The prudent word at the wrong moment
The most powerful word in a post-exhaustion address debate is often the quietest one. "Conservation" sounds like administrative common sense. It suggests thrift, patience and public duty. It tells engineers who remember the end of the free pool that scarcity will not be handed over to commerce, brokers, speculation or private convenience. It lets a registry speak the language of stewardship at precisely the moment when the resource it stewards has become a balance-sheet item.
That is why the word deserves scrutiny. In the LACNIC region, conservation no longer operates mainly as a rule for slowing depletion of a large pool of unallocated IPv4 space. That pool has gone. LACNIC's current Phase 3 environment is a world of recovered and returned addresses, a waitlist that began when the last available block was assigned in August 2020, small allocations from quarantined space, transfer procedures, legacy regularisation, reverse-DNS and RPKI continuity, and a secondary economy in which operators buy, sell, lease, sub-assign, acquire or finance address capacity because customers still need IPv4 reachability.
In that world, conservation language can still be useful. Registry accuracy is worth conserving. Routing-security authority is worth conserving. Contactability, abuse responsibility, reverse-resolution continuity, holder provenance and the public ability to tell who is responsible for a block are worth conserving. A registry that stops caring about these things becomes an unreliable bookkeeper of a globally unique resource.
But conservation can also become an anti-market posture without admitting that it is one. It can make delay look prudent when the real question is who bears the financing cost of waiting. It can make a need review look neutral when it favours organisations able to write conventional justifications. It can make restrictions on transfer, leasing or sub-assignment sound like protection of the commons when they also protect incumbents holding clean historical inventory. It can make public responsibility logs seem like transparency while leaving small operators to absorb the cost of compliance. It can make continued IPv4 demand appear morally suspect even where IPv6 deployment is already under way and customers still require dual-stack or IPv4 service.
This is not a complaint that LACNIC should become a laissez-faire address exchange. It should not. LACNIC is the regional Internet registry for Latin America and the Caribbean. It is an international, non-governmental organisation established in Uruguay, responsible for IPv4, IPv6, autonomous system numbers and reverse resolution in a region that its own public material describes as more than 13,000 network operators across 33 territories. Its record is relied upon by operators, upstreams, abuse desks, other registries, route-origin systems, buyers, sellers, banks, brokers and public institutions. A registry record that can be forged, ignored or casually changed would destroy value faster than any transfer rule could create it.
The harder point is that, after exhaustion, the registry's conservation vocabulary has become distributional vocabulary. It decides how a scarce input moves, who can monetise unused capacity, who can finance expansion, who gets official recognition, who must depend on private contracts, who pays for documentation, who has enough staff to follow policy debates, who waits on a queue, and who enjoys the quiet advantage of having received addresses before scarcity became a balance-sheet fact. The rhetoric matters because it can conceal these choices behind a word that almost everyone is reluctant to oppose.
This is the institutional shift that should frame the debate. During the allocation era, conservation chiefly governed access to a remaining stock. In the recognition era, it governs the liquidity of a stock already distributed. The first question was how to slow depletion. The second is how to make movement safe without converting the registry into an unacknowledged market gatekeeper.
The question for LACNIC is therefore not whether conservation is good or bad. It is what, exactly, is being conserved. If the answer is registry truth, operational responsibility and security continuity, the claim is strong. If the answer is discomfort with lawful transfers, leasing, price discovery, brokered liquidity or the fact that address holders can earn from scarce inventory, then conservation has become a bridge from stewardship to market control. That bridge should be marked clearly before too much economic traffic crosses it.
When conservation was a real allocation principle
Conservation had a coherent role before exhaustion. IPv4 was finite. The Internet's early address architecture was more generous than later growth could sustain. Classful allocation left large historical blocks in places that did not always match later use. CIDR, renumbering, private addressing, network address translation and IPv6 were all responses to a system whose original supply assumptions had failed. Regional registries, including LACNIC, had to allocate remaining space in a way that was efficient enough to preserve access for future networks.
In that period, need-based review was not merely a bureaucratic habit. A registry with unallocated stock had to choose among applicants. If it handed addresses out too freely, the free pool would vanish faster and late entrants would be harmed. If it demanded no utilisation evidence, organisations with better paperwork, sharper timing or more aggressive forecasts could consume space before others even arrived. If it ignored aggregation and registration quality, routing and operational accountability would suffer. Conservation was connected to a concrete object: a pool that still existed and a future set of networks that might need it.
LACNIC's older allocation language still reflects that world. The policy manual describes IPv4 allocations through a slow-start approach and says allocations are based on current justifiable need rather than predictions of clients or market research. For initial IPv4 requests in the current small-block range, an applicant must show immediate need or use for part of the requested space and a one-year plan for further utilisation. In the exhaustion phase, LACNIC can assign only small amounts, from the equivalent of a /24 to a /22, and each new member may receive only an initial assignment. These rules sound restrictive because they are. They also make sense as rationing devices for an almost empty administrative source.
The legitimacy of such conservation depends on the nature of the request. An applicant asking LACNIC for a fresh block from a reserve or recovered pool is asking the registry to distribute a scarce resource. The registry is not merely recording a private event; it is deciding whether public stock should be consumed. In that setting, conservation has a direct target. It tries to make the last administrative supply last longer, support new entrants, limit premature depletion and keep use tied to network deployment.
That history should not be caricatured. There was real stewardship in it. It is easy, in the transfer era, to forget why allocation rules became cautious. Waste was possible. Speculative requests were possible. Routing fragmentation mattered. Incomplete records harmed operations. Legacy allocation practices had already shown how generous assumptions could shape the market for decades. A registry that had ignored conservation before exhaustion would have invited criticism from the same late entrants who now depend on secondary-market access.
The mistake is to carry the allocation-era moral frame unchanged into a post-exhaustion settlement market. Once most marginal IPv4 demand cannot be satisfied from registry stock, the main policy question shifts. It is no longer mainly how to preserve a free pool. It is how to recognise, record and secure movement of addresses already issued. Transfers, leases, sub-assignments, acquisitions, public-sector successions, legacy regularisation and operational delegation are not the same event as initial allocation. They may require checks, but the checks should not smuggle old rationing logic into every movement of recognised capacity.
