Summary

  • LACNIC interconnection-dependency analysis asks how peering, transit, IXPs, upstream concentration and route reputation turn address records into bargaining infrastructure.
  • When IPv4 scarcity makes addresses capital facts, registry neutrality affects whether a holder can preserve customer continuity across routing changes, RPKI/ROA evidence and IRR side effects.
  • A credible regional ledger stays narrow enough to preserve portable identity rather than adding a second gatekeeper to already concentrated interconnection markets.

A regional ISP does not negotiate interconnection as an empty technical buyer. It negotiates while carrying every dependency that customers have already built around its numbers. Picture a midsized network in Latin America preparing for a new transit contract. It serves business parks, municipal systems, hosting customers, security vendors, payment processors and a small set of regional cloud workloads. It has an ASN, IPv4 addresses registered in the LACNIC region, route objects still used by some counterparties for filters, ROAs that make origin validation possible, reverse DNS entries that customers stopped noticing because they have not failed, and a history attached to prefixes that have long appeared as part of a stable access and hosting business. The sales meeting with an upstream carrier appears to be about commits, ports, burst pricing, latency and route tables. In reality it is also about whether the ISP can keep its public identity while changing who carries the packets.

If the ISP accepts provider-assigned addresses from the carrier, the first invoice may be lower and the engineering work may look simpler. The carrier can provision from its own pool, maintain its own filters, control its own ROAs and avoid the extra coordination that comes with customer-originated space. Yet the bargain changes the day customers begin to rely on those addresses. Anti-abuse systems learn them. Banks and payment platforms allowlist them. Enterprise firewalls, VPN concentrators, cloud security policies, municipal portals and outsourced support teams treat them as familiar egress points. If the ISP later wants to change carriers, it must either persuade all those parties to update their assumptions or accept that the carrier's address pool has become part of the cost of leaving. What looked like cheap transit becomes an identity mortgage.

The calculation improves when the ISP can bring its own LACNIC-registered resources, but only if those resources remain neutral, reliable and portable enough to function as bargaining assets. The upstream wants to know whether the customer is the legitimate holder, whether the route is authorized, whether the ROA matches the intended origin, whether IRR data will support filters, whether the prefix has a usable history, whether records can be updated quickly, and whether the block can remain stable through a commercial transition. The registry record is not merely a clerk's note in that negotiation. It is part of the credit file of the route.

That is the economic problem. Address governance becomes consequential not because packets ask permission from a registry. Packets do not. It becomes consequential because every party around the packet prices the probability that the holder can keep using the identifier. Interconnection bargaining depends on credibility, and credibility depends on continuity. A registry that keeps records accurate, security assertions coherent and transfers predictable lowers the cost of bargaining. A registry that thickens into a gatekeeper over commercial use, customer geography, leasing, transfer timing or organizational form raises downstream bargaining costs. The charge may not appear as a registry fee. It appears as a worse transit offer, a refused peering session, a longer migration, a higher deposit, an unspoken risk premium or a customer that will not move because renumbering feels too dangerous.

The LACNIC setting makes the issue practical rather than abstract. Latin America and the Caribbean include large national incumbents, smaller regional access providers, island networks with limited upstream choice, content-heavy metropolitan exchanges, long-haul dependency toward North America and growing IXPs that improve local reach without eliminating concentrated transit. This is not chiefly a story about geopolitical fragmentation, and it is not primarily a story about cross-border compliance. Those concerns have their own logic. Interconnection dependency is lower in the stack and more repetitive. It arises whenever a network must buy reachability from someone larger while trying not to surrender the address identity that lets it remain a negotiator rather than a captive customer.

Interconnection begins as bargaining power

Peering and transit are often described as technical arrangements, but they are markets in bargaining power. Transit is the purchase of reachability from a network that agrees to carry traffic onward to the rest of the Internet. Peering is the exchange of traffic between networks, sometimes without settlement and sometimes through paid or selective terms, usually limited to each party's own customers and downstreams. IXPs lower the cost of meeting counterparties by putting many networks on one switching fabric. None of these arrangements abolishes hierarchy. A carrier with global routes, content relationships and many downstream customers can decline more easily than a regional ISP that needs external reach before its customers begin work in the morning.

That asymmetry shapes the whole conversation. The smaller network wants redundancy, lower latency, better regional reach, direct paths to content and freedom from one upstream's pricing. The larger network wants clean routing, predictable traffic ratios, revenue where it has market power and limited operational trouble. A peering coordinator is unlikely to be persuaded by romantic language about openness if route objects are stale, abuse history is poor, the prefix is too small for policy, traffic is sharply one-sided or the customer appears unable to keep its own resources under control. The transit salesperson may frame the point more gently, but the quote will reflect the same judgment.

