A Phil Hall Op-Ed: So, I Asked ChatGPT to Predict 2027’s Housing Market

Share this article!

A Phil Hall Op-Ed: Did you ever get the feeling that artificial intelligence is often more effective than genuine human intelligence? This explains why too many people are relying on AI tools to get input on everything from writing books to romantic advice.

Thus, I decided to converse with my ChatGPT account about what next year’s housing market is going to look like. After all, can an AI tool accurately predict the trends, challenges and anomalies that will shape housing in 2027?

Allow me to share my Q&A with ChatGPT on the subject, which I admit up front had a lot more A than Q. Here is the exchange in its unedited entirety:

Q: What is your forecast for the US housing market in 2027?

A: If you mean “What do you think the housing market will look like in 2027?”, my base case is: better than 2026, but not a boom. I’d expect a market that is more active, somewhat more balanced, and still frustratingly expensive.

My 2027 forecast

• Home prices: roughly +2% to +4% nationally. Fannie Mae’s current expert panel expects about 2.2% appreciation in 2027, while J.P. Morgan is projecting about 3%.
• Mortgage rates: probably around 5.5%–6.5%, rather than a return to the 3%–4% rates of the 2010s. Fannie Mae’s current forecast has the 30-year mortgage averaging about 5.6% in 2027, although today’s rate environment makes that forecast particularly uncertain.
• Sales: I think this is where 2027 could show the biggest improvement. Fannie Mae’s latest forecast calls for total home sales to rise substantially in 2027, with existing-home sales reaching roughly 4.4 million.
• Inventory: likely better than the extraordinarily tight market of recent years, but not enough new construction to create a national glut.
• Affordability: still a major problem. Even if rates fall, prices aren’t likely to fall enough to make homes suddenly inexpensive.

The interesting thing about 2027 is that the market could become much more active without becoming much cheaper.

Right now, the combination of high prices and mortgage rates around 6½%–7% is suppressing transactions. Existing-home sales fell to a 14-month low in August 2026, even as inventory climbed to its highest level since 2019.

If mortgage rates gradually move toward the high-5s or low-6s in 2027, I think a lot of buyers who have been sitting on the sidelines will come back. At the same time, millions of existing homeowners will remain reluctant to sell because they’re sitting on much cheaper mortgages. That should prevent inventory from exploding.

The biggest wildcard

Interest rates. This is the variable I’d watch more closely than home prices.

If rates get down around 5.5%, 2027 could be a surprisingly good year for transaction volume. If they remain around 6.75%–7%, I’d expect another year of sluggish sales and modest price appreciation.

And there’s already considerable uncertainty: as of early September 2026, the 30-year mortgage rate was 6.85%, while expectations for Federal Reserve rate cuts have been pushed into 2027.

My shorthand prediction:
2027 = more houses for sale + more buyers + more transactions + modestly higher prices + still-poor affordability.

I would not expect either a 2008-style crash or another 2021-style frenzy unless the economy experiences a major shock.

If you’re thinking about this specifically from the perspective of real estate and mortgage lending, 2027 could actually be considerably more interesting than the headline national housing numbers suggest.

Q: Really? Can you tell me more?

Yes. And I think there is a more interesting story in 2027 than simply “prices go up 2%.”

The key thing: 2027 could be a normalization year

The housing market is currently in an unusual position. In August 2026, existing-home sales fell to a 3.98 million annual pace, while inventory reached 1.62 million homes, the highest level since late 2019. Yet the median price was still up 1.6% year over year, to $429,100.

That’s an important combination:

Weak sales + rising inventory + still-rising prices.

It suggests that the fundamental problem isn’t necessarily that people don’t want houses. It’s that the price of financing is keeping buyers and sellers from transacting.

I think 2027 could begin to break that logjam.

1. Mortgage rates will determine almost everything

I’m less optimistic than I was in my first answer about rates getting all the way down into the mid-5s.

Current forecasts have rates remaining above 6% through much of 2027. Fannie Mae’s more recent outlook has the 30-year mortgage around the low-6% range at the end of 2027.

