On a September morning in 2025, a retail account in Frankfurt could open an application, tap a ticker, and hold something that moved exactly like a share of a United States-listed company. The interface displayed the company's name. The price chart echoed the company's price. The order book agreed. Nothing on that screen disclosed that the holder had acquired no equity, no voting right, no claim on the issuer's balance sheet, and no line on the register of shareholders.
What the account held was a token issued by a third party, backed one-for-one by shares sitting in a custody arrangement, wrapped in a legal instrument that had never been negotiated with the company whose name appeared on the ticker. The issuer of the underlying equity had not consented. In at least one documented case, it had objected in public and asked for the practice to stop.
The objection is the story. Not the token. Not the chain. The chain is a delivery mechanism, and a slow one at that. The novel element is a legal question that has sat unresolved in securities law for decades and is now being stress-tested by a broker with a retail distribution channel and a founder willing to argue the point in public: can a third party manufacture and sell economic exposure to a security without the issuer's permission?
My answer, after a decade of dissecting protocols that promised more than their code could deliver, is that this question is not about technology at all. It is about consent. Consent is the layer nobody audits. The math holds, but the humans did not verify it.
To understand why the dispute matters beyond one broker's product line, you have to separate what was launched from what was claimed. What was launched is a European distribution of tokenized exposure to United States equities, targeted at non-American users, sitting inside the MiFID II and MiCA perimeter. What was claimed is considerably larger: that this constitutes a new asset class, a democratization of market access, and a structural improvement in how securities move. Those are different propositions with different evidentiary burdens, and the second one has not been demonstrated.
The product itself is not technically interesting. I want to be precise about that, because the crypto commentariat has spent weeks treating this as a technological event when it is a legal one wearing a technical costume. The architecture is a wrapped asset. A custodian or special purpose vehicle holds the underlying share. A third-party issuer creates an instrument that references that share one-to-one. That instrument is tokenized on a chain and distributed through a retail application. There is no new cryptography here. There is no novel consensus mechanism. There is no zero-knowledge proof doing what a signature could not. The innovation, insofar as the word applies, is regulatory arbitrage executed well enough to reach production.
The legal argument advanced by the broker's co-founder runs roughly as follows. A third party may create an independent instrument that references a security at a one-to-one ratio without touching the issuer's rights, the issuer's obligations, or the issuer's official register. The point at which issuer consent becomes necessary is the point at which the instrument seeks to alter the underlying share's rights, replace the official shareholder register, or impose new obligations on the issuer. Below that threshold, what exists is a referencing instrument, and referencing instruments have a long and unremarkable history: non-sponsored depositary receipts, listed options, structured notes, contracts for difference, total return swaps.
That argument is not frivolous. It is, in fact, the strongest part of the case, and I will return to it in the contrarian section because the bears have mostly ignored it.
The background that the argument elides is that non-sponsored depositary receipts have generated their own decades of litigation, regulatory inquiry, and issuer complaint. The existence of a precedent does not settle whether the precedent is a good fit for a new wrapper; it only establishes that the general shape has been contemplated before. A non-sponsored ADR trades on a regulated exchange, settles through established clearing infrastructure, and is subject to a disclosure regime that the issuer can, at minimum, observe and respond to. A tokenized equity reference instrument distributed through a mobile application to a European retail base settles on a chain, prices against an oracle, and may or may not be redeemable for the underlying. Those are not cosmetic differences. They are the differences that determine what happens in a dislocation.
The competing argument, voiced loudly by the chief executive of a media company whose equity is among the referenced names, is that a tokenized referencing market can decouple secondary trading from the issuer's capital formation. If a meaningful share of liquidity and attention migrates to a shadow venue denominated in a wrapper the issuer does not control, the price discovery that governs the issuer's cost of capital degrades. That is a real concern. It is also a concern that applies with equal force to the existing off-exchange volume in United States equities, which is now a majority of all trading. The issuer's complaint is not new. It is just newly visible because the wrapper is new.
What neither side has addressed, and what I find more consequential than either argument, is the operation of the custody layer. Two parties are arguing about the ownership boundary while a third layer โ the one that actually holds the shares and is contractually obligated to honor redemptions โ goes almost entirely undiscussed. That omission is where the risk lives.
