A small network does not experience IPv4 scarcity as an abstract debate about address policy. It experiences it as a deal that ought to close and does not. The prefix is routable. The buyer has customers waiting. A hosting company wants to bring its own addresses into a cloud platform, or a regional ISP wants to lease a modest block for an enterprise product, or a data-centre operator wants to acquire unused capacity from a company in another country. The engineering questions are manageable. The route can be announced, the abuse contact can be updated, the reverse DNS can be delegated, and the customer can be migrated. Then the transaction leaves engineering and enters the cross-border file.
That file is where the cost accumulates. A bank asks why a company in one jurisdiction is paying a registry in Mauritius in US dollars for resources used by customers in another jurisdiction. A registry asks whether the seller is the current recognised holder and whether the recipient can justify need. A cloud platform asks for assurance that a customer bringing its own IPv4 space will retain registry continuity and route-origin authority. A buyer asks for beneficial-ownership information, sanctions and politically exposed person screening, board resolutions, corporate certificates, tax status, powers of attorney and a legal opinion. A lawyer asks whether the local company registry can produce a document in the required format. A notary asks for wet-ink signatures. A translator asks whether a French, Arabic, Portuguese or local-language filing should be certified for an English-speaking reviewer. A central bank or commercial bank asks whether the foreign-currency remittance is allowed, documented and properly classified.
The address block has not changed. Its economic character has. It is no longer only a network identifier. It is a scarce operational input, a revenue support, a contract object, a compliance file and sometimes a balance-sheet asset. The compliance work does not merely surround the transaction. It prices it. Every unresolved question about lawful control, payment, tax, cross-border use, customer assurance or registry discretion becomes a discount, a delay, a holdback, a warranty, a legal fee or a reason for a smaller operator to give up before the market sees the supply.
AFRINIC makes this problem more visible than most institutions because its region is not one legal environment. It covers the African continent and nearby Indian Ocean economies, with common-law, civil-law and mixed systems; multiple official and business languages; uneven company registries; different telecom licensing regimes; capital controls; fragile banking corridors; sanctions exposure; public-sector procurement rules; state-owned operators; small private ISPs; multinationals; universities; exchanges; hosting firms; cloud customers; and legacy holders whose original files were created before anyone imagined IPv4 would become priced infrastructure.
Scarcity is the background. AFRINIC's exhaustion material records Soft Landing Phase 2 from January 2020, with small final allocations and utilisation checks. Its transfer policy says IPv4 transfers within the region can occur only from an existing AFRINIC member or regional legacy holder, that the source must be the current rights holder recognised by AFRINIC and not in dispute, and that the recipient must be an AFRINIC member that justifies need and signs the relevant agreement. Its fee schedule says accounts involved in transfers must be in good standing before Registration Services considers the transfer, while registry payments sit in a dollar and euro world mediated through banks and card rails. These are not merely administrative facts. They are the coordinates of a market.
Compliance can protect that market. A registry with weak verification becomes a venue for forged documents, captured contact records and revived shell companies. Public reporting on AFRINIC's 2019 address-record scandal showed why fraud prevention cannot be treated as decoration. But a registry can also turn verification into a private permission system. If it asks only what the ledger needs to know, it lowers cross-border verification costs. If it asks whether a lawful commercial transaction fits a discretionary theory of need, geography or virtue, it raises those costs and begins to behave less like a ledger than a gatekeeper. In a scarce-address economy, that difference is not philosophical. It is the difference between liquidity and capital control.
Scarcity turns cross-border delay into a capital charge
The economics starts with scarcity. While IPv6 exists, IPv4 remains embedded in customer equipment, enterprise firewalls, cloud platforms, hosting contracts, allowlists, security appliances, legacy applications and interconnection arrangements. Operators still need IPv4 because customers still need reachability across the existing internet. Once the free pool is constrained, addresses move from administrative abundance into priced reuse: purchase, lease, merger, financing, customer assurance and balance-sheet valuation.
AFRINIC's published exhaustion position changes the value of time. In Phase 2 the pool is no longer a comfortable source of large allocations for ordinary growth. The limits around small allocation and assignment sizes, efficiency checks and contractual-obligation checks tell members that the marginal address will often be acquired through reuse, transfer, leasing, acquisition or more intensive use of an existing holding. When supply is abundant, a slow form is an irritation. When supply is scarce, a slow form is a capital charge.
Delay enters the price in several ways. A buyer waiting for registry completion cannot deploy customers. A seller waiting for review cannot redeploy capital. A lessee waiting for assurance cannot commit a downstream contract. A bank waiting for legal comfort cannot lend against the revenue supported by addresses. A cloud customer waiting for BYOIP approval cannot migrate applications. The cost is not only the lawyer's invoice. It is the lost option to use, sell, finance or move a scarce input during the review period.
Cross-border delay is especially expensive because it is difficult to predict. A single domestic compliance file can be scheduled around one company's registry, one tax authority, one bank and one language. A cross-border address transaction may involve at least two corporate law systems, multiple tax assumptions, foreign-currency remittance, a registry incorporated in Mauritius, buyer-side diligence, seller-side authority, bank risk review, translations, notarisation, local counsel and technical acceptance by an upstream or cloud platform. The prefix can be ready while the file is not.
Markets capitalise that uncertainty. A buyer who does not know whether a transfer will close in three weeks or three months demands a discount. A seller who needs certainty may accept less. A broker prices the risk into fees. A lender adds covenants or refuses to treat the address-supported revenue as durable. A cloud customer asks for a stronger indemnity. If the transaction is small, the fixed compliance cost may consume the margin. The block may remain unused, leased informally or held defensively because the formal process is too costly.
