Summary

  • AFRINIC’s summary accounts prove a 2012 bad-debt expense of MUR 3,239,817, translated as USD 107,994 and rounded in narrative reporting to USD 108,000. The exact dollar expense rose by USD 82,419 from 2011, to about 4.22 times the prior figure, but the record does not disclose the debtor count, ageing, concentration, individual assessments or complete approval path behind it.
  • The annual report places unpaid fees, reclaimed resources and 51 closed members in the same aggregate sequence. It does not show that the USD 107,994 represented 51 accounts, that every closure involved a written-off balance, that every written-off account had resources reclaimed, or that any one act caused another.
  • The June 2013 draft member-meeting minutes record that the Board had not yet ratified the audit; a September-coded resolution later records Board approval; the auditor’s opinion is dated 29 October 2013. That is a visible chronology, not evidence of a defect. Aggregate audit approval and an unmodified opinion still do not reveal who authorized each bad-debt decision or how collection and closure cases were reconciled.
  • The strongest explanation is ordinary prudent accounting: old unpaid balances were assessed as uncollectible, contractual recovery measures were taken, aggregate results were disclosed, the Board approved the audit, and the external auditor issued an unmodified opinion. That case is plausible. It is also compatible with publishing a confidential, anonymized movement schedule that would let members test the accounting without identifying debtors.
  • AFRINIC is a private technical bookkeeper and coordinator. A receivable write-off, member closure or resource-register change is an accounting or contractual act, not a punishment, adjudication or exercise of sovereign power. The ledger should follow evidenced reality; it cannot decide guilt.

The entry, the closure and the registry action are not one event

At the centre of the 2012 disclosure is a precise accounting amount: MUR 3,239,817, translated in the summary statement of comprehensive income as USD 107,994. The accompanying narrative rounded it to USD 108,000 and called it the highest amount of bad debts AFRINIC had recorded. That is what the accounts establish. It is an expense recorded for the year, not a list of debtors, a count of default cases or an administrative order against a member.

The distinction is easy to lose because the annual report described the surrounding events in a compressed sequence. It said the written-off bad debts related to unpaid fees, said the associated resources had been reclaimed, and reported that 51 members were closed during the year. In prose, adjacency invites a single story: 51 members did not pay, their resources were taken back, and their combined debts became USD 108,000. Yet the public evidence does not supply the connective tissue for that story.

There is no account-level or cohort-level crosswalk, no number of receivables written off, and no statement that the closed-member population and bad-debt population were identical.

Three separate institutional ledgers are in view. The financial ledger contains invoices, collections, balances, impairment assessments and derecognition. The membership record contains contractual status and closure. The number-resource register records administrative changes associated with registry resources. A single member could, in principle, appear in more than one of those domains, but overlap must be demonstrated rather than assumed. Even the annual report’s aggregate wording cannot tell the reader whether all, some or none of the 51 closures corresponded to the balances making up the expense.

That separation is more than accounting etiquette. It prevents the public from treating a non-cash expense as if it were an act imposed on a debtor. A write-off removes a carrying amount once a debt has been assessed as uncollectible; it does not by itself show that every collection step was exhausted, that an underlying contractual claim vanished, or that a membership closure was justified. In the same way, a membership closure is not an accounting policy, and a registry-resource change is not a finding of fault. Each act needs its own rule, evidence, notice and authority.

AFRINIC’s own report showed why the combined picture mattered. It said the number of closed members was a matter of concern and had been researched. It also said resource reclamation associated with non-payment affected membership-fee growth and proposed a more systematic, structured fee-collection process for future years. Those statements acknowledge an operational mechanism: collection outcomes could influence both the financial base and the membership population. What the report did not publish was the research it referred to, an arrears ageing, a recovery ledger, or closure-level outcomes.

