APNIC and the economics of post-exhaustion legitimacy

The bargain changed when the pool stopped being the answer

The decisive institutional event in the Asia-Pacific address economy was not the discovery that IPv4 was finite. Engineers, operators and registries had known that for years. Nor was it only the formal exhaustion of the ordinary IPv4 pool. The more important turn came when exhaustion changed what APNIC could plausibly promise. Before that point, an operator with a convincing case for network growth could ask the regional registry for addresses and expect the answer, if the case satisfied policy, to include a new allocation from a common pool. After that point, the same operator might still have a real need, customers to serve and a defensible engineering plan, but the registry could no longer make scarcity go away. It could ration a small residual pool, record transfers, maintain the ledger, certify resources, delegate reverse DNS and keep the market from dissolving into conflicting claims. It could not restore abundance.

That change altered the source of legitimacy. In the allocation era, APNIC's authority could be explained through a recognisable public-resource story. The registry administered a finite technical resource on behalf of a region, evaluated need, discouraged waste and distributed addresses under policies developed through community participation. Scarcity mattered, but scarcity still had an administrative outlet. A request could be made, reviewed and, if approved, supplied. The core legitimacy questions were therefore allocative: were the rules published, were similar networks treated similarly, were large incumbents prevented from taking more than they needed, and could new entrants obtain enough address space to begin operating?

Post-exhaustion, that story is no longer sufficient. IPv4 scarcity did not disappear; it hardened into market infrastructure. Addresses became tradable, leasable and strategically retained. Holdings acquired balance-sheet-like significance. A clean registry entry could affect whether a block could be sold, financed, insured, routed, certified, defended in a dispute, used for reverse DNS, or accepted by a future counterparty. Old records became evidence. Transfer logs became market signals. Resource certification and registry contacts became part of the practical difference between a usable asset and a doubtful claim.

The question is therefore not whether APNIC can write a more persuasive brief for IPv6. IPv6 is the long-run technical answer to address scarcity, and the region needs more of it. But IPv6 progress does not abolish the installed base: access networks still using carrier-grade NAT, enterprise systems built around IPv4 assumptions, customers who must reach IPv4-only services, hosting markets that still price IPv4 capacity, and operators whose economics are shaped by the cost of scarce public addresses. A registry can advocate IPv6 without pretending that IPv4 has stopped being an operating asset.

The post-exhaustion legitimacy question is narrower and harder. Can APNIC remain a credible, limited and reviewable settlement layer for an IPv4 economy it no longer supplies in any meaningful quantity? Its authority now depends less on the romance of fair first distribution and more on the discipline of recognition: who is recorded as holding a resource, under what conditions, after what verification, with what remedies if the registry says no, and with what protections against the registry becoming a private economic regulator. That is a smaller role than moral stewardship of the Internet. It is also more economically powerful.

Allocation was frontier administration; recognition is settlement

In the frontier period of IPv4, number administration resembled the orderly distribution of a scarce input at the edge of expansion. The registry's task was to prevent chaotic appropriation, reduce conflict and match requests to demonstrated operational need. The language of conservation, aggregation and responsible stewardship had force because addresses were still available. If the pool existed, the central institutional problem was how to prevent over-assignment without starving real networks. A registry decision allocated something new.

Exhaustion changed the nature of the administrative act. APNIC's final-pool policy and later adjustments left eligible account holders with access only to small quantities from residual space. Recovered space could be redistributed under constrained rules, but it no longer functioned as a growth engine for carriers, cloud platforms, large access networks, hosting providers or fast-growing enterprises. A new operator might receive a modest starting block. An established operator needing substantial IPv4 capacity had to buy, lease, share addresses through CGNAT, redesign services around IPv6 where possible, or postpone growth.

In that world, the registry's scarce product is recognition.

Recognition converts a private agreement into a routable, registrable, certifiable and defensible claim. A buyer may have a contract, but without registry recognition it does not have a fully settled resource position. A seller may have an asset, but without clear records the asset is discounted. A lender or acquirer may see value, but not a clean chain of evidence. A routing counterparty may see an announcement, but not a reliable registration context. An abuse desk may see traffic, but not a trustworthy responsible party. A network engineer may be able to configure equipment, but the organisation may be unable to maintain reverse DNS, create resource certificates or satisfy due diligence in a future transaction.

This does not make IPv4 addresses property in the simple sense that land or shares are property. Their legal treatment differs across jurisdictions, and the registry system was not built as a title office. But economic life is not governed only by formal property labels. Many assets are valuable because a recognition system makes them transferable, financeable and usable. Bank deposits depend on ledgers. Securities accounts depend on settlement systems. Airport slots depend on recognised use rights. IPv4 resources now share part of that logic. They are operational resources whose market value depends on the stability and credibility of registry recognition.

The settlement analogy clarifies the institutional bargain. APNIC does not have to decide the social worth of every transaction. It does have to maintain a ledger that market participants can trust. A delay in recognising a transfer can impose financing costs. A denial can strand capital. An ambiguous documentation request can become a tax on smaller operators. A fee dispute can freeze liquidity if account standing is used too broadly. A hold on records may be necessary in a genuine dispute, but it can operate like a private capital control if the standards are vague. The frontier registry was judged by fairness in allocation. The settlement registry must be judged by predictability, narrowness, proportionality and review.

That is the core transition. APNIC's legitimacy after exhaustion is not the legitimacy of a benevolent distributor. It is the legitimacy of a constrained utility whose decisions sit between private capital, public routing stability and a heterogeneous regional Internet economy.

