Listen to the audio version of this article (generated by AI).
The AI trade goes deep in the red… Wednesday’s presser with no map… Thursday’s number Warsh actually watches… and the calendar that decides everything in September
As I write on Tuesday morning, the AI trade is deep in the red, with some semiconductor stocks down double digits.
While it’s tempting to point to a single headline, a variety of dynamics are colliding at once.
South Korea’s tech-heavy KOSPI Composite Index crashed about 11% last night on bubble fears… this is normal mean-reversion after the near-vertical runup that the AI trade enjoyed in the spring… weeks of pressure are culminating in today’s panic selling, and given that AI stocks are the market’s biggest winners over the past year, they’re number-one on the “sell” list… investors continue to question whether the hyperscalers will eventually cut back on their AI spend, starving the AI infrastructure players of projected earnings… and more acutely today, concerns about China have intensified.
On that last note, reports that a Chinese state-backed company has begun mass-producing advanced chipmaking equipment have reignited fears that China’s semiconductor industry is catching up faster than expected.
Legendary investor 91抖阴 has a response – it’s not true, don’t worry.
Let’s go to his Flash Alert in this morning:
There is a false narrative out there that Asia is taking over and that the U.S. tech sector will be overpowered by Chinese AI firms and Korean memory companies.
I need you to ignore it. The U.S. remains the leader in artificial intelligence, and the AI data center buildout is real.
In discussing the wider pressure on the AI trade and the recent pessimism surrounding it, Louis highlights a video he’ll be posting on that will come out tomorrow. In short, he says it’s very bullish for the AI trade.
Here’s Louis’ takeaway:
There are unscrupulous people trying to shake you out of your positions before the good news makes your stocks go higher…
Once you watch that video,
I need you not to worry about the gyrations in the market.
But while Louis is pointing toward sunnier days ahead, we still have a full plate of potential market-rattling events directly ahead…
Tomorrow brings the latest FOMC meeting from the Federal Reserve
The prevailing expectation is for a fifth straight meeting without an interest-rate change.
As I write on Tuesday, the CME Group’s FedWatch Tool puts the odds at 70.6% that rates remain in the 3.50% – 3.75% target range, with a 29.4% chance of a “surprise” rate hike.
Now, there’s no new Summary of Economic Projections tomorrow, which means no updated dot plot. The FOMC statement itself, and whatever Warsh says in his press conference – probably not much – are the only new information investors get from the Fed.
This means that tomorrow will likely bring a significant dropoff in content for Wall Street to chew on – and that’s by design.
As we’ve covered here in the Digest, Warsh has spent months tearing down the machinery former FOMC Chairs Ben Bernanke and Jerome Powell built up – the dot plot, the running commentary, the “here’s roughly where we’re headed” guidance that Wall Street organized entire trading strategies around.
At his confirmation hearing, Warsh put it as plainly as a Fed chair can:
I don’t believe in forward guidance. I don’t believe that I should be previewing for you what a future decision might be.
This isn’t a Fed Chair trying to signal to the market – it’s someone actively trying to get out of the way. So, the honest expectation for tomorrow isn’t a dramatic press conference – it’s closer to the opposite…
Fewer soundbites. More “the data will tell us.” And deliberate non-answers to reporters trying to bait him into previewing September and beyond.
But taking away the signal doesn’t take away Wall Street’s appetite for it
It just sends that appetite hunting somewhere else.
Fifteen years of forward guidance trained an entire industry – dot plot analysts, Fedspeak translators, hedge fund desks built around parsing a Fed chair’s exact phrasing – to expect a roadmap.
Deny them the map, and they don’t stop looking for direction. So, watch if they start overinterpreting: A pause before an answer… which question gets a clipped response and which gets a long, careful one… how his body language changes on different topics.
It will be interesting to watch how hard the market strains to find meaning in a man actively trying not to give it any.
Still, there’s a limit to how much you can actually read into a pause. So, there are likely two upcoming sources of “signal” that the market will begin to overweight: the FOMC minutes and the other 11 committee members.
As for those FOMC minutes, they provide a detailed, on-the-record account of the committee’s internal debate. Under Powell, the minutes were often an afterthought, since the press conference had usually already told you most of what you needed.
Under Warsh, that flips: with his statements running about half their old length and no color offered in the room, the minutes become the only detailed record of what the committee argued about, line by line.
This meeting’s minutes land on Wednesday, August 19 – we’ll circle back then.
Wall Street will also pay more attention to the other 11 voices on the committee. After all, Warsh going quiet doesn’t mean the Fed goes quiet.
Governors and regional presidents – Christopher Waller, Lisa Cook, Philip Jefferson, and others – have already been filling some of that vacuum with their own public remarks. Expect more of that from here, precisely because the chair isn’t doing it for them anymore.