This is why address scarcity and legal certainty move together. The more valuable IPv4 becomes, the more valuable reliable registry recognition becomes. The more discretionary or opaque registry recognition becomes, the more market value is diverted into proof, delay and risk allocation. An institution that reduces verification uncertainty increases the effective supply of usable capital. An institution that expands verification into open-ended approval reduces it.
The point is not that every delay is illegitimate. Some delays reveal forged documents, contested authority or payment problems. The point is that delay must be mapped to a specific ledger risk. If a pause proves that the seller is not the recognised holder, it protects the market. If a pause exists because the registry is reconsidering whether the business model, customer geography or commercial use is acceptable, it begins to ration a scarce asset through administrative discretion. Scarcity makes that distinction visible because every uncertain day has a price.
AFRINIC's region is not one legal document environment
AFRINIC's region is legally plural. That sentence can sound obvious until it is applied to a transaction file. The region includes common-law systems, civil-law systems and mixed systems. It includes English, French, Portuguese, Arabic and other business languages. Some jurisdictions have digital company registries with searchable current extracts. Others rely on paper filings, local registry searches or certificates that do not resemble the document a foreign reviewer expects. Some use concepts such as certificate of good standing; others rely on tax-compliance letters, business-registration extracts, notarised declarations, ministry confirmations or telecom-licence documents. Some corporate acts are public filings; others are private records.
An IPv4 transfer, lease or BYOIP assurance package therefore has to translate legal facts across systems. Does the company exist? Who can bind it? Who owns it? Is it in good standing? Has it changed name? Has it merged? Was the resource included in a sale of business? Is the signatory authorised by law, by board resolution, by power of attorney or by public appointment? Has the relevant licence moved? Are tax or withholding obligations triggered? Is foreign payment allowed? What document proves each fact?
Those questions do not have one regional answer. A French-language registry extract may be legally sufficient in its home jurisdiction but require certified translation for a bank or registry reviewer. A state-owned telecom operator may act through ministry approval rather than a conventional board resolution. A public university may need a council decision or government letter. A small ISP may trade under a name that differs from the registered legal entity. A company may have changed shareholders without changing legal personality. A merger may be structured as an asset sale in one country and a share transfer in another. If the proof standard assumes one corporate template, lawful organisations become non-compliant because their legal system looks unfamiliar.
AFRINIC cannot eliminate this pluralism. It is the cost of operating a regional ledger. The institutional question is whether the registry absorbs some of that complexity by defining functional proof categories, or pushes it onto each member through ambiguous document demands. A good registry asks what fact must be proved and accepts jurisdictionally equivalent evidence. A poor registry asks for a familiar form and leaves the member to discover, after delay, that its country does not produce that form.
The difference is distributional. Large groups can hire local counsel in every jurisdiction, obtain legal opinions, translate documents and produce a diligence binder. Small networks often cannot. They may have a general commercial lawyer, a finance manager and a founder who remembers the history. They may operate in a country where public offices are slow, online searches are incomplete and official documents are expensive to obtain quickly. A proof rule that looks neutral on paper can be regressive because legal form is not equally cheap to produce across the region.
National internet registries, where they exist in other regions, can sometimes reduce local-language and local-document friction by acting as a domestic interface. That is not a general answer for AFRINIC, and it creates its own delegation and accountability problems. For cross-border transfers and leasing, the central problem remains: a regional record must be trusted by parties who live under different legal systems. The registry must recognise equivalent evidence without becoming a judge of every local corporate-law dispute.
Legal pluralism also affects public-sector and procurement cases. A government agency buying cloud services, a state university leasing capacity, or a public telecom company transferring resources after restructuring may need procurement approvals, audit trails, budget authority and anti-corruption certifications that private firms do not need. If those documents are treated as irrelevant by the registry, public buyers face extra legal risk. If they are treated as an excuse for indefinite review, public networks become slow and expensive. The proper approach is narrow: identify the registry fact, identify equivalent evidence, and avoid turning legal diversity into institutional discretion.
Neutral ledgers reduce verification costs; gatekeepers manufacture them
A registry's economic value is not only that it records who holds which resource. It is that other parties can rely on the record without re-litigating the whole history every time a transaction occurs. A neutral ledger lowers verification costs. A discretionary gatekeeper raises them because counterparties must price not only the facts but the institution's future judgment.
AFRINIC's policy materials contain both possibilities. The transfer policy's requirement that the source be the current rights holder recognised by AFRINIC and not involved in a dispute is a ledger rule. It asks whether the record is clean enough to support a transfer. The requirement that the recipient be a member and sign the relevant agreement is also a ledger-administration rule. Good standing can be a payment and accountability rule. These requirements can be objective, predictable and useful.
But the same process can become a gate if need justification, regional theories, business-model suspicion or discretionary evidence demands become the real test. If a recipient must satisfy an open-ended assessment of need before receiving a scarce address block through market purchase, the registry is not merely recording a transaction. It is deciding whether the buyer deserves to participate. If a leasing arrangement is examined not for fraud, authority or contactability but for whether the registry approves the commercial model, the ledger has become a licensing office. If geography is recorded as metadata, it may be useful. If geography decides whether lawful use is acceptable, it becomes capital control.
The term mandate laundering describes this institutional move. A narrow administrative duty, such as keeping accurate records, is wrapped in language about stewardship, community interest or regional development and then used to justify broader control over transactions, commercial use and mobility. The expansion can sound responsible. It may even begin with a real risk. Fraud exists. Bad actors exist. Dormant records can be abused. But the cure becomes a different mandate: the registry stops proving facts and starts judging economic permission.
That shift creates a risk premium. Buyers ask whether a registry might later reconsider the transaction. Lessors ask whether customer geography might become a problem. Banks ask whether recognition can be impaired because a business model is unpopular. Cloud platforms ask whether a BYOIP customer can maintain registry continuity if the registry opens a review. Lawyers add warranties, indemnities and closing conditions. Brokers add fees. The market pays for discretion even when discretion is never exercised, because the possibility must be priced.
