicker, and more user-friendly to you. That is very clearly one of those industries that is going to have to adapt or go away as AI comes into our market. But I saw this Yarrow photo, and I thought that this was a good one given the sort of debate around AI right now, in that here's this great airplane that's all ready to take off, and that bull on the right-hand side seems to be getting bigger and bigger. That bull seems to be harder and harder to move out of the way of that plane trying to taxi.
And there are two things on that one. The first one is the government and what the government's response to AI is today and what it is going to be. In Q1 of this year, there were $131 billion of data center projects in the United States that were turned down by local municipalities. $130 billion of investment, of investment in our future, in jobs for electricians and construction workers that were turned down by local municipalities because they don't want data centers in their municipality. That government pushback on AI on data centers is growing quickly, and where it goes, honestly, is going to have a big impact on how quickly that plane flies and how high it flies. And we just know, we just know that it is not going to take off as quickly and fly as high as many people would either like to see it be or it could potentially go because the government will step in.
The Teamsters are going to step in and do something as it relates to autonomous vehicles, and they will do it before it's too late. There are 6 million truck drivers in America today, 6 million. You think the Teamsters are just going to let all those jobs go away because Waymo has come up with an autonomous vehicle that allows for all those jobs to go away?
It's really cool in San Francisco that there are a couple cabbies who now are working for Uber or whatever else. But the moment it starts to take away significant numbers of jobs, you know that the Teamsters will unionize, they will use their strength to slow that down, and they will put contracts in place that require a driver to deliver goods for Walmart. And Walmart, at that time, before autonomous is available, used across the country, will have to sign up to it.
Peter Linneman used an example last week of when he worked on the railroad as a kid. The plum job was being the shoveler of the coal into the engine. And he said clearly with electricity and with combustion engines, they no longer need to have someone on the train who shovels the coal into the train. But guess what still exists in union contracts for running trains in America? Someone to shovel the coal. And they sit on that train all day long and just pay.
While AI is ready to go, the government is going to come in, and it's going to slow it down. The other piece to it is how we do from a competitive standpoint against China. This slide is a really interesting one. This is May to June, so this is really current data. Look at monthly token usage in trillions. First of all, look at the growth that's going on in the use of tokens. From 46 trillion, and the blue is China and the light blue is the U.S., from 46 trillion tokens in China in the month of May to 98 trillion tokens in the month of June. U.S., from 37 trillion to 53. But look at the growth rate in China versus the U.S. And look at the growing disparity between China and the U.S. That's the counter to the government officials saying, whoa, slow down. We need to make sure that we're not displacing jobs. We need to make sure that we're using this technology effectively. The Chinese models are winning right now. And so the big question there is from a national security standpoint, what are we doing with all our data? And what are the Chinese going to do with all of our data if we end up using their models? Those are two things that I think very much we all need to keep in mind.
The one other piece to it is, and I'll show a slide on the stock market in a moment. The stock market is being pulled along by this investment in the future and all the forward cash flows that people expect our companies to make off of those future cash flows. This trend continues. It's Chinese companies that are going to get the benefit of that, not American companies.
Consumer sentiment has never been this low since they started doing consumer sentiment surveys in 1952, ever. Think about it, 1952 to 2026, and consumer sentiment today is lower than it has ever been. It is lower than it was during the tech bust, it is lower than it was during the great financial crisis, and it is lower than where it was during the pandemic.
And so you sit there, and you look at this data, and you say, well, how can it be that consumers are so down on the economy given the general health of the underlying numbers as it relates to GDP growth, as it relates to employment, as it relates to general opportunity? And you can put a political filter against it of how divisive our politics are today, and everyone is just saying, I don't have hope on the left or hope on the right. You can take a look at it that we're at war today, you can look at it that the economy seems to be benefiting only the wealthy and not the middle class, and not the lower class. But this is super concerning.
And at the same time, the stock market continues to go up. And we've never had such a delta between consumer sentiment and the stock market.
Typically, when you're sitting at those three points of the tech sell-off, the great financial crisis, and the pandemic, the stock market goes down with consumer sentiment. Today, the stock market continues to go up with consumer sentiment as it goes down. And that is something which is right?
Stock market right? It's the consumer's right.
