In my previous two notes, I made two claims that should be read together. First, I argued that the present registry model becomes unstable once registry power detaches from liability. Second, I argued that Internet number resources are not political property, but operator-held assets embedded in real networks. Put together, those two claims lead directly to the question addressed here: if the registry layer now sits above scarce, operator-held assets while carrying neither equivalent downside nor equivalent exposure, what is the real cost of that arrangement?

That cost is usually misunderstood because people look in the wrong place. They look at fees, transfer paperwork, meeting budgets, process overhead, and membership formalities. Those are visible costs, but they are not the real economic issue. The real cost appears in the conditions under which the asset exists. A structure built to coordinate low-stakes records now sits above scarce, transferable, financeable resources whose significance has grown far beyond the scale, liability assumptions, and institutional design of the bodies that administer them. What still presents itself as administration now functions, in practice, as a conditioning layer above assets.

The right place to begin is not with IPv4 valuation, but one layer deeper: with uniqueness itself. The true value of the Internet does not lie in number resources as such. It lies in the global uniqueness that makes interoperability possible. That uniqueness is what allows billions of endpoints, networks, services, and contracts to function as one Internet rather than many fragmented systems. If uniqueness collapses, interoperability collapses with it. If interoperability collapses, the economic and social value of the Internet collapses with it.

That point is decisive because uniqueness, by its nature, does not justify thick governance. It requires coordination, but only of a very narrow kind. The more the problem is reduced to global uniqueness, the less reason there is for broad discretion, representational mythology, sovereign-style posturing, or institutional self-importance. The more governance is inserted into that layer, the more the system is no longer protecting uniqueness but using uniqueness as a pretext for discretionary control.

A useful analogy is the Arabic numerals inside an IP address. They are used globally not because any institution successfully imposed them by force, but because they are simple, neutral, legible, and interoperable across civilizations. Their strength lies precisely in how little permission they require. No one needs a priesthood of numerals, a regional licensing authority for the number 7, or a specialized community speaking an esoteric language to validate whether the number 3 may continue to exist. The more the Internet’s uniqueness layer is pushed toward that sort of thin, neutral, universally legible coordination, the more stable it becomes. The more it is turned into a site of discretionary control, institutional gatekeeping, and quasi-political entitlement, the more unstable it becomes.

This is also why the present trajectory is conceptually wrong. No institution on earth is powerful enough to guarantee global uniqueness in the manner of a sovereign state. Not even the most powerful government can reliably impose one authoritative order across rival states, adversarial jurisdictions, and conflicting civilizations. Countries can disagree with one another while still depending on the same Internet. That means the correct end state for the uniqueness layer cannot be a thicker, more insulated, more government-like institution. That is a mathematical impossibility masquerading as institutional design.

The resilient direction is therefore the opposite. The registry layer should be as minimal as possible. What should remain is only the minimum coordination necessary for global uniqueness and interoperability. The registry layer should be simple enough that its continued operation appears as common sense rather than institutional mystique. The thinner and more legible the layer, the more stable it becomes. Anything beyond that, especially coercive, abusive, politically contested, or sovereign-style functions, introduces discretion, and discretion in this context is simply another word for arbitrariness.

Abuse management, discretionary enforcement, and contested judgments do not belong inside the uniqueness layer. They belong either in contractual arrangements among affected parties or in external sovereign structures that actually possess coercive power and corresponding accountability. If force is involved, it should sit with institutions designed to bear the consequences of force, not with registries that present themselves as coordinators while disclaiming meaningful downside. In that sense, the long-run endpoint is not conceptually mysterious. It is mathematical. Mathematical end states tend toward simplicity, elegance, and neutrality, not toward thick, messy, politicized institutional layers. The more a uniqueness system approaches minimalism, the more stable it becomes; the more it accumulates thickness, discretion, and political ambition, the more fragile it becomes. In today’s world, that end state would likely look far closer to a thin, transparent, distributed-ledger-style common record, or something functionally equivalent, than to the present model of large, expensive, discretion-heavy registry institutions.

