After SpaceX (SPCX.US) rose more than 3% against the trend, Morgan Stanley reiterated its $300 price target! The grand vision of Starlink and AI computing power underpins the outlook for the stock price to double.

date
15:24 18/11/2026
avatar
GMT Eight
Starlink's profitability explains why SpaceX has been relatively resilient, but for the stock price to deliver its doubling potential, it still depends on whether the computing power leasing business can deliver capacity to customers on schedule and continue to earn rental income sufficient to cover equipment investment and financing costs.
As of the U.S. stock market close on November 17, Eastern Time, SpaceX (SPCX.US), the "AI + space exploration" super-giant founded and personally steered by Musk, rose 3.16% against the trend to close at $141.70, compared with declines of 1.64%, 1.03%, and 0.74% in the Dow, S&P 500, and Nasdaq Composite Index, respectively, that day. SpaceX's counter-trend gain came before NVIDIA Corporation released its earnings report, and shortly after NVIDIA Corporation announced explosive results and future outlook, this AI chip giant drove a broad after-hours rally among leaders across the AI computing power supply chain. SpaceX, which has recently stepped up its buildout of an AI computing power leasing platform, also showed after-hours gains. Shortly after NVIDIA Corporation announced exceptionally strong results, the team led by Morgan Stanley senior analyst Adam Jonas reiterated the firm's most optimistic "Buy" rating on SpaceX and its $300 price target, implying 111.7% potential upside based on Tuesday's closing price. SpaceX's strong share price performance over the past week and past month prompted Morgan Stanley to detail in this research report the widely debated market question of "why SpaceX's stock has been relatively resilient recently, and whether this divergence can persist over the long term?" The firm also put forward the long-term bullish view that "the negative correlation between SpaceX and the AI sector is expected to weaken and SpaceX is beginning to decouple from the broader market." In the view of the Morgan Stanley analyst team, the Starlink technology platform, focused on cutting-edge satellite internet communications, already contributes the majority of SpaceX Group's own operating profit, and even if investors continue to lower valuations for the AI trade theme as a whole, market forces have reason to price this business separately. In addition, of Morgan Stanley's $300 bullish target price, $170 still comes from enterprise AI, meaning that the efficient delivery of AI computing power resources to leasing and the AI financing prerequisites associated with AI computing infrastructure still determine how far the stock price can go. Morgan Stanley's latest bullish research report covering SpaceX shows that in the third quarter, the connectivity business centered on Starlink contributed $2.367 billion in operating profit, up 99% year over year, accounting for about 85% of the group's operating profit. Home broadband subscriptions and continued payments from enterprise and government customers give investors a basis to price the satellite network separately even if they lower valuations for the AI computing power business. This is also why Morgan Stanley assigns a per-share valuation of $111 to the connectivity business built around the Starlink satellite internet platform, while attributing $170 of the $300 target price to enterprise AI; in addition, the remaining $19 per share comes from two businesses: the space business (Space) contributes $8, including rocket launch and other operations; social platform X and AI assistant Grok together contribute $11. Morgan Stanley's sum-of-the-parts valuation is: space business $8 + Starlink-centered connectivity business $111 + X and Grok $11 + enterprise AI computing power leasing/sales business $170 = $300 per share. Based on Tuesday's U.S. closing price of $141.70, the target price implies about 112% upside. Starlink satellite internet-related strong profit growth trajectory explains why SpaceX has been relatively resilient, but for SpaceX to deliver the doubling in share price that Morgan Stanley favors, it still needs to deliver computing power efficiently and on schedule, and ensure that customer rent payments continue to cover equipment and financing costs. After a rebound, the Philadelphia Semiconductor Index remains in a bear market; SpaceX first showed resilience, then rose against the trend From November 5 to 11, the Philadelphia Semiconductor Index rebounded from 9,011 points to 10,564 points, gaining more than 17% in six days. By Monday, November 16, selling resumed, with the Dow, S&P 500, and Nasdaq Composite falling 1.41%, 1.92%, and 2.49%, respectively; the Nasdaq 100 fell 2.67%, the Philadelphia Semiconductor Index fell 5.28%, and SpaceX fell only 1.26% over the same period. On Tuesday, the Nasdaq 100 fell another 0.72%, while the Philadelphia Semiconductor Index edged up 0.34% to 9,642.88 points, still 34.11% below its historical closing high of 14,634.72 points on June 22 and having entered a technical bear market as early as July. The following charts in the Morgan Stanley analyst team's report provide a more meaningful comparison than a single-day counter-trend gain: as of November 17, SpaceX had fallen only about 1% over the past week and about 3% over the past month, compared with a roughly 28% decline for NVIDIA Corporation over the past month over the same period. These chart data are enough to show that SpaceX suffered smaller losses in this selloff, but the duration is not yet sufficient to assert that the long-term correlation between these two stocks, and between SpaceX and AI-related technology themes, has already changed. Asian markets bore more concentrated selling pressure in chip stocks and broader technology stocks. On Monday and Tuesday, Taiwan's weighted index closed down 6.03% and 2.31%, respectively; South Korea's KOSPI