One of the main reasons we all know that the stock chart is doing the stock chart is because of the Mag 7. There's seven lions in there, tigers. And they are pulling this entire economy forward. 33% of the S&P 500 today is the value of those seven stocks. 33%. In that slide that I put up at the beginning of the mother bear with the baby bears hanging on, bear cubs hanging on at the back, and the Mag 7 pulling the economy forward, very clearly they are having an outsized impact on the overall economy and on the overall stock market. But that graph versus that graph and that disparity is wildly troubling because one of them's right.
And the other piece to it is if this one's right in the sense that the economy's going great and the future is bright, then what do we do with all the people who are voting there and their sentiment that things suck? What do they do with that anger and resentment? Because they're not sitting around saying I'm part of that party. They're not benefiting from that. One of the big things on consumer sentiment is oil. The price of oil and what you pay for it at the pump has a huge psychological impact on American consumers. It's the one price we all see sort of every day. You may not go to the grocery store and figure out what the cost of eggs is. You may not go to lunch and buy a Big Mac. But everybody every day drives down some street and sees $350, $450, $550 on that sign.
Interestingly, when oil was between $60 and $75 a barrel, the average American spent only 2% of take-home pay on fuel on an annual basis. While it's a number that sits in all of our heads all the time, it's actually not a significant piece of your wallet share as a U.S. consumer of what you spend on oil. But it has a huge psychological impact on whether you think inflation is running rampant or whether you think inflation is down, whether you think prices are high or whether you think prices are low. Two quick things on oil and why, up until last week, we were seeing oil recover from being over $110 a barrel down to under $70 a barrel.
I was biking with General David Petraeus here in Sun Valley last week because he was here for the Allen & Company Conference, and I asked him that question. I said, why is oil so low? He said, first of all, there are about 5 million barrels of oil a day that are going out of Saudi Arabia to the Red Sea and getting out into the world. And they are going to increase that pipeline from 5 million barrels a day to 8 million barrels a day, which will make it so that the Saudi supply is not as dependent upon going through the Straits of Hormuz.
Point one.
He said, second of all, there's more leakage through the Straits of Hormuz than you think. And the President obviously made an announcement a couple of weeks ago about the number of ships that are getting through, but Petraeus basically said there's more getting through than meets the eye. He said, third, the United States produces 20% of the world's oil and is essentially oil independent today, and so the U.S. being able to supply its own oil and being such a large producer has a huge impact on it.
And then the final point that he made, which was something that I had not known, was he said that the Chinese have basically released almost all of their strategic oil reserves to bring down their demand for oil and keep global prices low. And he said that the Chinese releasing their reserves as well as converting to non-fossil fuel energy in wind and solar at the rate that they have has pulled a lot of the demand out of the global oil market, which has allowed for oil prices to come back down.
A lot of different pieces to why we've gotten back to where we are. We've obviously seen oil spike back over the last week. The one other thing that many of you saw this morning, the inflation print for June came in lighter than many people thought. I would put forth that that probably allows the Fed to pause. Linneman's take on the Fed at this meeting was that they would not raise or cut at this meeting. They then take off in August, and then Linneman's projection is cut in September. That's a pretty contrarian view, but I remind anyone in the room that Peter Linneman was completely contrarian in 2024 as well as in 2025. At this time of the year, nobody was saying there would be rate cuts in the year, and in both years, Peter predicted three, and in both years, we got three. Peter's track record is pretty great. Who knows whether it will come through with rate cuts this year or rate increases.
Back to housing and the real estate economy and all of us sitting on that horse saying, man, I wish I were in that SpaceX plane taking off. We will have our day in the sun. There will be the day when fixed assets and long-duration assets are valued nicely by the markets. Some of us may feel right now, as AvalonBay and EQR are feeling, that the market is undervaluing their assets, but we all know, we're in a cyclical industry. If you go back and look at the returns in the stock market for the S&P from 1998 to 2001, the tech stocks go like that, and the commercial real estate REITs go like that. And then, all of a sudden 2001 hits and tech goes like that, and all the REITs go like that. We know that we're in a cyclical industry, and we know that we will have our day in the sun.
