Key Takeaways

Rosenberg says the economy is far weaker than record markets suggest. He argues "the economy is more fragile than people think," pointing to four straight quarters of real GDP near 1.5%, below the Fed's own 2% potential estimate.

He expects the Fed's next move to be a cut. Against the consensus, Rosenberg says a September hike is off the table and "their next move will be to cut rates," once the Fed shifts its focus from inflation to a weakening labor market.

Strong S&P earnings mask a soft economy. He notes broad national-account profits are running at about half the S&P 500's pace, and that ex-AI and ex-energy, corporate profits are growing less than 5%, while real household incomes are flat to slightly negative.

AI spending is doing the heavy lifting, at the old economy's expense. Rosenberg says "AI is sucking resources out of the old economy," with AI capex booming while non-AI investment is modestly negative.

The consumer is running on the stock market, not income. With real incomes flat, spending has held up only because the savings rate has roughly halved, which he flags as a key vulnerability if stocks stop rising.

Key Moments

00:21 - "Bear in the Bull Ring": Rosenberg's reputation and new book The self-described perma-bear who says he has never shorted a stock.

04:12 - Will the Federal Reserve cut rates next? Why Rosenberg thinks the next move is a cut, not a hike.

05:46 - Weak GDP growth and falling inflation Below-potential growth building slack and disinflation.

08:23 - Why the U.S. economy is more fragile than it looks The gap between the headlines and the trend.

10:08 - Treasury yields, the bond market and the term premium Why he says the yield rise is about uncertainty, not inflation.

13:53 - Why are S&P 500 earnings still so strong? Accounting, circular financing and the profit-income divide.

16:51 - AI spending, corporate profits and recession risk How a bifurcated economy shows up in earnings.

22:30 - The stock-market wealth effect and consumer spending The falling savings rate propping up spending.

25:34 - Why rising real rates are a threat The shock of higher real rates as growth slows.

Is the US Economy in a Recession? Why David Rosenberg Says It's More Fragile Than It Looks

A recession is usually defined as a broad, sustained decline in economic activity, and by that strict measure the US is not in one. But economist David Rosenberg, founder and president of Rosenberg Research and author of the new book Bear in the Bull Ring, told Wealthion in August 2026 that the distinction flatters the reality: beneath record markets and strong headline earnings, he argues, the economy is running below potential, real incomes are flat, and the strength is concentrated almost entirely in AI. He describes himself, with a smile, as "the first perma bear of all time who's never shorted a stock in his life," and this is his contrarian case, not investment advice.

Is the US economy in a recession?

Not officially, but Rosenberg argues it is far weaker than it looks. He points to three quarters of real GDP growth averaging below 1.5%, expects a fourth, and notes the Fed's own estimate of potential growth is 2%. When demand runs below supply for several quarters, he says, slack builds. His sharper point is on incomes: with no employment growth over the past 12 months and real disposable personal income fractionally negative year over year, "in any other time," he says, people would be asking when the recession starts. So while the label does not apply, "the economy is more fragile than people think."

Will the Fed cut rates?

Rosenberg's controversial call is that the next move is a cut. He says a September hike is off the table, with market-implied odds down to around 40% from 80% before the payrolls report, and that the Fed will then get boxed in ahead of the midterms and is unlikely to move in December. Right now the front end of the curve still prices in about one more hike, which he thinks will come out, making two- and three-year notes attractive. Looking further out, he says, "I'm actually leaning to the view that... their next move will be to cut rates," once the Fed shifts its attention from an inflation mandate it has missed for five years back to a weakening labor market. "That's when we get a rate cut."

Why does he think inflation is heading lower?

Because below-potential growth is building slack. Rosenberg argues aggregate demand running under aggregate supply for multiple quarters triggers disinflation, and he sees confirming evidence in the Fed's Beige Book: unlike the 2021 to 2023 period, when $2 trillion of stimulus meant consumers accepted price increases, businesses are now meeting resistance and even rolling back prices in sectors like hotels. He expects inflation to surprise to the downside for the next several months, with the economy still growing but below potential.

What is really driving long-term yields?

Not inflation expectations, in Rosenberg's view. He argues the rise in 10- and 30-year yields since late June has been driven almost entirely by the term premium and real rates, while inflation expectations have been range-bound for four years. The term premium, he says, reflects uncertainty: booming AI-related credit demand, a confused Iran policy, and a Treasury intervention on behalf of the yen (selling euros through the Fed's repo facility) that he thought unhelpful. He also faults the Fed's communication for adding to the uncertainty. The front end is all about Fed expectations; the long end is about confidence and policy uncertainty. For related bond-market views, see George Goncalves on stealth tightening and Michael Green on the bond market hiding a banking crisis.

If the economy is weak, why are S&P 500 earnings so strong?

