
Data as of 10 May 2026 unless noted; event references and yield snapshots elsewhere as dated inline.
Tokenized real-world assets surpassed $27.5B in May, roughly triple the figure a year earlier, but only $1.7B of that AUM is actively used in DeFi as collateral, in leverage loops, or in lending markets. The remaining $25.8B sits on institutional balance sheets, backs stablecoin reserves, or rests in DAO treasuries: tokenized, but not composable.
Treasuries dominate AUM; credit dominates usage
Tokenized treasuries account for the largest share of the market at $14.7B AUM, doubled year-on-year, yet only about 0.5% is deployed in DeFi — almost all on Aave Horizon, a permissioned Aave V3 market that requires institutional whitelisting on the collateral side. Tokenized treasuries are securities under US law, so KYC and whitelisting are written into the wrapper. Even when treasuries enter DeFi, they do so inside institutional perimeters. Outside DeFi, tokenized treasuries' main emerging use case is as backing for stablecoins. Ethena's USDtb holds ~90% of its reserves in BlackRock's BUIDL. Frax's frxUSD and Sonic Labs' USSD are collateralized by BUIDL, Superstate USTB, and WisdomTree WTGXX. OpenEden's USDO is backed by its own tokenized T-bills. These are still small relative to USDC ($76.5B) and USDT ($189B), both backed today by off-chain Treasuries, but if either migrated even a fraction of those reserves to tokenized wrappers, the $14.7B tokenized treasury AUM figure grow considerably.
Credit is smaller and more active: $3.2B AUM, growing 7x year-on-year, and accounting for 62% of all RWAs deployed in DeFi. The roster includes Maple's syrup tokens, Apollo's ACRED via Securitize, FalconX's funds on Morpho, Midas' mF-ONE, Hastra's PRIME (recently onboarded to Morpho), and a set of Centrifuge-issued credit pools. Reinsurance remains a marginal share by AUM at $329M, but 78% of it is active onchain. Treasuries, at roughly 4.5x credit's AUM and over 40x reinsurance's, sit at 0.5% deployed. Size and DeFi usage do not track.
The mechanism, in our reading, runs through yield. Tokenized treasuries pay around 3.5%; the credit, reinsurance, and active-strategy assets that have accumulated real DeFi deployment have so far been the higher-yielding ones — syrupUSDC at 4.7% (syrupUSDT at 4.2%), Hastra's PRIME around 6%, reUSD at 6%, FalconX's funds at 8%+, Midas mF-ONE near 9%, and OnRe's ONyc at 11%. To date, no permissioned-only or treasury-yield tokenized asset has accumulated material DeFi deployment, while every category that has shares two features: an open wrapper and a yield above the treasury baseline. We call the combined condition permissionless yield. Treasuries function as the reserve layer; credit, reinsurance, and alternative funds function as the composable yield layer.
Maple syrup and the permissionless playbook
Maple's syrup tokens are the clearest example of both conditions holding. syrupUSDC and syrupUSDT are yield-bearing tokens that earn from Maple's overcollateralized loans to crypto market makers and trading firms — freely transferable, KYC-free at the token level, ERC-4626-compatible (the standard vault-token format lending markets can list without bespoke integration), and cross-chain. Through 2024 and most of 2025 they paid in the 7–15% range, enough to anchor leveraged strategies across multiple venues. Today, with crypto borrow demand softer and stablecoin yields compressed market-wide, syrupUSDC sits closer to 4.7% and syrupUSDT to 4.2%. Of $1.8B in syrup market capitalization, $522M is deployed across five venues — Aave, Morpho, Kamino, Jupiter Lend, Fluid — making a single asset family responsible for nearly 30% of all RWAs in DeFi. On Aave, 99% of syrupUSDT on Plasma and 99% of syrupUSDC on Base are pledged as collateral, with curators such as Gauntlet running leverage loops on Morpho up to 5x. The asset went wherever the loops paid.
USCC, USTB, and the Horizon rotation
Inside permissioned venues, yield alone moves the collateral mix. Since launch, USCC — Superstate's Crypto Carry Fund, running long/short futures on BTC, ETH, SOL, and XRP plus staking yield, structurally closer to a hedge fund than a treasury product — anchored the Horizon collateral base, paying over 10% at inception and at times accounting for more than 90% of collateral on the venue. Institutions posted USCC, borrowed RLUSD or USDC at 80% LTV, and kept the carry exposure: a familiar leveraged trade with a tokenized fund as collateral.
