The quiet North American scarcity test
In September 2015 the American Registry for Internet Numbers announced that the free IPv4 pool in its region had reached zero. The announcement was bureaucratic in form and historic in substance. ARIN would still process requests; approved applicants could still seek addresses through a waiting list or through the transfer market. But the old North American assumption - that a qualified network could ask the registry for new IPv4 capacity and receive it from a common administrative pool - had ended.
That was the moment when an address block stopped being merely a registry entry. In economic terms it became a scarce input into production. In legal terms it became a contractual and policy-defined claim around a valuable asset. In operational terms it became a constraint on growth, customer onboarding, hosting density, cloud expansion, abuse management, routing hygiene and business continuity. In financial terms it became a balance-sheet question. The registry record still mattered, but the value was no longer created by the registry. It was created by networks, customers, routing acceptance, reputation, deployability and scarcity.
ARIN is a useful case because it is not an obviously broken registry. It has a mature policy process, a large base of sophisticated operators, a visible transfer process, contractual language that acknowledges defined rights, and a North American market deep enough to reveal price signals. Precisely for that reason, it shows the post-exhaustion problem more clearly than a crisis case would. The central question is not whether ARIN is competent. The question is what happens when a coordination body designed to administer number records becomes a gatekeeper around an asset class that private operators now buy, lease, finance and depend on.
The answer is not simple privatization rhetoric. IPv4 addresses remain unusual assets. They are not land, spectrum or equipment. They are unique numerical identifiers whose usefulness depends on a globally coordinated ledger and on the willingness of networks to route them. A block has no physical mass, but it can support a business. It can be priced, transferred, leased, filtered, reputationally impaired, collateralized in practice, contested in law and embedded in customer contracts. Its value comes from uniqueness plus usability. The registry does not create that value by decree; it records the claim that makes non-conflicting use possible.
That distinction matters. A registry that preserves uniqueness, accurate records, reverse DNS, RDAP, WHOIS, RPKI and orderly transfers performs a real coordination function. A registry that treats scarcity as a reason to keep discretionary power over capital is performing a different function. The first is infrastructure bookkeeping. The second is capital allocation. ARIN sits at the boundary between the two. Its waiting list, transfer rules, legacy-resource treatment, RSA structure and needs-based review process together form a post-exhaustion economic regime.
The post-exhaustion regime is best understood through institutional economics. Scarcity changes incentives; incentives change institutions; institutions that control valuable assets without matching liability become structural risks. ARIN's official materials are useful factual exhibits: the 2015 depletion notice, the IPv4 Waiting List page, transfer guidance, the Number Resource Policy Manual, the legacy-resource page and the Registration Services Agreement. They do not by themselves answer the political economy question. They show the machinery through which the question must be asked: can a regional Internet registry remain a neutral ledger when the record it maintains has become strategic capital?
The day administrative abundance ended
IPv4 scarcity was not a surprise. IPv4 uses a 32-bit address space, which gives fewer than 4.3 billion theoretical addresses before reservations, special-use ranges and operational constraints. The Internet grew from a research network into a commercial, mobile, cloud and platform economy. NAT slowed the exhaustion curve; CIDR improved aggregation; IPv6 was standardized as the long-run escape route. None of those developments prevented the market fact that usable IPv4 became scarce while the global economy still required IPv4 reachability.
ARIN's archived 24 September 2015 announcement was careful. It said ARIN had issued the final IPv4 addresses in its free pool and would continue to process and approve requests. It pointed to the waiting list and the transfer market. It also noted that future IPv4 space received from IANA or recovered from revocations or returns would be used to satisfy approved waiting-list requests. This was the administrative sentence that marked an economic break: from then on, new IPv4 supply in the ARIN region would no longer come from ordinary abundance. It would come from recycling, revocation, return, transfer or private market structure.
Scarcity changed the meaning of time. Before exhaustion, waiting was mostly a processing matter. After exhaustion, waiting became an allocation mechanism. A queue is not just a list; it is a way of deciding who bears delay, who can plan growth, who must buy, who must lease and who must redesign a network around scarcity. A quarterly distribution of recovered blocks may be fair in a procedural sense. It is not equivalent to elastic supply. The network that needs addresses for a product launch, a cloud cluster, an acquisition or a customer migration cannot treat an uncertain queue as capital planning.
Scarcity also changed the meaning of old holdings. A legacy /16 that once looked like an over-sized historical allocation became a material corporate asset. A university, manufacturer or early Internet participant that held more address space than its current direct network needs acquired optionality: sell, lease, renumber, hold, use as strategic reserve or use as bargaining power. A hosting company with enough IPv4 to serve customers without immediate market purchases gained an advantage over a newer competitor. A cloud platform with efficient address management could delay exposure to the spot market. What had been historical accident became capital structure.
This is the institutional problem at the heart of ARIN scarcity. The registry layer continues to speak in the language of allocation, assignment, operational need, policies and agreements. The market speaks in the language of price, liquidity, inventory, risk and opportunity cost. Both languages describe real things, but they do not carry the same assumptions. Allocation language assumes a steward deciding who qualifies. Market language assumes capital moving toward its highest-value use. The conflict between those assumptions defines the post-exhaustion IPv4 economy.
ARIN did not cause that conflict. In many respects it managed the transition more openly than weaker institutions would have done. Its transfer market is documented, its waiting-list rules are published, and its contractual terms are public. But transparency does not remove the contradiction. Once scarcity has created market value, administrative controls no longer feel like neutral order. They affect liquidity, pricing, small-operator access, balance-sheet value and the distribution of power between incumbents, entrants, brokers, lessors and the registry itself.
The break is easiest to see by comparing allocation with transfer. In the allocation era, ARIN was deciding whether an applicant could receive space from a common pool. In the transfer era, ARIN is often deciding whether a privately negotiated movement of already-issued space fits policy. The same vocabulary - need, utilization, stewardship, conservation - crosses from one world into the other. But the economic stakes change. The applicant is no longer simply asking for a low-cost administrative resource; the buyer may be committing capital, the seller may be monetizing an asset, and both may be embedding the transaction in a larger network, financing or acquisition plan. A rule that looked like rationing under abundance becomes capital control under scarcity.
That does not mean ARIN should become a passive notary. A registry that ignores forged authority, stale corporate records, hijacked blocks, inaccurate contacts, broken reverse DNS, disputed resources or fraudulent transfers would undermine the market it is supposed to support. The point is narrower. After exhaustion, legitimacy depends on separating ledger protection from industrial policy. The registry must be strong where the ledger would otherwise become false. It should be modest where market actors are deciding how to deploy their own capital.
