Valuation v0

point = $30M × documented drivers. The confidence band widens as confidence falls (±22% / ±35% / ±50%). Curated $NNN M+ overlays are class M opinions, secondary to the model range. Valuation basis is always model.

Clarke’s Slot Terminal fair-value figure is a labeled model: a documented heuristic so operators and investors can inspect an implied range and its drivers. It is not a live exchange price, an appraisal, or an offer to transact.

Formula

point = $30M baseline
      × arc desirability
      × coverage (GDP/pop heuristic)
      × occupancy (co-located satellite count)
      × remaining life / occupancy quality
      × operator tier
      × spectrum (bands, when known)
      × scarcity (congestion score)
      × license / brought-into-use

Confidence interval: ±22% (high) / ±35% (medium) / ±50% (low) around the point. High confidence is reserved for curated positions; government/military-only UCS users are flagged non-commercial and forced to low confidence.

Curated valueEstimate strings (e.g. $350M+) are a hand estimate / curated opinion (class M). They are secondary to the v0 range and do not replace it. Valuation basis is always model.

Inputs

Driver	Source	Notes
Arc desirability	Longitude band	Same corridors as the previous heuristic (Europe, North America, Asia, …)
Coverage GDP/pop	`src/lib/coverage-proxy.ts`	Public regional GDP and population shares, rounded. Not a measured beam footprint.
Occupancy	TLE-primary GEO longitude ±0.4°	Space-Track TLE when age/quality pass; UCS catalog fallback otherwise. Not an FCC/ITU assignment.
Remaining life	UCS launch date + expected lifetime	UCS dates are `M/D/YY`; v0 expands 57–99 → 1957–1999 and 00–56 → 2000–2056. Missing lifetime → multiplier 1.0 (no invented youth/expiry). UCS expected lifetime is design life, not remaining licensed life — many GEO birds fly past it.
Operator tier	Operator name vs tier-1 list	SES, Intelsat, Eutelsat, …
Spectrum	Curated band tags	UCS-derived rows often have no band data
Scarcity	Clarke congestion 0–100	Density, co-location, operator contention on TLE-primary occupancy longs
License / BIU	FCC SSAL + UCS occupancy	Paper filing vs in-orbit vs non-US operating

History

npm run seed:valuations writes 30 daily snapshots per slot into data/terminal.db. Until a real daily job exists, those points are a deterministic model path (smooth drift from the current v0 point), labeled backfill even when read back from SQLite — not observed trades. Re-run the seed after ingest or model changes. The sparkline is not on the default Terminal; it sits behind Experimental disclosure as “MODEL BACKFILL — not trades.”

What v0 is not

Not a live bid/ask or last trade
Not an ITU deed or FCC license valuation
Not advice to buy, lease, or file
Simulated capacity books are Experimental, labeled SIMULATED, derived from the same midpoint + congestion, and are not on the default Terminal first screen