中文版 · Blockchain learning guide · Previous article
Blockchain learning · Article BC074
A tokenised claim needs more than a correct token balance. Someone must identify the underlying asset, explain the claimant’s entitlement, maintain custody where relevant and connect a completed digital action to the real-world action being promised.
A useful review therefore compares records across that boundary. Does each claim map to the intended asset? Is the asset actually available under the stated conditions? Do the token record and entitlement record name the same eligible party? What evidence marks delivery as complete?
This lesson answers those questions through a fictional warehouse reconciliation. It contains no product recommendation, issuance procedure or instruction to transact. All organisations, crates, claims and programme rules are invented. The exercise does not establish that such an arrangement is legally available in any jurisdiction.
Choose your reading route
In this guide
- 1. The token belongs to a larger arrangement
- 2. Define the parties and the evidence they supply
- 3. Establish the baseline one-to-one mapping
- 4. Case A: an inventory shortfall
- 5. Case B: equal totals conceal a duplicate allocation
- 6. Case C: a valid token transfer leaves the entitlement condition unmet
- 7. Case D: reservation, delivery and closure occur at different observations
- 8. Ask who can resolve each discrepancy
- 9. Write the reconciliation result without overstating it
- 10. Mainland China context: a technical example is not permission to operate
- 11. Exercises with answers
- 12. Frequently asked questions
- 13. Sources and further reading
- Read sections 1–3 for the distinction between a representation and the full claim arrangement.
- Work through sections 4–7 for missing inventory, duplicate allocations, transfer mismatch and delivery timing.
- Use sections 8–10 to identify evidence owners and write a limited reconciliation conclusion.
- Complete the exercises and FAQ to test the arithmetic and the boundaries of what the records establish.
Back to contents · Next section
1. The token belongs to a larger arrangement
Tokenisation can place a digital representation on a programmable platform. The BIS and CPMI report published on 21 October 2024 examines such arrangements and stresses that governance and familiar risks, including custody and operational risks, remain relevant.
For a particular claim, the practical question is which relationships make the representation meaningful. A digital entry may refer to an asset held elsewhere, an obligation of an issuer, or a defined entitlement under an agreement. These are not interchangeable descriptions.
Our exercise uses one specified arrangement: a claim associated with a particular numbered crate of printing paper. It does not treat the token as a universal property title. The supplied programme rules define how its records are used to decide whether a delivery may proceed.
This narrower task differs from listing possible benefits of tokenisation. Faster record updates would be unhelpful if the wrong crate were allocated twice, the warehouse could not locate it or a transfer left the entitlement records inconsistent. The review must test the connections, not just the speed of the ledger.
Back to contents · Next section
2. Define the parties and the evidence they supply
The fictional issuer I maintains the claim specification. Custodian W holds the numbered crates and supplies inventory and delivery evidence. Register E records the programme’s eligible claimant for each claim. The token contract on network C records a token holder. Programme officer V handles unresolved discrepancies under the invented rules.
These roles are deliberately separate. A signature from I can identify I’s statement, but it is not a physical inspection by W. A token observation can identify the recorded holder, but it does not independently explain what E says. A warehouse count can establish a supplied inventory observation without settling which claimant is entitled to which item.
For this paper exercise, the observations are stipulated accurate where stated, and the programme rules are authoritative within the model. We are not testing real signatures, verifying a warehouse or making legal findings.
The model authorises a delivery only when the accepted token state and E identify the same eligible claimant, the specified crate is available, the claim is not already reserved or delivered, and an authorised release is recorded. These are invented conjunctive conditions: all must hold. They are not a claim that every real tokenisation arrangement uses two registers.
Naming the rules before examining discrepancies prevents us from inventing a convenient interpretation after seeing the result.
Back to contents · Next section
3. Establish the baseline one-to-one mapping
At reconciliation time T0, W’s confirmed available inventory is the set of four distinct crates A, B, C and D. Each is a whole numbered item in this exercise; there is no fractional claim or pooling rule.
The four open claims are:
- R1 maps to crate A; both the token record and E identify H.
- R2 maps to crate B; both records identify J.
- R3 maps to crate C; both records identify K.
- R4 maps to crate D; both records identify L.
Each named claimant is stipulated eligible. No claim is reserved or delivered. The relevant record observations are aligned to the same reconciliation cut, with their chain and register references supplied as consistent. No release has yet been requested or authorised.
The inventory count is 4 and the open-claim count is 4. More importantly, the mapping covers four distinct available crates exactly once. Matching totals are necessary for this particular baseline, but the identifiers are what establish its one-to-one allocation.
The baseline supports a limited conclusion: the supplied inventory and claim allocation agree at T0, and the holder fields agree. It does not show that every future request will complete, or that the custodian has no other obligations outside the stated exercise. Those wider claims are not supplied.
Keep that distinction in mind as we change one fact at a time.
Back to contents · Next section
4. Case A: an inventory shortfall
In this independent case, the four claim records remain as above, but a current confirmed inventory observation lists only A, B and C. The exercise states that D is unavailable for the promised delivery. We are not inferring absence merely from an incomplete search result.
