Summary

  • A first-time ISP seeking a /24 to /22 must show at least 25% immediate need and a one-year plan for 50% use. If recovered space cannot meet that need, a transfer recipient must pass the same justification logic before financing a private acquisition.
  • LACNIC’s own 2025 operational account put the smallest market block at approximately USD 8,000. That price is not a registry fee, and it is not universal, but it reveals a second gate that an incumbent with existing inventory does not face merely to remain in the market.
  • The defensible remedy is not a promise of IPv4 abundance. It is to measure first-time-recipient outcomes, make pre-approval and small-block transfer information more legible, and treat IPv6-first operation as a design path rather than proof that IPv4 compatibility has no entry cost.

The entry calculation begins with three numbers

Twenty-five per cent. Fifty per cent. About eight thousand dollars.

The first two numbers come from LACNIC’s Policy Manual. An ISP seeking an initial block between /24 and /22 must prove utilization or immediate necessity for at least a quarter of the requested space and submit a detailed plan showing at least half will be used within a year. If the applicant does not already hold a LACNIC-issued IPv6 block, it must request one at the same time.

The third number comes from an August 2025 article by LACNIC officials about operating after exhaustion. They wrote that the smallest IPv4 block available on the market cost approximately USD 8,000 and that many organizations could not afford it. The figure is a dated approximation, not a tariff and not a promise about today’s transaction price. It is still useful because it exposes the sequence confronting an entrant that needs IPv4 now: first prove the network case, then find the capital and a counterparty.

An established operator may also have to document utilization when requesting additional resources or receiving a transfer. Formal equality therefore exists. The asymmetry lies elsewhere. The established operator begins with customers, assignments, routing records, invoices, traffic and an address base from which it can grow. A new ISP must assemble credible evidence before the scarce input can help it win the customers that would generate the evidence.

That is LACNIC’s new-entrant disadvantage in its narrowest form. It is not that policy secretly says “incumbents first.” It is that a needs test designed to prevent speculation now sits in front of a market purchase. Prudence at the registry layer and scarcity at the market layer combine into a two-stage gate.

A needs test cannot allocate an address pool that no longer exists

Needs-based allocation made intuitive sense while a registry held an ordinary pool. The applicant demonstrated expected use; the registry conserved a common resource; an approved request produced an administrative allocation. The test determined who could receive a scarce public input.

After exhaustion, approval no longer completes that bargain. LACNIC’s transfer policy requires a receiving organization in the region to justify the IPv4 space as an initial or additional request under current rules. But the address block comes from another holder, not from a replenished common pool. The recipient must close a legal and commercial transaction, pay the seller on undisclosed terms, complete the registry process and absorb any operational rehabilitation the block requires.

The distinction matters. Justification controls eligibility; it does not finance the asset. Transfer registration can protect the ledger; it does not create a liquid market at a price a small operator can bear. Pre-approval through LACNIC’s List of Possible IPv4 Transfers can reduce one uncertainty, but it does not guarantee that a matching /24 will appear or that the applicant can purchase it.

LACNIC’s 2024 annual report shows why small-block access deserves attention. It recorded 197 transferred blocks, 25% more than in 2023, without an increase in total transferred address volume because more /24 blocks were involved. The same report recorded 95 IPv4 blocks, totaling 77,824 addresses, assigned alongside 659 IPv6 assignments. The market is moving in smaller units, and IPv6 issuance is far more frequent, but neither fact tells us how many first-time operators found an affordable and timely route to IPv4 compatibility.

The incumbent advantage is stored evidence

Address inventory is usually described as capacity. For an incumbent it is also an archive.

Customer assignments in WHOIS, historical utilization, routing announcements, reverse-DNS practice, abuse contacts and billing records make an established network legible. When that operator forecasts growth, it can point to a production system that already exists. If it conserves addresses, introduces carrier-grade NAT or purchases a supplemental block, it does so while its public services continue to work from an existing base.

The entrant’s documents are thinner because its business is younger. A signed customer letter may depend on the operator first demonstrating stable public addressing. A bank or investor may want evidence that address capacity will not obstruct launch. An upstream provider may offer assigned space, but that can increase dependence and complicate portability. A transfer seller wants certainty of closing. Each counterparty can rationally ask the entrant to solve a different part of the problem first.

