A New Reality for the AI Trade
Listen to the audio version of this article (generated by AI).
Microsoft eases AI fears… the market’s new AI split… Luke Lango’s Exit Ladder if the selling returns… what the Fed’s preferred inflation gauge just revealed…
Well, Microsoft (MSFT) just saved the AI trade – at least for the moment.
Yesterday after the closing bell, the software giant reported earnings that easily cleared Wall Street’s expectations.
Quarterly revenue reached $90 billion, well ahead of the $87.6 billion analysts expected. Adjusted earnings of $4.74 per share topped the $4.24 consensus estimate.
More importantly for the AI trade, Microsoft’s Azure cloud business accelerated.
Azure revenue jumped 43%, compared with 40% growth in the prior quarter and roughly 40% expected by analysts. Microsoft also revealed that Azure generated more than $100 billion in annual revenue for the first time, while its Microsoft 365 Copilot product surpassed 30 million paid seats.
As for capex, Microsoft kept its calendar year 2026 forecast unchanged. But on the call, finance chief Amy Hood added that she expects capex to keep climbing in fiscal 2027, pointing to “demand signals across our portfolio.”
In other words, Microsoft isn’t merely spending heavily on AI infrastructure. It’s seeing strong, accelerating demand for it – and telling investors more spending is coming precisely because that demand justifies it.
As I write on Thursday, MSFT is soaring nearly 17%.
Meta (META) also reported last night – and told a different story
Revenue rose 28% to a record $60.8 billion, edging past Wall Street’s $60.2 billion forecast. But earnings came in at $6.18 per share, well below the roughly $7.22 analysts expected – a 14% year-over-year decline in net income.
Part of that shortfall reflected one-time items: $2.4 billion of legal charges and $1.18 billion of severance costs tied to recent layoffs. But even stripping those out, expenses climbed 55% to $42 billion, badly outpacing revenue growth.
On the capex question, Meta’s spending continues to surge. The company spent $31.1 billion on capex during the quarter, up 83% from a year ago.
Looking ahead, it narrowed its 2026 capex forecast from $125 billion-$145 billion to $130 billion-$145 billion.
Technically, Meta didn’t raise the top end of its forecast. But by lifting the bottom end by $5 billion, management effectively told investors that spending is more likely to land toward the upper half of its previous range.
Shares are down more than 9% as I write.
The bottom line: Neither company is tapping the brakes on AI spending. But Wall Street decided one of them has proof the spending is working – and the other doesn’t, which accounts for the tale of two returns today.
Returning to Microsoft, did it just save the AI trade?
As I write on Thursday, the AI trade is soaring – a welcome rally after weeks of relentless selling pressure.
Coming into today, the Nasdaq was down about 10% from its record high set roughly two months ago, while chip stocks, as measured by the iShares Semiconductor ETF (SOXX), had collapsed by nearly 30%.
While multiple fears contributed to the bloodbath, the primary concern was that the tech giants were burning cash on AI infrastructure with no near-term return. That was bad for the hyperscalers, but also for the wider AI trade, as the fear was that, eventually, the lack of returns for the hyperscalers would force them to cut their spending, meaning less revenue for AI infrastructure companies.
Microsoft eased those fears last night.
As for capex, Microsoft reported $41 billion in property investments, including finance leases, which fell below the $42 billion figure Wall Street feared.
And on the return side, Azure’s full-year revenue soared past the $100 billion milestone for the first time. This gave Wall Street exactly what it needed – definitive evidence of commercial monetization from corporate AI transformations…which is also proof that the need for the broader AI infrastructure trade still exists.
On that note, SOXX is soaring almost 9% today as I write.
But here’s a critical note for your portfolio…
This morning is evidence of a new split within AI. Investors are punishing companies with abstract AI ambitions – like Meta – but aggressively rewarding foundational platforms showing immediate utilization and software margins – like Microsoft.
So, “show me the money” is the new reality in AI.
After the closing bell today – likely by the time you read this – Amazon (AMZN) will report earnings and update its capex guidance. We’ll see whether it adds to today’s AI comeback or throws cold water on it.
Now, in case “cold water” is the outcome, our technology expert has you covered…
How to protect your portfolio if today’s rally is just temporary relief
While AI stocks are surging today, after five straight weeks of grinding selling, one good morning doesn’t erase the possibility that the pain comes back.
Our technology expert Luke Lango has a plan for such an outcome.
Two semiconductor benchmarks – SOXX and the VanEck Semiconductor ETF (SMH) – both broke below their 50-day and 100-day moving averages earlier this week after weeks of steady selling.
