Summary

  • New-entrant disadvantage in the APNIC region is not simply that small or young networks face many costs. It is that several critical proofs must be assembled before revenue is secure: IPv4 supply, registry recognition, routing evidence, upstream confidence, abuse-contact readiness and customer credibility.
  • APNIC's current scarcity environment makes the distinction sharper. A new or existing member can receive only a small IPv4 delegation from APNIC's remaining pool; anything larger pushes the entrant toward transfers, leasing, cloud bridges or shared-address architecture before the business has proved itself.
  • The registry's legitimate pro-entry role is to keep resource records, transfers, routing-adjacent evidence and reviewable timing predictable. It should not become an investment committee, price regulator or gatekeeper over commercial models, because discretion tends to increase the very disadvantage it claims to manage.

The first customer asks for proof

The founder has not yet failed. That is the problem. A failed network at least leaves records. A mature incumbent leaves years of them: assignments, upstream letters, abuse-contact history, reverse-DNS patterns, route-filter experience, bank references, customer invoices, maintenance habits, old tickets and staff who know which explanations persuade which counterparty. The new network has a rack quote, a transit offer, a lawyer's company papers, a licence where one is required, a modest subscriber forecast and perhaps a first enterprise customer who wants an answer before signing. The customer asks a simple question: what public addresses will my service use, and will my traffic be accepted?

At that moment the business stops being a pitch deck and becomes an evidence problem. The founder can order routers before revenue. The founder can rent space before revenue. The founder can hire an engineer before revenue. But public IPv4 cannot be conjured from optimism. In the APNIC region, the remaining direct IPv4 path is deliberately small. New and existing members can still receive a maximum of a /23, or 512 addresses, from APNIC's remaining policy-limited supply. If the network needs more, the practical answer is not a larger fresh allocation; it is a transfer, a lease, a cloud design, carrier-grade NAT, or some combination of all four.

Each option has a timing problem. A transfer needs a source, price discovery, legal authority, account work, documentation and registry processing. A lease may solve the first month but raise questions about continuity, routing authority and address reputation. A cloud bridge can move early services online but may put public identity in the hands of a platform. Shared-address architecture can conserve scarce IPv4 but may weaken the product when customers need inbound reachability, static allowlists, clean reputation or precise attribution. None of these options is fatal by itself. Together, they mean a new entrant must finance credibility before the market has financed the entrant.

This is the economic mechanism behind new-entrant disadvantage. It is not a morality tale about heroic start-ups or obstructive incumbents. It is a sequence of proof costs that arrive before scale. The incumbent's advantage is not only that it has more addresses. It has old addresses whose history is already believed. It has upstreams that already accept its announcements. It has customers who have already built it into procurement files. It has bankable continuity because the market has seen it operate. A young network must buy or assemble the same confidence under scarcity.

APNIC sits in the middle of this story, but not as the author of every cost. The registry did not create all customer conservatism, all upstream caution, all cloud pricing, all IPv4 market value or all enterprise procurement habits. Its importance is narrower and more structural. APNIC is the public record through which scarce number resources become legible. Transfer recognition, Whois accuracy, routing-registry links, RPKI readiness, resource-holder status, NIR relationships and service timing can reduce uncertainty or add to it. In a scarce market, that difference decides who can enter cheaply and who must enter expensively.

Newness is expensive because credibility is cumulative

Markets often describe entry barriers as if they were a fence. For address-dependent networks, the better image is a staircase. Each step is individually defensible. A seller wants to know that the buyer can close. APNIC wants evidence to update the registry accurately. An upstream wants assurance that the announcing network controls the prefix. A customer wants static reachability, abuse response and continuity. A bank wants assets that can be explained. A public buyer wants low operational risk. None of these demands is irrational. Their sequence, however, favours incumbents because incumbents have already paid the fixed cost of being believed.

