The Catch With Onchain Equity

Imagine owning 10% of a private company. Somewhere, it keeps a list of shareholders showing your name and how many shares you own. That is its shareholder register: the record the company relies on when votes are counted, dividends paid and share transfers recorded.

In 2017, Delaware allowed companies to keep their official shareholder registers on a blockchain. The blockchain could be the authoritative record of ownership, rather than merely holding tokens representing shares recorded elsewhere. Nine years later, it remains a niche choice, used mostly by early-stage startups.

But putting shares onchain doesn't automatically make them easier to trade. Private companies can still restrict who owns their shares.

One company, MetaLeX, has a workaround: a separate token that can trade while the original shares remain registered to their owners.

That raises a bigger question: if blockchain was supposed to simplify share ownership, why does making those shares tradable add a layer between the investor and the company?

The clearest map of the options I have found comes from Gabriel Shapiro, a former BigLaw M&A lawyer who now runs MetaLeX, a small company building at the far end of it. On Galaxy's podcast in August, he sorted tokenised securities by a single test: where the legally definitive record lives. It is Shapiro's taxonomy rather than an industry standard, but it is useful.

At one end is the souvenir. Some early offerings, as Shapiro describes them, gave investors a token showing they had an entry on a transfer agent's books, while the books did all the legal work. The token proved something had happened and changed nothing.

Next comes what he calls pointer tokenisation, which is where much of the market sits. The token tracks a record someone else keeps, and the two are reconciled. Galaxy, a listed company, tokenised its own shares on Solana this way: Superstate is one of its transfer agents and records registered ownership while the token moves between allowlisted wallets.

The instruction token goes a step further. Moving the token is the instruction to update the books. This is the structure analysed in the final report of the committee that oversees the Uniform Commercial Code (UCC), the body of commercial law US states adopt, whose Article 8 governs how securities change hands. The committee concluded that Article 8 can already accommodate a digital token as the mechanism for transferring registered ownership of an uncertificated security, without treating the token itself as the security.

At the far end is what he calls constitutive tokenisation. The blockchain itself is the legally authoritative stock ledger. Changing its ownership state changes the legal record; there is no separate transfer-agent database that overrides it. This is what Delaware allowed in 2017.

Wrappers are different again. Rather than changing a company's shareholder register, they can give investors exposure to listed stocks without making them shareholders of the underlying company. Robinhood's stock tokens, as I wrote in July, are debt securities issued by a Robinhood company in Jersey and backed by shares held at a US custodian. Their holders get economic exposure, but no vote or direct claim on the companies whose stocks the tokens track. That is a different product from registering a private company's own shares onchain.

Franklin Templeton's OnChain US Government Money Fund shows why the categories get messy in practice. Launched in 2021, it was the first US-registered fund to use a public blockchain to process transactions and record share ownership. Franklin now describes the public blockchain as the fund's official system of record. Its prospectus is more precise: the transfer agent maintains that official record through a proprietary blockchain-integrated system, retains control over the records and can correct unauthorised transactions.

Galaxy made the same choice more plainly, with Superstate standing behind its token as the keeper of the register. Tokenisation changes where the asset can travel without removing the institution that stands behind the record.

That is the strongest argument for keeping an institution behind the token. When a court order arrives, a key is compromised or a record is wrong, there is somebody authorised to fix it and answer for it.

MetaLeX is the most complete attempt at the constitutive end I have found. A company's certificate of incorporation names smart contracts as its official stock ledger under those Delaware rules. Each holding is an NFT tied to a named holder, the signed agreement, a unit count, the officer who signed it and a record of every transfer. When a round closes, the investor's stablecoins and their shares change hands together. MetaLeX says it holds no key that can rewrite a company's cap table, and Shapiro's line is that MetaLeX is not your transfer agent; Ethereum is.

Shapiro says most companies using MetaLeX so far are early-stage startups whose securities don't trade at all.

That second detail matters more than it looks. I went in assuming the NFT was the thing people would trade. It isn't.

A token tied to a named holder and a signed agreement makes an excellent record and a poor trading instrument. Every new holder has to be checked, and every sale written onto the register under whatever restrictions the company has set.

So MetaLeX creates a separate tradable layer, not a new broker or custodian. A registered holder can lock part of their position and mint fungible tokens against it, which MetaLeX calls scrip. MetaLeX describes scrip as a security, with rights set by its terms. It can trade on a decentralised exchange or be posted as collateral while the underlying shares remain registered to the holder.

A scrip buyer is not automatically the registered holder of the underlying shares. An existing registered holder can burn scrip back into their registered position. A new holder needs an issuer-approved registration path before presenting scrip can put them onto the register.

The part that trades is no longer the registered position. Liquidity has introduced a layer between the trader and the register.

The route back onto the register can be written into smart contracts, which reduces reliance on manual processing. It is still a route whose conditions the issuer sets.

