In a small internet-service office, the decisive archive is often not digital. It may be a cardboard box under a desk, a filing cabinet in a finance room, a folder carried by a founder who still remembers why an old company name was used, or a set of scans kept by an accountant who has since retired. Inside are company-registration certificates, tax letters, invoices, regulator correspondence, a director's passport copy, a board resolution from a year when the company traded under another name, an early address-allocation email, and perhaps the contract by which a neighbouring network was absorbed.
That archive is not nostalgia. It is capital. If the operator wants to sell unused IPv4 space, buy a block from another holder, merge with a rival, reassure a bank, regularise an old record, update contacts after a reorganisation, lease capacity to a customer, or defend itself against a registry query, someone must turn that history into evidence. Documents must be found, scanned, translated, notarised, certified, sequenced and explained. Gaps must be bridged. A 2008 allocation message must be connected to a company in 2026. A founder's personal email must be connected to a current corporate authority. A business acquired in a distress sale must be connected to the number resources still routed by customers. The economic cost is not the PDF. It is the conversion of a messy operational past into a proof package that another institution will accept.
AFRINIC makes this problem unusually visible. The African regional internet registry sits at the meeting point of scarce IPv4, old records, institutional stress, litigation, receivership, contested governance, cross-border business structures and a region full of operators whose administrative histories often do not resemble the compliance departments of large cloud companies. Documentation in that setting is not clerical decoration. It is a market gate. It affects who can move address capacity, who can finance a network, who can defend a historical allocation, who can survive a merger review, and who is quietly priced out before a transaction begins.
The subject needs to be kept narrow. This is not the abuse-contact question. Abuse-contact policy asks whether reports about harmful traffic can reach a working desk, be routed to the right team and be escalated when an operator ignores them. That is a question of contactability and complaint handling. Documentation burden is different. It concerns the cost of producing legally and historically persuasive evidence when resource control, transfer, succession, financing or regularisation is at stake.
Nor is this mainly the identity-verification problem. Identity checks ask whether today's signer is who he says he is and whether he has authority to bind the holder. That is necessary, but it is not the heart of the burden. A passport check, company-director lookup or board resolution may prove today's authority. It does not by itself reconstruct a twenty-year chain of corporate continuity, historical allocation, merger succession, cross-border legal equivalence and transaction completeness. Identity and authorisation are boundary conditions. The economic issue is the fixed cost of producing an acceptable evidentiary record, and how that cost turns paperwork into a transaction-cost filter.
The filter can be useful. A registry that cannot demand proof becomes an invitation to forged letters, revived shell companies, captured contact records and hijacked historical blocks. Public reporting in 2019 described allegations of a large African IPv4 address heist involving dormant or defunct organisations, record manipulation and companies linked to a former AFRINIC official. A registry that sees such weakness cannot credibly accept informal assurances forever. It must ask for stronger evidence.
The danger begins when stronger evidence becomes open-ended evidence; when proof of control turns into proof of commercial virtue; when a missing historical document is treated as a reason to reopen the entire allocation; or when the same documentary burden falls on a multinational group with lawyers and a ten-person ISP with a box of paper. In that case fraud prevention and market access start to diverge. The registry may be saying "accuracy". The market may experience a tax on liquidity.
For an operator, the tax arrives before any formal rejection. It appears as legal fees, registry consultants, director time, translation, notarial costs, courier delays, uncertainty for a buyer, a lender's extra legal opinion and the opportunity cost of addresses that cannot yet be used, sold or financed. A large holder can amortise those costs over many transactions. A small network cannot. A proof standard that looks equal in a manual can be regressive in a market, because evidence production has fixed costs and archives are distributed unequally.
That is why AFRINIC's documentation burden should be analysed as institutional economics rather than etiquette. It decides whether trust repair becomes common infrastructure or a capital-rationing device.
The proof economy begins before the form
Most arguments about registry paperwork start too late, with the form a member must submit. By then the expensive work may already have happened. The real process starts with reconstructing a chain: original allocation, current holder, legal identity, operating continuity, signing authority, resource use, payment standing, contact history and, where relevant, the transaction that moved a business from one corporate shell to another.