IPv4 scarcity makes the staircase steeper. In the allocation era, a new provider could often treat address space as an administrative input. It still needed a plan and records, but the resource was not priced like productive capital. After exhaustion, addresses became both operational inputs and market assets. A new network that needs more than a small APNIC delegation must secure capital before the customer base is large enough to absorb the cost. That capital may be cash paid to a seller, recurring lease cost, larger cloud bills, support labour for shared-address architecture, or margin surrendered to an upstream that bundles public addresses with transit and hosting.

The incumbent's address stock, by contrast, behaves like inherited infrastructure. It may sit on the balance sheet indirectly or not at all, but it gives the company choices. It can reserve a clean block for a new customer. It can allocate public IPv4 only to high-margin services. It can lease out surplus. It can sell when the price is attractive. It can avoid a bad transfer market. It can run CGNAT with deeper safety stock. It can meet a public tender without first finding a seller. In economic terms, the incumbent has optionality. The entrant has a funding round disguised as network setup.

This is why official neutrality does not eliminate market bias. A rule can apply equally to every applicant and still preserve unequal starting positions. The rule may say that a member can get only a /23 from remaining APNIC space. The effect is not equal if one firm already holds a /16 and another has none. The rule may say that transfers require supporting information. The effect is not equal if one firm has a legal department, old invoices, staff who have closed transfers before and sellers who trust it, while the entrant is assembling its first transaction file. The rule may say that routing records are available to all. The effect is not equal if one firm's routes have been accepted for years and the entrant's first announcement triggers caution.

None of this means APNIC should compensate entrants by becoming a planner. That would be the wrong conclusion. The registry cannot know which new business model deserves scarce IPv4. It cannot inspect future revenue better than investors, customers or lenders. Its pro-entry responsibility is to keep the common record thin, predictable, fast and reviewable, so the market's necessary caution does not receive an extra layer of institutional friction.

The /23 ceiling changes what early growth means

The APNIC exhaustion page is a blunt exhibit. New and existing APNIC members can still receive IPv4, but the maximum is a /23. If more is needed, members are directed toward transfers. This is a rational conservation design in one sense: it preserves a small onboarding path for many networks. But it also changes the economics of early growth. A /23 is not a growth strategy for many access, hosting, enterprise-service or cloud-adjacent businesses. It is a starter ration.

The starter ration matters because it can make first connectivity possible. A new access provider can test systems, serve a limited customer base, run management services, support early business customers or satisfy a transitional need. Yet the ration also creates a cliff. The firm must decide quickly whether to compress demand behind shared IPv4, buy or lease more, steer customers into IPv6 where possible, rely on upstream-hosted services, or redesign the product around platform-controlled public identity. That decision often comes before the revenue curve is stable.

The incumbent sees the same /23 rule differently. It may not need the remaining pool at all. Its old holdings let it treat scarcity as portfolio management rather than launch risk. It can ration from inside, use old blocks whose reputation it understands, allocate by customer value and make transfer decisions when market timing is favourable. A new entrant cannot wait in the same way. It may need addresses because the first customer demands them. It may need addresses because an upstream contract assumes clean origin. It may need addresses because a regulator, enterprise buyer or bank service still treats public IPv4 as normal infrastructure.

That difference turns timing into capital. A new entrant that must buy early pays not only the address price but the uncertainty premium of needing to close now. It has less bargaining power with sellers. It has fewer references for brokers or counterparties. It has less operational history to show APNIC or an NIR if questions arise. It may be asked by its own investors why scarce addresses are being purchased before customer contracts are certain. The incumbent can often choose when to transact. The entrant must transact when credibility is due.

This is the point at which waiting-list analysis would distract more than it helps. APNIC abolished the IPv4 waiting list in 2019, and recovered space is folded into the remaining policy-limited pool rather than being a large alternative supply line. For a new network that needs real growth capacity, the central question is not queue position. It is how quickly the firm can convert money, documents and counterparties into recognised, routable, trusted public identity.

The address price is only the visible price

When managers discuss scarce IPv4, they often fixate on the market price per address. That price is important, but it is only the most visible part of new-entrant disadvantage. The full price includes search, diligence, closing risk, registry timing, routing acceptance, reputation checks, financing cost, contract delay and the operational plan for using the addresses without wasting them.