Waiting list economics
ARIN's IPv4 Waiting List shows what rationing looks like after depletion. A request that meets policy requirements can be placed on the waiting list for an approved block size. The applicant specifies the smallest block it would accept. Available addresses, typically from revocations for non-payment, are used to fill requests on a first-approved basis, subject to the sizes of the blocks ARIN has received back. The page states that organizations holding more than a /20 equivalent of IPv4 space in aggregate, excluding certain special-use space, are not eligible. The maximum aggregate an organization may qualify for at one time is a /22. An organization may have only one request on the list.
Those rules are defensible as fairness rules. They prevent large holders from consuming a small recycle pool. They direct limited returned capacity toward smaller or less-endowed organizations. They preserve an administrative route for entities that cannot or should not immediately enter the transfer market. But economically they also reveal the scale of scarcity. A /22 is 1,024 addresses. In a world of carrier-grade networks, cloud platforms, VPN operators, hosting companies, SaaS platforms and managed-security services, that is not a growth engine. It is a ration.
The waiting list therefore functions less as a replacement for the free pool than as a safety valve. It can help a small network obtain an initial or incremental block. It cannot supply the address needs of a fast-growing operator at scale. It can soften exclusion; it cannot abolish the market. Indeed, ARIN's own waiting-list page treats the transfer market as one of the other available IPv4 options. Receipt of IPv4 space through the waiting list, an 8.3 specified transfer or an 8.4 inter-RIR transfer removes an organization from the waiting list. A successful waiting-list distribution is therefore linked institutionally to the private market: using one route affects access to another.
The waiting list also imposes liquidity restrictions. ARIN states that address space distributed from the waiting list cannot be transferred to another organization for 60 months, except for 8.2 merger, acquisition or reorganization transfers. A filled waiting-list recipient must also wait 90 days before applying for additional space, unless a waiver is granted. These rules reduce speculative queue behavior, but they also make the address block less liquid. The recipient receives operational capacity, not freely movable capital. In institutional terms, ARIN is preserving the moral logic of distribution: the block was given because of demonstrated need, so it should not quickly become market inventory.
The problem is that need and capital are not separable after exhaustion. An operator that receives a /22 and deploys it efficiently has acquired a scarce input. If its business changes, if it is acquired, if it exits a product line, or if leasing demand exceeds internal demand, the economic value of that block remains. A 60-month hold does not erase value; it delays the ability to express value through transfer. That may be acceptable as a policy trade-off, but it is still a trade-off. It protects one fairness objective by limiting one form of liquidity.
There is another hidden cost: planning uncertainty. The waiting-list status page states that distribution order depends not only on request chronology but on the order, size and quantity of blocks that ARIN receives and places back into inventory. That means a qualified applicant is waiting for a stochastic supply event. The registry can publish the queue; it cannot guarantee the flow of returned space. For a small operator, that uncertainty changes behavior. It may overuse NAT. It may buy from a broker before it wants to. It may lease. It may postpone customers. It may accept addresses with weaker reputation history. It may spend management time on registry process rather than network growth.
This is the economic significance of exhaustion: scarcity does not merely raise prices. It changes organizational attention. The address plan becomes a board-level and treasury-level issue. The registry ticket becomes a growth constraint. A blocklist history becomes a due-diligence item. A ROA transition becomes part of acquisition execution. An item that used to be a back-office request becomes part of commercial strategy.
The waiting list also shows why administrative fairness is not the same thing as market access. A rule can be fair within the queue while leaving the broader economy highly unequal. The queue decides who may receive returned space under ARIN's rationing logic. It does not decide whether a new entrant can compete with a legacy holder, whether a rural ISP can finance growth, whether a cloud provider can avoid buying ahead, or whether a hosting business can survive a lease renewal shock. Scarcity has moved the center of gravity away from the registry's residual pool and toward market structure. A fair queue is useful, but it cannot substitute for liquid, predictable and safe movement of the much larger stock of already-issued addresses.
Transfers as capital allocation
The transfer market is the principal institutional bridge between registry order and market scarcity. ARIN's transfer guidance describes three main routes: transfers due to mergers, acquisitions and reorganizations under NRPM 8.2; transfers to specified recipients within the ARIN region under 8.3; and inter-RIR transfers under 8.4. In each case the registry record moves only when policy conditions are satisfied. The market may agree on price, but ARIN decides whether the registration change fits policy.
That is not unusual for an RIR. It is also not a normal asset market. In a mature property or securities system, the function of a registrar is usually to record a valid transaction, not to decide whether the buyer has a sufficient operational use for the thing being acquired. In ARIN's system, specified transfer recipients face a needs framework. The transfer page states that recipients within the ARIN region must demonstrate need for up to a 24-month supply of IPv4 addresses. The minimum transfer size is a /24. Recipients must have an updated and signed RSA and pay applicable fees. ARIN offers transfer pre-approvals based on projected need.
The Number Resource Policy Manual makes the capital-allocation function explicit. The current public version is presented as NRPM 2025.1, dated 3 March 2026. Section 8.5 says the receiving entity must sign an RSA unless it already has a current RSA on file; resources are allocated or assigned via transfer solely for use on an operational network; ARIN's minimum IPv4 transfer size is a /24; organizations without an ARIN IPv4 allocation qualify for an initial /24; larger initial or additional transfers require documentation that at least 50% of the requested block will be used within 24 months. Existing holders must have efficiently utilized at least 50% of cumulative IPv4 blocks to receive more. An alternative route allows organizations demonstrating 80% utilization of currently allocated space to receive transfers up to the total size of current ARIN IPv4 holdings, capped at a /16 equivalent in any six-month period.
These thresholds are the administrative skeleton of the North American transfer market. They give buyers a path and prevent purely passive accumulation through registered transfers. They also constrain capital. A buyer with financing, demand expectations and a strategic desire to secure long-term IPv4 supply cannot simply buy what it thinks the market price justifies. It must fit a registry definition of operational need. Capital alone does not prove need. In ordinary markets, the willingness to spend millions of dollars is itself a signal. In a needs-assessed transfer market, that signal is filtered through administrative judgment.
