Skip to main content
development environment — data here is not real and actions have no financial effect
Tokenization

Asset Tokenization

The general practice of issuing a token against an asset — what changes, what does not, and where the actual work is.

Definition

Asset tokenization is the representation of ownership, or a claim on ownership, as a digital register entry that can be issued, transferred and retired programmatically. It applies as readily to a fund unit as to a warehouse of copper. What differs between them is not the token; it is the legal structure, the custody arrangement and the servicing obligations underneath.

What it solves

  1. Manual transfer agencyMaintaining a register by hand is slow, error-prone and expensive, and the errors surface years later during a distribution or an audit.
  2. Opaque ownershipAn issuer that cannot say who holds its instrument today cannot service it correctly. A live register answers that question without a request to a third party.
  3. Servicing that does not scalePaying two hundred holders by hand is a week of work and a reconciliation. Computed from the register, it is a run that either balances exactly or refuses.

How it works

  1. Define the instrumentWhat the holder owns, what it pays, when it matures or redeems, and who may hold it.
  2. Choose the registerA permissioned token whose transfer rules encode the restrictions the instrument actually carries.
  3. Issue against considerationUnits are created when money has arrived and an allocation says who is entitled to what.
  4. Service and retireIncome, corporate actions and eventually redemption or maturity.

Architecture

A pure domain layer holding the state machines and the arithmetic; a service layer that is the only thing which touches the database; and an interface that authorises, acts, audits and revalidates. Critical invariants are enforced twice — once in the application and once by the database — because the ones that matter are worth checking in two places with two different failure modes.

Tokenization lifecycle

  1. DefinitionAsset class, legal structure, parties, documents.
  2. ConfigurationStandard, supply, decimals, restrictions, lockups.
  3. Primary issuanceOffering, subscription, allocation, issuance, settlement.
  4. ServicingValuations, distributions, corporate actions, reporting.
  5. RetirementRedemption, maturity, wind-down.

Supported token model

Selected per instrument by the token standard engine. Fungibility, fractionalisation and whether transfers depend on the recipient determine the standard. Every recommendation carries its reasoning, its alternatives and its warnings, so the choice can be reviewed rather than assumed.

Asset requirements

What must be true before this can responsibly be tokenized at all.

  • Clear title or a clear contractual claim.
  • A jurisdiction whose law governs the instrument.
  • A valuation basis and a review cadence.
  • An accountable servicer for income and corporate actions.

Compliance considerations

  • The instrument’s regulatory classification, decided before issuance and disclosed.
  • Investor eligibility encoded as policy, versioned, and snapshotted onto each decision.
  • Transfer restrictions enforced on-chain as well as in the application.
  • Records retained so that any historical decision can be replayed under the rules that governed it.

Investor workflow

  • Verify identity.
  • Read the documents.
  • Subscribe and fund.
  • Receive an allocation and units.
  • Hold, receive income, and exit where permitted.

Issuer workflow

  • Structure and document the asset.
  • Configure the token.
  • Publish policy and open the offering.
  • Allocate, issue and settle.
  • Service for the instrument’s life.

Payments

Every movement of money posts to a double-entry ledger in the same transaction as the state change that caused it. Balances are derived from entries and never stored, because a stored balance is a second source of truth that will eventually disagree with the first.

Lifecycle servicing

Servicing is where most of an instrument’s life happens and where most operational loss occurs. Distributions, corporate actions and redemptions are each modelled as their own state machine with their own approvals rather than as ad-hoc adjustments.

Secondary transfer

Where the instrument permits transfers, they are requested, checked, approved by somebody other than the requester, and only then confirmed against the register. An approved transfer is an intention; only a confirmed one changes a position.

Risks

Named plainly. An instrument whose risks are only in a footnote has been mis-sold before it has been issued.

  1. Structural riskA defective vehicle or an unenforceable claim cannot be fixed downstream.
  2. Operational riskServicing errors — a missed distribution, a wrong record date — are the most common real-world failure.
  3. Technology riskSmart contracts are immutable by design here; a defect found after deployment cannot be patched in place.
  4. Market riskThe value of the underlying moves, and tokenization does nothing to change that.

How we support it

  • Sixty-seven asset classes resolved from group profiles plus per-class overrides, so a correction applies across a group rather than being fixed once and forgotten nine times.
  • A launch flow that will not let an asset reach an offering without the structure, parties and documents that instrument requires.
  • Readiness derived from what has actually been recorded, not from a checklist somebody ticks.

Questions

Which assets are worth tokenizing?

Ones where the register is currently manual, the holder base is larger than a handful, and servicing is recurring. An asset with two holders and no income does not gain much.

How long does an issuance take?

The software is rarely the constraint. Legal structuring, custody and regulatory sign-off set the timeline, and they are measured in weeks to months depending on jurisdiction.

Can an existing instrument be brought onto the platform?

Yes — the register can be initialised from an existing holder list, and from that point every change is a recorded movement. The migration itself needs care, because it is the one moment the register is asserted rather than derived.

Next steps

Bring an instrument you are actually considering. Structuring something real is the only way to judge whether the model fits.