The $780B Bear Case

On June 3, 2026, Morningstar analyst Nicolas Owens initiated coverage with a fair-value estimate of about $780 billion — less than half the roughly $1.77 trillion IPO valuation at $135 a share — and called SpaceX "significantly overvalued." His discounted-cash-flow model assigns approximately $611 billion of enterprise value to the core launch and Starlink businesses and only about $170 billion, on a probability-weighted basis, to the AI businesses (xAI and X); together that lands near $780 billion. The gap to the IPO price, roughly $990 billion, is almost entirely a premium Owens declines to pay for the AI bet. Importantly, this is not a fringe view dressed up as analysis: it is an independent rating agency crediting the launch monopoly and the Starlink flywheel at more than half a trillion dollars and still concluding the deal is priced at roughly twice fair value. His recommendation is explicit — wait for "more attractive levels" after the IPO, because he expects the market to discount the premium once the roadshow's scarcity dynamics fade.

That its value is speculative, its moat unproven, and its downside real. Morningstar models the most ambitious leg of the AI plan — data centers in space — as a probability tree: a "moonshot" worth about $1.3 trillion but with only a 7% chance, a "no-go" that destroys more than $81 billion in value with a 43% chance, and a middle case carrying the balance. Probability-weighted, the AI businesses contribute only ~$170 billion — far below what the ~$1.77 trillion price implies. Owens calls xAI's economic moat "indeterminate" and the segment a "material threat of value destruction." The accounting backs the caution: xAI booked roughly $3.2 billion of revenue against a $6.355 billion operating loss in FY25 and is consuming billions in compute. The bull case on our /valuation/ page treats this optionality as the thing the multiple is buying; the bear case treats it as a trillion-dollar premium for an outcome whose most-likely-weighted result is mediocre and whose worst case burns capital. Both readings start from the same S-1 numbers.

That is Morningstar's stated view — Owens said investors will "have the opportunity to buy in at a more attractive price following the IPO" — but this site does not give investment advice, and the recommendation rests on a contestable premise. The "wait for a pullback" logic assumes the IPO price is inflated by roadshow scarcity that a post-IPO market will correct toward DCF fair value. That is reasonable if the dependency premium is sentiment. It is wrong if the premium is structural — if the market is correctly pricing an asset whose unremovability a cash-flow model cannot capture. Historically, founder-controlled infrastructure names with genuine moats have not reliably handed patient buyers a discount; some have, many have not. The honest framing is that "wait for the pullback" is a bet that Morningstar's frame (DCF) is the right one and this site's frame (scarcity premium) is the wrong one. Anyone acting on it is taking a side in exactly the disagreement this page exists to lay out. Do your own work and size to your own conviction.

Yes — on more than a bull might expect, which is part of why it is worth taking seriously. Morningstar credits the core launch-and-Starlink business at about $611 billion, validating the exact assets this site calls the moat; the disagreement is only about the AI premium on top. The agency also flags the same governance defects we document on our /people/ page — Musk's ~85% voting control through dual-class stock — and the same related-party concerns mapped on our /related-party/ page, noting the xAI merger "was not conducted at arm's length." So the bear case and the bull case are not arguing about the launch monopoly, the Starlink margin, or the control structure; they agree on all three. They are arguing about one thing: whether the dependency interlock those facts create is worth a trillion-dollar premium over the discounted cash flows. That narrowness is what makes the debate tractable — and what makes the $990 billion gap the cleanest single question in the offering.

You don't reconcile them; you choose a frame, because they measure different things. Morningstar's $780 billion is a discounted-cash-flow number — the present value of the cash the businesses are most likely to generate. The market's ~$1.77 trillion is a scarcity number — what buyers will pay to own infrastructure that four global systems cannot route around, priced at a fixed $135 with no range because there is no comparable to triangulate against. A DCF is the right tool for a company that competes; it is the wrong tool for one that has become a substrate, because it cannot price the option value of every counterparty contract awarded under no-alternative procurement, nor the cost to a regulator of removing the asset. The bear case is correct that the cash flows do not support $1.77 trillion. This site's case is that cash flows are not what the price is measuring. The $990 billion gap is the dependency premium itself — and whether you pay it is the whole trade. See /valuation/ and /dependencies/ for the other side of this exact argument.