A region too varied for one simple story

APNIC serves one of the world's most uneven Internet regions. Its service area covers advanced broadband economies, continental-scale markets, island states, small rural providers, mobile-first access networks, financial centres, carrier hubs, university networks, national research networks, hosting companies, cloud platforms, submarine-cable economies and territories whose connectivity depends on a small number of fragile routes. Seven economies are served through National Internet Registries: China, India, Indonesia, Japan, Korea, Taiwan and Viet Nam. Those NIRs work in local languages and domestic administrative contexts while operating within regional policy.

This variety makes post-exhaustion legitimacy unusually demanding. In a smaller and more legally uniform region, standardisation can look almost neutral. In the Asia-Pacific, uniform procedures can have unequal effects. A metropolitan cloud platform, a mobile carrier with in-house counsel and a broker familiar with registry practice may experience transfer recognition as routine compliance. A rural wireless provider, a small hosting company, an island access network or a new entrant operating across borders may experience the same process as a major burden on management time, legal costs and working capital.

The region also contains old and new Internet economies. Some networks hold substantial legacy resources because they arrived early in the IPv4 allocation era. Others grew after the frontier had already closed. Some markets have domestic NIR mediation; others deal directly with APNIC. Some operators can price legal ambiguity into a transaction; others cannot. Some can maintain staff for policy participation; others learn about a rule only when it blocks a transaction. These differences are not peripheral fairness concerns. They determine whether the post-exhaustion registry looks like common infrastructure or a club for repeat players.

APNIC's recognition layer is meant to lower transaction costs across this diversity. A buyer in one economy must be able to understand whether a block from another economy is cleanly registered. A cloud provider serving several jurisdictions must know how transfers and resource certificates will be treated. A small provider must know whether a fee dispute, a document gap or a change of corporate name can put operational continuity at risk. NIR members must know whether local mediation clarifies regional rules or obscures the costs attached to recognition.

Scarcity amplifies all of this. When addresses were allocated from a pool, procedural friction was partly softened by the fact that an approved request produced an operating input. After exhaustion, a buyer may already have negotiated with a seller, arranged financing, planned customer migration and committed engineering resources before registry recognition becomes decisive. The registry's procedure is no longer merely a queue for public resources. It is a condition of private investment.

That makes restraint more important, not less. The temptation in a diverse region is to expand discretion in the name of fairness, security, development, anti-abuse policy or national sensitivity. Some discretion is unavoidable. Fraud exists. Disputes exist. Documentation can be false. Legal obligations may intervene. But open-ended discretion does not fall evenly. Large repeat participants can absorb uncertainty, learn informal practice and hire advisers. Smaller networks pay for uncertainty through delay, worse transaction terms or lost opportunities.

Post-exhaustion legitimacy in the APNIC region therefore requires a paradoxical combination: enough verification to make the ledger credible, and enough limitation to prevent verification from becoming private control over scarce capital.

Address blocks became balance-sheet-like assets

IPv4 scarcity is often described as a temporary discomfort on the way to IPv6. That description is technically neat and economically misleading. A resource can be destined for gradual obsolescence and still remain valuable for decades. Copper loops retained value during fibre transitions. Airport slots remain valuable even as aircraft and scheduling systems improve. Carbon-intensive plants can still be grid assets while cleaner generation expands. IPv4 addresses have acquired similar durability because the Internet is not one centrally upgraded machine. It is a set of interdependent networks, customers, applications, procurement cycles, devices and business models that move at different speeds.

For many operators, IPv4 holdings are not merely a technical input. They are balance-sheet-like assets. They can be sold in distress, leased for recurring revenue, retained as strategic inventory, used to support customer acquisition, or valued in a merger. Holdings can affect enterprise value because they reduce dependence on expensive leases, tight CGNAT ratios or uncertain future purchases. They can reduce churn by improving user experience. They can support hosting, security services, VPNs, remote access, gaming, enterprise networking and other products where public or stable IPv4 still matters. They can lower operational complexity by reducing reliance on translation mechanisms.

The market price of IPv4 is not set by APNIC policy alone. It reflects global demand, available supply, broker activity, transaction size, routing history, reputation, abuse taint, geolocation expectations, legal clarity and the cost of alternatives. Yet registry recognition is a central component of price. A block with clean records, transferable status, usable reverse DNS arrangements and valid resource certification is more liquid than one surrounded by historical ambiguity. Buyers discount uncertainty. Lenders discount uncertainty. Brokers discount uncertainty. Compliance teams discount uncertainty. Scarcity turns registry confidence into economic value.

This is why APNIC's post-exhaustion role should be understood as market infrastructure rather than merely technical administration. When it recognises a transfer, it helps settle a transaction. When it maintains transfer logs, it contributes to price discovery and discipline. When it applies a lock to certain final-pool delegations, it shapes the timing of supply. When it asks a recipient for a use plan, it affects who can convert money into recognised capacity. When it treats a resource as in dispute, it can protect buyers from fraud or freeze an asset until the dispute is resolved.

None of this implies that APNIC should abandon verification. A ledger with no anti-fraud controls is not a settlement layer; it is an invitation to forgery. The value of the address market depends on confidence that the registry will not recognise fraudulent sellers, duplicate claims or unauthorised signatories. But the burden of verification must be visible. A slow transfer imposes financing and opportunity costs. A vague request for additional documentation favours repeat players. A broad account-standing condition can turn an ordinary billing or paperwork problem into a block on liquidity.

