Term Bridge

60 pairs

• Bank deposit

  →StablecoinFalse friend

  A deposit is a claim on a licensed bank, covered by a deposit guarantee scheme up to a statutory limit and embedded in a regulated payment system with defined recourse. A stablecoin is a claim on an issuer against a reserve; a MiCA e-money token carries a redemption-at-par right but no deposit guarantee. In an issuer insolvency the holder depends on how the reserve is legally segregated, not on a guarantee fund.

  CaveatTreasury policies that classify stablecoins as cash equivalents must state whose credit is being taken. It is issuer and reserve-custodian credit, not deposit protection.

• Bankruptcy remoteness

  →Trustless / non-custodial designFalse friend

  Bankruptcy remoteness is engineered legally: a restricted-purpose SPV, independent directors, non-petition and limited-recourse covenants, true-sale opinions. Non-custodial code means an operator cannot move your assets, which is a different and narrower protection. If the issuer or asset originator fails, a non-custodial token over that exposure still leaves the holder as an unsecured creditor.

  CaveatNon-custodial protects against operator theft, not against issuer insolvency. For real-world assets, the legal wrapper determines recovery, not the custody model.

• Bearer instrument

  →Self-custodied tokenFalse friend

  Both give effective control to whoever holds the thing, and both transfer by delivery. But a lost or stolen certificate has legal remedies: judicial cancellation, good-faith acquisition rules, orders against the issuer to reissue. A lost private key has none, and a transfer executed by a thief is technically valid and irreversible.

  CaveatPossession equals title is only half the analogy. The recovery machinery that makes bearer instruments tolerable in practice does not exist for private keys.

• Beneficial owner

  →Wallet owner / address controllerFalse friend

  Beneficial ownership in AML law identifies the natural person who ultimately owns or controls a customer, established through documentation and register extracts. Control of a private key is neither necessary nor sufficient for that: keys can be custodied, shared, or controlled by someone other than the recorded holder. On-chain data evidences control, not ownership.

  CaveatRecording an address as the beneficial owner satisfies no AML obligation. Identification runs to the person, and control of the key has to be evidenced separately.

• Best execution (MiFID II)

  →Aggregator routing / solver auctionFalse friend

  Best execution is a duty a firm owes its client, with an execution policy, monitoring and evidence obligations. A DEX aggregator or solver optimises a quote at a point in time and owes the user nothing. Neither transaction cost analysis nor a venue selection policy exists by default.

  CaveatAn aggregator best quote is not a best-execution record. Regulated firms trading on-chain still have to build the policy, the monitoring and the audit trail themselves.

• Cap table / shareholder register

  →Token holder registryFalse friend

  Both purport to answer who owns the company. For a German or Austrian GmbH the authoritative record is the shareholder list filed with the commercial register, and share transfers require notarial form, so GmbH shares cannot be tokenized. Tokenized equity generally works only where a statute permits it, for example electronic shares under the eWpG as amended by the Zukunftsfinanzierungsgesetz, Swiss ledger-based securities, or Luxembourg and Liechtenstein structures.

  CaveatFor most European private companies, cap table on-chain means a token tracking an economic interest in an SPV, not the shares themselves. Legal ownership of the shares stays with the SPV.

• Central securities depository (CSD)

  →The blockchain as ledger of recordFalse friend

  A CSD performs a notary function, sits at the top of the holding chain and its book entries have statutory effect. A blockchain is a database whose entries only have that effect where legislation grants it, such as the eWpG, Swiss ledger-based securities or the EU DLT Pilot Regime. Chains also lack the CSD apparatus for corporate actions, tax vouchers and regulatory reporting.

  CaveatWe do not need a CSD because we have the chain is a legal claim, not a technical one, and it is wrong in most jurisdictions absent an enabling statute.

• Circuit breaker / trading halt

  →Pause function / emergency shutdownFalse friend

  An exchange circuit breaker is a rule-based, pre-announced, market-wide halt applied by a neutral venue operator under published parameters. A pause function is a privilege held by whoever controls the admin key, exercisable at discretion and usually without notice. It protects the protocol against exploits, but it also means the asset is neither censorship-resistant nor operator-independent.

  CaveatA pause or freeze function is an exposure to the key holder and should be assessed as counterparty risk, not filed under investor protection.