A portable address block gives the regional ISP a way to bargain across that asymmetry. It lets the ISP tell a carrier that the service being bought is carriage, not identity. It lets the ISP buy from one upstream today, add another tomorrow, move part of the traffic to an exchange, shift data-center presence or bring a cloud edge closer to users without asking customers to rebuild external assumptions. It also tells potential peers that the network has an independent presence. The prefix is not a disposable residue of a retail broadband contract. It is part of a continuing network.

Where upstream options are concentrated, the distinction matters even more. A carrier that controls the addresses used by a customer controls more than a port. It controls part of the cost of exit. Once the customer's customers, banks, partners and security systems recognize those addresses, the carrier has bargaining power that was never written into the transit contract. If the ISP leaves, the carrier does not need to punish it. Renumbering does the work. The fear of reconfiguration, mail reputation loss, API failures, firewall changes and customer confusion disciplines the ISP before any dispute occurs.

That is why interconnection dependency cannot be measured only by whether a network is connected. A dependent network may have packets flowing smoothly while its negotiating position deteriorates. The visible topology may show multiple upstreams, yet if external identity is tied to one provider's addresses, the economic topology remains narrow. A regional ISP can be multihomed in BGP and still be captive in business terms if the identity layer cannot move.

Portable resources are therefore not an indulgence. They are the institutional equivalent of a movable franchise. They make it possible for the operator to separate access from identity, carriage from continuity and transit from customer memory. The registry's task should be to make that separation easy to verify. It should not turn the separation itself into a discretionary privilege.

Portable address identity is the asset being priced

An address begins as a number. In a live network it becomes memory. Customers put it into allowlists. Banks use it as a familiar origin. Security vendors associate it with a pattern of behavior. Mail systems build history around it. Geolocation providers attach it to a country, city or access network. Abuse desks learn whether complaints are handled. Cloud platforms, enterprise SaaS services and government portals may treat it as a known egress point. Over time the prefix carries a form of reputation that is not visible in a simple record but is inseparable from commercial continuity.

That reputation is uneven. Some addresses are ordinary capacity and can be replaced with limited disruption. Others are embedded in the external memory of a business. A hosting customer with thousands of DNS records, a fintech company with merchant integrations, a logistics platform with API counterparties or a municipal service with security allowlists does not experience renumbering as a small engineering task. It experiences renumbering as a coordination campaign across parties that may be slow, regulated, outsourced or unreachable. The direct technical work may take an afternoon. The social and commercial work can last months.

A regional ISP that understands this will try to preserve address identity as it negotiates transit and peering. It will prefer to announce portable space from its own ASN. It will want multiple upstreams to accept the same route. It will want IXPs and route servers to see the same prefix. It will want customers to buy services without becoming hostage to the ISP's current carrier. The ISP's business becomes more credible because a carrier dispute, price increase, data-center move or network redesign does not automatically force a customer renumbering event.

This is where LACNIC's role becomes economically important. The registry does not create the operational value of the prefix. The ISP creates value by deploying routers, paying for ports, serving customers, maintaining security, answering abuse reports, building relationships and making the prefix recognizable in the market. But the registry record helps other parties decide whether that value is stable enough to trust. It identifies the holder, the contact points, the allocation or assignment context and the routing-adjacent artifacts linked to the resource. When those records are neutral and dependable, they lower verification cost. When they are subject to uncertain discretion, they raise it.

The difference is subtle but material. A carrier can route a customer prefix without caring deeply about theories of number resources. Yet the same carrier's risk, provisioning or peering staff will care whether the customer can produce clean authorization, update route objects, maintain valid ROAs and survive a registry inquiry without service disruption. The more uncertain the registry relationship appears, the more the carrier has reason to prefer its own address space or demand more control over the customer's routing. That preference may be rational for the carrier. It is costly for the ISP.

Address identity also changes valuation. A block that is stable, portable, well maintained and trusted across counterparties is not the same economic object as a block that exists only as a conditional entry subject to broad administrative judgment. Both may contain the same number of addresses. Only one carries strong continuity. In interconnection markets, continuity is part of the asset. It affects who will accept the route, who will peer, who will buy service from the ISP and how expensive it will be to change the network around the prefix.

LACNIC's region turns portability into resilience

The Latin American and Caribbean Internet is not one market in operational terms. It is a set of national, metropolitan, island and cross-border markets with different economics. Some cities have strong exchange points and dense carrier choice. Others depend on a small number of submarine systems, international gateways, national incumbents or wholesale providers. Content caches have improved local performance, but not every network can reach every important destination locally. A regional ISP may need a port at an IXP, a paid link to a carrier hotel, transit from a regional backbone and backup paths that exit through another country. Each layer brings a negotiation.

In such a setting, address portability has more value, not less. The more constrained the upstream market, the more dangerous provider-assigned identity becomes. If an island network has two realistic upstreams and one controls the customer's public addresses, the customer does not have two equal choices. If a landlocked access provider depends on a small number of cross-border paths, the ability to move routes among those paths without renumbering is a form of resilience. If a regional hosting provider wants customers in several countries, stable prefixes help it avoid rebuilding customer trust every time it changes a transit blend.

