--- title: "SpaceX: From 'Rocket Man' to 'Landlord' — Who Will Foot the Bill for Musk's Next Trillion-Dollar Moonshot?" type: "Topics" locale: "en" url: "https://longbridge.com/en/dolphin/post/44451710.md" description: "Compute rental costs keep rising; the real test is orders and funding" datetime: "2026-10-21T05:48:59.000Z" locales: - [en](https://longbridge.com/en/dolphin/post/44451710.md) - [zh-CN](https://longbridge.com/zh-CN/dolphin/post/44451710.md) - [zh-HK](https://longbridge.com/zh-HK/dolphin/post/44451710.md) author: "[Dolphin Research](https://longbridge.com/en/dolphin.md)" generator: "portal-rs" --- # SpaceX: From 'Rocket Man' to 'Landlord' — Who Will Foot the Bill for Musk's Next Trillion-Dollar Moonshot? In the early hours of Oct 21 Beijing time (post-market Oct 20 ET), $SpaceX(SPCX.US) released its second post-IPO earnings: across-the-board beats with even stronger guidance. Highlights below: **① Revenue smashed estimates; full-quarter AI compute leasing nearly doubled revenue QoQ:** Q3 revenue was $145.8 bn, 13% above the Street (~$129 bn), up 161% YoY and accelerating vs. Q2 (+92%). All three segments—Aerospace, Starlink, and AI—beat. AI revenue reached $75.6 bn, 52% of total and above 50% for the first time, driven by full-quarter compute hosting and higher Grok revenue. **② GPM surged, but included one-offs and will normalize in Q4:** GP was $91 bn, above the Street (~$86.4 bn). GPM expanded 710 bps QoQ to 62.4% (vs. 55.3% in Q2). Drivers were threefold: 1) mix shift to high-margin compute hosting, with AI GPM up from 56.8% to 64.6%; 2) concentration of high-value aerospace missions, lifting GPM from 65.8% to 75.4%; 3) roughly 300 bps one-time tariff refunds at Starlink. Management guided Q4 GPM back to Q2 levels (~55%). This reflects Q3 one-offs and a favorable customer mix, so the Q3 peak should not be extrapolated linearly. **③ Strong operating leverage; GAAP OP turned positive and far exceeded expectations:** OP was $2.8 bn, more than 2x the Street ($1.25 bn) vs. -$0.14 bn in Q2. OPM rose ~21 ppt QoQ to 19%. **④ Segment performance in detail:** **a. Aerospace: high-value missions drove a revenue beat; Starship R&D ramped, but the segment neared breakeven** **Revenue:** Aerospace revenue was $1.77 bn, beating the Street (~$1.44 bn). It grew 6% YoY, reflecting strong mix and pricing. **Launch services:** Revenue reached $1.13 bn, ahead of expectations (~$0.8 bn), mainly on higher mix of high-value customers and better-than-expected pricing. **Launch cadence:** There were 12 customer launches (10 in Q2), slightly below the 13 expected. This included a Falcon Heavy launch of the Roman Space Telescope for NASA and six national security missions. Total launches were 35 (38 in Q2), including 23 internal launches (21 Falcon 9 Starlink missions + 2 Starship test flights). After the last Falcon 9 Starlink launch in Florida on Aug 26, Starship began taking over Florida Starlink deployment, and internal launch frequency continued to decline. **Launch pricing:** Per-launch customer revenue averaged ~$94 mn, up sharply from ~$65 mn in Q2. This was due to clustered high-priced missions such as Falcon Heavy and national security launches. New orders were also robust: three Dragon crewed missions (out to 2028), 18 national security launches, three Starfall missions, the Artemis 3 Starlink comms contract, and a Starship lunar payload contract. **With an almost monopolistic lift capacity (about 90% of global to-orbit mass), SpaceX retains strong pricing power.