Executive Summary
Five categories of public data — hyperscaler SEC filings, NASA engineering literature, xAI disclosed financials, neocloud disclosures, and FCC/merger filings — cross-referenced to surface a disclosure profile that institutional investors, D&O underwriters, and risk officers should evaluate before the S-1 is filed. Every major hyperscaler extended server useful life from 3 to 6 years between 2020-2023. SpaceX will inherit that convention for orbital compute assets operating in a categorically different physical environment. The question is whether the S-1 will explain what physical model supports it.
Key Metrics
The Five Layers
| Layer | Data Source | What It Shows |
|---|---|---|
| 1. The Terrestrial Precedent | Hyperscaler SEC 10-K filings | Industry-wide depreciation extension from 3 to 6 years — the convention SpaceX will borrow from |
| 2. The Orbital Asset Problem | Engineering literature / NASA / ISS operational data | Physical degradation vectors in space that have no analog in terrestrial depreciation models |
| 3. The xAI Integration | Disclosed financials / FCC dockets / press filings | xAI's burn rate consolidated into combined entity — $2.5B loss on $250M revenue obscured by merger structure |
| 4. The Cascade Model Failure | Hyperscaler earnings calls / neocloud disclosures | The justification for 6-year schedules requires hardware repurposing — impossible for fixed orbital assets |
| 5. Multi-Signal Convergence | All of the above + merger structure + FCC filings | Five signals from five data sources — each with a reasonable explanation — together forming a disclosure gap |
Convergence Summary
| Signal | Source | Data Point | Innocent Explanation |
|---|---|---|---|
| Depreciation precedent | Hyperscaler 10-Ks | 6-year convention established for terrestrial hardware under very different physical conditions | Common industry standard, defensible for CPU/GPU workloads |
| No actuarial history | Engineering literature | Zero multi-year orbital GPU deployments exist to base useful life assumptions on | Emerging asset class — estimates inherently speculative |
| xAI burn rate | Disclosed financials | $2.5B annual loss on $250M revenue, consolidated into profitable SpaceX balance sheet | Early-stage infrastructure investment with long-term payoff |
| Merger structure | Bloomberg / CNBC / FCC | Deal structured to avoid triggering debt repayment provisions; X platform included indirectly | Standard merger mechanics, legally sound |
| Valuation multiple | Reported / Kalshi | $1.5T target = ~94x trailing sales, ~500x trailing earnings; 76% IPO probability before Sept 2026 | Growth company multiple reflecting future optionality |
The data is always there. When the S-1 is filed, look past the Starlink growth charts and the xAI narrative. Find the line that says "useful life of orbital compute assets." Find the footnote that explains what physical model supports it. If that footnote does not exist, you have found the disclosure gap. The question is whether anyone is connecting it.
Subscribe for more cross-reference analyses like this.
Subscribe to the Weekly Brief