Search is costly because the seller's identity matters. A block with unclear authority, contested history, poor reputation, awkward routing history or incomplete records is not the same asset as a clean, well-documented block. Entrants are particularly exposed because they cannot easily absorb a bad first purchase. If the address range has mail reputation problems, geolocation residue, prior abuse reports or routing disputes, the entrant pays again through remediation and customer delay. An incumbent with deeper inventory can quarantine a troublesome block or use it for less sensitive traffic. A young network may have only one shot.

Diligence is costly because the buyer must prove that the transfer is legitimate and operationally safe. APNIC's transfer material describes transfers as movements of number resources from one legal entity to another, with APNIC processing requests according to policy and updating the Whois Database to reflect results. The page also notes that supporting information is needed and that requests may be delayed when information is missing or incomplete. These are ordinary recordkeeping statements. In the new-entrant setting, they become launch economics. A missing corporate document, unclear signer, incomplete evidence file or seller-side delay can postpone revenue.

Financing is costly because addresses are awkward assets under many accounting and lending practices. A bank may recognise their economic importance but hesitate over collateral treatment, registry dependence, transferability and continuity. An investor may understand that public IPv4 is needed for revenue but dislike spending scarce early cash on an asset whose legal language is not as clear as land, spectrum or equipment. A lessor may offer flexibility but charge for the uncertainty. The result is a credibility spread: the entrant pays more because counterparties know it must solve the address problem before it can prove the business.

Registry timing is costly because customers do not wait for clean institutional sequencing. A public-sector supplier may need the network live by a contract date. A hosting customer may need deployment before a seasonal demand spike. A wholesale buyer may need a prefix ready for acceptance in filters. A fintech or health service may require testing windows. If registry recognition, routing evidence or transfer closing is uncertain, the entrant must buy temporary bridges. Those bridges can be expensive: cloud public IP charges, provider-assigned addresses, NAT gateways, emergency leasing, duplicate transit, or a slower launch with fewer features.

The official transfer path is therefore necessary but not sufficient. It records the transaction in the public system, which is exactly what a registry should do. But the entrant's disadvantage is created by all the costs around the recording event. The narrower and more predictable APNIC keeps that event, the less extra premium the entrant pays. The broader and more discretionary it becomes, the more the market discounts young networks.

The first transaction becomes a reputation test

The entrant's first address transaction does more than add capacity. It teaches the market how to rate the company. A smooth transfer, clean routing update and calm customer launch can make the young network look more institutional than its age. A delayed transfer, awkward seller file, stale routing entry, reputation surprise or hurried lease can create a story that follows the company into its next negotiation. In scarce-address markets, first impressions are not cosmetic. They affect the cost of the next block, the confidence of the next upstream and the patience of the next customer.

This is another reason incumbents are protected by history. Their old mistakes are often absorbed inside a larger record. A stale contact, a noisy range or a slow update may be annoying, but counterparties can place it inside years of operation. For an entrant, the same event can look like character evidence. If the first announced range has abuse residue, if a route filter is not refreshed, if a reverse-DNS change is late, if a ROA is missing or if a customer has to ask twice who controls the prefix, the market learns caution. The entrant may be technically innocent and still pay the premium.

Address reputation is not only a security topic. It is a financing topic. A clean range shortens sales conversations. It lowers support risk. It helps a customer believe that a service will not be blocked by mail systems, fraud platforms, geolocation errors or upstream filters. A questionable range forces the entrant to spend scarce staff time proving that yesterday's use is not tomorrow's risk. Incumbents can assign cleaner stock to sensitive customers and push messy ranges into less visible uses. Entrants often lack that internal portfolio.

The first transaction also creates a record of operational competence. Did the company know which evidence to prepare? Did it coordinate legal authority and technical authority? Did it align the registry record, IRR entry, ROA, upstream letter, reverse-DNS plan and abuse contact? Did it give the customer a credible continuity explanation? These are management questions as much as engineering questions. The entrant is being judged on its ability to convert scarce capital into dependable service.