** **Launch and R&D services:** Recognized $640 mn, doubling QoQ (vs. $310 mn in Q2). Revenue is milestone/POC-based, and Q3 uplift came from Starship breakthroughs. Flight 13 deployed real V3 satellites for the first time, while Flight 14 achieved first orbit, deployed 26 V3 satellites, and executed a controlled deorbit burn. These corresponded to milestones under contracts such as NASA HLS. **Profitability:** Aerospace GPM rose ~10 ppt QoQ to 75.4% on high-priced missions and higher Falcon 9 reuse. This mix tailwind was meaningful and visible. Operating loss was only $7 mn, far better than the Street’s -$320 mn. Costs of two Starship test flights were largely booked as R&D (Aerospace R&D $1.23 bn, +14% QoQ), the main reason for the small segment loss; Adj. EBITDA turned positive at $360 mn. **b. Starlink: ARPU rebounded unexpectedly; B2B/Gov. led growth, while consumer subs missed** **Revenue:** Starlink revenue was $5.25 bn, +69% YoY and above the Street (~$4.86 bn). This was driven by both enterprise/government and consumer strength. **Enterprise & Gov.:** Revenue reached $2.13 bn (Street ~ $1.96 bn), +91% YoY, led by Starshield and aviation. Two USAF Starshield contracts signed in Q2 (Space Data Network Backbone ~$2.29 bn, Space-Based AMTI ~$4.16 bn) moved into execution. Aviation has signed ~12k aircraft (~1/4 of the available fleet, covering three of the four major US airlines). Installed aircraft exceeded 3,800 by end-Sep, with aviation terminal capacity lifted 4x this year and set to double again by year-end. Management expects enterprise & Gov. revenue to surpass consumer over time. **Consumer:** Revenue was $3.12 bn (Street ~ $2.84 bn), +57% YoY. This was above expectations even as subs trailed forecasts. **Subs:** Consumer users rose to 13.6 mn, a net add of 1.6 mn QoQ (vs. ~1.7 mn in Q2). This was below the ~14.25 mn expected. The main reason is the new V5 home terminal only just ramped, with prior capacity constraints. Inventory was pre-staged to retail ahead of holidays and not yet activated, with Q4 activations expected to rise materially. Also, from this quarter, 'Starlink users' are renamed 'Starlink consumer users', excluding hundreds of thousands of enterprise and Gov. users. This change clarifies reporting. **ARPU:** Consumer monthly ARPU was $69, +5% QoQ (vs. $66 in Q2), above the Street’s ~$65 and the first rebound in several quarters. New adds skewed to developed markets. Note ARPU includes subscription revenue only. On average users during the quarter, ~$0.4–0.5 bn of consumer revenue came from hardware and other non-subscription items. **Mobile:** For now, the business primarily wholesales direct-to-cell satellite capacity to carriers. MAUs exceeded 22 mn, +37% QoQ, reflecting rapid adoption. Next step is to build its own MNO: on Oct 8, SpaceX announced the acquisition of Grain’s 800 MHz low-band spectrum (expected to close in H1 2027), on top of EchoStar’s 2 GHz mid-band. The FCC has approved ~15k mobile V2 satellites (launches from 2027). Management plans to launch services with owned spectrum in early 2028, aiming to become a major US mobile operator. This is a structural shift from a pure satellite complement. By end-Q3 2026, in-orbit satellites were ~11.1k (new methodology includes ~400 approved deorbit satellites, old methodology ~10.7k; 2Q26 was ~10.2k). Flight 14 deployed ~26 Tbps downlink using roughly half its payload, equivalent to 10 Falcon 9 Starlink launches. **Margins:** Starlink GPM was 54.9%, up ~300 bps QoQ mainly due to one-time tariff refunds. Ex-refunds, it was roughly flat, indicating stable unit economics. OP was $2.37 bn with OPM of 45%, up 650 bps QoQ and well above the Street’s ~38.4%. Scale effects continue to expand for Starlink-like SaaS, and the ARPU rebound eases earlier concerns. **c. AI: full-quarter hosting plus Cursor consolidation tripled revenue; OP turned positive** **Revenue:** AI revenue was $7.56 bn, well above the Street (~$6.31 bn), +195% QoQ and +830% YoY. This reflects both hosting and applications. **AI solutions & infrastructure:** Revenue was $7.02 bn (Street ~ $5.84 bn), driven by two main factors. Hosting and Cursor consolidation were key. **① Compute hosting:** The Anthropic contract (~325k GPUs, 480 MW, ~$1.25 bn/month) contributed a full quarter for the first time, at ~ $3.75 