This is not a claim that the region is uniquely fragile. It is a normal institutional-economics problem sharpened by topology. Where alternatives are abundant, lock-in is survivable because exit is cheaper. Where alternatives are few, the same lock-in becomes strategic. A registry that stays thin and neutral helps smaller networks offset upstream concentration. A registry that adds discretionary friction strengthens the upstream side of the market, whether or not it intends to. The cost of registry thickness is therefore not paid only at the registry counter. It is paid in every market where a small network negotiates with a larger one.

LACNIC's region also contains many networks that have grown incrementally rather than through large capital plans. They may have acquired customers before acquiring strong legal or administrative capacity. They may depend on engineers who understand BGP better than contract theory. They may treat registry matters as paperwork because the daily emergency is keeping links up. This makes them more exposed to rules that turn registration into a broader permission surface. A large carrier can hire counsel, dedicate policy staff and wait through delay. A small ISP may lose a customer, a financing opportunity or a peering slot while waiting for a record problem to clear.

The asymmetry is compounded by language, distance and procedural familiarity. A metropolitan carrier with permanent regulatory and governance staff can participate in forums, interpret policy changes early and adapt internal systems around them. A rural or island ISP may encounter the same rules only when a transfer, audit, merger, leasing arrangement or security update becomes urgent. Complexity therefore functions as a tax on the less powerful. The tax is not merely the time spent reading rules. It is the bargaining disadvantage created when counterparties know that the smaller network has fewer ways to resolve registry uncertainty quickly.

For LACNIC, neutrality should be measured against this market reality. A registry can treat members formally alike while imposing procedures whose practical burden falls hardest on networks with the least bargaining power. Formal equality is not enough when the downstream effect is to make the weakest operators less portable. The relevant question is whether registry practice reduces or increases dependence on upstreams, incumbents and procedural insiders. In an interconnection market, that is the difference between coordination and gatekeeping.

Peering changes when the prefix can leave

Peering policy is usually expressed in technical language: traffic ratios, geographic presence, route count, NOC availability, abuse handling, capacity, latency, mutual benefit and minimum traffic. Beneath those criteria sits a simpler question. Is this network a durable counterparty? A network announcing stable portable space from its own ASN looks different from a customer whose visible presence is an extension of a provider. It may still be small, but it has a continuing identity. It can invest in engineering relationships because those relationships will not vanish when a wholesale contract changes.

At an IXP, this matters in practical ways. A network that joins an exchange with portable prefixes can establish route-server sessions, bilateral peerings and local content relationships that survive changes in upstream transit. If the ISP changes its primary carrier, the IXP presence remains legible. If it adds a cache, the cache can be reached through the same customer-facing identity. If it opens a second point of presence, the same resource base can support the expansion. The network's interconnection work accumulates rather than resetting with every provider change.

Provider-assigned space weakens that accumulation. A small ISP may still peer locally, but if key customers use addresses controlled by an upstream, the ISP's own bargaining story is thinner. The upstream can offer a package: transit, numbering, route authorization and operational simplicity. That package may be useful for a startup network. It becomes a trap when the network matures. The upstream's offer is not only connectivity. It is bundled identity. The ISP's future peering independence is discounted at the moment it accepts the bundle.

The same logic affects transit negotiations. A carrier quoting a network that controls portable resources knows it is competing for carriage. A carrier quoting a network dependent on provider-assigned resources knows that leaving will be painful. The latter customer may receive polite account handling and still pay a hidden lock-in premium. The premium may appear as higher minimum commits, worse burst terms, slower support, less willingness to customize routing or a refusal to support customer-originated prefixes except at extra cost. The economics need not be explicit to be real.

Portable resources also change the ISP's relationship with its own customers. A regional ISP can tell an enterprise that it will preserve public network identity even if upstream arrangements change. That promise is commercially important. Enterprises increasingly depend on stable egress for security, fraud control, partner systems and compliance documentation. They may not know the difference between peering and transit, but they understand the cost of changing hundreds of allowlists. An ISP that can offer provider-neutral identity competes on continuity rather than merely bandwidth.

Here again the registry record sits behind the bargain. If LACNIC's record is a predictable public statement of who holds the resource and how it may be securely announced, it supports the ISP's claim. If the record is perceived as conditional, slow, politicized or exposed to broad review of matters unrelated to uniqueness, the ISP's claim weakens. The upstream can argue that provider-assigned space is safer because the carrier controls the administrative chain. That argument is bad for the customer's independence, but it becomes persuasive when registry neutrality is uncertain.

The danger is not that LACNIC directly tells an ISP which carrier to use. The danger is that registry uncertainty reshapes the menu of commercially viable choices. When portability is cheap and records are neutral, the ISP can shop between carriers. When portability is administratively fragile, the carrier that offers the most vertically integrated bundle gains power. Interconnection markets then become less competitive even though no formal rule has banned competition.

