💰 SpaceX Comes Back To Earth

THE MAIN EVENTSpaceX Comes Back To Earth 🪂

Yesterday morning, the handcuffs came off 912 million SpaceX shares.

Those are shares held by employees and early investors, free to sell on the open market for the first time. June's IPO had put only about 639 million shares into public hands, so the supply of tradable stock more than doubled.

You would expect that to hurt. Thousands of people sitting on years of paper gains, suddenly free to cash out on the same morning.

Here is the strange part. The stock had already fallen more than 50% from its peak before any of them could sell a share. Then it closed up 6% on the day the lock came off.

Personally, I invest on fundamentals and ignore most technical signals. But a lockup expiry is not a chart pattern. It is a hard date when the supply of tradable stock changes, and that is worth paying attention to.

What A Lockup Actually Is 🔒

When a company goes public, only a sliver of it actually trades. In June, SpaceX sold about 5% of the company at $135 a share.

So what about the other 95%? It belongs to employees, early venture funds, and the founder, and it is locked down in two ways.

By contract. Insiders sign a lockup, usually 180 days, promising not to sell. That is what expired yesterday.

By law. Even after a lockup lifts, SEC Rule 144 caps how much officers, directors, and large holders can sell in any three-month window.

So the float usually drifts up, as insiders become eligible to sell a bit at a time. Yesterday was the exception, taking SpaceX from 4.9% to 11.8% in a single day.

Everybody Front-Runs The Date 👀

SpaceX already ran this experiment once this summer, in the opposite direction.

On July 7, it joined the Nasdaq-100. Every index fund tracking that benchmark then had to buy the stock, in size, on a date published weeks in advance. It is hard to invent a more bullish setup than buyers who are required to show up.

Yet the stock peaked on June 16, three weeks before it happened.

Now reverse it. A lockup expiry is a similar setup pointed the other way: a date everyone can see, after which a lot of new selling becomes possible.

Nothing forces those holders to sell, and JPMorgan's Doug Anmuth noted that eligibility “may not mean a tidal wave of selling.” But everyone without a lockup, including short sellers, can position ahead of the insiders.

What Usually Happens 📉

Two finance professors, Laura Field and Gordon Hanka, studied 1,948 lockup expirations.

They found that in the three days around the unlock, the average stock underperformed the market by 1.5%. Trading volume also jumped about 40% and stayed there.

But that is the average across nearly 2,000 companies, most of which nobody was watching. The ones everybody watches behave differently, because… everybody is watching.

For example, Facebook's first lockup expired in August 2012, freeing 271 million shares. The stock closed at $19.87, down 6.2%, a record low at the time.

Three months later came the one everyone dreaded: 800 million more shares. Facebook rose 12.6% that day, because the fear had been priced in for months and insiders mostly chose to hold.

The takeaway: Anticipation does the selling. The date just gets the blame.

So What Do You Do? 💡

SpaceX Prospectus on sec.gov

Check the lockup calendar before you buy any IPO. It is in the prospectus, it is free, and almost nobody looks. Buying a stock ahead of an unlock is a choice, and one you should only ever make on purpose.

For what it is worth, I thought this IPO was overhyped, and I did not buy it.

That being said, it's important to remember that momentum stocks like this bottom at maximum pessimism. There is plenty of that here: a first quarter the market did not like, a stock trading below its IPO price, and ~219 million shares sold short. Whether that is maximum pessimism is the open question.

Regardless, when a stock you like craters, the key is to ask what actually changed.

1. A low price with bad fundamentals is just a bad company.

2. A low price with bad technicals might just be an opportunity.

The first one you leave alone. The second one is worth your attention.

MARKET MOODRockets Lift Nvidia, Jalapeños Sink Chipotle 🌶️

Winners

Paycom ($PAYC) - Market Cap: $10.1B (Week-to-Date: +31.7%)

Paycom sells payroll and HR software, the least glamorous corner of the software world. It just turned in the quarter nobody expected: revenue of $531 million, profit 17% above expectations, and a full-year forecast raise. Now, nearly 46 cents of every dollar Paycom takes in turns into operating profit. Maybe glamour is overrated.

Palantir ($PLTR) - Market Cap: $374.7B (Week-to-Date: +26.7%)

Six weeks ago, Michael Burry said Anthropic was eating Palantir’s lunch, but it turns out the AI boom is big enough to feed two. Revenue grew 93% to $1.94 billion, US businesses grew 149%, and management delivered its largest full-year forecast raise ever. The stock had its best day in two years.