Each link can be messy for reasons that have nothing to do with fraud. An ISP may have begun as a family company, incorporated later, changed its trading name, brought in a shareholder, bought a competitor's customers, moved from a postal address to an industrial estate and migrated through several accounting systems. An early allocation may have been requested by a founder using a personal address because that was how small internet businesses worked at the time. A regulator's licence may have been renewed under a slightly different name. A tax certificate may use a punctuation variant. The registry record may preserve an abbreviation no one has used for years. A bank may ask whether the same economic enterprise has existed continuously, while the registry asks whether the resource holder of record is current, and a buyer asks whether a rival claimant could appear later.
The operator experiences this as a chain of translation problems. The company registry speaks one language. The telecom regulator speaks another. The registry database speaks a third. The lender's credit committee speaks a fourth. A lawyer then has to package the same history for a sale agreement, a board resolution, a registry ticket and a financing file. A large firm turns this into a diligence binder. A small network reconstructs it from memory, old emails and the signature habits of people who may no longer be available.
This is why the archive has economic value. An operator with clean corporate records, preserved allocation correspondence, board minutes, acquisition agreements, updated contacts and documented customer succession owns a more liquid resource than an operator with the same routed prefix and a weaker file. In routing terms the addresses may be identical. In market terms they are not, because the cost of proving control is different.
The effect resembles land markets with uneven title records, though IPv4 addresses are not land and should not be treated as ordinary property. They are globally unique identifiers administered through a coordination system. Still, the economic lesson is familiar. An asset whose use depends on trusted records is discounted when record proof is expensive. The address may route today. The question is whether it can be financed, transferred, regularised or defended tomorrow without turning every historical gap into a dispute.
Documentation burden is therefore not the price of paper. It is the price of converting operational history into acceptable institutional evidence. When that conversion is cheap, markets work better. When it is costly, markets concentrate around actors that can afford the conversion.
AFRINIC did not invent this burden. Every regional internet registry must verify resource holders, disputed claims, record updates, mergers and transfers. What distinguishes AFRINIC is the weight placed on evidence by its recent history. Allegations of record abuse made weak proof dangerous. Litigation and governance disruption made discretionary proof dangerous. A registry that had to recover confidence needed both more reliable files and less arbitrary process. Those two goals can coexist only if every document request is tied to a defined fact.
A certificate of incorporation proves that an entity exists. It does not prove that every current customer assignment should be reapproved. A board resolution proves authority for a transaction. It does not prove that a registry may judge the business model behind that transaction. A merger agreement may prove succession. It does not automatically convert a historical allocation into a fresh discretionary grant unless a clear rule says so. The whole issue lies in that boundary between evidence and discretion.
Scarcity makes paperwork a price
Documentation costs matter more because IPv4 is scarce. When addresses were plentiful, delay could be treated as administrative irritation. Scarcity turns delay into a price variable.
Networks still need IPv4 for customers, hosting, firewalls, interconnection, legacy systems and compatibility with services that cannot yet rely on IPv6 alone. As free-pool supply tightened, addresses acquired market value. That value may appear through a sale, a lease, a merger premium, a financing model or the revenue supported by existing customers. Once an operational identifier becomes a priced input, the cost of proving control over it becomes part of the asset's economics.
Consider a modest address block that could support a small hosting product, an enterprise connectivity service or a lease to a customer that still needs IPv4. On paper the transaction may be attractive. In practice it must survive evidence production. If the proof package costs thousands of dollars, consumes months and carries uncertain review, the transaction may fail. A larger block can absorb fixed costs. A smaller block cannot. The result is a minimum efficient scale for participation, created not by routing technology but by paperwork.
This scale effect matters in the AFRINIC region. Many operators are small or mid-sized. They may serve a city, connect schools, operate wireless access networks, host government systems, run enterprise circuits or manage regional data centres. Their holdings may be large enough to matter but not large enough to justify an open-ended legal exercise. If a transfer, regularisation or merger update requires the same documentary machinery as a much larger deal, the small holder faces a higher cost per address.
The same logic applies to buyers. A new entrant may need only a modest block. It may be willing to pay the market price but unable to price the evidence process. It may lack counsel familiar with Mauritius, experience with registry tickets, knowledge of historical policy categories, or cash reserves to survive delay. A broker, cloud company or specialist lessor can handle such processes repeatedly. The rule book may not say "large firms only". The transaction-cost structure can say it in practice.
