In August 2023, Telesat announced that its Lightspeed low-Earth-orbit constellation was “fully funded.” The company had locked in CA$2.5 billion in financing, signed MDA Space as prime satellite manufacturer, and secured a 14-launch agreement with SpaceX. The first satellites were to launch in mid-2026; polar and global broadband service would follow by late 2027. The hard part, it seemed, was behind them.

Three years on, the picture is more complicated. Service has slipped to early 2028. The legacy geostationary business that is supposed to bankroll the transition posted a 27% revenue decline in 2025 and is facing the expiration of major contracts. And in March 2026, Telesat quietly announced it was carving 500 MHz per satellite — 20 percent of total on-board capacity — away from commercial Ka-band and reserving it for Military Ka users: NATO allies, defense ministries, government agencies.

That last decision is the most revealing. It tells you something important about the economics of building a mid-scale LEO constellation in 2026.

The Chip Problem That Changed Everything

The immediate cause of the service slip is a hardware issue, not a funding crisis. The custom application-specific integrated circuits (ASICs) that power the Lightspeed payload are being developed by SatixFy, a company that MDA Space acquired in 2025. CEO Dan Goldberg told investors in March 2026 that the chip remains “one of the key schedule risks” on the program, even as he expressed confidence in MDA’s ability to execute now that it holds greater direct control over the supplier.

The revised plan calls for two Pathfinder satellites to launch in December 2026, followed by a heavy production cadence through 2027 and full global service by the end of Q1 2028. That is roughly a four-quarter slip from the original target. For a program of this scale, a three-month delay is manageable; what it does to the financial calendar is where the real pressure builds.

The Debt Wall That Won’t Wait

Telesat runs two businesses simultaneously: a mature geostationary fleet that generates most of its cash flow today, and a next-generation LEO constellation consuming most of its capital. The problem is that these two businesses are moving in opposite directions at exactly the wrong moment.

The GEO segment reported 2025 revenue of CA$418 million — down 27 percent year-over-year, reflecting the structural decline in DTH satellite television and rural broadband contracts that satellite internet is itself displacing. Management expects a further CA$90–110 million drop in GEO revenue in 2026 as additional contracts expire and one satellite reaches end of life. Telesat’s own guidance anticipates “structural challenges” in GEO becoming more acute through the decade.

Meanwhile, CA$1.7 billion in legacy Telesat Canada debt matures in December 2026. That is the same month the first Lightspeed Pathfinder satellites are supposed to lift off. Refinancing that debt with a shrinking GEO cash flow base and a LEO constellation still years from generating revenue is not a trivial exercise. Management has called it a “top priority,” with lender negotiations underway — but as of the Q1 2026 earnings call, no deal had been announced.

Trading Spectrum for Security

Into this environment arrives the Military Ka-band modification — and it starts to make sense as more than an opportunistic product add-on.

By dedicating 500 MHz of spectrum per satellite to Mil-Ka across the first 156 production satellites, Telesat is effectively creating a sovereign anchor tenant layer baked into the constellation hardware. The additional cost is modest — roughly US$25 million, less than 0.5 percent of total program cost. What it buys, in return, is the ability to pursue government and defense contracts from NATO members and allied nations that require physically reserved, non-commercial capacity. Telesat’s CEO has pointed to “once-in-a-generation increases in defense investments by allied countries globally” as the backdrop.

The commercial Ka-band bandwidth remaining on each satellite is reported at 2 GHz — a reduction from the original design, but still substantial for a high-throughput Ka-band system. Telesat frames the trade as minor. But it is also an acknowledgment that in today’s financing environment, a commitment from a sovereign defense customer carries a different kind of balance-sheet weight than a commercial capacity agreement.

This is not unique to Telesat. SDA contracts have been nudging commercial operators toward dual-use architectures for years. SpaceX took the opposite approach — Starshield is a physically and organizationally separate system, entirely distinct from the commercial Starlink network. Telesat is betting that integration costs less than segregation.

The Structural Gap

The deeper issue that Lightspeed exposes is what happens when a LEO operator sits in the middle of the market: too large to be a niche provider, too small to enjoy the economies of scale and vertical integration that define SpaceX’s cost structure.

Starlink manufactures its own satellites, launches them on its own rockets at marginal cost, and already serves several million subscribers — a revenue base that funds continuous constellation expansion. Amazon’s Kuiper carries the financial backing of a company with hundreds of billions in cloud and retail revenue. Both operators can afford to subsidize terminal hardware and absorb years of negative free cash flow without refinancing risk.

Telesat has none of those levers. It buys launches from SpaceX, outsources satellite manufacturing to MDA, and is transitioning from a legacy GEO business that competitors are actively eroding. The CA$1.1 billion commercial backlog it has assembled — with capacity agreements from Viasat, Orange, Vocus, and others — is meaningful, but it does not close the gap between today and first revenue.

The military pivot is the rational response to that gap. Sovereign customers sign longer contracts, have less price sensitivity, and lend institutional credibility to a program that commercial investors might otherwise price at a discount. The risk, as analysts and legal scholars have noted, is that embedding military payloads on commercial satellites complicates questions of targeting immunity in armed conflict — a concern the International Committee of the Red Cross has raised formally. Enterprise customers flying on the commercial payload share the same orbital bus as a military communications asset, with no physical separation.

Whether that tradeoff is the right one will become clearer over the next 18 months. Telesat has to refinance its legacy debt, get two Pathfinder satellites into orbit, and begin winning more government contracts — all before a commercial customer can buy a single gigabyte of Lightspeed bandwidth. It is a narrow path, but it is a real one.

Sources