The Evolution of Tokenization
In March 2012, an Israeli developer, Yoni Assia proposed colored coins, a way to mark specific bitcoins as representing something else, a share, a bond, a deed.
Unfortunately, the idea never left the whiteboard stage in any serious manner, but it named a problem every builder after him spent the next fourteen years solving. Which is how to take something real and illiquid and make it move like cash.
Fourteen years later, tokenized assets outside stablecoins carry $38.29 billion in freely tradable value and $369.44 billion committed, held by 1.79 million people.
This article entails a step by step worked out stages it took to reach this height in tokenization, and why it took this lon
What Is Tokenization?
A tokenized asset is a claim on something real, a treasury bill, a gold bar, a loan, a share of a hotel, represented as a token that lives on a blockchain.
The token itself is worth nothing on its own. Somewhere off-chain, a custodian, a fund administrator, or a trust holds the actual asset and agrees to honor the token as a claim against it, the same relationship a stock certificate or a bearer bond has always had to the thing it represents.
What the blockchain adds is narrow but real: it records who owns the claim and lets that ownership change hands in seconds, at any hour, without a broker sitting in the middle of the trade. Strip away the technology and this is an old idea. What's new is the speed it moves at, who gets to hold it, and how it gets used.
Early Builders and Early Struggles
Long before anyone believed tokenization could scale, a handful of builders had to prove it could work at all. Lucas Vogelsang and Martin Quensel built @Centrifuge in 2017 and spent its first four years on Tinlake, a system for financing real-world invoices and assets through tranched, revolving pools.
In mid-2021, Centrifuge ran the first RWA pool integrated with MakerDAO (now @SkyEcosystem), minting the first DAI ever backed by a real-world asset. It worked.
The same pattern showed up in real estate. In October 2018, a hospitality firm called 'Elevated Returns' closed an $18 million raise for the St. Regis Aspen Resort, selling 18.9% of a hotel valued at $224 million as one-dollar tokens to accredited investors through Templum Markets, the first major commercial real estate deal done on a blockchain.
A few months later, 31.6% of a $5.6 million Andy Warhol painting got tokenized and auctioned the same way.
Both deals worked exactly as planned and then they went nowhere, because nowhere is where a token goes once it's minted and there's no one left to trade it with.
@Paxos launched PAXG in September 2019, the first gold-backed token approved by New York's financial regulator. @Tether launched XAUT the same year, competing on Swiss vault storage against Paxos's Brink's London reserves. Neither product broke through to institutional attention at the time. Both are still running today, seven years later, carrying $1.9 billion and $2.7 billion respectively, proof that tokenization worked in commodities long before anyone believed it would work in treasuries.
Regulatory clarity lagged well behind. The SEC's 2017 DAO Report applied the Howey test to token sales for the first time, ruling that a digital token could be a security depending on how it was sold, not what it was called.
That single finding created the entire security token offering category and its accredited-investors-only cage. Issuers spent the next four years launching STOs into markets that legally could not include most of the people who might have traded them.
Thin books, forced holding periods, and unclear custody law killed most of the wave by 2021, the same year Augur peaked at 265 users and then collapsed to 37.
The Problem Faced
Assets kept getting tokenized but most of them just sat still. Today, 88% of all RWA market value concentrates in just 62 assets, and five products alone, @Figure's HELOC fund, @Circle's USYC, @Tether Gold, @BlackRock's BUIDL, and a fund by @Justokenglobal JMWH, account for roughly half the entire market. Per Forbes.
Everything else sits minted and mainly forgotten. Even inside that concentrated 88%, less than 10% of tokenized value is doing anything, deployed as DeFi collateral, used in lending, put to work.
The reason wasn't lack of demand. Capital wanted exposure to tokenized treasuries and gold well before 2024. The reason was that no institution would put its name on an asset unless a regulated entity it trusted was standing behind custody, transfer, and compliance. Nobody credible had proven that model at a measurable scale until @BlackRock did.
How Tokenization Finally Broke Through
Tokenization didn't need a better product. It needed one institution willing to put a name on the line that other institutions already trusted.
@BlackRock supplied that in March 2024 with BUIDL, tokenized and administered by Securitize, a compliance platform Carlos Domingo and Jamie Finn had built back in 2017.
