SpaceX’s US$2 Trillion Valuation Test: Can Starlink’s Profits, the AI Supercycle and Orbital Computing Justify the Price?

Author’s Note and Disclaimer:
Zion Zhao Real Estate | 88844623 | 狮家社小赵 | wa.me/6588844623 |  https://linktr.ee/zionzhao | https://www.huttonsgroup.com/vcard/R071443Z

Legal Disclaimer

This publication reflects the independent views of Zion Zhao and is provided for general educational and informational purposes only. It does not constitute property, investment, financial, mortgage, tax or legal advice, nor any guarantee of returns or outcomes. Information, forecasts and third-party sources may be incomplete, inaccurate or subject to change without notice. Readers should independently verify all material facts and seek appropriately qualified professional advice before making decisions. To the fullest extent permitted by Singapore law, the author makes no warranties and disclaims liability for losses arising from reliance on this material. All rights reserved.

SpaceX After the IPO: US$18.4 Billion in Quarterly Capital Spending, an Expanding AI Empire and the Risks Behind a US$2 Trillion Valuation

SpaceX After the IPO: A US$2 Trillion Bet on Rockets, Artificial Intelligence and the Future of Computing. Is SPCX a Buying Opportunity or a Valuation Trap?

An Investment Perspective on the AI Supercycle, Starlink's Cash-Generating Power, the US$60 Billion Cursor Acquisition, the Economics of Orbital Computing and the Price of Technological Ambition.

By Zion Zhao | Investment Research, Macroeconomics and Strategic Asset Allocation

16 October 2026 | All financial figures are in US dollars unless otherwise stated.

The Investment Paradox: An Extraordinary Company Is Not Automatically an Extraordinary Stock

SpaceX may be building the infrastructure of tomorrow, but Wall Street is already asking investors to pay for a substantial part of that future today.

That is the central dilemma surrounding Space Exploration Technologies Corp. (NASDAQ: SPCX), following its historic June 2026 initial public offering and subsequent correction from its early trading highs.

What began as a company pioneering reusable rockets has evolved into an increasingly integrated technology conglomerate spanning satellite communications, artificial intelligence, cloud computing, software development, advanced manufacturing and potentially orbital data centres.

The transformation is extraordinary. So are the financial commitments required to make it work.

SpaceX's second-quarter 2026 revenue surged 92% year over year to US$7.81 billion. Its Starlink-led Connectivity division remains highly profitable, while its emerging artificial intelligence business has secured substantial external computing contracts. Yet the company also reported a US$143 million GAAP operating loss and invested US$18.37 billion in capital expenditure during a single quarter (Space Exploration Technologies Corp., 2026a, 2026b).

This is simultaneously a remarkable growth story and an exceptionally capital-intensive investment proposition.

The market's enthusiasm is understandable. SpaceX offers exposure to two powerful structural trends: the commercialisation of space and the global artificial intelligence infrastructure supercycle.

But investors must confront an uncomfortable distinction. A company can possess exceptional technology, enormous commercial opportunities and talented management while still representing an unattractive investment at an excessive valuation.

SpaceX began public trading on 12 June 2026 at an IPO price of US$135 per share. By 15 October, its shares closed at US$148.62, approximately 10.1% above the offering price despite retreating substantially from their post-listing peaks (Space Exploration Technologies Corp., 2026b; Stock Analysis, 2026).

The correction therefore should not automatically be mistaken for undervaluation.

In my assessment, SpaceX deserves serious long-term consideration. However, its investment merits should be judged through business fundamentals, cash-flow economics, capital allocation and valuation rather than technological excitement alone.

The crucial question is not whether SpaceX can continue innovating.

It is whether the economic returns on those innovations will ultimately justify a market valuation approaching US$2 trillion.

From Rockets to AI: The Strategic Logic Behind the xAI Merger

The February 2026 acquisition of xAI marked a decisive transformation in SpaceX's corporate identity.

The all-stock transaction, which combined SpaceX and Elon Musk's artificial intelligence business at an estimated US$1.25 trillion pre-IPO valuation, initially appeared unconventional. A rocket manufacturer acquiring an AI laboratory and a social media platform hardly resembles traditional aerospace industry consolidation (Reuters, 2026a).

However, the strategic rationale becomes clearer when the transaction is examined through vertical integration.

SpaceX already possessed substantial capabilities in launch systems, satellite manufacturing, communications infrastructure and advanced engineering. xAI contributed AI models, engineering talent, GPU-based computing facilities, the Grok platform and X's digital distribution ecosystem.

Together, these capabilities potentially establish an integrated technology architecture connecting computing infrastructure, AI software, communications networks and orbital deployment.

This matters because artificial intelligence is increasingly becoming an infrastructure industry.

Developing advanced AI systems requires not only algorithms and engineering expertise, but also access to semiconductors, electricity, cooling, networking equipment, land, financing and dependable computational capacity.

Controlling more of that infrastructure can reduce dependence on external providers and improve operational coordination.

SpaceX is therefore pursuing something considerably larger than a traditional aerospace diversification strategy.

It is attempting to integrate significant parts of the technological supply chain that may support future computing, communications and autonomous systems.

The potential commercial advantages are substantial.

Proprietary infrastructure could support AI model development, improve software engineering efficiency and generate third-party cloud revenue. Starlink could provide communications infrastructure, while SpaceX's launch capabilities could eventually deploy specialised AI hardware into orbit.

