Elon Musk didn’t just pull off the largest IPO in history on June 12. He pulled it off at what may prove to be the single best moment in a decade to sell equity, and the closer you look at the numbers, the harder that is to write off as coincidence.

The raise

SpaceX priced at $135 a share, valuing the combined SpaceX-xAI entity at $1.77 trillion and putting $86 billion in cash on the balance sheet. By the close of day one, the market cap had run to roughly $2.1 trillion enough to make Musk, on paper, the first trillionaire in history. The deal priced without the usual investor roadshow, what Reuters called “take-it-or-leave-it” pricing. Underwriters reportedly warned clients that flipping shares on debut could get them frozen out of future allocations, a way of locking in the price rather than letting the market find it.

The burn rate the raise was covering for

Strip away the valuation headline and the case for urgency comes into focus. xAI, folded into SpaceX in February at a $1.25 trillion combined valuation, has been reported burning through cash at a pace few private balance sheets even Musk’s could sustain indefinitely: billions annually on GPUs, data-center buildout, and compute contracts, against roughly $250 million in revenue and $2.5 billion in losses over six reported months, by Motley Fool’s count. A $1.25 billion-a-month compute contract with Anthropic covering the Colossus 1 data center in Memphis helped the growth story, but growth-story revenue and profitable revenue are not the same thing when the burn is this steep.

Private capital can fund that kind of loss for a while. It cannot fund it forever, and it gets far more expensive to fund the moment credit markets tighten.

The multiple nobody has fully explained

SpaceX priced at somewhere between 67 and 112 times trailing sales depending on whose revenue estimate you use three to five times Nvidia’s multiple, and a number even sympathetic analysts struggled to justify. Morningstar’s Nicholas Owens put fair value near $780 billion, about 55% below the IPO price, pointing to index-inclusion mechanics forcing passive funds to buy regardless of price, and a genuinely tiny public float relative to the headline valuation. Ross Gerber, an existing shareholder, was blunter: the number reflects confidence in Musk personally, not in the underlying financials.

That gap, a company priced on faith in a founder rather than in its numbers is exactly the kind of gap that closes fast when market sentiment turns.

Why the “he knew” theory won’t go away

Nobody has reported that Musk timed this to a specific macro call. That claim doesn’t exist in any prospectus or filing. But the sequence of events is doing a lot of the talking on its own: an insider sale in December valuing the company at $800 billion, a merger two months later that folded xAI’s losses into a stronger balance sheet, and an IPO five months after that priced at more than double the December mark all while the underlying business was burning cash faster than almost anything else in the private market.

Set against a backdrop where historians of the 1929 and 1873 collapses point to structural downturns running anywhere from six to fifteen years to fully clear bad debt from bank balance sheets, the pattern looks, at minimum, well-timed. Whether Musk was reading those signals or simply moving because xAI’s cash burn made an IPO inevitable regardless of the macro picture is a question nobody outside his boardroom can answer. But the effect is the same either way: $86 billion in cash-burning, hard-to-fund private risk is now sitting on public shareholders’ books, at a 67-to-112x multiple, right as several analysts are already flagging it as overvalued.

The case for an 8-to-14-year downturn

The theory that this IPO was timed defensively only carries weight if the macro backdrop actually justifies that kind of urgency. Here’s the historical argument for why a coming downturn could run far longer than a typical recession and why some strategists think 8 to 14 years, not 8 to 14 months, is the right window to plan around.

Standard postwar recessions are short because central banks can cut rates and refill the system with liquidity. Structural, debt-backed collapses don’t respond the same way, because the problem isn’t a lack of liquidity it’s bad collateral sitting on balance sheets that has to be written down before credit can flow normally again. History offers three reference points:

• The Great Depression (1929): the physical contraction ran 43 months, but by most measures the economy didn’t return to its prior long-term trend line until 1942, 13 years out. Investors who moved to cash at the top reportedly took 15 years to break even.
• The Long Depression (1873): triggered by overbuilt railroad infrastructure, a 19th-century analog to today’s data-center buildout the initial contraction ran 65 months, with structural drag lingering anywhere from 6 to 23 years depending on the country.
• Dot-com and 2008: in both cases, recouping core value took roughly 5 to 7 years under normal liquidity conditions, but stretched past a decade in portfolios caught in a genuine “dash for cash.”

The economic argument for why this cycle, if it turns, could run long: fixed-investment cycles (the Juglar cycle) typically run 7 to 11 years, while infrastructure-investment cycles (Kuznets) run 15 to 25 years. A bubble concentrated in physical infrastructure data centers, GPUs, fiber, satellites sits at the intersection of both, which is the same alignment historians point to in 1873. On this reading, a full reset requires the 7-to-11-year fixed-investment cycle to clear before recovery can take hold, roughly matching the 8-to-14-year window.