@carlosdomingo had already run one of the first tokenized-security funds, SPiCE VC, which is part of why @Securitize existed to be chosen when BlackRock came looking for a partner.
BUIDL now holds $2.8 billion across 10 networks, custodied by BNY Mellon and audited by PwC, paying a 7-day yield of 3.42%. The effect was that BlackRock's name on a blockchain made every other asset manager's board take the category seriously in a way four years of STOs never managed to do.
Note: Franklin Templeton BENJI launched on @StellarOrg in April 2021, and became the first US-registered mutual fund to use a public blockchain but only as its official system of record, three years before BUIDL existed. The fund now runs on multiple chains, with: the EVM version, iBENJI, holds $1.72 billion on BNB Chain and Ethereum, while the original Stellar-native token holds $712.5 million.
Credit and Yield Infrastructure
Once the top of the stack had a credible wrapper, the middle of it needed somewhere for lenders and borrowers to meet. Sid Powell and Joe Flanagan built that in 2019 as @Maplefinance. It nearly ended the company.
In December 2022, a borrower called Orthogonal Trading defaulted on $36 million in loans after concealing its exposure to FTX's collapse, wiping out roughly 30% of all active loans on the platform in a single week.
Powell called himself "shocked and disappointed" and rebuilt around the lesson: diversify who's allowed to borrow, and never again take a delegate's word for its own exposure.
Maple has since originated more than $20 billion in loans and now carries $1.9 billion combined across its syrupUSDC and syrupUSDT products, the largest platform in tokenized private credit by a wide margin.
@Ondo came two years later, built by Nathan Allman, a Goldman Sachs digital-assets veteran, around OUSG, the first peer-to-peer transferable tokenized treasury token, and USDY, the first permissionless yieldcoin. USDY and OUSG now carry $2.145 billion and $449 million respectively, and Ondo has become the largest on-chain distributor of BUIDL itself. Allman died unexpectedly in May 2026.
The Oracle And Data Layer
None of this works without a way to get an off-chain price or NAV onto a blockchain in a form a smart contract can trust.
That job is majorly split across four networks:
1. @RedStone_defi: Securitize named RedStone its primary oracle partner in March 2025, and RedStone now delivers the daily NAV feed for BUIDL itself, along with Apollo's ACRED, VanEck's VBILL, and Hamilton Lane's SCOPE, the entire family of funds Securitize tokenizes.
Before that integration, Securitize's tokens had no live pricing at all, a fund share existed on-chain but couldn't be priced by a lending protocol in real time. RedStone's NAV feed is what let ACRED start earning yield on @Morpho and VBILL post as collateral on Euler. It now secures roughly $6 billion in on-chain value across 110 blockchains, almost all of it institutional fund pricing rather than crypto-native tokens.
2. @Pythnetwork took the opposite end of the market, speed over NAV complexity. It runs a pull model built for assets that trade constantly and need sub-second updates, equities, FX, commodities, now covering more than 750 US equity feeds plus Treasury rate feeds across several tenors.
@Ondo picked Pyth to price its own USDY yieldcoin, live across 65 blockchains, a direct answer to the question of who prices the token when Ondo's own treasury products need a feed independent of Securitize's stack.
3. @DIAdata_org built for verifiability specifically, source-to-contract transparency. Its xReal suite covers more than 100 RWA price points, stocks, ETFs, FX, bond yields, and it's the oracle Stellar and Ripple leaned on for their own RWA partnerships, an intentional bet that institutions moving slowly through compliance would rather audit a feed's full methodology than trust a name.
4. @Chainlink kept the job none of the other three touch, moving data and assets between chains that don't trust each other, not pricing a single fund. Its CCIP distributes DTCC Smart NAV data and underpins the Swift and UBS pilots run through Singapore's Project Guardian, which launched in May 2022 and ran its first live tokenized fund pilot with UBS in October 2023.
CCIP moved $4.9 billion in quarterly volume last quarter, up 353% year over year, and the network's total value secured now sits at $110 billion. Mantle alone moved over $2.5 billion in tokens onto CCIP this year. The DTCC is building Chainlink's infrastructure directly into its Collateral AppChain, targeting a Q4 2026 launch.
Distribution
Building the wrapper and selling it turned out to be two different jobs.
1. @Mantlenetwork began inside BitDAO, a treasury DAO backed heavily by the exchange Bybit that at its peak rivaled the Ethereum Foundation's own reserves.