Yet vertical integration is not automatically value-accretive.

Owning infrastructure also means assuming its operating costs, equipment depreciation, financial obligations and technological obsolescence risks.

A company leasing computing capacity can potentially reduce purchases when demand weakens. A company owning billions of dollars of data-centre equipment must continue managing those assets even when utilisation deteriorates.

There is also a distinction between genuine external revenue and internal demand generated by related businesses. Moving expenditure between subsidiaries does not create incremental consolidated revenue.

Ultimately, SpaceX must demonstrate that combining these operations produces greater economic value than the businesses could have generated independently.

The strategy is coherent. Whether the financial returns justify the integration costs remains an open question.

Why SpaceX Paid US$60 Billion for Cursor

If the xAI merger established SpaceX's position in artificial intelligence, the acquisition of Cursor signalled its intention to compete further up the technology value chain.

On 14 August 2026, SpaceX completed its acquisition of Anysphere, the company behind the AI coding platform Cursor, through an equity transaction based on an implied US$60 billion valuation (Space Exploration Technologies Corp., 2026c).

The acquisition was not simply a purchase of software capabilities.

It potentially connected three complementary components of the AI economy: computing infrastructure, foundation models and commercial applications.

Cursor provides AI-assisted software development tools that can accelerate coding, debugging and engineering workflows. Such capabilities are commercially valuable to external developers and potentially useful within SpaceX's increasingly complex engineering operations.

More importantly, a successful software platform could create recurring demand for the underlying AI infrastructure.

That addresses one of the central problems facing neocloud operators.

Building GPU facilities is expensive, but financial success requires those facilities to remain economically productive.

Acquiring a software platform with established customer relationships potentially creates an additional distribution channel for AI services and computational resources.

SpaceX could therefore benefit from infrastructure revenue, AI model usage and software subscriptions within a more integrated commercial ecosystem.

However, the US$60 billion valuation creates a formidable financial hurdle.

An all-stock transaction avoids an equivalent immediate cash outlay, but it does not eliminate economic cost. Existing shareholders experience dilution, and the acquired business must generate sufficient incremental value to justify the equity issued.

For perspective, a hypothetical 10% annual after-tax cash return on US$60 billion would require approximately US$6 billion in sustainable annual cash generation attributable to that investment.

This is an illustrative investment hurdle rather than a forecast of Cursor's future performance.

The acquisition also introduces competitive and governance questions. Cursor must retain developers, sustain product differentiation and remain commercially attractive despite intense competition from other AI coding platforms.

Its value will depend on pricing power, customer retention, software margins, inference costs and the durability of its development ecosystem.

SpaceX has purchased a potentially valuable demand engine for its AI infrastructure. But paying US$60 billion for that opportunity means execution must be exceptionally strong.

Strategic synergy does not excuse undisciplined acquisition economics.

The Neocloud Transformation: Why Google Matters More Than the Hype

Perhaps the most important development in SpaceX's investment narrative is its emergence as a major AI computing infrastructure provider.

Neoclouds specialise in supplying GPU-based computing capacity to AI developers and enterprises. Unlike established hyperscalers such as Amazon Web Services, Microsoft Azure and Google Cloud, they are frequently more concentrated in accelerated computing infrastructure.

Their commercial proposition is straightforward. They invest heavily in GPU clusters and supporting infrastructure, then monetise that capacity through customer contracts.

SpaceX increasingly resembles a diversified hyperscaler with a substantial neocloud business embedded within it.

Its agreement with Google provides concrete evidence that the infrastructure business has meaningful external demand.

Under a cloud-services agreement disclosed in June 2026, Google agreed to pay SpaceX US$920 million per month from October 2026 through June 2029 for access to approximately 110,000 NVIDIA GPUs and associated computing resources, subject to contractual delivery conditions and termination provisions (Space Exploration Technologies Corp., 2026d).

At the stated monthly rate, the contract represents approximately US$11.04 billion on an annualised basis.

That is a substantial commercial commitment.

However, annualised contractual payments are not equivalent to guaranteed recognised revenue. The Google agreement includes capacity-delivery requirements, possible fee adjustments and termination rights, including a provision allowing termination on 90 days' notice after December 2026.

This distinction matters enormously.

AI infrastructure companies can announce significant future contracted revenue while remaining exposed to delivery delays, customer concentration and renegotiation risks.

A large commercial contract validates demand, but it does not guarantee the profitability of the underlying assets.

SpaceX's second-quarter reporting showed US$2.56 billion of total AI segment revenue, including approximately US$2.19 billion classified as AI solutions and infrastructure. That reporting category includes cloud services alongside subscriptions, licensing and AI application revenues (Space Exploration Technologies Corp., 2026a, 2026b).

Consequently, the full amount should not be mistaken for pure GPU-leasing revenue.

The original investment narrative also references a projected US$41 billion annualised AI infrastructure revenue run rate by the end of 2026.

Although the company's expanding contractual commitments make substantial revenue growth plausible, that exact projection could not be independently reconciled with a definitive regulatory guidance disclosure reviewed for this analysis.

It should therefore remain a forward-looking estimate, not an established financial result.

Annualising total second-quarter AI revenue produces approximately US$10.24 billion, which is materially different from a US$41 billion exit run rate.

Reaching such an exit rate would require significant additional capacity deployment, customer monetisation and revenue recognition.