Let me decompose the structure the way I would approach it in an audit engagement. There are three functional layers, and each carries a distinct failure mode.
The bottom layer is the real equity, held in custody, presumably by a regulated custodian or inside a special purpose vehicle. This layer is boring until it isn't. Its failure mode is the failure mode of every custodial arrangement: counterparty solvency, operational error, rehypothecation, and the difference between what the custody statement says exists and what actually exists at the moment of a stress event. I have never seen a wrapped-asset structure fail because the token contract misbehaved. I have seen several fail because the custodian's assets and the issuer's liabilities did not reconcile under time pressure.
The middle layer is the legal wrapper: the instrument that grants the holder economic exposure without legal title. This is where issuer consent matters most, and it is the layer that determines what the holder actually owns. The holder owns a contractual claim against the issuer of the wrapper. That claim may be collateralized, may be bankruptcy-remote, may be senior or subordinated, and may or may not entitle the holder to demand delivery of the underlying. Every one of those attributes is a design decision, and none of them were disclosed at the point of sale in the manner that a securities disclosure regime would require. Absent that disclosure, the holder is not purchasing an asset. The holder is purchasing a set of assumptions about a contract they have not read.
The top layer is the token on the chain. It is the least important layer and the most discussed. Its function is transferability and programmability, and only one of those two functions is currently exercised. Transfers happen. Composability does not, because the wrappers are not, at present, integrated into lending markets, collateral systems, or decentralized exchange pools in any meaningful volume. A token that can only be bought and sold inside the issuing application's own perimeter is a database entry with extra steps. That is not an insult to the technology. It is an observation about deployment.
This is the same structural lesson I documented years ago when I took apart the metadata layer of a prominent profile-picture collection and found that the supposedly decentralized storage depended on a single cloud provider for critical retrieval. The contract was immutable. The artifact was not. The token was on-chain. The thing the token pointed at was not. The community called the finding pedantic, right up to the point where the retrieval path became a valuation input.
Provenance is a story we agree to believe in. The same sentence applies here. A tokenized share's provenance is a custody statement, a legal opinion, and a reconciliation process. The chain does not verify any of it. The chain merely records that a transfer occurred between two addresses, both of which were authorized by an entity whose solvency and operational competence remain unmodeled by the holder.
The regulatory analysis is where most commentators have gone wrong in both directions. Some have argued that tokenized equities are obviously securities and therefore the entire discussion is moot. Others have argued that because they are securities, everything is settled and the product is fine. Both miss the actual question. Whether the instrument is a security is not in dispute and never was. It obviously is. The Howey test โ investment of money, in a common enterprise, with an expectation of profit, derived from the efforts of others โ is satisfied on all four prongs with room to spare. The interesting question is not classification. It is authority: who holds the right to create a referencing instrument, and under what conditions does that right require the referenced party's consent?
Current securities law does not answer this cleanly for on-chain wrappers. It answers it partially for depositary receipts through a combination of registration requirements, exchange listing rules, and decades of accumulated practice. It answers it partially for options and structured products through the same accumulation. For tokenized wrappers distributed through a retail mobile application in the European Union, the applicable frameworks are MiFID II for the distribution and MiCA for the crypto-asset layer, and the two do not dovetail cleanly for an asset that is simultaneously a financial instrument and a token. MiCA's asset-referenced-token category was built with stablecoins and baskets in mind. Forcing a single-name equity reference instrument into that frame produces a classification mismatch with real consequences for reserve, disclosure, and redemption requirements.
The product's current geographic restriction to non-United States users is not incidental. It is load-bearing. Distribution into the United States would invite direct Securities and Exchange Commission scrutiny of a security-linked instrument issued without the referenced issuer's participation, and the absence of a settled legal path for that instrument would put the distributor in the position of testing the question through enforcement rather than through guidance. Nobody with a compliance department wants to be the test case.
So the operative uncertainty is not technical and not even economic. It is jurisprudential. Until a court or a regulator with authority rules on the consent boundary, every participant in this category is operating on an assumption. Assumptions are just risks wearing disguises.