The vocabulary matters. If registry decisions affect asset liquidity, they should be discussed in economic terms. A need test is not only a technical condition; it is a filter on market access. A transfer lock is not only a policy safeguard; it is a liquidity restriction. A fee is not only a membership charge; it can be a carrying cost of recognised holdings. A resource dispute is not only an administrative note; it may impair financing, sale or reorganisation. Post-exhaustion legitimacy begins by naming those effects plainly.

Transfers are where settlement power becomes visible

The transfer framework is the most obvious expression of APNIC's new role. Current policy recognises transfers among APNIC account holders, inter-regional transfers where the counterpart registry has a compatible policy, transfers involving historical resources, and movements connected with mergers or acquisitions. The rules set minimum sizes, require resources to be within the relevant managed range, expect the transferor to be the registered holder, exclude resources under dispute, and impose recipient conditions where policy requires them. APNIC maintains public transfer logs.

These conditions define the boundary between a real market and a rumour market. A private contract can move money and obligations between parties, but without registry recognition the economic life of the deal is impaired. The buyer may struggle to convince customers, transit providers, auditors or future counterparties that the resource is cleanly held. The seller may receive less because the buyer prices recognition risk. Brokers must reserve for delay or failure. Cross-border transactions add further uncertainty because another registry's policies and interpretations may matter. The private bargain depends on public-technical settlement.

APNIC's task in this setting is not to decide whether the buyer is socially deserving. Nor is it to maximise transfer volume at any cost. It is to maintain a credible settlement boundary: verify that the resource is eligible, the holder is authorised, the recipient meets published conditions, the counterpart registry can recognise the movement, and the public record will be updated accurately. That is enough power to protect the market. It is also enough power to distort the market if used expansively.

Need assessment is the most delicate example. In the allocation era, need assessment had an obvious justification: the registry was distributing new resources and had to prevent over-consumption. In a transfer market, the resource is being acquired from another holder, often for a substantial price. A recipient's willingness to pay is itself evidence that the resource has operational or strategic value. But pure willingness to pay is not the whole answer, because a registry may reasonably wish to discourage laundering of final-pool allocations, policy evasion or purely speculative receipt where the policy still requires operational use.

The legitimate solution is not to pretend the transfer market is an allocation queue. It is to make any need test specific, time-bounded and predictable. A requirement to show intended use within a defined period may be defensible if it prevents obvious warehousing and keeps scarce space connected to operational networks. It loses legitimacy if it becomes an invitation to judge business models. A cloud platform's plan, a carrier's CGNAT relief plan, a hosting company's customer-growth plan and a rural ISP's incremental expansion plan will not look alike. The registry's test should be whether the plan is plausible and relevant to the published policy, not whether staff would choose the same commercial strategy.

Inter-regional transfers raise the same issue in a different form. Compatibility protects the global recognition set from conflicting claims and policy arbitrage. But compatibility is also a liquidity boundary. If two regions interpret eligibility, need, sanctions, historical holdings or transfer locks differently, a block may be easier to move in one direction than another. Repeat participants learn those frictions. Occasional participants discover them through delay.

APNIC can strengthen legitimacy by treating transfer administration as a settlement service whose performance should be observable in aggregate. It need not disclose confidential documents or commercial terms. It can still report typical timelines, common causes of delay, denial categories, inter-regional coordination issues and the treatment of historical-resource cases. That would not turn the registry into a broker. It would make the boundary of recognition more legible.

The post-exhaustion registry is not legitimate because every transaction is approved. It is legitimate when a refusal is intelligible, tied to a published rule, proportionate to the risk and capable of review.

Leasing is the shadow allocation economy

If transfers are the formal market, leasing is the pressure valve. Many networks need IPv4 capacity but cannot or do not want to purchase blocks outright. A small access provider may need enough public addresses to reduce CGNAT pressure. A hosting company may need capacity for customers whose requirements may not last long enough to justify a purchase. A security, proxy or content-service provider may need distributed address space. An enterprise may need continuity during migration. Leasing turns a capital purchase into an operating expense and lets holders monetise space without permanent sale.

Leasing also exposes the limits of registry visibility. A lease may not change the registered holder. The lessee may announce the block, use delegated reverse DNS, depend on a lessor's routing arrangement, or operate under a more complex managed-service structure. Some leases are transparent and operationally clean. Others create confusion about abuse responsibility, routing authority, geolocation, reputation and renewal risk. The registry may not be a party to the commercial agreement, yet its records remain the reference point for outsiders trying to understand who is responsible.

The danger is not leasing itself. In a scarcity economy, leasing is an efficient adaptation. It avoids over-investment in an asset whose long-run value may decline as IPv6 adoption improves. It helps smaller operators preserve cash. It can bring idle holdings into productive use. The danger is a shadow allocation economy in which operational control detaches from public accountability. If Whois lists one organisation while another controls routing, customer assignment and abuse handling, the ledger's public value weakens. If leases are used to disguise de facto transfers that would not satisfy policy, the settlement boundary is undermined.

APNIC's legitimacy depends on proportionate recognition of this reality. Treating every lease as suspicious would punish ordinary commercial adaptation and push arrangements further into opacity. Ignoring leases entirely would also be a mistake, because operational control matters for abuse response, routing hygiene and market confidence. The useful questions are narrow: which registry records identify the holder, which contacts should reflect actual use, how should reverse DNS and resource certification be handled, when does a lease-like arrangement become a transfer-like change of control, and what evidence is needed when responsibility is contested?

The distinction between a ledger and a gatekeeper is practical here. A ledger-first institution asks what information must be accurate so networks can function and disputes can be resolved. A gatekeeping institution asks whether it approves of the commercial form. APNIC's authority is stronger when it focuses on the first question. If leasing creates routing or accountability risk, the response should be better responsibility signals, clearer operational contacts and certification guidance, not a licensing regime for business models.

