Listen to the audio version of this article (generated by AI).
The Fed holds rates steady… Warsh puts on a masterclass in saying nothing… why AI might keep a rate hike at bay… two more hyperscalers report today… last chance for Luke Lango’s AI MegaDeal Event
I wasn’t expecting new Federal Reserve Chairman Kevin Warsh to offer many details at his press conference today, but even I was impressed at how he was able to use so many words to convey so little meaning.
More on that in a moment…
First, in a 9-to-3 vote that exposed a growing “family fight” within the central bank, the Federal Reserve decided to hold its benchmark rate steady at the target range of 3.5% to 3.75%.
Beth Hammack of Cleveland, Neel Kashkari of Minneapolis, and Lorie Logan of Dallas were the three Fed presidents who voted for a quarter-point rate hike. This matches their stance from late April, when they opposed statement language that implied future cuts.
True to Warsh’s stated philosophy, the official FOMC statement completely omitted the formulaic forward guidance that market participants usually rely on to project future policy shifts.
While it clearly documented the 9-to-3 voting dissent, the statement text itself was a mirror copy of June’s release – purely stating the current economic facts without hand-holding investors. In other words, it’s exactly what Warsh had told us it would be – just the facts.
Turning to Warsh and his press conference, there was one thing he was explicit about – the Fed has not quietly accepted a higher target inflation rate than 2%.
From Warsh:
Five years of high inflation have left a mistaken impression that’s hard to shake, that the Fed’s implicit inflation target was somehow above 2%.
Let me reiterate: There is no soft inflation target, there is no soft implicit target, not on this committee’s watch. There’s only a target and it’s 2%.
Beyond that, Warsh performed a masterclass in saying nothing substantive.
He didn’t offer any real details or forecasts, preferring answers that largely went along the lines of, “We have an amazing team… we agreed on more things than we disagreed… we’re going to get the job done.”
Frankly, all Warsh did was talk about the group talking about inflation.
Here’s an example:
A lot of our focus was on trying to understand and identify underlying inflation dynamics amid (supply) shocks. We take these shocks seriously. There have been a series of them that have been hitting this economy.
We’re not looking through them and saying, ‘oh, they don’t matter.’ What we’re trying to understand is to what extent are these shocks broadening in their effects, broadening in their impact on prices that are quite far removed from it?
Now, despite Warsh’s refusal to provide forward guidance, the market is interpreting the three hawkish dissents as a sign that the Fed is creeping closer to a tightening cycle.
As I write immediately following the press conference, the CME Group’s FedWatch Tool puts the probability of a September rate hike at 53.3%.
I’m not so sure…
Why the bar for a hike is higher than it looks
Hawks who are already considering a September rate hike as a lock are missing something important.
Warsh has set a genuinely high bar for raising rates – and at first glance, that seems inconsistent with a point we’ve spent two years covering in this Digest…
AI is making many things more expensive.
The buildout has created enormous demand for GPUs, memory chips, networking equipment, electricity, cooling systems, transformers, and data centers. We’ve even tracked how rising memory costs are starting to show up in consumer electronics.
Viewed through that lens, AI looks unequivocally inflationary. More spending, more demand, higher prices. That’s certainly one way to read what’s happening.
But Warsh appears to have another…
Don’t forget what Warsh told the Senate
Earlier this month, during a Senate Banking Committee hearing, Sen. Jack Reed pressed Warsh on the AI buildout and the inflationary pressure it may be creating.
Reed noted that some Fed officials have grown increasingly concerned that the wave of AI investment could complicate the inflation outlook.
Warsh agreed that AI investment is likely to push prices higher over the next year. But then he repeated a point he’s made before – it’s easier to measure the effect the AI boom is having on demand than its effect on the economy’s ability to supply goods and services.
He then said:
I don’t view a one-time change in prices as necessarily being inflationary because I think there’s a supply response.
Will it increase measured prices over the course of the next 12 months?
I suspect it will…. Whether that’s inflationary or not, that’s up to the Federal Reserve.
Let’s be clear about what Warsh isn’t saying
He’s not arguing AI won’t raise prices – he explicitly expects it to.
But he’s then redirecting to a critical related question: Do those higher prices mark the start of a sustained inflation problem? Or are they simply the cost of building an economy that becomes far more productive down the road?
That distinction runs through nearly everything Warsh has said about AI over the past year.
