Diet Cokagne
Wandering Wednesdays #152
In 2023, Tom Hanks invented Diet Cokagne by mixing champagne and Diet Coke.
It sounds like something a weary parent might confiscate from a teenager.
And yet, folks tried it. Many liked it. Some even asked for the recipe.
Which brings us, naturally, to the SpaceX IPO.
Its S-1 reads like a lightly carbonated prospectus for Mars, magical thinking, marginal compute and memes. Diet Cokagne, in other words, for the capital markets. One part rockets that land. One part broadband that works. One part AI that lost five billion dollars last year and four billion in Q1’26. Add Mars, for color. Shake gently, stir vigorously, and serve to retail investors in a chilled flute marked forward looking statements.
Leave plenty of room for bubbles.
The astonishing thing is not that the whole offering is absurd. The astonishing thing is that some parts are not absurd at all.
SpaceX has pioneered reusable rockets. It has lowered launch costs. It has turned Starlink into an Internet service that normal people use, including people on airplanes who have discovered that on-air Wi-Fi need not be like sending snail mail.
There is a real business here. Several of them, in fact, cobbled together by someone who regards corporate structure as a quaint hobby for old-fashioned people. There is the rocket business and the broadband business, both of which make money. There is AI, which does not, but hopes to eventually. There is a social network hiding inside the AI business, wearing novelty glasses, glued onto the rocket company for obvious synergies. And there is the someday business of putting data centers in orbit, because terrestial ones require land, power, water, permits, politics, and neighbors unwilling to let it hum loudly beside their cul-de-sac.
So far, so plausible.
Then the prospectus opens the champagne.
… and reaches for the high heavens—the largest actionable total addressable market in human history.
Their words, not mine. The number quoted is $28.5 trillion, roughly the size of the US economy today, of which $26.5 trillion is AI. The rockets, in other words, are a rounding error in the spreadsheet. Mars is a footnote to a footnote.
The phrase is brilliant because each word may have been chosen by a different author. Largest came from marketing. Actionable from consulting. Total addressable market from VC. Human history from the man who has been gazing at Mars forever.
This is where the blend becomes dangerous. Starlink is the Diet Coke: mass-market, scalable, already succesful. The rockets are the glass: expensive, difficult, and the only reason the champagne isn’t on the carpet. AI is the champagne: intoxicating, fashionable, capable of making everyone at the party louder and more combative. Mars is the garnish. And Musk is the bartender—the reason a roomful of sensible adults agreed to try the cocktail in the first place.
The thing about Musk’s companies has never been that the numbers make sense. It's that enough investors decide the numbers are impolite—standing, as they are, in the way of destiny.
Tesla began as a bet that EVs could be desirable, not dutiful. That bet became a product, then a movement, then a stock, then a theology with quarterly deliveries. SpaceX is attempting something similar, except the article of faith this time is the rack, not the rocket. Launches, nat-sec contracts, Wi-Fi—all of these prologue to the AI business, the valuation kicker. The Mars renderings are the romance, the part that gets the room to fall in love. AI is what they're being asked to pay for. SpaceX is, by its own arithmetic, an AI enterprise that has left the atmosphere, priced on the continued willingness of public markets to squint heroically.
The risk factors write themselves.
We may not reach Mars.
We may reach Mars and discover that the addressable market consists mainly of rocks, silence, and one employee from finance asking about the reimbursement policy.
Our AI strategy depends on the continued availability of chips, power, customers, satellites, engineers, launch windows, regulatory patience, and the assumption that machines will soon assign work to other machines at sufficient scale to justify putting the racks in orbit.
Our founder owns forty percent of the company and controls eighty-five percent of the votes, an arrangement under which he can be removed as CEO only by persuading himself to do so.
Our CEO’s billion-share pay package vests only if the company is worth $7.5 trillion and a million people are living on Mars, conditions we are obliged to list as risks rather than the plot of a new Hollywood film.
Our business may be adversely affected by gravity.
You can mock all this. You probably should. A $28.5 trillion market has the same comic dignity as a child declaring he will earn pocket money by licensing oxygen. But satire should not become laziness in a better suit. The awkward fact remains that Musk has done this before: taken a category that respectable people found too hard, too capital-intensive, too regulated, or too ridiculous, and then made the respectable people update their slides.
That is what makes the IPO interesting. It is not a simple story of genius or fraud. It is a story about what happens when capital markets stop asking, What is this company worth? and begin asking, What future becomes more likely if we pretend this valuation is already true?1
Sometimes that is how bubbles form.
Sometimes that is how infrastructure gets built.
Often, inconveniently, it is both.
Which brings us back to Diet Cokagne. A sommelier would object. A doctor might object. A civilization with intact boundaries would certainly object. But someone takes a sip anyway. Then someone else asks for one. Before long a roomful of adults is holding glasses and saying, You know — this is not as bad as it sounds.
That may be the SpaceX IPO in one sentence.
If it works, we will call it vision. If it fails, we will call it overreach. If it sort of works—the likeliest outcome—we will call it capitalism: Diet Coke in one hand, champagne in the other, standing on a launchpad, reading the fine print by rocket fire.
P.S. This is not investment advice. SPCX listing on NASDAQ expected on June 12; participation in the lottery guarantees nothing. I've said what needed to be said.
Within days of the S1 filing, the asking price has been trimmed from two trillion dollars toward one and three-quarters—the market, it seems, pretending slightly less hard.
Down $250Bn on the upcoming IPO? Given the theme of your article on Diet Coke and champagne - someone has to say it: clearly the market is less 'frothy' for this product. :-)