This matters especially for smaller networks. They lease not because they have ignored IPv6 or failed to plan, but because buying scarce addresses can absorb capital needed for towers, fibre, radios, backhaul, routers and staff. If registry practice makes leasing precarious without making purchase realistic, the result is not conservation. It is concentration. Scarce IPv4 then moves toward those with capital and compliance capacity, not necessarily those with the most urgent operational need.

The registry cannot make IPv4 cheap. It can, however, avoid adding unnecessary uncertainty to the difference between holding, leasing, routing and responsibility. In a permanent scarcity economy, clarity is itself a form of access policy.

NIR mediation creates a two-layer legitimacy test

APNIC's National Internet Registry structure is one of the region's distinctive institutional arrangements. The seven NIRs serve large or complex domestic Internet communities in local languages and local administrative settings. They are not branches of APNIC, but they operate within the regional policy system and maintain their own membership and fee arrangements. In practice, many organisations choose between direct APNIC membership and local NIR membership. The existing NIR map is now effectively fixed, since new applications are no longer open.

The arrangement has obvious benefits. Local-language service reduces friction. Domestic institutional knowledge matters in economies with many networks and distinctive legal or administrative expectations. For a smaller operator that would struggle with English-language forms, overseas billing and unfamiliar procedures, an NIR can be a bridge into the regional recognition system.

Post-exhaustion changes the economics of that bridge. When the central problem was initial distribution, local mediation helped operators request resources. When the central problem becomes recognition of scarce assets, mediation can affect liquidity, fee transparency, dispute handling, account standing and policy voice. A member dealing through an NIR may experience APNIC policy through a national fee schedule, national documentation practice and national service culture. If the pass-through is clear, the arrangement can work well. If it is opaque, the operator may not know which cost comes from regional policy, which from the NIR, which from local law and which from administrative habit.

The risk is a two-layer legitimacy discount. APNIC may assume regional rules are clear because they are published. The NIR may assume local members understand domestic implementation because it is familiar locally. The member may find itself caught between them. A transfer delay may originate at one layer and be explained by the other. A fee dispute may raise questions about whether national account standing affects regional recognition. A documentation request may be framed as regional necessity when it is partly local practice. The result can be uncertainty over who is responsible for the friction.

This is not an argument against NIRs. It is an argument for treating them as part of the post-exhaustion settlement system, not merely as a service convenience. The more valuable IPv4 recognition becomes, the more important it is that NIR members can see how regional policy is applied, how fees pass through, how appeals work and how their interests are represented in policy deliberation. Local mediation should reduce transaction costs, not hide them.

The fixed NIR map adds a further question. Existing NIR economies retain a domestic mediation channel; other economies do not get that option. The reasons may be administrative, historical and practical, but the result is a region divided between mediated and direct relationships. If NIR service lowers barriers, non-NIR economies may lack that advantage. If NIR mediation adds opacity, NIR members may bear that cost. Either way, the distinction belongs in the analysis of legitimacy.

APNIC can make the two-layer system more credible by reporting the NIR dimension as a settlement issue. How long do NIR-mediated transfers take compared with direct APNIC transfers? How are fees and service charges explained to end members? How are delays or denials appealed? Are NIR members visible in debates over policies that affect address liquidity? Such questions may sound procedural. In a scarce-asset market, they affect price, access and trust.

In the allocation era, NIRs helped make the registry accessible. In the settlement era, they must also make recognition transparent.

Small operators and the cost of being legible

Post-exhaustion policy often looks neutral until one asks who can afford to comply with it. A large carrier or cloud platform can pay for legal opinions, broker due diligence, accounting advice, corporate documentation and staff time. It can keep people watching policy lists, attend meetings, maintain relationships with registry staff and respond quickly to additional requests. A small operator may have one engineer who also handles customer escalations, procurement, routing, billing and compliance. For that operator, the difference between a clear checklist and an open-ended request is not cosmetic. It is the difference between a manageable transaction and a stalled expansion.

APNIC's region includes many such institutionally thin networks: island access providers, rural wireless operators, local data centres, small enterprise-service providers, community networks and start-ups serving specialised markets. They may not be poor in an absolute sense, but they are fragile in administrative capacity. IPv4 scarcity hits them twice. First, they must pay market prices or lease addresses because the registry can supply only small residual amounts. Second, they must satisfy the same recognition procedures as better-resourced participants when they buy, lease, reorganise or defend address holdings.

The economics are unforgiving. Buying a block ties up capital that might otherwise fund towers, fibre, backhaul, radios, routers or staff. Leasing reduces upfront cost but creates renewal risk and may complicate responsibility records. CGNAT reduces address demand but can worsen user experience, complicate attribution, increase logging requirements and create support burdens. IPv6 deployment is necessary, but it may not remove customer demand for IPv4 reachability. The operator's problem is not ignorance of the future. It is the cash-flow cost of surviving the transition.

Registry recognition can reduce or intensify that cost. Clear transfer pre-checks, predictable documentation standards, timely processing, transparent fees and usable explanations lower the risk premium. Ambiguous standards raise it. If a broker or seller believes a small buyer may struggle with approval, the buyer pays in price, terms or lost opportunity. If a lender cannot assess whether recognition will complete, it withholds financing or charges more. If a leased block lacks clear operational contacts, the small network may face abuse and routing complications it cannot easily absorb.

