Tokenized real-world assets are the fastest-growing category in crypto and there is no agreed definition of what is being counted. The spread across published figures comes from four real methodological choices: as-of date, stablecoin inclusion, distributed versus represented assets, and asset-class scope. This piece works through a live example — USYC and BUIDL, two leaders with two denominators — and gives three questions that survive contact with any market-size claim.
Here are four published figures for the size of the tokenized real-world asset market, all from 2026: $26 billion, $31 billion, $38.17 billion, and, from a mid-year survey, a range of roughly $30 to $36 billion. Here are four published figures for the tokenized US Treasury segment inside it: $6.8 billion, roughly $12.88 billion, roughly $15 billion, and $16.21 billion.
None of these is a mistake. Every one of them is defensible. And that is a more interesting problem than any of the individual numbers, because tokenized RWAs are the fastest-growing category in crypto and there is no agreed definition of what is being counted.
Set them out with their dates, because the dates do a surprising amount of the reconciling work:
Notice something odd in that last line. The mid-May figure is lower than the early-April figure. In a category compounding at more than 50% a month, that ordering should be impossible — which tells you immediately that the two numbers are not measuring the same thing. Neither is wrong. They are counting different sets of products.
Start with the part that resolves easily. A category growing 56% month over month is a materially different size six weeks later. A figure published in April and a figure published in August can differ by a factor of two purely on elapsed time, with identical methodology on both sides.
This means a large share of the apparent disagreement in circulation is not disagreement at all. It is stale numbers being quoted without their as-of dates, in articles published later, cited by other articles later still. An RWA market size figure without a date attached is not a claim about the world; it is a claim about when someone last looked.
But timing does not resolve everything, and the residue is where the actual methodology lives.
The clearest illustration available is a pair of claims that look flatly contradictory and are not.
Circle's USYC is described as the leader in tokenized Treasury products, at $3.00 billion. BlackRock's BUIDL is separately and routinely described as the largest tokenized Treasury fund, at $2.68 billion. Both statements appear in credible coverage. Both are true.
The resolution is a denominator. On rwa.xyz's 9 August reading, the US Treasury debt segment as a whole stood at $16.21 billion, spread across 87 distinct products held by 63,010 unique addresses. Within that segment, USYC leads individual Treasury-backed tokens at $3.00 billion, with BUIDL second at $2.68 billion and Franklin Templeton's BENJI also among the largest. When a source calls BUIDL the largest, it is generally ranking within a narrower category — tokenized money market funds, or funds structured a particular way — rather than across every Treasury-backed token.
Two true rankings, two different denominators, one apparent contradiction. This is the entire tokenized-finance measurement problem in miniature.
Note also what that segment structure implies about concentration. If Treasury debt is $16.21 billion of a $38.17 billion total, a single asset class is roughly 42% of the market, and one analysis has argued the concentration is considerably higher on a wider definition of the segment. A market this concentrated in one asset class is not really a diversified tokenization market yet. It is a tokenized money market fund industry with some other things attached.
Beyond timing and denominators, at least four choices explain most of the remaining spread. This is a synthesis rather than any single provider's published taxonomy, and it should be read that way.
Stablecoins, in or out. This is the largest single fork and it changes the answer by an order of magnitude. Fiat-backed stablecoins are, by any literal reading, tokenized claims on real-world assets — mostly short-dated Treasuries and bank deposits. Most RWA trackers exclude them, on the reasoning that they are a payments instrument rather than an investment product and that including them would drown every other category. But that is a judgement call, not a fact, and a source that includes them is not miscounting.
Distributed versus represented. Does a position count when the token exists on a public chain and can be transferred, or also when an institution holds a tokenized record on a permissioned ledger that never leaves its platform? A great deal of institutional tokenization is the second kind. Counting it produces a much larger and much less liquid market.
Asset-class scope. Private credit, Treasuries, commodities, equities, real estate, and increasingly institutional funds. Trackers include different subsets. Private credit in particular is large, and its inclusion or exclusion moves the total substantially.
What is being measured. Total value locked, assets under management, notional issued, and outstanding value are not the same quantity, and coverage moves between them fluidly. A product that has issued $1 billion of tokens over its life and currently has $300 million outstanding can support either figure.
It would be easy to file all of this as pedantry — data providers disagree, journalists round, everyone moves on. It is not, for three specific reasons.
These numbers enter allocation decisions. Institutional allocators size positions relative to market size, because liquidity and exit capacity scale with it. A category that might be $16 billion or might be $38 billion depending on definitions cannot be prudently sized. The uncertainty is not in the forecast; it is in the present.
They enter regulatory comment letters. Arguments about whether a market is systemically relevant, whether it warrants a bespoke regime, and how urgent the rulemaking is, are all made with market-size figures. Regulators are being handed numbers that vary by more than a factor of two, frequently without as-of dates.
And they enter every pitch deck downstream. A growth chart is more persuasive than a growth argument, and a chart built by splicing figures with different methodologies produces a curve steeper than anything that actually happened — not through deception, usually, but through a series of individually reasonable citations.
None of this requires distrusting the data providers, who are mostly transparent about their methodologies for anyone who reads them. It requires reading the number as a measurement rather than a fact. Three questions survive contact with any market-size claim in this category:
One last note in the spirit of the argument. The $38.17 billion figure above is rwa.xyz's reading on 9 August 2026, and it was the freshest confirmed number available when this piece was written. If the dashboard shows something different by the time you read it, that is not a correction to this article. It is another data point for its thesis.
It depends on the date and the definition, which is the honest answer rather than an evasive one. Published 2026 figures include $26 billion, $31 billion, a mid-year range of roughly $30–36 billion, and $38.17 billion as recorded by rwa.xyz on 9 August 2026. The spread comes from as-of dates in a category compounding more than 50% month over month, from whether stablecoins are counted, from whether permissioned-ledger positions are included, and from which asset classes are in scope. Any figure quoted without a date and a methodology should be treated as approximate.
DrafterDaily reads crypto market data the way an analyst would — methodology first, headline second. One clear read a day.
Get the daily briefA race condition between the display logic and the signing buffer breaks the one promise a hardware wallet exists to make — that what the screen shows is what the key signs — without breaking any cryptography at all.
Coverage fixed on the $75 million exemption. The consequential provision is a conditional safe harbor that would let a crypto asset stop being subject to an investment contract — and a preemption clause almost nobody has read.
Staking participation hit a record 33.98% of ETH supply this month. Six researchers, including the Ethereum Foundation's Justin Drake, are proposing to burn validator rewards to zero if it reaches 50% — and their argument is not about yield.