Anne Hartnett

Hi, I’m Anne Hartnett and I am here with real estate economist Matthew Gardner to chat about our Q3 update. How are you, Matthew?

Matthew Gardner

I am well, Anne, how are you?

Hartnett

I’m great, thanks. So we’ve been waiting years for housing transactions to recover. At what point do we stop calling this a cyclical showdown and recognize that something more fundamental has changed in how Americans buy and sell homes?

Gardner

Great question. I mean, the short answer, as far as I’m concerned, is that we should have stopped calling it a cycle, or cyclical, about a year ago, because the forces that are holding sales down now aren’t really the kind that are cycle-fixes. So let me explain. When existing home sales, we know, fell from over 6 million a year in 2021 to about 4 million in 2023, and they’ve been stuck there for nearly three years now.

September numbers came in at 3.99 million. Now, there are three structural forces that explain this: One, the mortgage rate itself has become an asset that people just won’t sell. I mean, in the second quarter, 49% of outstanding mortgages carried a rate at below 4%. Now, a record 43% were originated between mid-2019 and mid-2022. That means that selling gives up that loan — and that’s not a cycle. It’s a balance sheet decision, and it decays very slowly. There are some who suggest that the share of mortgages at 6% or higher, which is around 22.5% in Q2, only hits 27% to 29% by mid-2027. That’s a slow grind.

Now secondly, the buyer pool has aged. Last year, a typical first-time buyer was 40 years old, and a typical repeat buyer was 62. And last year, first-time buyers were 21% of purchases. That’s a record low, and roughly half of their pre-2008 share. And a market that runs on equity rather than income, well, it transacts less often. And finally, the ten year Treasury, it’s running around 5% today, and the 3% mortgage was a product of a zero-rate world that no reasonable forecast brings back.

But this year, it gave us a small test of my belief. Briefly, back in February, mortgage rates dropped to below 6%, just. I think it was at the end of February. Lowest rate since 2022. But by June, they were back up to 6.5. Now they’re above 7.3. Now, at that point, when they were very low, NAR had forecast a 14% jump in sales. But what are we doing today? We’re running below last year. So the dip was far too brief to prove anything by itself. But in a normal cycle, cheaper money brings buyers back fast. This time it didn’t.

So the bottom line for me is that 6 million sales a year is the wrong benchmark. 2021 was an anomaly — a once-in-50-year financing, or refinancing, window. In 2019, we sold about 5.3 million existing homes. And my view is that the realistic norm, at least for the rest of this decade? Somewhere between 4 and 4.5 million. And we’ll get there through life events such as job changes, divorce, death, downsizing, not through a rate rally.

So for brokers, that changes their business model. In a 4 million-unit market, well, the pipeline isn’t buyers waiting for rates to fall. It’s people who have to move regardless of rate — relocations, estate sales, divorces, retirees. Now those transactions happen at 7%. If your marketing is built around a quote, wait for rates to drop, unquote, well, you’re building for a market, unfortunately, that may not show up.

Hartnett

Housing drives everything from construction and lending to consumer spending. Could several years of depressed transaction volume begin to weigh on broader economic growth?

Gardner

It already has, but interestingly, not in the way that most people assume. So let’s start with what counts. The sale of an existing home is just a transfer of an asset. It doesn’t add to GDP or growth. What counts is construction, remodeling and also the services around a transaction — that would be brokerage commissions, and also households buying new furniture, new appliances, that follow any move.

Residential investment is only 3.7% of GDP against a historic average of about five. Now, housing’s total share of the economy? About 16%. That’s actually still the lowest we’ve seen since 2019. Now, residential investment shrank for five straight quarters, falling at an annual rate of about 7.8% in Q1. It did tick up a bit in the second quarter.

So no doubt housing has been a drag. But the contrarian part is that the economy grew right through it. Nonresidential investment rose 8.5% in the second quarter, driven by AI equipment and also software. Consumer spending rose 3.4%. You see, homeowners are sitting on substantial equity, and that wealth supports spending even when they don’t move.

Low transactional volumes is a slow leak, not a recession trigger. Now the real risk, as I see it, is actually different. Housing’s historically been the economy’s shock absorber. A recession hits the Fed cuts rates to stimulate the economy, ultimately mortgage rates drop, housing turns and construction hiring leads to a recovery. Now that channel’s impaired. Ten-year treasury near five, inflation at 3.4, and in early October, markets are still weighing whether the Fed’s next move could be to hike in order to address inflation. But if the labor market weakens further — now remember, September we added just 29000 jobs — well, the Fed should be cutting rates, but I just don’t see that happening. So we may not get the usual housing-led rebound on the other side.