When BitDAO merged into Mantle in 2023, Mantle inherited that treasury whole, close to $300 million in stablecoins and 270,000 ETH at the time, now grown to $2.4 billion. Few competitors get to enter the distribution fight already holding the capital most startups spend years raising. Mantle now carries over 155 tokenized equities and more than $1 billion in DeFi liquidity, and on August 6 it extended that reach onto Solana through a CCIP-secured portal, the first time the network has pushed its RWA infrastructure onto a chain it doesn't own.
2. @Coinbase and @Binance sit on the collateral and aggregation side. BUIDL now trades on BNB Chain and is accepted directly as collateral on Binance.
3 @MEXC has taken the broadest approach of any single exchange, listing over 105 Ondo tokenized stock pairs alongside PAXG, XAUT, and core RWA infrastructure tokens, adding five more in August alone spanning AI infrastructure and rare earth names. It held 27% market share in tokenized gold trading in the first quarter of 2026, second globally. This month MEXC upgraded RealStocks, its brokerage-partnered product covering more than 7,000 US-listed stocks and ETFs, adding features designed to make token ownership mirror real shareholder rights.
4. The instrument growing fastest sits outside any single exchange's spot listings. RWA perpetual futures on @Hyperliquidx and Binance combined moved $61.7 billion in a single week at the end of July, equal to 99.2% of those platforms' Bitcoin perpetual volume in the same window. Tokenized equities made up 58% of that. On Hyperliquid, RWA perpetual volume now exceeds every other category on the exchange combined.
Regulation Catching Up
Regulation is moving on four fronts here.
1. The GENIUS Act, the first federal stablecoin framework in US history, was signed into law on July 18, 2025, settled and in force.
2. The CLARITY Act, which would set market structure rules for the rest of the industry, passed the House a day earlier and cleared Senate Banking Committee 15-9 in May 2026. It stalled after that on an ethics dispute, Democrats pushing for conflict-of-interest language covering federal officials' own digital asset holdings, Republicans pushing to move without it.
The Senate adjourned for its August recess on August 8 with no final vote, but Majority Leader Thune filed cloture on the motion to proceed before leaving, setting up a procedural vote for September 15, the day after the chamber returns.
That vote isn't passage, it's the motion just to start floor debate, and Senate staff have told reporters the ethics language remains unresolved. If September slips too, the midterm election calendar leaves little floor time left this year.
3. The SEC took a separate route. In March 2026, the SEC and CFTC jointly published a token taxonomy sorting digital assets into five categories, with only one, Digital Securities, remaining under full SEC jurisdiction.
Chair Paul Atkins built on that in April, announcing at the Economic Club of Washington that the agency was "on the verge of releasing" an Innovation Exemption, a sandbox letting firms trade tokenized securities on-chain for twelve to thirty-six months without full registration. Bloomberg reported the release as imminent in mid-May. By May 22, the SEC had pulled back, absorbing pushback from stock-exchange officials over investor protections. The most recent confirmed reporting, from mid-June, still had the exemption unreleased.
4. Nasdaq, the NYSE, and the DTCC skipped the wait entirely. The SEC had already approved Nasdaq's rule change to let Russell 1000 stocks and index ETFs trade in tokenized form through existing market plumbing in March 2026, followed by a similar approval for the NYSE in April.
The DTCC, which custodies roughly $114 trillion in assets, announced on May 4 that it would run a production pilot in July 2026 with more than fifty institutions including BlackRock, JPMorgan, and Goldman Sachs, covering the Russell 1000, major-index ETFs, and US Treasuries, with a fuller launch targeted for October. That rail keeps every traditional shareholder right intact, voting, dividends, the master securityholder file staying exactly where it's always been. The token is a wrapper around a settlement record that hasn't changed.
The SEC's stalled exemption would have opened a second, looser rail alongside it, tokens issued without the underlying company's consent, carrying economic exposure but not shareholder rights. Whether that second rail opens at all, and when, is still an open question. What isn't open is whether regulated tokenized trading of major US equities is coming, the DTCC pilot alone answers that.
Has the Problem Been Solved?
Score the diagnosis from earlier against where things stand today. Liquidity fragmentation is partially closed or at least in the process. Treasuries and private credit run deep, $16.2 billion and $7.3 billion respectively, spread across 87 and 2,543 assets.