The institutional justification is conservation. IPv4 addresses are unique; unused concentration can harm entrants; operational justification keeps resources connected to networks. The economic objection is that conservation policy can become liquidity suppression. If a firm cannot buy for long-term inventory, if a seller fears transfer delay, if a buyer's future use case is harder to document than its capital commitment, or if policy interpretation changes, price discovery becomes less clean. Assets become less valuable when transferability is conditional.
ARIN tries to reduce uncertainty through process. Its transfer guidance includes submission steps, documentation expectations, a signed RSA deadline, processing timelines after approval and fees, and routing-security transition checklists. It recognizes that transfer is not merely a billing event. Source organizations should edit or delete transferring prefixes from ROAs, review maxLength values, update IRR objects, coordinate reverse DNS and ensure the recipient understands its responsibilities. This operational detail is important. A transferred address block carries history: RPKI state, IRR objects, reverse DNS, abuse reputation, geolocation records, firewall allowlists and customer dependencies.
That operational complexity strengthens the case for a registry, but it also strengthens the case for market maturity. The more a transferred block behaves like a high-value operational asset, the less plausible it becomes to treat transfer as a minor administrative privilege. Due diligence around RPKI and IRR is the language of capital markets applied to routing. The buyer is not merely acquiring numbers. It is acquiring a continuity surface.
The transfer market also changes corporate structure. If a company can move addresses more easily through a merger or acquisition than through a specified transfer, deal lawyers will notice. A block can become a reason to buy an entity rather than a simple asset. A subsidiary may be kept alive because it carries a cleaner registration history. A seller may prefer a buyer whose corporate form fits ARIN's evidence expectations. A broker may screen not only for price and block size, but for documentation quality, officer authority, previous transfer history, reserved-pool status, dispute status and timing restrictions. This is how administrative rules become part of capital markets: not by setting prices directly, but by shaping the transaction structures that make prices executable.
Inter-RIR transfers add a further layer. ARIN's transfer page says such transfers can occur only between RIRs with reciprocal, compatible, needs-based policies. It currently identifies APNIC, LACNIC and RIPE NCC as approved for transfers with ARIN, while AFRINIC is not approved for that purpose. The practical result is that regional policy compatibility becomes a capital mobility condition. An address block may be globally routable in technical terms, but its recognized registration mobility remains regional and policy-dependent. The map of packet flow and the map of registry transferability are not the same map.
Legacy holdings and the North American inheritance
ARIN's region contains a large legacy-resource inheritance because the Internet's early administrative history was heavily North American. ARIN's legacy-resource page explains that before ARIN was formed in December 1997, IP address space and ASNs not administered by RIPE NCC or APNIC came into the ARIN-administered database. These earlier assignments are legacy number resources. At ARIN's formation, its Board decided to provide registration services for those legacy resources without requiring original holders to enter into a Registration Services Agreement or pay service fees.
That decision shaped the political economy of IPv4 scarcity. Legacy holders are not all the same. Some are universities or research institutions. Some are corporations with old networks. Some are technology companies whose address holdings became strategic assets. Some have reorganized, sold portions, signed agreements, deployed RPKI, or entered the transfer market. But as a category, legacy resources demonstrate the conversion of historical administrative abundance into modern capital advantage.
The legacy page states that organizations with legacy resources not under an ARIN agreement can maintain unique registration information in WHOIS/RDAP, update and manage publicly available data, manage reverse DNS delegations, maintain registry records through ARIN Online and access DNSSEC. It also states that organizations with legacy resources must be under an ARIN agreement to access ARIN's RPKI and Internet Routing Registry services. From 11 October 2007 through 31 December 2023, ARIN offered the Legacy Registration Services Agreement to organizations and individuals in the ARIN service region with legacy resources. The page says the legacy fee cap expired on 31 December 2023; organizations with an active LRSA entered before 1 January 2024 continue to have limited fees for legacy resources covered before that date, while resources covered under an ARIN agreement after 1 January 2024 are subject to annual Registration Service Plan fees.
This is a clean example of institutional adaptation under scarcity. ARIN could not pretend legacy space was ordinary newly issued space. It had to maintain the database and respect the historical position of holders. At the same time, modern services such as RPKI and IRR create incentives for legacy holders to sign agreements. The registry does not need to seize the resource. It can attach service value to contractual alignment.
The market consequence is uneven certainty. A legacy block outside an agreement may have a different risk profile from a block under the RSA. A block under an older LRSA may have a different fee profile from a block brought under agreement after January 2024. A buyer seeking transfer may need to clean up corporate authority, POC records, historical documentation, abuse reputation and routing state. A seller may discover that the most valuable thing it owns is also the hardest thing to document internally because it was obtained in an era when no one imagined a balance-sheet asset.
The North American inheritance therefore creates both liquidity and friction. It creates liquidity because many large blocks exist in entities whose current business may not require all of them. It creates friction because historical claims must be reconciled with modern transfer rules, corporate changes and registry documentation. The result is a market in which legal history, network engineering and capital planning intersect.
For small operators, the legacy inheritance can feel like a structural disadvantage. Early institutions received large blocks under informal or low-cost conditions. Later entrants must justify need, wait in line, buy at market prices or lease. The distribution may be historically explainable, but it is not competitively neutral. Scarcity turns history into market power. An incumbent's old /16 can finance expansion or generate leasing income; an entrant's lack of addresses becomes an upfront cost.
This is why the economics of IPv4 cannot be reduced to hoarding rhetoric. Some underused legacy space should move to higher-value use. But the mechanism matters. If movement depends on clear title-like expectations, predictable transfer, transparent registry records and market pricing, addresses can flow. If movement depends on moral condemnation, opaque pressure or expanding enforcement discretion, holders may freeze. Liquidity improves when owners believe they can transact without arbitrary loss. It deteriorates when they fear that visibility invites control.
The legacy question also exposes the difference between registration services and ownership rhetoric. ARIN can say, correctly, that its function is registry administration rather than ordinary property conveyance. A legacy holder can say, also correctly in economic terms, that a scarce address block has commercial value independent of the registry's preference for non-property language. The useful compromise is not to pretend that either side has the whole answer. The holder needs stable, transferable, secure registration rights. The registry needs accurate records, anti-fraud authority and contractual clarity around services. The market needs both sides to stop confusing legal vocabulary with operational reality.