The investigative task is consequently narrow. It is not to identify confidential debtors, reconstruct private disputes or infer motives from an expense. It is to determine what the public documents let members test. They let members verify the aggregate figure, compare it with prior-year numbers, observe the institution’s own account of unpaid fees and resource reclamation, and trace the later approval of the audit. They do not let members reconcile the expense to its underlying population. That unresolved boundary is where treasury governance begins.

A material change, with carefully limited arithmetic

The scale is clearest when the 2012 expense is compared with the figures disclosed beside it. AFRINIC recorded USD 25,575 in bad-debts expense for 2011. The 2012 exact amount of USD 107,994 was therefore USD 82,419 higher. It was approximately 4.22 times the 2011 amount, an increase of about 322.3 percent. Those calculations describe the size of the change. They do not reveal why the balances accumulated, whether collection practice deteriorated, whether a backlog was cleared, or whether management changed the threshold for judging an account uncollectible.

The June 2013 draft AGMM minutes add one piece of context. They record the Finance Director linking a deficit above USD 200,000 to the write-off of accumulated member debts from the previous three years and to higher travel costs. The annual report, in its own discussion, identified bad debts and a lower number of new members as the two main factors behind the 2012 deficit. Neither account allocates the deficit mathematically among those causes. The minutes suggest that the write-off reached back across accumulated balances; they do not supply the invoice dates, service periods or account composition needed to test that description.

Membership-fee income offers another scale reference. The 2012 accounts recorded MUR 79,080,963, or USD 2,636,032, in membership-fee income, compared with USD 2,453,781 in 2011. The USD 107,994 bad-debt expense was about 4.10 percent of that year’s membership-fee income. In a member-funded institution, four cents of expense for every dollar of that annual income measure is large enough to demand explanation. But it is not a collection rate. It does not mean 4.10 percent of invoices issued in 2012 went unpaid, because the public record does not establish that the expense came only from that year’s billings.

The draft minutes point in the opposite temporal direction by describing debts accumulated over three years.

The closing balance sheet provides a third comparison, and the caveat is even more important. Trade and other receivables stood at MUR 18,166,684, or USD 605,556, at 31 December 2012, up from USD 369,768 at the end of 2011. The bad-debt expense was about 17.83 percent of the 2012 year-end trade-and-other-receivables balance. That percentage is only a scale comparison between an annual expense and a broad closing balance. It is not a write-off rate, an allowance ratio, a default rate or a loss rate.

“Trade and other receivables” is wider than a schedule of membership debts, and a year-end stock is not the same object as a full-year expense flow.

The deficit makes the stakes visible without proving a causal allocation. AFRINIC reported a 2012 deficit of MUR 7,892,724, or USD 263,091, compared with USD 3,822 in 2011; the annual-report narrative described 2012 as its first reported deficit. The bad-debt figure was significant beside that result, and AFRINIC itself named bad debts as a main factor. Still, subtracting one figure from another would not reconstruct the institution’s finances. The report also discussed the lower number of new members, while the meeting minutes mentioned higher travel costs. The evidence supports materiality, not a single-cause theory.

Used properly, these ratios sharpen the questions. A fourfold year-on-year increase asks whether old balances were cleared in a concentrated exercise or whether new risks emerged. A figure equal to 4.10 percent of membership-fee income asks how costs and reserves absorbed the loss. A scale equal to 17.83 percent of broad year-end receivables asks what remained outstanding after the derecognition. None answers those questions. The numbers are diagnostic signposts, not verdicts on collection quality or individual members.

That is why a receivables movement matters more than another ratio. A useful schedule would start with gross opening membership receivables and the opening allowance, add new billings, subtract cash collections and credits, distinguish new impairment from actual write-offs, show any recoveries, and reconcile to the closing balance. Without that bridge, the public sees three isolated amounts—income, expense and closing receivables—without knowing the flow that connects them.

Five dates and the limits of the approval record

The public chronology begins before the accounting year. AFRINIC’s 2011 resolution register said external transactions over USD 100,000 required Board approval and created a Finance and Audit Committee for one year. Numerically, USD 107,994 exceeds that threshold. Institutionally, however, the comparison cannot carry the conclusion that the write-off required Board approval under that rule. A derecognition entry is not necessarily an external transaction. The register does not define the 2012 expense as a covered transaction, and the available documents do not establish the rule’s application to a non-cash accounting assessment.