APNIC cannot make every first transaction easy. Some checks are necessary because bad records damage the whole Internet. But it can make the test more objective. The more the entrant knows in advance about required information, expected timing, authentication, transfer state and correction paths, the less it must buy confidence from intermediaries. The more the process feels open-ended, the more the first transaction becomes a private exam in institutional fluency. Incumbents usually know the curriculum. Entrants must learn it while the meter is running.

This is why the thin-ledger doctrine is pro-entry even when it sounds austere. A thin ledger does not mean a careless ledger. It means the public record asks for what protects uniqueness, accuracy, routing trust and continuity, then stops. That stopping point matters. Every additional discretionary question becomes a possible story about the entrant's seriousness. Every objective rule becomes a defined requirement the entrant can satisfy.

Upstream trust is the second gate

Buying or leasing addresses does not by itself create reachability. The network must persuade upstream providers, peers, filters, security systems and sometimes customers that the route should be accepted. This is where newness becomes visible in operational form. A prefix from an old, familiar origin may pass through established filters with little drama. A prefix announced by a new ASN, through a new customer relationship, after a recent transfer or lease, invites more questions.

Those questions are not paranoia. Routing mistakes and hijacks can cause real harm. Upstreams have good reasons to require evidence that the customer controls the prefix and may originate it. They may inspect Whois records, letters of authorisation, IRR entries, RPKI ROAs, signed customer agreements, corporate identity and the history of the address range. The APNIC Routing Registry exists in this environment. It lets operators publish routing policy and announcements so other operators can use the data for route filtering, troubleshooting, planning and router configuration. APNIC's own description stresses that routing-registry information helps networks filter based on registered routes and check that the resource holder has control over routing entries for its resources.

For an incumbent, this evidence chain is routine. For a new entrant, it can be a launch gate. The firm may have the addresses but not yet the upstream confidence. It may have an LOA but not yet a trusted history. It may have an ASN but not yet a route acceptance track record. It may have a ROA but not yet all customer filters refreshed. It may be technically correct and commercially delayed at the same time.

This second gate explains why new-entrant disadvantage is not solved by cheap leasing alone. A leased prefix can provide early capacity, but customers and upstreams may ask who controls the ROA, who maintains reverse DNS, who answers abuse reports, who can change routing records, what happens at renewal and whether the lessee can survive a dispute. The less formal the arrangement, the higher the trust premium. A young network may accept these terms because it needs capacity. An incumbent may avoid them because it has internal inventory.

The same gate affects transfers. A transferred block must be made operationally believable. That means updating records, aligning route filters, creating or adjusting ROAs, cleaning up old IRR residue, checking reputation, changing reverse DNS and informing counterparties. Each task is small when staffed. Together they form an operating discipline. Incumbents have departments. Entrants have founders and contractors. The labour cost is part of the price of entry.

APNIC's pro-entry contribution here is practical. Keep the registry and routing-adjacent evidence coherent. Make resource-holder status reliable. Make transfer results visible. Make authentication relationships clear. Support RPKI and IRR use without turning routing evidence into a discretionary commercial approval system. The point is not to lower security standards for new entrants. It is to make objective proof cheaper to assemble.

Evidence favours the firms with history

Evidence is supposed to be neutral. In practice it often favours institutions with past transactions. A company that has held resources for years can show assignment files, utilisation records, invoices, tickets, routing history, abuse responses, staff continuity and old registry interactions. A company seeking to enter the market can show intentions, contracts under negotiation, technical design and capital commitments. Both sets of evidence may be honest. They are not equally persuasive.

This is the proof-before-revenue loop. The entrant needs customers to justify infrastructure, but customers want proof that infrastructure exists. The entrant needs addresses to win customers, but address sellers and financiers want confidence that the entrant will use and pay for them. The entrant needs upstreams to accept routes, but upstreams want registry and routing evidence. The entrant needs registry state to be updated, but the strongest evidence for future use is often the customer demand that depends on the update. Every loop can be solved, but each loop consumes time, money and managerial attention.