• Clearing house / central counterparty (CCP)

  →Smart contract as clearing layerFalse friend

  A CCP interposes itself between buyer and seller by novation and absorbs counterparty risk, backed by initial and variation margin, a default fund and a recovery waterfall. A smart contract holds collateral and executes rules; it takes no risk and has no capital. Where collateral proves insufficient, a CCP mutualises the loss under defined rules, whereas a protocol either socialises it onto users or simply accrues bad debt.

  CaveatThe smart contract is the clearing house ignores that a CCP core product is its balance sheet and default waterfall, which no protocol has.

• Credit rating

  →Protocol risk score / smart-contract auditFalse friend

  A credit rating is an opinion on probability of default issued by a regulated agency under a published methodology with a performance record measurable across cycles. A smart-contract audit assesses code against known vulnerability classes at one moment in time, and risk scores are vendor heuristics with no default statistics behind them. Neither addresses a borrower ability to pay.

  CaveatAn audited protocol can still lose all deposits through an economic exploit, a governance attack or an oracle failure, none of which is usually in the audit scope.

• Deposit guarantee / investor compensation scheme

  →DeFi cover protocol / mutualFalse friend

  Statutory schemes are funded by mandatory levies, backed by law and pay out on a defined trigger regardless of the operator willingness. Cover protocols are discretionary or parametric mutuals with limited capital, their own claims assessment, and frequently a token-holder vote on payout. Capacity is usually a small fraction of the sums nominally covered.

  CaveatCover capacity is correlated with the risk it insures. A systemic event is precisely the moment the mutual cannot pay, and the claims decision may rest with the same token holders who suffered the loss.

• Depositary (UCITS / AIFMD)

  →Custodial wallet providerFalse friend

  A UCITS or AIFMD depositary safekeeps assets, monitors the manager, oversees cash flows, and is liable to restore financial instruments lost in custody largely irrespective of fault. A crypto custody agreement typically limits liability to negligence and excludes protocol, chain and third-party events. The oversight duty has no crypto analogue at all.

  CaveatDepositary liability is among the strictest standards in fund law. Assuming a wallet provider carries an equivalent restitution duty is a material misreading of the custody agreement.

• Dividend

  →Distribution / protocol yield / rebaseFalse friend

  A dividend is a distribution of profit resolved by the company organs, ranking behind creditors, with a defined tax and withholding treatment. DeFi yield is usually protocol token emissions, borrower interest or trading fees; there is no profit, no resolution and no issuer owing you anything. Rebasing changes the balance rather than paying out, which produces different accounting and tax questions again.

  CaveatBooking protocol yield as dividend income misstates both the tax position and the risk. The cash flow can stop entirely without any insolvency event or corporate resolution.

• Financial statement audit

  →Proof of reservesFalse friend

  An audit is an opinion on financial statements as a whole under recognised standards, covering assets and liabilities, going concern and the control environment. A proof of reserves demonstrates control of assets at a point in time and normally says nothing verifiable about liabilities, encumbrances or borrowed assets. Merkle-tree liability proofs help but still depend on the operator submitting a complete liability set.

  CaveatA reserve attestation cannot detect the failure mode it is bought to detect: assets borrowed for the snapshot date, or liabilities omitted from the tree.

• Legal Entity Identifier (LEI)

  →Wallet addressFalse friend

  An LEI identifies a legal entity, is validated against company registers and must be renewed annually. A wallet address identifies a key pair and asserts nothing verified about who controls it. Address attribution is inference produced by analytics vendors, not registered identity.

  CaveatTreating an address as an entity identifier fails as soon as keys are shared, rotated, or held in an omnibus arrangement by a custodian.

• Money market fund

  →Yield-bearing stablecoinFalse friend

  A money market fund is a regulated fund with portfolio composition rules, liquidity buckets, stress testing and a defined redemption mechanism, and the investor holds units bearing NAV risk. A yield-bearing stablecoin is either a fund unit presented as a payment token or an issuer promise backed by a portfolio outside fund regulation. Under MiCA, issuers of e-money tokens and asset-referenced tokens may not grant interest, so a euro or dollar token distributing yield in the EU is generally something other than an EMT.

  CaveatIf it pays yield and calls itself a stablecoin, determine first what it legally is. Classification drives eligibility, accounting treatment and whether a regulated firm may hold it at all.