Registry neutrality does not mean passivity. A neutral registry can reject forged documents, record disputes, refuse an unauthorised transfer, require current contacts, insist on payment standing, preserve RPKI and reverse-DNS integrity, and publish clear procedures. Neutrality means that these actions are tied to the ledger function rather than to a moral or industrial-policy theory of who should use addresses, where customers should sit or whether monetisation is admirable.
AFRINIC's legitimacy problem is therefore partly a transaction-cost problem. The more it behaves like a predictable ledger, the less counterparties need to spend on legal buffers. The more it behaves like a gatekeeper, the more every cross-border deal must buy insurance against institutional preference. In a scarce market, that insurance is not free. It is paid through lower liquidity, higher spreads, longer diligence and unequal participation.
KYC turns a prefix into a customer-risk file
Cross-border address transactions increasingly resemble financial transactions because the same institutions ask similar questions. Who owns the company? Who controls it? Are any directors politically exposed persons? Are any beneficial owners sanctioned? Is the source of funds lawful? Does the transaction have a commercial rationale? Is the payment going to or from a high-risk jurisdiction? Is the counterparty a shell? Are there nominee shareholders? Is the customer using a cloud account, broker or reseller structure to hide the real user?
These are not registry questions in the traditional technical sense. They are compliance questions imported by banks, payment processors, cloud platforms, large enterprise customers, auditors and sometimes public procurement offices. Once IPv4 has market value, a block transfer or lease can trigger anti-money-laundering, sanctions, anti-bribery and tax processes that were not designed for internet-number resources. The resource is technical. The transaction is commercial. Compliance follows the transaction.
The burden is uneven because KYC has fixed costs. A repeat broker can maintain a standard onboarding package: certificate of incorporation, register of directors, beneficial-ownership chart, tax number, address proof, bank letter, sanctions-screening report, board resolution, sample contract, source-of-funds explanation and legal opinion template. A small ISP trying to complete one transaction may have to assemble those documents from scratch. If one shareholder is a trust, a state entity, a family company or a nominee structure, the file becomes more expensive. If documents are in another language, the cost rises again.
Banks add a separate layer. They do not only ask whether a transfer is legal. They ask whether they want to process it. In some African markets, correspondent banking relationships are thin, dollar liquidity is constrained, and compliance departments are cautious about unusual technology transactions. A bank may not understand why IPv4 addresses are being leased to a customer abroad, why a registry in Mauritius matters, or why a local ISP is receiving funds from a cloud customer in another region. The bank's uncertainty becomes delay, requests for invoices, contracts, tax analysis, proof of services and explanations of the address market.
Sanctions screening is one node in the file, not its main mechanism. Even when no sanction is present, the transaction must survive screening language built for banks, exporters and regulated service providers. The more discretionary the registry process appears, the more other institutions ask for comfort because registry uncertainty becomes part of their own risk.
Beneficial ownership is a good example. A registry may need to know who is authorised to act for the holder, not every economic beneficiary behind a corporate group. A bank may need beneficial-ownership information to process payment. A buyer may need it for anti-corruption policy. A public customer may need it for procurement. If these requirements are not separated, the registry file can become a repository for demands that belong to other institutions. That increases cost and confidentiality risk.
The proper registry posture is to define its own minimum. It should require enough identity, authority and control evidence to protect the record. It should not collect every KYC item a bank or cloud customer might require unless a defined legal obligation applies. It should record where legal compliance affects registry continuity, but it should not convert private banking standards into registry eligibility. Otherwise, de-risking by banks becomes de-risking by the registry, and an address holder can be priced out by compliance expectations that were never adopted as resource policy.
Corporate-form mismatch is a hidden tax on transferability
Many address transactions fail to look simple because the company history is not simple. A network may have started as a partnership, incorporated later, changed its trading name, bought a local competitor, transferred licences, reorganised under a holding company, taken investment, split hosting from access services, or moved assets after a family succession. The prefix may have kept routing throughout. The registry file may not have kept pace.
Cross-border transactions expose these gaps. A buyer in another jurisdiction wants to know that the seller can transfer what it claims to control. A lessor wants to know that the customer will not create a rival claim. A cloud platform wants a letter of authorisation that matches the registry record. A bank wants to know whether the holder is the same legal person that receives revenue. A lawyer wants to know whether a past acquisition transferred number resources explicitly or only "network assets" generally. If the earlier agreement did not name prefixes or ASNs, the parties now need a legal bridge.
Corporate-form mismatch is not fraud. It is the normal consequence of applying one registry logic to many legal systems and business histories. In one country, a change from sole proprietorship to company may be treated as a new legal person. In another, a business-name registration may coexist with a company registration. A telecom licence may be held by a subsidiary while resources are registered to a parent. A state-owned operator may hold assets through a statutory corporation rather than a Companies Act company. An insolvency sale may transfer customer contracts and equipment but fail to mention number resources because the practitioner did not understand them.
Each mismatch creates a proof cost. The parties may need a legal opinion explaining continuity, a supplemental agreement, a board ratification, a court clarification, a regulator letter, a tax clearance, a shareholder resolution or evidence of operational use. If the transaction is cross-border, the opinion may need to be understandable to foreign counsel and the registry. If the previous company was dissolved, there may be no board left to sign. If the founder is unavailable, old email evidence may be all that remains. The prefix still routes, but the market asks whether the legal chain is bankable.
Pre-transaction clarity matters. AFRINIC can reduce cost by publishing exactly what facts must be proved for name changes, mergers, acquisitions, transfers from legacy holders, transfers between existing members and transfers to new organisations. Its fee schedule already says transfer treatment differs depending on whether both parties are existing resource members or the recipient is a new organisation, and that good standing matters. The market also needs clarity on corporate evidence: whether a prefix must be listed in a sale agreement, whether a general sale of network assets can suffice, how court or regulator approvals are treated, and what evidence is equivalent across jurisdictions.