The other thing that I was thinking about as it relates to innovation in our industry, these three stocks are WeWork, Zillow, and Compass. The big one, WeWork, as many people remember, when WeWork was ready to go public, it had a market capitalization of $47 billion. It went to zero. When Zillow came out and went public, Zillow had a market capitalization of $30 billion. Zillow is today at $7.4 billion and is off almost 90% since its peak. And then Compass, the small one there, went public at a $9 billion valuation, and it is down 50% since its IPO.
The reason I put these three up there is not to poke fun at them, but to say that in our industry there are some companies that have gone and, if you will, broken the model. They've done something innovative. They've come up with a new way of thinking about either residential real estate, commercial real estate, or the data that sits below everything that we've done. And they have, in their moment in the sun, raised or been worth billions and billions of dollars. I think one of the big questions that we at Walker & Dunlop have and that we struggle with is as AI comes into our industry, how quickly do we invest in it, how quickly do we adopt it, and how quickly is it going to change our world? And we sat around with our board yesterday and had a very detailed presentation from our tech team on what we are doing.
And I said to everyone in the room, I said, look, the hardest part about this is I have no clue where this is going, and there's nobody else in the room that has a clue where it's going. We all have a sense. And, oh, by the way, if you aren't consistent in what you do, you could change your path every single day because there's another data point every single day that tells you to go left, go right.
And so, trying to figure out where this is going and how to embrace the technology and then implement the technology is, I'm certain, a big challenge for every single person in this room.
We're all waiting for that moment when our market starts to turn and come back up to the surface like this whale. And then when's it going to launch, and when are we all going to sit there and say, yay, finally our assets are worth more than books.
This is an interesting slide. This is from our research group, and this is the light brown line, or the kind of yellowish line is the S&L crisis. Many people in the room weren't around during the S&L crisis, don't remember it. I guarantee you my father remembers the S&L crisis quite vividly. The coin darker brown one is the great financial crisis, and the light blue one is our current cycle, the great tightening. And so what we did was we took the NCREIF index, indexed it to 100 on the left hand side and looked at the drawdown in value in the NCREIF index.
You can see in the GFC it went way, way down seven quarters into the GFC. You can see in the S&L crisis, it actually didn't come down nearly as much as it has in this cycle, nor in the GFC. And then we plotted out the recovery of values in the NCREIF index over the subsequent quarters. As you can see here on the GFC, it went way deep and then recovered very nicely. Obviously, the federal government stepped in, cut rates dramatically, put a lot of stimulus into the economy, and you can see the recovery curve. In the S&L crisis, not nearly as deep, a fall off and then a much more gradual recovery coming out.
And so you can see where we are 15 quarters after the peak value back in when we started this graph, and we plotted out on the light blue what are the returns if we stay between the S&L recovery and the GFC recovery, or what happens if we actually recover like we did in the GFC. And if we were to recover like we did, like we've got it plotted right here, that's a 16% return from 15 quarters out to 25 quarters out. Nothing incredible but pretty good.
And then if you recover like the GFC, we'll be up 34% or a 12% compound annual growth rate.
I think the one thing about this slide that you can pretty much bank on is that we are in that upward trend. Barring some, well, we actually go to war with Iran and don't have it be a military conflict. Something major happens to the economy. But if we continue to trend out of here, we're pretty much going to be either between those two lines or hopefully, even trend up a little bit to the right. Let's talk about migration.
Seems like, this is back to Yarrow photos, by the way, I want to give David credit. We've got his credit in here, but you can clearly tell when I'm toggling between just stock photography and Yarrow photography. But it feels like everyone is trying to move to Florida, and everyone is trying to move to Texas.
They are clearly two of the highest-growth economies in the country, and they are the recipients of a tremendous amount of migration. But look at this slide between, on the right-hand side you've got GDP growth, and on the left-hand side you've got population growth. The slide shouldn't surprise anyone as it relates to where people are moving in the United States.
You can see Florida up there, you can see Idaho up there, you can see Utah, you can see Texas at 8.8%, and that's population growth from 2020 to 2025. But I think looking at GDP growth over that period of time is also really important because there are a couple of states like Washington state, which hasn't had in-migration; it's only at 3.8%, but their economy has grown dramatically because obviously the companies that are based outside of Seattle, Microsoft and Amazon being the two big ones. And so I thought it was interesting to kind of look at Arizona, not quite as much in-migration in Arizona, but boy oh boy is the Arizona economy cranking.