The fundamental investment question is whether SpaceX can earn attractive returns on its AI infrastructure after accounting for depreciation, equipment replacement, electricity, financing and utilisation risk.

Revenue growth is necessary, but it is not sufficient.

The critical metric is not simply how many GPUs SpaceX owns. It is how much sustainable free cash flow those GPUs can produce over their economic lives.

The 10-Gigawatt Ambition: AI's Next Bottleneck Is Electricity

SpaceX's infrastructure ambitions extend well beyond its current operating capacity.

The company reported approximately 1.4 gigawatts of nameplate compute capacity at the end of June 2026, compared with 1.0 gigawatt in March and 0.4 gigawatts a year earlier.

During its August earnings discussion, management outlined ambitions to approach 10 gigawatts of computing capacity by the end of 2027, supported by substantial development of electricity generation, cooling and associated infrastructure (Space Exploration Technologies Corp., 2026a, 2026e).

Moving from 1.4 gigawatts to approximately 10 gigawatts would represent an expansion of more than seven times.

That ambition is commercially significant but operationally demanding.

At full continuous utilisation, 10 gigawatts of electrical load would correspond to approximately 87.6 terawatt-hours of annual electricity consumption, although nameplate capacity should not be confused with actual consumption or facility-wide energy demand.

The scale illustrates why electricity is becoming a defining constraint in the AI investment supercycle.

The International Energy Agency estimated that global data-centre electricity consumption reached approximately 485 terawatt-hours in 2025 and could approach 950 terawatt-hours by 2030 under its central outlook (International Energy Agency [IEA], 2026).

The challenge is not simply generating additional electricity.

It is obtaining dependable grid connections, installing transformers, securing suitable cooling systems, developing transmission infrastructure and coordinating construction schedules with computing equipment deliveries.

Power availability increasingly determines how quickly AI infrastructure can become operational.

This helps explain SpaceX's interest in greater control over on-site electricity generation, including potential in-house turbine development.

Such integration could reduce dependence on constrained equipment supply chains and improve coordination between data-centre construction and energy provisioning.

However, manufacturing turbines introduces its own industrial challenges.

Established manufacturers such as GE Vernova possess substantial engineering expertise, manufacturing capabilities and service networks.

SpaceX's propulsion and manufacturing experience may provide transferable advantages, but success in aerospace engineering does not automatically establish cost competitiveness in utility-scale power equipment.

The broader implication is that the AI supercycle is creating investment opportunities beyond semiconductor manufacturers.

Electricity generation, grid infrastructure, industrial equipment, cooling systems and engineering services are becoming increasingly important components of the digital economy.

Yet the same expansion introduces risks.

Data centres experiencing unreliable power, delayed energisation or insufficient cooling may fail to deliver contractual computing capacity despite having expensive hardware installed.

In a business where customers pay for available and reliable computation, electricity disruptions can become financial disruptions.

The next competitive advantage in AI may belong not merely to companies that can purchase advanced chips, but to those capable of powering and operating them reliably at scale.

Orbital Data Centres: Revolutionary Vision or Premature Economics?

SpaceX's most ambitious initiative involves extending artificial intelligence infrastructure into orbit.

In August 2026, NVIDIA announced that SpaceXAI was developing a first-generation Starmind AI satellite based on an optimised version of its Vera Rubin NVL72 architecture (NVIDIA Corporation, 2026).

This is a genuine development programme, not merely an unsupported technological concept.

Nevertheless, the distinction between engineering feasibility and commercial viability is fundamental.

Computing systems in space face constraints that conventional terrestrial data centres do not.

Electricity must be generated or stored onboard. High-performance processors produce substantial waste heat, which must principally be rejected through thermal radiation because space lacks the surrounding atmosphere needed for ordinary convective cooling.

Communications also present difficulties.

Training large AI models requires rapid movement of substantial quantities of data. Transmitting information between orbital facilities and Earth introduces bandwidth, latency and ground-station constraints.

The economics become even more challenging when hardware obsolescence is considered.

A terrestrial data centre can upgrade computing equipment through established supply chains. Replacing advanced processors in orbit may require new launches, specialised servicing or entire spacecraft replacement.

Recent technical research reinforces these concerns.

Turyshev (2026), in a research preprint examining orbital computing economics, found that solar generation, radiator requirements, launch costs, communications limitations and equipment replacement cycles impose substantial barriers to competitiveness with terrestrial infrastructure.

The analysis suggests that specialised space-native processing may become commercially useful sooner than general-purpose orbital computing for customers on Earth.

Because the study is a preprint, its numerical assumptions should not be regarded as settled peer-reviewed conclusions. Nevertheless, the physical constraints it examines are material.

There are also environmental and regulatory considerations.

Expanding orbital infrastructure increases the importance of collision avoidance, debris management and end-of-life spacecraft disposal. The foundational research of Kessler and Cour-Palais (1978) established how increasing satellite populations can amplify collision risks and contribute to cascading debris formation.

SpaceX's reusable launch technology could provide meaningful deployment advantages if future launch costs decline sufficiently.

However, lower launch costs alone cannot solve every challenge associated with orbital computing.

In my assessment, the nearer-term commercial applications are likely to involve satellite autonomy, onboard image analysis, communications optimisation and specialised edge computing.

These applications may benefit from processing information close to where it is generated.

General-purpose orbital AI data centres capable of competing economically with terrestrial facilities represent a significantly more demanding proposition.