The phrase "documentation burden" sounds like a complaint about paperwork. In a scarce-asset market, it is a distributional mechanism. Burden affects who can convert need into recognised resources, sell in distress, merge with a stronger partner, prove continuity after a corporate name change, or defend a holding after a founder exits. If the burden is necessary to prevent fraud, it should be defended and explained. If it exists because forms accumulated over time, it should be simplified.

An economically legitimate APNIC should therefore treat small-operator experience as an indicator of registry health. Not as charity, and not as a reason to weaken anti-fraud controls, but as a test of proportionality. A settlement layer that works only for large repeat players is not neutral. It subsidises scale.

Policy participation has the same problem. In principle, APNIC's policy process is open. In practice, meaningful participation requires attention, language confidence, time, familiarity with past debates and the ability to see how a sentence in policy will affect future transactions. Large operators and specialised advisers can treat this as a recurring function. Small operators often participate only when harmed, by which time the rule may already be embedded. Silence should not be mistaken for consent when the silent parties are those least able to follow the process.

Post-exhaustion legitimacy therefore requires more than open doors. It requires active testing of who can walk through them.

IPv6 is a partial exit, not an escape from the ledger

No serious analysis of post-exhaustion legitimacy can ignore IPv6. The long-run technical answer to IPv4 scarcity is a protocol with a vastly larger address space. APNIC has invested in measurement, training and regional encouragement. Many Asia-Pacific economies have made real progress, and some large mobile and access networks have achieved high IPv6 adoption. For new architectures, IPv6 can reduce exposure to the address market, simplify future growth and free networks from some of the compromises of shared IPv4.

But IPv6 progress does not abolish IPv4's economic power. It changes the shape of dependence.

The reason is interoperability. IPv6-only users still need to reach IPv4-only services, and IPv4-only users remain widespread across access networks, enterprises, applications and devices. Dual-stack operation requires maintaining both worlds. Translation mechanisms help but add cost, complexity and operational risk. Enterprise procurement cycles are slow. Customer-premises equipment, industrial systems, security appliances, management platforms, payment systems, gaming services, remote-access tools and embedded applications can remain IPv4-dependent long after a strategic plan declares IPv6 the future. A network may deploy IPv6 extensively and still need IPv4 for customer reachability, inbound services, hosting, logs, abuse attribution, partner connectivity or legacy customers.

CGNAT illustrates the point. Carrier-grade NAT lets many users share fewer public IPv4 addresses. It extends the life of IPv4, but it does not make public addresses worthless. It imposes equipment and operational costs, complicates attribution, can degrade some applications and often requires careful logging to satisfy legal or abuse-handling demands. The more a network relies on CGNAT, the more it values enough public IPv4 capacity to keep sharing ratios tolerable and support costs manageable. Scarcity remains even as engineering workarounds improve.

IPv6 also does not eliminate the market value of existing IPv4 holdings because transition is asynchronous. A block's value is determined by economically significant counterparties, not by the most advanced networks alone. If enough customers, suppliers, applications or jurisdictions still require IPv4, the address remains valuable. If global platforms support IPv6 but local enterprise systems do not, access networks still need IPv4. If adoption rises in one economy but lags in another, cross-border operators still maintain IPv4 capacity for reachability.

This is why APNIC's legitimacy cannot be reduced to IPv6 advocacy. A registry that says "deploy IPv6" is right about the long-run architecture. It is not answering the near-term capital problem. Operators must finance networks, serve customers, manage abuse, maintain legacy reachability and survive outages today. The IPv4 ledger remains economically powerful precisely because IPv6 is a partial exit: it reduces future dependence where adoption is high, but it does not release every operator from the installed base at the same time.

There is also a political risk in overusing IPv6 rhetoric. If a registry treats IPv4 concerns as legacy complaints rather than current market facts, it may under-invest in transfer transparency, historical-record cleanup, lease accountability, appeal mechanisms and small-operator support. The better posture is honest dualism. IPv6 is the strategic direction; IPv4 remains a scarce settlement asset during a long and uneven transition.

The irony is that successful IPv6 deployment can increase the need for disciplined IPv4 stewardship during the transition. As some networks reduce dependence, they may sell or lease holdings. Others may buy or lease to bridge slower customer bases. A declining asset can still require an excellent registry because decline is not smooth.

Operational recognition is more than the transfer table

The post-exhaustion ledger is not only a list of who bought what. It is a bundle of operational recognition services. Reverse DNS delegations help map addresses to names and remain important for mail, logging, diagnostics and reputation systems. Resource Public Key Infrastructure allows holders to create cryptographic attestations that a given autonomous system is authorised to originate a prefix. Whois and related registration data provide contacts, holder records and administrative context. Routing registries and public statistics add further layers of evidence. None of these systems is a perfect truth machine. Together they create the practical recognition environment in which address holdings are used.

This is why account actions can have market consequences beyond formal transfer approval. If an operator cannot maintain reverse DNS, update contacts, manage resource certificates or keep records current, the value of the resource declines. If a dispute freezes some functions but not others, the operator's commercial position changes. If a resource certificate cannot be maintained, counterparties may treat routes differently. If abuse contacts are stale because the actual user is a lessee while the registered holder remains the lessor, operational accountability suffers.

For APNIC, these services create both authority and duty. Authority, because networks, security teams, buyers and public institutions rely on registry signals. Duty, because those signals should be insulated from unrelated pressure. A transfer dispute should not casually become a routing-security disruption. A billing dispute should not casually become a reverse-DNS outage. A policy disagreement should not casually become a certification problem. There will be hard cases, especially where fraud, hijacking or unlawful conduct is alleged. But the default institutional design should separate recognition of existing operational status from adjudication of unrelated claims.

