Spectrum Valuation ($SATS) - Part 1
Part 1 of a multi-part series on the economics of spectrum and EchoStar’s spectrum portfolio
Let’s get that OnlyBands money!
How do we value spectrum? Voodoo math, picking a number out of our ass, or something else? How about all of the above.
Spectrum is infrastructure, and like any infrastructure, its value depends entirely on how it’s used, who needs it, and when. Strategic bidders value spectrum using DCF models that tie licenses to expected incremental revenue and cost savings. The basic idea is this: how much more capacity does this spectrum unlock (usually measured in Mbps per MHz per POP), and what does that mean in terms of real-world economics, such as more subscribers, higher ARPU, lower capex, or lower churn?
These models require a pile of assumptions: population growth in the license area, projected data usage, pricing power, deployment timelines, and the capex intensity if densification is chosen instead of spectrum purchase. On top of that, strategic premiums are layered in—scarcity value, adjacency to existing licenses, regional synergies, or even pure defense (keeping it out of a competitor’s hands).
Sooooo, in a sense, this is exactly like voodoo math, ha! Imagine doing all the analytical work to value a spectrum, and the CEO comes in and pounds the table and says, “I know we should pay $1.00 per MHz-POP, but we need to pay $1.50 or more because we can’t let those other schmucks have it.” Do you really want to be the Giants letting Saquon go to Philly?
E-L-G-L-E-S EAGLES!
Before you throw away your Finance 101 books in a huff, remember that auctions are not driven by the number of bidders but by the intensity of bidding. This is why in a lot of investment banking auctions, it is rational for the senior banker to identify not a huge number of bidders, but 2 right bidders who really want the asset.
So spectrum valuation is not a static number. It’s a moving target, shaped by the specific economics and priorities of each bidder. That’s why theoretical spectrum value can vary widely, even reasonably, across bidders, and why auction prices can swing depending on who shows up and what’s at stake.
Historical examples prove the point:
In 2015, DISH and others bid aggressively in AWS-3, driving up valuations.
The 2017 incentive auction underwhelmed because big buyers sat it out.
In 2021, Verizon and AT&T went all-in on C-band, desperate to catch up in 5G.
Back in 2017, Verizon paid $3.1B for mmWave licenses that, in hindsight, weren’t worth much. Both Verizon and AT&T went full retards here.
Shedeur Sanders going in the fifth round in the 2025 NFL Draft.
Also, one thing to keep in mind is that there hasn’t been a large-scale auction of spectrum since 2022 (3.45Ghz). Remember my analogy of Echostar’s spectrum portfolio to an oasis in a desolate desert? Are weary travelers thirsty now? How far is the next oasis in this desolate desert? Can you make it there before running out of water? How much is the water at the next oasis? What if your intel is wrong? What’s your cost if you’re wrong?
With that backdrop, here are five reasons spectrum value could rise and four reasons it might fall.
Here’s also a CTIA presentation on spectrum shortage in the U.S. if you want heartburn.
Bull Case: Why Spectrum Value Could Rise
1. AI and edge computing will drive low-latency spectrum demand.
As AI moves to the edge, powering autonomous vehicles, robotics, drones, and real-time agents, demand for mid-band spectrum grows. Fiber and cloud do the heavy lifting, but edge-native applications need guaranteed quality of service that only licensed spectrum can offer. The more intelligence and computation we push to the edge, the more valuable licensed spectrum becomes as a control asset.
2. Governments now see spectrum as a sovereign asset.
Spectrum isn’t just infrastructure, it’s strategic. It enables innovation, economic growth, and national security. U.S. policymakers increasingly treat it as a sovereign lever. That’s why the FCC is pushing hard on underused holdings like EchoStar’s: spectrum that sits idle is spectrum wasted, and it slows the digital economy’s progress.
3. Enterprise networks and spectrum leasing are breaking out.
Companies want private networks that are fast, secure, and built for their specific needs. To do that, they’re leasing spectrum, either directly from carriers or, in some cases, regulators. This opens up a new monetization path for spectrum holders. As enterprise demand grows, so does the value of even modest spectrum positions that can be leased or shared.
4. Spectrum supply is tight, and new clearing is slow.
Mid-band spectrum is the sweet spot for 5G, but it’s mostly locked up by federal agencies. The 3.1–3.45 GHz band has been delayed, and what does eventually get released will likely come with strings - shared access, limited rights, or geographic carve-outs. Clean, exclusive mid-band licenses are increasingly rare, and that scarcity drives value.
5. New use cases or new users of spectrum: D2D, cable offerings, fixed wireless
Hybrid models are emerging: SpaceX/T-Mobile, GlobalStar/Apple, and others are proving that terrestrial licenses aren’t just for ground-based networks. MSS spectrum could support direct-to-device use. That opens the door to new monetization - leasing, joint ventures, even resale value to satellite partners.
Cable companies will need their own network/spectrum in the near future if they want to control their own destiny in their growing wireless businesses. If they don’t start accumulating spectrum, they don’t have a wireless business; they have a marketing outfit! This introduces potentially new bidders into spectrum auctions. Can they really wait for another three years before deciding it’s time to enter the spectrum market?
Fixed wireless is a growing business for the major carriers. This offering is data-intensive, relying heavily on mid-band spectrum to deliver broadband to households. As adoption accelerates, it places sustained pressure on network capacity, increasing the strategic value of spectrum holdings, especially in suburban areas where fiber buildout is limited.
Bear Case: Why Spectrum Value Could Fall
1. Carriers are solving capacity with densification, not necessarily spectrum.
In dense markets, the spectrum problem is being engineered away with small cells, mmWave, and fiber densification. Meanwhile, Wi-Fi is offloading enterprise and indoor traffic. If carriers can keep expanding without new spectrum, their appetite for expensive licenses will shrink.
However, spectrum remains foundational. Even in dense markets, more spectrum means more flexibility and higher peak speeds, especially as new applications (AR/VR, real-time AI) emerge. Also, densification can lead to increased interference, which additional spectrum helps mitigate.
2. Regulators are leaning toward shared and unlicensed use.
CBRS showed that you don’t always need a full license to get good performance. The FCC’s push into shared models makes it easier for companies to build networks without buying licensed spectrum. That chips away at the economic scarcity that underpinned high auction prices.
However, critical applications require predictable, interference-free performance that only licensed spectrum can guarantee. Operators are more willing to invest in infrastructure when they have exclusive rights, knowing their investment is protected from interference and congestion. And finally, CBRS performance is often lower than fully licensed bands, and shared models can be complex to manage at scale.
3. Spectrum is a capital sink in a high-rate world.
Spectrum generates zero cash until it’s monetized. In a high interest rate environment, warehousing it becomes expensive. Financing is tougher, and spectrum-backed debt isn’t as attractive. If capital remains tight, holding spectrum turns into a liability.
4. New tech could undermine the whole premise.
Beamforming, mesh networks and AI tools are squeezing more out of existing spectrum. That would force a rethink of spectrum’s long-term value - especially for assets with limited use cases or poor geographic fit. A consideration for these technologies is that in data-heavy, capacity-constrained networks, neither fully substitutes for more spectrum. They're complementary tools in the optimization stack, especially when acquiring more spectrum is prohibitively expensive or politically constrained.
Conclusion
LET’S GO Spectrum! Fly, Spectrum, Fly!