There is a name on the share certificates of nearly every public company in America. It is not yours. It is not Vanguard's. It is not BlackRock's.
It is Cede & Co.
Cede & Co. is the nominee of the Depository Trust Company — itself a subsidiary of DTCC — and it is the registered legal owner of the overwhelming majority of all US public securities. DTC's depository assets crossed $100 trillion in mid-2025 and have continued to grow, all of it held in a single omnibus account. When you "buy a share of Apple," you are not, in the strictest legal sense, buying a share of Apple. You are buying a beneficial interest in a share that Cede & Co. holds on behalf of your broker, who holds it on behalf of you. The chain has at least three links, sometimes seven, and the ledger that says you own anything at all is reconciled overnight, in batches, against the master record at DTC.
This is the system whose subsidiaries processed $3.7 quadrillion in transactions in 2024 alone. It works. It has worked for fifty years. And it is the single biggest reason tokenization will not be won by crypto.
It will be won by the institutions that run the bug.
The popular case for tokenization sounds liberating. Equities, commodities, lending, real estate — all of it tokenized, all of it 24/7, all of it accessible to anyone with a wallet. Fees compress because the layers of intermediaries collapse. Markets become open. The privileged circle that decided who got access to what dissolves into a permissionless pool. It is, on its face, the same argument that powered Bitcoin in 2013 — money was for everybody now — extended to every other asset class.
It is also, on its face, half-true and structurally misleading.
It is true that BlackRock's BUIDL fund crossed $2.5 billion in tokenized money-market exposure across eight blockchains and is being listed on more, that tokenized US Treasuries surpassed $15 billion in early 2026, that Robinhood is rolling out tokenized US equities to European clients on its own Arbitrum-based Layer 2 with a stated ambition to extend the model to public and private equity. The directional bet is real. But the framing — liberation from intermediaries — confuses the wrapper with the system.
Tokenized equities today are mostly available to non-US persons or accredited investors. Backed Finance, xStocks, Robinhood EU — the on-chain ticker is permissionless; the access to it is not. KYC, accreditation, and Reg S walls remain intact. "Anyone can trade" turns into "anyone in jurisdictions where this is legal can trade," which is the same sentence US securities law has been writing for ninety years.
The fee critique is similarly fuzzy. Retail equity commissions have been zero at every major US broker for six years. The actual intermediary tax sits in places retail never sees: settlement, custody, securities lending, and the netting cycle inside DTCC. Tokenization compresses these costs meaningfully. But that compression accrues to the institutions that pay them — the broker-dealers, the prime brokers, the asset managers — not to the retail investor who already trades for free.
The retail story is not wrong. It is the cover story.
The actual story is what tokenization does to the back of the trade — and to the institutions that have spent fifty years building the back of the trade.
To understand why tokenization is a bigger deal for the major exchange groups than for any retail investor, you have to walk through what happens when you press "buy."
Today, in the United States, an equity trade settles at T+1 — one business day after execution. This is recent: the SEC compressed settlement from T+2 to T+1 on May 28, 2024. The reason it took fifty years to compress two days into one is not that nobody had thought of it. It is that compressing settlement requires every party in the chain to agree, and there are six parties in the chain.
The Broker-Dealer. Your interface. Schwab, Fidelity, Robinhood. Takes your order, decides where to route it. Roughly half of US retail equity orders are not routed to an exchange at all — they are sold to a wholesaler (Citadel Securities, Virtu) under Payment for Order Flow, which executes against its own book at a price matching or improving the National Best Bid and Offer.
The Exchange. NYSE, Nasdaq, CBOE, IEX, plus a long tail of Alternative Trading Systems and dark pools. Provides price discovery, runs the matching engine, surveils for manipulation. The listed franchises of global capital markets — owned increasingly by ICE (NYSE), Nasdaq Inc., CBOE Global Markets, LSEG, Deutsche Börse, JPX, HKEX, SIX.
The Clearinghouse. In US equities, the National Securities Clearing Corporation — a DTCC subsidiary. NSCC novates the trade: it inserts itself as counterparty to both sides of every transaction, so the buyer's exposure is to NSCC, not to the seller. If one side fails, NSCC makes the other whole. It nets the day's gross trades down to a single net position per member, a process that historically took two days, now takes one.
The Depository. The Depository Trust Company — also a DTCC subsidiary. Actual securities sit here. They have not moved physically since the 1970s, when the "paperwork crisis" of certificate-based settlement nearly broke Wall Street and led to the immobilization of share certificates inside DTC vaults. Ownership now changes by book entry. Cede & Co. is DTC's nominee — the name on the immobilized certificate.
