Summary
- Nearly US$2bn of satellite cost crossed out of "Satellites under construction" and into the caption that carries depreciation in Viasat's six months to 30 September 2026 — my subtraction on two printed captions of the same note.
- Depreciation in the cash-flow statement for those six months fell to US$482,224 thousand from US$510,371 thousand a year earlier, while that transfer happened.
- The Form 10-Q gives ViaSat-3 F3 "entered service" and gives ViaSat-3 F2 only "completed in-orbit testing and was integrated into the Company's existing satellite fleet", one day after the company's own filed letter said both had entered service.
Four captions, one direction of travel
The Form 10-Q for the period ended 30 September 2026, filed on 6 November 2026, prints property, equipment and satellites, net, at US$7,178,352 thousand against US$7,272,637 thousand at 31 March 2026, on a gross figure of US$12,439,748 thousand. It reports in thousands throughout. The line to watch is the one whose label carries the accounting assumption: "Satellites (estimated useful life of 7-17 years)", at US$5,371,396 thousand against US$3,410,741 thousand. Beside it, "Satellites under construction" falls to US$642,625 thousand from US$2,644,050 thousand, and "Construction in progress" falls to US$502,985 thousand from US$662,841 thousand.
Three movements point the same way. On the printed captions, and by my subtractions alone, about US$1,960,655 thousand arrived in the satellites caption, about US$2,001,425 thousand left satellites under construction and US$159,856 thousand left construction in progress. Net property fell over the same half-year, the ordinary result of depreciation and disposals running against a book that is still being built. What matters is the direction: cost leaves the waiting room and enters the room where it is depreciated.
The trigger is a condition, not a date
The policy sentence is short and it is doing all the work. "At the time satellites are placed in commercial service," Viasat states in the basis of presentation, "the Company estimates the useful life of its satellites for depreciation purposes based upon an analysis of each satellite's performance against the original manufacturer's orbital design life, estimated fuel levels and related consumption rates, as well as historical satellite operating trends." Depreciation runs straight-line over estimated useful lives of two to 38 years.
Read the sentence twice, because everything downstream hangs off it. The trigger is a condition of the spacecraft — placed in commercial service — not a launch, not a reflector deployment, not a filing date, not a press release. The policy says nothing about which spacecraft, on which day. It is a rule about a class, applied to a fleet of 24 in-service or operational satellites: 14 Ka-band; eight high-availability L-band, of which three contingency satellites are operational but not in commercial service; one S-band for the European Aviation Network; one Inmarsat-6 hybrid.
That the filing's own fleet count separates "operational" from "in commercial service" tells you the distinction is not decorative.
One spacecraft enters service; the other joins the fleet
The filing's own paragraph reads: "In September 2026, the second ViaSat-3 class satellite, ViaSat-3 F2, completed in-orbit testing and was integrated into the Company's existing satellite fleet covering the Americas. In August 2026, the third ViaSat-3 class satellite, ViaSat-3 F3, entered service covering the Asia-Pacific region." Two spacecraft, two formulations, one sentence pair.
On 31 August 2026 the company announced that ViaSat-3 F3 had entered service, "with capacity now available for customers operating across fast-growing markets in the Asia-Pacific region", and called the milestone "the beginning of revenue-generating operations". On 17 September 2026 it announced that ViaSat-3 F2 had entered service across the Americas, adding that "completion of the ViaSat-3 constellation has tripled the bandwidth available across Viasat's global fleet" — the company's own claim, and it belongs to the company. On 5 November 2026, in the shareholder letter filed as Exhibit 99.2, it wrote: "Both ViaSat-3 (VS-3) Flight 2 and Flight 3 entered service and are performing as expected." The same letter's achievement bullet reads: "Successfully commenced service on both the VS-3 F2 and VS-3 F3 satellites in the Americas and Asia-Pacific (APAC) regions, respectively." That letter went to the SEC on 5 November 2026.