RSA certainty, RSA asymmetry
ARIN's 2025 Registration Services Agreement is more sophisticated than the caricature of a registry contract. It defines "Included Number Resources" to include registration rights for IP address space and ASNs issued by ARIN, plus legacy resources specifically identified as subject to the agreement. It says services include registry entries, reverse name service, RPKI, maintenance of records and administration of IP address space. It grants the holder, subject to ongoing compliance, the exclusive right to be the registrant of the included resources within the ARIN database, the right to use them within the ARIN database and the right to transfer their registration pursuant to policy. It also says the holder acquires express contractual rights to the included number resources by virtue of the agreement.
Those clauses matter. They give ARIN-region holders more explicit contractual language than many casual discussions of number resources acknowledge. They support a market in which counterparties can understand what is being transferred: not metaphysical ownership of a number in nature, but a bundle of contractual registration, use and transfer rights within ARIN's registry system. That bundle is economically valuable because the Internet needs uniqueness and because ARIN's database is treated as authoritative for its region.
But the same RSA also reveals the asymmetry of the registry layer. Services and resources are subject to ARIN policies. Policy changes can become binding through ARIN's published process. ARIN may respond to governmental or judicial orders, including orders to stop services or terminate the agreement. Fees are a condition of service; if invoices remain unpaid long enough, ARIN may stop services, terminate the agreement and revoke included number resources. ARIN may review utilization when a transfer or additional address space is requested. It says it will not reduce services for lack of utilization except as provided, and has no right under the agreement to revoke included resources for lack of utilization; but it may refuse transfers or additional allocations if resources are not utilized in accordance with policy.
The disclaimers are equally important. The RSA provides services, included resources and registration on an as-is basis. ARIN disclaims warranties that services or resources will be uninterrupted, error-free, meet requirements or operate with the holder's configuration. Consequential and special damages are excluded. Aggregate liability is capped at the greater of the amount paid by the holder for services during the six months preceding the event or US$100. In bankruptcy language, the agreement states that none of the number resources, services or anything else provided by ARIN is or will be property of the holder's bankruptcy estate within the meaning of section 541 of the U.S. Bankruptcy Code.
This is the legal form of the institutional economics problem. ARIN recognizes contractual rights and provides a serious registration framework. Yet the holder's operational downside may be far larger than the registry's contractual downside. A cloud company, ISP, security platform or hosting company cannot measure the loss of a significant block only by six months of registry fees. The real loss may include customer churn, renumbering, routing disruption, firewall and allowlist changes, compliance incidents, geolocation problems, abuse remediation, engineering labor, litigation and lost revenue.
The point is not that ARIN is uniquely unfair. Standard-form infrastructure contracts often limit liability. The point is that scarcity changes the scale of consequence. When a registry record governed by limited-liability service terms determines the usability and transferability of high-value address blocks, the registry becomes a risk layer. Holders face a mismatch between the value they depend on and the remedy available if the registry layer fails, delays, misprocesses or becomes entangled with external orders.
This is why "owning" directly is not always the simple answer operators assume. A direct holder may have stronger registry visibility, but it also internalizes the registry contract, the policy surface, payment obligations, review processes, government-order compliance, POC security and termination mechanics inside the operating company. A lessee or downstream user faces counterparty risk instead, which may be better or worse depending on the lessor. The economically relevant question is not whether a name appears in the registry. It is where the registry-layer risk sits and who is best positioned to carry it.
The LRSA history sharpens the point. For years ARIN gave legacy holders a contractual path that was intentionally different from the standard non-legacy relationship. The fee cap made sense as a bridge between historical claims and modern services. Its expiration for new coverage after 2023 changed the incentives around bringing resources under agreement. Some holders will accept that trade-off because RPKI, IRR access, cleaner records and transfer readiness matter. Others may treat agreement status as new exposure. Both reactions are economically rational. Contract alignment supplies services and certainty, but it also brings the holder into a policy and fee structure that did not exist when many legacy resources were originally assigned.
Needs assessment after exhaustion
Needs assessment made intuitive sense when registries were distributing free or low-cost common resources. If a network wanted addresses from a limited pool, it should show operational use. The problem after exhaustion is that the same logic becomes a market constraint. In a transfer, the buyer is not receiving a gift from the common pool. It is paying a seller. The buyer's capital is at risk. The seller is giving up an asset. The registry is preserving uniqueness and policy order. Requiring an additional needs test therefore shifts the system from market exchange toward permissioned exchange.
ARIN's current transfer policy is not absolute central planning. It allows transfers, pre-approvals, inter-RIR movement and a six-month path for qualifying additional blocks under utilization thresholds. But the institutional logic remains needs-based. Address blocks should move to operational use, not to passive speculation. That principle has community support, and for understandable reasons. Nobody wants a thin financial class to lock up routing identifiers while real networks struggle.
Yet "need" is harder to measure than it appears. A cloud platform may need inventory before customers sign contracts. A data center operator may need addresses to make colocation products sellable. A security company may need clean blocks for customer segmentation, reputation isolation and deliverability. An AI infrastructure operator may need addresses for distributed services, not because each server requires one public IPv4 address but because customers, appliances, API endpoints, compliance systems and legacy software still expect IPv4 reachability. A buyer may need more addresses to avoid future exposure to price spikes. That is a business need, even if it is not immediately visible in a 24-month utilization spreadsheet.
Needs assessment is therefore an information problem. The registry must distinguish genuine operational planning from speculative warehousing. But the more discretion it exercises, the more it influences market structure. A strict interpretation favors incumbents with visible current utilization. A flexible interpretation favors strategic buyers. A documentation-heavy interpretation favors large firms with compliance staff. A slower interpretation favors those who can pay intermediaries or lease temporarily. A rule intended to conserve resources can unintentionally raise the cost of entry.
The institutional-economics critique is not that all needs assessment is illegitimate. It is that needs assessment after exhaustion should be understood as capital control. It decides how quickly scarce address capital may move, how much a buyer may accumulate, what business plans count as valid, and when private price signals are insufficient. In financial markets, capital controls may sometimes be justified. They still have costs: lower liquidity, higher spreads, more administrative arbitrage, weaker price discovery and incentives to use unofficial channels.
The IPv4 analogue of unofficial channels is not always illegal transfer. It may be leasing, reassignment, hosting arrangements, acquisitions structured around address holdings, cross-border transfer planning, shell entities with regional nexus, or complex operational justifications. When formal transfer is too constrained, market participants do not stop needing addresses. They search for structures that turn policy into a solvable transaction problem.