Rosenberg gives several reasons. First, national-account profits, which cover all companies, are running at roughly half the S&P 500's 28% pace, because the index is not the whole economy. Second, corporate income as a share of gross domestic income is at an all-time high (which he likens to a dot-com stock in 1999), while labor's share is at an all-time low (reminiscent of 2008), so profits have been "sapping the momentum out of all the other forms of income." Third, reported earnings have been flattered by AI-related circular financing, full expensing from recent tax legislation, and tariff remittances. Strip out AI and energy, he says, and corporate profits are growing less than 5% year over year. This connects to concerns about the AI spending boom cracking.

Is the AI boom masking a weak economy?

That is Rosenberg's central claim. He notes AI capex is about 7% of GDP and the fastest-growing part, up roughly 16% in real terms over the past year, but that the other 93%, including housing and commercial construction, is weak, and that non-AI capex is modestly negative. In his words, "AI is sucking resources out of the old economy," which still accounts for about half of capital spending. He contrasts this with the late-1990s internet buildout, when GDP grew around 4% and the whole economy moved in tandem; today, he says, growth is stuck near 1.5% and the economy is sharply K-shaped, across income, asset ownership, generations, and AI versus everything else. As he puts it, "AI is booming. It's one part of the economy," even though the coverage makes it sound like 93%.

What happens if the stock market stops rising?

This is the vulnerability Rosenberg keeps returning to. He explains the puzzle of real incomes being fractionally negative while real consumer spending is up about 2%: the gap is bridged by a falling savings rate, roughly halved from about 5% a year ago to around 2.5%. If households spent only their income, he says, spending growth would be near zero and the conversation would be about a recession starting, since the consumer is many times larger than AI. His question is what happens if the stock market stops going up or the savings rate stops going down, a classic mean reversion in which spending converges back to income. It need not produce negative GDP, he says, but it would widen the gap between supply and demand.

Why are rising real rates a threat?

Because they hit an already-slowing economy. Rosenberg warns that even as inflation expectations decline, bond yields may not fall if the risk premium does not ease. The danger, he says, is "the real interest rate going up... at a time when real economic growth is going down," which acts like an exogenous negative shock. For the opposite, more optimistic read on the economy's resilience, see Chris Galipeau on why the U.S. economy keeps defying expectations.

What Investors Should Watch

• Real GDP versus 2% potential: the sub-1.5% trend Rosenberg says is building slack and disinflation.
• The personal savings rate: which he calls the most important behavioral gauge in the national accounts, now propping up spending.
• The term premium and real yields: rising for reasons of uncertainty rather than inflation expectations.
• Ex-AI, ex-energy corporate profits: running below 5%, his measure of underlying earnings.
• Labor-market revisions and the labor force: the data he expects to pull the Fed toward a cut.

FAQ

Is the US economy in a recession? Not by the official definition, but David Rosenberg argues it is much weaker than it looks. He points to four quarters of real GDP near 1.5%, below the Fed's 2% potential estimate, no employment growth over 12 months, and flat-to-negative real incomes, saying that in any other era people would be asking when a recession starts.

Will the Fed cut rates? Rosenberg thinks so. He says a September hike is off the table and that the Fed's next move will be a cut, driven by data, once it shifts focus from inflation to a weakening labor market. He views the front end of the curve, especially two- and three-year notes, as attractive.

Why are S&P 500 earnings so strong if the economy is weak? Rosenberg attributes it to AI, accounting and concentration: broad profits are running at half the S&P's pace, earnings are flattered by circular financing, full expensing and tariff remittances, and ex-AI and ex-energy, profits grow under 5%.

Is AI masking a weak economy? In Rosenberg's view, yes. AI capex is booming at about 7% of GDP, but the other 93%, including housing and construction, is weak, and non-AI capex is modestly negative. He says AI is drawing resources out of the old economy.

What happens if stocks stop rising? Rosenberg warns the consumer is spending beyond flat incomes only because the savings rate has halved. If stocks stop rising or the savings rate stops falling, spending could mean-revert toward income, widening the gap between demand and supply.

Which expert and interview does this article reference? This article draws on Wealthion's interview with David Rosenberg, founder and president of Rosenberg Research: "The Economy Is More Fragile Than You Think."

Full Transcript (cleaned)

Speakers: Maggie Lake (Wealthion host) and David Rosenberg (founder and president, Rosenberg Research). ASR errors corrected (names, terms) and filler removed; meaning preserved. A brief partisan aside has been softened to its economic point.

David Rosenberg (cold open): I'm actually leaning to the view that their next move will be to cut rates, not to raise rates. I think the economy is more fragile than people think. AI is booming. It's one part of the economy. What happens if the stock market stops going up? What happens if the savings rate stops going down?