The operational risks deserve more space than they have received, because the commentary has focused on the legal disagreement and treated the mechanics as settled. They are not settled. Consider what happens in a sharp dislocation of the underlying equity โ a halt, a limit-down sequence, an overnight gap driven by news. The token trades continuously. The underlying does not. During the gap, the price relationship between the two is maintained by whatever mechanism the issuer has installed: an oracle, a market maker, or a discretionary intervention. If the oracle is stale, the token diverges. If the market maker withdraws, the spread widens without bound. If the intervention is discretionary, the issuer has acquired pricing authority over an instrument it also issues, which is precisely the conflict that securities regulation exists to constrain.
Now consider redemption. The one-to-one backing is a promise that, at some defined time and under some defined conditions, a token can be exchanged for the underlying or its cash equivalent. Promises of that shape have a well-documented failure mode. When holders of a redeemable claim all attempt to exit simultaneously, the backing must be liquid enough to meet the demand. If the backing is the underlying equity and the demand arrives during a market closure, the redemption queue extends. If the queue extends far enough, the secondary price of the token decouples from the underlying and the market discovers that the one-to-one relationship was never a law of nature. It was a liquidity assumption. I spent months modeling exactly this dynamic in the aftermath of an algorithmic stablecoin collapse, and the conclusion generalizes: any peg that depends on continuous confidence is a peg that terminates when confidence does.
The tokenized equity wrapper is not a stablecoin, and I am not claiming equivalence. But the redemption mechanic is structurally analogous, and the disclosure of that mechanic is the difference between a product and a trap. The exit liquidity is someone else's regret, and in a redemption queue the last holder discovers this at the worst possible moment.
There is a further omission worth flagging. The public discussion has not identified the custodian, the special purpose vehicle's jurisdiction, the bankruptcy-remoteness opinion, the oracle source, the chain, the bridging architecture if any, or the continuity plan for a custodian failure. These are not trivial details to be deferred to a technical appendix. They are the entire risk surface. A product whose value proposition is that it delivers exposure to a regulated asset, distributed through an application that markets itself on clarity and simplicity, has a duty to disclose the mechanics of the exposure in the same place it advertises the exposure. The absence of that disclosure is not a communications failure. It is a structural one.
The market structure consequences extend beyond the issuing firm. If the third-party wrapper model survives legal scrutiny, the value transfer runs from issuers and transfer agents toward brokers and on-chain distribution platforms. The transfer agent's role โ maintaining the authoritative record of who owns what โ is the specific function the wrapper model routes around. That is not an accident of implementation. It is the point. Any participant whose business is the authoritative register should be reading this dispute very carefully, because the question being asked is whether their function can be replicated by a private party with a custody account and a smart contract. The answer, technically, is yes. The answer, legally, is pending.
Exchange venues, by contrast, are unambiguous beneficiaries regardless of how the consent question resolves. A new tradable category generates listing revenue, data revenue, and order flow whether or not the legal foundation is stable. Infrastructure providers โ wallets, node operators, indexers โ benefit from additional on-chain assets regardless of their provenance. Decentralized finance stands to gain the most and is positioned to receive the least, because a wrapper whose legal status is uncertain cannot be safely accepted as collateral, and an asset that cannot be collateralized has no composability. It can be held. It cannot be used. That distinction is the difference between an on-chain asset and an on-chain receipt.
This is where I part company with the enthusiasts, and I want to be explicit about the mechanism of my disagreement rather than the conclusion. The claim is that tokenizing equities makes them more useful. Usefulness, in a financial system, is a function of what an asset can be pledged against, borrowed against, hedged with, and settled into. A wrapper that lives inside one application's perimeter satisfies none of those. It is useful for one thing: gaining price exposure through a channel the user already trusts, with better hours and lower frictions than the traditional rail. That is a real improvement. It is also a narrow one, and it does not require a blockchain. It requires a broker willing to operate outside market hours. The chain is the expensive way to get there.