Without that clarity, parties over-document. They hire lawyers to produce more evidence than necessary because no one knows what will be enough. Or they under-document, close the commercial deal, and discover later that the registry record cannot move. Both outcomes are inefficient. The first raises transaction costs. The second creates record inaccuracy and future disputes.
The hidden tax falls hardest on small and mid-sized networks. Their earlier transactions were often practical rather than legally perfect. They bought customers, towers, routers, leases and local reputation. They did not create an international data room for IPv4 assets. If AFRINIC's proof standards punish every historical imperfection without allowing equivalent evidence, transferability becomes a privilege of firms that had sophisticated counsel before scarcity made counsel necessary.
Notarisation, apostille and translation make time a tradable input
The most prosaic steps often set the economic timetable. A document must be notarised, witnessed, legalised, apostilled where the relevant countries participate in the apostille system, or authenticated through a slower consular route where they do not. A translation must be certified. An original must be couriered. A public office must issue a recent certificate because an older one is not accepted. A director must sign while travelling. A bank wants documents no older than three months. A registry ticket pauses until the file is complete.
These acts can be legitimate. Cross-border transactions need reliable signatures and documents. But their cost is not evenly distributed. In some jurisdictions notarisation is inexpensive and fast. In others it requires appointments, lawyers, stamps, tax payments or court-linked procedures. Apostille and legalisation can take days or weeks. Certified translation between English, French, Portuguese, Arabic and local languages can be routine in capitals and expensive elsewhere. A small operator outside the main business city pays travel, courier and time costs that a multinational simply absorbs.
The result is that time becomes a tradable input. A buyer willing to pay for urgent legalisation, specialist couriers and experienced counsel can close faster. A seller without those resources waits. A broker with standard templates can compress the timeline. A first-time participant learns by delay. A cloud provider with strict onboarding deadlines may reject a customer whose documents are lawful but slow. A bank may cancel a pending payment if documents do not arrive within its review window.
Evidence format mismatch is a particular source of waste. A reviewer may ask for a "certificate of incumbency" when the jurisdiction issues a company extract. A bank may ask for "good standing" when the relevant authority issues tax clearance instead. A registry may ask for a document naming directors when director authority is proved through a board minute and public annual return. A notary may certify a copy but not the legal effect of the document. A translator may translate names in a way that creates inconsistencies. Each mismatch triggers a new round.
The institutional answer is not to abolish certification. It is to publish acceptable equivalents. For each fact, the registry should identify primary documents, alternative documents and when a legal opinion can bridge the gap. It should say whether scanned certified copies are enough, when originals are required, how recent certificates must be, which languages require translation, what translation standard is accepted, and whether electronic registry extracts are valid. It should not leave each reviewer to improvise.
Predictability reduces price. If parties know the documents before signing, they can obtain them during negotiation. If they discover them after closing, compliance delay becomes a renegotiation. Holdbacks grow. Warranties expand. Lawyers debate who bears the cost. The buyer may demand a lower price because the seller's documents are not in the expected form. The seller may be forced to accept because the address block cannot be monetised otherwise.
In a market for scarce addresses, even ceremonial acts have economic consequences. The stamp on a document is not only a stamp. It is a condition of closing, a determinant of bargaining power and a contributor to the regional liquidity discount. A registry that understands this will design document standards for cross-border reality rather than treating every delay as private inconvenience.
Payment rails, VAT and foreign exchange controls sit inside registry risk
Cross-border compliance is not only about documents. It is also about payment. AFRINIC's fee material places registry billing in US dollars and euros, with online card payment in US dollars and bank transfer through Mauritius. It says remitters should bear bank charges, distinguishes Mauritius output VAT treatment for Mauritian and non-Mauritian members, and leaves members to raise withholding-tax obligations in their own countries. These details sound like finance-office housekeeping. In a transfer or leasing market, they are part of registry risk.
Good standing matters. AFRINIC's fee schedule says accounts of organisations involved in transfers must be in good standing before Registration Services considers the transfer. That rule can be sensible: a registry should not process significant record changes for parties that have not maintained basic account obligations. But payment standing is not always a simple willingness-to-pay question. A member may face foreign-exchange approval, bank de-risking, correspondent-bank delays, card limits, documentation requests, withholding-tax classification, public-sector payment cycles or local restrictions on remitting dollars.
For a large operator, these are treasury problems. For a small network, they can become transaction blockers. A buyer may be ready, a seller may be ready, and the registry file may be ready, but the payment trail may not. If a bank refuses or delays a wire because it does not understand the invoice, the member falls into risk. If a public-sector buyer must route payment through procurement and budget controls, timing becomes uncertain. If a local tax authority treats the payment differently from AFRINIC's invoice, the member may need a tax opinion. If bank charges are deducted and the registry receives less than invoiced, the account may remain short until reconciled.
Foreign-exchange controls make the problem sharper. Several countries in the region manage access to hard currency, require documentation for offshore payments or prioritise categories of imports. An invoice for internet-number registry services, transfer-related fees or membership obligations may not fit neatly into standard categories. A bank may require contracts, regulatory letters, evidence of service, tax forms or central-bank approval. If the payment relates to a cross-border address deal, the file may also need to explain why funds are moving between parties for a scarce digital resource that is not a domain name, software licence or physical asset.
Payment risk then feeds back into price. A seller may demand payment outside the registry process before taking compliance steps. A buyer may insist on escrow or staged release. A broker may hold funds until registry recognition moves. A lender may refuse to count expected revenue until payment standing is proven. Even if AFRINIC charges no current transfer fee between two existing resource members, the broader transaction still carries banking and tax costs. If the recipient is a new organisation, AFRINIC's schedule indicates membership and allocation-fee logic can apply. The market prices the full path, not merely the registry's line item.