A rough phase breakdown under this thesis: years 1–3 as shock and default; years 4–7 as stagnation, with weak hiring and tight credit; years 8–12 as the slow rebuild, once toxic infrastructure debt is fully written off and capital starts moving through verified, de-risked channels again.

It’s worth being direct about what this is: a historical-analogy argument, not a forecast with a track record. Mainstream economists are split on whether today’s AI-infrastructure buildout resembles 1873’s railroads (overbuilt, debt-fueled, due for a bust) or looks more like the early internet genuinely transformative capacity that gets absorbed once prices reset, without a lost decade attached. Nobody, including the strategists making this case, has a reliable model for calling the exact length of a downturn that hasn’t started yet.

What happens to the banks if the thesis is right?

If a structural downturn does hit, the mechanism isn’t abstract. A meaningful share of recent tech lending has been collateralized against things that aren’t physical assets projected cloud revenue, data-center contracts, GPU leases with book values that assume years of AI demand growth. A warehouse of depreciating servers doesn’t liquidate the way a building or a fleet does. If growth assumptions break, banks holding that paper would have to write down non-performing loans at scale, and under this thesis, the weaker regional and mid-size lenders would be the first to absorb the hit some likely folded into larger, systemically important banks, with cross-border lending tightening as institutions ring-fence domestic balance sheets. The playbook mirrors 2008, just with data-center debt in place of mortgage-backed securities.

The knock-on effect, on this reading, is a shift in what gets rewarded: after a shakeout like that, capital tends to flow toward whichever institutions can show clean, verifiable balance sheets and transparent infrastructure exposure and away from anything still carrying opaque, hard-to-value tech collateral.

Who actually eats the loss?

The core asymmetry in this thesis isn’t really about SpaceX. It’s about balance-sheet structure in a prolonged downturn generally. A business or household with 6 to 12 months of runway doesn’t survive a decade-long freeze it has to sell assets into a falling market, often at steep discounts, just to stay liquid. An entity sitting on tens of billions in already-raised, un-dilutable cash can instead sit still, or even buy distressed assets at those discounted prices, while others are forced sellers.

Under this framing, Musk’s position after the IPO is close to the best case for that dynamic: $86 billion in cash already banked, 82% voting control retained through super-voting shares, and the newly public shareholders retail investors, pension funds, and passive index holders required to buy in for index-tracking purposes now the ones directly exposed to whatever happens to a 67-to-112x-sales valuation if growth disappoints. If the downturn plays out the way the thesis predicts, that gap in staying power is where the pain concentrates: not evenly, but on whoever didn’t get to convert equity into hard cash before the reset.

Why size matters more than the depression-length number does

Set the 8-to-14-year figure aside for a moment, because there’s a tighter, more defensible version of the “this bubble is now bigger” argument: index inclusion.

Once a company the size of SpaceX joins major indices, passive funds are mechanically required to hold it in proportion to its weight not because portfolio managers chose to underwrite a 67-to-112x-sales valuation, but because tracking the index requires it. That’s the same dynamic Morningstar flagged when it put fair value at roughly 55% below the IPO price: a large chunk of demand isn’t valuation-driven buying, it’s rules-driven buying. It pulls in exposure from pensions, 401(k)s, and index-fund holders who never made an active decision to bet on this specific company.

That matters for how a correction would transmit, regardless of how long it lasts. A standalone overvalued stock can crater without dragging much else down with it. A stock this large, embedded this deeply in passive index weightings, moves differently, a sharp repricing shows up directly in the retirement accounts and index funds of people with no idea they’re exposed. That’s a real, mechanical amplification of the size argument, and it doesn’t require betting on a specific multi-year timeline to be true. It’s a statement about who’s holding the risk today, not a forecast about how many years it takes to unwind.

The bear case, plainly stated

If SpaceX’s AI ambitions underdeliver, or the tech-financing environment tightens the way it did in 2000 or 2008, the retail and index investors who bought in at $135 are the ones exposed to the downside. Musk, sitting on $86 billion in freshly raised cash and 82% of the voting power via super-voting shares, is not.

That asymmetry doesn’t require a conspiracy to be worth pointing out. It’s simply what the capital structure says, in plain numbers, about who is holding the risk and who is holding the cash.

This piece reflects market analysis and interpretation of public filings and reporting, not confirmed statements of intent by Elon Musk or SpaceX.