T-Mobile spent five years trying to get rid of the very spectrum Elon just called “the last key piece of complete mobile coverage in America.”

This 800 MHz band is Sprint’s old Nextel spectrum—the stuff that powered push-to-talk phones back in 2004. The Department of Justice (DOJ) forced T-Mobile to divest it as a condition of its $26 billion acquisition of Sprint. Dish initially had the right of first refusal at $3.59 billion, but gave it up because it couldn’t raise the money. Last year, T-Mobile finally sold it to an investment firm called Grain Management for $2.9 billion plus a spectrum swap.

Grain’s original plan was to sell it to utilities for private networks.

Fourteen months later, SpaceX bought the entire portfolio, and the carriers that missed out saw their stock prices fall by as much as 8% in a single afternoon.

This is why it matters so much. Starlink’s satellite signal at 2 GHz can reach your phone outdoors, but it fails as soon as it hits a wall. 800 MHz can penetrate buildings. Most phones already have this band built in. So satellites handle open skies, while this low-band layer handles indoors—and suddenly Starlink Mobile can cover all the areas served by traditional carriers, without a tower network.

Put these moves together: the $17 billion acquisition of EchoStar’s 50 MHz of 2 GHz spectrum. The FCC approved 15,000 Gen2 satellites this week. And now there’s a low-band coverage layer.

Verizon, AT&T, and T-Mobile paid $81 billion for C-band spectrum in a 2021 auction, then spent years and hundreds of billions of dollars deploying it on towers. SpaceX, meanwhile, pieced together a nationwide carrier from a distressed asset sale, a forced-divestiture castoff, and rockets it already owned.

The walkie-talkie spectrum from 2004 ultimately became the missing piece that could let satellites replace cell towers. No one who had dealt with it over the past two decades saw that coming.