Non-discriminatory payment alternatives are therefore a design issue. A registry serving legally plural and currency-constrained markets should offer payment methods that reduce avoidable exclusion: multiple currencies where practical, clear invoice descriptions, bank letters explaining the service, predictable treatment of bank charges, transparent cure periods for payment failures, and alternative channels when card or correspondent banking fails. It should distinguish unwilling non-payment from documented payment-rail failure.
The separation is important because payment holds can become informal enforcement. If a member's services, transfer processing or record updates can be impaired because a bank or foreign-exchange authority delays payment, the registry may unintentionally bring private banking risk into internet-number continuity. A narrow rule protects the ledger: accounts should be current, and non-payment should be cured. An overbroad rule converts payment friction into a cross-border access barrier.
Leasing and BYOIP expose the same proof problem
IPv4 leasing is often described as a workaround for scarcity, but from a compliance perspective it is also a proof market. The lessee wants to know that the lessor can maintain registry recognition, route authorisation, abuse contacts, reverse DNS and customer continuity. The lessor wants to know that the customer will not create reputational, payment or legal problems. The upstream, cloud provider or enterprise customer wants to know that the addresses can be announced, that the holder has authorised use, and that the record will not collapse during the contract term.
Bring-your-own-IP arrangements in cloud environments make the chain explicit. A customer may need a letter of authorisation, matching registry information, route-origin support, IRR or RPKI consistency, contact information, abuse-handling commitments and proof that the customer has the right to use the block. The cloud provider may not care about AFRINIC politics. It cares whether its platform will be blamed if a prefix is later disputed, revoked, held, polluted by reputation problems or challenged by another claimant. Registry uncertainty becomes platform risk.
Cross-border use magnifies this. A company registered in one AFRINIC-region jurisdiction may lease addresses to customers abroad, announce them through a network in another jurisdiction, or support a cloud deployment in a third. The technical internet allows this. The compliance file asks whether the arrangement is permitted, documented and durable. If out-of-region customer use is treated as suspicious or policy-relevant, downstream customers ask whether the same theory could affect the lease. Even if no action occurs, the risk is capitalised into contract terms.
Leasing also exposes the gap between legal control and operational control. The registry record may remain with the holder. The customer may operate services. The upstream may announce the route. A cloud provider may host applications. Abuse reports may go to one contact while payment goes to another. A sanctions or tax screen may focus on the customer, the holder, the reseller or the payment party. A public-sector procurement office may demand warranties from the immediate supplier even though registry continuity depends on the original holder. The address block becomes a multi-party compliance object.
This is why a registry that treats leasing as morally suspect creates costs beyond its own file. Lessors must explain registry risk to customers. Customers demand termination rights. Clouds require extra assurance. Brokers add protective clauses. Some transactions remain informal because parties fear that explicit documentation will invite review. That is a perverse result: a registry seeking control may push market activity into less visible forms.
The better approach is to recognise the registry facts that matter. Is the holder recognised? Are contacts accurate? Is the use authorised by the holder? Is there a dispute? Are routing-security objects consistent? Are abuse responsibilities reachable? Are payment obligations current? Does a court order or binding legal restriction apply? These questions protect continuity without licensing every customer relationship.
BYOIP also shows why preservation of last verified status is economically important. If a cloud deployment depends on a prefix, a temporary compliance query should not automatically interrupt route authorisation, reverse DNS or registry publication. The last verified operational state should be preserved while specific evidence is requested, unless there is a narrowly defined security or legal emergency. Otherwise the customer bears outage risk from a file dispute it may not control, and cloud platforms will price AFRINIC-region resources accordingly.
Markets price proof, continuity, reversibility and residual risk
The market does not buy only addresses. It buys a bundle: control, proof, continuity, transferability, reputation, routing usability, payment feasibility, tax clarity, legal comfort and a view about who bears unresolved risk. Two /24s can route equally and price differently because one has a clean file, a known holder, current contacts, no dispute, easy payment and a standard transfer path, while the other has a history of corporate changes, missing documents, uncertain customer geography, bank delays or unclear registry treatment.
This is common in asset markets. Legal certainty affects price. A warehouse with a clean title sells differently from one with a boundary dispute. A receivable owed by a strong counterparty finances differently from one subject to setoff. IPv4 addresses are not warehouses or ordinary receivables, but the market logic is similar: usable value depends on institutions that recognise control. When recognition is uncertain, the asset is discounted.
Cross-border compliance adds several layers to the discount. The market asks how much must be spent to show that the seller or lessor can act, whether registry services or route authorisation could be interrupted, whether a challenged deal can be unwound without destroying customer service, whether VAT, withholding and local tax treatments are understood, whether funds can move, whether documents can be accepted without delay, and whether an otherwise lawful transaction can be slowed for reasons not stated in advance.
Escrow and post-decision review can reduce some risk, but they should not become the central design. Escrow handles settlement risk between buyer and seller; it does not solve opaque registry discretion. Review handles adverse decisions after they occur; it does not lower the cost of routine transactions unless the standards are already clear. The larger value comes from predictable first-instance processing: objective evidence lists, published timelines, bounded holds and reasons that map to defined ledger risks.
Residual risk must be allocated. A buyer may accept the risk of future customer reputation problems but not past title defects. A seller may warrant authority but not the buyer's future regulatory treatment. A lessor may provide a letter of authorisation but not guarantee a cloud provider's internal rules. A registry may record recognised control but not adjudicate all beneficial interests under every jurisdiction. Clear allocation lowers price uncertainty. Ambiguous allocation raises it.