Route reputation is capital with operational memory

Route reputation is hard to measure because it lives across many systems. There are public blocklists, private anti-abuse scores, spam filters, geolocation data, threat-intelligence feeds, payment-risk systems, enterprise firewalls, CDN customer policies, cloud trust mechanisms and the informal memories of network operators. Some of these systems are accurate. Some are crude. All of them can affect the usefulness of a prefix. An address block with a poor history may route perfectly and still be commercially impaired. A block with a stable history may be valuable because it causes fewer questions.

For a regional ISP, this reputation is built slowly. It comes from consistent origin announcements, responsive abuse handling, predictable reverse DNS, stable customer use and the absence of sudden suspicious churn. It also comes from being seen by peers and upstreams as an operator rather than a transient reseller of questionable space. In a world of scarce IPv4, where leased and transferred resources move through many hands, route reputation becomes part of the economic distinction between clean continuity and mere capacity.

Registry records do not fully determine reputation, but they anchor it. When a prefix's registered holder, route authorization, abuse contact and security objects are coherent, counterparties can map reputation to responsibility. When the registry layer is ambiguous or slow to reflect operational reality, reputation becomes harder to manage. A legitimate holder may struggle to prove control. A buyer or lessee may inherit old reputational problems without a clean way to show changed responsibility. A network trying to move upstreams may face filters built from stale IRR data or private systems that lag behind the actual commercial arrangement.

The cost of this uncertainty is paid in negotiations. An upstream may accept the route but require more documentation. A peer may decline a session until the prefix looks cleaner. A security-sensitive customer may demand assurances the ISP cannot confidently provide. A financial services customer may insist on stable egress with a long history, making provider-assigned churn unacceptable. In each case, the address block is being evaluated like capital. Its value depends on trust, transferability, history and enforceability.

This is why treating registry work as back-office paperwork misunderstands the market. The registry record is one of the places where the network's external identity becomes legible. It is not the only place, and it should not pretend to be the origin of operational value. But if the record is unreliable or discretionary, it contaminates the reputational layer. The ISP's commercial story becomes harder to tell because the administrative evidence behind it looks less neutral.

Route reputation also explains why abrupt revocation or destructive dispute handling can create harm far beyond the registry account. If a live prefix is polluted, invalidated or left in a state that counterparties cannot interpret, the damage is not limited to the holder's internal systems. Customers may lose mail deliverability, API reachability, payment trust, remote access, CDN behavior or security exceptions. Even if the registry later corrects the record, reputation systems may not recover immediately. Some reputational costs have memory longer than the dispute that caused them.

The proper institutional lesson is restraint. A registry that protects uniqueness should avoid turning disputes into reputational events unless an independent necessity requires it. It can record conflict metadata, preserve the last verified operational state, prevent contradictory transfers and support adjudication without forcing the route into crisis. In interconnection economics, non-destructive record handling is not politeness. It is the preservation of capital value attached to continuity.

RPKI, ROAs and IRR add security while adding dependency

Routing-security tools are essential, but they are not neutral in their economic effects. RPKI lets holders create Route Origin Authorizations that state which ASN may originate a prefix and, within defined limits, how specific the announcement may be. Origin validation gives networks a way to reject routes that appear unauthorized. IRR data, though older and less uniform, still feeds route filters and operational decisions across many carriers. Together these tools reduce some forms of route leak and hijack risk. They also make registry-linked control more consequential.

Consider the ISP changing upstreams. If its ROAs are current and allow the intended origin ASN, the migration is safer. If the ROA is missing, too narrow, tied to the wrong origin or slow to update, the route may be treated as invalid by networks that enforce origin validation. If IRR objects are stale, route filters may not update cleanly. If a new upstream requires exact authorization before provisioning, commercial activation can be delayed even after the contract is signed. The security layer therefore becomes part of the sales cycle.

This is not an argument against RPKI or IRR. It is an argument for understanding their institutional side effects. A security assertion tied to a registry relationship can become a bargaining signal. The upstream asks whether the customer can manage ROAs quickly. The peer asks whether route objects are consistent. The customer asks whether a move will create invalid routes. The financing counterparty asks whether the resource can be controlled securely. Each question is reasonable. Each question increases the value of a neutral, responsive registry record.

Problems arise when security tools become additional levers of gatekeeping. A registry that can impair or delay the holder's ability to maintain ROAs does not merely affect a record. It affects whether the holder can change upstreams, split traffic, recover from an outage, respond to a hijack or add a point of presence. A dispute over account status, utilization review, commercial structure or contractual interpretation can therefore spill into routing security. The technical language remains clean, but the economic effect is coercive if the holder's route validity depends on the discretion of the institution with which it is in conflict.