ARIN's system remains more orderly than many alternatives because it has a recognized specified-transfer process. But the existence of a legal market does not eliminate shadow pressure. It merely determines how much demand flows through clean transfers, how much through leasing, how much through M&A wrappers and how much through operational workarounds. The tighter the gate, the greater the incentive to move around it.
Needs assessment also affects information quality. If applicants know that certain narratives fit policy better than others, they will optimize documents for approval rather than describe business reality with perfect candor. A firm's internal reason for buying may be option value, merger readiness, customer-confidence signaling or protection against future price shocks. The registry may prefer evidence of near-term network deployment. The result is a translation exercise. Some translation is unavoidable, but excessive translation creates a market in policy documentation rather than a market in address capacity. That is a waste of institutional attention.
The more mature approach would treat private capital commitment as evidence, not as conclusive proof. A buyer willing to pay market price for scarce capacity has revealed something real. The registry can still ask whether the source is authorized, whether the recipient exists, whether the transaction is fraudulent, whether the block is disputed, whether the records will remain accurate and whether the request violates explicit policy. But it should be cautious about substituting its own forecast for the business judgment of actors who will bear the economic loss if the addresses are not useful.
Leasing and shadow allocation pressure
Leasing is the practical answer many operators choose when purchase is too slow, too expensive, too uncertain or too capital-intensive. It is also the area where the old registry vocabulary is least adequate. If a resource holder leases addresses to customers, who is the true operational user? If a hosting company assigns addresses to dedicated servers, is that leasing, hosting or ordinary customer assignment? If a first-party lessor maintains the registry relationship and customers route or use downstream space, where does registry responsibility end and commercial responsibility begin?
The market does not wait for perfect doctrine. IPv4 leasing exists because addresses are scarce and because many users need capacity without buying a block outright. Leasing converts a capital expenditure into an operating expenditure. It allows faster deployment. It gives small firms access to addresses they could not afford to buy. It lets holders monetize idle inventory without permanently transferring it. It supports geographically distributed services, temporary campaigns, lab environments, VPNs, hosting, email, adtech, security and cloud workloads.
But leasing also creates risk. A leased block may have poor reputation. The lessor may lose registry standing. The customer may trigger abuse complaints. RPKI authority may not align cleanly with the routing party. Reverse DNS and geolocation updates may lag. A downstream customer may build business continuity on a resource it does not control. A brokered leasing chain may create multiple layers of promises without one party clearly responsible for long-term continuity.
This is why the leasing question is really a registry-layer question. Direct ownership can look safer because the operator's name appears in the registry, but direct ownership also places the operator under the registry contract and policy surface. Leasing can look weaker because the customer is not the registrant, but leasing from a strong first-party holder may move registry risk away from the operating company. Thin brokerage may pass risk through. Structured leasing may absorb and manage it. The difference is not the label "lease"; it is the risk architecture.
ARIN's formal policy framework is built around allocation, assignment, reassignment, transfer and registration. The commercial market increasingly thinks in terms of capacity, continuity, reputation, renewal, route authorization, abuse handling and capital efficiency. A rational operator asks: Can I keep customers online? Can I renew? Can I create ROAs? Can I maintain reverse DNS? Can I survive a dispute? Can I avoid renumbering? Can I price the risk? These questions are not answered by ownership rhetoric alone.
Leasing also introduces what might be called shadow allocation pressure. When the waiting list cannot supply enough addresses and the transfer market requires capital plus needs review, leasing becomes a parallel allocation system. It allocates by willingness to pay recurring fees and by counterparty access. That may be more efficient than a queue, but it is not automatically more equitable. It can concentrate control in large address holders and sophisticated lessors. It can also democratize access by letting small operators rent what they cannot buy.
The policy challenge is to distinguish harmful opacity from useful market structure. A registry should care about uniqueness, accuracy, fraud prevention, routing-security coherence and contactability. It should not pretend that every downstream commercial use can be morally adjudicated from the registry desk. If leasing is pushed into ambiguity, the market becomes less transparent. If leasing is recognized as a legitimate continuity and capacity tool, the registry can focus on record integrity and abuse-relevant accountability rather than trying to suppress the economic consequences of scarcity.
Leasing will also test how much the market values continuity over nominal control. Some buyers prefer registration in their own name because it simplifies governance and reduces counterparty dependence. Others may prefer a lease from a holder with deep inventory, clean operational processes, experienced abuse handling and the ability to absorb registry-side friction. The distinction resembles the difference between owning a data center and buying cloud capacity. Ownership supplies control, but it also concentrates risk. Service models supply flexibility, but only if the provider is strong enough to make continuity credible. IPv4 scarcity is pushing address capacity toward the same institutional trade-off.
For ARIN, the lesson is that suppressing leasing would not restore the old allocation model. It would merely push leasing into less visible forms: hosting bundles, managed services, reassignment chains, shell arrangements or opaque brokered structures. The better policy posture is to make legitimate leasing safer by encouraging accurate public records where appropriate, responsible abuse contacts, clear routing-security practices, and realistic education for downstream users. The registry cannot abolish the economics of scarcity. It can either illuminate them or drive them into the dark.
Small operators and the scarcity tax
The hardest case for ARIN scarcity is not the hyperscaler with a large treasury. It is the small ISP, regional hosting company, rural broadband operator, managed-service firm, enterprise network, school network, tribal network, Caribbean operator or start-up platform that still needs IPv4 because customers, vendors, legacy systems and global reachability still require it. IPv6 may be technically available, and many networks should deploy it. But IPv6 does not remove the need to reach IPv4-only services or serve customers whose equipment, software and counterparties remain IPv4-dependent. The dual-stack period is not a transition phase that ended. It is a durable operating condition.
For a small operator, IPv4 scarcity appears as a tax on growth. Buying addresses consumes capital that could otherwise fund radios, fiber drops, colocation, routers, support staff or security. Waiting-list supply is uncertain and capped. Transfer qualification requires documentation. Leasing adds recurring cost and counterparty dependence. Carrier-grade NAT reduces public-address demand but creates support costs, logging burdens, application problems and customer dissatisfaction. IPv6 helps, but it does not eliminate the compatibility burden.
This scarcity tax is not evenly distributed. Large incumbents often have historical holdings, more efficient internal utilization, better legal teams and the ability to buy ahead. New entrants must acquire at current market prices. A small rural ISP may be serving customers with thin margins while competing against firms whose address inventory was acquired decades earlier at negligible cost. The market price of IPv4 therefore embeds historical inequality.
