The 10.7% Anchor: A Tokenized Equity Statistic With No Source, No Definition, and No Denominator

A single number is circulating through RWA research notes this quarter: 10.7% of the tokenized equity market is "issuer-sponsored." The remaining 89.3% is not. The figure arrived without a data provider. No Messari, no Dune, no RWA.xyz attribution. No definition of what "issuer-sponsored" means. No statement of whether "market cap" is circulating float or fully diluted. No indication of whether the same underlying shares were counted twice across chains. I have audited enough of these claims to know what a sourceless ratio does. It does not describe a market. It anchors a narrative. The contract says 10.7%. The methodology says nothing. That gap is the actual story. Tokenized equities — on-chain claims against shares of public companies — became the newest sub-narrative of the RWA trade in 2025. The pitch is clean: a user in a jurisdiction without efficient US equity access can hold a token tracking Apple or the S&P, settled continuously, composable inside DeFi. That pitch collapses into three legally distinct architectures, and the distinction matters more than any aggregate statistic. Issuer-sponsored tokens: permissioned standards (ERC-3643, ERC-1400 class) with whitelists and transfer restrictions, backed one-to-one by physical shares held in a special purpose vehicle. These may pass through dividends, may transmit partial voting rights, and may — depending on how the SPV is drafted — offer bankruptcy remoteness. Synthetic tokens: over-collateralized debt positions, perpetuals, or off-chain contracts-for-difference mapped on-chain. No physical shares. No shareholder rights. Pure price exposure, plus counterparty risk. Custodial receipts: a centralized custodian issues an IOU token. The holder owns a claim on the custodian, not on the stock. The 10.7% figure only acquires meaning once you fix which of these three counts as "issuer-sponsored." The source never does. Here is the definitional problem. The term admits two readings, and they produce opposite conclusions. Under the narrow reading, "issuer-sponsored" means the listed company itself issued the token. By that definition, 10.7% is absurdly high. Almost no public company has the appetite or the legal machinery to issue its own on-chain equity. The real number would sit in the low single digits, possibly lower. Under the industry reading, it means a licensed tokenization entity formally issued the token and holds the underlying share one-to-one on its balance sheet or inside an SPV. By that definition, 10.7% is a ratio of licensed issuers to grey-market synthetics. If the first reading holds, the statistic overstates the maturity of issuer-backed infrastructure by an order of magnitude. If the second holds, the source has quietly merged "licensed issuance" with "self-issuance by the underlying company" — a category error dressed as a statistic. That ambiguity is the most important technical finding in the entire dataset, and it sits in a footnote that does not exist. Now the deeper question: why do synthetics dominate at 89.3%? The lazy answer is that markets prefer cheap derivatives. The correct answer is cost structure. Issuer-sponsored issuance carries fixed costs — licensing, custody banking, audit, legal opinions, SPV maintenance — that do not scale down. At current market size, a licensed issuer spreads those costs across a thin book and struggles to break even. Synthetic issuance bypasses securities registration by delivering exposure as a derivative. Its marginal cost approaches zero. It is not a better product. It is a cheaper one. So the 89.3% is not evidence that bad money drives out good. It is evidence that regulatory friction has a price, and the market found the cheapest route around it. That reframing predicts the future. If tokenized equity market cap grows tenfold, fixed compliance costs amortize and the issuer-sponsored share should rise — provided regulators first clarify the registration path. The 10.7% may be a snapshot of early-stage cost structure, not a terminal state. The verifiability asymmetry deserves a paragraph of its own. For synthetics, collateralization is checkable on-chain in real time. For issuer-sponsored tokens, you can verify supply on-chain and nothing about the shares off-chain. Based on my audit experience reviewing institutional custody architectures, the hardest thing to attest is the thing sitting in a bank vault. Reserve attestations arrive quarterly, go unaudited between windows, and are signed by the party collecting the fee. The 24/7 pitch deserves the same scrutiny. An issuer-sponsored token settles on-chain instantly while its underlying settles T+1, and its price discovery dies when the exchange closes. The chain never sleeps. The asset does. There is a second finding buried under the ratio, and it is larger. Nearly every tokenized equity product — sponsored, synthetic, or receipt — transmits no voting rights. The 10.7% versus 89.3% argument is a debate about the quality of rights that 99% of the market does not grant anyway. Dividend pass-through is the only economically material difference, and its annual impact is typically one to three percent — sometimes less than the funding-rate cost a synthetic carries. A token is a receipt until you read the SPV's bankruptcy-remoteness clause. Most holders never read it. And here is the risk nobody is pricing. The published narrative frames this as an investor-rights problem — synthetics deliver exposure without ownership. True, and understated. The legal problem is worse than the civil problem. A synthetic token tracking a US equity is, in the EU, very likely a contract-for-difference under MiFID II. In the US, it may be a security-based swap. Both classifications trigger licensing regimes most DeFi deployments do not hold. That is not a rights deficiency. That is an enforcement action waiting for a test case. Issuer-sponsored tokens fare no better on distribution. A one-to-one backed token sold to retail across borders still implicates prospectus requirements — the EU Prospectus Regulation, Section 5 of the Securities Act. Backing does not equal exemption. The honest framing: 89.3% of this market may be an unregistered securities derivative, and the other 10.7% may be an unregistered securities offering. The difference is the flavor of the violation, not the presence of one. I have watched this pattern before. During the TerraUSD audit, the marketing described a stablecoin and the mechanism described an unhedged short. The label and the structure diverged, and the structure won. Tokenized equities repeat the trick: the word "stock" does the marketing, the word "swap" does the plumbing. The bulls are right about one thing, and it deserves stating plainly. Synthetic architecture is not a degenerate shortcut. It is the only design that achieves real composability. Permissioned standards with whitelists and transfer hooks cannot circulate freely through AMMs or lending markets — that is the entire point of the whitelist. If your thesis is that tokenized assets should plug into DeFi, synthetics are the only vehicle that currently does. Second, 10.7% may be measuring the wrong thing. If large retail brokers in Europe wrap tokenized equity as contracts-for-difference — and several do — their products are statistically synthetic even though a regulated entity issues them. In that case 10.7% understates the share of the market that is institutionally issued. The statistic could be simultaneously too generous, by mixing definitions, and too harsh, by misclassifying regulated brokers. Third, and this is uncomfortable for my own position: if the SEC and MiCA clarify registration paths within twelve to twenty-four months, the minority becomes the scarce asset. Licensed issuer-sponsored tokens would carry a compliance premium synthetics cannot manufacture. The 10.7% would stop being a warning. It would become an inventory list. Strip the rhetoric and one question remains, and it is not about 10.7%. For any tokenized equity position, ask whether you hold price exposure or asset ownership. Those are different products, sold under the same ticker, governed by different law, and — in a liquidation — resolved by different courts. The market has spent a year arguing over a ratio it cannot source. It has spent almost no time on the fact that the category's legal status is unsettled across every major jurisdiction at once. When that resolves — and it will resolve through enforcement before it resolves through legislation — the percentage will change overnight. The investors holding the wrong architecture will discover they were never arguing about rights at all. They were arguing about whether the product would still exist.