In mature markets, these loops are softened by intermediaries. Brokers, lessors, cloud platforms, managed-service providers, transit carriers, consultants and local registry support can all help. But intermediaries are not free. Their fees or margins become the price of credibility. A large incumbent internalises many of these functions. It has staff, records, contacts and old address stock. A new entrant buys them piecemeal.

The Asia-Pacific region sharpens the asymmetry because operating environments differ so widely. A young provider in a high-income city may face high address prices and demanding enterprise evidence, but also better access to capital, brokers and skilled staff. A young provider in a lower-income or smaller market may face global address prices without global revenue levels. A network in an economy served through a National Internet Registry may benefit from local language and administrative familiarity, but may also face another service seam when cross-border customers, transfers or routing records are involved. A start-up serving customers across economies may need to make a single proof file legible to several types of counterparty.

The result is not a simple "small versus large" story. Some entrants are well-funded data-centre, cloud, enterprise or mobile challengers. Some incumbents are inefficient. Some small networks are technically excellent. But the structural bias remains. The market discounts the party whose evidence is not yet proven. IPv4 scarcity raises the price of buying enough evidence quickly.

This is why a registry that describes itself as neutral must pay attention to evidence burden. Neutrality is not only equal wording. It is also proportional proof, predictable timelines, clear reasons for delay, usable public records, objective correction paths and restraint in asking business-plan questions that a registry is not equipped to judge. A thin registry lowers evidence cost. A thick registry turns evidence into an entry tax.

Customer credibility arrives before customer revenue

The first serious customer often demands more than bandwidth. It wants assurance that the network can stay reachable, identify abuse, meet service-level commitments, support static addressing, maintain DNS and routing records, survive supplier changes and answer audit questions. That assurance may be more important than the technical commodity being sold. The customer is buying a counterparty.

This is especially true for enterprise, public-sector, financial, health, education, industrial and platform customers. They may not understand every detail of IPv4 scarcity, but they understand operational risk. They ask whether source addresses are stable. They ask whether remote systems can be allowlisted. They ask whether mail and security systems will behave. They ask whether the provider can produce logs. They ask what happens if an address block is withdrawn, a lease ends, a seller disputes authority or a route is filtered. These questions arrive before the entrant can point to years of successful service.

For the entrant, the temptation is to hide complexity. It can promise "Internet connectivity" and hope the customer does not ask about public identity. That strategy is dangerous. The moment a problem appears, the absence of clear address control becomes a trust crisis. A better entrant prices the issue honestly, but honest pricing can make the offer look expensive beside an incumbent whose old inventory is already amortised. The incumbent can include static public IPv4 as a feature, or at least present it as normal. The entrant must explain why normal became scarce.

Customer credibility therefore becomes another channel through which incumbency is preserved. Buyers compare service proposals, not registry histories. Yet registry history is inside the proposal. The provider with recognised resources, accepted routes and clean evidence can look boring. In infrastructure, boring is valuable. The provider still assembling public identity looks riskier even if its technology is modern and its service quality could be better.

This is where APNIC's recordkeeping role has a real but limited competitive effect. If the registry layer is predictable, a new entrant can turn a transfer or delegation into customer-facing assurance more quickly. It can show recognised resource status, routing evidence, contacts and security assertions. If the registry layer is uncertain, the entrant must sell through ambiguity. Incumbents can tolerate ambiguity because customers trust their history. Entrants cannot.

The policy lesson is modest. Do not make new firms prove more commercial wisdom than the market requires. Do not make address acquisition feel like a favour. Do not let scarcity rhetoric become a reason for slow, discretionary questioning. Record lawful, accurate changes of control. Support security evidence. Make disputes visible when they matter. Let customers, investors and operators judge the business model.

Timing protects incumbents without announcing itself

Delay is the most polite form of market power. It rarely looks like exclusion. It looks like a missing document, a pending verification, a support queue, an upstream review, a filter update, a seller's internal approval, a bank question, a cloud exception, a local registry handoff, a customer security review or a meeting that cannot be moved. For an entrant with cash burn, delay is expensive. For an incumbent with existing customers and address stock, delay is often tolerable.