Administrative fairness tools such as ARIN's waiting list partly address this imbalance. A /22 cap and exclusion of organizations holding more than a /20 equivalent aim to target scarce recycled space toward smaller players. But the scale mismatch remains. A /22 can help; it cannot transform the economics of a growing access network. Nor can it solve the broader problem that public IPv4 reachability remains a customer expectation in many contexts.
Small operators also face reputation and liquidity disadvantages. They may not have the staff to evaluate the abuse history of a leased or purchased block. They may not understand how RPKI transitions affect route acceptance. They may rely on brokers whose incentives are transaction-driven. They may lack bargaining power in lease renewals. They may find that the cheapest addresses are cheap because they carry deliverability or blocklist problems. Scarcity therefore creates a quality gradient, not only a price gradient.
The institutional lesson is that suppressing IPv4 asset value does not automatically help small operators. If price signals are suppressed through policy friction, incumbents may simply hold more. If transfer is difficult, sellers may avoid the market. If leasing is stigmatized, small operators lose a flexible access channel. If registries expand discretionary review, firms with compliance departments navigate better than firms without them. Moralizing scarcity can produce the opposite of inclusion.
A more realistic small-operator policy would accept that IPv4 is capital and then ask how market access can be made safer. That means clearer transfer timelines, predictable pre-approval, clean public data on returned blocks, stronger education around ROA and IRR transitions, better abuse-reputation due diligence, transparent facilitator standards, and recognition that leasing can be a legitimate access path when structured responsibly. The goal should be operability, not nostalgia for abundance.
There is a broader social consequence. Public-policy debates often treat small operators as beneficiaries of rationing and large holders as beneficiaries of markets. The reality is more mixed. A small operator benefits from a waiting-list block if it receives one in time. It benefits from a liquid transfer market if sellers can transact without fear. It benefits from leasing if counterparties are reliable and terms are transparent. It benefits from IPv6 if customers and vendors are ready. No single mechanism solves scarcity. A small operator needs a portfolio of access channels. Registry policy should avoid romanticizing one channel at the expense of the others.
Pricing, liquidity and the value of dormant space
IPv4 pricing is the visible surface of a deeper institutional transition. Public broker data and market commentary have for several years placed many IPv4 sale prices in the tens of dollars per address, with variation by block size, region, reputation, urgency and transaction structure. Leasing prices are usually quoted monthly per address and vary by term, support level, reputation, routing requirements and counterparty. The exact number matters less than the direction: IPv4 has a market price because it is scarce, useful and transferable enough for buyers and sellers to transact.
The important question is liquidity. A market can show high quoted value and still be inefficient if only a small fraction of holdings trade each year. Dormant address space may be economically valuable but practically sticky. Holders may not know what they own. Corporate records may be stale. Boards may fear selling a strategic asset too early. Legal teams may worry about title, tax, accounting, sanctions, export controls, customer dependencies or registry review. Network teams may resist renumbering. Sellers may prefer leasing income to sale proceeds. Buyers may wait for price drops that never come. Brokers may have incomplete visibility into real inventory.
Low liquidity has two effects. First, it makes price discovery noisy. A few urgent transactions can move expectations. Clean large blocks command premiums. Bad reputation discounts can be severe. Regional policy differences create arbitrage. Second, it preserves hidden wealth. If large holders do not need to sell, the market price does not force their assets into productive use. Scarcity remains real for entrants while unused or lightly used blocks sit inside organizations that may not be natural network operators.
This is where enforcement creep becomes economically dangerous. If holders believe that engaging the market invites intrusive review, retrospective questioning or moral suspicion, they may stay inactive. If buyers believe that approval is uncertain, they discount bids. If lessors fear that visible leasing will trigger policy pressure, they structure around visibility. A registry that tries to prevent speculative abuse can unintentionally reduce legitimate liquidity.
At the same time, a completely ungoverned market would create its own risks. Fraudulent transfers, hijacked blocks, forged corporate authority, abandoned POCs, stale ROAs, broken reverse DNS and dirty reputation histories can harm networks. A serious IPv4 market needs a trusted ledger and reliable operational transition. The argument is not for removing the registry. It is for limiting the registry to functions that make the market safer: uniqueness, record accuracy, anti-fraud controls, clear transfer status, dispute notation, routing-security support and neutral publication.
The valuation of IPv4 also affects telecom and infrastructure company finance. Address holdings can influence acquisition value, leasing revenue, collateral narratives, impairment analysis and strategic optionality. A company with significant unused IPv4 can fund expansion or improve resilience. A company without addresses must spend scarce capital before earning customer revenue. The asset is not always formally recognized on balance sheets in a straightforward way, but market actors know it matters.
The most mature view is therefore neither "IPv4 is just a database entry" nor "IPv4 is ordinary property." It is an infrastructure asset whose economic value depends on registry recognition, routing acceptance, clean operational state and contractual continuity. Liquidity improves when those dependencies are predictable. Liquidity worsens when they are discretionary.
The dormant-space problem is especially revealing. From a narrow conservation perspective, unused space should move to users. From a market perspective, holders need confidence before they expose themselves to transaction review, tax consequences, reputational questions and opportunity-cost decisions. From a registry perspective, the record must remain accurate and transfers must be legitimate. The policy task is to align these incentives. A punitive or moralizing approach may satisfy a sense of fairness but reduce the actual supply entering the market. A predictable, narrow, well-documented transfer environment is more likely to draw dormant space into productive circulation.
Liquidity also depends on information infrastructure. Market participants need to know whether a block has clean registry history, whether POCs are current, whether an agreement applies, whether ROAs exist, whether IRR objects need cleanup, whether reverse DNS is delegated, whether geolocation databases are likely to lag, whether the block is on major reputation lists, and whether prior use creates customer risk. Some of this information is outside ARIN's role. Some sits directly in registry services. The boundary matters. The registry should not become a broker, but it should recognize that better records lower transaction costs for the whole market.
Capital-control risk
Capital control is a strong phrase, but it describes a real mechanism. When a scarce asset cannot move unless an administrative body accepts the buyer's need, the seller's authority, the region's nexus, the documentation and the policy fit, the registry is not merely keeping records. It is controlling capital mobility. In ARIN's case that control is rule-bound and publicly documented. It is still control.
The risk is not only denial. Delay can be enough. A buyer may lose a deal if transfer approval takes too long. A seller may accept a lower price for a simpler counterparty. A broker may structure around the most predictable route, not the economically best one. A company may acquire an entity instead of buying a block because the corporate transaction better fits policy mechanics. A lessee may accept recurring cost because purchase approval is uncertain. These are capital-allocation effects created by registry rules.