Now where you feel the drag first is local and also sectoral. It’s not in GDP. It shows up in title, escrow, mortgage originations, moving companies, furniture retailers and obviously broker incomes, but also shows up in city and state budgets. And they depend on real estate transfer taxes. In Washington, excise tax, there being an example, now those are real jobs and real revenue, and they’ve been in a recession for three years. Well, the national economy hasn’t. So my long-winded answer to your very direct question is that low transactional volumes aren’t going to sink the US economy, but it has quietly created a recession inside the real estate services economy. It’s also removed the tool that we normally use to climb out of a downturn. That’s homebuilding.

Hartnett

As we all know, existing homeowners have substantial equity and low mortgage rates, while first-time buyers face record affordability challenges. Are we creating a lasting generational divide in homeownership and wealth?

Gardner

Wow. Yes we are. And the word generational may actually undersell how lasting it’s going to be.

The numbers: First-time buyers — 21% of purchases in NAR’s 2025 profile, a record low. Half, roughly, of the pre-2008 norm, roughly 40%. Median first-time buyer was 40. In the 1980s, it was the late 20s. Repeat buyers were on average 62 years old, median down payment of about 23%.Thirty percent paid all cash. First time buyers, median down was 10. So what does that mean? It means that there’s two different markets operating inside the same market.

So, how do I explain this? Here’s why it will last. Homeownership compounds. And I mean that buying your first home at 40 instead of 30, that’s not just delaying homeownership. It means 10 fewer years of owning a home, meaning 10 fewer years of amortization and appreciation. Now, NAR actually estimated that buying a home at 40 years old instead of 30, that costs roughly $150,000 in missed equity on the typical starter home. Now that $150,000, that’s the down payment for the next move-up purchase, and eventually, it’s the inheritance, right? So this is less a divide between generations, more of a divide within them than between them. I mean, millennials and Gen Z who’ve got home owning parents, they get help. They’re gifted down payments, cosigners and eventually a share of the largest intergenerational wealth transfer in U.S. history. But those without that help, they’re the locked-out ones. As boomers pass that equity down, as they will over the next 20 years or so, well the gap between young owners and young renters, that’s going to widen, not close. So today’s affordability problem becomes tomorrow’s inequality problem.

There’s a second point I want to make as well. Price isn’t the main enemy right now. I mean, the Case-Shiller numbers that just came out showed inflation-adjusted home prices falling for 13 straight months. Now what’s hurting first-time buyers is the payment. A 7.3% mortgage, plus insurance and taxes, and the down payment hurdle when rent eats into savings. So it’s very, very hard for the first time buyer.

Now the fix, and I’ve said this for years, is on the supply-side of the equation. Specifically, starter-home supply. Townhomes. Condominiums. Middle housing. Cottage housing. And that’s a job for city councils, not interest rates. And for brokers, the practical takeaway I would give them is that today’s first-time buyer is older, more financially sophisticated, often getting help from families, and also they are extremely payment sensitive. So assuming FHA and VA loans, seller-paid rate buydowns, down-payment assistance, well these aren’t niche tools anymore. They’re how this buyer actually gets to the closing table.

Hartnett

We focus heavily on rates and inflation. But with AI reshaping hiring and companies reassessing their workforce, could employment become a greater threat to housing demand than interest rates?

Gardner

I think it already is Anne, the margin anyway. And I’d absolutely tell your viewers who watch this, they need to watch the jobs report more closely than any Fed report.

Let’s look at September. Employers added 29,000 jobs. July was revised down to a loss of 10,000. Unemployment rate rose to 4.2%. And by the way, I think that’s actually a real understatement. In reality, the unemployment rate is around 7.6% today. But wage growth, it’s slowed to 3% annually. That’s the slowest increase since May of 2021. While inflation is running at 3.4%, meaning real wages are falling for the typical worker. And mortgages, they’re underwritten on income. Rates determine how much you can borrow. A job determines whether you borrow at all. So yes, the job number is very important.