Nvidia ($NVDA) - Market Cap: $5.30T (Week-to-Date: +9.1%)

SpaceX filed its first earnings report as a public company this week, and the most valuable line in it had nothing to do with rockets. Elon Musk told investors SpaceX will build its AI systems "exclusively" on Nvidia chips, scaling toward 10 gigawatts of computing power by 2027. Nvidia promptly added nearly $200B in market value.

Losers

AppLovin ($APP) - Market Cap: $112.8B (Week-to-Date: -15.2%)

AppLovin builds the software that decides which ads you see inside mobile games, and it has been a Wall Street darling for years. Revenue grew 53% last quarter, yet it still landed short of the company's forecast, its first miss since going public. CEO Adam Foroughi said the ad-targeting models simply did not improve as quickly as they normally do.

Chipotle ($CMG) - Market Cap: $42.7B (Week-to-Date: -9.4%)

This is Chipotle's second food-safety scare of the summer. A cyclospora outbreak dented sales in July, and now the CDC has traced salmonella back to a batch of recalled jalapeños that Chipotle and Qdoba had been serving. Chipotle already swapped suppliers in July, and the FDA says the risk has passed, but the stock lost its appetite anyway.

CVS Health ($CVS) - Market Cap: $122.8B (Week-to-Date: -7.9%)

The stores you walk past are the small part of CVS. The big parts are Aetna, the insurer it owns, and Caremark, which negotiates drug prices on behalf of employers. Aetna is finally working: profit hit $2.9 billion, nearly 3x a year ago. Caremark is not. It expects to lose employer clients next year, and investors seem to think that mattered more.

CHART OF THE WEEKToo Big To Ignore, Too Slow To Bubble ⚡

AI Buildout vs. Fiber Boom

AI Buildout vs. Housing Boom

Everyone has an opinion on AI spending, so it’s worth putting a real number on it.

Data centers should reach about 3% of GDP by 2027, up 10x from 0.3% in 2019. The fiber boom peaked at 1.2% in 2000, and housing peaked at 6.6% in 2005.

This buildout is more than double the fiber boom and less than half the housing boom. Enormous, but not unprecedented.

The fiber comparison also breaks down where it matters most. Those companies laid cable years before anyone needed it, and roughly 85% of it sat dark. AI has the opposite problem, because demand for computing is running well ahead of supply.

You will hear that a house collects rent for 50 years while a chip is obsolete in three. That is beside the point, because chips were never the constraint. A company can buy chips in a quarter. It cannot buy the power, the land, or the permits.

• Communities are saying no. More than 75 data center projects worth about $130 billion were blocked or delayed in the first quarter of 2026, matching the disruption of all of 2025 in just three months.

• Regulators are catching up. More than 300 state-level data center bills were filed in the first six weeks of the year.

• Power is the scramble. Every major hyperscaler has now signed nuclear deals, but the first one, a restarted Three Mile Island, is not expected online until 2027.

The takeaway: a bubble needs the ability to overbuild in a hurry, and this one physically cannot. That makes this buildout less like 1999 and more like a decade-long waiting list, which is a very different kind of investment.

FAST FACTSPremiums, Permits, And Paychecks 🧾

💊 Retirement Healthcare Costs $185,500: Per person, up 7.5% in a year. And 54% of pre-retirees think Medicare covers all of it. It doesn't. [Read]

🏠 Regulations Add $131,734 To Home Prices: That is 26.4% of the average sale price, and it has climbed 40% in five years while incomes rose just 18%. [Read]

📉 Job Switchers Are Getting Paid: Hiring slowed to 44,000 in July, the weakest in six months, but job changers got 7% raises versus 4.4% for job stayers. [Read]

📊 Young Day Traders Feel Like Failures: One in four men aged 18 to 29 now trade stocks daily, and 64% of them say it makes them feel like a failure. [Read]

🗽 New York's Second-Home Tax Hits In November: An annual surcharge on non-primary homes valued above $5 million will hit roughly 11,200 properties. [Read]

🧓 Gen Z Is "Retirement-Maxxing": They hold 3x what Gen X had in retirement accounts at the same age, and they start at 19. [Read]

🎓 Colleges Are Raiding Their Endowments: Withdrawals hit a record $33.4 billion and now cover 15.2% of the average operating budget. [Read]

🕵️ Burry Dropped Microsoft And His Oracle Short: He walked away from both, but still holds shorts against Nvidia, Palantir, Tesla and Micron. [Read]

👋 Google Lost Its Chief Scientist: Jeff Dean left after 27 years, and Demis Hassabis stepped down as DeepMind CEO on the same day. [Read]

WORDS TO REMEMBERStay Humble When The Market Fights Back 🧠