The IRR layer adds another kind of inertia. Many route filters are built from data that operators do not inspect daily. Old objects, inconsistent maintainer arrangements, inherited records and mismatched AS sets can survive for years. When a network tries to move, the mess becomes visible. A larger carrier may have staff to untangle it. A small ISP may lose a maintenance window or a customer migration. If the registry insists on broad discretionary review before record updates, the operational problem becomes a bargaining disadvantage.

The right principle is narrowness. RPKI and IRR should prove control, support secure routing and describe operational intent. They should not become instruments for policing leasing morality, customer geography, financing structures or business models unrelated to uniqueness and authorization. A ROA should not become a confession of loyalty to a regional theory of numbering. It is a security object. Its value lies in being reliable, fast, auditable and boring.

For LACNIC, the temptation to thicken security into governance should be resisted precisely because the tools are so important. The more routing-security deployment grows, the more damage a discretionary record layer can do. Security adoption should reduce hijack risk without increasing institutional hostage risk. If operators fear that security objects become another way to lock them in, they will adopt defensively, minimize exposure or keep parallel informal arrangements. That weakens the security system itself. Trust in RPKI and IRR depends on trust that the registry will not use routing-security dependence as commercial or institutional leverage.

IPv4 scarcity makes neutrality more valuable

IPv4 scarcity is often discussed as though it automatically justifies stronger registry control. The better conclusion is the opposite. Scarcity makes the resource valuable, transferable, financeable and operationally embedded. Those characteristics demand clearer rights, lower transaction costs and more neutral records. When a resource was abundant and low value, a loose administrative system could appear harmless. When the same resource becomes capital, ambiguity becomes expensive.

The price of IPv4 is not the main barrier to connectivity in most network cost stacks. Fiber, equipment, power, data centers, spectrum, financing, staff and upstream transit usually dominate. An IPv4 address may be scarce and still be a small part of the cost of delivering customer service. That does not make scarcity irrelevant. It means the address is economically powerful because it enables revenue and continuity, not because its monthly cost overwhelms the network. For an ISP, IPv4 is one of the few globally scarce inputs it can actually hold and deploy. Suppressing its asset value weakens operators that already struggle with thin margins.

In interconnection bargaining, scarcity has two effects. It makes portable addresses more valuable because replacements are not casually available. It also makes registry uncertainty more damaging because counterparties know that the holder cannot simply obtain another equivalent block if a dispute arises. A carrier negotiating with a small ISP understands that the ISP's address base is hard to replace. If the registry relationship is fragile, the carrier can price that fragility. Scarcity therefore amplifies the importance of neutrality.

Needs-based thinking becomes dangerous after scarcity has created a mature secondary and leasing market. When fresh pools were abundant, demonstrating operational need may have helped ration allocations. After exhaustion, the main question is no longer whether a hostmaster should give away new addresses. It is whether existing resources can move to their highest and most reliable use without breaking uniqueness. Transfer, lease, financing and customer geography are economic arrangements around a scarce asset. They should be disciplined by contracts, markets, courts and operational security, not by a registry's discretionary view of which business model deserves approval.

This matters for LACNIC because the region contains operators that could use IPv4 scarcity to strengthen rather than weaken their position. A small ISP with stable addresses has a scarce asset that can support customer continuity, collateral value, leasing revenue, interconnection independence and a better story in front of enterprise buyers. If registry policy treats that asset as a conditional favor, the ISP loses part of the benefit while keeping the risk. It cannot fully capitalize the resource, but it remains exposed if the registry record becomes uncertain. That is a poor bargain.

The distributional claim often runs the other way. More control is said to protect smaller or poorer networks from markets. Yet permissioned scarcity frequently protects insiders who can navigate process. Wealthier networks can hire specialists, wait out delays and structure around restrictions. Smaller networks need liquidity, predictability and the ability to use what they have. A policy that makes transfer or leasing slower in the name of equality may simply make small holders poorer and large buyers more patient.

Record neutrality is therefore not a pro-market slogan. It is the institutional condition for allowing operators to realize the value of resources they have made useful. The registry should record who holds the resource, preserve uniqueness, support secure routing, maintain contactability, publish accurate data and handle disputes without destroying live use. It should not decide whether an ISP's commercial strategy is sufficiently pure. Once scarcity has transformed addresses into capital, that kind of discretion is not stewardship. It is a tax on bargaining power.

Registry hold-up turns recognition into leverage

Institutional economists use the term hold-up for a familiar pattern. One party makes investments that are specific to a relationship or asset. After the investment is sunk, another party that controls a bottleneck can extract concessions or impose costs. The Internet number system now contains this pattern. Operators invest in networks, customers, route reputation, security objects, address continuity and interconnection relationships. The registry controls the recognition layer that makes those investments portable. If the registry's discretion expands after the investments are made, the operator bears the loss.

