He Wrote the Book on 100-Baggers | Chris Mayer on SpaceX, AI Reckoning, and Why Early Is Overrated

ELI5/TLDR

Chris Mayer wrote the book on stocks that go up 100x, and he has spent enough time studying them to be unimpressed by urgency. His core message: the great companies all suffer brutal drawdowns on the way up, the market keeps offering you fresh chances to buy them, and there is almost never a reason to rush in early at a stupid price. He looks at SpaceX going public at 145 times revenue and AI being stapled onto every product on earth and sees the usual pattern — a real thing wrapped in a temporary mania that will, at some point, get a reckoning. The discipline isn’t picking winners. It’s surviving the wait.

The Full Story

The conversation hangs on SpaceX’s IPO at a two-trillion-dollar valuation, which Mayer uses as a stress test for how to think rather than a stock to pitch. He has been in markets thirty years and the scars make him cautious about mockery. He remembers standing on a stage in 2004 ridiculing Google’s IPO valuation — roughly 80 times earnings, which then ran to 120 — and being made to look like a fool over time, because anyone who simply bought and held did fabulously well.

So he holds two thoughts at once. Google at IPO was under ten times revenue. SpaceX, at its 2.6-trillion peak, was around 145 times revenue, with no earnings to even discuss.

“It’s like in sports when they say the athletes get bigger and stronger and faster. It’s like in finance the bubbles get bigger.”

His conclusion is probabilistic, not prophetic: at 145 times sales the odds are poor, and markets being what they are, you will very likely get to own SpaceX one day at a fraction of this price. Amazon, the favourite example, fell 90% peak to trough. Expect any big winner to get cut in half at least once on its journey.

Two ways a stock gets cheaper

A useful distinction surfaces: a company can get cheaper because the price falls, or because the price stalls while the business keeps growing underneath it. Airbnb is his example — the stock has gone roughly nowhere since listing, but earnings and cash flow have compounded every year, so the thing quietly de-rated into something reasonable without the price doing anything dramatic. Both routes are fine if the underlying business is still expanding. The price action and the value creation are separate machines.

Don’t let the label do your thinking

Mayer keeps returning to general semantics — the Korzybski idea that the word is not the thing. “AI,” “quality,” “safe,” “must-own,” “TAM” are labels that quietly do your analysis for you if you let them. SpaceX is really three businesses — rockets, Starlink, and a data-centre/AI arm — and the word “AI” means something completely different at SpaceX than at Google than at IBM.

“Those labels should not do the work for you… Start taking apart the segments and assessing their competitive position, looking at their growth rates, looking how much capital they’re going to need, what kind of returns potentially on that capital.”

His field evidence is deflating. An expert-network call with a large software customer described that vendor’s new AI feature as a waste of time that added nothing. His golf app added an AI feature that summarises his round and recommends what to practise; he finds it actively worse than nothing. His read: companies are bolting AI onto products because they can and because it feels mandatory, not because it solves a problem. That gap — enormous usage, thin returns — is what eventually forces a rationalisation.

“I think at some point there’s going to be some sort of reckoning… a lot of those equities will get crushed and then the people who are a little more value-minded swoop in and pick off some of the long-term winners.”

The dot-com era is the template. Most .coms died; a few, like eBay, created enormous value; and some categories that flamed out early — pets.com — eventually worked once someone, Chewy, figured out the actual business. The interesting wrinkle is that the big AI winner may not be a tech company at all. It might be an HVAC firm or a storage operator that quietly harnesses AI and prints a margin nobody expected. That is what he is watching for.

Early is overrated

This is the spine of the talk and the thing his 100-bagger study taught him.

“If it is the real deal, you have plenty of time… there’s a lot of urge with investors — they want to get in really early — and there’s not necessarily any need for that.”

A genuine winner gives you many bites at the apple. You can start small, wait a few quarters to see whether the story actually shows up in the financials, and add into the inevitable drawdowns. If the stock doubles while you wait, fine. The urgency people feel is almost never justified.

His filter is to wait until the anecdote becomes a number. Lots of companies tell lovely unit-economics stories about AI; very few see those stories survive the trip down to organic growth or margin, because there’s too much other noise and expense in between. Yes, waiting for the print means the market has mostly figured it out and you give up some upside. But it cuts out the large pile of stories that never arrive. Validation in the numbers removes a chunk of downside risk, because now there is a real business.

Market cap is the wrong frame; TAM capture is the right one

Mayer’s preference is for companies still becoming rather than companies that have arrived. He warns against reading market cap in absolutes — a $30 billion company can be tiny against its opportunity or completely mature, depending entirely on the TAM and how much of it is realistically capturable. A wonderful, high-return business can live in a small TAM. The local Toyota dealer with the regional exclusive is “printing money” without owning the national market.

He punctures SpaceX’s own logic here: defining the company’s worth by claiming the entire $27 trillion “space and time” TAM, as though owning all of it were the base case. You don’t need the whole TAM to win.

The Buffett-ism he leans on: a company earning 15% on equity with no payout will, over five years, have its CEO deploy more capital than the entire business has accumulated in its history. So what management does with the next few years of incremental capital matters enormously. He trusts organisations with return-on-capital in their bones — Constellation Software is his example — over the ones speaking the language of market share, being first, capturing the land, where returns never enter the sentence.