bn (Q2 ~ $1.6 bn). The Google contract (110k GB300, 220 MW, $0.92 bn/month) began ramping in Sep and reaches full allocation in Oct. Reflection AI ($0.15 bn/month) began leasing in Jul. Ex-Cursor, this line was $6.41 bn, +192% QoQ, with hosting estimated around $4.5 bn. **② Cursor consolidated:** Closed on Aug 14, consolidated for ~1.5 months and contributed $608 mn, or ~$4.7 bn annualized. This adds a scaled AI dev entry point. **③ Grok model:** Grok revenue rose from ~ $0.6 bn last quarter to ~ $1.9 bn. This was supported by Grok 4.6 (8/12) and 4.7 (9/21), plus a persistent agent 'Grok Bot' rollout. By early Oct, paid WAU reached 0.8 mn, more than doubling in the past month, with enterprise deployment to be opened broadly within weeks. This momentum underpins application revenue growth. **X platform ads:** $0.54 bn, +47% QoQ but below the Street’s ~$0.66 bn. The recovery is ongoing albeit uneven. After completing AI ad infra rebuild in Q1 (automated buying, AI attribution, Grok-driven brand safety), ads have recovered for two straight quarters. Whether it can win back lost advertisers sustainably still needs to be monitored. **Margins:** AI GPM was 64.6%, up 800 bps QoQ (Street ~ 62.7%). OP was $430 mn, up by $1.7 bn QoQ and turned positive for the first time, driven by hosting premiums amid supply shortages. **⑤ Capex and cash flow: Q3 in line; Q4 guide nearly doubles; FCF remains deeply negative** Q3 Capex was $19.6 bn, broadly in line. AI accounted for $16.4 bn, Aerospace $1.6 bn, and Starlink $1.5 bn, with AI remaining the majority of spend. On compute, nominal capacity reached 1.7 GW by end-3Q26 (vs. 1.4 GW in 2Q26), slightly below the 1.74 GW expected. By end-Oct it was near 2.2 GW, with year-end at 2.5–3 GW (above prior 2 GW guide). Roughly 200–220k GB300 come online per month currently. From 2027, Vera Rubin becomes the primary ramp, shifting the supply curve further up. With capacity additions this quarter, management guided Q4 Capex to $35–40 bn, nearly double Q3. This implies full-year Capex of ~$83–88 bn, well above prior Street ~$65–70 bn. Operating cash flow improved QoQ to $9.6 bn on higher profits. But with $19.6 bn in Capex, FCF was roughly -$10 bn, underscoring heavy investment. **Dolphin Research’s take:** Overall, SpaceX delivered a broad beat: revenue +161% YoY and net margin up 37 ppt QoQ to 30.2%. Net income of $4.4 bn exceeded OP, helped by ~$1.0 bn other income, ~$0.9 bn interest income, and ~$0.4 bn tax benefits. On mix, the company has pivoted from 'rockets + satellite broadband' to 'AI compute as the largest incremental driver'. AI revenue topped 50% of total (52% vs. ~33% in Q2), signifying a new growth engine. OP turned positive mainly because AI shifted from 'cash burn for growth' to profit contribution. High-margin compute leasing underpins this turn and offers leverage. **More importantly, guidance massively beat: 1Q27 revenue guide far above expectations; targeting 13 GW compute and $500 bn–$1 tn ARR in 2027:** 1Q27 revenue guide of $33–37 bn is 40–60% above the Street (~$23.4 bn). This sets a high bar for the coming year and underscores demand strength. The $100 bn ARR target by end-2026 (annualizing Dec revenue) is unchanged. By end-2027, company-wide ARR could reach $500 bn–$1 tn, per management. The guidance rests on a more aggressive compute plan: near 3 GW this year locked in, 10 GW incremental in 2027 'highly certain', and a rising probability of exceeding 20 GW. Plans include adding ~20 GW power and cooling by end-2027 (aggregate ~23 GW). **Confidence stems from tight compute supply: unit pricing is slightly above the top end of the $30–50/W range, with new deployments achieving payback in under a year. Next year, payback could compress to ~6 months, with demand multiples of supply and H100 rental rates above initial levels.