• Net asset value (NAV)

  →Oracle price / on-chain price feedFalse friend

  NAV is a valuation of fund assets less liabilities, produced by an administrator under a valuation policy with a defined pricing source hierarchy, usually daily. An oracle reports an observed market price of a token. For a tokenized fund the two can diverge substantially, and the secondary market token price is not the NAV.

  CaveatUsing the on-chain market price as NAV for accounting or collateral purposes marks the fund to its secondary market discount, which is a different number with different consequences.

• Netting (close-out netting)

  →Batching / rollup compressionFalse friend

  Close-out netting is a legal construct that survives insolvency because statute and framework agreements such as ISDA or GMRA make it enforceable, reducing gross exposures to a single net claim. Batching transactions into a rollup reduces cost and message volume, not legal exposure. Atomic settlement in fact removes netting entirely.

  CaveatTreasury models built on netted funding break when moving to atomic on-chain settlement. Intraday liquidity requirements can rise by an order of magnitude.

• Novation

  →Contract migration / token upgradeFalse friend

  Novation replaces one contract with another by agreement of all parties, extinguishing the original obligations. Migrating users to a new contract version moves balances but extinguishes nothing legally. Rights against an issuer, and third-party consents such as those attached to security interests, do not transfer by redeployment.

  CaveatA protocol upgrade cannot novate an off-chain obligation. The underlying claim stays with whoever the original documentation names as obligor and obligee.

• Order book

  →Liquidity poolFalse friend

  An order book aggregates expressed intentions at specific price levels; depth is real, visible and can be pulled. A pool contains no orders at all: the price is a function of the reserve ratio and is imposed after the fact by arbitrageurs. Depth charts derived from a pool describe a curve, not demand.

  CaveatPool TVL is not depth. A large pool can still offer thin effective liquidity at the size you need, and the displayed price is stale until someone arbitrages it.

• Power of attorney / trading mandate

  →Token approval (allowance)False friend

  Both let someone else act on your assets. A mandate is bounded by its wording, revocable, and the principal has recourse against the agent for misuse. A token approval is a standing on-chain permission, commonly unlimited in amount and duration, that lets a contract move your entire balance, with no recourse if that contract is later compromised or maliciously upgraded.

  CaveatUnlimited approvals are the single most common cause of wallet drains, and they persist for years after the user has stopped using the application.

• Prospectus

  →Whitepaper / MiCA crypto-asset white paperFalse friend

  Both are the issuer disclosure document. A prospectus is approved by a competent authority and carries a statutory liability regime. A MiCA white paper is generally only notified, not approved, except that for asset-referenced tokens it is reviewed as part of issuer authorisation; content standards and liability differ. If the token qualifies as a financial instrument under MiFID II, MiCA does not apply and a full prospectus is required.

  CaveatA MiCA-compliant white paper does not exempt a security token from prospectus obligations. The classification question comes first and determines which regime applies.

• Regulated market / MTF

  →Decentralised exchange (DEX)False friend

  Both match buyers and sellers. A regulated market or MTF has an authorised operator with admission rules, market surveillance, transparency and reporting duties. A DEX is a contract with no party claiming that role, and where a security token trades on it the venue duties do not disappear merely because nobody performs them.

  Caveatverify: whether a permissionless DEX may lawfully host a security token is unresolved in the EU and the US. Do not assume a MiCA CASP authorisation covers trading in financial instruments.

• Repo (repurchase agreement)

  →Overcollateralised lending protocol positionFalse friend

  A repo is a sale with an agreed repurchase: title passes, the buyer may re-use the security, and close-out netting under a GMRA governs default. Depositing collateral into a lending protocol is closer to a pledge with an automated realisation right; there is no master agreement, no negotiated counterparty and no contractual close-out. The insolvency treatment of protocol-held collateral is untested in most jurisdictions.

  CaveatCalling protocol borrowing repo imports assumptions about title transfer, re-use rights and netting enforceability that no protocol currently supports.

• Securities lending

  →StakingFalse friend

  In securities lending a borrower takes title, posts collateral and pays a fee, while the lender keeps economic exposure and assumes borrower credit risk. Staking has no borrower at all: tokens are bonded to secure a network and rewards are protocol issuance plus transaction fees. The risks are slashing, validator failure and unbonding delay, not counterparty default.

