A geostationary communications satellite in shadow above a lit Earth limb, a single chartreuse marker on the solar array

On 12 August 2026 Telesat announced that a subsidiary had closed a US$120M secured term loan from an unaffiliated third-party lender. The borrower is a subsidiary of Telesat GEO Inc. that is a non-guarantor under the documents governing Telesat GEO's existing term loan and senior notes. The money is for general corporate purposes, the facility runs four years, and it prices off SOFR plus a margin. The share price fell around 13% the following day.

Read that borrower description again, because it is the whole article. Four months before Telesat GEO's Term Loan B and its 5.625% senior secured notes fall due in December 2026, new money entered the group at an entity the existing lenders do not have a guarantee from. The satellites did not change. The fleet did not change. What changed was where the claim sits.

What the quarter actually showed

The operating picture behind the borrowing is not obscure. In the quarter to 30 June 2026 Telesat's GEO segment produced revenue of C$78M, down 26% year on year. Satellite utilisation stood at 60%. GEO backlog was approximately C$900M, helped during the quarter by a five-year extension on a broadcast video contract. Consolidated revenue was C$79M against C$106.1M a year earlier, and the group reported a C$559M net loss, most of it a C$475M non-cash movement in the fair value of Lightspeed financing warrants together with foreign-exchange effects on US dollar debt.

The contrast inside the same group is the point. Telesat's LEO backlog stood at about C$1.1bn at 30 June, and roughly C$5.6bn pro forma once the Enhanced Satellite Communications Project – Polar contract signed with Canada's Defence Investment Agency in early August is included. Telesat has expanded the Lightspeed constellation from 156 to 225 satellites and raised its 2026 investment guidance to C$1.3bn to C$1.5bn. SpaceQ reported around C$2.3bn of legacy GEO debt to be refinanced ahead of the December maturities, against C$160.8M of cash at the legacy business.

So one half of the company holds a fifteen-year sovereign contract and a growing constellation. The other half holds an ageing fleet at 60% utilisation and the debt.

Value migrates before it is written down

This did not begin in August. Telesat moved 62% of the equity in Lightspeed out of Telesat Canada and into an unrestricted subsidiary of the parent, announced in September 2025. In January 2026 an entity acting for legacy bondholders sued in Ontario and New York, alleging breach of the credit agreement and a fraudulent transfer, on the argument that lender recoveries were sacrificed to preserve equity. Telesat called the claims without merit and said the distribution followed a robust governance process and complied with the debt documents and applicable law. That dispute is live and unresolved, and nothing here anticipates its outcome.

The legal merits are for the courts. The financing lesson is available now and is independent of who wins. A creditor lending at the operating-company level owns a claim on an entity, not on a spacecraft. Entities have perimeters, and perimeters have doors: unrestricted subsidiaries, non-guarantor borrowers, investment baskets, restricted-payment capacity. Every one of those doors was negotiated years earlier, in a document drafted when the fleet was performing and nobody was reading the definitions closely.

By the time a maturity wall is visible, the covenant package is fixed. The only question left is what the documents permitted, and that question was answered at signing.

Aviation learned this in the 1990s and 2000s, which is why the market moved decisively towards asset-level structures. A lessor holding an aircraft through a ring-fenced SPV under a true operating lease does not depend on the airline's group structure staying still. It holds title. It has security over an identified serial-numbered asset, registered in a public register, with contractual remedies that survive the lessee's distress.

Where the space equivalent is weaker, and where it is not

The obvious objection is that a satellite cannot be repossessed. That is true and it is not the whole story.

The international machinery is genuinely thin. The Protocol on Matters Specific to Space Assets to the Cape Town Convention was adopted in Berlin on 9 March 2012 and is still not in force: four signatures and one ratification against the ten required. There is no orbital equivalent of the International Registry that made aircraft security enforceable across borders. Anyone underwriting a spacecraft is relying on domestic security law, licence conditions and contract, not on a treaty.

What survives is control of the things that make the asset earn. A GEO satellite without its telemetry, tracking and command facility, its gateway earth stations, its filings and its licensed spectrum is an object in a slot, not a business. Those elements are terrestrial, identifiable, and capable of being owned separately and pledged. This is the practical reason ground-segment infrastructure prices differently from the spacecraft: you can stand next to it, and a receiver can take it.

The corollary is uncomfortable for corporate lenders. In a restructuring of a satellite operator, the recovery argument is rarely about the residual value of the fleet. It is about contracted backlog, the licences, the ground network, and whether any of those sit inside the perimeter you lent against.

What a credit committee should take from August

Three things, none of them specific to Telesat.

First, contracted duration is the collateral. A C$900M backlog against a fleet at 60% utilisation is a different asset from a fifteen-year sovereign contract, and the debt should have been sized to the former rather than to the segment's history.

Second, the guarantee perimeter is a live variable, not a static fact. Underwriting should test what the documents allow to leave, and price the optionality the borrower has retained. If new money can enter at a non-guarantor subsidiary, existing lenders are structurally behind it in the entity that received it.

Third, ownership beats a covenant. A finance lease or sale-leaseback in which the financier holds title through a dedicated vehicle, with direct security over the ground segment and an assignment of the transponder or capacity contracts, does not need to litigate the meaning of an unrestricted-subsidiary definition three years later. It is the structure aviation converged on for exactly this reason, and the argument for it in space is stronger, not weaker, because the enforcement regime is less developed.

None of this makes GEO uninvestable. It makes the difference between financing an operator and financing an asset a great deal more expensive to get wrong.

Sources