Delay also changes bargaining power. A seller waiting for registry approval bears market risk. A buyer waiting for transfer completion bears deployment risk. A lessee waiting for documents bears customer risk. A lender waiting for a clean record bears credit risk. A company trying to close an acquisition bears deal risk. If documentary review takes three months instead of three weeks, the cost is not only staff time. It is capital frozen in a market where optionality has value.
Liquidity depends on predictable closing. If parties know that a transaction will close within a defined period once defined evidence is supplied, they can price the deal. If review is open-ended, they discount. Sellers receive less. Buyers demand contingencies. Intermediaries earn more because they know how to manage the process. Smaller transactions disappear. The market becomes thinner, not because supply or demand is absent, but because the proof channel is expensive.
That is how documentation can operate as a hidden capital control. The registry need not ban transfers. It need only make proof uncertain, slow or broader than the ledger risk requires. Assets then become harder to move. Holders depend on tolerance. Buyers demand discounts. Unused resources remain trapped. Informal leasing and side agreements become more attractive than clean record updates. Excessive proof demands can therefore produce the opacity they claim to prevent.
Fraud prevention is necessary, but it must not become re-justification
AFRINIC's strongest argument for demanding documents is fraud prevention. It is also the argument most easily overextended.
Fraud prevention should focus on the facts that protect the ledger: who is the recognised holder; whether the entity exists; whether the signer is authorised; whether a chain of succession is real; whether the resource is disputed; whether documents are forged; whether a contact record has been captured; whether a dormant company's name is being abused; whether a court order affects the resource; and whether the transaction presented to the registry actually happened. These are evidence questions. A registry that fails to ask them invites hijacking.
Re-justification is different. It asks whether the current business use still resembles an earlier need statement, whether customers sit in the preferred geography, whether leasing fits an institutional view of address stewardship, whether utilisation plans look satisfactory, or whether a changed commercial model deserves continued recognition. Some of those questions may be relevant in a narrow free-pool allocation process, when a registry is distributing unallocated addresses at administrative prices. They are much more dangerous when applied to already allocated resources on which holders, customers and lenders have relied.
The distinction is not semantic. Fraud evidence is backward-looking and fact-specific. It asks whether a claimed chain is real. Re-justification is often forward-looking or normative. It asks whether the holder deserves the resource under present institutional preferences. Fraud evidence protects the market. Open-ended re-justification freezes it.
A proportional regime would therefore separate proof of recognition from proof of business virtue. If a company seeks to transfer a block after a merger, the registry needs merger documents, corporate records, current registration, tax or equivalent standing where relevant, board approvals and signatures from authorised parties. It may need evidence that the number resources were included in the transaction. It does not need to reopen every customer assignment unless a specific fraud or policy trigger justifies that inquiry. If a holder seeks to update a name after a rebrand, the registry needs evidence that the same legal person changed its name. It does not need to ask whether today's business would have qualified for the original allocation. If a buyer seeks to acquire space, the registry needs to know that the source is legitimate and that the buyer can be recorded. It should not turn the purchase into a moral examination of address commercialisation.
This discipline protects the registry as much as the member. Overbroad proof demands raise the stakes and invite litigation. A holder is less likely to fight a request for a missing board resolution than a demand to disclose customer usage across jurisdictions under threat of resource loss. Courts are less likely to be dragged into registry administration when the registry can show that each document maps to a defined ledger risk.
The public lesson of AFRINIC's recent history is not that documentation should be weak. Weak documentation is what allows old records to be manipulated. The lesson is that documentation must be specific. A registry that has suffered record abuse needs reliable evidence. A registry that has suffered legitimacy problems needs bounded discretion. It cannot repair trust by turning every proof request into a broad review of business acceptability.
This is also where the burden differs from identity verification. A check on today's signer is a necessary control against impersonation, but it is not enough to justify an archaeological review of the holder's entire commercial life. Conversely, a clean corporate-continuity file may still require a signature check before an update is processed. The two subjects touch, but they should not be merged. When they are merged, a narrow authority question becomes a reason for a much larger evidence demand.
Archive asymmetry turns equal rules into unequal costs
The phrase "equal treatment" can hide severe inequality when archives are unequal. Asking every member for the same document may be fair in a procedural sense and unfair in an economic sense if one member can produce it instantly while another must reconstruct it at great cost.
