Kevin Warsh's preferred inflation measure, the Dallas Fed trimmed mean, is running at 2.3% year over year. A year ago it was 2.7%.
At Jackson Hole the chairman said the underlying inflation trend is not coming down. David Rosenberg's point is that the chairman's own indicator says it already has, by 40 basis points.
"So it's just, I think, injected a lot of confusion as far as I'm concerned."
Rosenberg runs Rosenberg Research in Toronto and wrote the Merrill Lynch research note that people used to take off other people's desks; he has been on the dovish side of this argument since the dot plots turned.
The full segment is covered here so you can skip it.
Here are the 6 numbers that matter.
👤 Guest: David Rosenberg, President and Chief Economist at Rosenberg Research in Toronto, previously the author of the Merrill Lynch research note that carried his name
🎙️ Hosts: Tom Keene and Paul Sweeney, who anchor Bloomberg Surveillance
🧩 Other segments: Steven Major, Global Macro Advisor at Tradition; Alicia Levine, CIO of BNY Wealth; and Jay Goldberg of Seaport Research Partners
📰 Published: 16 September 2026 on the Bloomberg Surveillance feed
🟣 Apple Podcasts | 🔗 Episode page | ⏱️ length not available
Key Takeaways
The chairman's own favored inflation gauge is 40 basis points lower than it was a year ago, which is the opposite of what he said at Jackson Hole
The Dallas Fed trimmed mean is at 2.3%, a full percentage point below the core PCE deflator it normally tracks
Oil is not driving the Treasury selloff: crude is where it was on 30 April, when the 10-year yielded under 4.5%
The 30 September revision of the personal consumption expenditures deflator could revise the inflation trend down, and he is surprised the Fed moved ahead of it
Monetary, housing and labor-market inflation are all absent, and he says all three are disinflating
Labor costs are running at 1.4% year over year, against 2% a year ago
Nominal GDP growth decelerates from here, but does not go negative
Six years of fiscal stimulus turn into gridlock after the 3 November midterms
1. The Waller Camp
Keene set the segment up with a framing he took from Dominic Konstam, the previous day's guest: a Warsh School that reads the inflation trend as rising, and a Waller School that reads the same data as shocks that will pass. He asked which one Rosenberg belongs to.
"Are you in the Warsh camp or the Waller camp?" Rosenberg's answer: "Well, what do you think? I'm in the Waller camp." Keene said he was shocked; Rosenberg said he knew he would be.
2. Oil Isn't the Story
Rosenberg's first move was to separate the level of oil from the level of Treasury yields, because the two have moved apart.
"But WTI today is at the same level it was back on April 30th when the 10-year note was actually south of 4.5%."
"Is oil breaking out to new highs? No, you could argue it's at the high end of the range."
What did change was policy expectations, in two steps. The dot plots of 17 June moved from a cut this year to a hike, and the market now prices almost two. "And then Warsh set the bar really low for a rate hike at Jackson Hole."
"So a lot of this has been, as far as the Treasury market's concerned, a reset in Fed expectations."
3. Warsh Against Warsh
The core of Rosenberg's complaint is that the chairman's hawkish turn contradicts the indicator the chairman himself used to point at.
For a long time Warsh's preferred measure was the Dallas Fed trimmed mean. "And that's running at 2.3% year over year."
That sits a full percentage point below the core personal consumption expenditures deflator, which Rosenberg called unusual, because the two historically run right on top of each other.
He was careful to note that Warsh does not really trust any of the inflation measures, but that this was the one he went on record preferring before he turned hawkish.
"And this time last year, it's running at 2.7%."
"And yet then at Jackson Hole, he laments the fact that the underlying inflation trend isn't moving down. Meanwhile, it's 40 basis points lower than it was this time last year."
"So it's just, I think, injected a lot of confusion as far as I'm concerned."
4. The 30 September Reset
Asked whether a first increase since July 2023 — a date he had flagged in his own morning note — would be one and done or the start of two or three, Rosenberg said the question is not answerable yet.
"Well, I think it's going to be situational."
