SpaceX (SPCX.US) Starship reaches orbit for the first time, Wall Street unanimously bullish: Computing power leasing behind the scenes opens up valuation imagination

According to Zhitong Finance APP, on September 28, SpaceX (SPCX.US) conducted its 14th integrated test flight of Starship, launching from the Starbase in Texas. For the first time, Starship reached Earth's orbit and deployed 26 third-generation Starlink (Starlink V3) satellites using a device known as the "candy dispenser" on the spacecraft. The market is clearly focused on more than just this launch, as Wall Street investment banks have expressed bullish views.

On Monday, TD Cowen initiated coverage with a "Buy" rating and a $200 target price; together with RBC, Clear Street, and CLSA, the four institutions' target price range is between $200 and $250, corresponding to about 37% to 72% upside based on Monday’s closing price. During this round of bullish sentiment, what is being repeatedly mentioned is not rockets, but computing power leasing.

Starship Reaches Orbit for the First Time: All Satellites Deployed, Returns After Two Orbits

Reports indicate that this flight used the B21 booster and S41 spacecraft, both part of the third-generation Starship system; of the spaceship's six engines, one malfunctioned, and live commentators initially thought orbital insertion would be impossible. However, after reassessment, engineers confirmed an attempt could still be made. Approximately 25 minutes after launch, Starship achieved orbit and then spent roughly half an hour releasing the satellites one by one. The company established contact with all 26 satellites and Elon Musk confirmed that they were functioning normally.

The flight was not flawless. One engine prematurely shut down during ascent. The mission was originally planned to orbit Earth about six times at an altitude of 275 km for nearly 10 hours, but it was shortened to about 3 hours and only completed 2 orbits, with the remaining engines firing for a longer duration to compensate. The Super Heavy booster landed in the Gulf of Mexico, and no attempt was made to recover it with the launch tower’s robotic arms this time; the spacecraft performed its first deorbit burn and splashed down in the Pacific Ocean west of Chile.

The significance of this flight lies in the fact that, since the first integrated test in 2023, all previous 13 test flights had remained suborbital; reaching orbit is the basic threshold for launch vehicles. Aerospace engineer Dean Slayden called the delivery of 26 V3 satellites to their target orbit a "huge victory" and noted this marks Starship’s first revenue-generating commercial mission—these satellites are worth millions of dollars. Kathleen Curley, a senior analyst at Georgetown University’s Center for Security and Emerging Technology, was more reserved: the mission made significant progress but wasn’t a "super success," as SpaceX failed to achieve a 10-hour flight and an engine did malfunction “by SpaceX’s own standards and goals.”

Single-Satellite 1 Tbps: Leap in Capacity with V3

According to filings submitted by SpaceX to regulators, the V3 satellite is designed for a downlink capacity of about 1 Tbps, enough for each satellite to support around 10,000 households running 100 Mbps broadband simultaneously, and uplink capacity of 160 Gbps; compared to V2, downlink is increased about 10 times, uplink about 22 times, and the antenna supports 2,048 up/down beams (V2’s phased array supported 192 down and 144 up). Technical documentation from SpaceX states that a single Starship launch can deliver approximately 20 times the capacity of a Falcon 9 V2 launch into the constellation.

Weight is another barrier: each V3 weighs about 2,000 kg, double the previous generation; Falcon 9 cannot carry it. Third-party commentators are even more optimistic, though these are personal opinions: Rob Moller, Managing Partner at First Principles Group, stated on X, "With just today’s single Starship launch, SpaceX increased Starlink capacity in the past hour by more than the total added in the first year and a half of Starlink launches." Phil Bessel, former Apple executive and now Senior Director at Rivian, stated the user capacity design of a single V3 is 1.16 Tbps and inter-satellite optical links are 2.4 Tbps; ultimately, these links could create a highly routable backbone—routing more traffic and potentially bringing compute resources to space.

Musk hopes that once Starship is operating routinely, each launch will carry 60 V3 satellites; estimates say overall Starlink service will only markedly improve when about 1,000 V3s are in orbit, requiring at least 17 full-capacity launches. SpaceX has applied to the FCC to operate up to 100,000 V3 satellites; by September 2026, Starlink is expected to have launched about 13,000 satellites with about 11,000 in orbit, serving tens of millions of users in more than 160 countries and regions.

Wall Street Raises Price Target: Computing Power Is the Near-Term Engine

TD Cowen, led by analyst John Blackledge, initiated coverage with a "Buy" rating and a $200 target, stating SpaceX has "huge" opportunities in AI and space. The firm expects computing power leasing to contribute about 60% of revenue by 2027, with a 62% revenue CAGR from 2026-2031, and by 2031 LEO launches may approach 1,000.

TD Cowen also projects that around 35% of 2026 sales will come from computing-related business, which could overtake Starlink, the "crown jewel," as early as Q1 2027, making up 65% of overall revenue in 2028, with nearly half of planned compute capacity in the coming years to be leased to external companies. Deutsche Bank analyst Edison Yu gave a "Buy" rating and a $235 target price, estimating five signed clients represent an annualized revenue run rate of about $54.5 billion, and expects more large contracts as capacity comes online.