Large companies tend to maintain corporate-secretarial files, audited accounts, formal board minutes, signed acquisition agreements, document-retention systems, counsel and compliance staff. Directors are usually traceable. Names are more likely to be consistent across banks, regulators and contracts. Correspondence goes to role accounts rather than personal mailboxes. When a registry asks for evidence, the result is a polished file.
Small operators often have thinner archives. A company may have been founded by engineers rather than lawyers. Early decisions may have been made by email, invoice or handshake before formal resolutions became routine. Documents may sit in local government offices that are slow to search. A founder may have died, emigrated or fallen out with the current owners. A paper certificate may have been lost in an office move. A local acquisition may have been commercially real but not drafted with number-resource language. The network may have operated continuously for years, paying fees and serving customers, while its paper trail remained imperfect.
The registry sees missing documents. The economist sees archive asymmetry. Both views matter. Missing documents can be a fraud risk. They can also be the residue of doing business in markets where corporate formality was expensive, digital systems were immature and network operators were focused on keeping service alive rather than preparing for a future scarce-asset market.
Archive asymmetry is especially important for old allocations. A block issued in an earlier internet era may have moved through corporate changes before anyone understood how valuable IPv4 would become. Parties may not have listed prefixes in sale agreements because they thought they were buying "the ISP", including its customers, equipment, licences and operations. A founder may have requested resources under one name before incorporation under another. The evidence that would be obvious to create today may never have existed. A strict proof standard applied retroactively can punish historical normality.
That does not mean old allocations should be immune from scrutiny. They are precisely where fraud risk can be high, because dormant companies and stale contacts are easier to abuse. The answer is evidentiary humility, not evidentiary surrender. A registry should be able to reject a claim that cannot connect itself to the recognised holder. It should also define acceptable alternative proof: long-running payment history, consistent routing, customer continuity, tax records, regulator files, notarised director statements, audited accounts, old invoices, contemporaneous emails, court filings, public procurement records, bank letters or other documents that together show continuity. A single missing certificate should not defeat a strong evidentiary chain. A single old email should not defeat clear evidence of fraud.
Markets need this nuance because archive asymmetry creates a discount. Buyers and lenders pay less for resources whose proof depends on fragile files. That may be rational. But if the registry's proof rules are opaque, the discount spreads from weak files to the whole regional market. Everyone pays a governance premium because no one can predict which archive defect will matter.
AFRINIC can reduce that premium by publishing evidentiary categories. It can specify which documents are normally sufficient for name changes, mergers, acquisitions, liquidations, government entities, universities, subsidiaries, parent-company reorganisations, estates, cross-border transfers and legacy regularisation. It can identify which defects are fatal and which are curable. It can list alternative proof routes when historical documents are unavailable. Clarity does not weaken the ledger. It makes honest reconstruction easier and fraudulent reconstruction harder.
Mergers make networks into legal archaeology
Mergers and acquisitions are where documentation burden becomes most visible. The business logic may be simple: one ISP buys another's customers and network assets; a data-centre operator acquires a hosting company; a telecom group consolidates subsidiaries; a distressed network sells its operations; a bank finances a roll-up of regional providers. The registry logic is more exacting. Which legal entity held the resources? What was sold? Who signed? Were the resources included? Did liabilities move? Were both parties in good standing? Does policy treat the event as a transfer, a merger, a name change or a new request? Will the resulting record be accurate?
The gap between business logic and registry logic can be wide. Commercial parties may think they bought "the network". The registry may ask whether the agreement specifically lists the IPv4 prefixes and ASNs. The buyer may have acquired customer contracts and equipment but not the original resource-holding entity. A court or insolvency practitioner may have approved a sale of assets without understanding number-resource terminology. A bank may have taken security over network assets without checking whether registry recognition can move. The deal may close before the documentation question is answered.
That is legal archaeology. Lawyers and engineers dig through the transaction after the fact to determine whether the economic reality matches the evidence. Was the resource part of the business? Did the seller have authority? Did a board resolution approve the transfer? Did a regulator approve the licence change? Did tax clearance matter? Was the acquired company later dissolved? Are there creditors who could object? Did the buyer continue the same customers and routing? Each answer changes the burden.