  CaveatStaking yield is not a lending fee and largely reflects dilution paid by all holders, so the nominal APY overstates the real return relative to total supply.

• Settlement finality (Settlement Finality Directive 98/26/EC)

  →Block finality / probabilistic finalityFalse friend

  Legal finality means a transfer cannot be unwound, including in the insolvency of a participant, because a designated system and its rules say so. Chain finality is a statement about the probability of a reorganisation, not about legal effect. A confirmed block does not by itself make a transfer irreversible against an insolvency administrator or a court.

  CaveatTreating N confirmations as legal finality is the error. Unless the system is designated, or a statute such as the eWpG, the Swiss DLT Act or the DLT Pilot Regime attaches the effect, restitution remains legally possible.

• Shareholder meeting / proxy voting

  →On-chain governance voteFalse friend

  Shareholder voting is a statutory process with convocation requirements, agenda rules, minority protections, record dates and the ability to challenge resolutions in court. Token governance is plutocratic by construction, has no minority protection and no route to set aside a resolution. Delegation to a small number of large holders is the norm rather than the exception.

  CaveatGovernance token votes have no company law effect. A vote cannot bind an issuer or an SPV unless the constitutional documents expressly make it so.

• Stop-loss order

  →Liquidation thresholdFalse friend

  Both are supposed to cap losses at a level. A stop-loss is the holder instruction to their broker: it can be amended, cancelled, or fail to fill. A liquidation is the protocol right against the borrower, executed by third-party liquidators who are paid a bonus out of the borrower collateral, and it can be neither negotiated nor deferred.

  CaveatLiquidation is not a risk control you own; it is a right the protocol has against you, and the liquidation penalty is a real cost on top of the market loss.

• Trustee

  →Multisig / DAOFalse friend

  A trustee holds property for beneficiaries under fiduciary duties that are enforceable in court, with personal liability for breach. A multisig is a signing policy and a DAO is usually an unincorporated association of token holders. Neither owes fiduciary duties by default, and in several jurisdictions DAO participants risk being treated as partners with unlimited liability.

  CaveatIf a structure needs a fiduciary, it needs a legal person bound by fiduciary duty. A threshold signature scheme is an operational control, not a trust.

• Assignment of a claim

  →Token transferPartial

  Where a token represents a claim, transferring the token is intended to assign the claim. Many jurisdictions require form, notice to the debtor or a register entry for an assignment to be effective, and contractual anti-assignment clauses can block it. Statutes such as the German eWpG, the Swiss DLT Act and the Liechtenstein TVTG were written precisely to give the register entry that legal effect.

  CaveatWithout an enabling statute or a carefully drafted issuance, the token moves but the legal claim may not follow it, leaving holder and creditor as different persons.

• Bid-ask spread

  →Pool fee tier plus price impact plus MEVPartial

  Both measure the cost of immediacy. On-chain that cost decomposes into an explicit pool fee, price impact along the curve, gas, and value extracted by searchers and block builders. It is measurable per trade but not directly comparable to a quoted spread.

  CaveatBenchmarking execution against the pool mid price flatters the result, because the realised cost includes extraction that never appears in any quote.

• Broker / investment firm

  →Wallet plus DEX front endPartial

  Both are how an end investor reaches the market. A broker holds authorisations, owes appropriateness and best execution duties, and is answerable for errors and mis-execution. A front end is software: the user faces the protocol directly and carries execution, approval and key risk personally.

  CaveatA firm offering a front end to clients may still be performing a regulated service. The absence of custody does not by itself remove reception and transmission of orders from the analysis.

• Central bank money as settlement asset

  →Wholesale CBDC / trigger solutionPartial

  Both aim to settle the cash leg in the safest available asset. A wholesale CBDC would place central bank money directly on a distributed ledger; trigger or bridge solutions keep the cash in the RTGS system and use the chain only to initiate payment, which is what several European experiments actually implemented. The distinction determines whether settlement is genuinely atomic.

  CaveatTrigger solutions are not atomic DvP. A short window exists in which one leg has settled and the other has not, and legal finality sits with the RTGS system rather than the chain.

