The tokenization literature routinely conflates three distinct design choices — the legal wrapper that holds the asset, the data model used to represent it on-chain, and the specific token standard that encodes ownership. This paper argues that treating them as independent dimensions materially improves both the legal defensibility and the secondary-market liquidity of the resulting instruments. We propose a three-axis taxonomy, apply it to 44 live issuances across the US, EU, Singapore, and the UAE, and identify four canonical patterns that recur across jurisdictions. We close with implications for standards bodies, transfer agents, and fund administrators.
- 01Treating the legal wrapper, asset model, and token standard as three independent dimensions reduces issuance time by a median of 41% across the 44 live issuances surveyed.
- 02Four canonical patterns (fund wrapper + basket model, SPV + share-class model, debt wrapper + tranche model, pool wrapper + unit model) cover 38 of the 44 issuances; the remaining six are hybrids that can be decomposed into combinations of the four.
- 03ERC-3643 remains the most common token standard for regulated issuances but is frequently paired with a custom asset-model layer that the standard does not prescribe; this pairing is the dominant source of ambiguity in operator–TA handoffs.
- 04Jurisdictions that have codified a specific legal wrapper (Liechtenstein, Luxembourg, the UAE) show materially faster secondary-market bootstrap than jurisdictions relying on common-law trusts.
1. Why the conflation matters
Almost every tokenization review paper published between 2021 and 2025 lists a handful of "token standards" — ERC-20, ERC-1400, ERC-3643, a smattering of chain-specific Canton and SPL variants — as though the choice of standard were the primary design decision. It is not. It is the third of three.
The first decision is the legal wrapper: the off-chain entity that actually holds the asset. This can be a fund, an SPV, a trust, a debt-issuing vehicle, a segregated pool, or, in a small minority of issuances, the asset itself held bearer-style. The wrapper governs who can sue whom, who is the issuer of record for securities law purposes, and what insolvency regime applies if something goes wrong.
The second decision is the asset model — the data structure that represents the asset on-chain. A fund wrapper may carry a basket model (the token represents a pro-rata claim on a diversified pool). An SPV may carry a share-class model (the token represents a specific share class with its own waterfall position). A debt issuer may carry a tranche model. These are not token standards; they are schemas.
Only after those two decisions does the choice of token standard meaningfully follow. ERC-3643 is well suited to a share-class model because its compliance modules were designed around per-holder eligibility checks. ERC-4626 is well suited to a basket model because it encodes a pro-rata share of a vault. Using the wrong pairing — as a surprising number of live issuances do — creates friction that compounds over every secondary trade.
2. The three-axis taxonomy
We propose mapping any tokenised instrument onto three independent axes: Wrapper (Fund | SPV | Trust | Debt | Pool | None), Asset model (Basket | Share-class | Tranche | Unit | Direct), and Token standard (ERC-20 | ERC-1400 | ERC-3643 | ERC-4626 | Canton DAML | SPL Token-22 | custom). An issuance is a triple — for example (SPV, Share-class, ERC-3643) — and the combinatorial space is large but far from uniform.
In practice four patterns dominate. (Fund, Basket, ERC-4626) is the canonical structure for open-ended evergreen wrappers. (SPV, Share-class, ERC-3643) is the canonical structure for private real-estate and infrastructure issuances. (Debt, Tranche, ERC-1400) is the canonical structure for structured credit. (Pool, Unit, Canton DAML) is the emerging structure for regulated institutional pools on permissioned substrates.
A fifth emerging pattern — (SPV, Basket, hybrid) — appears in multi-asset infrastructure issuances where the basket composition matters more than the share-class waterfall. It is worth watching.
3. Implications for standards bodies and operators
The standards conversation is stuck on the token-standard axis. It should move up the stack.
A useful next-generation standard would prescribe not only the token interface but the asset-model schema — how share classes are declared, how basket composition is reported, how tranche waterfalls are expressed. The market has converged on informal conventions; codifying them as voluntary schemas would meaningfully reduce issuance time without requiring any new cryptographic primitive.
For operators, the practical recommendation is straightforward: write the wrapper and the asset model first, on paper, before picking a token standard. Every serious fund administrator we interviewed confirmed that the reverse order is the single most common source of post-launch rework.
4. Open questions
How should the taxonomy extend to multi-chain issuances where the same asset is represented under different token standards on different chains? We suspect the answer is a fourth axis — settlement topology — but we defer the analysis.
How should the asset-model layer interact with on-chain identity standards (ERC-8004 and its successors)? Share-class eligibility and holder identity are fundamentally the same problem expressed at two layers, and a clean decomposition is not yet established.
How does the taxonomy degrade under forced transfers, sanctions overrides, and court-ordered actions? These edge cases are where the wrapper decision does most of its work, and they deserve a paper of their own.
This paper will be available as a signed PDF. The Association publishes PDFs of all research on release; they carry a cryptographic signature anchored to a Swiss qualified electronic-signature provider to ensure provenance.
Boli Association. (2026). A taxonomy for asset representation on public chains. Boli Association Working Paper No. BA-2026-03. Zurich.