For a large transaction, the archaeology is expensive but manageable. For a small operator, it can be fatal. The cost of producing a clean succession file may exceed the economic value of the transaction, especially for smaller blocks. A buyer may walk away. A seller may keep unused addresses because sale is too hard. A financier may discount the business because a key resource cannot be reliably transferred. Customers may remain behind a fragmented network that would be better integrated into a stronger firm.
This is not an argument for ignoring merger evidence. Address blocks should not move merely because two parties say a deal happened. Fraudsters would exploit such laxity. The argument is for proportional proof and pre-transaction clarity. The registry should tell parties, in plain operational terms, what evidence is needed before closing. If resources must be listed explicitly, say so. If a board resolution needs particular authority language, provide a model. If both parties must be in good standing, say when that is tested. If a dissolved company creates a special path, define it. If a court-approved sale requires extra documents, define them.
Corporate continuity should be treated as a fact to be proved, not a trap to be discovered. The registry should want formal record updates to be cheaper and safer than informal workarounds. If formal regularisation is unpredictable, parties will shift economic control while leaving registry records stale. That is bad for everyone. The registry sees less accuracy. The buyer assumes more risk. Customers rely on records that no longer describe operational reality. A process intended to protect the ledger ends up preserving fiction.
Mergers also show why documentation burden affects investment. Consolidation can be healthy when it rescues weak networks, improves service and redeploys scarce address capacity. It can also be harmful if it concentrates resources without accountability. The registry's role is not to decide every industrial-policy question through ad hoc evidence demands. It is to make sure that the record reflects lawful change and that fraud is blocked. If that function is predictable, investors can price deals. If it is unpredictable, they discount the whole region.
Cross-border proof multiplies the cost
AFRINIC serves a region of many jurisdictions, languages, legal systems, corporate registries and administrative capacities. Cross-border documentation is not an edge case. It is part of the regional registry model.
A company incorporated in one country may hold resources used by customers in another, be financed by a bank in a third, use directors resident in a fourth and contract with a buyer or lessor in a fifth. Documents may need notarisation, apostille or consular legalisation. Some countries do not issue certificates in the format a foreign reviewer expects. Some company registries are online; others require local searches. Records may be in English, French, Arabic, Portuguese or local languages. Legal systems distinguish in different ways between registered office, trading address, tax address and licence address. Some corporate acts are public. Others are private. A single request for "proof of legal continuity" can become a cross-border project.
The multiplier is both monetary and temporal. Translation takes time. Notarisation takes time. Embassy legalisation can take weeks. Couriers lose documents. Public offices close for holidays. Directors travel. Banks require certified copies no older than a certain date. A registry ticket may sit while a holder waits for a government office. A buyer may not keep a price open indefinitely. A lender may not disburse until the resource file is clean. Delay becomes part of the transaction price.
The burden falls unevenly. A global group may have counsel in several jurisdictions. A small ISP may rely on a general commercial lawyer who has never handled number resources. A company in a country with efficient digital registries can produce documents quickly. A company in a country with paper records cannot. A registry that treats all delay as member failure misunderstands the region it serves.
Cross-border proof also creates interpretation risk. A document normal in one jurisdiction may look strange to a reviewer elsewhere. A company may use a trade name publicly and a different registered name legally. A government agency may not issue a "good standing" certificate but may issue a tax-compliance letter. A merger may be effected by share transfer rather than asset sale. A state-owned entity may require a ministry letter rather than a board resolution. A non-profit, university or municipal network may not fit a private-company template. If the proof framework is too rigid, legitimate organisations fail because they do not resemble the template.
The solution is functional equivalence. The registry should begin with the fact to be proved: existence, standing, authority, succession, resource inclusion, absence of dispute, payment status, operational continuity or consent. Once the fact is identified, different legal systems can supply different documents. A rigid list of forms should be supplemented by a functional proof list. That gives reviewers discipline without forcing every jurisdiction into one corporate model.
Cross-border proof is also where documentation burden intersects with development. If African and Indian Ocean networks are expected to attract capital, merge across borders, import address supply, sell unused resources, lease capacity and serve regional customers, the transaction process must recognise cross-border reality. The region is not a single filing office. Its proof rules should not pretend otherwise.