That distinction also prevents an inverse allegation. One cannot say the Board failed to approve the write-off merely because no account-specific resolution appears in the public register. The public record does not tell us which authority level the relevant policy assigned to a bad-debt schedule, whether management approved balances within a delegated limit, whether a committee reviewed a cohort, or whether Board consideration was embedded in materials that were not published. The absence of a public schedule is a disclosure gap, not proof that no authorization occurred.

At 31 December 2012, the financial statements recorded the bad-debts expense. That date defines the year-end accounting result, not necessarily the moment when every collection judgment or journal entry was made. The missing items include the policy in force, the dates of the uncollectibility assessments, the individual or cohort approvals, and the movement from allowance to derecognition. The annual report provides the outcome but not the decision trail.

The next public date is 21 June 2013. Draft minutes of the Annual General Members’ Meeting record the Finance Director presenting the 2012 audited accounts and stating that the Board had not ratified the 2012 audited report because it had been received only that Friday. The minutes also describe stricter measures on late payment and resource recovery. This is meaningful evidence that, at the member meeting, aggregate Board ratification had not yet occurred. It is not proof that no management approval existed for the accounting entries, nor is a draft minute a substitute for the underlying Board papers or debtor schedule.

The public resolution register then records Resolution 201309.184, under which the Board resolved to approve the 2012 Audit as presented by Ernst & Young. The code indicates September 2013. This supplies the missing aggregate approval step after the June presentation. It does not disclose a vote split, a proposal and second, a Board paper, a management representation letter, or an attached bad-debt schedule. Most importantly, approval of audited financial statements is not logically identical to per-account authority for every balance included in those statements.

The auditor’s opinion in the annual report is dated 29 October 2013. That produces a visible sequence: no Board ratification as recorded in June, a September-coded Board approval of the audit, and an auditor signature in late October. The sources do not explain the precise mechanics. Board approval may precede final auditor signing; the dates alone do not prove an irregularity, and the chronology should not be forced into one. What can be said is that the public documents allow the aggregate milestones to be ordered while leaving the transaction-level approval chain opaque.

The 2013 register also contains a new Delegation of Authority, identifies signatories and again states that transactions above USD 100,000 required Board approval. Those measures were adopted after the 2012 year end. They cannot be used retroactively to prove who was authorized to approve the 2012 bad-debt expense. Nor does the repeated threshold resolve the category problem: the evidence still does not establish that a write-off was the sort of transaction the threshold governed.

A later accounting-policy statement supplies similar context with a firm temporal boundary. AFRINIC’s 2013 audit report says trade receivables were carried at nominal value less estimated recoverable amounts, impairment was provided where objective evidence showed an inability to collect, and debts were derecognized when assessed as uncollectible. That policy describes a coherent accounting mechanism. But it is published for the later report. It cannot establish the exact 2012 policy wording, the evidence used for each 2012 balance, the dates of assessment or the approving authority.

The chronology therefore proves more than a blank page but less than a full control narrative. Members can see a year-end loss, a June presentation without Board ratification, a later Board resolution approving the audit, and the final auditor date. They cannot see the paper that moved the bad-debt population through management, any committee and the Board. A well-governed public account should connect those layers while acknowledging that different acts may lawfully require different approvals.

What the external audit establishes—and what it expressly leaves open

The independent auditor text is important evidence. It says Ernst & Young audited AFRINIC’s financial statements for the year ended 31 December 2012 under International Financial Reporting Standards and the Mauritius Companies Act 2001. It places responsibility for preparation and fair presentation on the directors, including the internal control needed to prevent material misstatement. It records an unmodified true-and-fair opinion and says proper accounting records had been kept, so far as appeared from the auditors’ examination.