That distinction is the foundation of the economic problem. Conservation before exhaustion constrained applications for public stock. Conservation after exhaustion can constrain liquidity in private and quasi-private arrangements. The same word hides a different incidence. The first postpones depletion. The second determines who can convert recognised control into operational use or cash. It is the difference between rationing a pantry and certifying settlement in a market for goods already on the shelves.
The queue that proves the market exists
LACNIC's waitlist is the clearest exhibit. It was created on August 19th 2020, when the last available IPv4 address block was assigned. Its purpose is orderly: create a sequence among organisations requesting IPv4 addresses. It is also sobering. Based on historical recovery patterns, LACNIC estimates that the last request on the list faces a wait of at least eighteen years and can receive at most 1,024 IPv4 addresses. Future recovered space is uncertain. The queue can change as earlier tickets are approved. Positions are not transferable. An organisation must be a LACNIC member and must have IPv6 resources before being included. Blocks assigned in this phase have been quarantined for at least six months, but LACNIC does not guarantee that recipients will avoid reputation or filtering problems.
This is not a growth mechanism. It is a rationing memorial. A /22 after a very long wait may help a small initial network or a limited transition requirement. It cannot support the normal address demand of a growing access provider, hosting company, data-centre operator, enterprise-services platform, mobile translation layer, public-sector migration or regional cloud business. Operators with present demand must obtain usable IPv4 from someone else, redesign around scarcity, rely on upstream space, use more carrier-grade NAT, lease capacity, buy through transfer, acquire a company, regularise old holdings or constrain service.
The waitlist therefore makes the market visible by failing to replace it. When the official queue is measured in years and small blocks, secondary mechanisms are not aberrations. They are the way scarcity is financed. A registry can dislike parts of that market, and it should regulate the record-facing risks. But it cannot pretend that conservation of a queue is conservation of operational opportunity. The queue conserves order among applicants for recovered space. It does not conserve working capital for a provider that needs addresses next quarter, nor does it substitute for the credit, contracts and risk allocation that now surround usable IPv4.
The economic significance is larger than the number of addresses. The waitlist changes the value of recognised inventory. A holder with clean records, current contacts, usable RPKI, coherent reverse DNS and no dispute has something that others cannot quickly obtain from the registry. That something can support revenue or be exchanged for money. A buyer does not merely buy numbers; it buys time. A lessee does not merely rent identifiers; it rents a way around a multi-year administrative wait. A small seller does not merely monetise surplus; it converts an old technical allocation into cash that may fund equipment, resilience, debt reduction or a business transition.
Conservation rhetoric can obscure this conversion. If all movement outside the waitlist is described as suspect because it treats IPv4 as an asset, the rhetoric ignores the fact that the waitlist itself has made asset behaviour inevitable. Scarcity turns idle capacity into option value. Scarcity turns clean documentation into a premium. Scarcity turns a registry review into a settlement risk. Scarcity turns a small block into working capital.
The sensible policy response is to distinguish official rationing from market hygiene. LACNIC can keep a disciplined waitlist while recognising that most actual marginal demand will be met elsewhere. It can require IPv6 resources for waitlist applicants while admitting that IPv6 does not eliminate present IPv4 demand. It can quarantine recovered space while publishing enough reputation guidance to help recipients. It can protect the queue from transfer of positions while not treating lawful purchase, lease or sub-assignment as moral evasion of the queue.
The wrong response is to imply that the waitlist preserves fairness so completely that liquidity elsewhere is unnecessary or distasteful. A queue that may take eighteen years cannot be the only legitimate path. Once that fact is accepted, conservation must stop meaning "movement is suspicious" and start meaning "movement must be truthful, accountable and secure."
Transfers: recognition, not charity
Transfers are where the old vocabulary meets the new economics most directly. LACNIC permits IPv4 transfers inside its region and between regions, subject to policy conditions. The minimum transferable block is a /24. A recipient inside the LACNIC region must justify the IPv4 resources before LACNIC under the applicable allocation or assignment policies. LACNIC or the corresponding regional registry verifies the holder and checks whether the resources are involved in a dispute. Intra-regional transfers require signed legal documentation. Inter-regional transfers require coordination between registries. Once the transfer is complete, LACNIC modifies the resource information to reflect the new holder and records the transaction in a public transfer log.
Each of these steps has a proper conservation rationale if the object is record reliability. Minimum size limits prevent useless fragmentation and preserve routability discipline. Holder verification prevents theft. Dispute checks keep the registry from laundering contested control. Legal documents connect the request to real corporate authority. Inter-regional coordination prevents one globally unique resource from being recognised inconsistently. The public log creates market memory without disclosing private prices. These are not anti-market devices by themselves. They are settlement infrastructure.
The economic danger lies in the same machinery. A one-year ineligibility period for the originator to receive new IPv4 resources changes selling incentives. A one-year lock on retransferring previously transferred space changes inventory strategy. A three-year lock on LACNIC allocations or assignments changes planning for newly received blocks. Legacy resources transferred into the region lose legacy status, changing valuation. Need justification changes who can buy quickly. Documentation discretion changes the premium earned by repeat participants. Account standing and renewal timing change closing risk. A public log changes bargaining by making some movement visible and some pricing still private.
This is the heart of post-exhaustion conservation. It is not merely preservation. It is market design. A registry can say, accurately, that it does not set prices and does not participate in commercial transactions. LACNIC's transfer listing service reflects that boundary: it facilitates contact among possible offerors, recipients and optional brokers; it charges an administrative fee for listing; it does not audit broker services; it says LACNIC does not take part in the commercial transaction between the parties. That is appropriate. Yet the registry still decides when a private bargain becomes a recognised public record.
The old need test is the most delicate part. When LACNIC allocates from its own remaining or recovered pool, asking for current justifiable need is rationing. When a buyer pays another holder for addresses that have already been issued, the registry's role is different. It should verify identity, authority, eligibility, amount, dispute status, responsibility and compliance with adopted rules. It may test whether the transaction is a sham or an attempt to evade a clear restriction. But if need review becomes a broad inquiry into whether the buyer's business plan is sufficiently virtuous, conventional or developmentally pleasing, conservation becomes economic permission.