The separation is especially important for small and NIR-mediated networks. Large operators can sometimes route around registry friction through engineering scale and commercial relationships. Smaller networks may depend more directly on ordinary registry functions. If a service interruption harms mail delivery, route validation or abuse response, the commercial damage can be immediate. Service continuity is therefore not only a technical uptime metric. It is part of legitimacy.

Operational recognition also shows why the registry's role must remain narrow. APNIC may be asked to solve abuse, fraud, sanctions exposure, speculative holding, IPv6 delay, market concentration, national-security concerns and competition policy. It has information relevant to some of these problems, but it is not designed to be a general regulator. Its comparative advantage is maintaining accurate, stable and reviewable number-resource records. If it tries to make every registry service a lever for broader economic policy, the ledger will lose the neutrality that gives it value.

Resource certification sharpens the point. RPKI can improve routing security by allowing networks to validate authorisation. But a certification system tied to registry standing also creates power. If certification can be interrupted for reasons unrelated to resource legitimacy or routing risk, it becomes a pressure mechanism. A disciplined registry should distinguish between actions necessary to preserve the integrity of certification and actions that merely use certification as leverage. The same logic applies to reverse DNS and contact maintenance.

In post-exhaustion economics, operational humility is not weakness. It is the source of authority.

Historical records are old paperwork with new capital value

Exhaustion increased the value of history. Blocks allocated before modern membership structures, inherited through early registry transitions, held by universities or enterprises, recorded under old names, or associated with organisations that have merged or disappeared can now be worth substantial sums. Historical resources are no longer just a database-cleanup problem. They are a recognition problem with market consequences.

APNIC policy recognises transfers of historical Internet resources and requires verification of legitimate holding where the resource is not already under a current account. The registry does not need to become a judge of every private bargain. Its role is to decide what evidence is sufficient to update records and what questions belong to external legal resolution. That restraint is important. A registry that reviews commercial price, business wisdom or moral desert would exceed its settlement function.

Historical holdings are nevertheless hard. Corporate entities disappear. Names change. Government agencies reorganise. University departments merge. Documentation is lost. Early records may not contain the administrative detail now expected. In a low-value environment, such ambiguity could sit unresolved. In a high-value environment, ambiguity attracts conflict. A dormant block may suddenly become the object of a sale, a creditor claim, an internal dispute or a third-party challenge.

The legitimacy test is how APNIC handles ambiguity without creating arbitrary windfalls. If it recognises claims too easily, it risks enabling capture by parties with weak authority. If it requires impossible documentation, it traps legitimate holders in dead records. If it treats every historical uncertainty as a reason for indefinite hold, it turns old space into unusable capital. None of those outcomes supports confidence in the ledger.

A sound approach distinguishes levels of evidence. Some cases will have clear corporate succession documents, historical correspondence, prior registry interactions, consistent routing history and current operational use. Others will have partial evidence that requires corroboration. Some will be genuinely contested and should be resolved by parties outside the registry before records change. APNIC should be clear about what it can verify, what it cannot decide and what evidence it needs for each kind of case.

Historical resources also test the moral language of conservation. Some early holders received large blocks under practices that would not be acceptable today. That history can look unfair to later entrants. But confiscatory rhetoric would damage confidence in the ledger and invite political struggle over past allocations. The better post-exhaustion approach is to make transfers possible, records accurate, responsibility clear and future policy predictable. Old-allocation regret should not become present-recognition improvisation.

This does not mean ignoring abuse or illegitimate capture. It means separating two questions: whether the original allocation would satisfy today's standards, and whether the current claimant has sufficient authority to be recognised. APNIC earns legitimacy by clarifying claims, not by re-litigating the entire history of the Internet.

Fees, reserves and the price of standing

Post-exhaustion legitimacy has a fiscal dimension. A registry must be funded. It needs staff, security systems, legal capacity, member support, data services, meetings, measurement, training and reserves for shocks. A zero-cost registry would not be resilient. The question is not whether APNIC should charge fees. The question is what costs the recognition utility should recover, how the burden is distributed and how account standing interacts with scarce assets.

In the allocation era, membership fees were principally the cost of participating in the registry system and receiving services. In the post-exhaustion market, they also become part of the carrying cost of recognised IPv4 holdings. If failure to pay, a dispute over fees or a disagreement over membership status can affect transfers, record updates, reverse DNS, certification or other recognition functions, the fiscal system has liquidity consequences. A fee is then not merely a charge. It is attached to the ability to keep an asset usable.

That makes proportionality essential. Fees must be paid and agreements respected. Free riding would undermine the utility. But essential recognition functions should not be restricted more broadly than necessary to protect fairness and the registry's integrity. Ordinary arrears, uncertain identity, suspected fraud and legal prohibition are different conditions. They should not be collapsed into one blunt category of bad standing. A small operator in temporary financial stress should not face the same consequences as a party attempting to transfer resources with false authority.

Reserves also require economic clarity. Reserves protect service continuity, legal resilience and operational stability. They can also become institutional ballast if they grow without a defined purpose. In a post-exhaustion market, members are entitled to understand whether reserves insure the settlement utility or fund broader institutional expansion. Training, research, measurement and community development may be valuable, but the legitimacy of funding them through charges tied to scarce recognised resources should be argued openly rather than assumed.

The NIR layer complicates the fiscal picture. Local fee schedules, service charges and pass-through arrangements may shape the actual cost of standing for end members. A direct APNIC member and an NIR member can both be subject to regional policy, but experience different fee structures and service paths. If those differences affect transfer timing, account standing or access to operational services, they are part of post-exhaustion legitimacy. Transparency is the remedy. Members should be able to see what they are paying for, which layer sets the charge and how disputes are handled.