Luke says the market has been pricing in a full basket of fears at once: hyperscaler capex peaking, open-source models commoditizing premium AI tools, circular financing arrangements in the AI supply chain, a possible re-escalation in Iran, and the chance that Fed Chair Kevin Warsh will hike rates rather than hold.
If those fears turn into a fresh wave of selling pressure, AI investors need a game plan – and that’s what Luke just provided to his Innovation Investor subscribers:
We are introducing a pre-defined, rules-based Exit Ladder as insurance against our fundamental bull thesis — four tranches of 25% each keyed to Fibonacci retracement levels, the 200-day moving average, and major prior support shelves on SMH and SOXX — not a forecast, but a hedge to protect open gains if the bounce never comes.
In other words, this is a pre-committed plan for reducing AI infrastructure exposure if the selling returns, decided in advance and executed without emotion.
Now, as I write, both SMH and SOXX are well above Tranche 1 (SMH $480-490 / SOXX $440-465). So, nothing has triggered.
If this bounce holds, the ladder simply goes unused – insurance nobody needed. But if the bears return, the plan is already in place.
The important takeaway: planning
Most investors approach a moment like this by trying to guess right – will this tech bounce continue? Will the U.S/Iran reach a permanent ceasefire? Will Warsh be able to prevent the Fed from raising rates this fall?
But having to guess correctly is a terrible foundation for a plan – it only works until the one time you guess wrong, and that one time is usually the one that costs you the most.
What seasoned investors like Luke do instead is separate the decision from the outcome. They decide in advance what would prove them wrong, and what they’ll do about it if that happens – before that moment has arrived and is clouding their judgment.
That’s the entire point of something like the Exit Ladder.
It’s not a prediction that stocks will fall further, any more than a smoke detector is a prediction that your house will burn down. It’s a decision made in a levelheaded moment about what action gets taken in an uncalm one.
And once that decision exists on paper, the only thing left to do is check the actual price against the actual trigger – no second-guessing, no hoping, no talking yourself into “just one more day.”
Define the risk before you’re in it, and by the time you’re in it, there’s nothing left to decide.
If you’d prefer to create and follow your own Exit Ladder instead of using Luke’s, that’s great. Just know what you’re doing, when, and why.
If you’re an Innovation Investor subscriber, click here to login and get all the details. And to learn about joining Luke in Innovation Investor, click here.
By the way, this afternoon, Luke went live, explaining why he thinks the biggest fortunes of the AI boom may not come from buying the household names everyone already knows – even the ones soaring today.
He detailed why he believes there’s another side to the AI story that many investors have overlooked. But it’s one that has the potential to generate returns that dwarf even some of AI’s highest fliers today.
If you missed his presentation, click here to catch a free replay.
What today’s inflation data told us
Let’s not overlook the June Personal Consumption Expenditures (PCE) that dropped this morning.
The headline number fell 0.1% in June, putting the 12-month headline rate at 3.7%, down from May’s 4.1%. Core PCE, which strips out food and energy, rose 0.1% for the month and now sits at a 3.3% annual rate.
Both readings still sit well above the Fed’s 2% target. But regular Digest readers know that headline and core PCE aren’t the numbers Fed Chair Kevin Warsh is watching.
As we’ve covered here in the Digest, Warsh has made clear he sees the standard PCE gauge as little more than a “rough guess” at real price pressure, since a single energy shock can ripple through the data and get mistaken for broad inflation.
His preferred measure is the “trimmed mean PCE” from the Dallas Fed. It throws out the most extreme price movers at both ends each month and averages the rest. It’s designed to show the inflation that’s actually sticking, not the noise from oil spikes or one-off shocks.
So, what did the trimmed mean show today?
It eased.
The 12-month trimmed mean came in at 2.2% in June, down from 2.4% in May and its lowest reading of the past six months. The one-month annualized figure cooled even more sharply, dropping to 1.4% after May’s 2.7% reading.
That’s notable, because it means the gap between core PCE (3.3%) and the trimmed mean (2.2%) – already wide – got even wider this month.
If you’re in Warsh’s camp, that gap is exactly the point: strip out the noise, and the underlying inflation trend looks a lot closer to the Fed’s target than the headline numbers suggest.
As I noted yesterday, if you think a rate hike in September is a lock, make sure you’re factoring this data into your assessment.
We’ll keep you updated on all these stories here in the Digest. For now, enjoy the rally.
Have a good evening,
Jeff Remsburg
(Disclosure: I own MSFT, AMZN, SMH.)