• Client asset segregation

  →Segregated versus omnibus walletsPartial

  Both distinguish assets held for clients from the firm own assets. On-chain segregation is directly verifiable: one address per client is provably separate, which is stronger evidence than an internal ledger entry. Omnibus wallets recreate the traditional commingling problem, and MPC or shared-key arrangements can make legal control ambiguous even when addresses look separate.

  CaveatAn address only the custodian can sign for is not client-segregated in a meaningful sense merely because it holds one client balance; control, not labelling, drives the insolvency outcome.

• Collateral and margin

  →OvercollateralisationPartial

  Both address counterparty credit risk by requiring assets against an exposure. On-chain lending typically demands 120 to 200 percent collateral up front rather than variation margin against a mark, and enforcement is continuous by third-party liquidators rather than by a margin call. There is no dialogue, no cure period and no relationship manager.

  CaveatThere is no margin call. The position is liquidated the moment the oracle reports a breach, including during an oracle outage or on a stale price.

• Corporate actions

  →Programmed distributions / contract upgradesPartial

  Distributions, splits and redemptions can be executed by a contract when the token is the register. What does not automate is the decision itself, record and ex-date conventions, withholding tax mechanics and notification duties. Optional and discretionary actions such as rights issues or tender offers still require off-chain elections.

  CaveatCorporate actions on tokens held via wrappers, bridges or lending pools frequently never reach the economic owner, because the contract sees the pool address, not the beneficiary.

• Correspondent banking (nostro / vostro)

  →Stablecoin payment railsPartial

  Both move value across borders. Stablecoin transfers remove the chain of intermediaries and the pre-funding of nostro accounts and settle continuously. They replace bank credit risk with issuer, reserve and chain risk, and the fiat on-ramp and off-ramp at each end are still banks with cut-off times and compliance queues.

  CaveatThe end-to-end payment is only as fast as the slowest ramp. An instant on-chain leg does not make the payment instant for the beneficiary.

• Custodian

  →Qualified custodian for keys plus smart-contract escrowPartial

  Both hold assets for the account of a client. On-chain, the object of custody shifts from a book-entry securities position to cryptographic key material and the ability to sign. Where the client self-custodies on a public chain, there may be no custodian in the legal sense at all.

  CaveatCustody of keys is not custody of the asset. MPC key sharding across a provider can leave it genuinely unclear who exercises legal control and into whose insolvency estate the position would fall.

• Custody statement

  →Block explorer view / wallet balancePartial

  Both show holdings. On-chain data is continuously verifiable by anyone and cannot be silently altered, which is evidentially stronger than a PDF from a custodian. It also carries no cost basis, no accruals, no counterparty names, and no legal status as a confirmation of anyone obligation towards you.

  CaveatAuditors and tax authorities want a statement from a responsible party. An explorer screenshot documents chain state, not your entitlement or the custodian acknowledgement of it.

• Escrow

  →Smart-contract escrowPartial

  Both hold value until conditions are met. A contract executes objectively verifiable conditions with no discretion and removes the risk that the escrow agent itself misappropriates the funds. It cannot interpret ambiguous conditions, weigh evidence or resolve a dispute, and it substitutes code and oracle risk for agent risk.

  CaveatEscrow disputes are usually about facts, not about payment mechanics, and a contract has no mechanism for determining facts.

• Fixed interest rate / committed facility

  →Protocol APY / utilisation curvePartial

  Both express the price of borrowing. Protocol rates are usually a variable function of pool utilisation, recomputed every block, and can move by hundreds of basis points within a day with no policy decision and no notice. Fixed-rate on-chain lending exists but forms a smaller, separately structured market.

  CaveatA quoted APY is a snapshot of an algorithm, not a rate you have contracted for. It is unsuitable as a funding assumption without a fixed-term wrapper, and there is no committed facility.

• Force majeure / market disruption event

  →Chain halt, reorganisation or fee spikePartial

  Both are events in which the market mechanism stops working as documented. Chain-level events break the assumption that a transaction can be submitted at all, which affects liquidations, redemptions and collateral transfers exactly when they matter most. Standard financial documentation does not define them.

  CaveatDisruption and fallback clauses in tokenized instrument documentation should name chain halts, reorganisations, fee spikes and oracle outages explicitly; a generic force majeure clause will be litigated.

