What spectrum is, why orbit makes it rivalrous, what the transaction record now proves about its value, and the framework to treat it as the asset it has become.
1.1 A company worth less than its licences
In the early summer of 2025, EchoStar Corporation appeared to be approaching the end. The company — heir to the DISH satellite-television business and to decades of accumulated American spectrum licences — was skipping interest payments and openly weighing bankruptcy protection, its share price reflecting a market consensus that the operating businesses could not carry the debt stacked upon them. The regulatory environment was actively hostile: the chairman of the Federal Communications Commission had publicly questioned whether EchoStar was genuinely using key licence holdings or merely warehousing them, and had set inquiries in motion that threatened the licences themselves.
What happened next is the founding parable of this article. Within eighteen months, EchoStar had sold spectrum rights to AT&T and to SpaceX in transactions totalling roughly forty-two billion dollars, retired its most pressing debt, and delivered the best shareholder return in the Fortune 500 — not by rescuing its businesses, but by conceding that the businesses had always been worth less than the rights beneath them. The satellites, the subscribers, the brands: all of it had functioned, in the end, as an elaborate holding structure for frequencies. The buyers were not purchasing cash flows. SpaceX, in particular, was buying the raw material for its next product — direct-to-device satellite service to ordinary mobile phones — and the seller’s distress mattered less to the price than the simple fact that suitable frequencies, held with the right priorities in the right market, exist in quantities that cannot be increased.
Three observations from this episode organise everything that follows. First, spectrum rights can be — and increasingly are — the dominant asset on a space or telecommunications balance sheet, capable of outvaluing the enterprises built upon them by multiples. Second, these rights are strange assets: they are not property in the ordinary sense, they exist only as entries in national and international registers, they can be diminished or extinguished by regulators without compensation, and their value is inseparable from queues, clocks and negotiations largely invisible to the financial world. Third, the people who must now make decisions about such assets — the operator choosing bands for a constellation, the investor pricing a licence portfolio, the administration deciding what to do with a national holding — mostly learned their professions in fields where none of this exists. This series of articles is written for them.
1.2 What spectrum actually is
Begin with the physics, because every commercial and regulatory peculiarity of spectrum descends from it.
Radio spectrum is the portion of the electromagnetic field usable for communication — conventionally, frequencies from a few kilohertz to a few hundred gigahertz. It is not a substance. Nothing is pumped, mined or consumed when a signal is transmitted; the “resource” is the opportunity to be received intelligibly, and that opportunity is what runs out. Two transmitters using the same frequencies, at the same time, in the same place, towards the same receivers, destroy each other’s signals. This — interference — is the entire foundation of spectrum scarcity, regulation and value. Spectrum is infinitely renewable and instantaneously rivalrous: the moment one use ends, the resource is whole again, but while two uses overlap, both can fail.
Scarcity therefore has four dimensions, and rights over spectrum are always, implicitly, rights defined along all four: frequency (which band), space (which territory or orbital geometry), time (always, or in defined windows), and direction (towards which receivers, from which transmitters — an increasingly important dimension as antennas grow more precise). A “spectrum right” is never simply a band; it is a bundle of permissions across these dimensions, conditioned on not causing harmful interference to others whose bundles rank ahead.
The second physical fact with permanent commercial consequences is that frequencies differ profoundly in character. Lower frequencies propagate far, bend around obstacles, penetrate buildings and rain — and offer little bandwidth, because bandwidth is a slice of frequency and the low bands are narrow. Higher frequencies carry enormous bandwidth but travel in increasingly fragile straight lines, absorbed by atmosphere, foliage and weather. Every band is therefore a distinct product with distinct economics. In satellite practice, the vocabulary of value runs roughly so: L- and S-bands (1–4 GHz), scarce, robust, penetrating — the natural home of mobile-satellite service and the direct-to-device contest, and the most expensive real estate per hertz; C-band (4–8 GHz), rain-resilient workhorse of legacy distribution, now the world’s most fought-over conversion target for terrestrial mobile; X-band (8–12 GHz), heavily governmental and defence-flavoured; Ku- and Ka-bands (12–40 GHz), the broadband constellations’ home, capacious and coordination-congested; and the frontier Q/V bands and above (40 GHz+), abundant, weather-vulnerable, and the destination of the next capacity race. Nothing in this hierarchy is static — technology continually renegotiates what each band can do — but at any moment, the hierarchy prices everything.