Shapiro's strongest answer deserves its full weight. Nobody has to hold scrip. A shareholder who wants direct ownership keeps the NFT, stays on the register and owns the share outright, with no depository, broker or software vendor between them and the company. Scrip is an opt-in for holders who want to trade or borrow against their position, and the terms for getting back onto the register are set in advance, in code. How far that loosens the company's control depends on how broadly an issuer can pre-approve new holders, which the public documentation does not settle.

Publicly listed shares offer a useful comparison, but they are not the same product. Buy 100 Apple shares through a US broker and your name will generally not appear on Apple's shareholder register. For shares held at DTC, the registered holder is Cede & Co., DTC's nominee. Your broker tracks your beneficial position, and dividends and voting materials reach you through that system. Eligible shares can instead be registered directly in your own name, although trading through a broker generally means transferring them back into the intermediated system.

That system emerged from a paperwork crisis. Late in the 1960s, trading volumes overwhelmed back offices that still moved paper certificates by hand, and the New York Stock Exchange closed every Wednesday to let clerks catch up. The Depository Trust Company, formed in 1973, reduced the need to move certificates between buyers and sellers. Shares deposited at DTC were registered to its nominee, while investors' interests could change hands through broker and depository records. DTC's onchain model is designed to preserve that arrangement: the assets remain in its custody while tokenised representations can move across approved networks, subject to controls including mint, burn, pause and clawback.

Shapiro reads that history as a technology problem. In the 1970s there was no such thing as a database nobody owned, so the register had to sit with an intermediary. Blockchains remove that excuse. The scrip layer suggests the harder constraint may be liquidity. MetaLeX has not recreated the broker or depository, but making the position easier to trade has still separated the liquid security from the registered shares.

Street-name ownership helped make trading in US public markets scalable. Scrip for private-company shares is a different legal instrument, but it answers a related problem: how to let an interest trade without recording a new shareholder on the company's official register after every transaction.

For issuers, the attraction is obvious. They can create a secondary market without giving up control over who reaches the register. For buyers, the trade runs the other way: they get a more easily tradable security, but their rights depend on its terms rather than direct registration as shareholders.

That separation has a cost. In 2013, T. Rowe Price publicly opposed Michael Dell's buyout of the company he founded and told its proxy adviser to vote against it. Its own voting system then generated instructions in favour, which travelled down a chain of intermediaries to Cede & Co., the shareholder of record. In 2016, a Delaware judge ruled that T. Rowe's funds could not seek appraisal for 26.7 million shares, while acknowledging the vote was a mistake. Shapiro tells this story himself as the case for registered ownership.

There is a question I could not answer : who votes the locked shares while their scrip is out in the market. If the underlying shares stay registered to the original holder, the person who sold the economics could keep the vote, unless the scrip's terms pass it through. The UCC committee flags the same tension whenever someone other than the economic owner sits on the register. MetaLeX leaves scrip holders' rights to terms each company writes, and I have not seen the terms of a live scrip instrument, so this is a design question to watch rather than a known outcome. If the answer is that the seller keeps the vote, it would be a version of T. Rowe's problem.

The first real test is close. MetaLeX has recently finished auditing a liquidity protocol, and Shapiro says an unnamed partner that tokenises SPV interests will run a market on it under existing private-market resale exemptions.

The UCC report includes a case study every general counsel should read. A fraudster drains a shareholder's wallet and transfers an instruction token to a buyer who pays in good faith. If the token transfer causes the issuer to register the buyer and the buyer qualifies as a protected purchaser, the buyer can take the shares free of the original holder's adverse claim.

What counts in law is the buyer reaching the register. The token is only how they got there. At the constitutive end, the token moving and the register changing are the same event, which is why it matters who can still intervene.

Constitutive tokenisation does not make administrative authority disappear. MetaLeX gives issuers powers such as force transfer, force burn, freeze and blocklist, with the option to renounce them individually. Its own standard, set out in a recent essay, is that "any power to intervene should have a defined scope and an accountable holder." That is a fair test, and it accepts the point: putting the register onchain moves authority into the rules governing who may alter the onchain record, where someone still holds it.

Privacy poses another problem. Putting shareholder records on a public blockchain risks exposing ownership information, which is why the UCC report notes that issuers running these systems normally keep holders' personal details in records off the platform. MetaLeX accepts the problem, writing that merely putting a cap table on a public blockchain "supplies none of those protections", and points to encryption and zero-knowledge proofs as the answer.

When the issuer's own documents make the blockchain its legally authoritative stock ledger, the onchain registered position can be the share. Making that share liquid is a different problem.

For private companies, MetaLeX's answer is scrip. Public markets developed street-name ownership, and DTC's onchain model preserves that underlying custody structure. These are different legal arrangements: MetaLeX has not recreated a broker or depository. What links them is the pressure to separate the registered position from the security or interest that can trade more easily.

The first MetaLeX scrip market will tell us more than another tokenisation announcement will. I’ll be looking out for who can buy the liquid security, what rights travel with it, who can turn it back into a registered position and who can reverse a bad transfer.

Putting the register onchain is possible, and has been for nine years. The harder question is whether a liquid market can exist without recreating a layer between the investor and the share. So far, even the furthest attempt has not escaped that trade-off.

See you next week.

James Smith