That equilibrium has shifted as USCC's APY compressed from roughly 15% to under 5%. USTB, Superstate's tokenized treasury product — same issuer but a product with a structurally different risk profile — has been gaining share inside the same venue. With the carry spread no longer compensating for strategy risk, capital rotated down the risk curve to a more conservative product on the same rails. The same compression is visible outside Horizon: syrup at 4.6% against a 3.5% treasury baseline is a different proposition than syrup at 9% against 4%.
Stress-tested on 18 April
On 18 April, a KelpDAO hack drove an unwind cascade through DeFi credit markets, and syrup-on-venues fell 70% from $1.76B to $522M in days. Maple processed over $800M in redemptions at par, with no defaults, no paused mints, no gating. The market unwound and prices held. The wrapper design that lets credit tokens flow into DeFi without friction also lets them flow out without breaking.
Whether syrup climbs back is the harder question. The $522M still deployed is a fraction of the pre-unwind level, and the rebuild would have to happen in a yield environment quite different from the one that drove the original accumulation. Whether existing deployment is stable, and whether new looping demand returns at a 100–150 basis-point spread over treasuries rather than a 500–1,000 basis-point one, is the most important near-term test for the composable yield layer.
Our reading
The two-stack picture is the right way to parse RWA growth. Treasuries are working as designed — reserves, not collateral — and the composable yield layer is where onchain adoption is actually accumulating. The most important variables to watch are wrapper innovations that could move treasuries and other asset classes into the composable layer, and yield compression that could push credit assets out of it. The next twelve months will resolve more of that question than the last twelve did.
We see two specific moves already underway. First, wrappers are evolving. Grove's Basin facility, launched mid-May, fronts up to $1B/day in stablecoins against approved redemptions of BUIDL and JTRSY, letting holders exit instantly while the underlying funds settle on traditional rails. Still institutional and gated, but it's the same wrapper-not-underlying move that turned syrup into collateral. Compress a treasury's settlement gap from days to seconds, and the asset begins to look usable as collateral.
Second, new categories are arriving with composability designed in. Tokenized public equities are the clearest case. Ondo Global Markets crossed $1B AUM less than eight months after launch, and together with Backed Finance's xStocks accounts for roughly $1.4B globally. Ondo has already announced Ondo Perps, where holders will be able to natively post Global Markets stock tokens as collateral. Ondo GM and xStocks work the same way structurally: shares held 1:1 with regulated broker-dealers, token holders get economic exposure rather than shareholder rights, and dividends are reinvested net of withholding tax — xStocks via an onchain rebase, Ondo via reinvestment into the underlying SPV. Neither product is available to US users today, though SEC engagement on the category has been visibly warmer this year.
Tokenized equities are more liquid than tokenized fund products and benchmarked against the most heavily traded reference markets in the world. However, they carry no intrinsic APY, and equity volatility is materially higher than anything in the current composable yield layer — which means tighter LTVs, larger haircuts, and a different risk profile for any lending market that wants to list them at scale. Most onchain activity sits on Solana, where xStocks holds 93% market share, and Kamino is the only major lending venue currently accepting them as collateral across eight tickers (TSLAx, SPYx, NVDAx, AAPLx, QQQx, GOOGLx, HOODx, MSTRx), with TSLAx the largest position at roughly $14M.
Total deployment is around $21M, against $1.4B in category AUM, well below credit and reinsurance but already a higher share than treasuries. Whether equities clear the volatility bar at venues beyond Kamino is the most concrete near-term test of how far the composable RWA stack actually extends.
What we're watching
- Syrup deployment trajectory — does the $522M stabilize or climb back, and at what spread to treasuries.
- Basin and the next wave of redemption facilities — whether treasury wrappers acquire enough liquidity flexibility to function as collateral.
- USCC vs. USTB share on Horizon — the cleanest single signal of how yield compression reshuffles deployment inside a constant access environment.
- Ondo Perps launch and tokenized-equity LTVs on Kamino and beyond — whether equity volatility can be priced into composable-yield-layer parameters.