Only after that is clear does the IPv4 question become legible in its proper form. IPv4 can no longer be understood as merely an identifier. It is scarce. It is leased. It is transferred. It supports revenue, infrastructure, services, customer relationships, financing, and continuity. Once scarcity became real and transferability became normal, the relevant question ceased to be whether the registry layer still performs an administrative function. It does. The real question became whether that administrative layer now conditions the economic reality of an asset class large enough to matter. It does that too.

A functioning asset system requires more than operational usefulness. It requires certainty. Certainty is what allows liquidity. Certainty is what allows capitalization. Certainty is what allows an asset to be financed, valued, transferred, collateralized, and integrated into broader business structures. If certainty is weakened, the asset is not formally confiscated, but part of its economic potential is suppressed. That suppression is a real cost, even when no one calls it one.

This is why the scale of IPv4 matters, and why the US$30—60 trillion range should not be dismissed as rhetoric. The number does not come from simply multiplying today’s thin transfer market by the global IPv4 pool. It comes from treating IPv4 the way capital treats a scarce, revenue-enabling base-layer resource. The logic is straightforward. There are fewer than roughly 3.1 billion usable IPv4 addresses. A single IPv4 often enables a service instance, customer endpoint, hosted workload, or network function capable of generating recurring revenue. In my earlier valuation note, I used a benchmark of roughly US$300 per month in cloud-service revenue enabled per IPv4, while current lease pricing is closer to roughly US$0.30 per month. That means the visible lease market reflects only around 0.1% of the enabled value. Once one stops pricing IPv4 as a low-value administrative artifact and instead prices it as a scarce access right embedded in recurring business activity, even a conservative implied value of US$10,000—20,000 per usable IPv4 address yields an aggregate asset base of roughly US$30—60 trillion. That is fixed scarcity multiplied by capitalized economic function.

The point here is not that every address should instantly trade at those levels today. The point is that the current visible market price is already institutionally discounted. IPv4 already trades as an asset, but under a framework that suppresses normal capitalization. Ownership is not fully recognized in property-like form, transfers are constrained by institutional friction, and needs-based thinking continues to distort price discovery long after scarcity has become the real governing fact. The visible market therefore, does not refute the higher valuation. It reveals the existence of a discount.

Once the asset layer reaches that scale, the registry layer can no longer be treated as merely clerical in economic effect. It becomes a layer sitting above a massive capital base while still speaking the language of low-value coordination. It does not need to seize the asset directly in order to impose cost on it. It only needs to preserve enough uncertainty in the recognition layer to prevent the resource beneath it from being treated with the certainty that serious capital requires. That is exactly what the current structure does. It keeps IPv4 operationally indispensable, but institutionally discounted.

This is the first extraction. The registry layer suppresses the full capitalization of the asset. It does so through a combination of non-asset rhetoric, conditional recognition, transfer friction, and the continued insistence that a scarce, transferable, revenue-enabling resource should still be treated as though it were merely an administrative allocation. The asset is not taken. It is discounted. Its upside is capped by institutional uncertainty. Liquidity is weakened. Capitalization is reduced. Certainty is held below the level the asset’s real economic role would otherwise justify.

This is not abstract. If you are a small or medium-sized operator holding just 1,000 IPv4 addresses, and each address carries even a conservative asset value of US$10,000, then the registry layer is sitting above at least US$10 million of your capital base. At US$20,000 per address, the figure is US$20 million. That is not symbolic. It is the difference between remaining a thin-margin operator and having a real base of scarce assets that can be recognized, financed, and capitalized. When the system suppresses that recognition, it is not merely imposing process on you. It is effectively taking millions of dollars of potential value away from the very operators who created real business value around those resources.

The second extraction is even more severe. The same operator whose asset is prevented from being fully capitalized remains fully exposed to registry risk. The operator depends on continued recognition, transferability, and administrative standing within the same registry layer. So the operator is trapped from both sides. On one side, the full upside of scarcity is denied because the asset is kept institutionally discounted. On the other side, the full downside of institutional dependence remains on the operator’s books because the registry still sits at the chokepoint of recognition. Value is capped from above, while risk is loaded from below.