Confidentiality matters here. Cross-border files may include sale prices, financing terms, customer contracts, licences, tax documents and personal information. A registry needs enough evidence to protect the record, but members need assurance that sensitive material will not be exposed casually or used for wider purposes. A clear statement of why each document is requested, how it is protected and what decision it supports lowers the cost of disclosure. Without that discipline, members rationally under-disclose or litigate.
The evidence market rewards scale
When proof becomes costly, a secondary market grows around proof production. It includes lawyers who understand local corporate law, registry consultants who know which documents tend to pass review, brokers who anticipate transfer friction, translators, notaries, investigators who reconstruct old records, and accountants who package continuity for lenders.
Such a market is useful when it creates reliable files. It is harmful when it becomes necessary because ordinary members cannot understand what the registry will accept. The difference lies in whether specialists add value or merely decode institutional uncertainty.
Scale changes the economics. A large address holder or frequent broker learns the process once and repeats it. It builds templates, relationships, checklists and internal records. Each new transaction becomes cheaper. A small operator faces the full learning cost on a single event. It may not know whether to hire a lawyer before approaching the registry, whether a notarised document is enough, whether a scan will be accepted, whether directors must sign in person, whether a prior name change must be regularised first, or whether a legal file will trigger questions about utilisation after the transaction documents are submitted.
This gives sophisticated actors a structural advantage. They may not be better network operators. They are better evidence producers. In a scarce-asset market, that advantage is powerful. It can allow them to buy under-documented resources at a discount, package them into cleaner files and capture the uplift. That is not inherently abusive. Markets reward those who reduce uncertainty. But if the uncertainty is produced by unclear registry standards, the uplift is a private gain from public opacity.
The evidence market can also change bargaining power. A small seller with weak documentation may accept a lower price from a buyer who promises to "handle the registry". The buyer then controls the proof process and may discover defects that justify repricing. A lender may require an expensive legal opinion before recognising address-supported revenue. A lessee may demand indemnities because the lessor's registry file is unclear. Each demand may be rational. Together they transfer value away from operators with weak archives.
Banks and investors understand this burden quickly. A credit officer does not need a philosophical view on whether addresses are property, resources, rights of use or service records. The bank asks simpler questions: can the borrower keep using the addresses; can the addresses support revenue; can the business survive a registry dispute; can the resource position be preserved through a merger or distress sale; and how much legal uncertainty should be priced into the loan?
For a small ISP, the answer affects capital. A fibre build, data-centre upgrade, tower expansion or enterprise-service platform may need financing. IPv4 holdings may not be formal collateral, but they are part of the revenue story. Customers depend on them. Hosted services use them. Firewalls, allowlists and contracts refer to them. If the bank sees the address position as uncertain, borrowing capacity falls. A clean file lowers financing cost. A messy file raises it.
This is the quietest economic harm. No public dispute occurs. No transfer is denied. The operator simply receives worse financing terms because the evidence environment around number resources is uncertain. A large operator can absorb that through balance-sheet strength. A small operator cannot. Documentation reform should therefore be treated as capital-market infrastructure, not only as member service.
AFRINIC should want expertise to be helpful but not mandatory. It can publish transaction checklists, sample resolutions, country-neutral proof categories, expected timelines, escalation paths and anonymised examples of accepted alternative evidence. It can distinguish ordinary defects from fraud indicators. It can allow pre-clearance questions before members spend money. It can provide a safe path for proactive record correction without turning every correction into a broad review. Clarity reduces the rent earned by procedural opacity.
The small-operator test
The fairest way to judge a documentation regime is to ask whether a competent small operator can comply without surrendering a disproportionate share of the transaction's value.
Imagine a regional ISP with a /22, an ASN, a mixed wireless and fibre footprint, ten staff, a local lawyer and no in-house compliance department. It has operated for fifteen years. It acquired a neighbouring network in 2016 through an agreement that mentioned customers, equipment and service contracts but did not list every number resource by prefix. The founder who requested the original allocation has retired. The company changed its registered address twice. It now wants to sell part of an unused IPv4 block, lease capacity to a customer, or use the resource position as part of a financing package for last-mile expansion.
What should the registry require? It should require evidence of the current legal entity, proof that the resource record belongs to it or its predecessor, evidence of succession from predecessor to current entity, authority of the signer, current contacts, payment standing and disclosure of any known dispute. It may ask for the acquisition agreement and supplementary evidence showing that number resources moved with the network. It may ask for a board resolution. It may ask for regulator records if they clarify continuity. It may ask for evidence that no rival claimant exists where the predecessor has been dissolved.