The third fact is jurisdictional: radio waves ignore borders. A geostationary satellite’s beam can illuminate a third of the Earth; even a modest low-orbit spacecraft sweeps across every territory beneath its inclination in a day. Interference is thus inherently international, and the governance of spectrum was forced to become international earlier and more completely than almost any other field of economic regulation — the International Telecommunication Union’s lineage runs to 1865, making it older than the aeroplane, the automobile and the electric grid it now sits beside in the family of critical infrastructure regimes.
1.3 Why orbit sharpens everything
Terrestrial spectrum management is hard; orbital spectrum management is harder, for reasons that compound.
Footprints are continental. A terrestrial licence can be cut to a country, a city, a cell. A satellite’s emissions cannot be so confined; its use of a band forecloses, or complicates, uses across vast areas and many jurisdictions at once. The unit of coordination is not the neighbourhood but the hemisphere.
Geometry creates chokepoints. The geostationary arc — the single ring, roughly 35,786 kilometres above the equator, where a satellite hangs stationary over one longitude — is a genuinely finite resource: useful slots over valuable landmasses, in valuable bands, are countable and long since claimed. Non-geostationary constellations escape the arc but inherit a subtler geometry problem: thousands of moving transmitters must share bands with each other and protect the arc above them, turning interference management into a continuous, computational, many-body negotiation. Orbital position and spectrum are not separate resources; they are one joint resource — this article will say “spectrum” throughout, but the asset is always spectrum-and-geometry.
The tragedy is anticipatory. Because rights accrue by queue, the rational strategy for every ambitious actor is to claim early and abundantly — before need, before financing, sometimes before intention. The result is the defining pathology of the current era: a queue containing, at the time of writing, filings for more than a million satellites, perhaps thirty times the number that will plausibly fly, lodged through administrations ranging from the great space powers to microstates whose regulatory sectors consist of a few dozen officials. The queue’s relationship to reality has become the central analytical problem of the field — and, for those who can read it, the central commercial opportunity.
And the stakes have militarised. Orbital spectrum now carries the communications of war and deterrence: proliferated defence constellations, protected tactical waveforms, the resilience programmes of every serious military. States increasingly read the register of filings the way they once read naval lists. Whatever else spectrum is, it is now an instrument of national power, and its regime is subject to the pressures that instruments of power attract.
1.4 The peculiar existence of a spectrum right
Here is the conceptual adjustment that newcomers from finance and engineering find hardest: nobody owns spectrum. Not companies, not individuals, not even states in the proprietary sense. What exists instead is a layered structure of recognitions and permissions, and the asset this article analyses is precisely that structure.
At the international layer, the ITU’s Radio Regulations — a treaty — establish which services may use which bands, and record, in the Master International Frequency Register, the specific frequency assignments of specific systems, filed by national administrations on behalf of operators, ranked by the date of filing. Recording confers not ownership but international recognition: the entitlement to operate free of harmful interference from later-comers, and the obligation to protect those who came earlier. The regime runs on clocks — a seven-year deadline to bring filed frequencies into genuine use, milestone schedules for constellations, suspension windows — and on a first-come-first-served queue whose priority dates are, as we will argue, the closest thing the asset class has to a title system.
At the national layer, each state licenses the use of spectrum on and over its territory: the operation of the satellite system itself (by its licensing state), the earth stations that anchor it, and — critically — the right to serve customers in each market, granted country by country. A satellite system with impeccable international paper and no market access is a beautifully documented inability to earn revenue; the reverse — national licences atop a weak international position — is revenue with a coordination time-bomb beneath it. The asset is always the stack: international recognition, operating authority, and market access, each with its own grantor, conditions and clocks.
Two properties of this structure drive everything. First, conditionality: every layer can be qualified, suspended or withdrawn by its grantor — for non-use, for breach, for policy change — generally without compensation. Spectrum rights are tenure, not title; their security is a function of use, compliance and political standing, which is why we treat defending rights as a discipline equal to acquiring them. Second, negotiated content: what a right actually permits is settled substantially through coordination — bilateral, technical, confidential negotiations with the systems ahead in the queue — so that two filings of identical vintage can confer radically different operational freedom depending on what their holders conceded in rooms no register records. Due diligence that stops at the register systematically misprices the asset.
