Damo: The SpaceX (SPCX.US) three-quarter report is just a “starter”; Starship 15 and the “physically intelligent economy” are the real catalysts for stock prices

The Zhitong Finance App learned that on the eve of SpaceX (SPCX.US)'s third quarter earnings report due to be released after the US stock market on Tuesday EST, Morgan Stanley published a research report to maintain SpaceX's “gain” rating and a target price of 300 US dollars. Analyst Adam Jonas's team believes that the core value of SpaceX being underestimated by investors is not that the fineness of its single quarter results exceeds expectations, but that it is playing a central role in building a “physical intelligence economy (PIE)” (Physical Intelligence Economy (PIE)).

The bank believes that the final size of PIE may be several times the global GDP and has a significant impact on national sovereignty; therefore, the direction of SpaceX's stock price will depend less on the subtle results of the three-quarter report and more on whether PIE, which is still fermenting in Musk's plan, can release some substantial progress.

Third quarter report forecast: revenue and computing power indicators are the focus

Morgan Stanley estimates that SpaceX's total revenue for the third quarter of 2026 will be US$13.05 billion, slightly higher than the market's consensus estimate of US$12.810 billion; profit before interest and tax (EBIT) is expected to be US$2,363 million, and the market forecast is US$2,396 billion; the adjusted EBITDA is expected to be US$6.931 billion, and the market forecast is US$6.837 billion; the adjusted diluted earnings per share is expected to be US$0.19, and the market forecast is US$0.16.

By business, Morgan Stanley expects revenue from the space business to be 1,570 billion US dollars, with market expectations of 1,458 billion US dollars; connectivity business revenue of 4.857 billion US dollars, market expectations of 4.917 billion US dollars; AI business revenue of 6.578 billion US dollars, and market expectations of 6.435 billion US dollars. Among them, the AI business includes advertising revenue of US$699 million and AI solutions and infrastructure revenue of US$5.879 billion. In terms of profit, the report predicts that space business EBIT will still lose 507 million US dollars, connectivity business EBIT will be 1,866 million US dollars, and AI business EBIT will be 1.05 billion US dollars. The gross margin is expected to be 56.3%, and the operating margin is 18.2%.

In terms of key operating indicators, the market expects consumer Starlink subscribers to be 14.293 million, and Morgan Stanley expects 14.288 million; the consumer monthly ARPU market is expected to be 66.74 US dollars, and Morgan Stanley is expected to be 63.64 US dollars; at the end of the period, the computing power market expects 1.7 GW, and Morgan Stanley expects 1.62 GW; revenue per watt (Rev/Watt, annualized by average computing power). The market is expected to be $16.61, and Morgan Stanley is expected to be $17.42.

The bank emphasized that revenue per watt is probably one of the most important and difficult variables to model in SpaceX's earnings report because the new cloud transaction contract period is short (about 6 months), and the deductive economy of the internal model is still unclear.

Investor sentiment: Positive bias, but extra-quarterly suspensions weigh on valuations

The report points out that investors' position sentiment before the three-quarter report was like a “roller coaster,” making it difficult to make accurate judgments. Overall, however, investors are slightly positive about the performance figures: computing power pricing is only rising or not falling, and market expectations may not fully reflect the value brought by Cursor, Grok Bot, and the announced new cloud cooperation orders.

However, factors other than quarterly results are still holding back quite a few investors from turning more active. The bank specifically mentioned that SpaceX's stock price soared from less than $150 to $172 in about two weeks. There was not much news during this period, and the total number of questions received from investors during this period was comparable to the number of questions received last Friday on the eve of the financial report, indicating that market sentiment is sensitive and highly divided.

In terms of strategy, Morgan Stanley bluntly stated that “market complacency is an opportunity.” According to the report, the rest of the year will enter a catalyst-rich, upward path. The starting point is Starship's 15th test flight and possible spacecraft capture (using launch tower robotic arms to capture the upper stages of the Starship), which is expected to be in the first half of November. This is probably SpaceX's biggest catalyst since its IPO, and is expected to reduce the risk of potential market space (TAM) for each major business.

In terms of valuation, based on the current stock price of 146.03 US dollars, Morgan Stanley believes that SpaceX's stock price has included 127 US dollars/share of space and connectivity business opportunities in its model, covering almost all or even all of this value, leaving only 19 US dollars/share left for the AI business. This is equivalent to about 2 times the 2028 EV/Sales valuation multiplier for the new cloud business.

Core Disagreement: Will the computing power be 5GW or 10GW by the end of 2027?

The report shows that investors lack confidence that SpaceX's computing power capacity will be “closer to 10GW than 5GW” by the end of 2027, mainly because it is uncertain whether future data center site selection, electricity acquisition, and construction will be faster than the industry. The seller's unanimous expectation was 6 GW, the Morgan Stanley model was 4.9 GW, and the buyer's expectations were also roughly in the 4 to 6 GW range. Although Musk said “closer to 10 than 5” during the last earnings call, the report said that no firm bets have been heard from customers approaching 10GW.