There are now four open claims against three available specified crates. The count shortfall is one. R4’s mapping identifies the affected item directly: its promised crate D is unavailable under the supplied facts.
The other three mappings remain intact within this case. It would be inaccurate to report that every claim has lost its underlying item. It would also be inaccurate to report full reconciliation merely because every token record is present and valid.
What should the worksheet say? “R4 has a confirmed availability discrepancy at this cut; R1–R3 retain their supplied item matches. The reason for D’s unavailability and the programme’s resolution are not supplied.” This identifies the evidence gap without inventing theft, insolvency or a recovery outcome.
A new token entry would not manufacture crate D. A fresher copy of the unchanged claim register would not resolve the physical shortfall. The relevant next evidence belongs to custody, the claim specification and the designated resolution process.
Back to contents · Next section
5. Case B: equal totals conceal a duplicate allocation
Return to the baseline inventory of A, B, C and D. In this independent case, the claim mapping changes: R1 maps to A, R2 maps to B, and both R3 and R4 map to C. No claim maps to D.
The counts still look reassuring: four available crates and four open claims. But there are only three distinct crates referenced by those claims. Crate C appears twice; D appears zero times.
Under the model’s one-whole-crate-per-claim rule, this is an allocation discrepancy. The same specified crate cannot satisfy both independent whole-item delivery claims as if it were two different crates. That conclusion follows from the supplied rule, not from a universal claim that all shared or fractional arrangements are invalid.
The unused D does not automatically cure the problem. R4 names C, and no substitution rule has been supplied. Treating D as an automatic replacement would silently rewrite the claim specification.
A proper reconciliation therefore checks both totals and the relationship between identifiers. Four equals four is a quantity check. A one-to-one match is a stronger structural check. The latter fails here even though the former passes.
Record the specific discrepancy, the affected claims and the missing authority for any proposed correction. Do not decide privately which claimant should receive C or whether someone must accept D; the exercise assigns unresolved disputes to V and supplies no outcome.
Back to contents · Next section
6. Case C: a valid token transfer leaves the entitlement condition unmet
Return again to the original baseline. A supplied technically valid token transfer changes R2’s token holder from J to M. Register E still identifies J. M is stipulated eligible, but the register has not changed.
The chain record can be correct about the transfer while the overall delivery condition remains unsatisfied. Our programme requires both records to identify the same eligible claimant. They now disagree.
Under that precise rule, neither J nor M currently satisfies the combined holder condition for delivery of B. J matches E but not the token record; M matches the token record but not E. This does not determine ultimate legal rights. It is the answer to the fictional programme’s stated processing condition.
It would be wrong to call the chain transfer invalid merely because a separate register is inconsistent. It would also be wrong to treat technical transfer success as automatic completion of every off-chain entitlement update.
The relevant question is what evidence and authorised process can reconcile the two records. The toy does not supply automatic synchronisation or permission for a reader to edit E. Until the discrepancy is resolved under the programme’s rules, the combined delivery condition remains unmet.
This case is a reminder to name the completion boundary. “The token holder changed” is a narrower claim than “all conditions for warehouse delivery now identify the new party.”
Back to contents · Next section
7. Case D: reservation, delivery and closure occur at different observations
Start once more from the baseline. Consider a supplied R1 redemption history with three later observation cuts. Here redemption means fulfilling this particular claim by delivering crate A, not exchanging a token for money.
At T1, H satisfies the holder and eligibility checks, an authorised release is recorded and R1 enters a reserved state. The reserved claim cannot be used for a second concurrent release under the model. Crate A remains physically at W, set aside for that release; no physical delivery acknowledgement has yet been supplied.
At T2, reliable evidence confirms that W delivered A to the authorised recipient for R1. The programme’s final closure record has not yet been observed. The model marks delivery as performed and prevents a second delivery against the same claim, while closure remains pending.
At T3, the reconciled programme records mark R1 closed and no longer outstanding. The remaining open claims R2–R4 correspond to B, C and D. There are three open claims and three available unallocated-to-completed-delivery crates in this completed state.
These cuts are a supplied history, not a deployment recipe. The model does not pretend that a physical handover and every digital record update occur atomically. At T1, a reservation is not proof of delivery. At T2, delivery evidence is not proof that the final closure record has already appeared.
A careful report distinguishes pending work from inconsistency. The intermediate states are permitted by the supplied lifecycle, with their stated protections. A real system would need evidence that its own rules cover interrupted handoffs and prevent conflicting releases; our fictional trace does not validate any implementation.
Back to contents · Next section
8. Ask who can resolve each discrepancy
The four cases point to different evidence owners. W can supply custody observations and delivery acknowledgements. I can explain the intended claim specification. E’s authorised process can establish the eligible-claimant record. Chain observers can establish the relevant token state. V receives unresolved disputes within the programme.
Those roles should not be collapsed into whichever party runs the most convenient dashboard. A dashboard can summarise evidence, but it cannot create the authority missing from a proposed change.
For Case A, the unresolved issue concerns D’s availability and the consequences under the claim arrangement. For Case B, the issue is the duplicated allocation and any authorised correction. For Case C, it is a cross-record mismatch. For Case D, the task is to identify which lifecycle stage the current evidence establishes.