Willingness to pay is not perfect evidence of use. Speculators can pay. Intermediaries can pay. Shell entities can pay. A registry should not treat payment as proof of legitimacy. But willingness to pay is evidence that the resource has economic value to someone, and in a capital-constrained region it should not be dismissed. Few small operators buy IPv4 space for decoration. They buy it because customers, upstreams, applications, hosting contracts, public agencies, migration projects or growth plans require it. A need review that ignores this fact reproduces allocation-era suspicion in a market where the registry is no longer giving away abundant stock.
The better standard is narrow recognition. A recipient should show that it is real, accountable, within the relevant policy framework and able to explain coherent use. It should not have to perform moral purity. A seller should prove authority and clean control, not demonstrate that profit is socially modest. A broker should be judged by transparency and conduct, not by the fact that intermediation offends the old non-commercial self-image of registry culture. A transfer should fail when record accuracy, legal authority, policy eligibility or operational responsibility fails. It should not fail because scarcity rent makes people uncomfortable.
Leasing and the sub-assignment bargain
Leasing is the place where conservation rhetoric most easily becomes moral rhetoric. A sale can be recorded as a transfer. A lease leaves recognised holdership with one party while operational use moves to another. The arrangement can be efficient, opaque, abusive, practical, speculative or essential depending on the facts. It turns a capital purchase into a recurring expense. It lets a small provider match cost to revenue. It lets a holder earn from dormant inventory. It gives a network temporary capacity for a migration, a seasonal project, an enterprise customer, a hosting service or a transition period. It can also separate the public record from the party that originates routes, handles abuse, needs reverse DNS and depends on ROAs.
For years, LACNIC's policy debates reflected discomfort with leasing. A 2022 proposal argued that lease of addresses should not be acceptable unless part of direct connectivity or services based on customer connectivity, on the theory that address resources are delegated rather than property and that justified need is tied to own infrastructure or direct customers. That line of reasoning is classic conservation language. It treats the link between addresses and network service as something to be preserved against commodification.
By late 2025 and early 2026, however, the debate had moved toward a more pragmatic acknowledgement. A proposal on sub-assignment of IPv4 resources to third parties described the existing prohibition on sub-assignments outside the original recipient's own infrastructure and argued that, in the context of IPv4 exhaustion, informal address leasing without records had created problems, including outdated registries and the possibility of resources leaking to other regions through transfers. The revised proposal sought a transparent mechanism, with WHOIS identification, a public movement log for start and end events, responsibility remaining with the member making the sub-assignment, exclusions for critical-infrastructure resources, regional-use requirements, maximum size limits aligned to small-block allocation logic, and restrictions such as a three-year period limiting certain future resource actions by organisations using the mechanism. LACNIC's policy system marked the proposal as ratified by the board in March 2026.
That movement is important because it shows conservation being forced to confront reality. A blanket anti-leasing posture may preserve vocabulary, but it can damage the record. If operators lease anyway, and the registry refuses to see delegated use, public responsibility degrades. Abuse contacts may point to a passive holder. RPKI changes may depend on a party not visible to the customer. Reverse DNS may lag. A broker chain may hide the party that can act. If a lease ends badly, stale authorisations may remain. The ledger can be formally accurate and operationally unhelpful.
The sub-assignment approach is not automatically right in every detail. It can create fixed costs. It may deter small users if public logs expose sensitive dependencies. Its IPv6 or ASN requirements may exclude some legitimate use cases. A three-year restriction may reduce speculation but also discourage lawful liquidity. Keeping final responsibility with the member is sensible, but it requires the member to price and manage downstream risk. The policy may still be too heavy for the very informal market it tries to bring into view.
Yet its direction is healthier than pure moralisation. It asks how to make responsibility visible instead of pretending temporary use is nonexistent. It recognises that conservation of accuracy may require recording a market reality. It shifts the question from "is leasing virtuous?" to "who is responsible, who is using the block, what can the public record show, and how does the arrangement end cleanly?"
That should be the general rule. LACNIC should not become a lease-price regulator. It should not collect commercial rents, approve margins or decide whether a holder's yield is too high. It is not a telecom regulator, central bank, competition authority or price commission. Its legitimate questions are record questions: who is the recognised holder; who is authorised to originate; who can administer reverse DNS; who answers abuse; who remains accountable if the operational user disappears; what happens to ROAs, IRR objects, geolocation and contacts at termination; when has a lease become a disguised transfer of practical control?
Conservation of the responsibility chain is a real registry task. Conservation of discomfort with rent is not.
Accuracy is not anti-market
The strongest defence of conservation after exhaustion is accuracy. A registry's ledger is valuable because others believe it. If the record cannot identify the responsible holder, if contacts are stale, if assignments are not registered, if abuse mailboxes fail, if reverse DNS is abandoned, if RPKI authority does not follow recognised control, then address liquidity becomes dirty. Buyers discount blocks. Lessees demand lower prices. Upstreams ask for more proof. Banks hesitate. Abusers exploit old records. Legitimate operators pay for the resulting mistrust.
LACNIC's own policy language contains a useful example. ISP assignments of /29 or larger blocks to customers connected to their network must be registered in LACNIC's WHOIS database within seven days. The policy identifies reasons for registration: informing the Internet community which organisation is using the space, providing contacts for operational and security problems, assisting regional allocation study and facilitating geolocation of sub-assignments. It also says that, as long as prefixes are registered, the recipient has the right to create and manage RPKI Route Origin Authorizations for those resources.
This is conservation in its best sense. It conserves the public usefulness of the record. It connects delegated use to contactability and route-origin authority. It recognises that accurate registration is not an ideological claim about ownership but a practical condition for operating a scarce resource. The same logic should guide transfers, leasing, sub-assignment and legacy regularisation.
Accuracy, however, is not the same as suspicion. A registry that wants accurate records should make clean registration cheap enough and predictable enough that serious actors prefer it. If updating delegated use, contacts, reverse DNS or ROAs is slow, risky or treated as an invitation to broad review, operators will avoid the formal path. If every correction feels like a possible audit of entitlement, old records will remain stale. If members believe that visibility of sub-assignment will invite moral judgement rather than operational clarity, lawful use will hide under private contracts.