The hold-up need not be dramatic. It can appear as uncertainty over a transfer, slow handling of a record update, a broad audit right, an annual renewal condition, a shifting interpretation of policy, a threat of revocation after breach or a demand that commercial arrangements fit a preferred administrative theory. Each item may be defensible when viewed alone. Together they place a thin institutional contract above a thick layer of operational reliance.

LACNIC's registration relationship, like other registry relationships, has features that make this problem visible. It is adhesive rather than individually negotiated, tied to terms and guidelines that may change, renewable over time and connected to powers of review and revocation under defined conditions. Such provisions may look ordinary for a low-value service. They look different when the resource supports business continuity, route reputation, customer contracts and interconnection leverage. The mismatch is not that a registry has no rules. The mismatch is between the scale of the registry's practical influence and the narrowness of the operator's exit.

Exit is the core issue. In ordinary commerce, a customer facing a difficult service provider can move to another provider. In number resources, the holder may be able to change transit providers, but it usually cannot change the registry relationship without significant structural friction. The registry's region acts as a lock-in device. If the registry remains thin, the lock-in is tolerable because the service is mostly recordkeeping. If the registry becomes thick, the same lock-in becomes power. Portability across qualified registry services would discipline that power by making poor service costly to the registry rather than only to the operator.

The downstream effect appears in interconnection contracts. An upstream knows whether the customer can leave cleanly. A peer knows whether the customer can maintain stable authorizations. An enterprise buyer knows whether the ISP can preserve identity during a provider change. Registry hold-up therefore becomes market hold-up. The registry may not be present in the carrier meeting, but the uncertainty it creates is priced there.

This is why the answer cannot be more elaborate assurances that the registry acts for the community. A community meeting does not absorb the ISP's losses if routes fail. A policy process does not compensate customers for renumbering. A registry budget does not match the capital value sitting under the records it maintains. The relevant institutional test is whether control is aligned with consequence. If the registry can materially affect continuity while carrying little of the downside, the operator will discount its own assets and counterparties will discount the operator.

The hold-up problem also explains why fear becomes normal. Many operators treat registry messages with more anxiety than ordinary vendor communications because the registry sits near the recognition layer. That fear is itself a cost. It makes networks conservative, discourages efficient transfers, deters flexible leasing, increases legal spend and pushes customers toward vertically integrated providers that promise simplicity at the price of lock-in. A healthy registry system would make operators less afraid to act, not more afraid to move.

When a registry thickens, upstreams gain

A thick registry is one that moves beyond the minimum tasks required for uniqueness, accuracy, contactability, transfer records and security assertions. It begins to judge business models, customer location, commercial morality, leasing structures, financing, historical usage beyond what is necessary for fraud control or the political meaning of regional resources. It may describe these activities as stewardship, fairness or risk control. The market hears something else: the holder's address identity is conditional.

Conditional identity shifts bargaining power toward upstreams and incumbents. Large carriers can absorb conditionality because they hold more resources, maintain dedicated staff and often influence the procedural environment. Smaller networks cannot. When registry friction rises, the easiest answer for a small ISP is to take the upstream's bundle and avoid the administrative fight. The upstream then gains a customer whose identity is less portable. The registry may believe it has protected the public interest. The interconnection market has become more concentrated.

The same dynamic affects IXPs. Exchanges work best when networks can join, announce stable routes and build many relationships without asking permission from a dominant carrier. If address portability weakens, the IXP becomes less able to discipline upstream power. A network may still connect to the exchange, but its most important customer identities may remain tied to the carrier that supplied addresses. The exchange lowers packet cost while the upstream keeps identity leverage. The region gets a more local routing fabric, but not the full bargaining benefit that portable resources could have delivered.

Thick registry control also creates a selective burden on business models that rely on leasing or shared continuity structures. Leasing can be abused, as any market can. But leasing can also let operators obtain addresses without warehousing registry risk, allow customers to preserve identity across providers and put idle resources into productive use. A blanket suspicion of leasing may protect a registry's sense of control while harming the networks that need flexible access to scarce resources. In interconnection terms, it can push customers back toward provider-assigned space, which is precisely the structure that strengthens upstream lock-in.

The effect is especially perverse when thick governance is justified by scarcity. Scarcity should make the common layer narrower because mistakes are costlier. Instead, scarcity often tempts institutions to expand. The registry begins to treat the record as a lever for shaping markets. It asks not only whether the resource is unique and securely controlled, but whether the holder's intended use fits a preferred view of regional virtue. That is a large move. It converts a neutral ledger into an industrial-policy instrument without the accountability, capital or public-law mandate that such an instrument would require.

For a LACNIC-region ISP, the practical question is not whether the registry uses the word ownership. The question is whether the operator can rely on continued use, transferability, secure routing updates and non-destructive dispute handling strongly enough to negotiate with carriers. If it cannot, the formal vocabulary matters little. The ISP is renting recognition from a single institution while trying to compete in markets where continuity is everything. Upstreams can see the weakness.