One day later, the Form 10-Q put the same spacecraft as having "completed in-orbit testing and was integrated into the Company's existing satellite fleet". Two documents of one registrant, a day apart, two vocabularies. They are not in contradiction — the filing does not deny that F2 entered service — and neither phrase is the accounting trigger. The trigger is the policy's "placed in commercial service", and no document read for this piece says which day that was for F2.
What the waiting costs
While a satellite waits, the money is capitalised. The same note lists what goes into a satellite's cost: construction and launch, launch insurance and insurance during in-orbit testing, the net present value of expected manufacturer performance incentives, monitoring and support costs, and interest during construction. That last item carries a clock. Viasat capitalised US$43.1 million of interest in the three months and US$105.5 million in the six months to 30 September 2026. A spacecraft sitting through in-orbit testing holds a financing cost that has not yet become a period expense.
The note also shows how much fleet value is written down before service begins. In fiscal 2024 the company took an approximately US$1.67 billion reduction to satellite carrying value on ViaSat-3 F1 and Inmarsat-6 F2, partly offset by approximately US$770.0 million of insurance claim receivables, with a final approximately US$10.0 million received in the first quarter of fiscal 2026. The depreciable base that eventually reaches the income statement is the base left after that reduction.
Seven further GEO satellites were under construction at 30 September 2026: GX7, GX8, GX9 and four Inmarsat-8 L-band safety service satellites. None of them carries a printed cost in the material read, and none carries a printed start date. That is the shape of the next transfer into the depreciating caption: real, scheduled and invisible in dollars until it happens.
The US$28 million that did not arrive
Now the line that makes this worth writing. Six-month depreciation in the cash-flow statement was US$482,224 thousand against US$510,371 thousand, a decline of US$28,147 thousand, about 5.5 per cent — my subtraction and my division on two printed lines. The caption that carries depreciating satellites rose by nearly US$2bn across the same six months, and the depreciation charge fell.
The two facts do not conflict. Depreciation is not a levy on new capital; it is a charge spread over the life of assets already in service, and a constellation's charge steps up as each spacecraft crosses the trigger. But a reader who assumes a US$2bn transfer into the serving caption must produce a visible step-up in the period expense has the mechanism backwards at this half-year mark. What the transfer produced, as printed, was balance-sheet geography.
The filing offers a second measure, and it is not the same measure. The adjusted EBITDA reconciliation prints combined depreciation and amortisation of US$335,379 thousand for the three months and US$650,620 thousand for the six, alongside amortisation of acquired intangible assets of US$65,007 thousand and US$130,025 thousand. The statements of operations print no depreciation-only line at all; depreciation sits inside cost of service revenues, cost of product revenues and SG&A.
My subtraction of those two reconciliation lines for the three months gives 270,372 thousand, and I label it as mine: the reconciliation does not state that the intangible line is the only amortisation inside the total, so that number cannot pass as a printed depreciation figure.
The card and the company
The Viasat entry in BTW's directory is a thin identity record, tracked from a handful of public references with no confirmed operator, and six earlier articles sit against it. It is not a financial document and nothing here treats it as one. The figures come from Viasat, Inc., the registrant of the 10-Q and the issuer of the announcements; the card records that BTW is watching the name, the filing records what the name is carrying.
The uncomputable part is the part worth carrying away. No per-spacecraft capitalised cost appears anywhere in the material read. No per-spacecraft depreciation commencement date appears. The transfer into "Satellites" cannot be split between F2, F3 and whatever else moved. Seven under-construction satellites will produce another transfer, unsized. And a six-month decline in depreciation tells a reader that the step-up had not arrived in the printed expense by 30 September 2026 — not that it never will.
Member Briefing
Read the full trend briefing
Members read the full briefing with the source notes behind every finding.
Strategic Circle
A daily briefing in your language, built from changes we have verified.
Join for USD 20/monthCancel anytime.
Members also get the full research library, a daily briefing on the networks they follow and the private forum.