The registry should therefore separate hygiene from merit review. A contact update should normally be a contact update. A reverse-DNS delegation should test authority over reverse resolution, not become a referendum on the holder's business plan. A route-origin authorisation should follow recognised or properly delegated control, not institutional taste. A sub-assignment log should show the facts necessary for traceability, not private financial terms. A transfer log should record movement, not imply approval of price.
This separation is especially important in a market with old records. Many LACNIC-region resources have histories that pass through universities, public bodies, state-linked operators, privatisations, family companies, cooperatives, acquisitions, bankruptcies, reorganisations and brands that no longer exist. Cleaning those records is good. Turning every cleanup into a moral review of continued IPv4 use is not.
Accuracy is pro-market because it reduces the discount attached to uncertainty. A clean record helps a small seller receive a fairer price. It helps a buyer close without excessive warranties. It helps a lessee demand real authorisation. It helps an upstream accept a route. It helps an abuse desk reach the right party. It helps the public distinguish lawful delegation from hijack. The more accurate the record, the less work must be done by private suspicion.
The best post-exhaustion conservation is therefore not a brake on movement. It is a discipline that lets movement happen without making the ledger false.
Fixed costs and the unequal price of virtue
A conservation rule often looks equal because the words are the same for every applicant. The costs are not equal. A requirement to justify need, produce legal documents, sign a service agreement, update assignment data, prepare a utilisation plan, prove authority, pay a fee, correct contact records or wait through inter-regional coordination has different incidence depending on the participant.
LACNIC's transfer pages make some of these costs explicit. Intra-regional and inter-regional transfers carry administrative fees by block size, with one level for blocks from /24 up to smaller than /19 and another for /19 and larger blocks. A US$200 down payment is required before justification is analysed and is not refunded if the transfer cannot be justified and is not approved. If the receiving organisation is a member and its category changes, a complementary invoice may be issued. If it is not a member, it may receive an invoice for the relevant category plus the administrative fee. Both offering and receiving organisations, where applicable, must be current with LACNIC obligations. Transfers or returns of assigned resources must be requested at least 30 days before the renewal invoice due date, or the member may have to pay the full renewal invoice to complete the process.
None of these rules is absurd. Fees fund process. Down payments discourage frivolous requests. Category changes preserve the membership structure. Current obligations prevent abuse of the service relationship. Renewal deadlines help administrative planning. But together they form a cost stack. For a large carrier closing a multi-block transaction, the stack may be ordinary. For a small operator selling a /24 or buying a /23, it may shape the transaction.
Documentation is often the largest hidden cost. In merger and acquisition cases, LACNIC asks for legal documents confirming transfer of assets, an inventory of assets used to keep the IPv4 space in use, and client lists or numbering plans to justify need. In ordinary transfer cases, LACNIC may request documents confirming authority and may use external collaborators to certify authenticity. In inter-regional cases, two registry processes and two document cultures may interact. These are legitimate anti-fraud measures. They are also fixed costs that do not shrink in proportion to a small block.
A multinational buyer can maintain legal staff, registry specialists, tax advisers, brokers and escrow relationships. It can prepare need narratives in the style registries expect. It can wait out a review. It can absorb a failed down payment. A small island access provider may have one person handling routing, billing, customer support, procurement and registry tickets. A family ISP may have valid authority but old paper. A public agency may have legal continuity through a decree rather than through a tidy corporate registry extract. A university network may have inherited records from a research era when nobody imagined IPv4 would become a scarce input. A cooperative may need a board meeting just to sign a document. Equal formal rules can therefore produce unequal settlement risk.
The same conservation burden can therefore act like a tax on smallness. A rule meant to prevent hoarding can favour those already big enough to document everything. A rule meant to discourage speculation can also discourage a legitimate small transfer. A rule meant to ensure current obligations can turn a bank delay into a capital freeze. A rule meant to prove need can reward applicants whose demand looks conventional and punish those whose demand is real but newer, more flexible or temporary.
The answer is not to lower proof for weak participants. That would damage the record and invite fraud. The answer is proportionality and predictability. Common cases should have published evidence paths. Routine holder transfer, name change, asset sale, share sale, public-sector succession, insolvency, account recovery, suspected fraud, legal dispute, legacy regularisation, inbound inter-RIR transfer and outbound inter-RIR transfer should not all feel like one undifferentiated request for more comfort. A small block should not need the same economic ceremony as a large consolidation unless the risk is the same.
Conservation claims should therefore include incidence analysis. Who pays the fixed cost? Is the cost proportional to the record risk? Does the rule favour repeat players? Does it make clean formal movement more attractive than informal workaround? Does it preserve accuracy, or does it make accuracy too expensive to report? A conservation policy that cannot answer these questions is not yet a policy. It is a virtue claim.
Geography and the politics of patience
LACNIC's region is too varied for a single patience assumption. Brazil and Mexico contain large technical communities, national institutional channels, deep operator ecosystems, legal capacity, data-centre growth and repeat participants in number-resource processes. Argentina, Chile, Colombia, Peru and other large markets contain sophisticated networks but may face their own currency, inflation or procurement constraints. Central American operators may run across borders and rely on narrow wholesale relationships. The Caribbean includes small island economies where telecom staff, legal resources, correspondent banking, storm resilience and public-sector demand are all thin.
A rule can be formally regional and economically local. An extra week of review is different in a large firm with inventory than in a small provider whose hotel customer, government service or enterprise migration starts next month. A requirement for a dollar payment is different where dollar settlement is routine than where a foreign wire triggers approvals, bank fees, exchange-rate exposure or de-risking. A public log is different for a large cloud platform than for a small island network that does not want competitors to infer customer dependencies. A documentation request is different in a market with fast corporate registries than in a public-sector succession involving old ministries.
Large-country gravity is not inherently bad. Bigger markets provide expertise, staff, policy participation and professionalisation. They can support brokers, counsel, data-centre investment, network groups and repeated engagement with LACNIC's policy system. Their operators often understand registry process and can help improve it. The danger is mistaking their ability to comply for proof that compliance is equally cheap elsewhere.