That isn’t cheap.

But psychologically, there’s a huge difference between:

6.8% → 6.1%

and

6.8% → 4.5%.

The first could revive the market. The second would create another housing boom.

I think the first scenario is much more likely.

2. Sellers may finally become less powerful

This may be the biggest change.

For years, sellers have had an enormous advantage because inventory was so scarce. That’s beginning to change. August inventory was up nearly 6% from a year earlier, and the market had reached 4.9 months of supply.

If that trend continues into 2027, buyers will have something they haven’t had much of since before COVID:

choices.

That doesn’t necessarily mean falling prices.

Instead, I expect more:

• price reductions
• seller concessions
• mortgage-rate buydowns
• inspection contingencies
• longer marketing times
• negotiations over closing costs

In other words, the effective price of a house could fall even while the published median price rises.

That’s an important distinction.

3. I don’t expect a national housing crash

There are certainly markets where prices could decline substantially. But nationally, I don’t see the ingredients for a 2008-style collapse.

The biggest reason is that today’s homeowner is generally in much better mortgage shape than the homeowner of 2006–08.

There’s also still a structural shortage of housing in many parts of the country.

So my base case is:

Prices don’t collapse. The market simply becomes less irrational.

That’s actually a fairly big change.

4. New construction could become increasingly important

Builders have something individual homeowners don’t: they can manufacture inventory.

If existing-home sellers remain reluctant to give up their low-rate mortgages, builders can capture buyers by offering incentives.

I expect builders in 2027 to become increasingly aggressive with:

• mortgage-rate buydowns
• closing-cost assistance
• upgrades
• smaller homes
• townhomes
• entry-level products

Fannie Mae’s forecast has single-family housing starts increasing in 2027 after a decline in 2026.

That could make new construction surprisingly competitive with existing homes in some markets.

And here’s the part I think is particularly interesting for mortgage lending

The mortgage business could have a much better 2027 even without a huge housing boom.

Think about what has happened:

2022–26:
High rates → fewer refinances → fewer purchases → extremely low origination volume.

If rates merely drift down toward 6%, you could get a substantial increase in purchase activity.

And there’s an enormous amount of latent housing demand.

People who postponed buying in 2022, 2023, 2024, 2025 and 2026 don’t necessarily disappear. Some eventually have children, get married, change jobs, receive inheritances, move for work, or simply decide they’re tired of renting.

So you could have:

modest price appreciation + falling rates + increasing transactions

which is actually a pretty favorable environment for lenders.

But there will be a geographic divide

I wouldn’t make a national prediction and assume it applies everywhere.

Some markets could have a very good 2027.

Others could be flat or declining.

And Connecticut/New England could be particularly interesting because housing supply remains relatively constrained. Connecticut’s market is currently considerably tighter than the national market; the state’s median sale price was about $450,000 during the first eight months of 2026, and the Hartford area was still experiencing declining inventory.

So, a national +2% housing market could conceivably mean:

• Connecticut: +3–5%
• some Northeast markets: +4–6%
• oversupplied Sun Belt markets: -2% to +2%
• particularly overheated markets: potentially considerably worse

That’s why I would be cautious about saying “the housing market” will do one particular thing in 2027.

My rough scenario analysis

My money is on the base case.

And there’s one thing I’d watch above everything else over the next 12 months: the 10-year Treasury yield. Mortgage rates aren’t controlled directly by the Fed; longer-term Treasury yields and mortgage-market spreads matter enormously. The recent jump in the 10-year Treasury toward 4.9% is one reason mortgage rates have moved back toward 7%.

If the 10-year settles substantially lower during 2027, the housing market could surprise on the upside.

If it stays around 5%, 2027 could turn out to be another year of “almost recovery.”

And that distinction matters enormously for banks, mortgage companies and real-estate lenders, because transaction volume—not just home prices—is what ultimately drives much of their business.

# # #

Post-script: Do you think ChatGPT knows what it is predicting? Please share your thoughts in the comments section below.

Phil Hall is editor of Weekly Real Estate News. He can be reached at [email protected].