This was the week the AI trade stopped being a chip story and became everyone’s story.

Wednesday, Marvell’s $115 billion bet started looking shakier the closer you read the numbers.

Thursday, it jumped the fence into software. Palantir, CrowdStrike, and Okta all sold off together, erasing roughly $40 billion off Palantir alone.

And Palantir did it despite posting the best quarter in company history.

Revenue crossed $2 billion for the first time ever, up 96% year over year. The stock still fell double digits.

The market isn’t questioning whether Palantir can grow. It’s questioning what’s slowing down while it does, and whether that question now applies to the whole group.

That’s exactly what today’s Playbook digs into.

Missed Wednesday’s deep dive on Marvell?

Thursday’s selloff pushed the Nasdaq into correction territory as investors dumped technology and AI-linked names, while the Dow fell more than 1,800 points. Attention now turns to Friday’s U.S. jobs data, along with earnings from companies such as Moderna and Enbridge. The bigger takeaway is that this market has suddenly become much more sensitive to rates, policy, and economic data after months of AI-led optimism.

SpaceX potentially reaching $570 a share by the end of 2029, roughly quadrupling from current levels, if Starship scales rapidly and AI infrastructure becomes a major growth engine. The article assumes SpaceX could approach $2 trillion in annual revenue by 2030, helped by Starlink, launch services, and entirely new markets. In plain English, the upside is enormous, but so is the execution risk behind those forecasts.

Marriott beat Q3 expectations with EPS of $3.02 and revenue up 8%, but the more interesting story is its capital-light business model. Gross fee revenue rose 14%, while hotel conversions accounted for more than 40% of openings, allowing Marriott to grow without funding most hotel construction itself. The catch is valuation, with the stock trading around 43 times earnings. The takeaway is that Marriott increasingly looks like a high-margin brand and loyalty platform, not simply a bet on hotel occupancy.

Your client doesn't care how their site was built. They care that it launches on time, runs on their own domain, and keeps working long after you hand it over.

With Lovable, you describe what your client needs and get a working version you can show them in the first meeting. Then you take it all the way to a live site on your client's domain, built to be found on Google. Lovable's project monitoring feature spots bugs and suggests fixes, so your name stays on work you're proud of.

Start today and take $5 off your first month.

Build for clients, $5 off

_{^{Discount of $5 applies to first month of any paid plan. You need to sign up via this webpage in order to claim the discount. The discount applies to new and returning users without an active paid plan. Terms apply.}}

Three things happened to AI-adjacent names this week, and none of them individually explain Friday.

Marvell’s leverage story cracked on Wednesday. CrowdStrike and Okta fell alongside Palantir on Thursday. And Palantir posted the best quarter in its history and still got sold.

The easy read is that investors are just nervous and selling good news on principle.

That read is lazy.

Pull the numbers apart and there’s a real signal sitting underneath the noise. This Playbook is here to find it.

Start with the headline numbers.

Palantir didn’t just grow this quarter. It detonated growth.

Revenue hit $2.31 billion in Q3, up 96% year over year and 20% from the prior quarter.

US commercial revenue grew 148% year over year to $984 million, overtaking US government revenue of $953 million for the first time in company history.

That is the line Alex Karp has been chasing for three years.

Management raised guidance into it too, lifting full-year revenue growth to roughly 91% and commercial growth guidance to 143%.

Morgan Stanley raised its target to $233 from $205. DA Davidson reiterated Buy at $250. Bank of America held Buy at $255.

So no, this week did not prove Palantir’s business is slowing. It proved the opposite.

Here’s what the headline revenue didn’t tell you.

Palantir closed $1.46 billion in total contract value from US enterprises this quarter. Last quarter, that number was a record $2.13 billion.

That’s a sequential drop of roughly 31%.

Revenue is what already happened. Contract value is what’s coming next.

Karp’s own explanation is that deals are getting shorter in duration but expanding fast once signed.

That’s a real possibility. It’s also exactly what a company would say either way.

