Forgive me, I have been listening to loads of French music recently.
(In fairness, when do I not? But still…)
My current “workday soundtrack” opens with Göttingen by Barbara.
Göttingen helped post-war France see German children as “us, too.”
Barbara’s message?
Having a sense of commonality—of sameness—is crucial.
That’s because shared experiences give us a common language.
And that is the speedway to empathy.
Man, everyone needed extra empathy to move forward in post-war Europe.
It feels like we could use it today, too.
I am telling you this because this is how I think about why I write here—and what I aim to do more generally, starting with clients:
If we can “get on the same page” about our interpretation of what’s happening around us (which does not mean blindly agreeing with each other), we can try to make decent, reasonably informed decisions together.
That starts with me sharing my thoughts as transparently and authentically as I can, so you can cleanly decide what resonates with your best understanding of things.
Hopefully, that allows us to proceed through time and space without losing each other in the process, as the world inevitably makes it its mission to surprise us.
Let’s try it.
There are a million things that feel “up in the air” right now—and worth talking about.
Some feel immediately important…
Like whether we’re getting more oil out of the Middle East any time soon.
Or (and this one is being pulled to the front of the queue recently) what’s going on with interest rates?
What you see in the graph above is simple.
Since 2020, we went from extraordinarily low interest rates across maturities (well in the bottom 10 percent of observations since the turn of the 21st century) to rates today that are stubbornly high, especially longer-dated ones, like the 10-year and 30-year yield.
For reference, the 10-year Treasury yield was at 4.61% and the 30-year Treasury yield at 5.14%, as of earlier this week. (This is an easy one to “get on the same page” about: it’s a market price. It is what it is, as they say.)
There are two main reasons people talk about this at length these days:
The impact of lending rates that feel like a real burden relative to the bliss of borrowing just a few years back.
Whether you’re looking to buy or sell a home, to make a decision on long-dated corporate capital expenditure, or you’re just trying to value private assets, there are a lot of “suspended states” out there.
Do the deal now. Wait to see what happens. Recognize losses. Hope for relief soon. You name it.
The cause behind these high (and rising anew) rates. This is the source of endless speculation.
There is the fear of higher inflation now and later, which is a painfully obvious one to anyone pulling up to a gas station or going to a grocery store.
There is the plain concern that Western governments (including the U.S.) appear hellbent on adding to their bar tab, even during “the good times.”
And then there are fun ones, the clever ones—like the concern that the manic AI buildout is actually causing inflation (they call it “chip-flation,” which is fun to say) and adding massive amounts of debt supply to markets, thus pushing interest rates higher.
Sweet. Aren’t we lucky?
(If you want more on the situation with interest rates, the great John Authers over at Bloomberg is hard to beat.)
This takes us to our main topic for today: the AI economic inflection point.
AI matters because it can change the economics of labor. And when two-thirds of the economy is consumption—so much so that we call people “consumers”—labor income is the collateral under everything else.
Before I say anything else, let me start by quoting Andrea Eisfeldt, who said to me on TREUSSARD TALKS, "We're all at the starting line of this huge race, and nobody really knows where we're going after this."
I am deeply aware that this applies to me.
What follows is a picture of things to look out for along the way, not a set of predictions.
But as we flip the pages of this script, it will help us sort out what kind of an adventure we’re on.
In the meantime, here’s a headline from this week’s news:
Meta Begins 8,000 Global Job Cuts in AI Efficiency Push.
I’ve been thinking about this quite a bit recently (I just gave a presentation on what AI means for markets).
Let’s ground ourselves in a few facts.
There are roughly 18 million people in the U.S. whose job is some form of “office and administrative support.”
Another 3.5 million are “general and operations managers.”
All in, there are roughly 86 million white-collar workers in the United States, totaling approximately $7 trillion in wages.
That’s all according to the U.S. Bureau of Labor Statistics (BLS).
Now, extraordinarily rough estimates—the types of numbers I’d only trust to bar napkins—suggest that around 30% of the core tasks that define what these 86 million people are hired to do could be substantially compressed with AI: the same work, in a fraction of the time.
That maps to somewhere between $2 trillion and $2.5 trillion in wages annually earned by humans around you.
Friends.
Neighbors.
Family members.
The key question here is whether (i) we all get massively more productive—thereby “more than earning” our ongoing keep in the workforce, even to the point of pushing paychecks up—or (ii) CFOs all around us come to the conclusion that removing a couple of trillion dollars of ongoing labor expenses from corporate ledgers is worth a shot.
