SpaceX hits record low: US$124 and a Pentagon bet

SpaceX closed at an all-time low of US$123.99 on Friday, below its US$135 IPO price and down nearly 40% from the US$201.80 high it touched on June 16, capping a six-day losing streak before a modest premarket bounce to US$125.53 on Monday. The company went public on the Nasdaq under the ticker SPCX on June 12, 2026, in the largest IPO in history, raising roughly US$75 billion at a US$1.75 trillion valuation. For investors, the drop is a live case study in what happens when a story stock loses altitude: a company can execute one of the most successful debuts ever and still slide below its offer price six weeks later.

The potential catalyst investors are watching is a reported deal with the U.S. Defense Department to supply AI computing capacity. According to the Wall Street Journal, if the talks materialize, the arrangement would mirror agreements SpaceX already has with Anthropic and Google, which are projected to generate roughly US$26 billion in annual revenue once fully ramped. That figure is the crux of the bull case, and it points to the real question hanging over the stock. SpaceX is no longer being valued as a rocket company. It is being valued as an AI infrastructure company, and the market is deciding in real time whether that pivot is worth US$1.75 trillion.

What exactly happened

A record IPO that ran out of thrust

SpaceX's debut was historic by any measure, opening at US$160.95, a 19.2% premium to its US$135 offer price, and briefly pushing its market capitalization past US$2.6 trillion at the June 16 peak. Then it reversed. The stock gave back nearly 40% from that high and dipped under the IPO price, a swing that reflects sentiment and positioning more than any single disclosure. The most recent earnings report, released May 7, showed a quarterly loss of US$1.27 per share on US$4.69 billion in revenue, a reminder that the combined entity is currently unprofitable. Several institutional investors, including Atwood & Palmer and Marquette Asset Management, added small new positions in the second quarter, but nothing at a scale that would arrest the slide.

The AI pivot underneath the ticker

The reason the Pentagon talks matter so much is structural. SpaceX's AI computing business centers on the Colossus I and II data centers in Memphis, which arrived through the February 2026 all-stock merger with xAI, Elon Musk's AI venture. That merger transformed a profitable rocket-and-satellite operation into a loss-maker carrying enormous AI ambitions, including a longer-term plan to place solar-powered data centers in orbit. That orbital vision depends on Starship, the fully reusable rocket meant to cut the cost of reaching orbit from thousands of dollars per kilogram to hundreds. Starship's progress is not smooth: its most recent test flight was scrubbed after some engines failed to ignite, and the rescheduled attempt targets a full sequence of launch, stage separation, boostback, landing, and the deployment of 20 Starlink V3 satellites. The compute revenue is the near-term story; the rockets are the long-term one.

Market context

SpaceX's swoon plays out against a specific 2026 backdrop: a wave of mega-IPOs from companies whose valuations rest on AI infrastructure promises rather than current profits. SpaceX was the first of that trio to reach public markets, and its post-IPO reversal is being read as a referendum on how much investors will pay for AI optionality attached to a cash-generating core. Underneath the loss-making combined entity sits Starlink, which passed 10 million subscribers and generated US$11.39 billion in 2025 revenue at EBITDA margins above 60%, the engine funding both Starship development and the AI losses. The market is effectively pricing two companies at once: a profitable connectivity utility and a speculative AI-and-launch venture bolted on top.

Analysts remain constructive even as the stock falls, which is itself a signal worth weighing. Piper Sandler upgraded SPCX to a hold on July 17, and the broader consensus sits at Moderate Buy, with four Strong Buy ratings, 24 Buy, eight Hold, and one Sell. Recent price targets range widely, from Susquehanna's US$170 at Neutral to Needham's US$250 at Buy, with Wedbush at US$190, RBC at US$225, and a consensus average of US$234.78, nearly double where the stock trades now. That gap between price and target captures the central tension: Wall Street models the AI revenue as largely inevitable, while the market, for now, is demanding proof. The reported Defense deal is one piece of that proof, but the Anthropic and Google agreements underpinning the US$26 billion projection carry their own conditions, including compute commitments that can be terminated on short notice.

What this means for investors and business owners

1. A great IPO is not the same as a great entry point. SpaceX executed the largest, most oversubscribed debut in history and still trades below its offer price weeks later. The general principle is that the quality of a company and the quality of a purchase price are separate questions. Buying at a euphoric open, when a stock trades at a premium to an already-stretched IPO price, exposes investors to multiple compression even when the underlying business performs. Discipline on entry price protects returns more reliably than conviction in the story.
2. When valuation rests on future revenue, every catalyst becomes binary. The bull case hinges on roughly US$26 billion in projected annual AI revenue that has not yet fully ramped. That makes each data point, a Pentagon deal, a Starship test, an earnings print, a swing factor rather than a footnote. For investors, the lesson is that story stocks trade on expectations, so their volatility is structurally higher. For business owners, it is a reminder that valuations built on projected contracts are fragile until those contracts convert to recognized, recurring revenue.
3. Understand what you are actually buying inside a conglomerate. SPCX bundles a highly profitable Starlink utility with a loss-making AI-and-launch venture. An investor buying the ticker is buying both, whether they want to or not. The broader principle is that conglomerate structures blend cash cows and cash burners under one price, and the market's mood determines which side dominates the multiple on any given day. Before buying a diversified entity, separate the segments and ask which one you are really underwriting.
4. Terminable contracts are softer than they look. The AI revenue thesis leans on agreements like the Anthropic compute deal, which reporting indicates can be canceled on short notice. Revenue that a counterparty can walk away from quickly deserves a discount relative to locked-in, long-dated contracts. The strategic takeaway for any business is that the durability of a revenue stream, not just its size, determines its value. Investors should read the cancellation terms before capitalizing a projection into a valuation.
5. Analyst targets describe a destination, not a timeline. The consensus target near US$235 sits almost double the current price, yet the stock keeps falling. Price targets express where analysts believe a stock should trade if the thesis plays out, not when or whether it will. Treating a target as a promise invites disappointment. The disciplined use of analyst research is as an input to your own model, weighing the assumptions behind the number rather than anchoring to the number itself.

Sources

• SpaceX (SPCX) Stock Hits All-Time Low, Can a Pentagon Deal Turn It Around?
• SpaceX blasts off with a record-breaking $75 billion IPO
• SpaceX IPO: 8 Things To Know Before It Goes Public
• SpaceX IPO: Valuation, Timeline and Investment Options
• SpaceX IPO 2026: SPCX Stock Analysis, Valuation & Investment Thesis