One of the big reasons for that is the Taiwan Semiconductor Plant that's being built outside of Scottsdale. That plant, right now, the projections are that 125,000 people will work in that ecosystem once the plant is built and all the feeder companies/investment are built up, 125,000 people. That's a huge job creator and GDP grower in Arizona by having that Taiwan Semiconductor Plant coming in.
Two other data points on this one, New York and California aren't on that slide, and I would assume that people would think that even without the migration to those states that their GDP still grew. In California, they've lost 5% of the population or no, 51 basis points, sorry, less than 1%, my bad. But their GDP is up 14% over that period of time. Not on that list but at 14% and New York has lost 1% of its population and has still generated 11% GDP growth over that period of time. Both California and New York, good GDP growth but obviously have some real issues as it relates to out-migration of their citizenry.
Let's go to housing for a moment and the question of whether you're either going to build a house like that beaver or whether you're going to live in an apartment like one of those bees. If you look at this slide, this is signal versus noise, folks. Take a look at the average, this is net worth by age cohort in the United States. Net worth by age cohort in the United States. 55 to 64 age cohort. The average net worth of an American between 55 and 64 is $1.5 million, average. But the median is $364,000. To those of you who can't remember back to when we learned the difference between the average and the median, the average is take everyone together including Elon Musk, divide it by the population and boom, you get $1.5 million. That's the Elon Musk effect of pulling everything up because he's got a trillion dollars in that calculation. If you go to the median, it takes that person in the middle. 50% above, 50% below and what is his or her net worth? That's $364,000. If someone walks into your office and says the average net worth of an American between 55 and 64 is $1.5 million, I go, I'm a home builder because everyone's got a ton of money. They got $1.5 million on average, and they can go buy my home. Someone else walks into my office and says the median net worth is $364,000, and I go, maybe I'm in the multifamily industry because they don't have a whole lot of money to put a down payment onto a home, and they're more of a renter than a buyer. Super important to look through the data and make sure you're seeing the data point that drives your business decisions.
Similarly, this slide, when we were in Chicago at the Zell conference, David Schwartz stood up and mentioned this slide and said this is Willy's money slide. I think this is the multifamily money slide. Rent versus own. The light blue lines in the background are the average cost of a single-family home in America. You can see it's gotten up to, on the right-hand side, $410,000. Note it was going down and has reverted back up. Note it has gone down and it's reverted back up. Also, the black line going through it is your monthly P&I expense on a mortgage. As mortgage rates have gone up, that number has gone up, and as you can see it was coming down quite nicely. It has reverted back up as well. The light blue line in the middle is the average monthly rental payment for someone renting in the United States, which as you can see has been right around $1,800, a little bit less than $1,800, all the way back to May of 2022. The cost of renting is light blue, the principal and interest on your mortgage is the black, and as you can see, those two are widening back out right now.
Multifamily has a built-in, intrinsic, right now, cost competitiveness against single-family that should make it have tailwinds as long as that delta is there. Single-family is unaffordable, multifamily is affordable on a relative basis in the United States today. You will obviously hear in a second, Ryan Marshall talk about why Pulte Homes is doing well, and you will also hear Dallas Tanner talk about why single-family rental is doing well, and then you will hear Steven DeFrancis talk about why Cortland is doing well in the multifamily space. Which one of those lines wins going forward will have a lot to do with how Pulte, Invitation Homes, and Cortland all do over the next decade.
The lock-in effect on mortgages. This is a really, really interesting thing for everyone to keep in mind. I talked about the cost of ownership versus renting. One of the things that I think a lot of people have forgotten about the lock-in effect on mortgages is what that has done to the household cash flow in America. So, Linneman has a stat, which is your disposable household income against your total net debt. And so if you go back to 2019, the DPI, disposable personal income, so you take what you owe from a debt standpoint, you take what you make, and what is left over. The DPI in America in 2019 was 84.75%. Basically, the disposable personal income, you had 15% to go and spend on other things once you've taken care of all of your debt obligations, your mortgage, your credit card expenses, any student loans, et cetera. Today, it is 75%. Every single American has picked up 10% of additional disposable income. And the main reason for that is the lock-in effect on single-family mortgages. Everyone went out and put a 2% or 3% mortgage on their home and locked it in.