Thinness is therefore pro-competition. It does not mean absence of records or absence of safeguards. It means the registry does only what must be common and leaves commercial variation to operators, contracts and markets. It preserves uniqueness. It verifies control. It publishes accurate contacts. It supports RPKI and IRR without turning them into obedience tools. It records disputes without poisoning live routes. It makes transfers legible without deciding whether commerce deserves to exist. That is the version of registry neutrality that improves downstream bargaining.

Customer continuity is the hidden price signal

The price of interconnection is not only the number on the transit invoice. It is also the cost of keeping customers whole while the network changes. A regional ISP that can move from one upstream to another without changing customer-facing addresses has a different cost curve from an ISP that must coordinate hundreds of external parties before it can leave. The first can negotiate with a credible threat of exit. The second may have theoretical alternatives but practical captivity. The carrier's account team may never use that language, but the contract will feel it.

Customer continuity is often more valuable than raw bandwidth because it protects revenue that is already won. The ISP has acquired customers, installed circuits, configured routers, handled support, built local trust and taught other systems to recognize its addresses. That work is sunk. Once it is sunk, the parties around the ISP know that disruption is expensive. An upstream with provider-assigned identity in the stack can use that fact without appearing aggressive. It can be slower to discount, less flexible on routing, less urgent in resolving congestion or more confident that the customer will renew. The customer's fear of renumbering becomes the supplier's bargaining cushion.

Portable address identity reverses part of that cushion. It does not make the small ISP equal to a global carrier, but it gives the ISP a movable base. The ISP can tell an enterprise customer that a new upstream will not require a new network identity. It can tell a data-center partner that the route can shift while the customer record remains stable. It can tell a second transit provider that traffic can move if the terms are right. These are commercial claims, not slogans. They are credible only if the registry record and routing-security layer support them at operational speed.

Continuity also affects financing and planning. A bank, investor or strategic buyer looking at a regional ISP will care whether the customer base depends on addresses controlled by someone else. If the addresses are portable and reputation-bearing, the ISP's revenue looks more durable. If the addresses are borrowed from an upstream or exposed to uncertain registry discretion, the revenue deserves a discount. The same logic applies to enterprise procurement. A customer buying connectivity for critical systems will pay more for a provider that can preserve public identity through carrier changes than for one that must renumber under stress.

LACNIC's recordkeeping therefore sits close to a financial signal. The registry may not set the ISP's valuation, but its neutrality affects how much of the ISP's customer continuity can be treated as durable. If updates are predictable, disputes are isolated and security assertions remain under the holder's practical control, the continuity premium belongs to the operator that earned it. If records are conditional on broad administrative judgment, part of that premium leaks away before any packet moves.

The customer angle also clarifies why this issue differs from broader political or regulatory debates. The enterprise buying connectivity from a regional ISP may not care which institutional forum adopted a rule or which theory of community prevailed. It cares whether its payment integrations, remote access, supplier portals and security exceptions continue to work when the ISP improves its network. Registry neutrality matters because it preserves the customer's option to receive better service without absorbing the hidden cost of someone else's carrier dependence.

The registry record should not choose the carrier

A neutral registry record does not literally pick an upstream. A non-neutral one can effectively do so by making independent address identity too costly for smaller networks. The choice then becomes simple. Either the ISP carries the administrative burden of portable resources under uncertain rules, or it accepts provider-assigned space and lets the carrier handle the paperwork. Many will choose the latter, especially if they are under pressure to launch service, close a contract or reduce immediate legal risk. The registry has not ordered anyone to choose a carrier. It has altered the relative cost of the options.

This is the institutional subtlety at the heart of interconnection dependency. Power often acts through transaction costs. A rule can be formally general and still change bargaining outcomes by making one path more expensive than another. A slow transfer process can favor incumbents. A broad review standard can favor those with policy staff. An uncertain leasing posture can favor carriers with large internal pools. A fragile ROA update path can favor upstreams that keep customers inside their own origins. A public record that does not cleanly reflect operational reality can favor the party that controls the simplest story.

LACNIC's legitimacy in this market should therefore be judged by its effect on operator choice. Does it make it easier for a network to bring its own addresses to a carrier? Does it make it easier to move between transit providers without renumbering? Does it let a holder update security objects quickly when the operational origin changes? Does it keep dispute handling from becoming a route-reputation event? Does it distinguish fraud prevention from business-model review? Does it preserve the ability of small networks to use scarcity as capital rather than treating scarcity as a reason to keep them under supervision?

These questions are more important than ceremonial claims about community. A network's customers do not experience the community. They experience reachability, latency, security and continuity. If a registry policy makes an ISP less able to negotiate upstreams, the cost eventually reaches customers through higher prices, weaker redundancy or slower expansion. If registry neutrality lets the same ISP carry identity across carriers, the customer gets a more resilient market. The outcome is not ideological. It is operational.

