Amazon (NASDAQ:AMZN) has never been shy about spending big when it sees a control point forming.
The $10.8 billion Globalstar deal may be the clearest example yet—not because of satellites, but because of the scarce layer it secures: licensed spectrum. On the surface, this looks like another entry in the Bezos-versus-Musk space saga. That is the easy version.
The more revealing version is that Amazon may have paid $10.8 billion for something much scarcer than satellites: licensed spectrum and a cleaner path to own another layer of digital infrastructure.
The company’s latest earnings call makes that reading harder to ignore. Management grouped low earth orbit satellites with AI and chips as major investment areas, said commercial service for Leo is expected in 2026, and described a network that can connect directly to AWS rather than the public internet. That changes the frame. This is not just about launching objects into orbit. It is about owning the pathways that make those objects commercially useful.
Amazon already has cloud, chips, logistics, devices, and enterprise relationships. Add spectrum to that stack, and the Globalstar deal starts to look less like a side project and more like a quiet move for control.
The Scarcity Story Is Spectrum, Not Satellites
Most investors will first see metal in orbit. That is natural. Satellites are visible, dramatic, and easy to compare with Starlink.
But the harder asset to replicate is spectrum. Satellites can be built, financed, and launched over time. Spectrum is different. It is regulated, limited, and politically sensitive.
Once direct-to-device connectivity becomes a real market, the value of licensed spectrum rises fast because it becomes the gatekeeper for connectivity at scale. That is why the market may be underreading the Globalstar transaction. Amazon is not merely buying transmission hardware. It is securing rights that make a future satellite network far more valuable.
The article you shared already pointed in this direction. Globalstar’s spectrum rights became more valuable as Apple and SpaceX pushed deeper into satellite-to-phone services.
That is the key clue. The race is not only about who has more satellites. It is also about who controls the licensed routes into billions of devices. Amazon’s purchase gives it infrastructure, existing relationships, and globally coordinated wireless assets in one stroke. That bundle matters because scarcity tends to collect pricing power over time.
In that sense, the Globalstar deal looks less like a catch-up move and more like a preemptive land grab in a market that is still being defined.
Amazon Leo Turns This From Theory Into A Real Platform
A deal like this would be easier to dismiss if Amazon were still at the concept stage. The latest earnings call says otherwise. Management said Amazon Leo has already launched 180 satellites, with more than 20 launches planned in 2026 and more than 30 in 2027.
It also expects to launch commercially in 2026. That is a very different setup from a speculative idea. It suggests the satellite layer is already being assembled. The details matter here. Amazon described Leo Ultra as an enterprise-grade terminal capable of 1 Gbps download speeds and strong upload capacity.
More importantly, management said Leo will offer secure private networking that bypasses the public internet and connects directly to AWS.
That one line opens the story up. Amazon is not building a consumer novelty. It is extending cloud infrastructure into underserved regions. Amazon also said dozens of commercial agreements are already signed, including AT&T, JetBlue, and others. That makes the Globalstar deal look like acceleration, not entry.
This Fits Amazon’s Much Bigger Control-Stack Strategy
The earnings call did more than update Leo. It placed satellites alongside AI and chips as core investment pillars. Management said Amazon expects about $200 billion in capital expenditures, mostly tied to AWS. That is a staggering number.
This matters because it puts the Globalstar purchase inside a broader pattern. Amazon is not just spending on separate opportunities. It is building a vertically integrated infrastructure stack.
The company already has AWS at scale, custom chips like Trainium and Graviton, and AI infrastructure through Bedrock. Add Leo and Globalstar, and the stack extends from silicon to signal. That vertical logic is easy to miss because each business is often analyzed separately. Investors talk about AWS one day, retail the next, and satellites after that.
Amazon appears to be thinking more holistically. A company that owns compute, connectivity, and customer endpoints can build tighter economics over time. The Globalstar deal strengthens that thesis because spectrum is one of the few inputs Amazon cannot manufacture.
In a year of massive capital spending, this acquisition looks like a targeted move for a bottleneck asset.
The Market Still Has Reasons To Stay Careful
None of this means the story is risk-free. SpaceX still has a major lead in satellite scale. Starlink already operates around 10,000 satellites, with hundreds dedicated to cellphone connectivity. Amazon is earlier in deployment.
The direct-to-device market is also still forming. Most users still rely on traditional telecom networks. Satellite connectivity today is often positioned as a backup rather than a primary solution. That means adoption curves and pricing models are still uncertain.
There is also the financial side. Amazon expects about a $1 billion cost increase tied to Leo in the near term. While some costs will be capitalized later, this is still a capital-intensive strategy.
Execution risk, regulatory hurdles, and competitive pressure remain real.
So, the balanced view is simple: the Globalstar deal improves positioning, but it does not eliminate uncertainty.
Final Thoughts
The cleanest way to read this deal is that Amazon did not spend $10.8 billion just to join the satellite race.
It spent that money to strengthen control over a scarce and regulated layer of global connectivity. The earnings call reinforces this view. Amazon is already building out Leo, signing enterprise agreements, and integrating connectivity with AWS. At the same time, investors should keep expectations grounded. The opportunity is large, but so are the execution risks.
From a valuation standpoint, Amazon is not cheap.
The stock currently trades at approximately 3.8x EV/Revenue, 18.7x EV/EBITDA, and 34x P/E. These multiples suggest the market already prices in quality, scale, and future optionality. The open question is whether satellite connectivity becomes another durable growth layer within that framework.
For now, the stock reflects confidence—but not full certainty.
Disclaimer: We do not hold any positions in the above stock(s). Read our full disclaimer here.