AFRINIC's institutional history makes residual risk more salient. Litigation, receivership and contested governance did not stop the internet from functioning, but they reminded counterparties that registry recognition depends on a private institution under domestic law and court supervision. Cross-border parties respond by asking for more comfort. That comfort costs money. A registry that wants to rebuild legitimacy should therefore focus not only on public governance statements but on the ordinary commercial question: can parties predict what risk remains after the evidence file is complete?
The answer shapes liquidity. If AFRINIC-region resources are perceived as harder to transfer, lease, finance or assure than comparable resources elsewhere, the market applies a regional discount. That discount is not paid by abstract speculators alone. It is paid by African networks trying to monetise unused capacity, by new entrants trying to obtain addresses, by customers seeking cloud assurance, and by operators whose financing depends on predictable continuity.
Fixed costs make equality regressive for small networks
Formal equality can be economically unequal. A rule that requires every party to provide the same legal package, the same translations, the same certifications and the same review steps may appear neutral. In practice it imposes a higher per-address cost on smaller networks and first-time participants than on large carriers, cloud firms and repeat brokers.
The reason is fixed-cost incidence. A legal opinion costs roughly the same whether the transaction covers a small block or a large portfolio. A notarised board resolution takes director time regardless of block size. A beneficial-ownership chart, certified translation, tax opinion, bank explanation and registry consultant may cost the same for a /24 as for a much larger block. A large deal can absorb the cost. A small deal may not. The result is a de facto minimum transaction size.
Small networks face other disadvantages. Their records may be less formal. Their founders may have handled early registry matters personally. Their finance teams may not know how to explain IPv4 transactions to banks. They may lack in-house counsel. They may serve customers whose contracts were not written with address transferability in mind. They may operate in countries where public registries are slow or where hard currency is scarce. They may need the transaction proceeds or leased revenue precisely because they do not have large cash reserves.
Repeat players experience the opposite. A broker or large carrier can standardise the process. It can maintain template board resolutions, standard KYC packs, counsel relationships, translation vendors, registry contacts, escrow providers and customer-assurance language. It learns which evidence is accepted and which wording causes delay. It can price uncertainty across many deals. It can wait. Its compliance literacy becomes a competitive asset.
This does not make brokers or large firms villains. Intermediation can reduce search costs, organise files and protect parties from mistakes. The problem arises when the registry process is so uncertain that specialised intermediation becomes mandatory. Then complexity transfers market power to actors who understand the gate. The small operator may hold the scarce asset, but the repeat player controls the path to monetising it.
The regional development effect is serious. Small and mid-sized networks often serve towns, enterprise niches, schools, local hosting markets and access gaps that larger networks treat as secondary. If they cannot efficiently buy, lease, sell or finance IPv4 capacity, their growth is constrained. If they sell only at a discount because compliance is hard, capital shifts away from them. If they avoid formal transfers because they are too costly, registry accuracy suffers. If they depend on brokers for every change, bargaining power moves out of the operating network.
A good cross-border compliance design therefore includes a small-operator test: whether a competent operator with modest resources can understand the evidence before starting, submit functional equivalents from its jurisdiction, cure defects without facing a general resource threat, access payment alternatives if bank rails fail, obtain a decision within a published time and appeal an adverse action without destroying customer service. If not, the system may be neutral in words and regressive in effect.
The test is not sentimental. It is the economic test of whether the registry lowers transaction costs for the region or creates a market architecture that favours scale. In a scarcity economy, fair access to verification is as important as fair language in policy.
Brokers, carriers and clouds arbitrage compliance literacy
Every complex market creates specialists. IPv4 is no exception. Brokers find counterparties, organise due diligence, advise on registry process, arrange settlement mechanics and help parties avoid obvious mistakes. Large carriers and cloud firms maintain compliance teams that can answer customer-assurance questions quickly. Specialist lessors know how to package letters of authorisation, routing objects, abuse commitments and customer contracts. This expertise has value.
It also creates an arbitrage. When the rules are clear, expertise competes on service. When the rules are opaque, expertise competes on access and prediction. A broker who knows how a registry reviewer interprets an ambiguous evidence requirement has an advantage over a holder who only knows the facts of its own business. A cloud platform that can demand extensive assurance shifts the cost to the customer. A large carrier with internal legal and compliance capacity can close deals that smaller networks cannot even price. Compliance literacy becomes a market asset in itself.
The arbitrage can be efficient up to a point. It is better for a small operator to use a competent broker than to lose a transaction through avoidable error. It is better for a cloud customer to learn the required BYOIP file from a repeat platform than to guess. But the institutional goal should be to make specialist help useful, not indispensable. If the only way to complete an AFRINIC-region transaction is to hire insiders, the registry has failed to publish a usable process.
Cloud assurance is a strong example because it is downstream from the registry yet shaped by registry risk. A cloud provider may require the customer to prove route authorisation, ownership or right of use, abuse contactability and non-conflict. If AFRINIC's own standards are vague, the cloud provider writes stricter private standards to protect itself. The customer then faces two gates: the registry and the platform. If the registry later opens a review, the platform may suspend BYOIP eligibility even if the customer has done nothing wrong. Registry ambiguity becomes platform conservatism.
Large carriers behave similarly. They may announce customer space only after reviewing registry records, letters of authorisation and route objects. If cross-border use is politically or administratively sensitive, they ask for more comfort. If they fear de-aggregation, reputation contamination or dispute risk, they price accordingly. The operator seeking transit or hosting may have to explain legal facts to a technical counterparty that has no appetite for registry controversy.
This is how private compliance ecosystems amplify registry discretion. The registry may think a review is narrow. The market hears uncertainty and multiplies it: bank, broker, carrier, cloud, customer, lender, auditor. Each institution adds a layer. The original registry question may concern one document. The downstream cost can affect a whole commercial chain.