Out-of-region use shows the issue clearly. NRPM section 9 permits ARIN-registered resources to be used outside the ARIN service region, but out-of-region use as justification for additional number resources requires a real and substantial connection with the ARIN region and the same type of resources used within the region, including at least a /22 used in-region for IPv4. The policy lists possible evidence such as physical presence, staff, assets, services to regional residents, meetings, investment capital and incorporation, while stating that incorporation alone is insufficient and that the weight of factors is determined by ARIN.
This is a reasonable attempt to prevent paper claims on ARIN resources by entities with no regional substance. It is also discretionary regional capital control. A global network may have customers and infrastructure across borders. A North American entity may operate services in Latin America, Europe, Africa or Asia. A multinational may centralize procurement. A cloud platform may route capacity where demand appears. Regional nexus rules shape which corporate structures can acquire and hold address resources efficiently.
The broader institutional question is whether RIR regions should behave like capital jurisdictions. IPv4 itself routes globally. Packets do not respect RIR borders. Business models increasingly cross regions. Yet registry policy remains regional because the RIR system is regional. That creates arbitrage and friction. Inter-RIR transfers require reciprocal and compatible policies. Some RIRs accept some resource types and not others. Different needs tests, fees, holding periods and contractual assumptions affect where address capital can move.
ARIN is often viewed as one of the more functional nodes in this system. That makes its capital-control role more consequential, not less. If the most mature transfer market still requires administrative need, regional nexus and policy-defined use, the global market remains structurally permissioned. Investors, operators and lessors must price not only address scarcity but policy mobility.
Capital controls can be defended when they preserve system integrity. They become dangerous when they preserve institutional power at the expense of productive use. The difference lies in narrowness, transparency and accountability. A narrow rule prevents duplicate registration or fraud. A broad discretionary rule decides whether a business model deserves address capital. The former protects the ledger. The latter governs the market.
There is also a political economy risk. Once a registry becomes used to controlling capital mobility, it may mistake control for legitimacy. Each additional rule can be described as community policy, conservation, stewardship or anti-abuse discipline. Some will be justified. Others may simply preserve the authority of the registry and the preferences of incumbents who have learned to operate under its procedures. A mature market should ask the same question of every control: what concrete ledger harm does it prevent, and is the cost to liquidity proportionate?
This question is not anti-registry. It is pro-ledger. Duplicate registration, forged transfers and corrupted records would destroy value faster than policy restraint would. But capital controls that are too broad can also destroy value by freezing movement, discouraging visibility and pushing demand into shadows. The narrow ledger function and the broad capital-allocation function must be kept analytically separate. Otherwise every argument for record integrity becomes an argument for institutional discretion.
Registry-layer risk
The registry layer is often invisible until something fails. Most days, WHOIS and RDAP respond, reverse DNS works, ROAs validate, invoices are paid, POCs remain current and transfer tickets move through process. On those days the registry looks like plumbing. Scarcity changes the evaluation standard. A system that controls valuable assets must be judged not only by normal days but by stress: disputes, insolvency, litigation, political pressure, sanctions, governance failure, cyber incidents, data errors, compromised accounts, mistaken revocation, RPKI misconfiguration and policy shocks.
ARIN's official materials show several stress surfaces. The RSA allows ARIN to follow government or judicial orders concerning number resources or services, including orders to stop service or terminate the agreement. It allows suspension or termination for defined causes. It limits damages. The transfer process requires authority documentation. The waiting list can remove tickets for non-payment. Legacy holders outside agreement lack access to certain services such as RPKI and IRR. Each of these rules may be rational. Together they show that a holder's continuity depends on a registry-controlled stack.
Registry-layer risk is not the same as ARIN-specific misconduct risk. It is structural. Any registry that maintains the recognized ledger has potential choke-point power. It may not own the economic value, but its records help determine who can use, transfer and secure that value. If the registry's liability is small relative to the holder's potential loss, risk is externalized to operators. If policy changes bind holders, future governance becomes part of present asset value. If registry services are necessary for clean routing-security operations, service access becomes a commercial dependency.
This is why the distinction between protecting the ledger and protecting the gatekeeper is crucial. The ledger must be protected. Duplicate registration would be destructive. Fraudulent transfers must be blocked. Accurate contact and routing-security data matter. Reverse DNS, RDAP, WHOIS and RPKI continuity are operationally important. But none of that requires treating registry discretion as inherently legitimate whenever scarcity is involved. A neutral ledger is infrastructure. A discretionary gatekeeper over capital is an economic actor.
In the ARIN context, the best version of registry accountability would preserve the boring functions and narrow the discretionary ones. It would make transfer review predictable, timelines measurable, data quality auditable and dispute procedures independent. It would avoid using policy as a substitute for market price. It would recognize that holders and operators bear real downstream continuity costs. It would treat IPv4 as capital without pretending that capital status abolishes the need for accurate registry coordination.
The alternative is a slow legitimacy problem. Operators tolerate registry power when they perceive it as neutral, technical and proportionate. They resist when they perceive it as discretionary, moralized or detached from liability. Scarcity raises the stakes because every policy decision can redistribute value. A waiting-list rule affects entrants. A transfer rule affects liquidity. A legacy-fee rule affects old holders. An RPKI eligibility rule affects security posture. An out-of-region rule affects multinational capital structure. Each decision is institutional economics, even when written as registry administration.
Registry-layer risk also has an information-security dimension. POC authority, ARIN Online access, officer acknowledgements, signed documents and account status are not mere administrative details. They are controls over assets that may be worth millions of dollars in market terms and far more in operational continuity. A compromised account, stale POC, careless corporate reorganization or missed invoice can become more than a clerical defect. It can become a business-continuity event. The more valuable IPv4 becomes, the more the registry interface resembles a financial-control surface.
The correct response is not panic. It is discipline. Companies should treat number-resource governance as part of risk management. They should know which resources are held directly, which are legacy, which are under RSA or LRSA, which are leased, which are reassigned, which have ROAs, which depend on third-party POCs, which have reverse DNS dependencies, which are subject to transfer restrictions and which could be affected by merger or bankruptcy events. Registry-layer risk cannot be eliminated. It can be inventoried, priced and allocated.