• Front-running / market abuse

  →Maximal extractable value (MEV)Partial

  Both describe extracting value by acting on knowledge of an incoming order. In regulated markets this is prohibited conduct by a firm that owes duties to a client. MEV is an emergent property of public mempools and block construction, carried out by parties owing duties to no one, and the MiCA market abuse provisions do not map cleanly onto searcher and builder behaviour.

  Caveatverify: whether specific MEV strategies constitute market abuse under MiCA is unsettled. Do not assume the practice is either clearly permitted or clearly prohibited.

• Fund administrator

  →Oracle plus accounting smart contractPartial

  Some administrator functions can be executed by contracts: unit registry, fee accruals, subscription and redemption processing. Valuation, independent pricing, financial statement preparation and the control environment cannot. In regulated fund structures the administrator remains a contractual and often licensed party.

  CaveatAutomating the mechanics does not remove the need for an independent valuation function. An oracle fed by the manager is not independent pricing.

• Fungibility of securities

  →Fungible token standardPartial

  Units of a fungible token are technically interchangeable, mirroring the civil law concept of fungible goods. In practice, units and addresses can be tainted by sanctions listings, issuer freeze functions or venue policies, so acceptance becomes selective. Traditional markets solve this at account level; on-chain the transaction history travels with the asset.

  CaveatA token can be technically fungible and commercially non-fungible at the same time. Provenance risk attaches to the units you received, not only to counterparties you chose.

• ISIN / WKN

  →Contract address plus token IDPartial

  Both identify the instrument. An ISIN is assigned under ISO 6166 by a national numbering agency and is the key for reporting, settlement and reference data; a contract address identifies one deployment on one chain. Tokenized securities normally carry both, and regulatory reporting still requires the ISIN.

  CaveatOne instrument can have several contract addresses across chains and bridged wrappers. The address identifies a representation, not the security.

• KYC / AML onboarding

  →Wallet whitelisting / allowlistPartial

  Whitelisting is the on-chain enforcement layer for an off-chain KYC decision: only verified addresses can receive or hold the token. It does not replace customer due diligence, ongoing monitoring, PEP and sanctions screening or the institution risk assessment. Permissionless tokens have no such layer at all.

  CaveatA whitelist proves an address was approved once. It does not prove the person behind it is still eligible, still exists, or still controls the key.

• Market maker

  →Automated market maker (AMM)Partial

  Both provide continuous two-sided liquidity. A designated market maker quotes under an obligation, manages inventory actively, and can widen or withdraw; an AMM follows a fixed curve and cannot decline a trade. AMM liquidity providers are systematically adversely selected by arbitrageurs, measured as loss-versus-rebalancing.

  CaveatAMM liquidity does not withdraw in a crisis because it cannot. The pool keeps selling into the move and absorbs exactly the toxic flow a human market maker would have avoided.

• Reconciliation

  →Shared ledger / single source of truthPartial

  A shared ledger genuinely removes the need for two parties to reconcile their copies of the same position. It does not remove reconciliation between chain state and internal books, custodian statements, the fiat leg, or between an asset and its wrapped representations. New reconciliation surfaces appear at every bridge and every off-chain event.

  CaveatFirms that budget zero reconciliation effort after tokenizing find it has moved rather than vanished, into chain-to-ERP, bridge-to-base-asset and token-to-register checks.

• Rehypothecation

  →Looping / recursive collateral use / restakingPartial

  Both describe re-use of the same collateral to support several exposures, lengthening the credit chain. On-chain the re-use is visible and quantifiable, which is an improvement over opaque collateral chains. But there is no regulatory cap and no consolidated view of who bears the terminal risk when the base asset moves.

  CaveatTransparency does not shorten the chain. Leverage built from looped positions unwinds simultaneously because every participant shares the same liquidation logic and the same oracle.

• Sanctions screening / name matching

  →Know Your Transaction (KYT) / chain analyticsPartial

  Both screen counterparties and flows against risk. KYT adds address-level provenance and cluster attribution, which has no traditional analogue. But attribution is probabilistic, vendor-specific and not reproducible in the way a list match against a legal name is.

  CaveatTwo vendors will score the same address differently. Treating a KYT risk score as a sanctions hit, or its absence as clearance, is not a defensible control on its own.

