Management priced its 555.6 million primary share offering (plus a 15% over-allotment option) at a rigid $135 per share. This implies a market valuation hovering near $1.78 trillion on a tight, controlled float. The stock is now at about $200 per share a 2.72 trillion valuation.

To evaluate if the $1.78T entry point is reasonable, we must review what a simplified version of our internal model (SpaceX - Valuation - RC) requires over the next decade versus the "Future Targets" management outlined in their pitchbook.

This model extracted the best-case scenario, Starlink reaches 41.5 million subscribers, Starship scales to 350 commercial launches, and the AI segment secures 25,000 enterprise customers. Yet, even when we assume these milestones are met, the baseline numbers show that management's long-term corporate goals are detached from current industrial constraints.

This SpaceX valuation model uses a very simple Sum-of-the-Parts (SOTP) Discounted Cash Flow (DCF) approach to estimate the company's value, projecting financials from 2025 through 2035. The model is simplified to focus on the main drivers and get the point across without confusion.

The model concludes that SpaceX has a total Equity Value of approximately $456.2 billion. The valuation is heavily dominated by the Starlink division.

The gap between the calculated $456 billion and the $2 trillion trading market cap suggests that the market may be pricing in significantly more aggressive growth assumptions, higher terminal multiples, or additional business lines not captured in this specific financial model. But in all honesty we believe that beyond a lot of the assumptions we’ve made the markets aren’t just irrational but outright crazy!

Total Enterprise Value (EV): $468.7 billion

Net Debt: $12.5 billion (deducted from EV to arrive at Equity Value). (This debt estimate is much lower than current obligations

The model's valuation relies on aggressive growth in satellite internet adoption and launch frequency:

Our model relies on Starlink scaling to 41.5 million users by 2035E while strictly maintaining a consumer ARPU of $110/month and an enterprise ARPU of $500/month.

The roadshow presentation reveals the real operational cost of buying this user base. Starlink has surpassed 10.3 million active subscribers across 164 countries, capturing roughly 75% of all active maneuverable satellites in orbit. This drove Connectivity revenue to $11.4 billion in 2025 (50% YoY growth), producing a Segment Adjusted EBITDA of $7.2 billion. To hit these numbers, management is heavily discounting service. The roadshow notes that their international expansion is causing a structural shift toward a "declining average ARPU". They claim a global total addressable market of $1.6 trillion, split between Broadband ($870B) and Mobile ($740B), assuming they can connect 3.3 billion people.

However, expanding into lower-income regions directly threatens our model's fixed $110 ARPU floor. Our model assumes hardware margins improve from -20% to +15% over the forecast window. However, to push next-generation V3 Broadband satellites (boasting 1,024 Gbps per satellite, deployed via Starship in H2 2026), SpaceX must continue to absorb significant near-term subscriber acquisition costs.

The presentation emphasizes that SpaceX controls over 80% of global mass to orbit and has achieved a 95%+ booster reusability rate across ~650 total historical launches. Yet, the roadshow financials expose the underlying economic reality: the launch division is a capital-heavy business designed to subsidize internal networks.

Financial Breakdown is as follows. External launch revenue in 2025 was a modest $4.1 billion, growing just 8% YoY from 2024. Because internal Starlink launches generate no inter-segment revenue, the Space segment's Adjusted EBITDA fell from $1.2 billion in 2023 to just $700 million in 2025. This drop was driven by a massive surge in R&D spending, which doubled to $3.0 billion to support Starship's development.

The roadshow outlines a goal to cut launch costs by 99%+, dropping from an early baseline of $18,500/kg down to Falcon 9's $2,700/kg and Falcon Heavy's $1,400/kg, eventually hitting pure fuel costs with Starship. Our model factors in this optimization, scaling Starship to 350 annual flights and dropping the price per launch to $10 million by 2035E. For this to generate meaningful profit, external demand must be highly elastic. If the market fails to expand beyond commercial telcos and defense primes, SpaceX will have spent billions on infrastructure to deflate its own end-market pricing power.