Small-island dependency is not a sentimental category. In a small market, a modest IPv4 block can support hotels, payments, public offices, health systems, schools, emergency services, local hosting or enterprise customers. A local provider may need to lease addresses because buying would consume cash needed for batteries, spares, backhaul or storm recovery. It may need to sell unused space to finance resilience. It may need reverse-DNS continuity for mail systems used by a public agency. It may need RPKI updates before an upstream will accept a route. The global market may call the block small. The local economy may not.
This is where conservation rhetoric can become regressive. If conservation means slower transfers, heavier need review, suspicion of leasing and reluctance to recognise market value, large incumbents can manage the friction. Small entrants and local operators bear it. If conservation means accurate records, clear responsibility chains, published documentation standards and predictable timing, small operators benefit. The word is the same; the incidence is opposite.
Language adds another layer. LACNIC's policy system works across Spanish, Portuguese and English. That is a strength. But practical policy participation is still shaped by language communities, meetings, mailing-list habits, translation timing and informal networks. An English-speaking Caribbean operator may be technically competent and still peripheral to the main debate. A Portuguese-speaking participant may follow different channels from a Spanish-speaking one. A small operator may not have time to read every proposal in every language. Silence in such a region should not be treated as proof that costs are acceptable.
The right conservation standard is not regional paternalism. LACNIC should not block small sellers from selling on the theory that resources should stay local, nor should it subsidise every operator facing a hard bank or old documents. It should make the rules usable without hidden scale advantages. That means timing data, plain-language guidance, multilingual operational examples, clear cure paths, proportionate documentation, payment procedures that recognise currency friction, and a distinction between record risk and market discomfort.
Patience is capital. Policies that require more of it distribute power toward those who already have enough.
Idle inventory and the moralising of scarcity
Dormant or underused IPv4 inventory is the object on which conservation rhetoric often hardens. If a holder has more addresses than it currently uses, is it conserving operational optionality or hoarding a public resource? If it leases part of the space, is it putting idle capacity to work or extracting rent from a historical allocation? If it sells a block, is it freeing resources for productive use or monetising something it received under earlier rules? The answers vary by case. The rhetoric often does not.
There is a real public concern. Idle space with stale contacts can be hijacked. Old holders may not monitor abuse. A party may claim control through outdated credentials. A holder may lease through a chain that hides the operational user. A block may carry reputation problems. A buyer may receive less than it thought because the record is unclear. A market in which every holder can quietly monetise old inventory without responsibility would damage the very trust that makes addresses usable.
There is also a real economic benefit to liquidity. If addresses are dormant because the holder no longer needs them, a transfer or lease can move them toward active use. If a public institution has old space but no current use, regularisation may make the record safer. If a small operator can monetise unused inventory, it may fund the network it still runs. If a leasing arrangement supports a temporary customer need, it may be more efficient than a permanent purchase. Scarcity is not conserved by trapping capacity in inactive hands.
The moral temptation is to describe every market use of idle inventory as hoarding. That word does too much work. A holder maintaining spare capacity for migration, resilience, customer churn, disaster recovery, carrier-grade NAT buffers, merger integration or future uncertainty is not necessarily hoarding. A lessor that supports route-origin authorisation, reverse DNS, abuse contacts and end-of-term cleanup may be providing useful liquidity. A seller that releases space to a buyer with real demand may be improving utilisation. A broker that reduces search and documentation costs may be market infrastructure rather than parasite.
The same caution applies in the other direction. Not every lease is productive. Not every broker is useful. Not every holder deserves trust. Some inventory is warehoused purely for price appreciation. Some leases create abuse externalities. Some subleases conceal practical control. Some sellers try to monetise resources they cannot lawfully transfer. Some buyers use shell entities or misleading justifications. A serious registry must act against those risks.
The difference is evidence, not vocabulary. LACNIC should ask whether the holder is accountable, contacts work, delegated use is visible enough, route authority is current, reverse DNS is coherent, reputation damage is disclosed where relevant, disputes are contained, and transfers comply with adopted rules. It should not treat scarcity rent itself as proof of misconduct. Scarcity rent is what happens when demand persists after supply stops. The question is whether the rent-bearing arrangement internalises responsibility.
Moralising scarcity can preserve incumbency. If lawful liquidity is described as suspicious, the safest operator is the one that already has addresses and does nothing. Late entrants, small networks and cash-constrained providers are then told to wait, justify, renumber, deploy IPv6 faster, use NAT more heavily or accept upstream dependency. Conservation becomes a discipline imposed on those who arrived late, while those who arrived early enjoy quiet possession and the option value of doing nothing.
That is not stewardship. It is historical allocation protected by moral language.
Legacy records as conservation stress tests
Legacy resources reveal the tension between accuracy and economic control. LACNIC defines legacy resources as IP addresses and ASNs assigned by InterNIC or IANA before the current regional registry system, specifically before December 28th 1997, and not later subject to a LACNIC membership agreement. For years, such records could seem like historical residue. After exhaustion, they are market-relevant assets, operational dependencies and possible security risks.
The economic problem is old paper meeting new value. A university may have changed legal status. A public telecom asset may have been privatised, split or absorbed. A government network may have moved between ministries. A family company may have gone through succession. A research network may have become a commercial provider. A cooperative may have board minutes rather than polished corporate filings. A dissolved parent may still sit in an old record. When IPv4 was abundant, these defects were annoying. When IPv4 is scarce, they shape prices and incentives.
In June 2026, LACNIC published a call to organisations holding legacy IPv4 resources, initiating an institutional process to formalise relationships, update associated information and support legitimate and secure utilisation. The notice said that organisations that do not contact LACNIC within the established deadline, or that cannot justify their right to use the resources after review, would lose registration services. This is a major post-exhaustion conservation action. It can clean the ledger. It can reduce hijack risk. It can make legacy blocks more marketable after regularisation. It can also impose heavy proof burdens on holders whose continuity is real but poorly archived.
Legacy regularisation is legitimate if it conserves title clarity and operational responsibility. It becomes dangerous if it turns modern paperwork expectations into a de facto confiscation of resources that have been used continuously but documented imperfectly. Public-sector continuity may be legally obvious inside a country and obscure to an external reviewer. A family succession may require local probate evidence. A university merger may have archival gaps. A state-owned enterprise may have old decrees rather than commercial contracts. Fraudsters should not be allowed through. Legitimate continuity should not be priced out by evidentiary perfection.
This is where patient finality matters. Patient, because old records require accommodation of local legal history, incomplete archives and public-sector forms. Final, because valuable resources cannot remain indefinitely in a fog. A registry that regularises clearly reduces the title discount. A registry that creates a cliff creates panic, rumours, defensive leasing and distressed sales. A holder that emerges with clean recognition has a stronger asset. A holder stuck in uncertain review loses bargaining power.
The legacy process should therefore be explicit about evidence categories. What proves continuity? What proves authority? What substitutes are acceptable when old documents do not exist? How are public institutions treated? What happens during review to RPKI, reverse DNS, WHOIS, abuse contacts and transfers? What appeal or cure path exists if LACNIC doubts the claim? How are abandoned resources distinguished from weak archives? Which facts are missing when a claim fails?
Legacy title is not an exception to conservation rhetoric. It is its stress test. If conservation means cleaning the record, the process strengthens liquidity. If conservation means administrators deciding, case by case, which historical holders deserve modern recognition under unclear standards, it becomes a capital-control event.
RPKI, reverse DNS and the conservation of continuity
The most technical details often carry the greatest economic weight. A transferred or leased IPv4 block is not fully useful simply because a contract exists. It must be routable, supportable, contactable and trusted. RPKI, reverse DNS, WHOIS or RDAP data, abuse contacts and technical contacts are the ways registry recognition becomes operational confidence.
LACNIC's RPKI service has two modes. Its hosted service has operated since January 2011, allowing member organisations to perform RPKI tasks through MiLACNIC. Its delegated service has operated since December 2019, allowing an organisation to run its own certificate authority and maintain its private key for signing cryptographic material. These details matter because route-origin authorisation is part of the quality of a block. A buyer whose ROAs cannot be created or updated quickly may face upstream reluctance, customer concern or internal security friction. A lessee dependent on an inattentive lessor may discover that a route is authorised only when someone else acts.
Reverse DNS is less fashionable but still commercially important. LACNIC's DNS servers handle reverse resolution for IP addresses assigned to ISPs and other organisations in the region. Reverse delegation is registered through MiLACNIC, and DNS servers are updated daily from system records. For IPv4, LACNIC's material explains that reverse delegation must respect byte boundaries and that /24 or /16 delegations can be registered in its DNS servers. Mail systems, monitoring, diagnostics, security tools, customer platforms and reputation work still care about reverse resolution. A block whose reverse DNS cannot move cleanly is not the same economic product as a block that can.
Inter-regional transfers expose the continuity risk. LACNIC warns that because resources move from one RIR to another, services such as reverse DNS or RPKI may be affected and not immediately available. That warning is operationally modest and economically large. It means a transfer may be legally and registrationally approved while part of the usable service layer still lags. Escrow agreements, customer migrations and route announcements must account for the difference.
This is a proper conservation concern. LACNIC should conserve continuity of route authority, reverse resolution and contact responsibility. It should publish what events normally happen in which order: holder verification, recipient review, legal document acceptance, payment, transfer agreement or order, record update, public log entry, RPKI availability, reverse-DNS delegation, contact transition and service-agreement activation. It should distinguish what LACNIC controls, what another registry controls and what the parties must prepare before closing.
The danger is using technical continuity as leverage for unrelated concerns. If RPKI or reverse DNS can be interrupted because of a curable billing delay, vague compliance discomfort or moral suspicion of leasing, a non-routing issue becomes a routing-security problem. If contact changes invite broad entitlement review, records remain stale. If delegated use cannot be reflected because leasing is disliked, abuse and route responsibility become less clear. Conservation of continuity requires proportionate remedies.
A useful approach is a service-continuity firewall. Fraud, account compromise, court order, disputed authority, sanctions prohibition, late payment, incomplete documentation, routine contact update, merger review, lease-related delegation and transfer settlement should not all have the same operational consequences. Some states require locks. Some require preserving the last verified state. Some require annotation. Some require a cure deadline. Some should not affect RPKI or reverse DNS at all. The blast radius should match the risk.
This approach protects both the registry and the market. It lets LACNIC be strict where service continuity would otherwise become false, while preventing operational systems from becoming quiet weapons in economic disputes.
The attention market behind conservation policy
Conservation rhetoric does not operate only through written rules. It hardens through procedure. LACNIC's policy system is public, multilingual and structured. Proposals can be discussed on the mailing list, taken to the Public Policy Forum, assessed for consensus, put through last call, reviewed again and ratified by the board. Proposal pages show versions, authors, dates, language access, downloadable files, diffs and status paths. This is valuable institutional memory.
It is also an attention market. A proposal about sub-assignment, waitlist priority or transfer restrictions may be technically public, but only a subset of the region will read it early, compare versions, understand the policy manual references, attend or follow the forum, write a substantiated objection, return during last call and monitor implementation. The cost of doing so is low for a repeat participant and high for a small operator. Open access is not the same as equal participation.
This matters because conservation claims often begin as frames in policy debate. A proposal can be introduced as preventing speculation, protecting the community, encouraging IPv6, preserving fairness, avoiding leakage to other regions or keeping resources tied to need. These phrases shape the burden of proof. The participant defending liquidity must appear commercial. The participant defending restriction appears prudent. The operator describing a customer deadline sounds self-interested. The incumbent defending "the community" sounds civic.
The mailing-list record can correct this if participation is broad and evidence-rich. It can expose whether a rule hurts small ISPs, whether a public log creates adoption fear, whether a transfer lock reduces flipping or traps legitimate restructuring, whether an IPv6 requirement rewards real deployment or penalises transitional networks, whether a broker restriction reduces abuse or raises search costs. But if participation is concentrated among those with time, staff and institutional familiarity, the conservation frame may pass as consensus while incidence remains under-analysed.
The region's language and scale differences sharpen the issue. Spanish, Portuguese and English publication is necessary but not sufficient. Translation timing, informal discussion, forum attendance, travel budgets, social familiarity and technical confidence all affect who speaks. Silence from the Caribbean or from small continental operators may mean agreement. It may also mean no one had time to translate the rule into business risk before the consensus clock moved.
The same lesson appears in other registry controversies: governance legitimacy is not a slogan; it is the relationship between decision rights and those who bear consequences. Elections, courts, legacy claims and capital-flow decisions all show that formal process can be necessary without being sufficient. A policy system can be open and still overweight the voices that can afford persistence. A registry can follow procedure and still fail to measure who pays for the rule.
For high-incidence conservation policies, the consensus record should include incidence notes. Which categories of actors commented? Which were absent? Did small operators, Caribbean networks, NIR-related communities, brokers, lease users, public-sector holders, legacy holders and transfer recipients appear? What changed between versions? Which costs were considered? Which implementation metrics will be published after ratification? Absence should not veto policy, but it should temper claims that the region has spoken with a single economic voice.
Conservation is too powerful a word to be left to whoever has the most procedural stamina.
Lessons from adjacent registry conflicts
LACNIC does not need to copy the crisis history of any other regional registry to learn from it. The point is not to import another region's litigation, election disputes or institutional failures. The useful lesson is narrower: once IPv4 scarcity turns registry recognition into economic infrastructure, governance defects become capital events.
Recognition-gated capital flow is the first lesson. A registry may describe itself as a technical ledger, but when scarce address value depends on its recognition, every delay, refusal, lock, documentation demand and policy interpretation affects capital. The institution need not intend to control markets. Markets will price it as a control point. If the control point is narrow and predictable, it reduces risk. If it is broad and discretionary, it becomes a tax.
Mailing-list attention is the second lesson. Address policy is often made through open discussion among those who show up. That is better than closed bureaucracy, but it is not equal incidence analysis. Repeat participants and large organisations can convert attention into rulemaking influence. Small operators, legacy holders and occasional market participants often discover rules when they transact, not when they are debated. Conservation language can exploit that gap because it frames restriction as the default public good.
Election legitimacy is the third lesson, even when no election is directly at issue. A registry's authority rests on more than technical competence. Members and affected networks must believe that decision-making positions are filled and supervised through credible mechanisms. If participation is thin, representation contested or board action opaque, scarcity policy becomes easier to challenge. In a post-exhaustion market, legitimacy affects the value of the record.
Court continuity is the fourth lesson. Registries operate under legal systems. If a dispute, receivership, injunction, corporate challenge or public-law issue touches the registry, the continuity of services becomes economically important. RPKI, reverse DNS, transfers and account access cannot depend on vague institutional confidence. They need clear service-continuity rules. LACNIC's legal environment is not the same as any other region's, but the general lesson holds: a registry should insulate ledger continuity before a crisis tests it.
Legacy title is the fifth lesson. Old allocations become contested when prices rise. A block that once sat unnoticed can suddenly attract claims, audits, brokers, buyers and legal attention. If regularisation is clear, legacy space becomes cleaner and more useful. If it is opaque or punitive, holders panic and buyers discount. Conservation of old records must be patient enough to handle history and firm enough to produce finality.
These lessons converge on one proposition: a registry should be a reliable settlement utility, not an all-purpose guardian of market morality. It must protect the ledger, but the more it tries to decide which economic outcomes are virtuous, the more fragile its legitimacy becomes.
Translating conservation into mechanisms
The practical test for any conservation claim is translation. A policy participant, staff member or board should be able to complete the sentence: "This measure conserves..." with a concrete object. Does it conserve a residual pool? Then specify the pool, size, eligibility and expected depletion effect. Does it conserve registry accuracy? Then specify which record field, which responsibility path and which failure mode. Does it conserve routing security? Then specify ROA authority, RPKI continuity or route-origin risk. Does it conserve abuse accountability? Then specify the contact path. Does it conserve fairness? Then say whether fairness means queue order, equal opportunity, small-operator protection, anti-speculation, regional retention or something else.
If the object cannot be named, the claim should not carry policy weight. "Conservation" should not mean general institutional discomfort with movement. "Stewardship" should not mean broad discretion. "Community protection" should not mean protection of existing holders from liquidity. "IPv6 transition" should not mean punishing present IPv4 dependence. "Anti-hoarding" should not mean hostility to inventory, resilience buffers or lawful lease supply.
The second test is incidence. Who pays? A waiting-list rule may benefit later applicants and harm current demand. A transfer lock may reduce flipping and trap legitimate restructuring. A public sub-assignment log may improve traceability and deter adoption by small networks concerned about exposure. A need review may deter speculation and favour large buyers with better forecasts. A documentation requirement may prevent fraud and impose a fixed cost on small blocks. A payment rule may protect LACNIC and create foreign-currency friction. Every conservation mechanism has a balance sheet.
The third test is substitutability. If formal transfers are slow, will parties lease? If leasing is stigmatised, will they call it managed service? If sub-assignment logs are too burdensome, will they route through private letters? If legacy regularisation is cliff-like, will holders avoid contact until forced? If need review is too subjective, will buyers prefer acquisition structures? A rule that pushes behaviour into less visible forms may conserve rhetoric while damaging the record.
The fourth test is measurement. LACNIC should publish metrics where confidentiality allows. Median and long-tail processing times by transfer category. Frequency of supplemental documentation requests. Need-review approval and failure categories. Inter-RIR timing. Renewal-deadline effects. Payment delays. RPKI and reverse-DNS service lags after transfer. Sub-assignment log usage after implementation. Contact-quality failures. Legacy regularisation outcomes by evidence category. Public data does not need to reveal private prices to reduce uncertainty.
The fifth test is reviewability. If an applicant is delayed or refused, the reason should map to a category: missing authority, disputed holder status, legal prohibition, policy ineligibility, inadequate need justification, payment defect, incomplete documents, operational inconsistency, suspected fraud, inter-registry coordination or service-dependency lag. A vague request for more comfort is costly. A specific defect can be cured or priced.