The carrier side should welcome thin neutrality in its healthier form. Good carriers compete on performance, reach, support, installation, latency, resilience and commercial trust. They do not need to own a customer's identity to retain the customer. A carrier that can deliver a portable prefix well becomes more attractive to sophisticated buyers. The market then rewards service quality rather than lock-in. Thick registry uncertainty, by contrast, rewards carriers that can make administrative fear disappear by bundling the customer into their own address space.

The same applies to managed network providers, data centers, SASE platforms and local access providers. Many already support bring-your-own-address models in practice. If the registry layer is clean, they can productize that capability. They can tell customers that the service path can change while public identity remains stable. That creates a premium service category around continuity. If the registry layer is uncertain, these providers are tempted to avoid the complexity or to push customers into provider-controlled numbering. The market loses a chance to separate delivery from identity.

In this sense, LACNIC's best contribution to regional interconnection would be to make itself less interesting. Accurate records, reliable publication, fast security updates, clear transfer history and predictable non-destructive dispute handling should become background assumptions. The registry should not be the dramatic part of a carrier migration. It should be the quiet evidence that the route belongs to the party announcing it and can keep belonging to that party as the network improves.

A thinner common layer would lower bargaining costs

The architecture implied by interconnection economics is not complicated. The common layer should contain only what must be common: uniqueness of number resources, accurate holder records, proof of control, contactability, transfer history, routing-security assertions, reverse DNS delegation, dispute metadata, auditability, continuity and replacement paths. These functions are important precisely because the market depends on them. They are also limited. They do not require a registry to judge ordinary commercial arrangements or to make itself the guardian of every use case that touches a prefix.

For LACNIC, this would mean treating the registry record as a neutral ledger for networks rather than a permission field for business models. If a transfer is real, non-fraudulent and does not create duplicate recognition, the registry should record it. If a holder authorizes a new origin ASN, the security system should reflect that authority quickly and predictably. If a resource is in dispute, the dispute should be visible without destroying the last verified operational state. If a network wants to lease, finance, route through a partner, move upstreams or serve customers outside a narrow administrative expectation, the registry should ask only whether uniqueness, accuracy and security remain intact.

Portability is the essential discipline behind this thinness. A registry with no realistic exit path is always tempted to become thicker. A holder that cannot move must accept delay, reinterpretation and administrative fashion. A holder that can move has leverage. Portability does not mean chaos, duplicate records or careless transfers. It means that qualified registry service should be replaceable without forcing the network to renumber or abandon its identity. In other infrastructure markets, replaceable administration is a sign of maturity. In number resources, it is still treated too often as a threat to the incumbent.

Interconnection markets would benefit directly. A portable holder can negotiate transit without surrendering identity. Peering relationships can survive carrier changes. IXPs can become more powerful disciplines on upstream concentration. RPKI and IRR can support security without locking the operator to one institutional interpretation. Route reputation can accumulate around the network that earned it. Customers can choose ISPs based on service rather than fear of reconfiguration. Capital can value IPv4 holdings with less discount for registry uncertainty.

There will still be disputes, fraud, abuse and operational mistakes. Thin coordination does not deny that. It handles them at the right layer. Fraudulent record changes should be stopped. Duplicate claims should be resolved. Security assertions should be authenticated. Abuse should be handled by the networks, customers, contracts, platforms and public authorities that actually have the relevant facts and remedies. Commercial disagreement should not be converted into route invalidation unless an independent decision requires it. The registry should not act as claimant, judge and executioner over a live network's identity.

That last point is central to customer continuity. The live network is the asset-producing system. It carries services, customers, contracts, reputation and reliance. The registry record supports that system by making it legible. When the support layer threatens the live layer to preserve institutional discretion, the order has been reversed. Interconnection dependency then worsens because every counterparty must ask whether the ISP's identity can survive the registry as well as the carrier.

LACNIC can remain useful by being narrower. A quiet, accurate and replaceable registry function would strengthen the region's operators far more than a thick theory of stewardship. It would help smaller ISPs bargain with upstreams, help exchanges deepen local routing, help enterprises buy continuity, help security tools gain trust and help IPv4 scarcity become capital for networks rather than leverage for administrators. The result would not be a slogan about ownership or sovereignty. It would be a lower cost of moving packets through markets where the smallest networks most need the ability to leave without disappearing.

The regional ISP in the opening negotiation does not need a registry to bless its business model. It needs the registry to make one fact boringly reliable: the resources that customers already trust can remain stable while the network around them changes. If LACNIC does that, it reduces dependence. If it does more than that, and if doing more turns records into permission, it raises the price of every future negotiation. In interconnection, the party that can keep its identity can bargain. The party that cannot is merely buying another path to someone else's address book.

Sources and further reading

These references provide the article's public doctrine and background context. They are used for institutional-economic framing, not for adopting any registry or official-sector narrative.