LEO viability · head-to-head

Four LEO systems on the same denominators, each carrying its tier —disclosed fact, desk derivation, soft estimate, or honest hollow. The verdict publishes only when disclosed facts support it reliably; otherwise we withhold or suppress it rather than guessing. Move a slider to re-run the model.

Publishable
IRR(-8.5% vs hurdle)

PUBLISHABLE: every ancestor is disclosed or a transparent slider (no un-anchorable hole, no soft analyst stand-in). Front-loads the EUR 2bn at t0 (conservative). This is the go-forward continuation IRR, not a full-program IRR (Gen1 original build cost is undisclosed).

disclosed 8derived 15estimated 8hollow 0
Satellites
disc
Capacity (gross)
der
Capacity (sellable)
der
Cost / sat
der
Capex / sellable Gbps
der
Revenue
der
Replenish /yr
der
Assumptions · 8 sliders

Return is a go-forward IRR: the disclosed €2bn continuation capex against a revenue ramp (current LEO revenue × growth multiple × EBITDA margin). These sliders are that model's inputs.

Satellite design service life6.5 yr
5 yr7.5 yr
Assumed LEO EBITDA margin at maturity35.0%
15.0%60.0%
Steady-state LEO revenue as a multiple of FY25-263.00×
1.00×6.00×
Years to steady-state LEO revenue5 yr
2 yr10 yr
DCF horizon (constellation operational to ~2034)9 yr
5 yr15 yr
Assumed sellable fraction of gross capacity10.0%
3.0%30.0%
Additional replenishment spend within horizon€0/yr
€0/yr€500M/yr
FX rate EUR→USD (for cross-system comparison)1.08×
0.95×1.25×
Funding & scope history →
Publishable
Cash flow · QIRR · modeled

PUBLISHED verdict, disclosed-anchored: Connectivity self-funds — adj. EBITDA exceeds segment capex by ~$1.2B in the quarter. No assumptions needed.

disclosed 14derived 9estimated 10hollow 1
Satellites
disc
Capacity (gross)
disc
Capacity (sellable)
der
Cost / sat
est
Capex / sellable Gbps
der
Revenue
disc
Replenish /yr
—
Assumptions · 6 sliders

The verdict is disclosed free cash flow — no model, no assumptions. These sliders drive only the illustrative IRR, which rests on an estimated cumulative capex (hence withheld).

Satellite design service life5 yr
3 yr7 yr
Opex fraction (0 — the input is already EBITDA)0.0%
0.0%30.0%
Years to steady EBITDA (near-steady already)2 yr
1 yr6 yr
DCF horizon8 yr
5 yr15 yr
Assumed sellable fraction of gross capacity7.4%
2.0%15.0%
Modeled annual replenishment on top of capex$0/yr
$0/yr$6.00B/yr
Funding & scope history →
Withheld
IRR

COMPUTABLE but NOT verdict-publishable: the capex is now the disclosed $5.2B, but the return still hangs on the $4B 2032 revenue FORECAST × an assumed EBITDA margin — a forward forecast, not disclosed actuals. Per §9.1 the engine computes a number for the sandbox but the verdict gate stays red.

disclosed 18derived 12estimated 14hollow 4
Satellites
disc
Capacity (gross)
est
Capacity (sellable)
est
Cost / sat
est
Capex / sellable Gbps
est
Revenue
—
Replenish /yr
—
Assumptions · 8 sliders

IRR is modeled on the disclosed $5.2B program capex against Telesat's Aug-2026 $4B 2032 revenue forecast × an EBITDA margin. Verdict stays withheld — that revenue is still a forward forecast, not disclosed actuals.

Assumed mature EBITDA margin on 2032 revenue60.0%
40.0%90.0%
Opex fraction (0 — steady figure is already EBITDA)0.0%
0.0%30.0%
Years from now to the 2032 steady EBITDA6 yr
3 yr10 yr
DCF horizon14 yr
8 yr20 yr
Assumed sellable fraction of installed capacity5.0%
3.0%15.0%
Assumed replenishment spend within horizon$0/yr
$0/yr$600M/yr
Falcon 9 price per launch (no Telesat figure — analyst-anchored)$70.0M
$50.0M$100M
Installed capacity per satellite (desk scenario)40 Gbps
25 Gbps60 Gbps
Funding & scope history →
Suppressed
IRR

Verdict SUPPRESSED (null) — but no longer from a data hole. With cost now anchored, the treadmill (~$3.1B/yr) exceeds net revenue at the default $3B/yr scenario, so the DCF has no positive root: at Quilty's cost, Kuiper does not close on any disclosed-scale revenue, and no IRR exists to show. It only turns positive once you slide revenue above break-even — a level Amazon has not disclosed. Move the revenue/opex/bus-cost sliders to find where it closes.

disclosed 11derived 14estimated 11hollow 0
Satellites
disc
Capacity (gross)
der
Capacity (sellable)
der
Cost / sat
der
Capex / sellable Gbps
der
Revenue
—
Replenish /yr
der
Assumptions · 7 sliders

Bus cost is now anchored to Quilty's analyst estimate (~$1.75M/sat, movable $0.5–2M), so cost/sat and the ~$3.1B/yr replenishment treadmill compute. The IRR verdict stays suppressed: at Quilty's cost the treadmill exceeds net revenue at any disclosed-scale scenario, so no IRR exists to publish. Slide revenue up to find where it closes — but that revenue is a scenario, not a fact, which is why no verdict is published.

Assumed design service life5 yr
3 yr7 yr
Assumed steady-state annual revenue$3.00B/yr
$500M/yr$12.0B/yr
Assumed opex as fraction of revenue35.0%
15.0%60.0%
Assumed years to steady-state revenue6 yr
2 yr12 yr
Assumed DCF horizon15 yr
10 yr25 yr
Assumed sellable fraction of gross capacity5.0%
2.0%30.0%
Gen1 satellite bus manufacturing cost (Quilty analyst estimate)$1.8M
$500k$2.0M
Funding & scope history →
Derivation

Every value traces to a disclosed source, a movable assumption, a deskderivation, or an honest hollow. This is the full working at the current slider settings. Change a slider and reopen to see it move.

Coverage meter reads left→right by evidence strength: disclosed →derived → estimated →hollow. Money normalizes to USD with a movable FX rate (OneWeb carries a EUR→USD slider; the others report in USD). Sliders recompute the model live in your browser.