A bent-pipe satellite was a single-tenant building in the sky. A flexible-payload satellite is a multi-tenant building. The difference decides whether the asset is financeable, and on what terms.
For the three decades from the first commercial GEO comsats to the mid-2010s, a satellite's entire economic value rode on one commercial assumption: the customer who signed the anchor capacity deal at procurement would still be there fifteen years later, paying the same price, on the same beam plan. That assumption has broken more than once — long-haul telephony, analogue DTH, bulk IP backhaul in Africa — and every time it broke, the satellite behind it was worth a fraction of its book value.
The payloads built since roughly 2020 do not depend on that assumption any more.
Single-tenant to multi-tenant
A legacy "bent-pipe" payload is hard-wired at the factory. Coverage, frequency plan, channel count and per-channel bandwidth are all fixed at contract signature — typically three to four years before launch and fifteen to twenty years of on-orbit life. If the traffic pattern changes, the operator either flies the satellite half-empty or takes the write-down.
A software-defined payload — of which Thales Alenia's Space INSPIRE is the reference platform and Airbus's OneSat is the direct competitor — performs the RF processing in a digital engine, forms beams electronically rather than through fixed feed horns, and treats the entire coverage and frequency plan as a load on the flight software. In-orbit reconfiguration takes minutes.
What actually changes in orbit
The reconfiguration is not theoretical — digital processors have been flying since Inmarsat's I-6 generation entered service, and the fully software-defined platforms extend the same idea to the whole payload. Typical in-orbit changes:
| Dimension | Practical use |
|---|---|
| Frequency | Move from C-band to Ku-band over a region after a regulator re-farms spectrum for 5G. Avoid a multi-billion write-down of affected transponders. |
| Beam shape | Collapse three medium-gain beams into one high-gain beam over a disaster zone or conflict area. Triple downlink EIRP to the priority footprint, revert after. |
| Beam position | Follow a moving cruise ship, deployed force, or temporary event. Maintain a single anchor-tenant contract over the asset's whole life. |
| Capacity | Double bandwidth allocated to a gateway beam during peak business hours; halve it overnight. Smooth peak-hour congestion without new hardware. |
| Segmentation | Isolate a sovereign tenant on its own transponder subset with separate encryption and TT&C. Sell the same satellite to civil and government customers without cross-contamination. |
The aviation analogue
Aviation finance has lived with the airframe / engines split for forty years. The lessor owns the airframe; the operator maintains the engines and the hot section to contractual return conditions; hours-on-wing define what the asset is worth at redelivery. The same logic maps cleanly onto space infrastructure.
On a ground-station sale-leaseback, the civil structure, pedestal, waveguide runs and reflector are the "airframe" — 25-to-30-year life, structural residual. The feed, low-noise amplifier, high-power amplifier, modems and monitor-and-control software are the "engines" — refreshed at tenant cost every seven to ten years, with tested return conditions at lease end. A ground-station lease is written in that vocabulary.
“GEO is real estate at a specific longitude. LEO is fleet. MEO sits between. Each regime finances like its aviation counterpart — a Grade-A building, a fleet of regional jets, a small widebody portfolio.”
What this means for the lessor
A software-defined satellite is a materially better residual-value asset than a bent-pipe. End-of-first-lease repositioning is a software load, not a capex event. Tenant-replacement risk is lower because the next customer can be served without reconfiguring the hardware. Depreciation schedules driven by technological obsolescence — the old risk of a C-band-only payload becoming stranded when the market moved to Ka — materially soften.
| Profile | Residual basis | Second customer | Re-lease optionality |
|---|---|---|---|
| Bent-pipe GEO, legacy | Tied to the first customer's contract | Rarely available without a new payload | Effectively none |
| Software-defined GEO | Can be underwritten against re-use | Served by a software load | Material, once in-orbit data supports it |
The underwriting caveat is honest: the on-orbit reliability curve for this generation of payloads is still under-sampled. Both reference platforms have run years behind their original schedules: the first OneSat, for Inmarsat, was planned from 2023 and has slipped to 2026, and CNES lists the first Space INSPIRE launch for 2027. Early-life anomaly data on both lines has to build up before anyone can extrapolate residual values with confidence.
But the direction is clear. The asset class is becoming bankable for a leasing vehicle because the payload has stopped locking the satellite to a single customer. That is what flexibility really buys you.
Sources
Thales Alenia Space — Space INSPIRE product documentation. Airbus Defence and Space — OneSat brochure. Euroconsult, Satellites to be Built & Launched 2024. ESA ARTES programme documentation. SES and Inmarsat operator annual reports (2024, 2025). Accessed 29/06/2026.