The use of EBITDA for Starlink is also interesting. For something like terrestrial fiber, I can imagine thinking that there’s a lot of depreciation on the books, but that most of the equipment keeps working after the depreciation period or is cheaper to replace than it was to buy, and that the right of ways and attachments don’t really depreciate. But Starlink satellites are actually gone at the end of their useful life.
I have not dug into the filing to see how this really breaks down.
Starlink satellites can likely have their life extended by quite a bit, like past constellation satellites did. But it doesn’t really matter as SpaceX is still upgrading capability like crazy. Starship will mean a factor of 10-100x Starlink capacity expansion.
I mean, the SpaceX bet is that what you mention for terrestrial fiber
> or is cheaper to replace than it was to buy
will also hold true for cost of mass to orbit. There's a lot riding on making that prediction come true for SpaceX, hence all the CapEx going into Starship.
I have not dug into the filing to see how this really breaks down.