A signed statement can improve attribution of a report without making the reported physical fact true. Similarly, an automated update can be executed exactly as programmed while relying on a mistaken input. The worksheet should identify the evidence source, its scope and the observation time rather than treating automation as a substitute for those details.
Good escalation is specific: name the failed condition and the record needed to address it. “Blockchain problem” is too broad to tell the custodian, issuer or register operator what actually needs examination.
Back to contents · Next section
9. Write the reconciliation result without overstating it
A useful report has four short parts: the cut being examined, the records compared, the conditions that passed, and the discrepancies or unresolved fields.
For the baseline, it might state: “At T0, the supplied four open claims map one-to-one to four confirmed available numbered crates. Token and entitlement-holder records agree for each claim. No delivery has yet been authorised. This is an allocation and record-consistency result under the stated programme assumptions.”
For Case B, it should instead identify R3/R4’s duplicate reference to C and the unused D. For Case C, it should name J and M’s mismatched roles rather than reporting a generic transfer failure. For Case D at T2, it should say delivery is evidenced while final closure remains pending.
Do not extend a limited reconciliation into an audit of every obligation, a guarantee of future performance or a legal opinion. A snapshot can be accurate within its scope and still omit other claims, liens, agreements or later events. Our baseline excludes such unknowns by stipulation for teaching; a real report would need to state its actual coverage.
The value of the exercise is not a green badge. It is a traceable explanation of which connection works and which requires further evidence.
Back to contents · Next section
10. Mainland China context: a technical example is not permission to operate
As checked on 12 October 2026, the PBOC-led notice 银发〔2026〕42号 of 6 February 2026 addresses virtual-currency and real-world-asset tokenisation activities. It prohibits mainland RWA tokenisation activities and related services as specified, with an exception for relevant business authorised by competent departments and conducted through specified financial infrastructure. It also replaced the 2021 notice.
This brief context is not a legal clearance or an interpretation that our fictional warehouse arrangement qualifies for an exception. The lesson supplies no issuance, marketing, investment or operating instructions. Readers assessing an actual activity need the current applicable rules and appropriate professional guidance.
Cross-border examples do not erase that boundary. A project discussed in another jurisdiction does not establish availability or permission for a reader in mainland China. The classroom model remains a way to understand evidence and institutional responsibilities.
Back to contents · Next section
11. Exercises with answers
Exercise 1: Count and identify the shortfall
Case A has four open claims and available crates A, B and C. What is the shortfall, and which claim’s specified item is affected?
Answer: One available specified crate is missing relative to the four claims. R4 names D, which the case explicitly states is unavailable. Do not infer that A, B and C are also missing.
Exercise 2: Why does four equal four fail?
Case B has four crates and four claims. What additional check reveals the problem?
Answer: Count references by crate identifier. A and B each appear once, C twice and D zero times. The intended one-to-one allocation fails despite matching totals.
Exercise 3: Can D be substituted automatically?
Can the reviewer resolve Case B by assigning unused D to R4?
Answer: No substitution authority or rule is supplied. The reviewer can identify the possible question but cannot rewrite the claim terms. V’s process must address the unresolved discrepancy under the invented programme.
Exercise 4: Who passes the combined holder condition?
In Case C, the token names M and E names J. Who satisfies the stated two-record condition?
Answer: Neither currently does. Each matches only one record. This is a processing-condition result, not a determination of their real legal rights.
Exercise 5: Which event establishes physical delivery?
In Case D, is reservation at T1 enough to report delivery? Is the closure record already observed at T2?
Answer: No to both. Physical delivery is supplied at T2; final closure is observed at T3. The report should retain those separate milestones.
Back to contents · Next section
12. Frequently asked questions
Does an accurate ledger prove the warehouse inventory?
It can accurately record a statement about inventory without independently inspecting it. The evidence connecting the statement to the physical items must be identified separately.
Does tokenisation always require two claimant registers?
No. That is this exercise’s explicit rule. Other arrangements may define authoritative records differently. The lesson is to discover the actual rule rather than assume every transfer completes every related update.
Does a discrepancy prove fraud?
Not by itself. It identifies a mismatch or unmet condition within the evidence. Cause, responsibility and remedies require further information and the applicable process.
Is the baseline a guarantee that every claim will be fulfilled?
No. It is a scoped reconciliation at one cut under supplied assumptions. Future availability, operational performance and legal enforceability are separate questions.
Back to contents · Next section
13. Sources and further reading
Sources checked on 12 October 2026. The warehouse programme and all numerical cases are original fictional teaching examples.
- BIS–CPMI, Tokenisation in the context of money and other assets: published 21 October 2024; concepts, governance and risk context.
- PBOC-led notice 银发〔2026〕42号: 6 February 2026; the dated mainland context described above.
- Digital Assets, Stablecoins and Tokenised Finance: broader Singapore background, distinct from this reconciliation workshop.
A tokenised claim is understandable when the record, asset, entitlement, responsible parties and completion evidence can be connected explicitly. Matching a token count is only the beginning of that explanation.