SpaceX's proposed demonstrations from 2027 onward should therefore be assessed as technical milestones rather than evidence of immediate commercial profitability.

Orbital AI deserves serious research attention, but not an indiscriminate valuation premium.

For investment purposes, it is more appropriate to treat the technology as a long-duration strategic option until operating performance and economic viability are demonstrated.

The Financial Reality: Strong Revenue Growth, Uneven Profitability

SpaceX's second-quarter financial results reveal a company expanding rapidly while carrying significant investment costs.

Consolidated revenue reached US$7.814 billion, up from US$4.071 billion a year earlier.

The company reported a GAAP operating loss of approximately US$143 million and a net loss of US$541 million.

Importantly, the operating loss narrowed from US$1.943 billion in the immediately preceding quarter and US$970 million in the corresponding prior-year period.

It would therefore be inaccurate to characterise the latest quarter as evidence of uniformly deteriorating operating profitability (Space Exploration Technologies Corp., 2026a).

The segment breakdown is particularly revealing.

Connectivity generated US$4.291 billion in revenue and US$1.656 billion in operating income.

AI contributed US$2.561 billion in revenue but recorded an operating loss of US$1.257 billion.

The Space division generated US$962 million in revenue and an operating loss of US$542 million.

These results demonstrate that SpaceX has an established profitable commercial foundation, but its most ambitious expansion businesses continue consuming substantial resources.

The Space segment's reported revenue also understates the strategic importance of its launch capabilities because launches supporting the group's own satellite operations do not necessarily create consolidated external revenue.

That is one reason conventional aerospace valuation frameworks are insufficient for assessing SpaceX.

The more revealing accounting distinction concerns adjusted EBITDA.

SpaceX reported approximately US$3.538 billion in consolidated adjusted EBITDA during the quarter, including US$1.146 billion from AI.

The AI segment's positive adjusted EBITDA was encouraging, particularly following negative adjusted EBITDA in the preceding quarter.

But investors should examine what the metric excludes.

The segment reported approximately US$1.885 billion in depreciation and amortisation and US$516 million in share-based compensation adjustments.

Removing these expenses substantially improves reported adjusted profitability.

That does not mean the underlying costs are economically irrelevant.

Depreciation reflects the accounting consumption of previously acquired assets. Share-based compensation can dilute shareholders. Meanwhile, replacement expenditure eventually requires additional financing or operating cash flow.

Adjusted EBITDA is useful for evaluating certain aspects of operating performance, but it is not a substitute for free cash flow.

A data-centre business cannot ultimately reward shareholders merely by reporting strong adjusted earnings while continuously spending more cash than it generates.

The investment case strengthens when revenue growth converts into sustainable operating profits and free cash flow, not simply when non-GAAP profitability improves.

US$18.4 Billion in Quarterly CapEx: The Most Important Number in the Investment Thesis

SpaceX's most striking financial figure may not be its revenue growth.

It may be its capital expenditure.

During the second quarter of 2026, the company invested approximately US$18.37 billion, compared with US$10.11 billion in the preceding quarter.

Of that amount, US$15.83 billion was allocated to AI, representing approximately 86% of total quarterly capital expenditure (Space Exploration Technologies Corp., 2026a).

This illustrates the extent to which SpaceX's investment priorities have shifted toward computing infrastructure.

The company is increasingly exposed to the same structural risks affecting specialist neocloud operators such as CoreWeave and Nebius.

Those risks include high upfront investment requirements, hardware obsolescence, customer concentration, competitive pricing pressure and potential refinancing difficulties.

SpaceX benefits from diversified operations and significant financial resources, but its data-centre business is not immune to the broader economics of AI infrastructure.

Indeed, its first-half 2026 cash-flow statement deserves careful consideration.

SpaceX generated approximately US$3.47 billion in operating cash flow during the first six months of the year.

Over the same period, capital expenditure reached approximately US$28.48 billion.

Subtracting capital expenditure from operating cash flow produces a simplified free cash flow deficit of approximately US$25.01 billion (Space Exploration Technologies Corp., 2026b).

The calculation does not capture every possible financing arrangement, but it clearly demonstrates that the company's investment programme was not yet internally self-funding.

To be fair, growth capital expenditure must be distinguished from maintenance expenditure.

Constructing infrastructure today can support revenue and cash generation over many future years.

However, the quality of that investment depends on utilisation, contract durability, operating costs and asset replacement requirements.

AI computing equipment is particularly exposed to rapid technological change.

New generations of accelerators may offer substantially better performance per watt and lower costs per unit of computation, potentially reducing the commercial competitiveness of older facilities.

This creates an important paradox.

The faster AI technology improves, the greater the opportunity for AI adoption. But the same progress can accelerate the economic obsolescence of existing infrastructure.

For a company investing tens of billions of dollars, that tension is central to future shareholder returns.

The question is whether SpaceX can recover its investment and earn an appropriate return before existing equipment loses a substantial portion of its economic competitiveness.

Starlink: The Profitable Engine Behind the Bigger Ambition

Amid the excitement surrounding AI, Starlink remains central to SpaceX's financial investment case.

The Connectivity segment generated approximately US$4.29 billion in quarterly revenue and US$1.66 billion in GAAP operating income during Q2 2026.

Its operating margin approached 39%, demonstrating considerably stronger established profitability than the group's AI and Space divisions.