** **Thus, management prefers spot short-term rentals, betting on rising $/W; hosting contracts also allow early termination (e.g., Google can be terminated by either party with 90 days’ notice after end-2026). This is optimal in a shortage, but lowers revenue visibility and shortens contract duration, the largest valuation discount for AI.** **Cash and marketable securities were only $89.9 bn at quarter-end, insufficient for this pace of investment. SpaceX expects to raise capital by H1 2027.** In the four-plus months since the IPO, SpaceX shares saw a near half-off drawdown and rebound: from ~$202 close ($225.6 intraday) down to ~$108 in early Aug, then back to ~$163 in mid-Oct. This corroborates the post-Q2 view: like TSLA, SpaceX is priced by 'event catalysts' rather than near-term earnings. Long-dated bets like space compute and 10 GW AI expansion carry most of the EV, with little current profit, making the stock a 'probability-weighted option'. Over the past two months, catalysts across legs landed almost simultaneously. This clustering reinforced conviction and narrowed uncertainty bands. **① Starship Flight 14 reached orbit for the first time (Sep 28):** This was the 14th Starship launch and the 3rd V3 flight. The vehicle entered ~275 km orbit and deployed 26 operational Starlink V3 satellites, with per-satellite downlink of 1 Tbps and uplink of 160 Gbps. These are ~10x and ~22x V2 levels respectively. One Raptor shut down early, shortening the mission from six orbits (~10 hours) to ~3 hours and ending with a controlled splashdown in the Pacific. The booster soft-splashed down in the Gulf of Mexico as planned, and the spacecraft completed its first deorbit burn and controlled reentry. The 'orbit + payload deployment + controlled reentry' loop is proven, achieving the main objectives. In parallel, SpaceX launched three missions within 13 hours covering crewed (NASA Crew-13), commercial, and military payloads. These included an on-orbit test of Google TPU chips, boosting sentiment for orbital data centers. **② New AI hosting customers:** At end-Q2, only Anthropic revenue was recognized for hosting. Subsequently, Google, Reflection AI, and two unnamed customers signed, totaling ~$545 bn ARR for five customers. This corresponds to ~1.2–1.4 GW compute at an average ~$400–450 bn per GW ARR. The company also signed new cloud services agreements this quarter. Management noted per-watt pricing slightly above the top end of $30–50. New capacity payback is under one year, reflecting persistent supply tightness. **③ Cursor closed; Grok iterated rapidly:** **On Aug 14, SpaceX completed the all-stock $60 bn acquisition of Cursor:** Cursor became a wholly owned subsidiary, consolidating ~1.5 months in Q3 and contributing $608 mn. From Q4 it will be fully consolidated. The deal brings ~$4 bn ARR (about 75% enterprise; >64% of Fortune 500 use). Cursor’s interaction data has been used to further train Grok, improving model performance. **Meanwhile, Grok moved to monthly iterations:** On Aug 12, Grok 4.6 launched (1.5T params, 500K context, $2/$6 per mn tokens). The 'Artificial Analysis' score was 61, matching GPT-5.6 Sol and entering the top tier. Grok Bot launched in Aug, entering the computer-use agent space via cloud VMs. Paid WAU reached 0.8 mn by early Oct, more than doubling in a month, keeping narrative heat high. Grok 4.7 launched on Sep 21, lifting params to 2.1T (~40% above 4.6), maintaining a strong value-for-money pricing strategy. Iteration velocity remains elevated. Musk previewed that Grok 4.8, 4.9, and 5.0 will arrive by year-end. He said 4.9 'may reach frontiers' and 5.0 'could beat any model', with Grok 5 (6T+ params, trained on Colossus 2) expected in Dec. SpaceX’s AI business is extending from 'selling compute' to 'selling models and apps'. Near term, revenue still leans on premium hosting, but Grok monetization, especially enterprise, grows more important over time. The long-term aim is to close the loop across compute, data, models, and