Maggie Lake: Hello and welcome to Wealthion. I'm Maggie Lake. Joining me today to discuss the outlook for markets is David Rosenberg, founder and president of Rosenberg Research. Hi David, great to have you back.

David Rosenberg: Thanks, Maggie. Always great to be on with you.

Maggie Lake: Many viewers know you, but for newcomers, you have a reputation as a bit of a bear, and you're leaning into it with a book coming out, Bear in the Bull Ring.

David Rosenberg: It's really a survival kit for Wall Street, a book about my 40 years in the financial industry as an economic and financial adviser, a collection of the data points and experience that shaped my views and convictions, and how that influences not just how I interpret the economic data and invest around it, but how to live life. It's called Bear in the Bull Ring because of my reputation. It's interesting: I was hired by Gluskin Sheff, a household name in Canada. Ira Gluskin was like Canada's Warren Buffett, an iconic long-only equity manager who would never have hired a perma-bear to sit with his portfolio managers, but they'd followed my research for years. People who really know me, my 2,300 clients, know it's just a label. I might push back when it gets extreme, but there's no perma-anything in this business, because "perma" is short for permanent, and ours is an industry of fast change. Still, I've embraced the perma-bear label, because the worst thing you can be in my business is ignored. And I'm the first perma-bear of all time who's never shorted a stock in his life.

Maggie Lake: There's a whole industry built around being constantly bullish on equities, and that doesn't quite fit anymore, so we love catching up with someone willing to push back against consensus. Congrats on the book. This was a critical data week before September. CPI came in tame. Did that take a Fed rate hike off the table?

David Rosenberg: I think it took September off the table. Fed futures odds are down to about 40%; before the payroll number they were around 80%, then 60, then a toss-up before the CPI, now 40. It'd be unusual for the Fed to hike unless it wanted to shock the market, and then it gets stuck ahead of the midterms, and it's rare to move in December when liquidity is thin and everyone's closing their books. So there's a real opportunity at the front end of the curve, which is still priced for at least one more hike; I think that comes out, and the two- and three-year notes are a very good place to be. Looking beyond that, and this is controversial, I'm leaning to the view that their next move will be to cut rates, not raise them, and it'll be driven by the data. From a supply-demand perspective, we've had three quarters in a row of real GDP growth averaging below 1.5%, and I think Q3 will make it four. People say it's not negative, so it's not a recession, but the Fed's own estimate of potential growth is 2%. When aggregate demand runs below aggregate supply for multiple quarters, it builds slack and triggers disinflation. The latest Beige Book showed a lot of consumer resistance to higher prices, which you didn't see in 2021 to 2023 when there was $2 trillion of stimulus in household pocketbooks; back then consumers accepted price increases. Now businesses in some sectors, like hotels and motels, are rolling back prior increases. So inflation should surprise to the downside for the next several months, with the economy growing but below potential.

Maggie Lake: The employment number wasn't clean, some said.

David Rosenberg: It's funny that when bears adjust the data, we're accused of data mining, but when bulls exclude things and say state and local government doesn't matter, well, tell that to those employees. Even with all the adjustments, the number came in below consensus, and there's no denying the downward revisions month after month. So the economy is more fragile than people think. I was surprised how strident Kevin Warsh was; how can you call the economy solid or resilient just because you hear that on media and in Wall Street research? In 40 years I've never heard a 1.4% real GDP trend described as solid. It's running below potential, which is what matters for inflation. The Fed had no smoking gun today, and I don't think it will get one. The question is whether the labor force keeps contracting; at some point it stabilizes at a lower level, and then the unemployment rate hooks up. The Fed isn't paying much attention to the labor market right now; it's focused on an inflation mandate it's missed for five years, which is looking through the rearview mirror. My assumption is that over the next several months and quarters it shifts attention back to the labor market, and that's when we get a rate cut.

Maggie Lake: The bond market was also nervous about inflation, with the long end moving up. Are they behind the curve too?

David Rosenberg: When you look at 10s and 30s, the run-up in yields since late June has been driven by the term premium, by real interest rates. Inflation expectations have been range-bound for the better part of four years. So it hasn't been an inflation-expectations story, even if that's what you hear on the financial news; it's been a term-premium story. You could argue that reflects booming AI-related credit demand, or the confused US-Iran situation. I didn't think the Treasury Secretary did anyone any favors intervening on behalf of the yen through the back door of the Fed's repo facility by selling euros. And I found the Fed's communication confusing: when Kevin Warsh was asked why the dissenters wanted to raise rates, he said he'd let them speak for themselves, which is fair, but he didn't explain why he refrained from raising, and didn't answer what he was looking at. So the term premium is about uncertainty, and more uncertainty feeds the longer end. The front end is all about the Fed; the two- and three-year notes aren't complicated, they're about expectations of Fed policy. The longer end is about confidence and policy uncertainty. What's undermining the back end is inflation uncertainty, and fiscal uncertainty, not core inflation itself, which isn't plunging but is going in the right direction. Remember, when Paul Volcker left office in 1987 there was no 2% target, and core inflation was about 4%; we're below that now. So at the low end of the curve it's not an inflation story, it's uncertainty, and uncertainty correlates with the risk premium in bonds.

