Bull & Bear

Bull and Bear

Verdict: Avoid — at a $1.75T IPO mark, the operating crown jewels are real but the price already pays for the AI option, and the governance structure is engineered to leak any upside before it reaches a minority holder. Bull's strongest evidence is that Starlink is genuinely a top-10 global telecom hiding inside a launch monopoly; Bear's strongest evidence is that the controller owns 85.1% of votes on 51% of equity, a board director's firm is counterparty to $20.2B of SpaceX-guaranteed xAI lease obligations, the CFO's bonus trigger was switched from free cash flow to Adjusted EBITDA weeks before the S-1, and the inaugural Starship V3 flight scrubbed the day after the S-1 dropped. The single tension is the AI segment: Bull frees it; Bear charges full freight for a capital vortex. Until AI capex/revenue starts bending toward 100% on a trailing-twelve-month basis and the first Section 404 attestation is clean, there is no version of this name where the minority underwrites the risk and captures the upside.

Bull Case

Bull target: $2.25T equity value (~+29% from the $1.75T IPO mark, ~+49% from the $1.51T May 2026 private mark) over 18 months. Method: sum-of-the-parts — Starlink ~$900B (38× forward $24B FY27 EBITDA on subs scaling toward 18–20M), Space ~$350B (Starship V3 commercial debut + NSSL Phase 3 cash), xAI ~$1.1T (Anthropic-anchored Colossus + ~12% of OpenAI's implied private mark), less a 12% holdco discount. Disconfirming signal: Starlink US residential ARPU falls below $50/mo with quarterly net adds below 1.5M and HughesNet net adds turn positive for two consecutive quarters.

Bear Case

Bear downside target: $750B enterprise value (~50% derate from the $1.5T private mark; ~57% from the $1.75T IPO ask) over 12–24 months. Method: Bear SOTP — Connectivity $300B (12–15× FY26E segment EBITDA, half-credit for ARPU floor under a Kuiper price war), Space $125B (replacement + a partial Starship option), AI $100B (Anthropic-anchored revenue × 12× with execution and governance overlay), less a 20% holdco / controlled-company discount; the $11.8B X-deal goodwill is treated as an impairment candidate, not a value-add. Cover signal: clean PwC Section 404 attestation in the second post-IPO 10-K with no remediated material weakness, combined with AI capex/AI revenue below 200% TTM, Starlink ARPU stable above $60 for two consecutive quarters with sub adds intact, and an audit-committee-blessed protocol gating future Musk-affiliate transactions.

The Real Debate

Verdict

Avoid. Bear carries more weight because the case does not rest on disputing Bull's strongest fact — that Starlink is a top-10 telecom built on a launch monopoly — it rests on what the IPO mark forces the buyer to pay for on top of that crown jewel, and on a governance architecture that determines who captures any upside. The decisive tension is AI: Bull treats xAI as a free option, but at a $1.75T mark with capex running 397% of segment revenue in FY25 and roughly 940% in Q1-26, the buyer is paying full freight for an unproven AI bet inside a controlled-company structure where a director's firm is counterparty to $20.2B of SpaceX-guaranteed lease obligations and the CFO's bonus trigger was rewritten from FCF to Adjusted EBITDA weeks before the S-1. Bull could still be right if the next four quarters show AI TTM capex/revenue bending below 200% with a clear revenue inflection and Starship V3 delivers a paying-customer payload to orbit in 2026 — both are achievable, and the operating engine underneath has earned the benefit of the doubt on milestones before. The durable thesis breaker that would change the verdict is a clean first Section 404 attestation with no remediated material weakness plus an audit-committee-blessed protocol gating future Musk-affiliate transactions; the near-term evidence marker is V3 reaching orbit with a commercial payload in calendar 2026. Until that combination prints, this is a bull-case bundle priced at zero margin of safety inside a structure engineered to leak value to the controller — an institutional pass, not a short, and not a private-market entry at the IPO mark.