Those conclusions support a strong presumption that the aggregate financial statements were presented in accordance with the stated reporting framework. They are part of the benign explanation for the write-off, not an inconvenient detail to be dismissed. Any responsible analysis has to credit the external opinion and the later Board approval of the audit.

The audit text also states its own boundary. The auditors considered internal control in order to design audit procedures; they did not express an opinion on the effectiveness of internal control. That means the opinion cannot be cited as proof that AFRINIC’s fee-collection process was well designed, that every notice or cure period was fair, that every recoverability assessment was adequately documented, or that each debtor-level action had the correct approval. Financial-statement assurance and a control-effectiveness opinion are different products.

Nor does the true-and-fair opinion publish the composition of the USD 107,994. The reader still lacks the number of accounts, the age of the balances, their concentration by safe anonymized band, the notices issued, the payment plans or disputes considered, and the evidence that made management conclude collection was no longer probable. The opinion may be consistent with all of those records existing privately. It does not make them visible.

This is not a demand for the auditor to disclose protected member files. It is a demand to respect the division of responsibilities. An auditor can provide financial-statement assurance while the institution provides governance transparency. The Board can approve an audit while management explains the receivables movement. One control does not have to impersonate another.

The same principle applies to “proper accounting records.” The phrase reports what appeared from the audit examination. It does not turn every underlying administrative action into a publicly justified decision, and it does not answer how the finance ledger was reconciled to membership closure and resource recovery. Those are narrower questions that an anonymized management disclosure could answer without reopening the audit opinion.

An institution weakens its own accountability when it makes an audit opinion carry more political or operational weight than the opinion claims. The better reading is modest: the opinion materially strengthens confidence in the aggregate accounts, while the explicit control limitation preserves a legitimate space for scrutiny of collection design, approval routes and registry consequences. Both propositions can be true at once.

The unresolved 51: why a missing crosswalk matters

The number 51 is memorable, which is precisely why it must be handled carefully. The annual report says 51 members were closed during the year and notes the number with concern. It says the matter was researched, connects reclaimed resources associated with non-payment to membership-fee growth, and proposes a more systematic collection process. Yet the report does not disclose the promised research or identify how many written-off balances belonged to closed members.

The 51 cannot serve as the denominator for any calculation involving USD 107,994. Dividing the expense by 51 would create an apparent average debt per closed member, but the premise would be unsupported. Some closed members may not have been part of the write-off. Some written-off accounts may not have been among the closures. A single account might contain several invoices, and the number of debtor accounts is itself unknown. An elegant quotient would merely conceal a false assumption.

The gap also runs in the other direction. The annual report’s statement that associated resources had been reclaimed does not prove that every account in the expense population underwent resource recovery, or when any recovery took place. It does not establish that reclamation extinguished any remaining contractual claim. Accounting derecognition, membership status and resource administration must therefore remain separate columns until evidence connects them.

Why does the mapping matter if debtor names properly remain private? First, it would show whether the financial loss was concentrated in a small number of large balances or dispersed across many small ones. Concentration affects the kind of control response that is useful. Second, it would show whether closure was a common, rare or unrelated outcome for the written-off cohort. Third, it would clarify the timing: whether balances were written off before, at or after a membership and registry action. Fourth, it would reveal whether post-write-off recoveries occurred without implying that derecognition had been improper at the time.

None of this requires publishing a debtor identity, resource block, country, company or person. A safe crosswalk can operate in cohorts. It can state the count of accounts in broad value and age bands, show how many had membership status changed, show how many were associated with a completed registry-resource action, and state how many later generated recoveries. Small cohorts can be combined to prevent re-identification. The institutional choice is not total secrecy or exposure; it is careful aggregate design.

The public record’s reference to research on closures makes the omission more noticeable. If the institution investigated why members were closed, a summary of findings could have explained whether collection design, changing membership conditions, accumulated debts or other causes mattered, without naming cases. Because that research is not public in the available record, readers cannot tell what management learned or how the proposed systematic process addressed it.