• Settlement (T+2)

  →Atomic settlement / instant on-chain settlementPartial

  Both describe the moment the asset leg and the cash leg are exchanged. On-chain, the two legs can be made conditional on each other inside a single transaction, which removes the settlement window T+2 creates. What does not carry over is the machinery T+2 pays for: multilateral netting, a settlement-fail and buy-in regime, and a statutorily defined moment of finality.

  CaveatFaster is not automatically better. Atomic settlement eliminates netting, so gross funding requirements rise sharply and both legs must be pre-funded before the trade can settle at all.

• Share class / unit class

  →Token class / partition (ERC-1400)Partial

  Partitions and multi-class token designs replicate the idea of differentiated rights within one instrument, and fee levels, distribution waterfalls and transfer restrictions can be encoded. But the rights themselves derive from the fund documents or articles of association, not from the code. Divergence between the two is a live legal risk in every issuance.

  CaveatWhere code and terms disagree, the terms usually prevail in court while the code prevails in practice. Closing that gap is a drafting problem, not an engineering one.

• Transfer agent / share registrar

  →On-chain register / registry smart contractPartial

  Both maintain the authoritative record of who holds what and process transfers. A smart contract can execute the entries, but in most jurisdictions the legal register is still operated by a person: under the German eWpG a licensed crypto securities registrar is mandatory, and the chain is merely the technical medium of the register.

  CaveatDeploying a token contract does not create a register. Without a licensed registrar and a statutory basis, on-chain balances are evidence of a position, not proof of title.

• Treasury bills

  →Tokenized treasury productPartial

  Tokenized treasury products are almost always fund units or notes referencing short-dated government paper, not direct holdings of the bills themselves. Return, duration and sovereign credit come from the underlying; the token adds a fund or SPV layer, fees and continuous transferability. The cash leg of a redemption still follows banking hours and cut-offs.

  CaveatRound-the-clock secondary transferability is not round-the-clock liquidity at NAV. Primary redemption follows the underlying market calendar, and the secondary price can trade at a discount.

• Value date / cut-off time

  →Continuous 24/7 operationPartial

  Chains settle continuously and have no business calendar, which removes cut-off risk on the token leg. Everything attached to the token still has one: fiat funding, valuation points, accounting periods and staff. Continuous operation relocates the problem to intraday liquidity and to who is available at three in the morning on a Sunday.

  CaveatA 24/7 market funded by a five-day banking system is a maturity mismatch. Liquidations occur at weekends precisely when the cash leg cannot be replenished.

• Bank deposit

  →Tokenized depositDirect

  A tokenized deposit is a deposit: same bank, same claim, same guarantee scheme, recorded on a distributed ledger instead of a conventional core banking system. Transferability is normally restricted to the issuing bank network or a consortium precisely in order to preserve that legal status. It is the closest available representation of commercial bank money on a ledger.

  CaveatThe trade-off is reach. A tokenized deposit is generally not freely transferable to any wallet, so it does not compose with public DeFi the way a bearer stablecoin does.

• Delivery versus Payment (DvP)

  →Atomic swap / HTLC / DvP in one transactionDirect

  The economic principle is identical: no delivery without payment. On-chain it is enforced by transaction atomicity rather than by a CSD and a settlement bank acting in sequence. The constraint is that both legs must sit on the same ledger, or be linked by a hashed timelock or bridge construction.

  CaveatCross-chain DvP and on-chain-against-fiat DvP are not atomic. Principal risk reappears exactly where practitioners assume it has been designed away.

• Haircut

  →Collateral factor / liquidation thresholdDirect

  The same concept applies: the lender values collateral below market to absorb price and liquidity risk. On-chain the parameter is public, uniform across all users and changes only by governance decision. In traditional markets it is bilaterally negotiated and can be varied by the lender at short notice.

  CaveatPublic and uniform parameters mean every borrower is liquidated at the same price level, which makes cascades more likely rather than less.

• Special purpose vehicle (SPV)

  →Issuance SPV / tokenization vehicleDirect

  The same instrument performs the same job: isolate an asset, issue claims against it, define the payment waterfall. Tokenization changes how those claims are recorded and transferred, not the structure itself. Jurisdiction, tax treatment and insolvency analysis remain the deciding factors.

  CaveatThe token is only as good as the SPV. Most RWA risk is credit, structural and jurisdictional, and none of it is affected by which ledger records the units.