Other institutions moved on the same day: RBC reiterated "Outperform" and $225 price target, calling the 14th flight an important milestone for Starship launch capability and V3 deployment; Clear Street maintained "Buy" and $217 target, saying the mission enhances SpaceX’s high-capacity constellation capability but engine issues remind of execution risks; CLSA’s first coverage gave an "Outperform" rating with a $250 target; Bernstein SocGen maintained "Buy" and $248 target. According to Tipranks data, the average analyst price target is $233.

The details of compute contracts are key to understanding the current expectations. Anthropic pays $1.25 billion per month to rent compute from SpaceX’s Colossus data center in Memphis; Google Cloud’s compute contract is $920 million per month, effective next month; Reflection AI pays $150 million per month; an unnamed client will pay $1.11 billion monthly starting December—totaling about $3.4 billion per month, or $40.8 billion annualized, if fully executed. However, only three of the five disclosed clients have had their identities publicly confirmed; a fourth, revealed in July, is speculated by some analysts to be the US Department of Defense; these leases also carry "exceptionally short" 90-day break clauses, making it relatively easy for clients to exit.

The parent company’s performance provides a floor. According to its financial summary, SpaceX reported Q2 (August 4 release) revenue of $7.8 billion, up 92% year-over-year, exceeding the market expectation of $6.81 billion; adjusted EBITDA was $3.5 billion, up 191%; net loss was $541 million, with the AI segment bringing in $2.6 billion (up 247%). CFO Brett Johnson says the company is expected to reach $100 billion annual recurring revenue (ARR) by end of 2026; management plans to scale compute infrastructure to 15-20 GW by end of 2027, with Musk moving the goal of $1 trillion revenue from 2031 up to 2030.

This compute power story originates in a merger this February: SpaceX acquired xAI via an all-stock deal (worth about $250 billion, with a combined valuation of about $1.25 trillion), and in July the AI business was officially renamed SpaceXAI, with Colossus becoming a public asset. Musk made the logic clear—orbital data centers: the company has applied to the FCC for up to 1 million in-orbit compute satellites for AI, planning to launch via Starship as early as the end of next year.

On the Other Side: $1.9 Trillion Valuation, 20x Price-to-Sales Ratio, and Unlocking Wave

The other side of the stock is valuation. As of the September 28 close, SpaceX's total market cap was roughly $1.92 trillion; per the original text, the company's share price is down 10% YTD but up 1% for the month. Calculations show that 2025 revenue is around $18.7 billion and still at a loss, while Wall Street expects it to jump to $44.8 billion and $108.3 billion over the next two years—even if all is achieved, the 2027 price-to-sales ratio would be nearly 20x, which is not cheap for a capital-intensive firm, and the forward P/E is about 204x.

"Big short" Michael Burry stated at the company’s IPO that nothing in the prospectus supported even a $1 trillion valuation, later calling it "a sliver of space business, a niche telecom business, a struggling social media platform, plus a ‘lite version of CoreWeave.’" Share supply is a more pressing concern: the company only offered about 4% of shares in the IPO, and the lockup is expiring—one batch unlocked on September 24, with up to 328.4 million shares becoming tradable on both October 9 and 24, and potentially up to 1.3 billion shares flooding the market after the Q3 report; Musk himself is locked up through next June, but employees and institutional investors are not.

Financing costs are also rising: according to the same report, shortly after the IPO, the company issued $25 billion in bonds with coupons of 5.35% to 6.65%; with the Fed having just raised rates and signaling more hikes, future bond issues to fund compute and launch capacity construction will face a tighter path to profitability.

Retail investor sentiment is picking up. On Stocktwits, SPCX's retail sentiment index rose from "bearish" the previous day to "neutral," with message volume up 74% in 24 hours; some long-term holders commented "the money will come—every major house analyst confirmed it," and another user said the launch pad flames and splashdown fireball were "absolutely normal at this stage of testing," describing the 14th flight as a "historic success."

Trailing Rivals and NASA's Bet on Starship for Lunar Landing

The plight of competitors only highlights SpaceX’s position. Blue Origin's "New Glenn" had its first flight in January, but a May mission failed during powered landing and the rocket was destroyed, as was its only launch pad, pushing refly to year-end; ULA’s "Vulcan" has been grounded since February due to a booster anomaly. Amazon's Leo constellation last saw a successful deployment on July 2, with a total of 396 satellites launched—a stark contrast to Starlink's over 11,000 in orbit; its goal to begin commercial service by mid-2026 has already expired.

Even more important is NASA’s reliance. According to industry media estimates, Starship is the only manned lander for the Artemis lunar program, requiring 10 to 20 "refueling Starship" flights to transfer cryogenic methane and LOX in orbit—a process with no precedent to date; SpaceX has twice delayed orbiting propellant transfer demonstrations, with the latest target “no earlier than the end of 2026,” while Artemis III aims to conduct an unmanned lander orbital demonstration by the end of 2027, and Artemis IV plans the first crewed moon landing in early 2028—both dependent on this demonstration’s success.

Returning to Musk’s timetable: on Sunday he wrote, “Starship is 2 to 3 years from hourly flights”; the longer-term goal is 10,000 launches per year by 2030—as a reference, Falcon 9 flew 165 times in 2025, already a company record, accounting for about half of the world's orbital launches that year. TD Cowen’s projection of nearly 1,000 LEO launches by 2031 equates to "a bit less than three launches a day." The outcome of the engine anomaly investigation will determine if Flight 15 can attempt catching the ship with the tower robotic arms as planned.