The Transfer Agent. Computershare, EQ, AST, Broadridge. Maintains the issuer's official shareholder register. Handles dividends, proxy voting, corporate actions. The TA's books say Cede & Co. owns most of Apple. The brokers' books say their customers beneficially own that holding. The reconciliation between the two happens in batches.
The Custodian Bank. BNY Mellon, State Street, JPMorgan Asset Servicing, Citi. Holds securities in street name on behalf of institutional clients. Manages sub-custody chains for cross-border holdings. Provides securities lending desks, corporate action processing, regulatory reporting.
Six parties. Each with its own ledger. Each with its own reconciliation cycle. Each with a fee.
The point is not that these parties are unnecessary. They are extraordinarily necessary. The system processes a quadrillion-and-a-half dollars a year with vanishingly low failure rates because each party is specialized, regulated, and liable for its slice. The point is that at no point in the day, intraday, does any single party have a real-time gross view of who owns what.
DTCC sees the netted result at end-of-day. The transfer agent reconciles overnight. Brokers see their own customers but not other brokers' customers. Custodians see their clients but not the street. The only entity that sees the whole position in real time is — nobody.
This is fine when settlement is T+1 and netting absorbs noise. It becomes a problem the moment you ask the system to do anything new: real-time risk management, collateral mobility, intraday liquidity, cross-border atomic delivery-versus-payment, securities lending priced on a live order book.
Every one of those things is what tokenization is for.
Here is the part of the story the retail-democratization framing misses entirely.
Cede & Co. is not a glitch. It is the legal architecture that lets the modern stock market exist at the speed it does. Without an omnibus nominee, every trade would require a registry update at the issuer's transfer agent — millions of registry events per day, each of them touching the official shareholder list of every public company. The nominee structure decouples trading from registry. You can buy and sell a thousand times a day; the registered owner — Cede & Co. — never changes. Registry events happen at the perimeter, when securities move between brokers or out of street name.
The nominee structure is the bug and the feature. It makes high-velocity trading possible. It also makes corporate governance broken, securities lending opaque, and post-trade visibility a nightly batch problem.
Consider corporate governance specifically. Because Cede & Co. is the registered owner, your right to vote your Apple shares is a chain of delegations: from Cede & Co., to the DTC participant (your broker), to Broadridge (the proxy distribution monopoly), to you. Vote pass-through rates for retail are abysmal — a substantial fraction of beneficial shareholders never receive a ballot, and a smaller fraction return one. Corporate governance in America is, in practice, exercised by the institutions that custody the shares, not the people who beneficially own them. This is not a feature. It is a side effect of an omnibus-nominee architecture that solved a 1970s paperwork problem and never got revisited.
A tokenized securities ledger eliminates this by construction.
On a tokenized chain — whether it is Ethereum, a permissioned DTCC chain like Project Ion, or SIX's SDX — every transfer is a registry event. The chain is the official shareholder list. There is no Cede & Co., because there is no need for an omnibus nominee. The trading layer and the registry layer are the same layer. Settlement is atomic. The reconciliation problem disappears not because we got better at reconciliation but because there is nothing left to reconcile. Proxy voting becomes a smart-contract function call. Securities lending becomes a public order book. Intraday risk becomes a query against live state.
This is not a retail benefit. It is an institutional one — with a retail side effect that finally fixes shareholder democracy as a downstream consequence.
The popular framing of tokenization is a story about disruption — crypto-native protocols replacing legacy intermediaries. The actual story is about absorption. The legacy intermediaries are building tokenization themselves, on their own timelines, using their own regulatory wrappers, capturing their own value.
DTCC's Project Ion went live in a parallel production environment in August 2022, processing over 100,000 bilateral equity transactions per day on distributed ledger rails alongside DTC's classic systems. Citi, Goldman, and BNY Mellon are among the participants. The objective is not to disrupt DTCC. The objective is to migrate DTCC onto distributed ledger rails while keeping every existing relationship, every existing fee schedule, every existing regulatory privilege.
SIX Swiss Exchange has been further along for longer. SDX — SIX Digital Exchange — is the most established fully regulated tokenized securities exchange and central securities depository in the world. It runs as a parallel rail to the legacy SIX system. Bonds, equities, structured products, all issued natively on chain. UBS, Pictet, Sygnum, Hamilton Lane have all issued or settled tokenized instruments through SDX. Two newer regulated venues are now live alongside it: 21X in Germany, the first MTF licensed under the EU's DLT Pilot Regime by BaFin, Bundesbank, and ESMA, which began trading in September 2025; and BX Digital in Switzerland, the first Swiss DLT trading system authorized by FINMA. The category is no longer a single-venue category.