Because a company priced like Palantir isn’t allowed to just beat the past. It has to keep proving the future is still accelerating too.

This week, the future side of that equation wobbled, and that’s why the stock fell on a quarter that smashed every revenue estimate.

There’s a second layer here, and it’s valuation, not operations.

Even after this week’s drop, Palantir isn’t cheap by any normal software measure.

Morgan Stanley’s own $233 target comes with an Equalweight rating, not a bullish one. That’s a target that says “fair value,” not “buy more.”

A cheaper stock shrugs off a soft bookings quarter. Palantir can’t.

The multiple leaves no cushion for anything less than flawless, and this week wasn’t flawless.

Palantir’s week had two distinct legs down, and they tell different stories.

The first came right after earnings, a 6 to 8% drop that took the stock from the high $190s into the $170s as traders digested the contract-value number.

That part was Palantir-specific. The second leg wasn’t.

On Thursday, the stock fell another 10%, landing near $155, as CrowdStrike, Okta, and Palo Alto all dropped the same day on a broader rotation out of AI software.

The NASDAQ fell more than 5% that session. This wasn’t just a Palantir story anymore.

Palantir is now down 13% year to date and sitting roughly a quarter below its October 9 record high.

The level to watch heading into next week is whether $150 holds.

If it does, this starts looking like a sector-wide flush that found its floor. If it doesn’t, the bookings story stops being the headline and the valuation reset becomes the story instead.

If total contract value snaps back toward that $2 billion-plus range, this becomes a one-quarter wobble. If it stays depressed, the market starts treating this quarter’s revenue beat as the last clean one for a while.

Karp’s explanation for the contract slowdown is structural, not demand-related. The next two quarters either confirm that shorter deals expand the way he says, or they don’t.

CrowdStrike and Okta falling the same day changes the read entirely. If the whole group stabilizes together next week, this was a flush. If Palantir falls further while peers recover, the bookings story is uniquely Palantir’s.

This week’s selloff traces back to the midterm result and a political mood that’s turned skeptical of AI buildouts generally. Any fresh headlines on that front, in either direction, will move this group faster than earnings will.

I covered PLTR back on Oct 12 in Wall Street Upgraded PLTR. The Options Desk Didn’t Cheer. Options traders were skeptical of an upgrade then. Now the skepticism has moved from sentiment to the actual contract numbers, a more serious version of the same doubt.

I keep coming back to the same thought with this one.

Palantir did not have a bad quarter. It had a quarter too good to hide the one thing still soft underneath it, and then it got caught in a selloff that had nothing to do with Palantir at all.

That’s the real lesson of this week. Marvell’s leverage, Palantir’s bookings, Thursday’s software rout. Three different stories that are actually one story: the market repricing every AI-adjacent name as a single trade instead of individual businesses.

For months, a beat was enough. This week proved a beat alone doesn’t cut it anymore, and it proved something else too.

And honestly? When CrowdStrike and Okta fall on the same day for the same reason, that’s not a stock problem anymore. That’s a crowded trade unwinding.

Palantir is still one of the best growth stories in software. Nothing about this week’s numbers changes that.

But being the best growth story and trading like the safest one are not the same thing, and this week is the market figuring out which one it’s been paying for.

Wallets are returning, stablecoin supply is rising and institutional money is still flowing in. Our free guide explains why we’re accumulating SOL now, plus the key metrics and risks to watch.

Get Free Guide

This week rewarded patience over reaction, twice.

Marvell’s dip on Wednesday wasn’t the end of the AI infrastructure story. Palantir’s dip on Friday isn’t the end of its growth story either.

The lesson repeating all week is simple. A single stock’s selloff tells you one thing. A whole group selling off together on the same day tells you something different, and most investors don’t pause long enough to tell which one they’re looking at.

Going into the weekend, resist the urge to score this week by Thursday’s close.

Score it by what you’d still want to own in three months.

If this week’s Playbook saved you from a bad entry, send it to one friend who needed it too.