agent products. This integrated approach can support durable economics. **④ Low-band spectrum + FCC approvals clarify the mobile roadmap:** On Oct 8, SpaceX announced a ~$6–8 bn purchase of Grain Management’s US-wide 800 MHz low-band spectrum (up to 14 MHz paired) for coverage and indoor penetration. Together with EchoStar’s ~65 MHz 2 GHz mid-band for high-capacity throughput, Starlink now has a 'low + mid' terrestrial spectrum combo. Coupled with FCC approval for 15k direct-to-cell satellites, positioning shifts from 'satellite fill-in' to a 'fourth facilities-based operator'. This expands optionality and strategic leverage. Near-term revenue is limited: satellites alone cannot solve urban high concurrency and indoor penetration, and the ~80 MHz total spectrum is about one-fifth of the Big Three’s. The Big Three also refuse MVNO leases, implying a need to build terrestrial networks (see Starlink section). Still, holding spectrum is an 'option' in itself. The more credible the self-build threat, the stronger the leverage to force MVNO access from incumbents. **Bottom line, the print itself was strong: revenue and profit both beat, OP turned positive, and guidance far exceeded expectations. Post-earnings stock action faces near-term unlock overhang, while mid-term depends on key milestones: full Starship reuse (Flight 15 dual-stage catch), V3 rollout pace and mobile ground network plan, whether 2027 compute expansion orders and financing match, and whether the $500 bn–$1 tn ARR by end-2027 materializes. Specifically:** **a. Near term: ~1.3 bn-share unlock follows the print, the most concentrated technical overhang** On Aug 6, the first 911.5 mn shares (~7% of Mkt Cap) unlocked after Q2, but selling pressure was modest. Prior declines had absorbed marginal supply, and shares rose 16% the next day (Aug 7). Per lock-up terms, the peak unlock arrives on the 2nd trading day after Q3 results (around Oct 22): 28% of vested employee equity unlocks alongside institutions. This is ~1.3 bn shares, ~10% of total shares outstanding, the largest since listing. On Dec 8 (180 days post-IPO), the remaining employee shares unlock, potentially lifting float to ~40% of total. The stock has already rebounded ~50% from the lows to >$160 pre-earnings (~$2.2 tn Mkt Cap). This time, profit-taking willingness is stronger than in Aug. But guidance that far beat expectations offers buyers more confidence and a stronger bid. **b. Launch: Flight 15 dual-stage catch is the largest post-IPO positive catalyst** Massive Starship cost-downs are foundational to V3 deployment at scale and orbital data centers. Musk said Flight 15 is expected in Nov and will attempt simultaneous booster and ship recovery with the tower. FAA approval is expected to cover return and tower catch. Musk also said 'success probability is higher than failure' (previously self-assessed at 50–60%). If successful, it will be the first fully recovered orbital launch vehicle. Afterward, early 2027 will re-fly the second stage and attempt on-orbit refueling. The goal is weekly launches in H1 and twice weekly in H2, starting deployments of mobile V2 and Starmine satellites. The new Louisiana pad targets a first launch in 2028 and scale-up in 2029. This marks a cost watershed for V3 scale-out and beyond. Without reuse, V3 launch cost is ~$600–900/kg. With 5–6 reuses, it can drop below $200/kg, dramatically improving unit economics and confidence in long-term visions. **c. Starlink: near term watch V3 rollout; mid-term, how to solve the mobile 'ground network'?