What should it not require by default? It should not require the operator to relitigate whether the original allocation would be approved today. It should not demand customer-by-customer use evidence unless a specific fraud or policy trigger justifies it. It should not treat missing 2016 wording as fatal if other evidence shows continuous operation and no rival claim. It should not force the operator into months of uncertainty without a clear list of missing facts. It should not make the operator hire specialists whose fees consume the economic value of the transaction.
If the operator can comply through a clear evidence bundle, the regime passes the small-operator test. If the operator cannot know what is enough until after spending heavily, the regime fails. If only a broker can make the file acceptable, the regime may be creating broker dependency. If the registry can use the file to reopen business-model questions unrelated to the requested change, the regime becomes a control device.
The small-operator test is not sentimental. It is a market-efficiency test. Small networks often serve customers and places that large networks ignore. They may hold modest resources that could be redeployed more productively if transactions are possible. They may need financing. They may be acquisition targets in rational consolidation. If documentation rules make them illiquid, the region loses more than administrative neatness. It loses network investment.
The test also prevents a common policy mistake: designing evidence requirements around the best-resourced members and then calling the result neutral. A regional registry serving diverse legal and business environments should design for both the operator with a data room and the operator with the cardboard box. The standard of reliability can be the same. The routes to reliable proof should be flexible enough to reflect reality.
Receivership raised the burden of confidence
Receivership is meant to preserve continuity. It is not meant to make every later document request suspect. Yet in a registry context, receivership inevitably changes how evidence is interpreted.
When a court appoints a receiver to maintain an organisation, arrange governance steps and restore lawful operation, counterparties ask who has authority during the interim period. Can staff process transfers? Can member status be updated? Can a document signed during the period be relied on later? Can a board restored after a contested process approve rules that affect address mobility? These are not abstract legal questions. They enter transaction files.
Public reporting on AFRINIC's governance disputes, including concerns over voter documentation and powers of attorney during election attempts, showed how evidence of authority can become politically and economically decisive. The point here is not to make those episodes the framing for all registry action. It is narrower. When authority records are contested in the institution's own governance, members and counterparties will be more sensitive to authority demands imposed on them. The registry must therefore be unusually clear about who asks for documents, under what authority, for what fact and with what route of review.
Board legitimacy matters because boards approve budgets, procedures, policies and institutional posture. A documentation rule applied by an institution emerging from contested governance may be formally valid and still carry a market discount if members doubt its stability or motives. Buyers may delay. Lenders may demand extra comfort. Holders may comply while reserving legal arguments. The cost appears not only in court filings but in hesitation.
Recovery claims, budgets and strategic plans can reduce uncertainty, but only if ordinary members can see how evidence requests work. A high-consequence request should state the purpose, authority, required facts, accepted documents, timeline, cure path and appeal route. It should separate curable defects from severe consequences. It should identify when a file is being reviewed for fraud rather than ordinary completeness. It should make clear when business-model issues are outside the scope of the document request.
The receiver-era lesson is modest but important: continuity needs paperwork, and paperwork needs legitimacy. A registry emerging from court supervision should not ask members to bear indefinite proof burdens while its own authority chain remains difficult to inspect. Mutual evidence discipline is the only credible path. The registry should ask members for records in the same spirit in which members and markets ask the registry for process: not as theatre, but as the price of trust.
This is also why documentation reform should not be presented as a favour to members. It is institutional self-protection. Narrow, predictable proof standards reduce disputes, make staff decisions easier to defend, help courts understand contested cases, and lower the suspicion that every request hides a wider agenda. A recovering registry cannot afford ambiguity to be its main control mechanism.
Proportional proof would protect both sides
A better documentation regime would not be lighter in every case. It would be more structured.
The first principle is fact mapping. Every request should identify the fact to be proved: legal existence, resource holdership, signing authority, corporate succession, merger inclusion, payment standing, contact control, operational continuity, absence of dispute, court restriction or transfer consent. A document should not be demanded because it is administratively familiar. It should be demanded because it proves a defined fact.