And what that has done is it's driven consumer spending. And people forget about that. They say, oh, great, everyone's got a cheap mortgage, and that's really good because the housing industry is nice and stable. 9 percentage points of increased disposable income for the average American is massive. It allows them to go to Walmart. It allows them to run back to the mall, as they say. Monthly retail sales in the United States in May were $639 billion. Everyone has been thinking the U.S. consumer was going to give out, that their debt was going to be too high.
One of the big false narratives out there as far as signal and noise, people say there's a trillion dollars of debt outstanding on credit cards. There is a trillion dollars. But people are using credit cards as an ATM machine. They're not paying the interest on it. They're paying for things with it, and then they're paying it off. If you look at credit card defaults and delinquencies, and I follow this number very, very closely, while they went way, way down after the pandemic because everyone got checks from the federal government and paid off their credit card balances. And so the default and delinquency numbers went through the floor. Where they sit today is at the historic long-term average. They are not elevated.
And so the consumer is held in there very nicely. And so go back to that consumer sentiment slide. You're like, hold on a second. You're not getting consumer defaults. A lot of people have locked in long-term mortgages, which has brought down their overall debt exposure, and they're able to go to the mall and spend money. What gives? And that's back to, is it the stock market that's telling you the future is strong, or is it the consumer that's saying the future is not strong? On the hospitality side of things, all of that increased consumer spending is driving a bifurcation in the hospitality space between the very, very high end and the lower or middle end. If you look at resort properties around the world and what they're being sold for, $2 million, $3 million a key, the high-end luxury space is off the charts. Absolutely off the charts. The lower end and the general business hotel, not doing quite as well. Rev par in that category is down. RevPAR in the high end of the segment is up 4.4% year on year. Seeing this increased consumer spending at the high end going extremely well as it relates to hospitality.
Let's run back to a couple more fundamentals on multifamily. This slide, everybody in the room who's in the multifamily space knows. The light blue annual starts. The dark line is your annual deliveries. Starts peak, then comes down. Deliveries peak. We've been waiting for the deliveries to burn off.
Obviously, you've got both of them coming back into trend, and that's going to start to have a real impact on where rents go. On absorption, nice part about this, first half of 2026, second highest absorption we've had since 2010. 2010.
Remember, that's 10. 16 years ago. First half of this year, the second-highest absorption we've had in 16 years. And only to be beaten by last year's first half of the year. Really good absorption on the multifamily side. And here's a slide on vacancy.
Look at this. We've had negative vacancy, or increases to the vacancy numbers on a national basis since December of '21. All of those up above, up above, up above, are increased vacancies on a national basis. And then finally, for the last four months, we started to see it go back down. And reductions in national vacancy. You can look at the absorption numbers and the vacancy numbers and say, we're finally chewing through that excess supply and getting ourselves to a more normalized market where you can start to push rents.
This slide is confusing. Bear with me for two seconds. But I pulled the Denver starts, deliveries, and rent growth because it's so emblematic of where we've gone. Look at Denver starting from 2015 to 2018. That's that first red box up on the left. Building about 10,000 units a year in Denver. Okay? Those then flew into deliveries, which then flew into rent growth of three, 2.78 in the bottom left-hand corner there. Everyone in the room knows this. You get the starts going, then we supply them, and then we get to what that does to rent. The 10,000/year got you to 3% rent growth. Then all of a sudden it fell off in '19 through '20 where it went from 10,000 units a year down to 7,000 units a year.
And once those got delivered, what did that do? When you had that decreased supply, rent spiked. Look at those rent growths between 2021 and 2022.
That was a combination of both the supply as well as the pandemic, right? Everyone knows that. That 15%, 16% rent growth in Denver was due to in-migration as well as reduced supply previous to the pandemic.