I have made a version of this argument before, in a different context, when I was asked whether AI agents executing contracts autonomously represented a new security paradigm. My answer then was that the security properties of the system are determined by the boundaries you impose on the non-deterministic component, not by the sophistication of the component. A model that interprets ambiguous instructions will eventually interpret an instruction in a way you did not intend, and the only defense is a deterministic constraint at the interface. The same principle governs wrapped assets. The chain does not make the wrapper safe. The constraints at the boundary โ disclosure, custody segregation, redemption terms, consent โ make the wrapper safe. Remove the constraints and you have automated a risk you did not measure.
Now to the part the bears will not like.
The bulls are right about the legal precedent, and the bear case has largely failed to engage with it. The proposition that a third party may construct a referencing instrument without the referenced issuer's consent is not an invention of the crypto industry. It is the established shape of a very large portion of modern capital markets. Total return swaps reference equities without issuer consent. Listed options reference equities without issuer consent. Contracts for difference reference equities without issuer consent. Non-sponsored depositary receipts reference foreign equities without issuer participation, and they trade in volume every day. The issuer's complaint that its capital formation is being affected by activity in instruments it does not control is a complaint that has been made against the derivatives market for forty years and has never been accepted as a basis for prohibition.
If the consent requirement were imposed, the consequences would reach far beyond tokenized equities. It would imply that the referenced party holds a veto over the creation of any referencing instrument, which is not the law and has never been the law. The bulls are correct that the distinction between using an issuer's name in a fabricated prospectus and building an independent instrument that references a listed security is a real distinction, and the second is generally permitted.
Where the bulls go wrong is in treating legal permissibility as sufficient. Permissibility answers whether you may build it. It does not answer whether the thing you built is safe to hold, whether its mechanics are disclosed, whether its custody is sound, or whether its holder understands what they own. A product can be lawful and still be mis-sold. A structure can be permitted and still be fragile. The bear case has been arguing the wrong question, and the bull case has been treating an answer to that wrong question as vindication for everything else.
There is also a version of the bull case that deserves more credit than it has received: the customer genuinely benefits. A retail investor outside the United States who wants exposure to a United States equity faces currency conversion, market hours, brokerage access barriers, and settlement delays measured in days. A tokenized exposure product reduces all four. If the wrapper is properly collateralized, transparently custodied, and redeemable on disclosed terms, the user gets something real. The demand is not manufactured. It is just currently served by a structure that has not published the terms that would let a careful buyer evaluate it.
Which brings me to what I consider the actual fault line, and it is not between the broker and the issuer.
The fault line is between narrative and mechanics. The narrative says ownership. The mechanics deliver exposure. The narrative says decentralized. The mechanics depend on a custodian, an issuer, a market maker, and an oracle. The narrative says demand-driven. The mechanics require a legal theory that has not been tested. Every one of those gaps is a place where a holder's expectation and a holder's entitlement diverge, and divergence of that kind is the precondition for every market failure I have ever documented. Correlation is the comfort of the unprepared.
What should a reader do with this, in a market where survival matters more than return?
Track the consent question as your primary variable. Not the token's price, not the launch metrics, not the competitive positioning of the issuing venues. The single event that reprices this entire category is the first authoritative legal or regulatory determination of whether a third-party equity wrapper requires the referenced issuer's authorization. If the answer is yes, the model must be rebuilt around licensing agreements and the economics change materially. If the answer is no, the model becomes replicable by anyone with a custody account, and the issuing firms' advantage compresses to distribution. Either outcome is significant, and neither is currently priced because neither is currently known.
Second, demand the custody disclosure before you demand the price. Who holds the underlying, in what entity, in what jurisdiction, under what segregation, with what audit cadence, and with what redemption waterfall. If those answers are not published, the wrapper is a promise, and promises are only as good as the balance sheet behind them. A public chain is a poor substitute for a public disclosure.
The disagreement between a broker founder and a media chief executive will be resolved in a courtroom, a regulatory order, or a statute, and the resolution will define the boundary of the tokenized asset category for a decade. What will not resolve itself is the gap between what the wrapper is called and what the wrapper is. That gap is a design choice, and design choices are reversible. The only question is whether the industry reverses it before a redemption queue does it for them.
Value is consensus; truth is optional. In this case, the consensus is forming faster than the disclosure, and the holders are the ones who will discover which of the two they actually bought.