The registry should avoid financing broad ambitions through the leverage of scarce IPv4 recognition. It should be able to explain which fees recover direct service costs, which support general registry operations, which fund community functions and which build reserves. It should show why reserves are sufficient but not excessive. It should distinguish the cost of settlement from the cost of mission expansion.

Legitimacy is not produced by being cheap. It is produced by being predictably priced for a narrow and necessary function.

Policy participation after abundance

The policy process changes when the pool is no longer the main object of policy. During the allocation era, policy determined access to future assignments. After exhaustion, policy determines market conditions for existing holdings: transfer eligibility, documentation standards, locks, recipient tests, historical-resource treatment, NIR implementation, public logs, account conditions and operational services. The stakes are more financial than they once were.

Consensus processes have real virtues. They avoid top-down command, let technical communities adapt rules to operational reality and give participants a way to resist capture by governments or private firms. APNIC's policy process is formally open, with public discussion and documented decision-making. But post-exhaustion consensus faces a classic political-economy problem: concentrated interests participate more consistently than diffuse ones.

This does not make consensus illegitimate. It means consensus should be interpreted carefully. Silence from small operators may mean agreement. It may also mean lack of time, language barriers, uncertainty about consequences or inability to see the effect of a proposal until it becomes operational. A change in transfer requirements may look technical to participants in the room while altering the cost of capital for those outside it. A documentation rule may sound like common sense to advisers and large networks while imposing real costs on occasional participants.

Post-exhaustion policy legitimacy therefore requires affected-party analysis. Who benefits from a proposed rule? Who bears documentation costs? Does it favour holders over buyers, large networks over small, direct APNIC members over NIR members, or domestic transactions over cross-border ones? What behaviour is the rule trying to prevent? Is the restriction time-limited or subject to review? How will APNIC know whether it worked? These are not academic questions. They are the economic content of policy once addresses have become scarce assets.

The process need not become paralysed. The registry cannot run a full regulatory impact assessment for every adjustment, and address markets will not wait for perfect information. But it can ask economic questions before adopting rules that affect liquidity. It can publish implementation experience after a rule takes effect. It can make proposals easier to understand for members who do not live inside policy forums. It can avoid mistaking procedural openness for actual participation.

NIR members deserve particular attention. If national mediation reduces language barriers, it should also transmit concerns upward. If NIR members are affected by regional transfer rules, their practical experience should be visible in regional debate. If local fee schedules or documentation practices alter the effect of regional policy, those differences should be disclosed. Otherwise, the policy process sees the NIR as an institutional participant while the economic burden falls on many end members.

Appealability is the final component. A policy process can be open and still produce individual mistakes. Transfer denials, documentation disputes, historical-resource claims and account-standing actions need review channels that are accessible, timely and reasoned. Appeals should provide written reasons, allow correction of errors and create institutional memory that improves future decisions. After abundance, legitimacy is not only who can speak before a rule is made. It is also who can challenge how the rule is applied.

The service-continuity firewall

A registry's legitimacy is often invisible until governance fails. Then every quiet assumption becomes a market risk. Who can sign? Who directs staff? Are transfers still processed? Are reverse DNS and resource certification still reliable? Are board disputes separated from registry operations? Are legal costs consuming reserves? Are members informed? These questions sound remote in normal times. In a post-exhaustion market, they belong in the basic economics of confidence.

The Asia-Pacific registry has not needed to be defined by a severe governance collapse. That is not a reason to ignore the lesson. Scarce IPv4 recognition makes service continuity more valuable. Ordinary registry functions should be protected from governance disputes, leadership transitions, election fights, litigation, budget arguments and political pressure. The more market value depends on registry recognition, the more important it is that the ledger operate through institutional stress.

The firewall has several layers. Technical systems must be resilient and documented. Staff authority for routine operations should not depend on contested political direction. Financial reserves should be sized for continuity, not institutional vanity. Board and executive roles should be clear. Conflict-of-interest controls should reflect the fact that policy and recognition decisions can affect private asset values. Emergency powers should be defined and time-limited. Members should know which services continue under stress and which decisions require extraordinary review.

Service continuity should also be separated from moral judgement about members. If a holder is controversial, under investigation or involved in unrelated disputes, the registry should still preserve accurate records and essential services unless a defined legal or policy reason requires restriction. The ledger's value comes from recording reality, not rewarding virtue. A registry that uses operational services to punish disfavoured networks will quickly lose legitimacy as a neutral recognition layer.

The firewall must extend to NIR relationships. If an NIR faces domestic governance stress, members should understand how regional recognition continues. If APNIC itself faces stress, NIRs need clear continuity expectations. A member should not discover during a crisis that no one can explain who is responsible for maintaining records, certification, reverse DNS or transfer processing.

Legal obligations are part of the same problem. APNIC operates under Australian law and within a region affected by geopolitical tension, sanctions regimes, law-enforcement requests and state interest in network infrastructure. It cannot ignore lawful constraints. But post-exhaustion legitimacy requires legal compliance to be as narrow and transparent as possible. Essential registry functions should not be interrupted more broadly than law and policy require. If a resource holder is subject to a legal restriction, the registry should act within defined bounds and preserve network stability where possible. If a compliance concern is unrelated to resource legitimacy, it should not become an all-purpose lever over transfers, reverse DNS or certification.