Transparency reduces the arbitrage. Published documentary standards, anonymised examples, service-level targets, clear rejection reasons, risk tiers, pre-clearance channels and processing statistics make the market less dependent on private lore. They do not eliminate specialists. They shift specialists from gate interpreters to service providers.
Fraud control is evidence work, not private economic licensing
AFRINIC has real reasons to care about proof. Public reporting in 2019 alleged large-scale manipulation and sale of African IPv4 address records involving dormant or defunct organisations and companies linked to a former AFRINIC official. Whether one focuses on the reported dollar value, the number of addresses or the institutional embarrassment, the lesson is clear: weak records invite theft. A registry that cannot verify holdership and authority undermines the very market it is supposed to support.
Fraud control, however, is evidence work. It asks whether the claimed holder is real, whether the signer is authorised, whether the chain of succession is genuine, whether documents are forged, whether there is a duplicate claim, whether a court order affects the resource, whether contacts have been captured, whether payment standing is accurate, and whether the transaction presented actually occurred. These questions protect uniqueness, accuracy and reliance.
Private economic licensing is different. It asks whether the registry approves the commercial use, the customer geography, the leasing model, the buyer's business plan, the price, the investment motive or the idea that addresses have monetary value. These questions may be attractive to an institution that sees itself as stewarding regional scarcity. But they are not ordinary ledger questions. They turn the registry into a licensing authority for capital movement.
The danger is that fraud control and licensing can use the same documents. A customer list may help investigate a specific fraud claim. It can also become a pretext for judging whether customers are in the desired geography. A business plan may help assess a free-pool allocation request. It can also become a pretext for deciding whether a market transfer should be allowed. A beneficial-ownership chart may help verify authority. It can also become a pretext for reputational judgment. The same file can protect the ledger or expand the mandate.
This is where boundaries matter. AFRINIC should be able to demand strong evidence when a dormant company appears, a signature is suspicious, a transfer chain is incomplete, a court order exists, or a resource is disputed. It should be cautious about demanding customer-by-customer use data, future utilisation plans or moral explanations for leasing when the transaction concerns recognised resources and lawful counterparties. If a prospective policy truly governs commercial use, it should be clear, narrow, adopted through proper process and compatible with reliance interests. It should not be smuggled into individual document review.
Mandate creep becomes capital-control risk when the registry can delay, deny or condition movement of scarce assets based on standards that are broader than fraud prevention and less accountable than public law. The registry is not a central bank, telecom regulator, tax authority, sanctions agency or competition tribunal. It may cooperate with lawful authorities where required. It may preserve accurate records. It may flag disputes. It may require evidence. But it should not use essential registry functions to decide who may monetise, lease, finance or move resources because it dislikes the economics.
Fraud control and market liquidity are not enemies. Clear proof standards can increase liquidity by making buyers trust the record. The enemy of liquidity is open-ended discretion presented as compliance. A disciplined registry reduces fraud by asking better questions. An undisciplined registry reduces liquidity by asking all questions.
Institutional stress raises the cross-border risk premium
AFRINIC's recent institutional history matters for cross-border compliance because counterparties price institutional continuity. The registry has faced allegations of internal record abuse, conflict over Cloud Innovation's large holdings and leasing model, litigation in Mauritius, a period without normal board function, court-appointed receivership, contested election processes, external concern about continuity and continuing disputes. These facts should not become the whole frame of every AFRINIC article, but they do affect transaction costs.
The NRO statement on the official receiver presented receivership as a continuity safeguard. Internet Governance Project commentary argued that receivership could demonstrate resilience through rule-of-law correction, while also recognising the seriousness of the governance challenges. Later reporting on election problems and external interventions showed that confidence was not restored by a single procedural step. For cross-border counterparties, the lesson is simple: when the institution that recognises control is under stress, legal comfort becomes more expensive.
A buyer asks whether a transfer completed during one governance period will be respected later. A lender asks whether registry recognition could be impaired by litigation, a board dispute or a policy change. A cloud provider asks whether route authorisation and registry records will remain stable if a holder is challenged. A public-sector customer asks whether procurement can rely on the resource position. A bank asks whether payments to or from a party connected to contested registry resources carry reputational or legal risk. Lawyers respond with more conditions.
The issue here is narrow: institutional stress increases the private cost of proving that a cross-border transaction is safe. If AFRINIC were a perfectly trusted, predictable ledger, counterparties would need less additional assurance. Because its history includes both record-fraud concerns and discretionary-enforcement concerns, counterparties ask for more. The region pays through slower transactions and discounts.
Institutional stress also changes how ordinary document requests are interpreted. In a high-trust environment, a request for an updated corporate extract looks routine. In a low-trust environment, the same request may be read as the first step toward a broader review. Members may over-lawyer responses. Buyers may wait. Banks may seek written assurances. Staff may become cautious. Caution lengthens timelines, and timelines become price.
The repair is not a public-relations campaign. It is procedural discipline. Every high-consequence request should state the authority, the fact to be proved, the accepted documents, the timeline, the cure path, the effect on existing services, confidentiality treatment and appeal route. If the matter is fraud-related, say so in a bounded way. If it is ordinary completeness, say so. If last verified status or existing routing is unaffected, say so.
Cross-border markets do not require perfection. They require predictability and remedies. Investors, buyers and operators can price known legal systems, known document lists and known timelines. They struggle with private institutions that possess essential recognition power but do not define how that power will be exercised. AFRINIC's institutional recovery will be judged partly by whether routine transactions become boring again.
Due process turns compliance from a threat into infrastructure
Compliance becomes economically destructive when it is unpredictable, irreversible or broader than the risk it addresses. Due process is the mechanism that prevents this. It is not courtroom theatre imported into technical administration. It is the set of practical guarantees that allows parties to plan: notice, reasons, evidence standards, time limits, cure opportunities, impartial review and preservation of continuity while disputes are resolved.