The most aggressive claim in the roadshow presentation is the $28.5 trillion long-term TAM, which allocates an astronomical $26.5 trillion to AI ($22.7T in Enterprise Applications, $3.8T in Infrastructure and Subscriptions). To justify this number, the presentation relies on highly volatile, short-term arrangements:

Management highlights a Cloud Services Agreement with Anthropic, generating a $1.25 billion monthly fee ($15B annualized) through May 2029. However, the presentation concedes this agreement is "subject to certain conditions" and represents raw compute leasing across their 1.0 GW Colossus training clusters rather than organic adoption of their own software. Qualitatively, this means their AI revenue is exposed to significant counterparty concentration and cancellation risks.

While our model assumes an initial AI margin of 40% expanding to 80%, the S-1 GAAP reconciliation reveals a deeply unprofitable segment. The AI division posted an operational loss of $6.4 billion in 2025, with a Segment Adjusted EBITDA loss of $1.2 billion. In Q1 2026 alone, the division lost another $2.5 billion. Furthermore, the roadshow admits that the segment’s historical revenues in 2023 and 2024 were driven "substantially" by the legacy X advertising business, which is actively declining (down 12% in 2024).

The presentation outlines a partnership with Cursor, carrying an option to acquire them at an implied $60 billion valuation. It also details a joint venture with Tesla and Intel ("Terafab") to design and fabricate proprietary chips. These capital commitments explain why consolidated capital expenditures surged from $4.4 billion in 2023 to $20.7 billion in 2025.

The ultimate irony of the SpaceX prospectus is that the segment carrying the most absurd valuation weight, the $26.5 trillion AI TAM, is precisely the vertical where SpaceX holds its weakest, most exposed competitive position. Management is pitching a flawless transformation from rockets to software, but a strategic reality check reveals that this is the specific business line most likely to trigger a severe multiple re-rating.

On Earth, the ultimate return on investment (ROI) for generative AI remains highly debated, with hyperscalers spending hundreds of billions on capex while struggling to show durable enterprise software profitability. SpaceX expects to solve this by launching AI into orbit to escape land, energy, and cooling bottlenecks. However, moving depreciating, radiation-exposed hardware into a vacuum does not solve the fundamental utility problem of AI; it simply adds an extreme capital-intensity premium to unproven economic returns.

Even if SpaceX successfully pulls off a "Starlink-like" move (fully vertically integrating space-based data centers with Starship logistics) they enter the market as an absolute lightweight compared to terrestrial hyperscalers (Microsoft, Alphabet, Meta). Alphabet alone recently issued an $85B equity chunk to fund its data infrastructure. SpaceX is attempt-funding a hardware clone war with a $75B public raise that is heavily spoken for by legacy bridge loans and an existing $235B cash gap.

The assumption that SpaceX can operate an uncontested global AI toll road completely ignores geopolitical reality. China's state-backed telecom apparatus (e.g., Huawei, ZTE) already leads global physical telecom infrastructure rollout across emerging markets and will aggressively undercut space-based subscription pricing. Furthermore, Beijing maintains an absolute choke point over the physical AI supply chain, dominating rare earth extraction, refining, component sub-assemblies, and battery supply chains critical for the "Terafab" and Colossus facilities. SpaceX's space cloud can be structurally strangled or priced out before it reaches scale.

While Space and Starlink enjoy profound cost advantages built on twenty years of rocket iteration, SpaceX has no structural moat in AI. Consumer AI switches seamlessly across APIs; enterprise workflows are intentionally architected to multi-home models to prevent platform lock-in. SpaceX is burning $2.5 billion a quarter on a commoditizing asset-heavy layer (GPU compute rentals to rivals like Anthropic) while running a horrifying -183% operating margin

This IPO lowers the standard for what it means to be a public shareholder. You are not buying cash flows, you are funding an ‘technocracy.