This timing asymmetry is central to APNIC-era new-entrant disadvantage. A scarce resource whose recognition takes longer than the sales cycle creates lost deals. A transfer that closes after the customer deadline is economically different from a transfer that closes before it. A route that is technically valid but not yet accepted by an upstream filter is not fully useful. A ROA that exists but has not propagated through operational trust is not the same as accepted reachability. A public IPv4 plan that depends on a future transaction is not the same as addresses already in hand.

The entrant pays for timing through buffers. It may lease temporary addresses while waiting for a transfer. It may buy a more expensive cloud design because customer launch cannot wait. It may accept provider-assigned addresses, weakening portability. It may overbuy because a second transaction would be too slow. It may delay marketing until the network identity is stable. It may discount early contracts to compensate customers for perceived risk. These costs do not appear on APNIC's fee schedule, but they are registry-layer-adjacent costs when timing depends on record recognition and evidence.

Incumbents can use timing passively. They do not need to lobby against entrants. They can simply enjoy the fact that old inventory is already accepted. They can respond to tenders quickly. They can expand a customer within existing blocks. They can wait for transfer markets to improve. They can force entrants to be the party explaining scarcity. Advantage is preserved by accumulated readiness.

This is why a pro-entry registry discipline cares about service levels, reasons and reviewability. A delay may be necessary. Fraud control, duplicate claims, legal authority and security integrity matter. But necessary delay should be legible. Entrants can plan around known checks. They struggle with uncertain discretion. A registry that publishes clear requirements and processes objective records quickly reduces the incumbent timing premium. A registry that treats each case as an open-ended judgement increases it.

NIR seams can lower friction or add another proof layer

APNIC's region is unusual because National Internet Registry arrangements sit inside the regional system for several large or distinct economies. These institutions can make the registry layer more local. They may provide language support, familiar service channels, local currency billing, domestic community knowledge and administrative proximity. For entrants, that can lower search cost and make the first resource file less intimidating.

But local service seams can also become another proof layer. A network that buys, leases, routes or serves across borders may need evidence to be accepted by APNIC, an NIR, an upstream outside the local market, a foreign customer, a broker, a cloud provider and a bank. If records, timelines, terminology or expectations differ, the entrant becomes translator. Incumbents often have the staff and history to handle translation. Entrants pay consultants or learn under deadline.

The seam matters especially for transfers and routing acceptance. A block's registry path, prior holder, NIR context, route history and current operational use can affect buyer confidence. A local registry may be helpful in validating the resource. A cross-border counterparty may still demand APNIC-visible clarity. An upstream may care less about local institutional nuance and more about whether filters, IRR entries, ROAs and contacts line up. The entrant must make all audiences believe the same story.

This does not make NIRs bad. In a region as large and varied as Asia-Pacific, local registry layers can be valuable. The danger is not locality; it is opacity. If locality reduces friction, entrants benefit. If it adds uncertain discretion or translation burden, entrants pay. The difference depends on whether the registry system behaves like a coordinated record layer or a set of semi-independent approval cultures.

APNIC's pro-entry task is to make local and regional evidence converge. Resource-holder status, transfer outcomes, routing authority, contact records and security assertions should be explainable across the region. The entrant should not need private knowledge of institutional culture to make a valid resource state legible. Thin coordination does not mean weak coordination. It means strong common evidence with limited discretionary control.

Leasing and cloud bridges are useful, but not neutral

Because early address acquisition is expensive, entrants often use bridges. They lease IPv4. They rely on provider-assigned space. They launch services behind a cloud platform's public identity. They use NAT-heavy architectures. They buy only enough public IPv4 for premium customers. These bridges can be rational. Without them, many services would never start. The mistake is to treat them as neutral substitutes for durable resource control.