Everything outside those two categories is still thin (few categories have more than a billion dollars value though). Custody trust is solved for the BlackRock and Franklin Templeton tier and unsolved below it. Retail access remains closed to roughly 97% of tokenized assets, still gated behind accreditation requirements that trace directly back to the SEC's 2017 Howey ruling.
That's the optimistic read. A July 2026 BeInCrypto Intelligence report, built on RWA.xyz data across more than 7,000 tokenized products, found a harder number underneath it: 56% of tokenized assets worth more than $100,000, roughly $32.9 billion, recorded zero weekly on-chain transfers. Existing on a blockchain and moving on a blockchain turned out to be two separate achievements, and right now the market has only managed the first one.
Some of which is by design, a tokenized treasury bought to hold for yield isn't supposed to trade daily any more than a savings bond is. But it means the honest scorecard isn't "partially closed," it's closed at the top and barely open everywhere beneath it, five products holding roughly half the entire market's value while the other 6,000-plus mostly sit still.
What's Next
Every serious forecast for this market disagrees on the number and agrees on the direction.
McKinsey's base case puts tokenized assets at $2 to $4 trillion by 2030. Ark Invest says $11 trillion. BCG and Ripple say $9.4 trillion by 2030, rising to $18.9 trillion by 2033. Standard Chartered says more than $30 trillion by 2034.
The distance between those numbers isn't really a disagreement about whether tokenization grows, every one of them implies roughly 100x growth from today's market. The disagreement is about definitions, whether stablecoins count, whether deposits count, how wide a net "tokenization" is allowed to cast.
Institutions agree with that forecast and hedge against it at the same time. Coinbase and EY-Parthenon's January 2026 survey of 351 institutional decision-makers found 73% planning to increase digital asset allocations this year, with regulatory clarity cited as the top driver by 65% of respondents. The same survey found regulatory clarity is also the top concern, cited by 66%, the exact thing pulling allocations forward is the thing everyone says isn't settled yet.
A separate EY survey found institutions aren't waiting for that to resolve before moving: 83% of institutional investors now plan to hold tokenized bonds by the end of 2026, up from just 33% two years earlier.
That shift is already visible in the plumbing, not just the surveys. BUIDL became tradable directly through @UniswapX in February 2026. The following month, it plugged into Chronicle's proof system, letting anyone verify BlackRock's underlying treasury holdings in real time straight from BNY Mellon's custody records. Sky's Spark Liquidity Layer now automatically shifts $1.5 billion across BUIDL, Anemoy, and @Superstateinc based on which one is paying the best yield that week, a job a treasury desk used to do by hand.
DWF Labs founder Andrei Grachev argued tokenized equities have already rewired trader behavior, letting crypto traders rotate into stocks without leaving the venue they started in, and predicted on-chain shares and commodities get "much bigger" this year. @Falconfinance's Artem Tolkachev framed the mechanism plainly: composability and exit design are what actually connect a real-world asset to crypto liquidity.
The industry is splitting over how to get there. One camp is building ownership-first, permissioned rails, tokens that stay inside a compliance perimeter the whole time.
The other is building composability-first, wrapping a permissioned asset like BUIDL so it can circulate through permissionless DeFi anyway.
Call it the difference between RWA 1.0 and RWA 2.0 again, except now it's not a theory, it's two competing product roadmaps, and @Circle's USYC overtaking BlackRock's BUIDL as the largest tokenized treasury fund earlier this year is evidence for which camp is currently winning on distribution, if not yet on brand.
The first version held an asset and represented it, buy, hold, redeem, nothing more, the same job the St. Regis Aspen tokens and Maple's earliest lending pools were built to do. The second version puts that same asset to work. Tokenized real-world assets deposited across DeFi platforms hit $7.4 billion in the second quarter of 2026, more than three times the year-ago level, with yield-bearing stablecoins and tokenized treasuries making up the largest share.
Maple's syrupUSDT already runs live through @Aave on Mantle, production infrastructure moving real capital today.
The assets that mattered in 2018 were the ones nobody could trade. The assets that will matter next are the ones nobody has to choose between holding and using.
This article although comprehensive does not necessarily complete and backdate every incident that happened in the history of tokenization. Do not treat as a reason to invest. Every image and information here is public and searchable.