The capital-control analogy is imperfect but useful. A state can restrict currency movement; a registry can restrict recognition movement. The registry's power is narrower, but in an IPv4 market it can still freeze value. An organisation unable to transfer a block may be unable to raise cash, complete a sale, exit a market, restructure debt or merge. If the reason is fraud prevention or legal necessity, a hold may be legitimate. If the reason is vague discomfort or bargaining leverage, legitimacy decays quickly.

The legitimate registry is not the one that always keeps services unchanged. It is the one that can explain precisely why a restriction is necessary, how long it may last, what function it affects and how the decision can be reviewed.

What APNIC should be judged on now

The central risk after exhaustion is not that APNIC will fail to distribute abundant IPv4. That era is over. The risk is that settlement power expands without the economic discipline appropriate to a constrained utility. Several practical tests follow from the legitimacy transition.

Transfer processing should be treated as market infrastructure. Aggregate timelines, delay reasons, denial categories, inter-regional coordination problems, historical-resource cases and differences between direct and NIR-mediated pathways should be visible. Confidential documents and commercial terms can remain protected. What matters is whether members can see how the recognition system performs.

Documentation standards should be predictable. Applicants should know in advance what evidence is normally required for corporate authority, merger and acquisition cases, historical-resource claims, recipient use plans, account standing and dispute clearance. Additional evidence may be necessary in particular cases, but the reason should be stated. A checklist cannot solve every dispute, yet it can reduce arbitrary burden.

Account standing should not become an all-purpose lever. Fees must be paid, contacts must be accurate and agreements must be respected. But essential recognition functions should not be restricted more broadly than necessary. The registry should distinguish ordinary arrears from identity uncertainty, fraud risk, legal prohibition and conduct that directly threatens registry integrity.

NIR pass-through transparency should be treated as regional legitimacy, not local housekeeping. Local fees, service timelines, appeal routes and policy communication should be clear enough for NIR members to understand the economic conditions attached to recognition. A two-layer structure is defensible only if the layers reduce friction rather than obscure responsibility.

IPv6 advocacy should continue without becoming an excuse to neglect IPv4 market duties. The region needs IPv6 deployment, but the registry still has to maintain a credible IPv4 settlement layer during the transition. CGNAT, leasing, dual stack and legacy reachability are not embarrassing leftovers. They are part of the economic reality the ledger must serve.

Policy participation should be assessed by affected-party reality, not only formal openness. If a rule affects small operators, island networks, rural providers, NIR members, late entrants or cross-border buyers, their likely costs should be discussed before adoption. Consensus is more credible when silence is not automatically treated as consent.

Conflict controls should reflect the asset value now attached to registry decisions. Disclosure, recusal where necessary and confidence in staff neutrality matter more when a policy or recognition decision can move value. The registry should not be run as if it were merely distributing forms.

Service continuity should be protected from governance and political stress. Transfers, registration data, reverse DNS, resource certification and member support are the recognition utility on which the market depends. A registry that remains boring under stress does more for regional trust than one that expands its mission in calmer times.

These tests point away from interventionism. APNIC's authority is safer when it is narrower, better measured and easier to challenge.

The constrained bargain

APNIC's post-exhaustion legitimacy rests on a bargain that is easy to state and hard to maintain. The region accepts that the registry will not restore IPv4 abundance; that IPv6 is the strategic direction but not an immediate economic escape; and that transfers, leases, historical holdings, CGNAT and address-sharing mechanisms will remain part of the operating economy for years. In return, the registry must provide a credible recognition layer: accurate records, predictable transfers, limited verification, transparent fees, operational continuity, reviewable denials, NIR clarity and policy processes that understand economic consequences.

That bargain differs from the allocator bargain. The allocator claimed authority because it distributed new resources fairly. The post-exhaustion steward earns authority because it prevents the recognition system from becoming either chaotic or overbearing. It must be strong enough to reject fraud, preserve the integrity of records, coordinate with other registries and protect routing-related trust. It must be modest enough not to become a moral allocator after the allocative resource is gone.

The distinction matters because scarcity invites mission creep. When a resource is valuable, every stakeholder wants the registry to solve a preferred problem: stop speculation, punish abuse, accelerate IPv6, help small networks, restrain large platforms, satisfy governments, police leasing or shape regional development. Some of these goals are legitimate public concerns. Most exceed the comparative advantage of a number registry. APNIC can support better information, stable records, fair procedures and secure routing signals. It should be cautious about becoming a private economic regulator for the IPv4 market.

The Asia-Pacific region needs the registry precisely because the region is diverse, unequal and interconnected. A common ledger lowers transaction costs across languages, legal systems, network types and market scales. It helps a small operator prove what it holds, a buyer understand what it is buying, a network validate routing authority, abuse teams find contacts and NIR-mediated communities connect local service to regional recognition. It helps the transition to IPv6 proceed without pretending IPv4 dependence has vanished.

But the ledger is useful only if it remains trusted. Trust after exhaustion is built by restraint. The registry should verify, not moralise; record, not command; settle, not allocate winners in a market it no longer supplies. It should make denials understandable and correctable, keep fees tied to the cost of a necessary utility, protect essential services from institutional politics, and let the market move addresses toward use while guarding against fraud and conflicting claims.

The frontier phase of IPv4 gave APNIC authority through distribution. The settlement phase gives it authority through disciplined recognition. APNIC earns authority by being a constrained settlement layer, not by becoming a moral allocator or private economic regulator. If it remembers that bargain, post-exhaustion scarcity can be governed without turning the registry into a gatekeeper over private capital. If it forgets, the region's scarcest resource will be confidence in the institution that records the ledger.