benchmark index, in which SK Hynix and Samsung Electronics account for more than 50% of market capitalization weight, fell 3.59% and 2.59%, respectively; and Japan's Nikkei 225 fell 2.42% and 1.34%, respectively, for cumulative two-day declines of 8.20%, 6.09%, and 3.73%. Uncertainty over NVIDIA Corporation's leadership transition and growing uncertainty surrounding large AI financing projects tied to NVIDIA Corporation due to Jensen Huang's resignation as CEO, a new round of U.S. chip export restrictions in the pipeline, and investors moving funds ahead of Anthropic's planned listing together affected the tone of the AI trade in Asia. Before NVIDIA Corporation's earnings disclosure, Goldman Sachs Group, Inc., Citi, and UBS Group AG respectively cut their NVIDIA Corporation target prices from $300, $285, and $300 to $220, $225, and $240, all maintaining "Buy" ratings. Goldman Sachs Group, Inc. was concerned about a decline in the valuation multiples investors are willing to pay, while UBS Group AG also lowered its 2028 revenue forecast, reflecting the potential real earnings impact of export reviews and delivery delays. NVIDIA Corporation's after-hours earnings alleviated demand concerns while raising the bar for the financing quality of computing power projects. Fiscal 2027 third-quarter revenue was $115.8 billion, up 103% year over year, with data center revenue of $107.6 billion, up 110% year over year. Vera Rubin contributed nearly $20 billion in revenue in its first mass-production shipping quarter. Fourth-quarter revenue guidance has a midpoint of $130 billion, above the market expectation of about $125 billion, includes an extra week, and excludes data center computing revenue from China. Even more significant, NVIDIA Corporation management raised its fiscal 2028 revenue growth forecast from 70% to about 75%. The board made clear it has no intention of asking Jensen Huang to return as CEO, while interim CEO Colette Kress said new investments and guarantees will be more prudent, with particular emphasis on whether customers can secure land, power, and data centers. NVIDIA Corporation rose nearly 5% after hours, while CoreWeave and Nebius at one point rose more than 7% and 9%, respectively. For SpaceX, robust GPU demand is favorable for computing power leasing, but whether it can energize on schedule, deliver computing power resources efficiently, and collect rent determines how much cash flow this round of demand can . How does Morgan Stanley derive the $300 target price? Starlink contributes current profit, while enterprise AI determines the valuation ceiling Starlink is the strongest financial basis for SpaceX's relative resilience. In the third quarter, connectivity business revenue was $5.248 billion, up 69% year over year, and operating profit was $2.367 billion, up 99% year over year, accounting for about 85% of the group's operating profit. Service fees charged to home, enterprise, and government customers give the company recurring revenue even when chip procurement cycles fluctuate. Morgan Stanley further estimates, based on expected 2030 operating levels, that a single fully loaded Starship launch of V3 broadband satellites, with about 5% of total capacity ultimately sellable, could support roughly $300 million to $400 million in annualized revenue, with an investment payback period of about 20 weeks as measured by incremental EBIT. This is the modeled return after satellites enter operation. What determines the speed of payback is whether new bandwidth can be deployed to regions with paying demand and how much capacity customers actually purchase. The $300 target price's reliance on enterprise AI is higher than the label of "star commercial space stock" suggests. Morgan Stanley's sum-of-the-parts valuation assigns $8 per share to the launch and space business, $111 to the connectivity business, $11 to X and Grok, and $170 to enterprise AI. Enterprise AI, already discounted by 50% for execution risk, still accounts for 56.7% of target value. The model forecasts through 2040, using a valuation date of September 30, 2027, a weighted average cost of capital of 11.3%, and a cost of equity of 12.1%. Based on the November 16 closing price of $137.36 cited on the first page of Morgan Stanley's research report, the target price corresponds to 118.4% upside. Based on a static estimate of approximately 13.581 billion Class A and B shares disclosed by SpaceX in October, the $300 target price implies an equity market value of about $4.07 trillion. Goldman Sachs Group, Inc. maintained "Buy" with a $245 target price over the same period, Citi maintained "Buy" and raised its target price from $240 to $280, and Cantor maintained "Overweight" and a $345 target price, corresponding to about 73%, 98%, and 143% upside based on Tuesday's closing price, respectively. SpaceX's share price is still near its issue price, but Morgan Stanley's judgment on its earnings outlook is more optimistic than in the early post-listing period. Based on the roughly $140 share price used in the report, SpaceX trades at about a 4% premium to its $135 issue price. At the same time, since early July, Morgan Stanley has raised its 2027 and 2028 earnings per share forecasts by about 80% and 60%, respectively, to $3.95 and $9.91. Morgan Stanley still leaves room in its computing power business: it expects capacity in operation of 8.3 gigawatts by the end of 2027, below Musk's 13-gigawatt target, and assumes capacity rises to 14 gigawatts by the end of 2028, while revenue per watt falls from $34 in the first quarter of 2027 to $21 in fiscal 2028. Under this set of assumptions, and excluding the one-time gain from the reversal of the valuation allowance on deferred tax assets, SpaceX corresponds to a 2028 P/E of about 17.5x, a slight premium to Morgan Stanley's estimated 16x for the S&P 500 Index. Visible Alpha's consensus market expectations are more optimistic about SpaceX's earnings growth outlook over the same period, so