Box 1.1 — The first spectrum fortune: Tonga, 1990. The discovery that sovereignty itself could be monetised in this regime belongs to the Kingdom of Tonga, which — guided by the satellite entrepreneur Matt Nilson — filed in the late 1980s and early 1990s for a string of geostationary slots it had no intention of using itself, becoming for a time one of the largest claimants of orbital positions on Earth and leasing the rights to commercial operators for millions of dollars. The industry called it scandalous; one aggrieved operator parked a satellite in a Tongan slot in protest. But no rule had been broken: the regime’s founding principle of sovereign equality entitles every administration to file, for anyone. Tonga’s play is the ancestor of today’s flag-of-convenience filings through Rwanda, Papua New Guinea and the Solomon Islands — and the first demonstration that in this asset class, the queue position is the asset.
1.5 What the transaction record now proves
For most of the space age, the claim that spectrum was valuable rested on inference: operators guarded their filings, regulators rationed licences, and everyone behaved as though the rights mattered enormously — but arm’s-length prices were rare, and the asset stayed dark to financial analysis. The years 2020–2026 changed that. Three clusters of transactions created, for the first time, a public comparables record, and each cluster teaches a distinct lesson.
The C-band clearances taught that regulatory migration is a priced event. When the United States determined that the lower C-band should pass from satellite distribution to terrestrial mobile use, the resulting auction raised on the order of eighty billion dollars, and the incumbent satellite operators — chiefly Intelsat and SES — received billions in accelerated relocation payments for clearing on schedule. The lesson generalises: when spectrum moves between uses, the terms of transition — who is compensated, for what, on what timetable — are themselves an asset-pricing mechanism, and incumbency in a band coveted by better-funded industries is not merely a threat but, negotiated well, a monetisable position. The sequel is already running: in July 2026 the FCC adopted rules to auction a further 160 megahertz of the upper C-band, with incumbent clearing by mid-2031 — the template applied again, with the terms of accommodation once more determining who captures the conversion value.
The EchoStar transactions taught that spectrum can be the whole company. Roughly forty-two billion dollars for licence portfolios — from a seller whose operating businesses the market had valued at a fraction of that sum — established the extreme case: enterprise value residing almost entirely in regulatory positions. They also established the modern buyer profile. AT&T’s purchase was classical: a terrestrial carrier consolidating mid-band. SpaceX’s was the signature of the new era: a satellite operator paying tens of billions for terrestrial mobile licences, because direct-to-device service — satellites transmitting in the bands that phones already hear — collapses the old boundary between satellite and terrestrial spectrum into a single market. After EchoStar, no analysis of satellite spectrum value can stop at the satellite bands.
The direct-to-device land rush taught that scarcity premiums attach to suitability, not abundance. The D2D contest — SpaceX building on acquired terrestrial licences, AST SpaceMobile on partner carriers’ frequencies, incumbent MSS holders defending L- and S-band positions, and regulators improvising frameworks case by case — has concentrated extraordinary valuations on the narrow set of frequencies that combine mobile-device compatibility, satellite feasibility and available rights. The market is pricing a specific technical-regulatory conjunction, and repricing it with every regulatory pronouncement: a single supportive or sceptical statement from a commission chairman now moves D2D operators’ equity in a day. Spectrum value, the record shows, is not a slow variable.
A fourth lesson threads through all three clusters and previews: the rules themselves are now the volatile input. In July 2026 the United States replaced the treaty-era power limits governing how low-orbit constellations must protect geostationary satellites — limits the international process had declined to revisit on any near timetable — with performance criteria of its own design. At a stroke, identical international rights became worth different amounts in different jurisdictions, and every valuation of a GSO or NGSO position acquired a regulatory-fragmentation term. This asset class has entered the period in which its legal architecture is openly contested.
1.6 The framework: spectrum as an asset with a lifecycle
The organising idea is that spectrum rights are best understood — by all three of its audiences — through an asset lifecycle with six phases, each governed by identifiable rules, clocks and failure modes:
Creation. Rights begin as filings and applications: band selection, administration choice, the drafting of the technical envelope, the securing of a priority date. Decisions taken here — often by engineers, often early, often casually — fix the asset’s fundamental quality forever.
