There’s this theory that AI adoption will be “massively disinflationary.”
In that as the technology improves, we will see big production gains that lower the costs of everything.
Simply put, supply will greatly outstrip demand, making things cheaper.
This sounds like a good outcome for inflation and an environment that can invite lower interest rates, including cheaper mortgage rates.
But skeptics argue that AI could lead to structural unemployment that rate cuts can’t fix, meaning lowering rates would only help the wealthy and potentially overheat the economy in the process.
AI Is Expected to Lead to a Positive Supply Shock
An article came out yesterday in the Financial Times with asset manager Mike Hunstad arguing that AI could lead to “one of the biggest positive supply shocks we’ve ever seen.”
It sounds great on paper. AI makes everything less expensive and more abundant, leading to growth in the economy without the nagging inflation.
What’s not to like? That would mean we could also lower interest rates, something soon-to-be Fed chair Kevin Warsh has argued as well.
It seems to parallel the late 90s tech boom driven by the advent of the Internet, which allowed then-Fed chair Alan Greenspan to keep rates steady instead of raising them as growth exploded.
Back then, the federal funds rate doubled from 1994 to 1995 to cool the economy, but subsequently the Fed was able to hold rates flat because the Internet acted as a positive supply shock.
However, we all know that in the late 1990s and early 2000s, the dotcom bubble popped.
Partially because the Fed began raising rates again to cool an overheating stock market and excess demand, driven in part by the wealth effect of the Internet.
You can see some parallels today with AI stocks and wild valuations, but it pales in comparison to the dotcom era.
The other major difference is AI seems to be taking jobs away, while the Internet created jobs.
While there is hope that it eventually leads to job gains, as the Internet did, it might get worse before it gets better. And it could take quite a bit of time.
How AI May Affect Mortgage Rates
What this all means is AI may initially displace a lot of workers and lead to a positive supply shock. But it’s job cuts first, production gains later. Potentially way later.
This means higher unemployment, but also higher growth. And what’s unique this time is the Fed may think that cutting rates won’t actually induce new hiring if there aren’t skilled workers in the nascent AI space.
As such, you might have a situation of wait-and-see, which the Fed has kind of been doing for a while now after a large series of hikes followed by some cuts.
The predicament is that cutting rates might just exacerbate that K-shaped economy where the wealthy get even wealthier, and the low- and middle-class get worse off.
So standing pat or cutting a little bit more might be the move, as opposed to massive rate cuts.
If they cut too aggressively, it may lead to even more spending in the AI/tech space and more of that dotcom-era exuberance.
With valuations already high, whether it’s a house or a tech stock, this won’t be the desired outcome from the Fed.
They can’t cut their way into more tech jobs if people don’t have the necessary skills, at least not quickly. Nor can they risk inflation surging higher again.
This all kind of leads to a firm policy stance, though given Warsh was hired by Trump, who hasn’t been shy about wanting rate cuts, more cuts are certainly possible.
The end result is maybe slightly lower mortgage rates versus current levels.
We had a 30-year fixed sub-6% recently, before the war with Iran broke out, and perhaps we get back there late this year if that conflict subsides.
After that, it wouldn’t be shocking to see mortgage rates settle around those levels, though perhaps reach deeper into the 5s.
That would be just fine for the housing market, creating more demand without fueling a speculative frenzy again.
How the Timing Could Play Out
- Rates flat/down slightly in next 12 months as unemployment rises and war tensions ease, despite continued growth and strong AI investment
- Rates potentially up in mid-to-late 2027 if AI hype creates excessive demand for data centers, chips, energy, etc., putting upward pressure on inflation
- Eventually we see a typical market correction due to the exuberance, albeit not as bad as dotcom
- Downturn leads to a series of rate cuts and by extension mortgage rates come down further
- But over longer time horizon AI adoption creates an environment of more jobs and low inflation similar to the established Internet era
(photo: Saundra Castaneda)
Before creating this site, I worked as an account executive for a wholesale mortgage lender in Los Angeles. My hands-on experience in the early 2000s inspired me to begin writing about mortgages 19 years ago to help prospective (and existing) home buyers better navigate the home loan process. Follow me on X for hot takes.