Overall we believe SpaceX acts as a bailout mechanism. It bought $131M in Cybertrucks at retail price (accounting for 18% of US Q4 registrations) to prop up Tesla metrics. The $250B xAI merger allowed underwater Twitter lenders (like Morgan Stanley) to be paid in kind through SpaceX equity. Board member Antonio Gracias (Valor Equity) is tied to $20B in "failed sale-leasebacks" resting on SpaceX's balance sheet.

From accounting perspective any expected financial obligations must be accounted for but there’s a loop hole where if a transaction is deemed ‘improbable’ it can be left out of accounting projections. This indirectly incentivises management to state crazy expectations without having to underwrite them now leaving room for disappointment. Musk’s package grants 1 billion shares tied to a $7.5T market cap and a Mars colony. By legally classifying the Mars milestone as "improbable," SpaceX avoids immediate expense recognition, yet Musk can vote the shares and borrow against them today.

The IPO mechanics reveal Wall Street's investment bankers’ complete loss of leverage to mega-IPOs and the aggressive funneling of risk to retail and passive investors.

Investment Banks didn't really value this, they priced it. Musk dictated the $135 fixed price. Top banks accepted fees below 0.75% ($500M total) and engaged hyping up the stock like promotors.

20-30% of the deal ($15B) is retail-allocated, oversubscribed 7:1. Fidelity dropped minimums to $2k. UK retail is entering via Mark Financial on a bare-minimum 6-month solvency guarantee. Brokers like SoFi are threatening $50 fees and bans for flipping within 30-120 days.

They are just looking to dump the stock on passive ETF funds. They are locking retail in until passive flows arrive. Nasdaq altered rules to fast-track SpaceX into the NDX-100 in 15 days, and Russell in 5 days, forcing hundreds of billions in price-insensitive buying. (S&P refused to alter rules).

The money is still not enough. The $75B raised covers less than a third of the required capital. Cape Fear Advisors estimates a $235B cash gap through 2030. The first $20B immediately retires the bridge loan used to absorb xAI's debt. This is just series A of public dilution.

The roadshow presentation confirms that the numbers behind this offering do not match traditional asset-valuation principles.

The business is burning billions on capital-intensive AI infrastructure while trying to pass itself off as an enterprise SaaS provider to secure a trillion-dollar multiple.

Our internal model maps out a highly optimistic, best-case growth path, and it still cannot comfortably bridge the gap to the market's $1.78T premium.

This website and all materials contained herein are issued by Regal Capital Botswana (Pty) Ltd ("Regal Capital") for informational purposes only and do not constitute investment advice, a solicitation, or an offer to buy or sell any financial instruments. The content is general in nature and does not consider specific investment objectives, financial situations, or individual needs. While Regal Capital strives for accuracy, no representation or warranty, express or implied, is made regarding the completeness, reliability, or accuracy of the information. Any opinions or estimates reflect the author’s judgment as of the publication date and are subject to change without notice. Regal Capital, its affiliates, directors, employees, agents, or contractors shall not be liable for any direct, indirect, or consequential loss or damage resulting from the use of or reliance on this website. Users should seek independent financial advice before making any investment decisions.

All intellectual property rights in this website are proprietary to Regal Capital or authorized third-party sources. No content may be reproduced, distributed, or transmitted without prior written permission. Investing involves risks, including potential capital loss. Past performance is not indicative of future results, and investments in foreign currencies are subject to exchange rate fluctuations. The strategies or instruments mentioned may not be suitable for all investors.

By accessing research materials, users acknowledge that Regal Capital and its affiliates may hold positions in securities discussed. Research is provided “as is” without warranties of accuracy or completeness and should not be considered a recommendation or solicitation. Users are encouraged to conduct their own due diligence and consult financial, legal, or tax advisors. Regal Capital is not liable for any trading losses, lost profits, or other damages resulting from reliance on its research. Use of this website is governed by the laws of Botswana, and users agree to the jurisdiction of its courts for any disputes

For more information visit our website at http://www.regalcapital.org or contact email [email protected]