Leasing provides flexibility but creates renewal and authority risk. Who controls the ROA? Who updates routing records? What happens if reputation problems surface? Can the lessee show customers sufficient continuity? How quickly can the range be replaced? Does the lessor have incentives aligned with the entrant's customer contracts? A lease may solve the scarcity problem while creating a credibility dependency. The entrant has capacity but not full independence.

Provider-assigned addresses create a different dependency. They can be cheap and quick, but they make exit harder. If the entrant changes upstream, the addresses may not move. Customer systems, allowlists, DNS, security tools and reputation histories can become attached to someone else's numbering. The entrant buys early simplicity by surrendering future portability. Incumbents with their own holdings do not face the same trade at the same intensity.

Cloud bridges are similar. A young company can avoid some address acquisition by using public cloud services, managed NAT, load balancers and platform-owned public IPs. This can be excellent engineering. It can also turn the platform into the holder of public identity. The entrant becomes a tenant in another network's address economy. When public IPv4 charges, egress architecture, account controls, reputation events or compliance requests arise, the platform's rules shape the entrant's choices.

Shared-address architecture can conserve scarce IPv4 but may degrade the product. CGNAT and similar designs can increase logging burdens, support complexity, abuse attribution problems, application failures and customer frustration. A large incumbent can absorb those costs over a broad base. A new entrant may see each edge case threaten reputation. The customer does not blame scarcity; the customer blames the young provider.

These bridges are therefore part of the disadvantage, not merely remedies for it. They allow entry while preserving dependency. They let the entrant start, but often with a higher variable cost, weaker independence or lower product quality than an incumbent with recognised stock. The registry should not ban these arrangements simply because they complicate the old allocation story. It should record reality where recordkeeping is needed, support clarity around delegation and control, and avoid treating commercial flexibility as suspicious by default.

Incumbent advantage is an option portfolio

For incumbents, IPv4 scarcity is not only an operating constraint. It is an option portfolio. Existing holdings let a firm choose among internal use, customer allocation, reserve, lease, sale, merger value, collateral discussion, premium product design and strategic patience. The holder can wait for better prices, keep blocks for high-value customers, or use address depth to make procurement promises competitors cannot match.

This is not necessarily bad behaviour. A company that built infrastructure and maintained records has real reliance interests. It may need reserve capacity for continuity, disaster recovery, customer churn, network redesign and future products. Accusing every incumbent of hoarding would miss the point. The economic issue is not moral guilt. It is asymmetry. The same scarce resource that is a launch hurdle for entrants is a strategic option for incumbents.

The option value becomes visible in customer negotiations. A buyer asks for public IPv4 and predictable routing. The incumbent can say yes from inventory. The entrant must price a transfer, lease or bridge. A public body asks for continuity across a contract term. The incumbent can point to its history. The entrant must explain contingencies. A lender reviews a network acquisition. The incumbent's recognised address base can support valuation. The entrant's plan to acquire addresses later can look speculative.

Registry language can either clarify or distort this asymmetry. If the registry treats resources as mere administrative permissions, it may understate the real capital value of incumbent holdings and the real entry cost borne by challengers. If it treats the registry as owner-like gatekeeper, it increases uncertainty and makes capital more expensive. The better approach is asset-realism without registry sovereignty: recognise that scarce IPv4 has market value and operational reliance, while keeping APNIC's role to accurate records, uniqueness, security assertions, transfer legibility and continuity.

This distinction matters for future articles on incumbent optionality and asset capitalisation, but it is already present here. New-entrant disadvantage exists because incumbents' old holdings function as options while entrants' first holdings function as obligations. The incumbent can choose. The entrant must prove.

The ledger should not become an investment committee

Scarcity invites paternalism. If IPv4 is valuable and new entrants are disadvantaged, it is tempting to ask APNIC or any RIR to decide who deserves scarce addresses, whether a business plan is socially useful, whether a lease is acceptable, whether a buyer has enough need, whether a customer geography is proper, or whether an operator is using addresses in the right way. This temptation should be rejected. It makes the registry more powerful and entrants more dependent on discretionary judgement.