The date he is watching is 30 September, when the Bureau of Economic Analysis carries out a full revisit, reset and revision of the personal consumption expenditures deflator.
What he wants to know is the direction of the rewrite: "And my sense is how much will the trend be revised down?"
"That's why I'm so surprised that Kevin Warsh started sounding so hawkish in light of the fact that September 30th is a very important date in terms of what this newly revised deflator data are going to show."
He also pointed at the chairman's new task forces, one of them on inflation, whose results are not known. Keene asked whether he had been invited onto one. He had not, and Keene said that was not a surprise either.
The program's own caution was that the situation is fluid. It was not long ago that the market was priced for a cut and the Fed's own March dot plot carried one, so the pricing can shift back depending on the economy.
5. Where Disinflation Is
Keene asked him to do what page three of the old Merrill Lynch Rosenberg report did — run through the inflation series and say which ones show trend disinflation. Rosenberg took it series by series.
"Look, the whole thing comes down to the oil price and whatever spillovers we've had into airfares and delivery services." He named financial services and telecom services as other idiosyncratic movers.
On money: M2 is just over 5%, but "money velocity peaked in March and is rolling over." He stopped short of calling it monetary inflation.
On housing: "New home prices are down 1% year over year. Rental rates nationwide are still declining, albeit moderately. There's no inflation out of the housing market."
On labor: "When labor costs are running at 1.4% year over year, this time last year they were 2%."
His appeal to authority was to three former chairs. "This is the sort of stuff you see that deep thinkers... like Greenspan and like Bernanke and like Yellen would be talking about right now."
The question he leaves the hawks with: "So how are you going to get any sustained inflation from oil when you're not seeing monetary, housing, or labor market inflation?"
"In fact, you're seeing disinflation in those three principal areas of the economy."
6. Nominal GDP Decelerates
Keene, having praised the firm's retail-sales note, asked whether nominal GDP comes down from what he has been calling a banana-republic level.
"I expect that nominal GDP growth will be decelerating."
The retail data are the reason he is not more bearish than that. The report was good, but it came off a weak July, and he puts real retail sales below 2% annualized for the third quarter — neither too hot nor too cold, in his phrase.
The bigger swing factor is fiscal, and it arrives after the election. "But keep in mind that one thing that's going to happen post-November the 3rd, after the midterms, is we get fiscal gridlock that's going to replace six years of unrelenting fiscal stimulus, and that's going to cut in aggregate demand."
He still expects underlying inflation to come down, with the usual caveat about what oil prices do.
"So I'm in the camp that thinks that nominal GDP growth is not going to go negative." Forced to choose between acceleration, deceleration and no change, "I think it's going to decelerate."
Bonus Insights
Keene's sign-off was a job offer of sorts: "I think we need you on the task force."
Keene referred to the firm's analysts as the gnomes of Toronto and praised the retail-sales note the firm had published that morning.
The Merrill Lynch anecdote is a piece of research-industry history. Keene described page three of the Rosenberg report, with a column of what he called 400 gazillion inflation parameters, and said people used to steal the note off other people's desks.
Rosenberg's bottom line is that the Fed has tightened into a reset of expectations rather than into actual inflation: the chairman's own preferred gauge has fallen 40 basis points in a year, money, housing and labor are all disinflating, and the revision due on 30 September may show the trend was lower than the committee thought when it moved.
Products, Companies & Tools Mentioned
Rosenberg Research (His firm in Toronto, and the source of the morning note on retail sales and the July 2023 comparison)
The Federal Reserve (Warsh's Jackson Hole framing, the June dot plots and the new task forces, one of them on inflation)
The Bureau of Economic Analysis (Whose 30 September revision of the PCE deflator he says should have been waited for)
The Federal Reserve Bank of Dallas (Publisher of the trimmed mean measure, at 2.3% against 2.7% a year ago)
Merrill Lynch (Where he wrote the report Keene says people stole off desks)
Listen to the full episode:
If this was worth your time, send it to someone who has to have a view on this.
Get the latest market chatter as it happens: