The address bill that arrives before the customer does
The case for IPv4 leasing begins in a place where policy language is least useful: a customer deadline. A small access provider in the Caribbean has a hotel group that still needs public IPv4 for reservation systems, monitoring and guest services. A payments processor wants predictable reachability for old integrations. A public agency has firewalls and remote-access appliances that were procured around IPv4 assumptions and will not be redesigned before the contract starts. The provider has IPv6 in parts of the network, and carrier-grade NAT carries much of the residential base. None of that removes the immediate requirement. The customer does not ask for a lecture on transition. It asks whether the service will work.
Buying addresses would be cleaner in theory. It would also tie up cash before the revenue has arrived. A seller may want quick settlement in dollars. A bank may ask why a small island company is wiring money to a broker or foreign holder for an asset that does not resemble normal equipment. The legal review may take longer than the sales cycle. The transfer may require documents from companies in different jurisdictions. The provider may be choosing between IPv4 capacity and batteries, backhaul, spares or tower work before hurricane season. Scarcity does not arrive as an abstract curve. It arrives as a queue of bills.
The LACNIC waiting list is not an operational answer to that queue. After its last available IPv4 block was assigned in August 2020, the registry was left with recovered space and a rationing mechanism. LACNIC has said that the last request on the list faces an estimated wait of at least eighteen years and can receive at most 1,024 addresses, with the timing uncertain because future recoveries cannot be predicted. An applicant must already have IPv6 resources. Recovered space is quarantined before release, and the recipient may still have to rehabilitate reputation problems. A /24 can matter. It cannot rescue a contract that starts next month.
Leasing enters through that gap. It turns an address block from a capital purchase into a monthly input, closer to bandwidth, colocation or a managed service. It lets a network match cost to revenue, bridge a migration, serve a temporary project or avoid spending scarce cash on an illiquid asset. For a large company, this may be convenience. For a smaller operator, it can be market access.
Yet the same transaction creates a second economy beside the public registry record. The registered holder remains in the book. The operating user may be someone else. The route may be originated by a different ASN. Reverse DNS may be handled by a third party. Abuse complaints may need to reach the lessee rather than the holder. RPKI authorisation must follow operational use, not simply the name in the registry. A lease is therefore both working-capital relief and shadow allocation: scarce productive capacity is allocated by contract while recognised holdership stays where it was.
That is not a scandal by itself. Nor is it a footnote to the transfer market. Leasing is the private answer to two public mechanisms that cannot satisfy every case: fresh allocation from a pool that has gone, and permanent transfer of recognised holdership. In the LACNIC region, the answer is shaped by small-island dependency, hard-currency friction, large-country demand, brokers, language, legal-document variation and the long tail of IPv4 demand. The institutional question is narrow but important: how can the registry make responsibility visible without trying to regulate rent?
Scarcity moved allocation into contracts
In the old allocation story, the registry and the economic user were usually close together. A network justified need, received resources, recorded assignments, maintained contacts and used the addresses in its own service or in ordinary downstream customer relationships. LACNIC's authority was rationing authority over a common pool. The registry asked whether an applicant qualified because the registry was distributing scarce stock at administrative cost. The public record and the productive use of the resource were not identical, but they usually pointed in the same direction.
Exhaustion loosened that relationship. LACNIC remains the regional Internet registry for Latin America and much of the Caribbean. It still provides registration services, Whois and RDAP access, reverse DNS, RPKI and the policy machinery through which rules can be changed. Those functions become more important after exhaustion, not less. But the registry no longer supplies most marginal IPv4 demand. The next usable address for a business normally comes from another holder, an upstream provider, a transfer, a corporate transaction, a recovered block, a brokered arrangement, a lease or an engineering workaround.
Leasing is the most revealing case because the market moves use while the record often stands still. In a permanent transfer, recognised holdership eventually changes. In a lease, the holder remains the holder and grants another party the right to use the block for a term. The agreement may include payment, route-origin authorisation, acceptable-use limits, reverse-DNS duties, renewal rights, deposits, reputation cleanup and termination rules. Economically, the lease allocates scarce address capacity. Administratively, it may leave the registry's holder field unchanged.
That distinction is the heart of the shadow market. The word "allocation" has a policy meaning in registry vocabulary, and a lease is not a registry allocation. But it allocates productive capacity all the same. A hosting company can sell servers because leased addresses are available. A Caribbean ISP can win a business customer because someone else's registered block can be routed under contract. A regional cloud provider can bridge demand while deciding whether to buy. A holder can earn yield from dormant inventory without surrendering future optionality.
This private allocation layer is not foreign to the internet. Downstream assignment has always existed. Broadband customers receive addresses. Cloud users receive addresses. Universities and carriers assign pools internally. The sensitive point is reached when the operational user is no longer a minor customer inside the holder's network but an independent operator using the space as a production input. At that point the registry record may identify the first responsible party but not the party that can fix a route, answer a complaint or explain a customer-impacting failure.
Markets price that gap. A lessee discounts space where route-origin support is slow or reverse DNS is uncertain. A holder charges more for clean inventory and responsive operations. A broker earns money by reducing search costs, or by monetising opacity when search is hard. Upstreams ask for letters of authority or RPKI state. Customers ask why a prefix geolocates to the wrong country or carries old blacklist damage. Auditors ask whether address capacity is controlled, rented or merely hoped for.
LACNIC should read this as institutional information. Leasing grows when operators need continuity more than a symbol of holdership, when they lack capital to buy, when transfers are too slow for a temporary need, when holders prefer income to sale and when small networks cannot wait for recovered space. None of these motives proves evasion. Together they show that post-exhaustion IPv4 is allocated through more than one channel. The registry's problem is not that private contracts exist. It is that the public record must still be useful when they do.
The missing actor in the public record
The decisive distinction in a leasing economy is holder versus user. The holder is recognised by the registry, controls authenticated changes and remains accountable for the resource relationship. The user is the network placing the addresses into production, assigning them to customers, relying on their routes, receiving revenue from their use and suffering if the block is withdrawn, misrouted or polluted. Sometimes the two are the same. A lease separates them.
The separation can be efficient. A serious lessor may be better at registry-facing work than the lessee: maintaining account security, issuing route authorisations, managing reverse DNS, monitoring reputation, coordinating geolocation correction and handling abuse escalation. A lessee may need capacity for two years but not want to buy. A holder may have space that would otherwise remain idle. There is social value in putting dormant addresses to work, especially in a region where waiting-list supply is symbolic rather than practical.
The separation can also hide responsibility. The public record may show the holder's corporate name and contacts. The route may originate from the lessee's ASN. The abuse traffic may come from a downstream customer the holder has never met. Reverse DNS may be delegated to the lessee, maintained by the lessor, outsourced to a managed provider or left stale. RPKI may sit in the holder's account while the operational network waits for a new ROA. A broker may have arranged the lease without controlling the holder account. A sublease may add another layer. The record can be accurate in the narrow sense and still fail the operational question: who can act now?
That missing actor is what LACNIC should care about. Not the rent. Not the lessor's margin. Not whether leasing feels too financial. The legitimate public concern is that delegation should not break the functions that make number resources usable: contactability, routing authority, reverse resolution, abuse handling, uniqueness, dispute clarity and the ability to distinguish authorised use from opportunistic use.
The registry need not become a warehouse of lease contracts. It should not ask for prices, customer lists, side letters, financing terms or every downstream assignment. That would create confidentiality problems and invite arbitrary economic judgement. The useful question is narrower. When delegated use is large enough, independent enough, route-originating enough or long-lived enough to affect third parties, can the public layer show the responsible path from holder to operator?
The answer will differ by case. A one-month migration bridge is not the same as a multi-year lease of a route-originating /20 to an independent provider. A managed hosting customer is not the same as a brokered sublease. An ordinary retail assignment inside an access network is not a shadow allocation problem. A lease that transfers nearly all practical control for years may deserve different treatment from a small temporary delegation. The categories matter because blunt rules push legitimate use into euphemism.
LACNIC's existing services already point in the right direction. Whois and RDAP make registration data queryable. Reverse DNS ties address resources to operational naming. RPKI lets holders publish routing-security assertions. Transfer procedures recognise when permanent movement of recognised holdership occurs. Leasing extends the same logic into a more awkward zone. If the holder and user diverge, coordination needs become more precise, not less important.
The public record does not need to tell the whole commercial story. It needs to prevent the commercial story from making responsibility disappear. A market in which every operator relies on private documents, broker promises and informal reputation is expensive and fragile. A registry that sees enough to preserve the responsibility chain can remain narrow while keeping its record relevant.
Working capital, not merely cheaper addresses
IPv4 leasing is often described as a cheaper alternative to buying. That is sometimes true and often incomplete. The better description is working-capital relief. A permanent purchase brings a large payment forward. It requires due diligence, legal confidence in the seller, settlement capability, registry recognition and operational transition before the addresses can support revenue. A lease turns the same capacity into recurring expense. It can be matched to a customer contract, a product launch, a migration window or a seasonal demand spike.
The difference matters more in Latin America and the Caribbean than it might in a spreadsheet built for a large North American or European cloud buyer. A company may earn in local currency and pay for address capacity in dollars. Inflation may make cash preservation rational. Currency controls may complicate settlement. Banks may ask why a foreign counterparty is receiving a payment for internet numbers. Public agencies may require procurement steps that outlast the technical need. A small island operator may face correspondent-banking delays or de-risking that make an ordinary wire unpredictable. Address scarcity is therefore entangled with payment infrastructure.
Leasing converts those frictions into monthly risk. It is not free, and it is not always cheaper over the life of a service. A long lease can exceed the cost of purchase. Renewal can become expensive. The lessor may raise rates, sell the block or withdraw from the business. Yet the lease can match cash outflow to cash inflow. If a hotel group pays monthly, the address cost can be paid monthly. If a security customer needs a two-year deployment, a two-year lease may be rational. If a carrier needs overlap while customers move to IPv6-capable services or new addressing plans, permanent purchase may be wasteful.
The lease also buys a wrapper around the addresses. A good lessor does not merely point to a prefix. It supports RPKI changes, reverse-DNS delegation, abuse escalation, geolocation correction, reputation monitoring, evidence of authority and end-of-term cleanup. Those services are not decorative. They make the addresses bankable. A bare block with uncertain authorisation and stale reverse DNS may be cheap only because the lessee will pay later in outages, support calls and customer distrust.
This explains why a first-party lessor with clean space and disciplined operations can charge more than a brokered chain with weaker guarantees. The customer is not paying only for numbers. It is paying to avoid the hidden costs of uncertainty: an upstream that rejects a route, a mailbox that lands on a blacklist, a government customer that refuses a geolocation mismatch, a support ticket that nobody in the chain can answer, a lease end that leaves old route objects behind.
The danger is dependency. A lessee may build customer promises on addresses it does not control. It may have weak renewal rights. It may lack direct registry standing. It may depend on the holder's account status, staff availability, payment behaviour and policy exposure. If a dispute between the holder and another party affects RPKI or reverse DNS, the lessee's customers may be harmed without having caused the defect. If a broker disappears, the lessee may discover that the most important relationship was never with the broker at all.
The public-policy issue is therefore not the existence of rent, but the visibility of dependency. An operator can choose to lease if it understands who controls route-origin authorisation, who can change reverse DNS, who answers abuse, whether the holder is in good standing, whether the block has reputation damage, whether subleasing is permitted and what happens at termination. It cannot price what it cannot see. The registry cannot answer every commercial question. It can make enough of the responsibility chain visible that the private bargain is not blind.
Island economics and the price of waiting
LACNIC's region makes the leasing question less generic than a global IPv4 essay would suggest. The region contains large economies with dense technical communities, deep corporate buyers and established brokers. It also contains small Caribbean states and territories where network operations are thinly staffed, import-dependent, disaster-exposed and often tied to sectors such as tourism, finance, public administration, education, maritime services and remote work. A few hundred public IPv4 addresses can matter materially to local resilience.
For a small island operator, the cost of delay is different. A block that looks small by global standards may support a cluster of hotels, a payments firm, health systems, public offices or an emergency-service platform. A missed renewal, routing defect or reverse-DNS failure can become a local economic problem, not just an engineering inconvenience. When staff are few, the same engineer may handle procurement, peering, outage response, policy paperwork and customer escalation. A rule that looks modest on a mailing list can become an expensive fixed cost in a small office.
Currency and banking frictions sharpen that cost. The operator may not have easy access to dollar settlement. A foreign payment may require approvals. A small invoice may attract a fee large enough to matter. A de-risking event at a correspondent bank may interrupt a payment channel. A lessor in another country may not understand why a payment that looks late is actually trapped in banking process. The result can be operational risk created by financial plumbing.
Leasing can help precisely because it lowers the entry cost. It lets a small network buy time. But it can also deepen dependency if the lease chain is opaque. A larger operator can demand direct holder confirmation, test ROAs before production and negotiate exit rights. A small operator may accept the cheaper offer because it is available. It may later learn that reverse DNS takes weeks, RPKI changes depend on a party it never met, reputation is poor or renewal terms have changed because a broker's supplier changed.
This is not an argument for a protected island quota or for excusing every payment failure. It is an argument for institutional design that recognises how scale changes friction. Clear lease-hygiene guidance, multilingual explanations, predictable support channels and public responsibility signals reduce small-operator dependency. Suspicion without usable alternatives does the opposite. Small networks will still lease, but through channels that may be harder to verify.
The same point applies to policy participation. English-speaking Caribbean operators may have less regular presence in a region where Spanish and Portuguese dominate much public discussion. Public-sector networks and cooperatives may not watch policy lists. A regional cloud customer using leased space may not know that a debate about delegated use can affect its continuity. Silence in an open process should not be read too quickly as indifference. It may reflect distance, language, staff scarcity or the rational choice to run the network rather than follow every procedural change.
LACNIC does not need to become a development bank to account for these realities. It can keep the registry problem narrow: make authority easier to verify, make operational contacts easier to find, ensure that route-origin and reverse-DNS responsibilities follow real use, and avoid remedies that turn a curable paperwork problem into a customer outage. The small-island case shows why leasing is not merely a speculative convenience. It is sometimes the way scarcity is kept from becoming an entry barrier.
Brokers and the economics of opacity
Brokers exist because IPv4 supply is fragmented and information is uneven. A holder may not know who needs addresses. A lessee may not know which holders have clean inventory. A buyer may need a size, term or region that no obvious seller advertises. A broker can search, match, translate, hold deposits, draft authorisation documents and explain the steps by which addresses become usable. In a thin market, that work has value.
The value can be benign. A good broker reduces transaction costs, filters unserious offers, checks authority, warns about reputation, coordinates documents and keeps a small operator from walking into a broken chain. In that role the broker is a market-maker for scarce infrastructure. It can be especially useful where operators lack time, language reach or specialised legal support.
The incentive can also be misaligned. A broker is paid to close. It may prefer speed over diligence. It may know only the immediate supplier, not the original holder. It may package several blocks from different holders as a single service. It may rely on letters of authority without confirming the holder's authenticated control. It may understate RPKI limits, reverse-DNS delays, geolocation damage or termination risk. A strong broker sells knowledge. A weak one sells the customer's ignorance back to the customer.
Lease chains create three risks. The first is authority risk: does the party selling use actually have permission from the recognised holder, and does that permission cover the route, term, origin ASN, region and customer class? The second is service risk: who can create or amend ROAs, update reverse DNS, answer abuse reports and correct contacts? The third is continuity risk: what happens if an intermediary fails, is not paid, loses authority, changes its business model or disappears?
The longer the chain, the more diligence the lessee must do. Large buyers can demand direct holder confirmation, insist on test ROAs and require explicit operational service levels. Smaller networks often cannot. They face the oldest problem in markets: the participants least able to evaluate quality are the ones most likely to need the cheaper product. In leasing, the discount may be an undisclosed transfer of risk.
First-party leasing is therefore structurally different. If the registered holder leases directly and supports operations, fewer parties can disclaim responsibility. The lessee can evaluate one holder's registry standing, support competence and contract terms. The holder has a direct incentive to preserve reputation because the next customer will inherit the prefix history. The public record is closer to the operating reality. First-party leasing is not automatically safe, but it is easier to make accountable than a chain built on thin authority.
LACNIC should not try to license brokers or certify commercial deals. That would exceed the narrow registry function and could create a false impression of safety. The better route is to reduce the value of opacity. If lessees and upstreams can verify holder identity, route-origin authority, operational contacts, reverse-DNS delegation and dispute status, brokers must compete by adding real value rather than by controlling mystery. A transparent responsibility layer would not eliminate brokerage. It would discipline it.
The same logic explains why a public transfer interface is relevant without being sufficient. A contact-finding mechanism for permanent transfers recognises that scarcity creates market search. Leases create the same search problem, but with more operational ambiguity because recognised holdership may not move. A registry can remain outside the commercial bargain while still improving the facts that make the bargain trustworthy.
Route-origin authority is the title test in practice
For a leased prefix, route-origin authority is often the moment when paper meets the network. A contract can say that the lessee may use the block. An upstream still wants proof that the route is authorised. Operators increasingly look for RPKI validity or at least consistency between the claimed holder, the origin ASN and any supporting authorisation. The most carefully drafted lease is weak if the route cannot be made credible.
In a simple case, the holder publishes a ROA authorising the lessee's ASN, or the holder's upstream originates the route under a clear arrangement. In a less simple case, the lessee uses one upstream today and another tomorrow; the lessor must amend authorisations quickly. In a brokered case, the party speaking to the lessee may have no direct access to the holder account. In a sublease, the operational origin may be two steps away from the entity with registry credentials. Each additional handoff increases the chance that the route record and the commercial promise diverge.
RPKI changes the economics because it makes some hidden weaknesses visible. A stale or missing ROA may not have mattered much when route acceptance depended mostly on bilateral trust and loose paperwork. As route-origin validation becomes more common, an invalid route can create reachability problems that customers notice. A lessee may discover that the cheapest block is cheap because the holder cannot or will not provide timely RPKI support. A lessor may discover that poor route hygiene reduces the future value of its inventory.
Letters of authority still matter, but they are no substitute for an authenticated, maintained route-origin path. A letter can be forged, stale, overbroad or inconsistent with registry state. An upstream may accept it, reject it or require further confirmation. A public or authenticated signal tied to the recognised holder lowers this uncertainty. It does not need to publish the rent, the customer contract or the commercial motive. It needs to show that the origin is authorised by the party able to authorise it.
The route question also distinguishes temporary leasing from disguised permanence. A short migration bridge with tightly managed route state is one thing. A multi-year arrangement in which an independent operator originates a large block, assigns it to its own customers and bears all operational responsibility is another. The latter may still be a lease. But it has public effects closer to a durable reallocation of use. Treating both cases as invisible customer assignments makes the record less useful.
End-of-term hygiene matters as much as start-up. A stale ROA can leave an old origin looking authorised after the lease ends. A stale IRR object can confuse upstream filters. A stale geolocation correction can mislead customers. A stale reverse-DNS delegation can create mail and security problems. A responsible lease is temporary only if exit is operationally managed. Otherwise the next user inherits a residue of the last bargain.
For LACNIC, the route-origin lesson is not that every lease should be approved in advance. It is that material delegated use should have a verifiable authority path. The holder remains the anchor. The operational user should not be forced to depend on informal favours when a route change is needed. The public layer should make it hard for unauthorised users to masquerade as lessees and hard for authorised lessees to be stranded by invisible dependencies.
Reverse DNS, abuse and reputation are capital too
IPv4 addresses carry history. They may have been used for mail, hosting, residential broadband, VPNs, security scanning, content delivery, gambling, fraud, spam or nothing at all. They may appear in reputation databases. They may geolocate to the wrong country. They may have stale reverse zones. Their abuse contacts may be old. The market calls the result clean or dirty space, but the economic meaning is broader: reputation is a shadow balance sheet attached to the prefix.
Leasing makes that balance sheet more complex because use is temporary and responsibility is divided. The lessee wants the space to work now. The holder wants to preserve future value. The downstream customer may create abuse. The broker may have promised quick use without understanding the block's history. The next lessee may inherit the consequences. If nobody owns the cleanup duty, the prefix's value declines and the market compensates with deposits, discounts, shorter terms or stricter acceptable-use clauses.
Reverse DNS is often treated as a minor administrative service until it fails. For mail, enterprise security tools, monitoring systems and some customer audits, reverse resolution is part of trust. A lessee that cannot get reverse DNS changed may find the block technically routable but commercially weak. A holder that delegates reverse DNS without cleanup may leave stale names after the lease ends. A registry that treats reverse DNS as a simple holder service may miss the fact that operational responsibility has moved.
Abuse handling is similar. A public contact that reaches only the holder may be insufficient if the holder cannot investigate the lessee's downstream customer. A contact that reaches only the lessee may be insufficient if the holder must enforce the lease or terminate repeated abuse. The right model is layered responsibility: the holder remains accountable for the resource relationship, while the operational user is reachable for incidents it can actually fix. The exact public presentation can preserve privacy. The path must not be dead.
Reputation risk also makes subleasing more dangerous. A holder leases to a broker, the broker leases to an operator, the operator assigns space to a reseller, and the reseller serves end customers. Every layer can point to another. Meanwhile complaints accumulate. By the time a new lessee asks why the block is discounted, the old chain is gone. This is how shadow allocation becomes externality transfer. The party that captured the short-term revenue may not bear the long-term cost.
LACNIC should not become a reputation agency. It should not score blocks, adjudicate every abuse complaint or decide whether a customer class is respectable. But it can insist that contact data remains usable, that the holder cannot disclaim all responsibility merely because a lease exists, and that material delegated use has an escalation path. These are record-truth questions, not content regulation.
The market will then do much of the work. Lessors with clean inventory and reliable operations will charge a premium. Lessees that create abuse will face higher deposits or termination. Brokers with thin chains will be discounted. Upstreams will prefer prefixes with coherent authorisation. The registry's job is not to replace that pricing. It is to make sure the facts needed for pricing are not hidden behind a public record that names only the first actor.
Shadow allocation without the theatre
"Shadow allocation" can sound accusatory. It should be treated as descriptive. The shadow is the difference between recognised holdership and actual productive use. In a lease, the holder stays on the record while another party uses the addresses for a term. The market has allocated scarce capacity by contract rather than by registry transfer. That is shadow allocation even when the contract is lawful, the holder is accountable and the lessee is responsible.
The category is useful because it separates two questions that are too often blurred. The first is whether leasing exists because operators need flexibility, capital efficiency and temporary capacity. It does. The second is whether the public record remains truthful enough when use moves outside formal transfer recognition. That depends on the responsibility chain. Condemning all leasing misses the first fact. Celebrating all leasing misses the second.
There are weak forms. A party may route space without credible permission. A broker may sell a block it does not control. A holder may leave stale contacts and collect rent while abuse accumulates. A lessee may sublease into high-abuse services and disappear. A lease may be used to avoid a transfer that would better reflect permanent control. A forged authorisation may turn a hijack into something that looks like a commercial delegation. These cases damage both market trust and the registry record.
There are strong forms. A holder leases directly to an operating network under clear terms. The lessee's origin ASN is authorised. Reverse DNS is delegated or maintained under service commitments. Abuse reports reach a mailbox that can act. Reputation history is disclosed where relevant. Renewal and termination are clear. Exit cleanup removes stale route and DNS state. The arrangement is still a private allocation of use, but it is not a public-record failure.
The difference is responsibility, not virtue. A registry that treats rent as the problem will push serious lessors to rename leases as managed services, customer assignments or partnerships. A market that treats responsibility as optional will reward cheap opacity until failures accumulate. The middle position is more durable: acknowledge that leasing is a scarcity instrument, then make the responsibilities that affect third parties visible and enforceable through ordinary registry expectations.
This approach also avoids false precision. No public system will reveal every lease. Some assignments are too small. Some customer relationships are legitimately confidential. Some operational delegation happens inside normal network service and should not carry registry process. The aim is not a perfect map. It is to prevent the registry ledger from becoming irrelevant where delegated use is material.
The phrase "shadow allocation" is valuable only if it leads to better design. It should not be used to dramatise leasing. It should be used to ask practical questions. Does the holder remain accountable? Can the operator be reached? Is the route authorised? Is reverse DNS coherent? Is abuse escalation real? Is the arrangement temporary in operation, not only in contract? Are remedies proportionate if something breaks? If those questions are answered, the shadow becomes less dangerous.
Large-market gravity inside a regional registry
The LACNIC region is not a flat market. Brazil, Mexico, Argentina, Chile, Colombia, Peru and other large economies contain carriers, hosting providers, cloud customers, managed-security firms, enterprise networks, government projects and data-centre growth. These actors can absorb more IPv4, pay higher prices, hire specialist counsel, maintain broker relationships and follow policy detail. Their demand shapes the terms available to smaller networks.
Large demand is not illegitimate. A cloud platform serving regional customers may need IPv4 continuity. A hosting provider may put dormant space to productive use. A carrier may need migration overlap. A content platform may improve local performance. Liquidity allows underused capacity to move toward use. The mistake is not that large actors lease or buy. The risk is that their gravity sets the market's default conditions.
Large operators can demand cleaner space, stronger ROAs, better reverse-DNS support, longer terms and better renewal options. They can buy when leasing becomes expensive. They can hold spare inventory. They can verify broker claims. A small access provider competing for the same pool may receive weaker blocks, shorter terms or more reputation work. If formal transfers are slow or uncertain, large actors can route around the process more easily. If leases are opaque, large actors can afford to investigate them. Opacity is regressive in precisely this way.
LACNIC's transfer architecture influences the lease market even when leases are outside transfer recognition. If transfers are predictable, a firm that needs long-term control may buy. If transfers are slow, document-heavy or uncertain, even long-term users may lease. If need justification is interpreted narrowly, some buyers will prefer operational use without holdership. If inter-regional transfers involve uncertain timing or service transitions, a lease may become the bridge or the substitute. Leasing is not merely an alternative to transfers. It is partly a derivative of transfer design.
Large-market pull also affects regional politics. If address capacity appears to flow toward the largest economies or global platforms, smaller members will ask whether scarcity is becoming extraction. If the answer is to discourage leasing broadly, small operators that depend on leases may suffer first. If the answer is to make responsibility and flows more legible, the debate can focus on actual harms: stale contacts, weak route authority, poor reverse DNS, broker-chain failures, transfer friction and unequal access to clean inventory.
The registry should not publish lease prices or private counterparties. It can still improve aggregate visibility. It can report transfer timing, common documentation bottlenecks, operational-contact defects, reverse-DNS delays, route-authorisation problems and support categories related to delegated use. It can distinguish market facts from anecdotes. Scarcity politics becomes less dangerous when the facts are strong enough to discipline slogans.
This is especially important because a regional registry is not a competition regulator. It cannot decide that large markets should get less capacity or small markets should pay less. It can reduce avoidable transaction costs that hurt smaller actors more. In practice that means predictable processes, clearer guidance, better public-contact models and a responsibility chain that does not require every lessee to become an expert investigator.
Why transfers cannot absorb every lease
It is tempting to say that any serious long-term need should become a transfer. Sometimes it should. If practical control has permanently moved, if the user bears the economic risk indefinitely and if the holder's role has become only nominal, transfer recognition may better match reality. But this answer is incomplete. Transfers solve finality. Leases solve timing, financing and flexibility. The difference is not semantic.
A transfer is a closing. It requires confidence in the seller, legal authority, documents, settlement and registry recognition. It gives the buyer cleaner control but also concentrates registry-facing obligations. It may be the right choice for a business that needs durable capacity, has capital available and can wait for completion. It is less attractive for a temporary project, a customer trial, a migration overlap, a public-agency contract with uncertain renewal or a small operator that cannot afford to trap cash in a block.
The lease also allows specialised risk allocation. A holder experienced in RPKI, reverse DNS, abuse handling and reputation can keep the registry-facing layer while the lessee focuses on service delivery. This is not always inferior to direct holding. For some small operators, direct holdership may be symbolically satisfying but operationally costly if it brings documentation, payment, policy and account-security burdens that the operator is poorly staffed to manage. Leasing can be a rational division of labour.
Transfers and leases also respond differently to uncertainty. A company expecting growth may lease while it tests demand. A network migrating away from IPv4 may lease for overlap rather than buy an asset it hopes to use less. A company in an unstable currency environment may prefer recurring dollar exposure to a large upfront payment. A public-sector buyer may not be able to purchase address assets quickly but may procure a service that includes address use. These cases are not attempts to evade the registry record. They are economic choices under constraint.
That said, transfers discipline leases. If transfer paths are clear, timely and objective, leases will be used mainly where they make economic sense: temporary need, cash-flow management, specialist operation, or transitional capacity. If transfer paths are uncertain or heavy, leases will also absorb demand that should have moved into recognised holdership. A growing shadow market can therefore signal not only private preference but public friction.
The boundary should be based on control and third-party risk. A lease that leaves the holder actively responsible, supports the operator's route and maintains public contacts can remain a lease. A lease that quietly transfers all practical control for years, allows subleasing, leaves the holder passive and makes the public record misleading deserves closer scrutiny. The point is not to force every lease into transfer. It is to prevent transfer-like control from hiding indefinitely behind lease vocabulary.
This boundary is where LACNIC can be useful without becoming a price regulator. It can publish criteria for material delegated use: duration, block size, independent route origination, operational control, subleasing, contact responsibility and end-of-term process. It can say which facts matter because they affect the public record. It can leave the commercial bargain alone.
Visibility without rent control
The hardest line is the one between making delegated use visible and regulating rent. Scarcity invites anger because prices rise and holders can earn from space received under older conditions. It is easy to treat rent itself as suspicious. That is poor institutional economics. A scarce input with continuing demand will produce rent through purchase, lease, bundled service or strategic inventory. The useful question is whether the rent-bearing arrangement internalises responsibility or exports risk to others.
LACNIC should not set prices, approve margins or decide whether a lease is commercially elegant. It is not a telecom regulator, competition authority, court, central bank or price commission. It does not know the fair monthly value of a /24 in a Caribbean hotel network, a Brazilian cloud cluster, an Argentine enterprise migration or a temporary security deployment. It cannot see every alternative, financing constraint, reputation issue or customer deadline. If it tries to police rent, serious actors will move into less visible structures and the public record will be poorer.
Business-model approval would be a similar mistake. A registry should not decide that buying is morally superior to leasing, that a holder is wrong to monetise idle space, that a lessee must prove sufficient IPv6 virtue before using IPv4, or that a large customer is less deserving than a small one unless a clear policy rule directly applies. IPv6 deployment is essential. It does not erase the current IPv4 cash-flow problem. Confusing transition advocacy with discretionary hostility to IPv4 use will not accelerate transition. It will raise the cost of shortage management.
Visibility is different. Visibility asks whether third parties can rely on the public and operational record. Who is the holder? Who is authorised to use the block? Who originates the route? Who can amend the ROA? Who controls reverse DNS? Who receives abuse complaints? Is the delegation temporary or effectively permanent? Is there a dispute? Are contacts reachable? Does the holder remain accountable? These are registry questions because they affect coordination.
The rule should be nexus. If an issue affects record truth, uniqueness, authority, route security, reverse DNS, contactability, dispute status or compliance with an adopted rule, LACNIC has a reason to act. If the issue is price, margin, commercial motive, customer preference or discomfort with leasing as a business, LACNIC should step back. A narrow registry can be strict. A broad registry becomes a gatekeeper whose discretion the market must price.
This line also protects against capture. Large incumbents could use anti-leasing arguments to raise rivals' costs. Brokers could use complexity to preserve margins. Lessors could invoke confidentiality to hide weak controls. Lessees could use "temporary use" to avoid accountability. A registry that focuses on responsibility rather than commercial virtue is harder for any one group to manipulate.
The line will not settle every case. Some leases will resemble transfers in substance. Some subleasing chains will conceal control. Some holders will resist contact updates. Some complainants will call ordinary leasing evasion because it suits them. Hard cases need evidence, categories and review, not theatrical language. The more clearly LACNIC defines the responsibility facts that matter, the less it must rely on instinct.
A responsibility chain, not a contract registry
A workable leasing posture starts with the chain. The first link is recognised holdership. The holder remains accountable for the registry relationship, account authority, resource status and the truthfulness of the public layer. Leasing cannot become a way to collect income while disclaiming responsibility. If the holder remains the holder, it remains the public anchor.
The second link is the operational user. Where a lessee independently operates the space, the relevant contact path should reach the party that can investigate and act. That does not mean publishing every customer or exposing sensitive infrastructure. It means abuse, routing and operational notices should not die in a holder mailbox if the holder has no practical visibility into the lessee's use. A layered model can distinguish holder, operator and service contacts without publishing the lease itself.
The third link is route-origin authority. If the lessee or its upstream originates the prefix, the holder must authorise that origin through a maintained path. RPKI should be current. IRR data should be consistent where used. Letters of authority should not contradict authenticated records. Start-of-lease and end-of-lease route procedures should be ordinary, not improvised. A stale authorisation after termination is a live risk.
The fourth link is reverse DNS. A lease should make clear whether the holder manages it, delegates it to the lessee or uses a provider. Public delegation should reflect that arrangement. Changes should have service expectations. Termination should include reset or transfer steps. For mail, hosting and enterprise services, reverse DNS can be as commercially important as the address block itself.
The fifth link is abuse and reputation. The parties should know who receives complaints, who investigates downstream users, when the holder can intervene, how repeated abuse affects the lease, who pays for cleanup and what happens to deposits. LACNIC does not need the commercial terms. It can require that the contact path is real and that the holder does not vanish from responsibility.
The sixth link is continuity. Remedies should match defects. A stale contact should trigger correction. A missing ROA should trigger route-authorisation repair. A private payment dispute should not automatically corrupt the public record. A forged authorisation should trigger urgent protective action. A genuine competing claim should pause relevant changes. A legal prohibition should be followed. The ladder matters because leased space often supports innocent downstream customers.
The seventh link is exit. Leasing is temporary only if termination is managed. ROAs, IRR objects, reverse DNS, geolocation, abuse contacts and customer assignments must be cleaned up. The holder's next customer should not inherit the last user's residue. The lessee's customers need realistic renumbering windows. The market will price lessors and lessees by how well they handle exit, if the facts are visible enough.
This chain can be supported by guidance rather than heavy permission. LACNIC could distinguish ordinary downstream assignment, internal corporate use, first-party leasing, brokered leasing, managed hosting, subleasing and migration bridges. It could publish expectations for material delegated use and sample responsibility models. It could clarify that the aim is accurate responsibility, not commercial approval. Such guidance would help holders know what responsible leasing requires, lessees know what to demand, brokers know what they must substantiate and staff know what questions are within scope.
The outcome would be more market discipline, not less. Serious lessors would have an advantage. Thin chains would be easier to detect. Upstreams would receive stronger signals. Small operators would have a checklist. The registry would become less exposed to ad hoc judgement. The shadow allocation economy would still exist, but its public risk would be lower.
Market facts worth publishing
Before the region forms strong opinions about leasing, it should measure what can be measured without exposing confidential contracts. The first category is operational-contact quality. How often do abuse or technical contacts fail? How often are they stale? How often does the named holder lack a path to the actual operator? These are not private commercial details. They are indicators of whether the public record works.
The second category is route-authorisation friction. How often do delegated-use prefixes experience ROA mismatch, stale authorisation, route-object confusion or upstream rejection because authority is unclear? LACNIC may not know every lease, but support cases and authenticated requests can reveal patterns. Route-origin failure is where private delegation becomes public risk.
The third category is reverse-DNS performance. How often are delegations requested for operational users? How often do changes lag? How often do stale zones create problems after lease end, customer migration or transfer? Reverse DNS is a registry-adjacent service that shows whether the public layer keeps pace with use.
The fourth category is transfer substitution. If operators lease because transfer is too slow, uncertain or document-heavy, that is market feedback. LACNIC can publish transfer timing by category, including long-tail delays, documentation rounds, inter-regional coordination issues, payment-related delays and service-transition lags. Predictable transfer channels help confine leasing to cases where leasing is genuinely the better economic tool.
The fifth category is small-operator access. Are smaller LACNIC-region networks leasing because they cannot buy? Are they more likely to accept brokered chains? Do English-speaking Caribbean operators understand guidance as easily as Spanish- and Portuguese-speaking participants? Are public-sector networks able to verify authority without expensive help? These questions do not justify price control. They justify clearer guidance and better visibility.
The sixth category is reputation carryover. How often do buyers, lessees or waiting-list recipients encounter blacklisting, geolocation errors or abuse-history problems? LACNIC already recognises that recovered space can require rehabilitation. The same economic fact applies to leases. The registry should not score address reputation, but the region should understand how much operational cost is being carried from one user to the next.
The seventh category is broker-chain failure. Without collecting private terms, LACNIC can learn from complaints and support patterns whether certain intermediated arrangements produce more stale contacts, unclear authority or abuse escalation failures. That would support targeted guidance and avoid treating all leases alike.
Measurement is not punishment. It is insurance against bad policy. Without facts, one side will point to abuse and extraction, the other to continuity and efficiency. Both will be partly right. A registry that publishes operational facts can keep the debate tied to risk rather than rhetoric.
Policy must hear the absent lessee
The people who debate number policy are not always the people who live with leased-number risk. Holders, large carriers, national registries, technical specialists and regular policy participants may be present. Lessees, small hosting firms, public agencies, security vendors, enterprise customers and island operators may be absent. Yet they are often the ones harmed first when a delegated-use arrangement fails.
The absent lessee carries practical exposure. If a lessor's ROA is wrong, the lessee loses reachability. If reverse DNS is slow, the lessee's customers complain. If the lease ends abruptly, the lessee renumbers. If abuse contacts point to the wrong party, the lessee may be blamed too late to contain damage. If a registry intervention freezes a holder's services broadly, the lessee may become collateral damage. A debate that treats only holders as affected principals will miss the working economy of the resource.
This does not delegitimise the policy process. It means rules that affect delegated use should include economic impact analysis. A proposal touching contacts, RPKI, reverse DNS, transfer eligibility, holding restrictions or delegated-use guidance should ask who pays the compliance cost, who is likely to be absent, whether the rule improves responsibility visibility, whether it pushes legitimate use into opacity and whether small operators can comply without specialist support.
Plain-language education would help more than suspicion. A small operator considering a lease should know the questions to ask before signing. Who is the registered holder? Can the holder prove authority? Who creates ROAs? Who manages reverse DNS? What abuse contact will be public? Is subleasing allowed? What happens if the holder changes status? What is the renewal process? How is reputation handled? What exit window exists? This is market hygiene, not legal advice.
Such guidance would reduce information asymmetry. Brokers would still operate, but customers would be less dependent on broker mystique. Lessors with sound controls would benefit. Thin chains would be exposed. The registry would improve record truth indirectly by helping market participants demand it.
The language matters. If leasing is described as inherently suspect, people will avoid the word. If it is described as delegated operational use with responsibility duties, serious participants have reason to disclose operational facts. The public record becomes more truthful when the vocabulary matches reality.
Watchpoints as the shadow market matures
The first watchpoint is euphemism. If the region treats leasing as a normal scarcity instrument with responsibility obligations, serious participants will have reason to maintain clean contacts and route authorisation. If the language becomes punitive, the same activity will be renamed as managed service, customer assignment or partnership. The market will not disappear. The vocabulary will.
The second watchpoint is RPKI failure. As route-origin validation becomes more common, poor ROA management will become more visible. Disputes over origin changes, stale ROAs, lessor response times and lease-end cleanup will show whether delegated use is being operated as infrastructure or merely sold as inventory.
The third watchpoint is reverse DNS and mail reputation. Hosting, mail, enterprise and security customers often discover address quality through reputation systems, reverse resolution and geolocation. A lease market that treats these as afterthoughts will create churn and distrust. A mature one will price them as service obligations.
The fourth watchpoint is broker concentration. If only a few intermediaries understand where usable supply sits and how to navigate regional risk, small networks will pay an information premium. Better public signals can reduce that premium without eliminating brokerage.
The fifth watchpoint is transfer friction. If transfers are predictable and well documented, leasing will occupy a healthier role as temporary capacity and capital flexibility. If transfers become slower or more subjective, leasing will absorb demand that should have moved through recognised holdership. The size of the shadow market will then reveal public friction as well as private preference.
The sixth watchpoint is small-island resilience. Caribbean and smaller-market operators will test whether the system works under real constraints: thin staffing, foreign payments, storms, undersea dependency, small customer bases and multilingual participation. A leasing posture that works only for large-country operators will fail the region even if it looks orderly.
The seventh watchpoint is remedy discipline. When a lease creates a problem, does the response match the defect? Contact errors should be fixed. Fraud should be stopped. Route-authorisation errors should be corrected. Abuse should be escalated through responsible parties. Customers should not be destabilised unnecessarily. The market will judge leasing by how failures are handled.
A narrow registry, a cleaner market
LACNIC's strongest posture is neither hostility to leasing nor surrender to opacity. It is a narrow bargain: protect the record, make delegated responsibility visible where third parties rely on it, and leave rent to the parties unless a clear rule or legal duty is directly implicated. That bargain accepts the economics of scarcity without letting scarcity hollow out the public coordination layer.
The record matters because IPv4 value depends on more than scarcity. A block is valuable because it is unique, recognised, routable, supportable, contactable and capable of clean operational delegation. If the public record becomes detached from use, the market will rely on private documents and informal reputation. If the registry tries to control too much, the market will hide. The stable middle is a ledger that sees enough and judges less.
For the island provider in the opening scene, the distinction is practical. It needs temporary capacity, credible routing, reverse DNS, abuse handling and renewal confidence. It does not need a registry to decide whether the lease price is tasteful. It does not need a sermon about IPv6 while customers still require IPv4. It does need a public environment in which the lessor's authority can be trusted, operational contacts are real and a private dispute does not unexpectedly break customer service.
For LACNIC, the institutional interest is equally practical. A registry that denies leasing will lose sight of a material part of post-exhaustion use. A registry that regulates leasing prices will overstep and invite evasion. A registry that treats leasing as delegated operational use with responsibility obligations can strengthen the record. It can support IPv6 transition while acknowledging present IPv4 dependency. It can help smaller operators by reducing hidden risk instead of pretending scarcity can be allocated away.
IPv4 leasing is therefore best read as financing and operational delegation under scarcity. It is working-capital relief because it converts a large purchase into recurring cost. It is shadow allocation because use moves through private contracts outside ordinary transfer recognition. It is responsibility risk because holder and user can diverge. All three facts are true at once. The policy error is to choose only one.
The region will live with IPv4 scarcity for years. The waiting list will not clear immediate demand. Transfers will remain important but not sufficient for temporary and capital-constrained use. Large markets will keep pulling capacity. Small islands will keep needing continuity. Brokers will intermediate. Lessors will earn yield. Lessees will build services on borrowed numbers. The public question is whether those arrangements remain visible enough for the internet to know who is responsible.
LACNIC can protect record truth without pretending leasing can be wished away. It can insist that holders remain accountable, operational users be reachable where use is material, RPKI and reverse DNS match authorised use, abuse paths work, lease chains do not hide responsibility and remedies remain proportionate. It can avoid the temptation to judge rent, commercial virtue or business-model desirability. That is not a weak registry. It is a disciplined one.
Scarcity made IPv4 valuable. Leasing will continue because it solves real timing and cash-flow problems. Shadow allocation will become safer only if the public layer can illuminate the responsibility chain. LACNIC's job is to keep that illumination narrow, reliable and dull. It should not decide who deserves value. It should keep the record truthful enough that value can be used without making responsibility disappear.