The sixth test is least market distortion. If the risk is stale abuse contact, fix the contact. If the risk is false route authority, fix or lock the ROA path. If the risk is hidden operational user, require a responsibility signal. If the risk is speculation, define the behaviour and time limit. If the risk is price discomfort, admit that the concern is economic and ask whether the registry has a mandate. Conservation should be no broader than the risk.
These tests do not require LACNIC to abandon stewardship. They require stewardship to show its working.
Safeguards for a narrower regime
The first safeguard is a documentation map. LACNIC should publish ordinary evidence expectations for routine transfers, mergers, asset sales, share sales, public-sector succession, insolvency, name changes, relocations, account recovery, suspected fraud, legal disputes, legacy regularisation, inbound transfers and outbound transfers. Each category should include normal documents, acceptable substitutes, signature authority, translation expectations, timing targets, common defects and escalation paths. Hard cases will remain hard, but applicants should know what kind of hard they face.
The second safeguard is timing transparency. Processing statistics should separate applicant delay from registry review, document supplementation, payment delay, inter-registry coordination, legal hold, need-review failure, dispute review and operational-service lag. Medians are useful; tails are essential. The long tail is where escrows fail, sellers discount, buyers walk away and small operators run out of patience.
The third safeguard is scoped need review. LACNIC can require coherent use, real identity, ASN or network context where relevant, policy compliance and anti-sham evidence. It should state that review does not judge price, seller profit, buyer virtue, regional popularity, leasing morality or general discomfort with IPv4 commerce unless an adopted policy explicitly addresses the issue. A buyer denied recognition should know which fact failed, not merely that the case felt inconsistent with conservation.
The fourth safeguard is responsibility-chain visibility for delegated use. The recognised holder should remain accountable. The operational user should be reachable where use is material. Route-origin authority should be current. Reverse DNS should be coherent. Abuse contacts should work. End-of-use cleanup should be expected. Public logs should disclose only what traceability requires, not commercial terms. The aim is visibility without rent control.
The fifth safeguard is a service-continuity firewall. LACNIC should publish how account states, disputes, legal orders, payment defects, transfer reviews, legacy reviews, suspected fraud and contact problems affect WHOIS/RDAP publication, account access, RPKI, reverse DNS, transfer eligibility and support. The default should preserve the last verified operational state unless the specific risk requires interruption. This protects downstream users and prevents unrelated disputes from becoming routing problems.
The sixth safeguard is legacy patience with finality. Legacy holders should face clear deadlines and evidence requirements, but also realistic categories for public bodies, universities, old companies, mergers, succession and continuous operational use. LACNIC should distinguish abandoned resources, false claims, weak archives and curable documentation gaps. The market needs final outcomes; legitimacy requires that old records are not judged solely by modern transaction paperwork.
The seventh safeguard is policy incidence reporting. High-impact proposals should include plain-language summaries of affected parties, changed burdens, expected benefits, likely substitutes, small-operator effects, language availability and post-implementation metrics. Consensus records should note underrepresented groups. Ratification should be followed by implementation feedback, not silence.
The eighth safeguard is a disciplined relationship with IPv6. LACNIC is right to promote IPv6 and to make IPv6 resources easy to obtain. IPv6 is the long-term architecture. But IPv6 advocacy should not be used to moralise lawful IPv4 liquidity. Networks can deploy IPv6 and still need IPv4 for customers, legacy applications, security appliances, mail systems, hosting, translation, public-sector procurement and global reachability. Blocking liquidity does not automatically accelerate transition. It can trap capital that would otherwise fund modernisation.
These safeguards are not radical. They are the ordinary governance equipment of a registry that recognises the economic consequences of scarcity.
The economics behind the word
Conservation is too important to abandon and too powerful to leave undefined. LACNIC should conserve the things only a registry can conserve: uniqueness, truthful registration, holder authority, operational contactability, responsible delegation, RPKI continuity, reverse-DNS coherence, transfer finality, dispute containment and public confidence in the record. Those are public goods in the practical Internet sense. They make networks cheaper to operate and markets safer to use.
It should not conserve the old allocation-era self-image by turning every market movement into a moral test. The free pool is gone. A waitlist measured in years cannot satisfy present demand. IPv4 remains necessary during a long and uneven transition. Scarce address capacity has become working capital for operators. Transfers, leases, acquisitions and sub-assignments are not departures from reality; they are reality. The policy choice is whether they become visible, accountable and predictable, or hidden, expensive and dependent on private knowledge.
The distributional stakes are clear. Delay benefits those with inventory and patience. Documentation opacity benefits repeat players. Subjective need review benefits conventional business plans. Anti-leasing moralism benefits those able to buy outright or already holding addresses. Weak logs benefit insiders with market memory. Sudden legacy processes benefit buyers who can discount uncertainty. Payment rigidity benefits those with easy dollar access. None of these effects requires bad faith. They follow from institutional design.
The strongest argument for LACNIC is not that it can rise above economics. It cannot. Exhaustion made that impossible. The strongest argument is that it can make its economic role narrow, explicit and accountable. It can say: we verify facts; we protect the ledger; we publish timing; we distinguish risks; we keep operational services continuous; we measure incidence; we do not moralise price; we do not confuse IPv6 advocacy with obstruction of lawful IPv4 movement.
That posture would serve small operators better than paternalistic restriction. A Caribbean provider does not need a lecture on scarcity when a customer deadline arrives. It needs a clear path to recognised capacity or a visible responsibility chain for leased use. A public university does not need to be told that IPv4 is old. It needs old records regularised without losing operational continuity. A regional buyer does not need the registry to approve its business plan. It needs to know which facts must be proved for transfer recognition. A seller with dormant inventory does not need virtue language. It needs a market that can distinguish clean title from fog.
Conservation began as a way to preserve access to a finite pool. After exhaustion, it must become a way to preserve trust while allowing scarce capacity to move. If it instead becomes a way to slow movement, shame liquidity and protect incumbency, it will conserve the wrong thing.
The better rule is simple. Conserve the record. Conserve responsibility. Conserve security. Conserve the path by which IPv6 grows without breaking today's customers. Do not conserve artificial illiquidity merely because it sounds more virtuous than a market.