Tenure. Rights mature and persist only through use and compliance: coordination completed, deadlines met, milestones satisfied, licence conditions kept. Tenure is active; neglect is the primary way value dies in this class.
Encumbrance. Every right carries negotiated and inherited burdens — coordination agreements, protection obligations, national conditions — which define its true operational content. Mapping encumbrances is the heart of diligence.
Valuation. Value is a function of band quality, priority position, geographic and market-access coverage, encumbrance burden, deadline and milestone risk, counterparty durability (not least the sponsoring administration), and regulatory weather.
Transfer. Rights change hands through the M&A of holding entities, leases, sharing and hosting arrangements, and compensated migrations — never through a clean title transfer, always with regulatory consent embedded. Structure follows the regime’s peculiarities.
Loss. Rights lapse, are suppressed, are trimmed by milestone review, are rendered worthless by coordination defeat or regulatory change, or are extinguished for breach. The loss modes are enumerable, and most are foreseeable years out — which is why the professionals in this field think in clocks.
Across all six phases runs the discipline this article will insist upon until it becomes reflex: date everything. Priority dates, bringing-into-use deadlines, milestone gates, licence terms, comment windows, conference cycles. The field’s practitioners earned their reputation for pedantry honestly — in an asset class where a missed date can vaporise a position that took a decade to build, the calendar is the risk register.
1.7 Three vantage points
The lifecycle looks different depending on where one stands, and this article is written to be read from three chairs.
The operator experiences spectrum as a design constraint that arrives too late. The mission is conceived, the constellation sized, the business plan financed — and then the regulatory questions surface, at which point the queue’s realities can force redesign or worse. We argue for inverting this sequence: band and filing strategy as inputs to system design, the priority date pursued as aggressively as any engineering milestone, coordination staffed as a core competence, and the calendar of clocks integrated into programme management. For the operator, the lifecycle framework is a checklist of what must not be discovered late.
The investor experiences spectrum as an asset without an information infrastructure: no exchange, no ticker, no standard diligence, comparables that arrive years apart, and value concentrated in documents — filings, coordination agreements, licence conditions — that data rooms under-produce and generalist advisers under-read. We supply the missing method: how to inventory a target’s stack, read its encumbrances, price its clocks, and weigh the political durability of the administrations and regulators behind it. For the investor, the lifecycle framework is a due-diligence protocol and a valuation model.
The administration experiences spectrum as a portfolio it may not know it holds: legacy filings and positions accumulated across decades, commercial suitors seeking sponsorship, treaty conferences demanding national positions, and — for smaller states especially — a genuine strategic choice between hospitality and credibility whose poles are Tonga and Luxembourg. The administration’s treat the national spectrum office as what it has become: a manager of sovereign assets and an instrument of industrial and security strategy. For the administration, the lifecycle framework is an audit template and a policy map.
The three readerships are not merely accommodated; their interaction is the subject. Operators hold what investors price; administrations grant what operators hold; investors fund what administrations regulate. Most professional failures in this field are failures to understand the chair across the table — the operator who cannot explain its regulatory moat to capital, the fund that cannot read a coordination agreement, the ministry that prices its sovereignty in filing fees.
1.8 The stressed regime
A final framing note: this asset class is entering its period of maximum institutional stress, and strategy that extrapolates the regime’s past stability will misprice its future.
The stresses are four, and they interlock. The queue crisis: a million filed satellites, speculative filing as rational strategy, and enforcement machinery built for a slower, smaller era. The flag-of-convenience system: administrations competing on fees, speed and laxity for the sponsorship business, with supervision migrating — as it did in shipping — away from the flag and towards the markets served. Regulatory fragmentation: the largest national regulator has demonstrated that it will overtake the treaty process when dissatisfied with its pace, and others will note the precedent; the single international rulebook is becoming a base layer beneath diverging national regimes. And the conference cycle: WRC-27 in Shanghai will contest direct-to-device allocations, new mobile-satellite spectrum and the mid-band frontier with terrestrial mobile, under explicit great-power competition, with WRC-31 already visible as the venue where this cycle’s studies become the next cycle’s rules.
None of this counsels despair about the regime; regimes under stress are also regimes in motion, and motion is where strategy earns returns. But it does counsel the disposition this article tries to teach: read the primary record, price the clocks, respect the politics, and treat every rule as a variable with a date on it.