Writing in The Wall Street Journal last November, he argued investors were badly underestimating AI’s economic impact:
AI will be a significant disinflationary force, increasing productivity and bolstering American competitiveness.
Productivity improvements should drive significant increases in real take-home wages.
He explained why productivity carries so much weight in his thinking:
A 1-percentage-point increase in annual productivity growth would double standards of living within a single generation.
That’s not the language of someone fixated on the latest inflation print. It’s the language of someone thinking in terms of how technological revolutions reshape economies over decades.
Legendary investor 91抖阴, editor of , made a similar point recently, coming at it as an investor rather than a policymaker.
Here’s Louis:
Remember, AI is not inflationary. It’s temporarily inflationary in memory. That’s a freak thing because of a bottleneck.
But other than that, no. AI is going to be creating incredible productivity gains, lots of GDP growth, and that’s that…
The Fed’s not increasing rates. So, whatever the talking heads on TV have been telling you, that’s not happening.
Different vocabulary, same underlying read: strip out the one-off bottleneck, and what’s left isn’t inflation, it’s productivity.
What this does – and doesn’t – mean for rates
The takeaway here shouldn’t be “Warsh is a camouflaged dove who will be cutting.”
During Warsh’s confirmation hearing, Sen. John Kennedy suggested the new Fed Chair’s optimism about AI implied a more dovish approach to monetary policy.
Warsh rejected that framing outright:
That is not how I would characterize the story on AI.
The more accurate read is that Warsh is raising the bar for what counts as inflation serious enough to justify a hike.
That’s not nothing for AI investors, who’ve watched their positions get hammered over the past two years each time Wall Street has gotten spooked by the threat of higher-for-longer rates.
Growth stocks live and die on the discount rate, and few sectors are more rate-sensitive right now than AI, given how much of its valuation is built on cash flows still years away.
So, if Warsh needs to see inflation broaden, or productivity disappoint, or expectations come unanchored before he treats AI-driven price increases as the “bad” kind of inflation, that’s a materially higher bar than the market has been pricing in.
Of course, Warsh is just one vote within the FOMC. Today showed us that three other members are already willing to cast the opposite vote.
Will Microsoft and Meta follow Alphabet’s lead?
By the time you read this, we’ll likely already know.
Microsoft (MSFT) and Meta (META) both report earnings after the bell today, and their capex guidance should be released along with their earnings.
Last week, Alphabet (GOOGL) maintained – and actually raised – its capex plans. Wall Street punished it, sending it almost 7% lower the next day.
The question tonight is whether Microsoft and Meta get treated the same way.
Will they hold or raise their own AI infrastructure spending? And if they do, does Wall Street punish them the way it punished Alphabet?
Even if Wall Street punishes them, will it finally reward the smaller AI capex companies on the receiving end of those hundreds of billions of dollars?
We’ll report back tomorrow.
Your final reminder about tomorrow’s AI MegaDeal Event
As we just noted, tonight, Microsoft and Meta will report how much more they’re willing to spend chasing AI dominance.
Most investors watching those numbers already know the standard playbook: buy the companies on the receiving end of all that capex – the chipmakers, the data center operators, the picks-and-shovels names getting paid by Microsoft and Meta’s spending.
Luke doesn’t disagree with that trade. As we’ve been tracking here in the Digest, he’s been looking for the technical bottom of the recent AI infrastructure selloff, looking for the “back up the truck” moment.
But Luke will also agree that this is just half the picture – because some of that hyperscaler capex doesn’t get billed at all. It gets paid out in acquisitions.
Google didn’t just buy chips for its AI research – it bought DeepMind outright. Meta didn’t just pay a vendor for training data – it wrote a $14.3 billion check for a stake in Scale AI. Microsoft didn’t just rent compute to a partner – it committed $23 billion to OpenAI back when almost no one outside Silicon Valley had heard of the name.
That’s a different way to profit from the same trillion-dollar spending spree: not selling the giants what they need, but becoming the thing that the giants decide they need to own.
Luke’s research concludes this pattern is about to repeat, at a larger scale than any of those three examples. And tomorrow at 1:00 pm Eastern, at his AI MegaDeal Event, .
Register now, and you’ll also have the option to join Luke’s VIP text list. If you do, he’ll send over his bonus report, , free.
So, last call! and we’ll see you tomorrow.
Have a good evening,
Jeff Remsburg
