A sector that sells “disintermediation” still runs on names. This is not a hall of fame. It is a map of who touched which layer — idea, standard, issuance, venue, reserve, depositary — so you can tell a protocol founder from a transfer agent when a press release blends them.

The idea layer

Nick Szabo gave the English. Smart contracts in the mid-1990s. Property titles that could survive a bad registry. He did not ship a tokenized equity. He shipped the suspicion that a title should be harder to fake than a filing cabinet.

Yoni Assia wrote the 2012 note that made Bitcoin look like a register for other things. Meni Rosenfeld made that note look like an engineer’s spec. Vitalik Buterin sat on those drafts before he built a machine where the token did not have to piggyback on a satoshi.

J.R. Willett insisted the new assets should ride on Bitcoin rather than fork it, then proved people would fund the ride.

Fabian Vogelsteller turned a messy wiki draft into ERC-20, the socket every later “real” token still resembles even when the compliance hooks are bolted on.

Mike Hearn, in early Bitcoin circles, spent time on smart property and timelocks — the notion that a ledger could govern access to a thing, not only a balance. The implementations moved on. The instinct did not.

The first regulated experiments

Patrick Byrne and the Overstock / Medici / tZERO line of work matter because they refused to treat a share as a toy. They bought market plumbing (an ATS path, a router) and tried to issue company paper on a ledger while the ICO market was doing the opposite. Byrne’s public claim was blunt: equity tokens would need a venue that securities law could see.

Whether any single ATS “won” is beside the point. The template — license first, token second — is the template the serious end of the market still uses.

The compliance decade

Polymath treated issuance as software with restrictions (ST-20, later a dedicated chain). Harbor treated a transfer as a question a contract had to ask. TokenSoft and the ERC-1404 conversation made the “no” readable. Securitize collected the licenses that make “the token is the record” something a fund counsel will discuss — transfer agency, broker-dealer, ATS — and became the name institutions cite when they want a share class to live on a chain without leaving the 1940 Act conversation.

These firms are easy to skip in a pioneers list because they do not look like 2012. Skip them and you cannot explain 2026.

The reserve layer that made the charts honest

Stablecoin issuers — Tether on Omni first, then the regulated cousins — accidentally trained the world to ask “what is in the box?” Tokenized treasury and money-market issuers reused the question for funds. BlackRock (BUIDL), Franklin Templeton, Ondo, WisdomTree, Superstate, Figure, Spiko and others turned NAV into something a token could inherit instead of invent.

On the equity-wrapper side, Ondo Global Markets and Backed / xStocks filled screens with hundreds of names. They did not settle the legal debate. They gave the debate a tape.

The pipes

DTCC / DTC is not a startup. It is the place American securities already live. A no-action letter in December 2025 and live production work in 2026 moved tokenization from “issuer experiment” to “depositary service.” Nasdaq and NYSE writing rules for tokenized entitlements on the same book as the traditional share is the venue half of that move.

Chainlink, Pyth, and the marks coming off issuer desks are the reason a second price exists at all. An equity wrapper without a second price is a blog post.

How to read a “top tokenization companies” list

Ask which layer the name occupies.

• Idea or standard: you are reading history.
• Issuer of wrappers: you are reading a product catalog. Check the documents.
• Transfer agent / broker-dealer / ATS: you are reading the adult stack.
• Depositary / exchange: you are reading whether the traditional market intends to keep the record.
• Oracle / data: you are reading whether anyone can audit the print.

We keep a live, sour sector tape on /data. This page is the human index for that tape.