Governance, trust, and the BS detector

SpaceX tests corporate-governance limits — Elon appoints the board, shareholders can’t sue, it’s effectively his company and you’re along for the ride. Mayer is old-school enough to dislike even dual-class stock, yet honest enough to admit the tension: a fully rational board never lets SpaceX get to the last-launch-or-bankruptcy moment that made it. Backing entrepreneurial energy means tolerating concentrated control, which is also exactly where insiders have historically enriched themselves at shareholders’ expense.

So he falls back on reading the people. Modest compensation, no salary grab, a CEO whose wealth comes from the stock rising rather than from pay and options, small clues like not flying first class or not driving a flashy car — no single one means anything, but the pattern does. He wants boards of large owners plus people who genuinely add value, not “independent directors” who by definition own little and aren’t engaged. The anecdote — a footnote, a media appearance, a story from someone who’s dealt with them — is the underrated input. There are still people running these companies.

The real lesson: drawdowns and the long wait

The emotional core comes via a Worldly Partners paper, Generational Investing, on stocks that rose 100x since 1972. The findings validate his whole career: 82% of those eventual mega-winners lost more than half their value at some point, the average drawdown was 65%, and there were eight years between highs — and still they returned 533x on average.

“I think you have to approach it by knowing this is the way stocks behave.”

He frames the disappointment investors feel right now through an idealisation-frustration cycle: people hold an ideal — stocks that mostly tick upward — that doesn’t match a reality of repeated halvings and long dead stretches. He pairs it with Wes Gray’s “Even God Would Get Fired” study: build the perfect five-year portfolio with foreknowledge of the winners, and it still suffers 35%+ drawdowns. Even a divine stock-picker gets fired by impatient clients. The long-term mindset is hard precisely because everything in the culture — quarterly measurement, sloshing passive and algorithmic flows, the FOMO around new IPOs — pushes the other way.

Key Takeaways

The best long-term winners are not smooth compounders. Across stocks that rose 100x since 1972: 82% suffered a 50%+ drawdown, the average drawdown was 65%, and there were eight years between highs. Expect to be cut in half at least once.
“If it’s the real deal, you have plenty of time.” A genuine winner offers many entry points; early is overrated. Start small, wait for proof, add into drawdowns.
Wait for the anecdote to become a number. Management AI stories rarely survive the trip to organic growth or margin; once the financials confirm it, downside risk drops sharply even if you give up some upside.
A stock can get cheaper two ways: price falls, or price stalls while earnings grow underneath it (Airbnb). Separate price action from value creation.
Don’t let labels — “AI,” “quality,” “TAM,” “must-own” — do your analysis. Break the company into segments and assess competitive position, growth, capital needs, and returns on that capital.
Read market cap relative to TAM, never in absolutes. A high-return business in a small TAM beats owning a sliver of a giant one. You don’t need the whole TAM to win.
Buffett-ism: a company earning 15% on equity with no payout will, over five years, deploy more capital than it accumulated in its entire prior history. Next-few-years capital allocation is decisive.
Favour companies with return-on-capital in their culture (Constellation Software) over those speaking only of market share and being first.
For concentrated-control founders, judge the person: modest pay, no salary grab, wealth tied to the stock not extraction. Patterns of small frugality signal motivation; no single clue is proof.
“Even God would get fired” — even a perfect-foresight portfolio suffers 35%+ drawdowns. Underperformance and waiting are features, not bugs.
Expect an AI reckoning: huge usage today, thin returns, a coming rationalisation that crushes weak equities and lets the patient pick off survivors. The biggest winner may be an ordinary business, not a tech name.

Claude’s Take

This is a genuinely good conversation, and it earns the score by being the opposite of what its clickbait title promises. There is no SpaceX hot take, no AI doom-mongering — just a 30-year practitioner using the loudest stories of the moment as props for durable, unglamorous discipline. The 100x drawdown statistics (65% average drawdown, eight years between highs) are the kind of fact that should be tattooed somewhere, because they reframe the entire emotional experience of holding winners.

The BS-filter notes: the value-investor’s perpetual hedge is on display. “It’s probably overvalued but markets make me look like a fool, so I’ll stay humble” is intellectually honest but also unfalsifiable — he gets credit whether SpaceX triples or halves. The Google anecdote cuts both ways: he mocked it and was wrong, which is the whole point, yet the takeaway he draws (“wait, you’ll get another chance”) is the opposite of what actually rewarded Google holders, who were rewarded for not waiting. He half-acknowledges this. And the “7% of volume is human” stat he repeats with an audible shrug — he flags it as possibly-fake-but-feels-true, which is the right amount of skepticism, but it still gets aired.

What lifts it above the average finance podcast is the texture: the useless golf-app AI, the welder made a millionaire, the bank that put its biggest customers on the board, the dental-practice-versus-national-TAM image. Concrete, memorable, transferable. Docked from a 9 because it’s two regulars riffing comfortably and occasionally drifts (the Mars incentive-comp tangent, the “rational exuberance” riddle they punt on). But the signal-to-noise is high and the frameworks are real. Worth the hour.