Either way, for a lot of people, it just got trickier to answer the question: “what would you say you do here?”
I really don’t know which way this is going to go.
I am only going to suggest that what may be an obvious privately rational decision (firing a bunch of expensive workers) could have meaningful societal implications (having a bunch of people all looking for work in a shifting job market, all at once).
We call that sort of thing negative externalities in economics.
What could be driving some of the bad stuff?
As we all know, a large fraction of Americans cannot cover a $400 emergency. The official statistic is 37% of us, according to the Federal Reserve SHED 2024.
Then, there is the open question about how the market would react to a meaningful reduction in 401(k) contributions. As I discussed with Mike Green last year, the so-called passive-investing revolution has been predicated on steady contributions to the stock market from biweekly paychecks. We haven’t seen what happens when unemployment rises and stays high for a while since the GFC.
And then, the big one… Payroll taxes are about 36% of federal revenue, and individual income taxes are about 49%. Most of that individual income tax base is wages and retirement income—roughly one-fifth is business and investment income. In other words, whatever else the U.S. government is, fiscally, it’s largely a “percentage-of-labor-income” business. (Korinek & Lockwood, Brookings Institution, January 2026). That takes us back to the fiscal question. It also might illuminate why the government may be looking at things like tariffs and other forms of “consumption taxes.” Maybe someone somewhere has noticed this funding “dependence on income.”
And then there is the housing stock.
We are finally back to where we were, right around the peak of the mid-2000s.
Home prices are roughly 5 times household income, using national median figures.
The normal way that people think about this graph is in terms of affordability. Homes are expensive for people who work for a living.
When it comes to risk, the normal (and painfully obvious) way to think about these things is that houses are collateral for mortgage loans: If the borrower gets in a tight spot, “the bank” can take the home, sell it, and move on (In the modern world, the bank doesn’t own the mortgage, so there are other players in the middle, but same idea).
That’s all good and well, but it does assume that “the bank” can find a buyer for the foreclosed home at a reasonable bid.
That pool of potential buyers is a function of those people’s income.
Selling a house to someone who’s out of work doesn’t exactly solve the problem.
So, in the aggregate, there is an argument for looking at this graph as a risk metric, not just an affordability metric, if there is a credible risk that unemployment could rise.
At the risk of bluntly connecting dots, that’s one way to map out a nasty scenario for the residential real estate market.
Again, not a prediction. A pre-mortem, i.e., a non-absurd answer to the hypothetical question “how could this go badly?”
Remember, the idea isn’t that bad things can’t happen.
It’s that you can reasonably expect to “live to fight another day” if they do.
And if you’re looking for a bunch of worthwhile “ifs” to explore, the potential implications of AI on the economy and markets are a pretty decent place to spend your time these days.
In the meantime, here is what I would look at, to keep tabs on which way this is going.
Unemployment. That’s obvious, but maybe too obvious. Loads of people may keep their job, even if they’re stagnating in their careers.
Declining vs. rising real wages. Productivity gains could translate into higher labor earnings, after inflation. Loss of bargaining power could lead to faster loss of workers’ purchasing power.
Housing market health. At the intersection of interest rates and labor market conditions, home prices and volume of transactions could reveal the fingerprints of what’s going on in households around the country.
I hope this is helpful.
And if you’d like to see the full deck, it’s available for download below.
I would love to hear your thoughts.
Jonathan
Data Sources:
Wages at Core AI Risk: Annual payroll = BLS headcount × BLS mean annual wage (BLS OEWS May 2024, Table 1). Not a directly published BLS figure. Core-task AI exposure = mean genaiexp_estz_core by SOC 2-digit group from Eisfeldt, Schubert, Taska & Zhang, “Generative AI and Firm Values,” Journal of Finance, forthcoming 2026 (artificialminushuman.com). Scores reflect GPT capabilities assessed March 2023; they measure primary-task AI replaceability, not predicted displacement rates.
Housing data: Harvard Joint Center for Housing Studies, Table W-13 — Ratio of Median Home Price for Existing Home Sales to Median Household Income, 1990–2024.
All figures provided for informational and educational purposes only, as-is. Not investment advice.
Disclaimer: All content here, including but not limited to charts and other media, is for educational purposes only and does not constitute financial advice. Treussard Capital Management LLC is a registered investment adviser. All investments involve risk and loss of principal is possible.
Full disclaimers: https://www.treussard.com/disclosures-and-disclaimers.