On AI, and I want to be careful here because the data is genuinely mixed. AI is not showing up as mass layoffs in the aggregate numbers, but it is showing up as a hiring freeze at the entry level. I just finished a Stanford study that found that employment in people aged between 22 and 25, if they are working in AI-exposed occupations, or at least that’s what they want to do, it’s about 19% where it would be otherwise. And if you look at the unemployment rate for 20 to 24-year-olds, what do you see? It was 8% in September against 4.2% overall. But there’s other research out there that finds recent graduate unemployment is still within the range it’s held for the last four years. So it’s not a collapse, but a narrowing on-ramp.

What does this matter for housing? Well, the on-ramp is the first-time buyer pipeline. The 26-year-old who doesn’t land the analyst job? Well, they don’t rent a one-bedroom apartment alone. And five years later, they don’t buy the condo. The effect of AI on housing is delayed household formation, and that’s going to take years to show up in the sales data.

Now there’s a second-order effect as well: AI investment. And we know it’s massively capital intensive, not labor intensive. Companies are spending hundreds of billions of dollars on data centers, but at the same time they’re trimming their head counts. Now that’s GDP growth without payroll growth. It supports stock portfolios, also the luxury market, but it thins out the middle of the buyer pool.

Now here’s a twist on rates. A weak labor market normally brings rates down. Now, in September, job numbers came out. Treasury yields actually did fall. So a soft job market partially self-correct mortgage rates. However, the truly dangerous scenario is weak jobs and sticky inflation at the same time, because then the Fed can’t help, and that’s where we are today.

Think about it. Lousy job numbers. CPI at 3.4. Mortgage rates above 7%. This is not a hypothetical. It’s roughly where we are today. So the bottom line is that at least I’d say for the next 12 months, employment is the bigger swing factor to watch. Rates set the price of housing, jobs set the size of the buyer pool.

Hartnett

Atlanta continues to see population and economic growth, yet home sales are declining. Is affordability becoming a ceiling on the market’s growth, or is there enough underlying demand to sustain it?

Gardner

I’d actually push back gently on the premise, because growth is slowing too, and affordability is a big reason why. So you think about the larger metro Atlanta market. That’s what, 29 counties? That added about 62,000 people in the year ending July of 2026. Now that’s the third largest gain in the country. That’s real. But the Atlanta Regional Commission’s newest estimate shows that growth in the 11 counties that make up the core metro area slowing to about 1% this year, down from 1.2% a year before. Now, their explanation was that job growth is flattening. That makes sense as the core metro area has most, well, the most jobs, certainly, and it saw jobs declining about 0.3% between September 2025 and September of 2026. And low housing supply and high prices are acting as a brake on population growth, especially people moving in from other states.

Now, the region’s own planners are telling us affordability isn’t just capping home sales, it’s actually capping the growth that’s supposed to drive home sales, meaning jobs.

Now, on the sales side, in September, the core metro showed closings down about 13% from a year ago, median price up 0.3% to $399,950, I believe. In the greater 29-county market, area closings were down about 11% from a year ago, median price flat at just over $387,000. Now, the number I focus on, and I do, is pending sales. Now the core metro, down 21%, and roughly a quarter of listings had a price cut. 4.7 months of supply, that technically is a balanced market, but one that’s edging towards becoming a buyer’s market. So Atlanta’s problem isn’t a lack of demand, it’s that its competitive advantage as a large, relatively speaking, affordable Sunbelt Metro, it’s eroded. I mean, a $400,000 median sales price at 7.3% isn’t cheap relative to local incomes. And of course, Atlanta competes with Charlotte, Raleigh, Nashville and the Texas metros for the same mobile households.

Is there enough underlying demand? Yes. Over the long run. Atlanta is again the sixth-largest metro. The city itself is growing faster than in prior decades, driven largely by apartment development in the Midtown and also along the Beltline. And the employer base is actually very diversified. So I don’t see a price collapse. I see prices flat to slightly down in nominal terms and actually falling modestly in real terms for the next year or two, at least until incomes catch up. But what would chang my view is if I saw more entry-level supply. Unfortunately, however, building permits remain well below the historic averages. And Atlanta’s growth model actually — it’s interesting — has always been build outward, keep it cheap. When that stops working, demand doesn’t disappear. It just goes somewhere else.

So brokers, I say this is a negotiating market. Price discipline on listings, creative financing on the buy side, that matters far more than chasing volume.

Hartnett

As always, Matthew, thank you for recapping Q3. I look forward to seeing you on screen soon as we go over real estate predictions for 2027.