The report also pointed out that many investors are still used to viewing SpaceX's business in isolation according to their own background (aerospace and defense, Internet, telecommunications), and are just beginning to understand the interconnectedness between businesses. Other than a few investors with close ties to Musk's companies, few seriously discuss “physical AI” as a future opportunity, although Morgan Stanley believes this is the most obvious TAM expansion direction for AI and satellite communications. The question about Terafab is also still limited to estimating the company's recent capital expenditure burden.

In terms of the new cloud business, the bank said that the recent transaction price is about 30 to 50 US dollars per watt, and the industry pricing index will still rise until the end of the year; although hyperscale cloud vendors and the new cloud business are expected to bring in more than 30 GW of incremental computing power capacity in 2027, which will help balance supply and demand. The company may lease up to half of the 2GW of computing power to Xinyun customers at the end of 2026. Two new deals were previously announced in the second quarter earnings report and industry conference; as the computing power rises above 5GW by the end of 2027, subsequent transaction announcements are worth paying attention to. The report also mentioned that the company mentioned many times that the payback period for computing power investment in the new cloud business is about 1 year, which indicates that the supply and demand for computing power will still be quite tight until next year.

Earnings Conference Call and Follow-up Catalysts: Starship, Starlink V3, Grok and More New Cloud Businesses

The report predicts that SpaceX management will update the verbal guidance from the second quarter conference call: capital expenditure for the fourth quarter is expected to be close to the third quarter; computing power is “higher than 2 GW” at the end of the year; the initial guidance for computing power by the end of 2027 is 5 to 10 GW; Musk previously said it was “closer to 10 than 5.” The company may also provide details such as the date and target of the 15th flight of the Starship, and the pace of deployment of the Starlink V3 satellite through Starship. The report estimates that each Starship launch is equivalent to the Falcon 9 Starlink launch capacity for more than 1 to 2 months, depending on the loading ratio.

During the conference call, the AI business was the focus of observation. The third quarter saw a number of AI milestones, including the completion of the Cursor acquisition, the release of Grok Bot, and significant improvements in Grok 4.6 and 4.7 to move closer to the cutting edge of the model. This quarter may be the first time investors can actually observe and infer economic benefits. The bank recommended focusing on Grok and Cursor's user growth, token calls, potential new enterprise transactions, and penetration into existing Cursor customers — about two-thirds of Cursor customers are Fortune 500 companies.

On the Starlink side, the report predicts that after the 14th test flight and deployment of the first batch of 26 V3 satellites, the company will provide more details of the V3 plan, including the pace of deployment and balance with the V2 mobile satellite. SpaceX recently announced the purchase of 800MHz spectrum from Grain Management, pending approval from the US Federal Communications Commission (FCC). Morgan Stanley believes that the company's mobile phone directly connected to the satellite business model is still relatively uncertain. It is expected that management will provide incremental information, but they may intend to remain open to reserve the right to choose.

On the Starship side, the 15th flight is expected to take place in early November, and the market is concerned about whether to test spacecraft capture. If successful, this may be SpaceX's biggest stock price catalyst since its IPO, and also a major verification point for reusability, paving the way for the company to increase its launch pace in 2027. The company may also give general expectations for future test flights, but there isn't necessarily a clear timeline, including the first simultaneous capture of a spacecraft and booster in the same flight, the first launch from Florida, and plans to deploy more Starlink satellites. The 16th flight is expected to take place before the end of the year; if the 15th flight does not attempt capture, the 16th flight may be the first capture opportunity; if the 15th flight succeeds, it is possible to see the first time that the same flight captures the spacecraft and booster.

Additionally, Grok 4.8, 4.9, and 5.0 are expected to be released before the end of the year. Musk said on X on September 14 that Grok 4.8 will be a significant improvement over 4.7. 4.9 “may reach the current cutting-edge model level”, and 5.0 “may be better than anything, wait and see.” More new cloud deals are also seen as potential catalysts: the Morgan Stanley model expects SpaceX's computing power to be around 2.1 GW by the end of the year and rise to about 5 GW in 2027, opening up space for more deals.

Funding, lockdown, and other concerns

The report sees the need for additional financing as a key overhang in market sentiment, particularly the recent new cloud vendor debt incident. The Morgan Stanley model expects SpaceX to issue 60 billion, 80 billion, and 90 billion US dollars of bonds in 2027, 2028, and 2029, respectively. Currently, it is expected to issue ordinary investment grade bonds, but the company has various financing options. The report says this is the biggest risk of its model, and it doesn't want to downplay this; however, if there is a hypothetical scenario where SpaceX is unable to finance, it may have a broader impact on the economy and the rest of the investor mix.

In terms of the lockdown period, investors no longer regard the release of the ban on restricted shares in batches as a core concern, because the impact of the lifting of the ban so far has been relatively limited. The ban on 1.3 billion shares was lifted after the third quarter earnings report, and a total of 2.4 billion shares were lifted before the end of the year; however, more than 20% of these may have been absorbed by passive index purchases, and may even be more than the shares actually sold.

The report also mentioned that there is no more new information about the new Roadster jointly developed by Tesla and SpaceX other than admitting that the presentation was impressive. This is the first major product jointly designed by the two companies, and it is also a major demonstration of the engineering capabilities and potential for collaboration between the two companies.