But this doesn’t mean we don’t have a big data point to trade this week
If Warsh isn’t going to hand the market anything meaningful tomorrow, the next real data point arrives a day later: Thursday’s June PCE report.
Early estimates from Truflation see headline PCE cooling to around 3.7% year-over-year, down from May’s 4.1%, while core PCE is expected to hold roughly flat near 3.3%.
Sounds great, right?
Yes – except there’s a limit to which Warsh will care.
Last month, we told you the new Fed chair doesn’t lean primarily on the standard PCE reading that dominates the headlines. Instead, he watches the trimmed mean PCE, published monthly by the Dallas Fed.
It strips out the most extreme price movements on both ends of the distribution – the one-off spikes and drops that come from a single volatile category – and averages what’s left.
In May, core PCE ran at 3.4% while the trimmed mean sat at 2.42%, nearly a full percentage point cooler and just above the Fed’s 2% target.
That gap hasn’t been a one-month fluke. The trimmed mean’s 12-month reading has held in a tight 2.3%-to-2.4% band for six straight months now, even as headline PCE swung from the high-2s into the low 4s over that same stretch.
So, while the press will be watching Thursday’s PCE data and will be eager to hypothesize on our rate path if it comes in cool, the real issue will be if that trimmed mean’s steadiness continues.
If there’s a significant narrowing or widening, that’s what will move Warsh.
In the meantime, Wall Street is increasingly eyeing September
The Fed’s September meeting brings a fresh Summary of Economic Projections – a new dot plot – that will be genuinely contested.
Nine of 18 FOMC participants penciled in at least one 2026 hike back in June. Warsh himself submitted no projection at all – the same silence he’s likely to maintain tomorrow.
How will those same participants vote next month?
As I write on Tuesday, the CME Group puts a 57% probability on a “hold” in September. But about 23% odds of a quarter-point cut, while about 20% odds of a quarter-point raise.
We haven’t seen this degree of interest-rate uncertainty in years – which means every data release between now and mid-September carries more weight than it used to, precisely because there’s no guidance filling the gaps between them.
But it all starts with tomorrow’s FOMC meeting.
Don’t overlook the other potential market-movers this week
The Fed and the PCE data aren’t the only reasons that tomorrow and Thursday matter.
Both days also carry some of the most consequential corporate earnings of the year – and they land right on top of a debate we’ve been tracking closely in this Digest…
Whether Big Tech’s AI spending is still worth the price of admission.
Tomorrow afternoon, after the market closes, Microsoft (MSFT) and Meta Platforms (META) report earnings. Thursday after the close, it’s Amazon (AMZN) – with Apple (AAPL) also reporting that day, though it doesn’t carry the same hyperscaler capex story as the other three.
Alphabet (GOOGL) reported last week, and its results set the tone for what’s coming. As we covered here in the Digest, Alphabet beat handily, but the stock still fell about 7% the next day on fears of overspending on AI.
Is that what we’ll see this week, too?
The market’s patience for that kind of spending isn’t unconditional anymore. So, if any of these three raises its capex guidance again, don’t be surprised by the same “beat the numbers, sell the stock” reaction that hit Alphabet last week.
Keep your eyes on how the hyperscaler capex forecasts affect the AI infrastructure trade
Our technology expert Luke Lango, editor of , has argued that holding – or raising – capex guidance should support AI infrastructure stocks, not sink them, since it’s the clearest demand-side validation the market gets.
His read on Alphabet’s post-earnings slide was that it came down to a macro-driven selloff – an oil spike and a jump in Treasury yields that hit the same day. So, strip out the macro noise, and a capex raise should be a bullish signal.
Bottom line: While Warsh will spend this week trying to say as little as possible, three of the four largest AI spenders in the world will be doing the opposite – telling investors exactly how much they intend to spend next and finding out in real time whether Wall Street approves or not.
One more thing: two days until Luke’s AI MegaDeal Event
In 2023, venture firm Spark Capital wrote a $75 million check into an obscure AI startup with almost no revenue and no ticker symbol. That startup was Anthropic.
Today, Spark’s stake is worth an estimated $7 billion. Nearly 100-to-1, on a company that didn’t exist on any exchange when they bought in.
Luke has been studying that pattern closely – and he thinks it’s about to repeat, just with different names attached. To be clear, being early isn’t a guarantee of monster returns. But it’s the only way to be a part of such rocket ships before the price reflects what everyone already knows.
That’s the idea behind Luke’s free event this Thursday, July 30, at 1 p.m. Eastern – right as this week’s hyperscaler-AI-spending debate hits its peak.
Luke will walk through his , plus reveal one specific company he believes fits the pattern today.
To join Luke for free, .
For now, all eyes are on tomorrow’s FOMC meeting and Warsh’s commentary. We’ll report back.
Have a good evening,
Jeff Remsburg
(Disclosure: I own MSFT, AMZN, GOOGL, AAPL)