The missing crosswalk also protects AFRINIC from overstatement. Without it, critics may assume all 51 closures were debt cases and that every resource recovery corresponded to the write-off. Supporters may assume the opposite—that every written-off balance followed a complete sequence of notice, cure, closure and reclamation. Both stories outrun the evidence. A reconciliation would replace competing narratives with bounded facts.

A private technical ledger cannot become a punishment system

The authority question must be stated directly. AFRINIC is a private technical bookkeeper and coordinator. Its records organize contractual membership, financial receivables and number-resource administration. They do not create sovereignty. A write-off does not pronounce guilt. A membership closure does not become a public-law sanction merely because the registry’s operation matters. A resource-register change is not adjudication or confiscation.

This distinction gives the ledger a demanding but narrow role. The ledger must follow reality: invoices must correspond to actual contractual obligations; recoverability estimates must rest on evidence; write-offs must follow the applicable accounting rule; and administrative records must reflect decisions reached under the relevant agreement and delegated authority. Accuracy and traceability are therefore essential. But accuracy is not a licence to convert bookkeeping into enforcement beyond the contract.

Narrow authority also changes how the 2012 loss should be investigated. The question is not whether AFRINIC punished 51 entities; that premise is unsupported and institutionally mistaken. The questions are whether the receivables were genuinely assessed as uncollectible, whether the correct rule was applied, whether the schedule followed a visible authority chain, and whether separate closure and resource processes can be reconciled. These are treasury and coordination questions.

Resource administration can have operational consequences for network operators. That is why precision, notice, a meaningful opportunity to cure, accurate records and reversibility where technically possible matter. The potential consequence does not enlarge AFRINIC into a sovereign authority. It makes restraint more necessary. A private coordinator should act within a documented purpose, make no broader judgment than the record supports, and preserve a route to correct mistakes.

The member-funded bargain adds a second accountability line. Members supply the fee income that supports the registry. When unpaid balances become expense, the loss may reduce reserves or shift pressure toward paying members, although the public sources do not quantify any individual burden. Members are entitled to understand the aggregate movement and the controls governing it. Their interest in that explanation does not override legitimate debtor confidentiality.

Official AFRINIC documents are authoritative for what AFRINIC recorded, reported and resolved. They do not, through self-description alone, prove that an administrative step was necessary, proportionate, fair or completely authorized. Conversely, a missing public document does not prove the step was improper. The disciplined position sits between institutional deference and accusation: credit what the record establishes, identify what it cannot establish, and ask for the least intrusive disclosure capable of closing the gap.

This framework also guards against moralising an accounting loss. An overdue invoice is a contractual receivable, not evidence of misconduct. An impairment reflects an estimate of recoverability. A derecognition decision reflects the accounting treatment of an assessed-uncollectible amount. Each may arise from ordinary financial administration. None should be used to label an unnamed member or to imply intent.

Visible accountability is consequently not theatre. It is the mechanism that keeps a consequential technical administrator within its private role. If the institution can show which rule governed the receivable, which authority approved the treatment, how accounting categories reconciled, and how errors or later recoveries were handled, confidence rests on an inspectable process rather than on claims of institutional status.

The strongest countercase is ordinary prudent accounting

The benign explanation deserves to be presented in its strongest form. Management may have accumulated old unpaid membership balances over several years, taken collection and contractual resource-recovery steps, assessed the remaining receivables as uncollectible, and recognized the loss in the accounts. The annual report disclosed that the amount was the largest then recorded and did not hide the accompanying concern over member closures. Management presented the accounts to members, the Board later approved the audit, and Ernst & Young gave an unmodified opinion.

On this reading, the 2012 entry is a sign of prudence rather than weakness. Carrying receivables that are no longer recoverable would overstate assets. Writing them off allows the balance sheet and annual result to reflect management’s best-supported assessment. The large year-on-year increase could represent a delayed clearing of accumulated debts rather than a sudden collapse in collection. The draft meeting minutes’ reference to debts from the previous three years is compatible with that explanation.