That is why this is not ordinary asymmetry. It is double extraction. The registry layer and the insider structure around it do not merely sit above the operator. They simultaneously suppress the full recognition of operator-held scarcity and preserve operator exposure to a separate administrative power center. The result is that the operators who create the value do not receive the full asset upside, yet still bear the operational and commercial downside if the registry layer becomes unstable, politicized, captured, or coercive.

This point matters most for operators, especially small and medium-sized operators. For many of them, Internet number resources are the first form of true scarcity they have ever been able to hold rather than merely consume. For decades, many operators survived by selling low-margin connectivity while the deeper capital layers of the Internet sat elsewhere. IPv4 scarcity should have changed that. It should have given operators, perhaps for the first time, direct exposure to a scarce base-layer resource with genuine asset characteristics. But instead of allowing that scarcity to be fully capitalized by the operators who built value around it, the registry layer and its surrounding insider structure continue to suppress recognition, constrain liquidity, and retain decisive influence over the conditions of use.

It is important to understand what follows from that over time. I am certainly not the first operator or the first affected party to be harmed by this structure. A system built on repeated asymmetry, repeated procedural insulation, and repeated externalization of downside does not suddenly produce one isolated conflict by accident. It produces many, over years and then over decades. Most of those harmed are weaker, smaller, poorer, more isolated, or more easily exhausted. They are buried one by one in process, cost, delay, and language. Their losses remain private, and their silence is then misread as proof that the system is legitimate. In that sense, I am certainly not the first victim of this structure. At most, I am one of the first survivors: one of the first who had the resources, the incentive, and the refusal necessary to remain standing long enough to make the pattern visible.

Operators should read that in practical rather than rhetorical terms. Each operator should ask a simple question. If the registry layer came for your resources, directly or indirectly, and the result was the effective destruction of your business, do you actually have the resources to resist? Do you have the capital to absorb years of disruption, litigation, delay, uncertainty, and reputational pressure? Do you have the time, legal capacity, institutional knowledge, and endurance to fight for as long as necessary? Most do not. That is the point. A system like this does not survive because it is just. It survives because most of the people it harms are not in a position to fight for long enough to expose what it is doing.

This is why the problem is not only one of economics, but one of operator survival. If you are a small or medium-sized operator, you should not read this as an abstract theory about governance. You should read it as a direct question about whether you can survive becoming the next example. The structure does not need to attack everyone. It only needs to make clear, again and again, that if it comes for you, the burden will fall overwhelmingly on your side. That is enough to discipline the market through fear, silence, and anticipatory compliance. In that sense, even operators who have not yet been targeted are already paying the cost.

The strongest defense of the present system is procedural. It says that accountability in the RIR system is provided by governance rather than by liability: the process is open, members elect boards, and policies are developed by the community. That is the strongest version of the present model’s self-description. It is also inadequate. Participation is not authorization. An open process does not mean those who show up represent all affected parties. It means a subset participates. A door being open does not mean the people inside the room somehow represent everyone whose businesses, assets, and continuity depend on the outcome.

Nor does calling the registry “of the members” dissolve the problem. The registry is a legal entity. It signs contracts. It enforces status. It can suspend, terminate, and revoke resources under those contracts. Those are not abstract community functions. Those are actions taken by a distinct entity with operational authority. Members, operators, and affected parties are not the same set. Not all resource holders are members. Not all affected parties participate. And not all participants have equal influence over outcomes. Collapsing all of these into “the community” does not resolve the problem. It obscures it. The same public discussion has also shown that when proposals challenge the underlying allocation of power, they are frequently redirected into narrower channels or declared out of scope. At that point, “open process” functions less as a mechanism of correction than as a boundary around what kinds of questions the system is willing to consider.