Starlink's subscriber base reached approximately 12 million at the end of June 2026, doubling from roughly 6 million a year earlier (Space Exploration Technologies Corp., 2026a).

That combination of subscriber expansion and positive operating income is financially significant.

Unlike an experimental technology initiative, Starlink already operates a large-scale commercial service with recurring revenue.

It provides broadband connectivity across consumer, enterprise, maritime, aviation and government markets.

The company also benefits from substantial vertical integration in satellite production, launch operations and network deployment.

Nevertheless, subscriber growth should not be assessed in isolation.

Reported average revenue per user declined to approximately US$66 per month from US$85 in the comparable prior-year period.

That may reflect geographical expansion, customer mix and lower-priced international services rather than fundamental weakness.

However, the decline highlights the importance of revenue quality and operating efficiency.

Future profitability will depend on maintaining attractive network economics while expanding coverage and replacing satellite assets.

Starlink's potential extension into mobile connectivity provides another meaningful growth opportunity.

In October 2026, SpaceX announced plans to acquire approximately US$8 billion of low-band wireless spectrum interests from Grain Management, subject to regulatory approval, as part of its broader Starlink Mobile ambitions (Reuters, 2026b).

The transaction could strengthen SpaceX's ability to combine satellite-based connectivity with terrestrial mobile infrastructure.

However, commercial execution requires spectrum approvals, network deployment, compatible devices and effective competition with established telecommunications operators.

It would be premature to assume that Starlink can immediately replicate the economics of conventional mobile carriers.

Still, the strategic opportunity is substantial.

Starlink provides an existing commercial platform through which SpaceX can pursue global connectivity expansion, enterprise contracts and additional communications services.

In my assessment, this is the strongest established component of the company's investment proposition.

Investors may be captivated by orbital AI, but Starlink is currently doing much of the financial heavy lifting.

The US$100 Billion Balance Sheet: Strength Does Not Guarantee Capital Efficiency

SpaceX's substantial liquidity provides an important advantage over more financially constrained competitors.

At 30 June 2026, the company held approximately US$93.52 billion in cash and cash equivalents, together with US$6.49 billion in short-term marketable securities.

Its combined liquid resources therefore totalled approximately US$100.01 billion.

Reported current and long-term debt and finance leases amounted to approximately US$39.36 billion, leaving a simplified net liquidity position of about US$60.65 billion before other liabilities and commitments (Space Exploration Technologies Corp., 2026b).

This provides substantial flexibility.

The company can fund infrastructure development, navigate operational setbacks and pursue strategic initiatives without relying exclusively on immediate operating profitability.

However, the liquidity balance must be interpreted correctly.

Much of SpaceX's financial strength reflects external capital raising, including approximately US$85.68 billion in net IPO proceeds and US$25 billion in senior notes issued during June 2026.

Those funds improve financing capacity, but they should not be confused with cash generated from recurring commercial operations.

An IPO increases liquidity while diluting ownership. Debt financing supplies capital while creating interest and repayment obligations.

A company can therefore possess an exceptionally strong balance sheet while still facing substantial capital allocation risks.

The relevant issue is not merely whether SpaceX has enough cash to invest.

It is whether management can deploy that cash at returns exceeding the company's cost of capital.

This consideration becomes particularly important during periods of elevated interest rates and market volatility.

The October 2026 pressure on AI infrastructure equities illustrates how rapidly financing concerns can affect valuations across the sector.

Financially leveraged infrastructure providers may encounter difficulties when borrowing costs rise, collateral values weaken or investors become less willing to fund prolonged negative free cash flow.

SpaceX is better positioned than many smaller competitors to absorb such volatility.

Nevertheless, its shareholders remain exposed to capital expenditure risk, valuation compression and potentially significant dilution through future acquisitions.

Financial flexibility is valuable.

Financial discipline is indispensable.

Valuation: Has the Market Already Capitalised Too Much of the Future?

The largest challenge for prospective SPCX shareholders is reconciling the company's strategic potential with its valuation.

At an illustrative US$2 trillion equity valuation, SpaceX trades at approximately 64 times second-quarter revenue annualised over four quarters.

This is not a conventional trailing-twelve-month multiple, nor does it represent a precise live valuation. It is a simplified measure designed to illustrate the magnitude of future growth expectations.

The implication is nevertheless clear.

Investors are assigning substantial value to revenue and profitability that have not yet been demonstrated in historical financial results.

A premium valuation is not inherently irrational.

Companies possessing powerful competitive advantages, expanding addressable markets and attractive long-term returns on capital can deserve higher multiples than mature businesses.

SpaceX arguably has several such advantages.

Its launch capabilities support satellite deployment. Starlink provides recurring connectivity revenue. Its engineering expertise allows extensive vertical integration. Its AI infrastructure contracts demonstrate commercial demand, while software acquisitions may create additional distribution opportunities.

However, these advantages must be evaluated against capital intensity, technological uncertainty and the economic costs of expansion.

Consider an illustrative scenario.

At a US$2 trillion equity valuation, generating a sustainable 4% annual free cash flow yield would require approximately US$80 billion in annual free cash flow attributable to shareholders.

At a hypothetical 25% free cash flow margin, that would imply roughly US$320 billion in annual revenue.

Neither figure is a prediction. The calculation simply demonstrates how much future profitability may be necessary to support a mature valuation of that magnitude.

