People talk about tokenization as if it were invented when a large asset manager put a fund on a public chain. That is a chapter. It is not the book.

The book starts as a property-title problem, becomes a Bitcoin metadata trick, becomes a fundraising craze, becomes a compliance product, and only then becomes something a depositary will put into production. If you skip the middle, you cannot read the present.

Before there was a chain to write on

In 1994 and again in 1997, Nick Szabo described “smart contracts” — agreements that could be embedded in the software and hardware that already controlled valuable things. His teaching example was a vending machine. The machine takes a coin and releases a can because the mechanism makes theft more expensive than the can. He wanted that logic on houses, securities, and names.

In 1998 he pushed further with secure property titles: a public record of who owns what that would be harder to forge than a paper registry in a bad year. He did not have a global replicated ledger that strangers would trust. He had the shape of the problem.

That shape is tokenization. A title that can move with a key.

2012: coloring a satoshi

On 27 March 2012, Yoni Assia published a note titled “bitcoin 2.X (aka Colored Bitcoin).” The proposal was almost childlike. Take the smallest unit of bitcoin. Mark it. Treat the mark as a claim on something else — a share, a bar, a contract. The bitcoin network would carry the mark for free because it was already carrying the coin.

By December 2012, Meni Rosenfeld had written the idea into a specification. Genesis transactions. Color tracking through the transaction graph. A year later Assia, Rosenfeld, Vitalik Buterin and others put more flesh on “BitcoinX.” The New Scientist would print the thought that a share or a gold bar might live as a colored coin.

Nothing about 2012 was a market. It was a method. The method said: you do not need a new chain for every new asset. You need a way to point at one.

J.R. Willett had already been circling a cousin of the same idea. In early 2012 he circulated what he called a second Bitcoin white paper: use Bitcoin as a base layer and issue new rules on top of it, rather than launching a competing chain for every experiment. “Alternate block chains compete with bitcoins financially, confuse our message to the world, and dilute our efforts,” he wrote. The implementation became Mastercoin, later Omni. In the summer of 2013 the project ran what the industry still treats as the first token sale. Thousands of bitcoin moved to an Exodus address. A second-layer asset class had a treasury.

Tether would later issue on Omni. That single fact is easy to skip and expensive to skip. The largest tokenized dollar of the next decade did not begin on a purpose-built securities chain. It began as a colored claim on Bitcoin.

Counterparty arrived in 2014 by burning bitcoin to mint its own unit, then issuing assets on top. Open Assets, Colu, ChromaWay, and Coinprism all tried to make coloring operational. Most of those names are footnotes now. The footnotes matter. They prove the instinct predated Ethereum.

2015: a standard that made issuance cheap

Ethereum gave the instinct a virtual machine. In November 2015 Fabian Vogelsteller opened issue 20 on the Ethereum improvement process and turned a draft circulating around Vitalik Buterin into ERC-20 — a common interface so a wallet, an exchange, and a contract could treat one token like another.

ERC-20 did not know what a security was. It knew transfer and balanceOf. That ignorance was the feature and the later lawsuit. Suddenly anyone could issue a unit that moved like money. 2016 and 2017 filled with offerings that used the standard as a fundraising loudspeaker. The word “token” spent two years meaning “new ticker,” not “old claim.”

The hangover is part of the history. Tokenization of real assets had to crawl out from under a boom that had used the same file format for a different job.

2015–2017: a retailer tries to issue a stock

While the ICO tape ran, a parallel experiment tried to stay inside securities law.

Overstock, through a subsidiary that became tZERO, spent 2015 proving it could move a cryptosecurity and a cryptobond on a ledger it controlled. In December 2016 it closed a Series A preferred raise in which a slice of the money — about $1.9 million of a $10.9 million round — arrived through shares issued on that ledger. It was not the modern public-equity wrapper market. It was the first time a listed American retailer had let subscribers take company paper across a blockchain rail in a regulated raise.

Patrick Byrne, then Overstock’s chief executive, spent 2017 saying the quiet part at conferences. He wanted equity tokens that were securities, and a venue allowed to trade them. “I think it’s a historic event. We’re opening a new type of capital market,” he told the press when tZERO’s alternative trading system came into view. The ATS was the point. Without a regulated place to meet, a blockchain share is a private database with extra steps.

The lesson from that stretch is not that tZERO “won.” The lesson is that the first serious issuers already knew the bottleneck was the venue and the transfer agent, not the hash function.

2017–2020: security-token plumbing

Polymath (2017) tried to make issuance a product with ST-20 and later ERC-1400-style compliance hooks. Harbor designed a regulated-token pattern that asked a contract for permission before a transfer cleared. TokenSoft and others pushed ERC-1404 so a restriction could return a reason code instead of a silent revert. Securitize went after the unglamorous licenses: broker-dealer, transfer agent, later an ATS — the stack you need if the token is going to be the record rather than a pointer to a record.

Those years were slow on purpose. The ICO market had taught regulators to read a white paper as a prospectus. Builders who wanted to tokenize a real fund had to look boring. Boring is how transfer agents get hired.

2020–2024: the fund becomes the beachhead

Public-equity wrappers remained legally awkward in the United States. Cash-like funds were less awkward. Tokenized treasuries and money-market vehicles found buyers who already understood NAV. BlackRock’s BUIDL, Franklin’s onchain share classes, Ondo’s cash products, and a dozen cousins turned “RWA” into a ticker people could quote without explaining a building.

This was not a sidetrack. It was the first time the reserve layer was something a money-market accountant already knew how to count. Tokenization stopped being a demo when the underlying had a daily NAV and a custodian who would put their name on the letter.

2025–2026: the pipes move

Two facts changed the tone of the desk.

First, tokenized public equities stopped being a curiosity. Multiple issuers put hundreds of names on-chain as wrappers — economic exposure to listed stocks and ETFs, backed in custody, traded when the cash session was closed. Ondo Global Markets and xStocks became names a directory could sort. The legal object was still often a wrapper, not the listed share itself. The tape, however, was real enough to have a premium and a discount.

Second, the depositary of the American securities market asked for, and received, permission to try the other object. In December 2025 the SEC staff gave The Depository Trust Company a no-action letter to run a three-year tokenization service on a defined set of highly liquid DTC-custodied assets — Russell 1000 names, major-index ETFs, U.S. Treasuries — with the token carrying the same entitlements as the traditional position. DTCC ran live production exercises in July 2026 with large banks and managers. A commercial window was aimed at the second half of 2026. Nasdaq and NYSE wrote rules so a tokenized entitlement could meet its traditional twin on the same book when CUSIP, symbol, and rights matched.

Frank La Salla, DTCC’s chief executive, put the condition in public: the benefits only arrive “if market infrastructure provides a robust foundation.” That is a transfer-agent sentence. It is also the sentence 2012 did not have a building for.

What the road actually taught

Every decade reused the same three failures.

People issued a token before they could say what the holder owned.

People printed a price before they had a second price to check it against.

People advertised 24/7 ownership of a thing that still lived in a market with hours.

The present tense of tokenization is those three failures getting numbered. That is why this site exists as an observatory rather than as a canvas. History did not need another whiteboard. It needed a place that would keep the claim, the reserve, and the print on the same page.