** **Broadband:** Management said hundreds of V3 satellites are needed for users to clearly feel capacity gains, targeted by mid-2027. This requires Starship to enter high-frequency deployments from Flight 15/16. At full load, a single V3 launch adds >20x the capacity of a Falcon 9 Starlink mission. Musk expects constellation bandwidth to reach 100x today’s within ~3 years once V3 is fully ramped. **Mobile:** With Grain low-band plus EchoStar mid-band, the spectrum puzzle is mostly set. Next watch whether SpaceX participates in the Upper C-band auction in Apr 2027 (~160 MHz, the largest mid-band block in the US in years). For ground networks, management’s plan is: satellites carry most traffic on S-band, 800 MHz handles indoor penetration, and small cells are added at Starlink-equipped homes, enterprises, and towers, reusing sites. The target is service launch in early 2028. This may be lighter than the market’s prior '>$100 bn nationwide self-build' assumption, though spend is undisclosed; scaled commercial rollout likely not before 2029. Near term, track V3 deployment milestones, mobile V2 launches, and the specialized meeting where the ground network plan will be detailed. Execution here is key. **d. AI: $100 bn ARR looks secure; for 2027, watch orders, financing, and pricing** In Q2, we estimated that without new megadeals, AI ARR by year-end would be ~$50 bn, short of the $100 bn target. New hosting contracts in Q3 largely closed the gap, and management reaffirmed the target. **Updated Dec 2026 ARR breakdown:** ① Compute hosting: five disclosed customers total ~$54.5 bn annualized, plus newly signed agreements this quarter and leases from added capacity as year-end compute rises to 2.5–3 GW. ② Cursor: ~$4.7 bn annualized in the consolidation period; at 8–12% MoM growth, year-end ~$6–8 bn. This assumes continued enterprise demand. ③ Grok: year-end ~$8 bn, with Grok 5 expected in Dec, potentially a step-change in capability. This underpins application ARR. ④ Ads: ~$2–2.5 bn annualized, reflecting two quarters of AI-led recovery. This remains a smaller contributor. ⑤ Starlink + Aerospace: year-end ~$30 bn. This adds base connectivity and mission revenue to the mix. Totaling ~$100.5–103 bn, plus new agreements and added capacity, the $100 bn target looks very likely. Execution risk remains but is mitigated by demand-supply imbalance. **Looking to 2027:** 1Q27 revenue guidance annualizes to $132–148 bn, making $100 bn ARR just a starting point. Musk further suggested end-2027 ARR could reach $500 bn–$1 tn. To get there, four questions remain: **1) What does $500 bn–$1 tn ARR imply?** It is 5–10x the $100 bn end-2026 target. At the low end, with 13 GW compute, 60% external (about 8 GW), and $50/W, hosting ARR would be ~$390 bn. Add applications at ~$80 bn and connectivity plus aerospace at ~$45 bn for a total of ~$515 bn. This is the lower-bound scenario. At the high end, hosting alone would require 17–18 GW external sales, beyond the optimistic 20 GW external capacity. This demands massive Grok Bot and Cursor scale-up, effectively a long-dated option. **2) Where does the money come from?** Q4 guidance lifts 2026 Capex to $83–88 bn. Adding 10 GW in 2027 at $30–35 bn per GW implies $300–350 bn, while cash and marketable securities were under $90 bn at Q3-end. Management prioritizes operating cash flow, supplemented by debt (discussions ongoing) and industry partnerships, maintaining investment-grade ratings. Equity financing was not mentioned. New capacity begins contributing cash from H1 2027, but relative to spend it is modest. Without new funding, cash likely runs out by Q1 2027, implying large debt raises by Q4 or H1 next year. **3) Will orders keep up?** The 1Q27 guide is based on contracts visible over the next six months, covering only through H1 next year. With 60% external, year-end would require ~8 GW in external contracts. We estimate ~1.2–1.4 GW already signed, leaving a >5 GW gap. Part of this gap is intentional, as management prefers spot short-terms and avoids locking future capacity. Given demand multiples of supply, matching should improve each quarter. Still, we need to see actual lease signings before drawing firm conclusions. **4) How long can high pricing last?