What ARIN tests
ARIN tests whether a post-exhaustion registry can remain a trusted coordination layer while operating inside a market it does not fully control. It has several advantages: a sophisticated operator community, public documentation, an established transfer process, explicit contractual rights language, visible legacy-resource guidance and enough market activity to support price discovery. If the RIR model can adapt anywhere, ARIN should be one of the easier cases.
That is precisely why its unresolved tensions matter. The waiting list is fair but small. Transfers are real but needs-assessed. Legacy rights are recognized but stratified by agreement status and fee history. The RSA grants contractual rights but limits remedies and preserves policy dependence. Leasing solves practical access but sits uneasily beside allocation vocabulary. Small operators receive procedural protection but still face market scarcity. Global networks can use ARIN resources out of region but must satisfy regional nexus rules. The registry is transparent, yet transparency does not eliminate the economic effect of its controls.
The core post-exhaustion challenge is to move from scarcity denial to scarcity governance. Scarcity denial says IPv4 should be treated as administrative stewardship because recognizing asset value might harm the community. Scarcity governance says the asset value already exists; the policy question is how to make that value move safely, transparently and productively. The first approach preserves moral authority for administrators. The second approach asks administrators to become narrower, more accountable and more technically focused.
For operators, the practical lesson is direct. IPv4 strategy is no longer a procurement afterthought. It belongs in capital planning, M&A diligence, risk management, customer-continuity planning and board-level infrastructure policy. A firm should know what it holds, under which agreement, with what POC authority, what ROAs, what IRR objects, what reverse DNS, what abuse history, what transfer restrictions, what lease obligations and what renewal or registry exposure. A firm that treats IPv4 as a ticketing issue is mispricing its own risk.
For sellers, the lesson is that dormant space is not automatically liquid. Clean records, corporate authority, reputation remediation, routing-security cleanup and registry-process readiness all affect value. For buyers, the lesson is that price is only one term. Transferability, needs qualification, RPKI transition, blocklist history, geolocation, reverse DNS, contractual continuity and future policy exposure all matter. For lessors, the lesson is that the market will increasingly distinguish between thin intermediaries and continuity-bearing structures.
For ARIN and the broader RIR system, the lesson is harder. The more valuable IPv4 becomes, the less sustainable it is to rely on old categories. "Allocation" does not describe what happens when a company pays market price for a block. "Need" does not capture strategic inventory. "Community" does not by itself justify capital control. "Stewardship" does not answer liability. The registry's legitimacy will depend on whether it can remain neutral infrastructure rather than becoming the institution that decides which forms of address capitalism are morally acceptable.
The North American case is therefore not a minor regional story. It is a test of institutional adaptation in a mature market. ARIN's comparatively orderly environment removes many excuses. There is no need to invoke emergency, collapse or chaos to see the underlying tension. The ordinary rules are enough. A waiting list, a needs test, a transfer guide, a legacy page and a contract can quietly define who bears scarcity, who can move capital, who can monetize history, who pays for uncertainty and who controls the record.
Watchpoints
Several indicators will show whether ARIN's scarcity regime is becoming more market-compatible or more permissioned.
The first is waiting-list behavior. The size, age and fill pattern of the waiting list will show how much small-operator demand remains outside the transfer market. If recovered blocks are sporadic and small while demand persists, the waiting list will remain a symbolic fairness mechanism rather than a material supply channel. If non-payment revocations or returns increase, the quality and reputation history of recycled blocks will matter more.
The second is transfer velocity. A healthy market should have predictable approval times, clear documentation standards and enough completed transactions to support price discovery. If transfer volumes stagnate despite demand, the likely causes will be price expectations, policy friction, seller reluctance or buyer qualification uncertainty. If transfers concentrate among large players with sophisticated counsel, small-operator access may be weakening even if the policy is formally open.
The third is legacy conversion. The retirement of the legacy fee cap for new agreement coverage after 1 January 2024 changes incentives. Some holders may sign agreements to obtain RPKI or IRR access and facilitate transfers. Others may avoid new contractual exposure. The balance will reveal how holders value modern registry services versus contractual autonomy.
The fourth is leasing normalization. If leasing becomes more transparent, with better routing-security practices, clearer abuse handling and stronger contractual continuity, it can relieve scarcity pressure. If it remains stigmatized or informally structured, it may become a shadow allocation layer with uneven risk. The market's demand for first-party leasing and continuity assurance suggests that many operators already see registry exposure as a business-continuity issue, not merely a purchasing issue.
The fifth is needs-assessment evolution. ARIN's 24-month need framework and utilization thresholds are the heart of its permissioned transfer market. Any policy change that loosens, tightens or clarifies those requirements will directly affect liquidity. The central question is whether the policy community begins to treat capital commitment as evidence of need or continues to treat administrative documentation as superior to price signals.
The sixth is registry liability and dispute design. As IPv4 values rise, the gap between contractual remedies and operational downside becomes more visible. A registry does not need unlimited liability to be legitimate, but it does need credible accountability when its actions affect high-value operational assets. Independent dispute handling, preservation of last verified operational state and clear limits on self-help will become more important.
The seventh is out-of-region policy. Global networks do not map neatly onto RIR borders. If ARIN's real-and-substantial-connection test remains predictable, it can preserve regional legitimacy without freezing global operations. If it becomes too discretionary, regional nexus will be priced as a capital mobility risk. The same applies to inter-RIR compatibility: a global IPv4 market cannot mature if registry borders behave like unpredictable customs posts.
The final watchpoint is language. Institutions reveal their adaptation through vocabulary. If ARIN and the RIR community continue to describe IPv4 mainly as allocated resources under stewardship, the conflict with market reality will grow. If they describe IPv4 as registry-recognized operational capital requiring accurate, neutral and accountable ledger services, the model can evolve without abandoning coordination.
ARIN's post-exhaustion economy is therefore not a side issue in Internet governance. It is the frontier where technical bookkeeping, private capital, small-operator access and institutional legitimacy meet. IPv4 scarcity did not destroy the registry function. It made that function more important. But it also made the function more dangerous when bundled with broad discretion.
The North American lesson is clear. The ledger matters. Uniqueness matters. Registration accuracy matters. Routing-security continuity matters. But the address block's economic value now lives in the market and in the networks that use it. ARIN's task is to protect the ledger without mistaking ledger protection for ownership of the capital that scarcity has created. If it can do that, it will remain a coordination institution in an asset world. If it cannot, the market will keep searching for ways to route around the gatekeeper while still depending on the record.