The better pro-entry answer is narrower. APNIC should act as a ledger, not an investment committee. It should protect uniqueness. It should keep Whois and related resource records accurate. It should process transfers according to objective requirements. It should support routing-adjacent evidence, RPKI, reverse DNS and contactability. It should make reasons for delay clear. It should isolate disputes without punishing uninvolved customers. It should provide predictable correction and review paths. It should make NIR seams legible. It should not decide whether an entrant's commercial model is worthy.

This is not a laissez-faire fantasy. Fraud control still matters. Duplicate claims still matter. Sanctions and legal orders may matter. Security emergencies may matter. Abandoned or dirty resources may need careful handling. The distinction is between record integrity and economic permission. A registry can verify that the claimed change is real. It should be cautious about judging whether the resulting business is desirable.

The reason is institutional competence. APNIC does not bear the entrant's capital risk. It does not pay the seller. It does not compensate the customer if deployment fails. It does not know future demand better than the market. It does not finance routers, towers, backhaul, cloud regions or support desks. It can make the common record more reliable, but it cannot allocate capital without becoming a source of capital-control risk.

For entrants, this boundary is crucial. A predictable record layer lowers the cost of being believed. A discretionary gate raises it. The firms best able to navigate discretionary gates are rarely the youngest or most competitive. They are the ones with staff, counsel, consultants, legacy evidence and time. If APNIC wants a region where new networks can challenge incumbents, the most useful institutional virtue is not generosity. It is restraint.

A pro-entry discipline for APNIC scarcity

New-entrant disadvantage cannot be abolished by policy language. Scarcity is real. IPv4 has market value. Customers will keep asking for compatibility. Upstreams will keep requiring evidence. Sellers will prefer buyers who can close. Banks will ask hard questions. Incumbents will retain the advantages of history. A registry cannot make these facts disappear without pretending to own the market.

But APNIC can lower the part of the disadvantage that comes from the registry layer. The discipline is practical. Requirements should be known before a transaction starts. Supporting information should be proportionate to record accuracy, not a wandering inquiry into commercial merit. Transfers should have clear status, reasons and timelines. Routing evidence should be easy to align with registered control. RPKI and IRR practices should make objective trust cheaper. Contact records should be useful without becoming broad enforcement hooks. NIR relationships should reduce language and service friction without creating hidden variance. Fees should reflect necessary registry functions. Adverse actions should be narrow and reviewable.

The market will still decide which entrants survive. That is normal. The question is whether the common registry layer adds avoidable premium to entry. A thin, reliable APNIC record lets a young network convert capital into recognised resources, recognised resources into accepted routes, accepted routes into customer proof, and customer proof into revenue. A thick, slow or discretionary record layer breaks that chain and rewards the firms that no longer need it urgently.

This is also why official stories about IPv6 as the long-term solution do not settle the entrant question. IPv6 is necessary and abundant, but a new provider still faces customers, applications, public tenders, security tools, allowlists and reputation systems that treat IPv4 compatibility as current reality. The entrant must operate in that world, not in a future diagram. A policy that assumes the future has already arrived will underprice the evidence costs of entering the present.

The fairer entry environment is not one where IPv4 is made artificially cheap or where APNIC chooses winners. It is one where scarcity is visible, transfers are legible, routing trust is objective, registry timing is predictable and address control can move without unnecessary institutional theatre. In that environment, incumbents still keep the benefits of their history, but they do not receive an additional subsidy from opaque process.

The founder at the beginning of this story does not need inspiration. The founder needs a public identity that counterparties will believe before revenue is secure. That identity is made of scarce addresses, accurate records, accepted routes, security assertions, clean reputation, reachable contacts and a timeline that customers can trust. In the APNIC region, those pieces are the entry ticket to much of the modern Internet.

If APNIC is wise, it will treat that ticket as a recordkeeping problem rather than a licence to inspect ambition. New entrants should win or lose because customers, capital and operations judge them, not because the common ledger adds a hidden incumbency premium. The registry may record scarcity. It should not compound it.

Sources and Further Reading