Maggie Lake: If the economy is tepid and growing under potential, why are earnings so strong?

David Rosenberg: A couple of ways to answer. There are four sources of income in the national accounts: personal income, corporate income, rents and dividends. Leave rents and dividends aside. National-account profits are not S&P 500 profits; they're not weak, but they're running about half the 28% you're seeing in the S&P 500, because they include all companies, not just the fastest-growing 500. Equity investors pay for corporate earnings, not personal income, but personal income ultimately supports the 70% of the economy that is the consumer, and real disposable personal income is fractionally negative year over year, about minus 0.1%. How can you call the economy strong with no employment growth and no real income growth over 12 months? Corporate earnings have been sapping the momentum out of all the other forms of income, to the point where corporate income as a share of gross domestic income is at an all-time high, which looks like a dot-com stock from 1999, while labor's income share is at an all-time low, like a 2008 bear market. Income and wealth inequality is at a three-standard-deviation extreme, which I think helps explain the strong showing of more populist candidates in the primaries; capitalism and democracy always dance precariously on a pin. And even within the S&P 500, a lot of the earnings are AI-driven, with circular financing: companies own each other's stock, and there are financing arrangements. The recent tax bill allowed full-year expensing, so when company A sells to company B, company A books the revenue immediately while company B expenses it at once, which is a boost. There were also tariff remittances that added a couple of points. So it's not all about what the economy did. Nominal GDP is around 5% to 6%, yet profits are growing several times that pace, and there are reasons for that. It's also lopsided: the energy sector's earnings skyrocketed because of the US-Iran tension and the uncertainty over the Strait of Hormuz. Strip out the AI effect and energy, and corporate profits are running less than 5% year over year. People call that data mining; I call it analysis. The lopsided nature of earnings mirrors the lopsided economy.

Maggie Lake: Kevin Warsh pointed to AI capex as evidence of resilience.

David Rosenberg: AI capex is about 7% of GDP and the fastest-growing part, up about 16% in real terms over the past year, but there's another 93% of GDP. Is housing resilient? I don't think so. Commercial construction? No. Ex-AI, capex is modestly negative, in what we used to call the old-economy industrial side. So AI is sucking resources out of the old economy, which still represents about half of capital spending. AI is booming; it's one part of the economy, though you'd think it was 93% of GDP from the coverage, when it's 7%. Back in the late-1990s internet and telecom buildout, GDP was growing around 4%, not 1.5%, and even though tech capex led, everything moved in tandem. Nobody talked about a K-shaped economy in 1998 or 1999. We do today, and the K isn't just high-end versus low-end consumer; it's AI capex versus everything else, asset owners versus non-owners, and it's generational, with young people bearing the brunt of a no-hire, no-fire labor market.

Maggie Lake: Talk about the wealth effect and the consumer.

David Rosenberg: The equity wealth effect at the high end, and parts of the middle, creates a bizarre situation. Real disposable income growth is fractionally negative year over year, yet real consumer spending is up more than 2%. How do you get a two-point gap between spending and income? If the consumer had to live within its means, spending growth would be zero over the past 12 months, not plus two, and we'd probably be debating whether a recession is starting, because the consumer is bigger than AI by a factor of 10. The link runs from AI to the stock market to the equity wealth effect to spending, and it shows up in the savings rate. The savings rate is the most important behavioral gauge in the national accounts, that decision by 130 million households about how much to save versus spend. A year ago it was close to 5%; it's been cut in half, to about 2.5%. That arithmetic explains the wedge between zero income growth and 2% spending growth. But ask yourself: what happens if the stock market stops going up, or the savings rate stops going down, and we get a classic Bob Farrell rule-number-one mean reversion, with spending converging on income? That's a plausible base case. It doesn't necessarily produce negative GDP prints, but it reinforces the gap between aggregate supply and demand. I sound like my friend Lacy Hunt reaching for the blackboard; he may have a different view on the bond market than I do. It could be that bond yields don't go down if the risk premium doesn't ease, so you could have inflation expectations falling and yields not doing much. And it's a problem to have the real interest rate going up at a time when real economic growth is going down. That's like an exogenous negative shock on the economy.

Maggie Lake: Fantastic, David. This is why we like to talk to the bear, not perma, but sometimes bear, always open-minded and happy to be contrarian. Congrats on the book and the upcoming ETF; come back when it launches.

David Rosenberg: Thanks a lot, Maggie. Enjoy the rest of the summer.