No single request is discriminatory. Together they form a circular proof burden: demonstrate use to qualify, qualify to negotiate, finance the block to launch, and launch to produce the strongest evidence of use.

An incumbent acquired its inventory under earlier conditions and is not at fault merely for retaining it lawfully. Nor would forcing productive holders to surrender addresses recreate a fair entry market without imposing new disruption. The relevant institutional question is more restrained: can a formally neutral registry observe when its procedures interact with historical endowments to raise the minimum viable cost of entry?

IPv6-first is necessary, but compatibility still has a price

LACNIC’s answer to exhaustion is increasingly IPv6-mostly. That is technically coherent. In 2025 LACNIC reported more than 13,000 member organizations, including more than 1,600 operating exclusively with IPv6. Its guidance describes IPv4 as a complementary service for destinations and customers that still require it.

The cohort should not be misread. “IPv6-only member” means the organization had no IPv4 assigned by LACNIC; it does not prove that the organization owns no IPv4 from another source, failed commercially or even intended to become a retail ISP. A 2023 LACNIC study found 483 such members in WHOIS data. Of those, 343 had an ASN, 261 announced some or all of their IPv6 prefix and 222 did not announce the prefix. The study itself did not claim that a single cause explained those outcomes.

Still, the growth of this category shows that membership and IPv4 possession have separated. An organization can join the regional Internet system, receive IPv6 and participate in policy without receiving the compatibility asset that many customers, platforms and devices still assume.

IPv6-mostly architecture can reduce the amount of IPv4 an entrant needs. It cannot make every legacy destination reachable without some translation or IPv4 complement. Carrier-grade translation adds hardware, logging, port management, abuse attribution and support work. Provider-assigned IPv4 may be a workable bridge, but it gives an upstream provider leverage over numbering continuity. LACNIC’s 2025 article said leasing LACNIC-assigned IPv4 was not allowed and could expose the resource to revocation, while leasing blocks from elsewhere could create geolocation, traceability and support problems.

The practical goal is therefore not to preserve an IPv4-first design. It is to reduce the quantity of IPv4 required without denying that the final quantity has a price.

Registry fees show the difference between participation and acquisition

LACNIC’s 2026 ISP schedule lists an annual fee of USD 633 for the Nano category, covering holdings smaller than /22, and USD 1,055 for the Micro category, covering holdings smaller than /20. The fee is based on the highest category determined by assigned IPv4 or IPv6 resources.

Those amounts fund registry membership and services. They are not the private price paid to acquire a transferred block. Conflating the two would understate the entrant’s capital requirement. A newcomer may pay a registry fee, an administrative transfer fee, professional or facilitator costs, and the seller’s price, then carry the technical cost of making a previously used block operational. The transfer log records parties, dates and resources, not the private price.

The missing information is not evidence of misconduct. Confidential consideration is ordinary in asset transactions. Yet the absence of aggregate price and outcome data makes it harder for policy participants to see whether small-block access is improving or merely becoming more active on paper.

Neutrality needs an incidence report

LACNIC cannot manufacture additional IPv4, guarantee a cheap seller or treat every startup forecast as established demand. It can make the entry path observable.

A useful annual incidence report would distinguish first-time IPv4 recipients from organizations expanding existing holdings. It would publish, in aggregate, request-to-pre-approval time, pre-approval-to-transfer time, approved block sizes, completions, expiries and withdrawals. It could show how many recipients already had LACNIC IPv4, how many operated with IPv6 only, and how often the smallest transferable unit matched the approved need. None of this requires disclosure of private prices or company strategy.

Clearer guidance could also separate the evidence needed for registry approval from the evidence a seller, bank or upstream may demand. A published validity period and status vocabulary for recipient pre-approval would help counterparties understand what LACNIC has verified and what it has not. Small operators would still need capital. They would face fewer avoidable sequencing uncertainties.

This is a test of restrained institutional legitimacy. A neutral registry need not guarantee equal commercial outcomes. It should be able to explain when the same rule lands differently on firms with and without inherited inventory, and it should publish enough evidence for the community to decide whether that incidence remains acceptable.

The new entrant’s disadvantage is not hidden in one hostile clause. It appears between the clauses: prove a network, obtain approval, find a block, raise the money, complete the record, and only then use the resource to make the network easier to sell.

Sources