JPMorgan's Onyx (now Kinexys) processes hundreds of billions of dollars annually in tokenized intraday repo. Goldman runs DAP. HSBC runs Orion. BNY Mellon launched a tokenized collateral platform with Goldman in 2024. BlackRock's BUIDL is accepted as margin at Crypto.com, Deribit, FalconX — meaning a US Treasury position can be posted as collateral in a perpetual swap on a Caribbean-licensed exchange, in real time, on a Sunday. That is a multi-trillion-dollar efficiency unlock the moment it generalizes.
None of this is being built by crypto-native protocols. It is being built by the incumbents the crypto-native protocols were supposed to replace.
The reason is not that the incumbents are smarter. The reason is that tokenization without a regulated wrapper is, for most institutions, untouchable. The 1933 Securities Act, the 1940 Investment Company Act, ERISA, Basel III — these are the ground rules institutional capital operates under, and a token on a permissionless chain is a compliance impossibility for almost every pension fund and insurance company on earth. The wrapper is the product. The wrapper is what the incumbents already own.
There is a second, less remarked-upon advantage the incumbents enjoy: their regulators are partners, not adversaries. Nobody at LSE, ICE, NYSE, DTCC, or BNY Mellon is suing the SEC, the CFTC, FINMA, or BaFin. The opposite is true. These firms sit on advisory committees, file comment letters that get incorporated into final rules, and build their tokenized rails in real-time consultation with the agencies that will eventually supervise them. The same five years that crypto-native firms spent in litigation, depositions, and Wells notices, the incumbents spent in working groups, sandbox pilots, and DLT regimes co-designed with the regulator. By the time the formal rules arrive, the incumbents' platforms are already compliant by construction. Crypto-native protocols, even the well-funded ones, will spend their next five years trying to retrofit compliance into systems that were originally designed in opposition to it. That asymmetry compounds, and it is not a technology gap. It is a relationship gap.
What they did not have, until they built it themselves, was the rail. Now they have both.
The DEX architecture matters here, because it accidentally proved a thesis that decades of equity-market structure papers had argued in the abstract.
For most of the last forty years, the conventional wisdom in market structure was that vertical disintegration was efficient. Specialization drove down costs. Each layer of the trade — execution, clearing, settlement, custody — would be best served by a specialist firm focused on that layer. The unbundling of the LSE in the 1980s, the demutualization of exchanges in the 2000s, the rise of independent ATSs and dark pools, the separation of clearinghouses from exchanges — these were all moves in the direction of more specialization, more layers, more interfaces.
DEXes ran the experiment in reverse.
A perpetual swap on Hyperliquid is matched on Hyperliquid's order book, custodied on Hyperliquid's chain, cleared by Hyperliquid's risk engine, and settled in Hyperliquid's stablecoin balances. Every layer is the same entity. Every layer is the same state machine. There is no clearinghouse because there is no settlement risk; the trade settles atomically into the user's collateral balance. There is no custodian because the user's collateral is the chain. There is no transfer agent because there is no off-chain registry to reconcile.
This integration is precisely why Hyperliquid took share from venues that outsource each layer. Latency collapses when there are no inter-firm message passes. Capital efficiency collapses when collateral can be reused across products in a single account. Failure modes collapse when there is no counterparty risk between the matching engine and the settlement layer.
Crypto did not disrupt the exchanges. It demonstrated, in a controlled experiment with billions of dollars of real risk capital, what their next form looks like.
NYSE, ICE, Nasdaq, LSEG do not need to read the white paper. They can read the order book. The lesson is that the value chain is recompressing, and the firms that re-integrate vertically — execution + clearing + settlement + collateral on a single ledger — will own the post-trade economics that DTCC, NSCC, and the transfer agents currently fragment.
ICE owns NYSE. ICE owns ICE Clear US, ICE Clear Europe, ICE Clear Credit. ICE owns the data and analytics franchise it built out of the IDC, BondPoint, and TMC Bonds acquisitions, consolidated into ICE Bonds and ICE Data Services. The pieces are already inside one corporate parent. Tokenization is the technology that lets them run as one rail.
The same is true at LSEG (which acquired Refinitiv), at Deutsche Börse (which owns Clearstream and Eurex), at JPX, at HKEX. Each of these is one merger or one platform decision away from a fully vertically-integrated tokenized stack.
This is the move. This is what tokenization unlocks for them.
If you take the value chain seriously, the question is not whether tokenization wins — every major incumbent is already building it — but who captures the value when it does.