SpaceX could potentially generate substantial revenue across satellite connectivity, AI services and other technological opportunities.

But investors must distinguish plausible long-term potential from the probability, timing and cost of achieving it.

The appropriate valuation approach is therefore a sum-of-the-parts analysis.

Starlink should be evaluated using subscriber economics, operating margins, network investment and recurring cash flow.

AI infrastructure requires analysis of utilisation, customer contracts, hardware depreciation and returns on invested capital.

Launch and aerospace operations should be assessed through manufacturing economics, programme development costs and external contract opportunities.

Cursor and other software capabilities should be valued through customer retention, recurring revenue, competitive differentiation and sustainable software profitability.

Orbital computing deserves consideration as a strategic option, but its valuation should reflect substantial uncertainty.

Corporate governance also matters.

SpaceX's dual-class share structure provides concentrated voting control, potentially supporting long-term strategic decision-making while limiting the influence of ordinary public shareholders (Space Exploration Technologies Corp., 2026f).

Major acquisitions involving related business ecosystems further reinforce the importance of transparent disclosure and disciplined capital allocation.

The result is an investment proposition containing both exceptional opportunity and substantial complexity.

The risk is not necessarily that SpaceX fails technologically. It is that the market has priced in too much commercial success, too quickly.

My Investment Outlook: Strategically Bullish, Valuation-Disciplined

I believe SpaceX has the potential to become one of the most consequential infrastructure companies of the coming decades.

Its combination of launch capabilities, satellite communications, AI infrastructure and software integration creates opportunities that few individual companies can replicate.

But I do not believe investors should approach its shares as an automatic purchase simply because the stock has corrected from previous highs.

In my investment framework, several developments would materially strengthen the valuation case.

The AI segment must demonstrate sustained improvement in GAAP operating profitability rather than relying predominantly on adjusted EBITDA.

Capital expenditure should increasingly translate into recognised external revenue, stronger operating cash flow and attractive utilisation-adjusted returns.

Starlink should maintain healthy subscriber growth, service reliability and operating margins while developing new enterprise and mobile markets.

Cursor must demonstrate that the US$60 billion acquisition creates genuine incremental economic value rather than merely increasing the size and complexity of the group.

SpaceX's ambitious orbital computing initiatives should progress through verifiable engineering milestones before investors assign substantial near-term earnings expectations.

Most importantly, the company's valuation should offer an adequate margin of safety against delayed commercialisation, competitive pressure and weaker-than-expected cash generation.

That margin can emerge through a lower market price, stronger financial fundamentals or a combination of both.

I would therefore retain SpaceX on a priority research watch list while remaining selective about the price at which I would initiate exposure.

Beyond SpaceX, I would also monitor Rocket Lab and MDA Space for differentiated opportunities within the developing space economy.

Semiconductor suppliers, established hyperscalers, electrical infrastructure providers and specialised AI cloud companies may offer alternative ways to participate in related technological trends.

However, diversification across companies associated with artificial intelligence does not necessarily eliminate correlated market risk.

Portfolio construction should remain grounded in valuation, financial resilience, business quality and exposure to common macroeconomic factors.

The objective is not simply to own as many companies associated with future technologies as possible.

It is to acquire durable economic assets at prices offering attractive risk-adjusted prospective returns.

Conclusion: The Future May Belong to SpaceX, but the Price Still Matters

SpaceX is no longer simply a reusable rocket company.

It is becoming an integrated infrastructure enterprise operating at the intersection of space transportation, satellite connectivity, artificial intelligence, enterprise software and advanced computing.

That strategic transformation is real.

Its growth is measurable.

Its commercial opportunities are substantial.

But so are its capital requirements, execution risks and valuation expectations.

The second-quarter financial results tell the story clearly. Revenue grew 92%, Connectivity remained strongly profitable, and AI infrastructure expansion accelerated dramatically.

At the same time, consolidated GAAP earnings remained negative, capital expenditure reached US$18.37 billion, and first-half free cash flow was deeply negative under a conventional simplified calculation.

Its substantial liquidity offers financial resilience, but the ultimate investment outcome will depend on whether capital is converted into sustainable economic returns.

SpaceX's future orbital ambitions may eventually produce entirely new categories of infrastructure and commercial opportunity.

Yet an investment thesis cannot be built solely on the assumption that every technological ambition will become commercially successful.

The appropriate distinction is between what the company has already achieved, what it has credibly contracted to deliver and what it still hopes to accomplish.

My conclusion is that SpaceX represents a compelling long-term business proposition, but not an unquestionably attractive investment at any price.

For now, I favour disciplined observation over valuation-driven enthusiasm.

I would prefer stronger evidence of AI infrastructure profitability, moderating capital intensity or a more attractive valuation before committing significant additional capital.

The market may ultimately reward SpaceX for reshaping multiple industries.

But investors should not confuse the commercial importance of an enterprise with the financial returns available to its shareholders.

The greatest investment opportunities do not necessarily come from identifying the companies capable of building tomorrow. They come from identifying when tomorrow's economic value can be purchased at a rational price today.

That is the standard against which I would judge SpaceX.

And it is why, despite its extraordinary potential, SPCX remains a company I would study closely before deciding to buy.