** Unit pricing slightly above $30–50/W top end with sub-1-year payback implies 3–5x faster paybacks than peers. This is premium pricing driven by tight supply. Management’s logic: model progress raises the 'useful intelligence' per watt, with H100 rentals now above initial levels. Hence, favor spot rentals and reprice on renewal, effectively holding a call option on price in a shortage. **Key metrics are prices on new deals and renewals. The risk is a supply-demand inflection when new capacity such as Vera Rubin ramps in 2027, which could compress today’s premiums.** We view this as a beat on results and an even larger beat on guidance. Near term, the ~1.3 bn-share unlock could cause technical noise; mid-term, the stock’s main driver shifts to whether 10+ GW of new data centers launch on time and leases match spend. As long as new capacity leases at current prices, sub-1-year paybacks make heavy Capex a growth accelerant rather than a burden. This underpins the bull case into 2027. At the current ~$1.95 tn Mkt Cap, valuation is under 4x the low end of end-2027 ARR and under 2x the high end. It looks 'cheap' if ARR is delivered. If end-2027 ARR hits the $500 bn lower bound, even at a conservative 30% net margin and 15–20x P/E, implied EV is ~$2.3–3.0 tn. That leaves notable upside vs. today. Should the unlock drive a pullback, we see a medium-to-long-term buy-the-dip. Positive catalysts include a successful Flight 15 dual-stage catch, delivering the $100 bn ARR by Dec, Grok 5 outperformance, execution of a debt-heavy 2027 financing plan, and new compute leases with 'no price cuts and longer terms'. Separately, Musk openly addressed a potential merger with TSLA for the first time: he personally thinks it is a good idea, but the independent special committees have not agreed on any terms. TSLA would require approval by 50% of all outstanding shares, a high bar, so the market is overestimating the odds by treating it as a near-certainty. **Earnings charts:** \ Risk disclosure and statements for this article: [Dolphin Research disclaimer and general disclosure](https://support.longbridge.global/topics/misc/dolphin-disclaimer) **Dolphin Research archive:** **Deep dives:** **<**[**From daydream to megabucks: is SpaceX really that sci-fi?**](https://longbridge.cn/en/dolphin/post/41771513)**\>** **<**[**SpaceX: sky mesh, on the way to dominance?**](https://longbridge.cn/en/dolphin/post/41874137)**\>** **<**[**SpaceX: AI burn continues, is 'space compute hegemony' the ultimate weapon?**](https://longbridge.cn/en/dolphin/post/42313269)**\>** **<**[**SpaceX: space compute, another Musk pitch or the real future?**](https://longbridge.cn/en/dolphin/post/42514128)**\>** **<**[**SpaceX’s challengers: can Bezos and China’s teams catch up?**](https://longbridge.cn/en/topics/39769288?app_id=longbridge&utm_source=longbridge_app_share&share_track_id=ed3db402-f682-49b5-901e-60a99aa584bd&invite-code=4NOXYT&locale=zh-CN&community_badge=1&profile_following_followers_activities=1)**\>** **<**[**Musk’s new trump: can SpaceX remake 'space economics'?**](https://longbridge.cn/en/topics/38852185?app_id=longbridge&utm_source=longbridge_app_share&share_track_id=7573e7e7-3d3e-44a5-b65a-61f4e4842dfc&invite-code=4NOXYT&locale=zh-CN&community_badge=1&profile_following_followers_activities=1)**\>** **Earnings reviews:** **\** ### Related Stocks - [SPCX.US](https://longbridge.com/en/quote/SPCX.US.md) - [UXSD.SG](https://longbridge.com/en/quote/UXSD.SG.md) - [SPCH.US](https://longbridge.com/en/quote/SPCH.US.md) - [SPCF.US](https://longbridge.com/en/quote/SPCF.US.md) - [SPAL.US](https://longbridge.com/en/quote/SPAL.US.md) - [SNK.US](https://longbridge.com/en/quote/SNK.US.md) - [SSPC.US](https://longbridge.com/en/quote/SSPC.US.md) - [SPCU.US](https://longbridge.com/en/quote/SPCU.US.md) - [LOFF.US](https://longbridge.com/en/quote/LOFF.US.md) --- > **Disclaimer: This article is for reference only and does not constitute any investment advice.**