The answer, consistent with the thesis of Three Doors One Hallway, is a barbell.
The middle thins. Matching, clearing, settlement, transfer agency — these become commodity infrastructure on a unified ledger. NSCC, DTC, Computershare, and the dozens of mid-tier custody and reconciliation businesses that exist to reconcile their disagreements compress dramatically. The fees that flow to them today flow to a single integrated ledger tomorrow.
The front fattens. Distribution wins. The interface to retail and institutional capital — the Robinhoods, the Schwabs, the BlackRocks, the banks — becomes more valuable, not less, because the asset universe accessible through that interface expands by orders of magnitude. Tokenized equities, tokenized private credit, tokenized real estate, tokenized commodities, tokenized everything. Whoever owns the user owns the toll.
The bottom fattens too. Collateral mobility, securities lending, repo, prime brokerage on a unified ledger become a multi-trillion-dollar efficiency unlock. The firms that own the inventory and the credit relationships — BlackRock, JPMorgan, BNY Mellon, State Street, Goldman — capture this. Hyperliquid's lesson here is the most economically important: when collateral is composable across products in a single account, the institution that custodies the collateral wins disproportionately.
What sits in the middle — the specialist clearinghouses, the standalone transfer agents, the reconciliation vendors, the post-trade tech stacks — is where the compression happens. The Computershares of the world will not disappear, but they will be repriced.
This is the same barbell Three Doors One Hallway described for the front of the trade. The same barbell The Great Migration described for the M&A wave. The same dynamic, applied recursively to every layer of the financial stack: distribution and inventory fatten at the edges; the middle thins.
Tokenization is the latest instance. It is not the last.
The endpoint, two to five years out, looks the same wherever matching, clearing, settlement, and data already sit inside a single corporate parent.
Take NYSE/ICE as the worked example. ICE runs a tokenized listings venue where issuers issue native tokenized equities. The same entity provides matching, custody, clearing, settlement, and corporate-action processing on a single ledger. The same entity licenses access to broker-dealers and asset managers, who build apps on top — retail interfaces, institutional execution algos, structured product wrappers, on-chain ETFs. The same entity sells the data feed to the buy side.
It is an exchange. It is a clearinghouse. It is a custodian. It is a transfer agent. It is a settlement bank. It is the platform on which a thousand financial apps are built.
It is, in other words, what Hyperliquid is for crypto perps. What Binance is, internally, for retail crypto. What Stripe is for payments. What Apple is for mobile.
The same construction works at LSEG (London Stock Exchange + LCH + Refinitiv data), at Deutsche Börse (Eurex + Clearstream), at JPX (Tokyo + JSCC), at HKEX (Hong Kong + HKSCC + OTC Clear), at SIX (which has already shipped the early version as SDX), at Nasdaq, at CBOE, at CME. Each of these is one platform decision away from running its own integrated tokenized stack — and each of them already owns the regulatory wrapper and the existing relationships that let it deploy at institutional scale.
The institutions that get there first will own a structurally larger share of capital markets than they do today, because they will have collapsed six fee layers into one and captured the surplus. The losers are not the crypto-native protocols. The losers are the standalone clearinghouses, transfer agents, and reconciliation vendors that exist as separate businesses today because the value chain has been fragmented for fifty years and never had a reason to recombine.
This is what is actually being built. The retail-access story is the marketing copy.
The first time someone explained to me what Cede & Co. actually was — not a clerical detail but the legal nominee for over $100 trillion of American wealth — I assumed I had misunderstood. I had not. The plumbing of the modern stock market is held together by a single nominee account at a single subsidiary of a single quasi-private utility, reconciled overnight, against ledgers maintained by parties who do not see each other's books.
The system works. It has worked for fifty years. It is also, evidently, the bug — the single biggest piece of operational debt in the global financial system, the thing that makes every other improvement (real-time risk, intraday liquidity, atomic DvP, mobile collateral) impossible.
Tokenization is what fixes it.
The framing that tokenization is a retail liberation event is the marketing layer. The framing that it is a crypto-native disruption event is the press release. The actual event is an incumbent re-integration of a value chain that fragmented in the 1970s and is now, for the first time, re-mergeable. The institutions that built the bug will own the fix. They will charge for it.
The people who think they are buying a share of Apple will continue to buy a beneficial interest in a tokenized claim on a registered tokenized share, custodied by an entity that sounds a lot like Cede & Co. The chain will be shorter. The ledger will be faster. The fees will be lower. The owners of the rail will be the same.
That is what tokenization actually means.