Tokenized assets are busier than the data shows

51 min ago5 min readDepending on who you ask, the share of tokenized real-world assets actually being used in DeFi is under 1%, or 7%, or 11.7%, or close to 20%. Every one of those numbers was published this year. Every one of them is defensible. None of them measures the same thing.The low one gets most of the airtime: of the roughly $51 billion in tokenized real-world assets on public blockchains, this estimate suggests only a single-digit percentage actually does anything. It gets repeated as proof that onchain finance is still a toy. All this tokenized "value," and almost none of it working, at least publicly.The critique isn't baseless. An asset that moves onchain, pays fees to get there, and gains no productivity in return is a worse product than the one it copied. But the number being used to prove that critique is close to meaningless. And not because it's too low. It's that both halves of the fraction are theater.Where the numbers come fromThe sub-1% figure measures three tokenized money market funds, not a market: BlackRock's BUIDL, Circle's USYC and Franklin Templeton's iBENJI hold $7.2 billion between them and have roughly $50 million deployed. Widen the basket, and it becomes 11.7% on DeFiLlama's data, or about 19% using CoinShares' $7.4 billion Q2 count against RWA.xyz's $38 billion total — same market, same quarter, a 20x spread, because nobody has agreed what the question is.The denominator was never going to moveAccording to Bernstein’s research, about 47% of the $51 billion in tokenized real-world assets onchain is private credit. Private credit doesn't move much in traditional finance either; tokenizing it changes neither its redemption calendar nor its holder base. Counting it in the denominator of a composability metric is a category error, not a disappointment.Then there's the trophy tier. The early flagships that generated the "RWAs are here" headlines BUIDL, Apollo's ACRED and others like them — shipped wrapped in enough transfer restrictions and whitelist gates that they couldn't function as collateral even if someone wanted them to. Someone involved in one of those products told me flatly that it was "a terrible product."Three reasons utilization looks low, and only one is a failureRestricted by design. Some assets genuinely cannot be used: whitelists, transfer agents, accreditation gates, no permissionless path to onchain lending markets. This is the failure everyone assumes is the whole story.Parked by intent. Some assets can be used and simply aren't, because of who holds them and why. A foundation holding BUIDL on its balance sheet to get BlackRock to deploy on its chain is buying a headline, not a yield strategy. More broadly, BUIDL is what you buy when you want BlackRock's name on the position; its competitors, at a similar yield, only get minted when someone wants something BUIDL doesn't have — more DeFi composability, or more DeFi subsidies. So the challenger's holders are selected for wanting to use the thing. Comparing utilization across those two cohorts tells you why people bought, not what the token can do.Used invisibly. Some of it is working where the metric can't see. Binance and Franklin Templeton have BENJI functioning as off-exchange collateral. BUIDL is accepted as derivatives margin. Kraken takes tokenized equities as collateral for leveraged trades. In each case, the asset does genuine collateral work while sitting with a custodian rather than in a smart contract, hard for DeFi trackers that count value locked in contracts they can read.Strip out what was never mobile, correct for who's holding and why, add back what works off-contract, and utilization of the assets actually built to move looks far closer to 20% than to 8%, roughly where CoinShares' independent count already lands, before any adjustment at all.The real bottleneck is settlementThe more interesting problem is that even when a tokenized asset is usable, the thing that makes onchain collateral valuable remains hard to do with it.Looping is the clearest case: deposit collateral, borrow against it, buy more, repeat. For crypto-native assets, it's nearly frictionless — yield-bearing stablecoins and liquid staking tokens mint instantly and freely, flash loans build and unwind a position atomically, curators are comfortable with high LTVs, and the liquidity is deep. That stack is battle-tested.Tokenized real-world assets break it, because they don't settle instantly. They settle T+1, T+2, or on a redemption calendar, so the loop has to run sequentially, once per cycle. Building a 4x position on a T+1 asset takes around 8 loops, so 8 days in, 8 days out.I saw this firsthand. I worked on bringing Apollo's ACRED onchain as collateral on Polygon PoS, alongside Securitize, Gauntlet and a Morpho-powered vault. It worked, to an extent. Some looped it, and it proved the thing could be done at all. It also proved what still had to be solved before institutional flows would follow — chiefly on-demand liquidity for atomic redemption. Redemptions on a fund like that come quarterly. Loop it 4x, and unwinding completely can take a year. The redemption path simply didn't make sense, and that isn't a smart contract problem or a demand problem. It's a duration mismatch between an instrument built for quarterly liquidity and a market that clears in seconds.That's the actual reason RWA utilization is low, and it's nowhere in the debate.It's also what newer entrants appear to be solving. Protocols like 3F, built on Morpho on the Ethereum mainnet, replace sequential looping with an onchain auction in which specialists front the full capital for the target leverage in one shot. Twenty settlement cycles collapse into one.What actually changes the numberThe regulatory sequence is also finally running in the right order. The GENIUS Act mattered less for tokenized securities directly than for what it unlocked upstream: credible stablecoin rails, the prerequisite for the institutions that would hold these assets to be onchain at all. You cannot have a collateral market without settlement money. CLARITY, which would do the direct work on market structure, has stalled in the Senate — which is why the administrative path matters more near-term. Rules being written at the CFTC and SEC quietly remove the compliance anxiety that pushed issuers to over-restrict these assets in the first place. As that clears, the gates come down.Count the collateral, not the press releasesA tokenized Treasury fund that can't leave a whitelist is a filing cabinet on a blockchain. A tokenized asset that can be posted as collateral, borrowed against, levered and unwound in a reasonable window is the actual product.The number that matters was never how much got tokenized. It's how much got used — and how quickly someone could get back out. When the industry starts publishing that one, the theater is over.Note: The views expressed in this column are those of the author and do not necessarily reflect those of CoinDesk, Inc. or its owners and affiliates.12345678910Anvil: The Missing Collateral LayerAnvil: The Missing Collateral LayerAnvil is a shared on-chain collateral layer built on a programmable letter of credit: reserve assets as a guarantee -no loan, no interest, keep custody & yield.Jul 29, 2026Anvil is a shared on-chain collateral layer built on a programmable letter of credit: reserve assets as a guarantee -no loan, no interest, keep custody & yield.Why it matters:Anvil is a shared on-chain collateral layer built on a programmable letter of credit: reserve assets as a guarantee -no loan, no interest, keep custody & yield.View Full Report