Bounded holds are essential. A registry may need to pause a transfer if the source holder is disputed, a signature appears forged, a court order is presented, payment standing is unclear, or the recipient has not supplied required documents. But the hold should identify the affected resource, transaction or record. It should not spill into unrelated services or unrelated resources. A dispute about one transfer should not impair existing reverse DNS, RPKI, WHOIS or RDAP publication for unrelated holdings unless a specific integrity emergency exists.
Preservation of last verified status is equally important. If a holder was recognised yesterday and a document query opens today, the ordinary presumption should be continuity while the defined issue is resolved. This protects customers who are not party to the compliance dispute. It also reduces litigation incentives. A holder is more likely to cooperate with document requests if cooperation does not carry immediate existential risk. A customer is more likely to accept AFRINIC-region resources if registry review does not automatically threaten service.
Transparent documentary standards turn compliance into infrastructure. The registry should publish standards for transfers, leases where relevant, BYOIP support letters, mergers, acquisitions, name changes, legacy-resource handling, payment standing, authority checks, beneficial-owner questions where legally required, translations, notarisation and alternative evidence. It should identify what is mandatory, what is situational and what is not required. It should provide examples of accepted equivalent documents from different legal systems without turning those examples into an exclusive list.
Appeal and review should be practical, not ornamental. A member should be able to challenge an adverse documentary decision without waiting so long that the transaction dies. Reviewers should have authority to order recording, narrow the request, preserve continuity or require clearer reasons. Courts remain available for genuine legal disputes, but registry-function disputes should not require full litigation whenever a reviewer misreads a foreign document.
Due process also protects the registry. Staff decisions become easier to defend when they follow published standards. Courts receive a clearer record. Members understand the difference between fraud investigation and routine incompleteness. Banks and cloud providers can rely on written procedures. The registry reduces accusations that it is using compliance as leverage. Administrative discipline is institutional insurance.
The compliance file should therefore be designed as a trust-lowering device: it lowers the amount of private inquiry each counterparty must perform. If it instead raises private inquiry by making every party fear unbounded discretion, it has failed. The goal is not less compliance. It is compliance that reduces total system cost rather than exporting cost to the weakest participant.
A lower-cost registry design separates record-keeping from licensing
The institutional design principles follow from the economics. AFRINIC should treat cross-border compliance as a transaction-cost problem in a scarce-address market, not as a moral contest over who deserves mobility. The registry's job is to make lawful and operational reality legible while preserving uniqueness, accuracy, contactability and continuity. It should not become a private licensing authority for commercial models, customer geography or capital movement.
Evidence requirements should therefore be narrow. Each request should map to a defined fact: legal existence, recognised holdership, signing authority, succession, payment standing, transfer consent, dispute status, court restriction, contact control, routing-security authority or customer-use authorisation where strictly relevant. If a document does not prove one of those facts, it should not be required by default.
The last verified status should be preserved while ordinary evidence defects are cured. Registry publication, route-security services, reverse DNS and record continuity should continue unless a defined integrity emergency exists. Adverse action should be limited to the smallest affected record, transaction or resource, and irreversible action should require fraud, abandonment, duplicate claim, court order, severe breach or independent review. This lowers the cost of cooperation and protects customers.
Documentary standards should be transparent and jurisdictionally flexible. AFRINIC should publish functional evidence categories with accepted equivalents for different legal systems, languages and public-sector forms; define translation, notarisation, apostille or legalisation expectations; state document age limits and electronic-record rules; and explain when a legal opinion or alternative evidence can bridge older files. Predictability is the cheapest compliance reform.
Time limits should be real. Ordinary transfers, name changes and authority updates need target processing periods. Pauses should identify missing facts, not merely request more information. If legal escalation is needed, the category should be stated. Markets can price known timelines. They discount unknown ones.
Payment alternatives should be non-discriminatory. Members in foreign-exchange-constrained or bank-de-risked markets need clear invoice descriptions, multiple payment channels where feasible, documented cure periods, treatment for bank-charge shortfalls and a way to distinguish payment-rail failure from refusal to pay. A regional registry should not let correspondent-banking weakness become accidental exclusion from address mobility.
Due process should be built into holds. A hold should state what it affects, why it exists, how to cure it, how long review should take and how to challenge it. Service-specific effects should be separated: a transfer question should not automatically become an RPKI, reverse-DNS, WHOIS or account-continuity question. Compliance must be bounded or it becomes a threat.
Above all, record-keeping should be separated from moral and economic licensing. AFRINIC may record facts about holder location, contact location, payment status, dispute status and lawful authority. It should not deny or delay registry recognition merely because addresses have monetary value, are leased, are financed, are used by customers abroad, are moved for consideration, or are held as operational capital, unless a clear and prospective rule tied to a legitimate registry function specifically applies. Markets, courts, tax authorities, banks and telecom regulators have their own roles. The registry should not launder their mandates into its own.
The conclusion is deliberately institutional rather than heroic. AFRINIC does not need to solve every legal, banking and tax problem in its region. It cannot. It can, however, decide whether it will lower or raise the cost of trust. A neutral ledger lowers cross-border verification costs by telling parties what the record means, what evidence changes it, what continuity is preserved and what remedies exist. A gatekeeper raises those costs by keeping discretion vague and forcing every transaction to insure against institutional preference.
In the economics of scarce IPv4, that choice is capital allocation. If proof is narrow, transparent and bounded, addresses can move toward productive use with less fraud and less fear. If proof is broad, discretionary and entangled with moral licensing, addresses become less liquid, small networks pay more, intermediaries gain power, and registry administration begins to resemble private capital control. The address block still routes. The question is whether the institution around it lets the market treat it as usable infrastructure or prices it as a risky permission.