---

Investment Research Disclaimer

This article represents the author's independent commentary for general educational, informational and investment research purposes only. It is not personalised financial advice, an offer to buy or sell securities, or a guarantee of investment performance. The analysis draws upon publicly available information as of 16 October 2026, with primary financial statements covering the second quarter of 2026. Historical financial results, management targets, illustrative valuations and forward-looking statements should not be interpreted as assurances of future outcomes. Securities investments involve market volatility, liquidity risk and potential loss of capital. Readers should independently verify material information and consult appropriately licensed financial, legal or tax advisers where necessary. References to companies and individuals are made for objective analysis and do not imply endorsement, affiliation or commercial association.

---

References

American Psychological Association (APA), 7th Edition

International Energy Agency. (2026). Key questions on energy and AI. https://www.iea.org/reports/key-questions-on-energy-and-ai

Kessler, D. J., & Cour-Palais, B. G. (1978). Collision frequency of artificial satellites: The creation of a debris belt. Journal of Geophysical Research, 83(A6), 2637–2646. https://doi.org/10.1029/JA083iA06p02637

NVIDIA Corporation. (2026, August 24). SpaceXAI adopts NVIDIA Vera CPU to accelerate agentic AI at massive scale[Press release]. https://investor.nvidia.com/news/press-release-details/2026/SpaceXAI-Adopts-NVIDIA-Vera-CPU-to-Accelerate-Agentic-AI-at-Massive-Scale/default.aspx

Reuters. (2026a, February 2). SpaceX acquires xAI in record-setting deal as Musk looks to unify AI and space ambitions.https://www.reuters.com/business/musks-spacex-merge-with-xai-combined-valuation-125-trillion-bloomberg-news-2026-02-02/

Reuters. (2026b, October 9). What is low-band spectrum that SpaceX acquired for Starlink Mobile?https://www.reuters.com/business/what-is-low-band-spectrum-that-spacex-acquired-starlink-mobile-2026-10-09/

Space Exploration Technologies Corp. (2026a, August 4). SpaceX reports second quarter 2026 results [Earnings release, Exhibit 99.1]. U.S. Securities and Exchange Commission. https://www.sec.gov/Archives/edgar/data/1181412/000162828026052515/earningsreleaseq22608042.htm

Space Exploration Technologies Corp. (2026b, August 4). Quarterly report on Form 10-Q for the quarter ended June 30, 2026. U.S. Securities and Exchange Commission. https://www.sec.gov/Archives/edgar/data/1181412/000162828026052535/spcx-20260630.htm

Space Exploration Technologies Corp. (2026c, August 14). Current report on Form 8-K: Completion of the Anysphere, Inc. acquisition. U.S. Securities and Exchange Commission. https://www.sec.gov/Archives/edgar/data/1181412/000162828026056945/spcx-20260814.htm

Space Exploration Technologies Corp. (2026d, June 5). Free writing prospectus: Cloud service agreement with Google LLC. U.S. Securities and Exchange Commission. https://www.sec.gov/Archives/edgar/data/1181412/000162828026041150/spacexagreementfwp.htm

Space Exploration Technologies Corp. (2026e, August). SpaceX second quarter 2026 earnings call transcript. The Motley Fool. https://www.fool.com/earnings/call-transcripts/2026/08/11/spacex-spcx-q2-2026-earnings-call-transcript/

Space Exploration Technologies Corp. (2026f). Registration statement on Form S-1. U.S. Securities and Exchange Commission. https://www.sec.gov/Archives/edgar/data/1181412/000162828026039276/spaceexplorationtechnologi.htm

Stock Analysis. (2026). Space Exploration Technologies Corp. (SPCX) stock price history.https://stockanalysis.com/stocks/spcx/history/

Turyshev, S. G. (2026). Orbital data centers: Spacecraft constraints and economic viability [Preprint]. arXiv. https://arxiv.org/abs/2604.27197

From Starlink to Orbital AI: Is SpaceX Building the Next Global Infrastructure Giant or Pricing In Too Much Future Growth?

🏙️ What Does SpaceX’s US$2 Trillion AI Bet Mean for Singapore Property Owners and Investors?

Global technology revolutions do not just reshape stock markets. They can influence employment, capital flows, financing conditions and real estate demand.

SpaceX’s expansion into artificial intelligence and infrastructure highlights economic forces that may also affect Singapore’s property landscape, including business investment, expatriate housing demand, borrowing costs and investor confidence.

For homebuyers, understanding these trends supports informed purchasing and financing decisions. For sellers, market timing and competitive pricing remain essential. For landlords and tenants, evolving employment and relocation patterns may influence rental demand. For property investors, sustainable rental yields, capital preservation and long-term fundamentals matter more than market hype.

Global opportunities require local expertise.

As your Singapore real estate salesperson, I provide data-driven market insights, property comparisons, strategic pricing and personalised guidance to help you buy, sell, rent or invest confidently.

Your property deserves more than a transaction. It deserves a well-informed strategy.

📞 Zion Zhao 赵峻慷 | +65 8884 4623

🔗 https://linktr.ee/zionzhao

Like, save, share and follow my social media channels for Singapore property insights, market analysis and investment perspectives.

Make informed decisions today. Build lasting property value for tomorrow.