Confidentiality also supplies a reasonable explanation for the lack of names and case details. Membership accounts can contain commercially sensitive information, disputes and payment arrangements. Publicly exposing debtors could be both unnecessary and damaging. An aggregate disclosure may have been judged sufficient in the annual report, especially alongside an external audit.

The policy context published for 2013 is consistent with a conventional progression: carry trade receivables at nominal value less estimated recoverable amounts, recognize impairment where objective evidence indicates inability to collect, and derecognize debt when assessed uncollectible. Although that later statement cannot prove the exact 2012 policy or its application, it shows that the accounting concepts at issue are ordinary rather than inherently suspect.

Nothing in the available evidence defeats this countercase. There is no proof of crime, corruption, fraud, illegality, bad faith or punitive intent. There is no basis to identify a debtor or infer that AFRINIC intended an accounting entry as a sanction. The analysis must remain a disclosure and control inquiry.

But the benign case does not answer the narrower questions. An unmodified opinion is not a control-effectiveness opinion. Board approval of the audit is not a published per-account authorization. Confidentiality does not prevent anonymized ageing and concentration bands. Aggregate disclosure does not reconcile the 51 closures with the write-off population. Responsible accounting and fuller governance transparency are not competing conclusions; the latter could strengthen the former.

The appropriate response is therefore not accusation, reversal or exposure. It is a request for a control-quality record proportionate to the loss: an anonymized receivables movement, a description of the objective evidence and policy version, an approval timeline, and a cohort-level crosswalk among bad debt, membership status and resource administration. If ordinary prudent accounting is the full explanation, that record should make the explanation easier to sustain.

The control record that would settle the public question

A useful disclosure would begin with the accounting movement. It would state the opening gross membership receivables and opening impairment allowance, then show new billings, cash collections, credits, recoveries, new impairment, amounts written off and the closing figures. The categories must be defined so a reader does not confuse an allowance estimate with a derecognition or a closing balance with an annual flow.

Next should come ageing. Balances could be placed in time buckets without names, with the number of accounts and total value in each. Concentration bands could show whether the expense was driven by a few large exposures or many smaller ones. Where a bucket is too small to preserve confidentiality, it could be combined with another. The object is not to reverse-engineer identities but to reveal the risk shape.

The institution should identify the policy version applied at 31 December 2012 and summarize the objective criteria for an inability-to-collect assessment. It could describe the types of evidence considered—without publishing case files—and explain whether disputed balances, active payment plans or incomplete notices were excluded. The public record presently does not answer those questions, so the disclosure should distinguish a documented rule from any later policy statement used only for context.

The authority chain should be presented by level and date. Management approval, any Finance and Audit Committee review, and any Board action should be separated. The schedule should state the thresholds or classifications that sent a cohort to a particular authority. It should not imply that the USD 100,000 external-transaction rule applied unless the governing instrument actually classified a write-off that way. It should also distinguish approval of the debtor schedule from approval of the audited financial statements.

Finally, a confidentiality-preserving crosswalk should place the three ledgers side by side. For the written-off cohort, it could show how many accounts were still active, closed, or otherwise unresolved; how many were associated with a completed resource-register action; and whether the timing was before or after derecognition. A separate view of the 51 closures could show how many, if any, fell within the written-off population. Later recoveries could be disclosed in aggregate, alongside a statement of whether any residual contractual claim remained.

One timeline should reconcile the year-end accounting date, the June 2013 presentation, the September-coded Board resolution and the 29 October auditor signature. The timeline need not allege that one sequence was required. It should simply say what each milestone approved and what it did not.

This control design is deliberately modest. It does not demand names, invoices, resource blocks, countries or dispute records. It asks the institution to demonstrate that the financial ledger, membership register and resource administration can be reconciled internally and described safely in public. Such a record would reduce speculation, make the benign countercase testable, and show that private technical coordination remains evidence-led and bounded.