This matters because the registry layer no longer sits above a trivial technical function. It now sits above a large share of the economic substrate of the Internet itself. UNCTAD reports that the digital economy is expanding at roughly 10—12% annually, is expected to generate more than two-thirds of new value creation over the next decade, and is already central to global growth. The coordination premise on which the registry layer sits therefore, underpins a multi-trillion-dollar digital economy. The real cost of registry fragility is no longer limited to what it extracts from operators. In its present form, it places at structural risk the uniqueness layer on which a very large share of global digital value depends.

The imbalance becomes even clearer when one compares the scale of value below the system to the scale of responsibility above it. At the narrowest visible level, the gap is already extreme. A resource that can support hundreds of dollars per month in enabled business value is still institutionally priced as though it were worth only a tiny fraction of that. Yet when things go wrong, the same registry layer or its defenders still fall back on symbolic liability logic, including caps as low as US$100 or the greater of the previous six months of fees. That means the upside suppressed beneath the system is measured in thousands of times more value than the downside the system is willing to bear. Once one includes the broader business exposure attached to continued recognition of the resource, such as contracts, customers, financing, infrastructure, and continuity, the economic damage caused by registry-layer uncertainty rises by orders of magnitude beyond that again.

This is why the issue is larger than unfairness. It is inefficiency. The value extracted by the registry layer and its surrounding institutional class is almost certainly far smaller than the wealth destroyed by the structure itself. The system does not merely redistribute value upward. It destroys value on the way. By suppressing liquidity, weakening capitalization, and forcing operators to bear registry risk without granting them full asset recognition, it imposes a deadweight loss on a base layer whose scale is already measured in the tens of trillions. The social and economic damage is therefore larger than the rents captured.

There is a reason this pattern feels familiar. It resembles the logic of many twentieth-century communist states, where property was said to belong collectively to “the people,” yet practical control remained concentrated in the administrative machinery of the state. On paper, ownership was universal. In reality, effective power sat with the institution that controlled recognition, allocation, and use. Those who actually produced value bore scarcity, inefficiency, and the consequences of bad decisions, while the administrative class controlling the system remained comparatively insulated. The language of collective governance did not eliminate asymmetry. It concealed it.

The registry layer now risks the same pattern. Operators create the value and bear the downside, while an administrative structure with limited downside retains decisive influence over the conditions under which that value remains usable. The economic result is predictable. Those who build and produce remain poorer than they should be, while those who sit above recognition capture influence without bearing equivalent exposure. But the deeper problem is not just unfairness. It is inefficiency on a very large scale.

This is also why NRS has recently built a case archive for documented incidents involving affected operators, disputed resources, procedural abuse, economic losses, and unresolved outcomes. Its purpose is not rhetorical. It is evidentiary. A structure like this survives by fragmenting harm, isolating victims, and making each case appear exceptional. A documented public archive does the opposite. It turns scattered incidents into a visible pattern and allows operators, legal teams, journalists, and policymakers to see that these are not isolated anomalies, but repeated manifestations of the same structural imbalance. In that sense, the archive does not merely record abuse. It reduces the darkness on which the system depends.

If the answer to “would you survive it?” is no, then the issue is no longer theoretical. It is already operational. I have written a note about why operator-first continuity structures matter in practice: prolonged registry uncertainty is survivable only with capital, endurance, legal capacity, institutional understanding, and contingency planning. Most operators should not have to learn those lessons through direct exposure.

That is the misunderstood cost of registry double extraction. It is not what the registry charges in fees. It is the gap between what it controls and what it bears, applied to a now-enormous capital base. That gap was once small enough to ignore. It no longer is. Once that becomes visible, the question is no longer whether the registry layer once had historical utility. It clearly did. The question is whether its present form remains economically coherent in a world where number resources are no longer merely recorded, but capitalized.

If it does not, then the future will not be decided by sentiment, tradition, or institutional self-description. It will be decided by operators seeking structures that better align control with consequence, recognition with responsibility, and scarcity with those who actually bear the downside of reality. Before that future arrives, however, each operator should ask the most concrete question of all: if you were next, would you actually survive it? If the answer is no, then the problem is no longer theoretical. It is already present.