SpaceX的2万亿美元估值大考：星链单季赚16.6亿美元，AI豪掷158亿美元，SPCX究竟值不值得买？

SpaceX上市后的2万亿美元豪赌：从火箭、人工智能到未来算力，SPCX究竟是逢低布局的投资良机，还是高估值陷阱？

作者：赵峻慷（Zion Zhao）｜投资研究与战略市场分析

SpaceX或许正在建设属于未来的科技基础设施，但投资者如今支付的高昂估值，已经提前反映了大量尚未兑现的增长预期。

自2026年6月完成首次公开募股（Initial Public Offering，IPO）以来，美国太空探索技术公司（Space Exploration Technologies Corp.，简称SpaceX，纳斯达克股票代码：SPCX）已不再只是一家专注于可重复使用火箭与星链（Starlink）的航天企业，而是逐步转型为横跨航天科技、人工智能（Artificial Intelligence，AI）及云计算的垂直整合科技集团。

通过整合xAI，以及斥资600亿美元收购人工智能编程平台Cursor，SpaceX正试图将AI算力基础设施、自有人工智能模型与开发者软件整合至同一个商业生态系统（Space Exploration Technologies Corp., 2026a）。

战略蓝图令人瞩目，但其背后的财务代价，更值得投资者审慎评估。

2026年第二季度，SpaceX营收同比飙升92%，达到78.1亿美元。

其中，以星链为核心的通信业务（Connectivity）贡献42.9亿美元营收及16.6亿美元营业利润。人工智能业务则实现25.6亿美元营收，但仍录得12.6亿美元营业亏损（Space Exploration Technologies Corp., 2026b）。

星链仍是集团成熟的盈利引擎，而人工智能则代表增长潜力更高、同时资本需求更庞大的战略布局。

SpaceX与谷歌（Google）签署的算力服务协议，每月合同金额高达9.2亿美元，证明其人工智能基础设施已获得重要的外部商业需求。

然而，合同仍涉及算力交付条件及终止条款，因此不能将合同金额直接视为必然实现的营收。

市场关注的410亿美元AI业务年化营收运行率（Annualised Revenue Run Rate），目前仍属于前瞻性预测，而非已经确认的财务业绩（Space Exploration Technologies Corp., 2026c）。

与此同时，管理层提出到2027年将算力容量提升至接近10吉瓦（Gigawatts，GW）的目标，进一步凸显其基础设施扩张的庞大规模。

真正的挑战，在于增长需要付出多大的资本代价。

SpaceX第二季度资本支出（Capital Expenditure，CapEx）高达183.7亿美元，其中绝大部分投入人工智能基础设施。

2026年上半年，经营活动现金流约为34.7亿美元，但资本支出达到284.8亿美元。按照简化计算方式，自由现金流（Free Cash Flow，FCF）缺口约为250亿美元（Space Exploration Technologies Corp., 2026b）。

尽管公司持有约1000亿美元现金及短期有价证券，雄厚的财务实力仍不等于投入资本必然能够创造理想回报。

SpaceX与英伟达（NVIDIA）合作研发太空轨道人工智能计算系统，更为其长期发展带来了革命性的想象空间。

但电力供应、散热管理、太空辐射、通信限制及设备更新成本等挑战，意味着轨道计算的商业可行性仍有待验证（NVIDIA Corporation, 2026）。

以约2万亿美元的假设性估值衡量，投资者实际上正在为尚未充分兑现的未来增长和盈利能力支付巨额溢价。

我的投资判断是：长期看好SpaceX的战略方向，但必须坚持估值纪律。

在考虑建立较大投资仓位之前，我更希望看到美国公认会计原则（GAAP）下的盈利能力持续改善、人工智能基础设施使用率提高、星链维持稳健利润率，以及资本支出强度逐步趋于合理。

SpaceX值得列入长期优先研究观察名单，火箭实验室（Rocket Lab）及MDA Space等太空经济企业，同样值得持续关注。

但技术领先，并不代表任何估值都合理。

真正伟大的投资机会，不只是提前发现谁有能力建设未来，而是在合理的价格下，买入未来能够创造的真实经济价值。

现阶段，SpaceX无疑是一家极具潜力的企业，但SPCX是否值得买入，最终仍取决于战略执行、现金流创造能力，以及投资者付出的价格。

未来或许属于SpaceX，但长期投资回报，始终取决于估值与基本面能否相互匹配。

---

参考文献（References）

采用美国心理学会第七版（APA 7th Edition）格式。为确保学术严谨性及方便核实，保留英文文献原题、机构名称与官方链接。

NVIDIA Corporation. (2026, August 24). SpaceXAI adopts NVIDIA Vera CPU to accelerate agentic AI at massive scale.https://nvidianews.nvidia.com/news/spacexai-adopts-nvidia-vera-cpu-to-accelerate-agentic-ai-at-massive-scale

Space Exploration Technologies Corp. (2026a, August 14). Current report on Form 8-K: Completion of the Anysphere, Inc. acquisition. U.S. Securities and Exchange Commission. https://www.sec.gov/Archives/edgar/data/1181412/000162828026056945/spcx-20260814.htm

Space Exploration Technologies Corp. (2026b, August 4). Quarterly report on Form 10-Q for the quarter ended June 30, 2026. U.S. Securities and Exchange Commission. https://www.sec.gov/Archives/edgar/data/1181412/000162828026052535/spcx-20260630.htm

Space Exploration Technologies Corp. (2026c, June 5). Free writing prospectus: Cloud service agreement with Google LLC. U.S. Securities and Exchange Commission. https://www.sec.gov/Archives/edgar/data/1181412/000162828026041150/spacexagreementfwp.htm