The second principle is tiered risk. A low-risk name correction should not face the same burden as a contested transfer from a dissolved entity. A routine merger between two active members in good standing should not face the same scrutiny as a claim based on a dormant company. A transfer involving a block with no dispute history should not be treated like a case involving forged documents. Risk tiers let the registry concentrate scrutiny where it protects the ledger.
The third principle is alternative evidence. Where historical documents are unavailable, a holder should be able to submit a bundle: payment history, routing continuity, old invoices, customer contracts, regulator records, tax records, public filings, notarised statements, bank documents, court records or contemporaneous correspondence. The registry can weigh the bundle. It need not accept every bundle. But it should say which kinds of alternative proof can satisfy which facts.
The fourth principle is prospective clarity. If future transfers require explicit listing of number resources in sale agreements, publish that rule clearly. If board resolutions must contain particular authority language, provide sample wording. If both parties must be in good standing, state when standing is tested. If translations must be certified, define certification. If documents expire after a period, say so. The cheapest documentation burden is the one avoided by writing the right document before the transaction closes.
The fifth principle is cure before penalty. Missing documents should trigger a request, explanation and cure period. Persistent failure may justify a status flag or refusal to process the specific transaction. Severe consequences should require independent grounds such as fraud, abandonment, duplicate claims, court order or serious contractual breach. A missing certificate should not become a general resource threat.
The sixth principle is time discipline. The registry should publish target processing periods for each risk tier and explain pauses. If a file is incomplete, the holder should know exactly what is missing. If legal escalation is required, the holder should know the category and expected timing. Markets can price known delay better than uncertain delay.
The seventh principle is confidentiality with auditability. Transaction files contain sensitive information. Customer lists, sale prices, financing terms and private contracts should not be casually exposed. But decisions should be auditable. The registry can publish aggregate statistics: number of transfers, merger updates, name changes, rejected files, average processing time, common missing documents, fraud escalations and review outcomes. Aggregate evidence builds trust without revealing private files.
The eighth principle is separation from business-model judgment. Documentation should prove control and continuity. It should not become a channel for deciding whether leasing, out-of-region customers, address monetisation or financing structures are desirable unless a clear and prospective policy squarely governs the transaction. The registry may record facts. It should be cautious about turning evidence files into commercial approval files.
Such a regime would protect AFRINIC as well as members. It would reduce forged transfers, lower support burden, improve records, make litigation less attractive and give courts a cleaner trail if disputes arise. Proportional proof is not a concession to weak compliance. It is the architecture of a credible ledger.
The real ledger is the cost of proof
The economics of documentation burden is easy to underestimate because paperwork looks small beside litigation, board elections, receivership and public conflict. But paperwork is where institutional power meets ordinary operators. Most members will not litigate against a registry. Most will not take part in governance campaigns. Most will encounter the institution through a ticket asking for documents.
That ticket can do two things. It can make the ledger more truthful by asking for the narrow proof needed to record reality. Or it can make the market less free by turning evidence production into a costly and uncertain test of acceptability. The same request can be either, depending on scope, timing, proportionality, confidentiality and remedy.
For AFRINIC the stakes are high because the region's IPv4 economy is already constrained. New free-pool supply is limited. Transfers and leasing matter. Small operators need capital. Old records need repair. Fraud must be prevented. The registry's own continuity has been tested. In that environment, a document is not just a document. It is a price signal, a risk allocation and sometimes a barrier to entry.
The operator with the cardboard box understands this before the policy room does. It knows that a missing certificate can delay financing, that a retired founder can become a bottleneck, that a merger agreement written by a local lawyer can be reinterpreted years later, that a bank may discount revenue if registry recognition is uncertain, and that a large competitor can afford the proof exercise more easily. It knows that the cost of proving continuity can decide whether scarce IPv4 becomes productive capital or trapped inventory.
AFRINIC's documentation problem should therefore be framed with institutional discipline. Fraud prevention is necessary. Evidence is necessary. Archive repair is necessary. But proof requirements must be limited to the facts the ledger needs, designed around regional legal diversity, scaled to transaction risk, supported by alternative evidence, protected by cure periods and insulated from open-ended commercial judgment. Anything else turns documentation into an invisible capital control.
The registry that gets this right will not merely have tidier files. It will lower the cost of trust. In a scarce-address market, that is the economic function that matters.