But then look at what ends up happening. You get to '21 through '23, and we're back up at that 16,000, 17,000 units a year, and lo and behold, where do rents go? Flat to negative. You sit there, and you say, what's Denver really look like? If Denver is building 10,000 units a year, starting 10,000 units a year, you should get yourself to about 3% rent growth. If Denver falls down to 6,000 to 7,000 units a year, we should be seeing heightened rent growth, 5, 6, 7, maybe not the 16, 15.
But then the moment that we get Denver supplying at 15,000, 17,000 units a year, rents fall through the floor. We've got this data on every single market in the country. We track this. If you're looking at a market, ask us about this data. But this is a leading indicator, and you can see where Denver's gone for '23, '24, down in the 3,000, 4,000. You can only imagine that it reverts back to maybe not the rent growth that we're reporting there, but clearly high single digits rent growth.
Supply and absorption. The interesting thing about this is to look at the T12 numbers. We've got T12, Phoenix, Austin, Charlotte, Denver, all negative on a T12 basis of deliveries and absorption. But in the last quarter, in Q2, you can see deliveries and absorption get back to a positive. If you're looking at T12 numbers, you probably don't like those markets. You're looking at T1, and we like those markets. 32%, 67%, 191% in Denver, Colorado. Big change in that market from being -16 over T12 to significantly more absorption in Q2 than there was over the last year.
On this slide, just from a borrowing standpoint, I've made this comment before. Most of the people in this room are borrowing short-term, and when I say short-term, 5-year, not 10-year. You can see here this is our Fannie Mae loan durations. Going back to 2020, we did $20 billion of agency originations in that year, and all of it was 7- or 10-year money, and the majority of it was 10-year money.
You can now look at our 2025 numbers, and you can see that almost half of what we did was 5-year duration. The thing to keep in mind here is that all that 2020 paper is maturing in 2030, and all that 2025 paper is maturing in 2030. There will be a pileup of maturities in 2029, 2030, and 2031 as this year, last year, and the year before of shorter durations pile up on top of the longer durations that were done previously. So, as we talk to clients, it's just do you want to be in that refi moment, or do you want to get over it or ahead of it?
Similarly, this is agency spreads. Agency spreads have tightened and tightened and tightened and tightened, and you can see how tight agency spreads are today in 2026. The question I'd ask you to ask yourself is, you think about a 450, 460, 10-year. How much conviction do you have that goes up or goes down? If you've got a lot of conviction on it, congratulations.
You know a lot more about the markets than I do. But if you kind of sit there and say it's going to sit at 450, might go up a little bit, might go down a little bit, but it's at 450, that 2026 spread is due to people wanting agency paper. That's why spreads are so tight, because there's not that much supply.
Go back to this slide. If you get all the 2020 maturities and the 2025 maturities piling up in 2030, there will be a huge supply of paper, which means that spreads widen. Just something to think about as it relates to your financing cost. You can't control the base rate, but where spreads go will have a lot to do with the amount of agency paper that's being sold into the secondary market. Real quick on Fannie and Freddie privatization. I took this picture because there's a train and it's on the tracks, but there's a guy there with a gun pointing at the conductor that might knock it off the tracks, and there's an attractive woman there that might distract the conductor to do something else.
It seems like every time the privatization of Fannie and Freddie gets on the tracks, it kind of finds some reason to fall off. There's no doubt that Director Pulte is very focused on growing the top line and the bottom line of both GSEs. There is no doubt that the management teams, and many of them here at Fannie and Freddie, are focused on running those businesses as best as they possibly can be run. I would just put this. I would put forth that they will stay off the tracks for as long as single-family borrowing costs stay above 6%, because with the president's focus on affordability of single-family housing, with interest rates above 6% on a single-family loan, and Ryan will talk to you in a moment about Pulte buying down rates to get them down well below 6%, but without rate buy-downs, with interest costs above 6% on a single-family mortgage, I don't think that the federal government can take the risk that if they were to privatize Fannie and Freddie and the spreads on their bonds were to widen out, that would then be that the borrowing costs of the U.S. consumer would widen out, and that wouldn't meet the president's intention of privatizing Fannie and Freddie. As long as borrowing costs are heightened for the single-family consumer, I would think that the privatization of Fannie and Freddie is off the rails.
Welcome to Sun Valley. Thank you all for being here.