at the same share price and also excluding the tax gain, the P/E is about 11.2x. In other words, after excluding the one-time gain from the reversal of the valuation allowance on deferred tax assets, SpaceX's 2028 expected P/E is about 17.5x, slightly above the firm's estimated 16x for the S&P 500 Index over the same period; Visible Alpha expectations correspond to higher 2028 earnings, so at the same share price and also excluding the tax gain, SpaceX's expected P/E is about 11.2x. For investors, whether SpaceX is cheap relative to the broader market ultimately depends on whether new computing power can be delivered on schedule and achieve sufficiently high utilization so that profit growth is enough to absorb the impact of falling revenue per watt. Based on Morgan Stanley's own latest forecasts, investors are still paying a modest valuation premium for SpaceX. Only when the company delivers on the market's more optimistic earnings expectations will the relative discount implied by the current price be supported by fundamentals. Morgan Stanley also lists two factors that would change shareholder returns. The first is the Terafab wafer manufacturing plan. The U.S. tightening of advanced chip technology controls has increased the strategic value of domestic supply capability, but Morgan Stanley's model includes capital expenditure for the project of $15 billion in 2027, a cumulative $170 billion through 2030, and a cumulative $490 billion through 2035. Shareholders need to see where the money is going, capacity utilization, and the payback path. The second is SpaceX's future capital relationship with Tesla, Inc. According to Morgan Stanley's conversations with investors, even some bulls who agree with the industrial logic of a potential combination are holding off on building positions due to the timing of a transaction, the views of the independent committees of both boards, and potential equity dilution; Morgan Stanley lists this uncertainty as the biggest current holding concern. From ground data centers to space-orbit AI computing power ambitions, which SpaceX businesses are already generating strong profits? Ground computing power leasing already generates revenue, and profitability depends on whether the customer payment period can cover the equipment financing period. As of the end of September, SpaceX's rated computing capacity reached 1.7 gigawatts. The start of cloud service contract execution, together with Cursor's consolidation and Grok's growth, pushed quarterly revenue for the superintelligence segment (the SI segment focused on AI businesses) to $7.560 billion and operating profit to $431 million. Anthropic disclosed that its contract with SpaceXAI extends to 2029, with maximum spending of $120.5 billion, or about $3.1 billion per month once all capacity is delivered, but after the initial irrevocable period, either party can terminate with 90 days' notice. Correspondingly, SpaceX has obtained binding commitments for about $40 billion of Vera Rubin equipment finance leases. Put together, the investment implications become clearer: delivery of large AI rack leases is only the starting point for revenue; sustained utilization, customer renewals, and construction cost per watt determine how much return five-year equipment leases leave for shareholders. Rocket reuse first improves Starlink's economics and also provides launch capacity for orbital computing. The 15th Starship flight completed the first return of the upper stage from orbit and its capture by the launch tower, and deployed 26 V3 satellites; recovery of the first-stage booster still awaits a breakthrough. The previous first commercial orbital mission also deployed 26 V3 satellites, with new downlink capacity about ten times that of a Falcon 9 Starlink mission. If reuse turnaround accelerates, SpaceX can add sellable bandwidth at lower cost. SpaceX's recent focus on the smartphone direct-to-cell business expands the potential customer base, with satellites filling coverage gaps and ground-based 5G base stations involving Qualcomm handling capacity demand in densely populated areas; the two sides plan to begin deployment in the U.S. starting in 2027. The longer-term Starmind plan would send optimized Vera Rubin systems into sun-synchronous orbit, powered by CECEP Solar Energy and cooled by radiation. More than EUR60 million in orders have already been placed for the first space-use CECEP Solar Energy cell production line equipment, and SpaceX also purchased $151 million of Tesla, Inc. Megapack as fixed assets in the third quarter. Overall, in the view of the Morgan Stanley analyst team, Starlink already has profits, ground computing already has leases, and orbital computing still must use tests to verify heat dissipation, lifespan, and the in-orbit cost per unit of effective computing power. Morgan Stanley also views subsequent in-orbit tests and whether China and other participants increase investment as corroboration for testing this direction. This business structure also explains why, after looking through holdings, SpaceX remains an important part of the AI trade. At the end of the third quarter, NVIDIA Corporation increased its SpaceX stake by about 7.71 million shares, with a position value of $19.683 billion, accounting for 40.73% of its 13F portfolio and surpassing Intel Corporation to become its largest holding. SpaceX is also Alphabet's largest 13F holding, with a market value of $83.734 billion. For investors holding both NVIDIA Corporation and SpaceX, Starlink introduces operating profit from a different source, yet the two companies' computing power businesses still jointly depend on GPU deliveries, power, and end-customer payments. Next, the most telling indicator for verifying "decoupling" will be whether Starlink's operating profit can continue to grow when global equity market chip and semiconductor sectors come under pressure again, and whether new ground computing rent can cover new financing burdens.