BUIDL vs BENJI vs OUSG vs USDY: The Race for On-chain Cash

In March 2026, Circle's USYC overtook BlackRock BUIDL as the world's largest tokenized Treasuries fund, and as of July 23, it leads with approximately $2.96 billion compared to BUIDL's $2.52 billion. If the market were read solely by the ranking of assets under management —AUM, the total money each fund manages— the conclusion would be that BlackRock is no longer winning, and that reading is deceptive. The tokenized Treasuries fund market grew from about $1.1 billion in March 2024 —when BUIDL debuted with 100 million in seed— to nearly $16 billion as of July 23, 2026: more than 14 times in two years, and it already functions de facto as the default "cash drawer" for treasuries operating on-chain. Within that race, there are two wars that an AUM ranking hides —DeFi integration and cost— plus a separate geographical board where USDY plays. BUIDL matters less today for its size —it no longer leads— than for being the real guarantee of on-chain crypto: collateral in Euler, derivatives guarantee in Binance, and reserve for Ondo's OUSG and Ethena's USDtb stablecoin. Franklin Templeton BENJI competes on the cost axis with the lowest fee in the category, at 0.15% tied with OUSG. This article puts BUIDL, BENJI, OUSG, and USDY in the same dated table —AUM, growth, fee, net yield, and availability— and explains what makes a tokenized fund the default collateral for an on-chain treasury.

Why the AUM ranking no longer tells you who's in charge

A tokenized Treasuries fund invests in very short-term U.S. public debt —Treasury bills and overnight repos— and represents each share as a token on a blockchain instead of on a traditional manager's registry. The appeal is twofold: it pays the yield of public debt (around 3.3% annually as of July 2026) and is moved, custodied, and settled 24 hours a day on the same network where the investor's money already resides. This is why crypto treasuries use it as a cash substitute: a balance that generates yield without leaving the chain.

Throughout 2024, the picture was simple: BUIDL, launched by BlackRock in March of that year through the tokenization firm Securitize, came to control nearly 46% of the market at its peak in May 2024, a figure that fueled the "BlackRock dominates tokenization" narrative. Two years later, this no longer describes the reality. With the market having multiplied and several competitors operating at scale, BUIDL's share dropped to around 16% as of July 23, 2026, and Circle's USYC —inherited upon the acquisition of Hashnote in January 2025— surpassed it in AUM in mid-March 2026, maintaining the lead ever since.

What is interesting is how USYC won that top spot: through its instant convertibility against USDC—Circle's stablecoin—and its use as margin collateral on Binance and other derivatives platforms, with Circle's distribution plumbing pushing its own reserve product. Its DeFi integration also exists—USYC is, in fact, accepted collateral in Aave Horizon—but what gave it the lead was commercial power. That nuance is key: the number at the top of the ranking measures the issuer's distribution strength more than the token's utility as a system component. Only the latter decides which fund becomes indispensable once the marketing fades.

What makes a tokenized fund accepted collateral?

The variable that no AUM ranking captures is whether a lending protocol—or a stablecoin issuer—accepts that token as a guarantee. This depends on three conditions evaluated one by one:

• Technical integration. The token must fit the standard the protocol runs on (usually ERC-20 on Ethereum, sometimes wrapped in an ERC-4626 vault that automates yield), have reliable price oracles, and audited contracts. A token without clean integration doesn't get in, no matter how large the fund behind it is.
• Redemption liquidity. A protocol accepting collateral needs to be able to liquidate it quickly if the loan deteriorates. If the fund only allows redemptions on business days or with restricted windows, the collateral is fragile; if it offers near-immediate exchange for a stablecoin, it is solid. Exit speed matters more than size.
• Issuer risk and legal clarity. Who custodies the assets, in which jurisdiction, with what bankruptcy-remote structure, and which regulator oversees it. A fund backed by BlackRock and custodied by BNY clears an institutional protocol's risk committee with less friction than a young issuer, even if both hold Treasury bills underneath.

This capital grows through a path not seen in the snapshot of each fund individually. Real-world asset (RWA) deposits parked in DeFi protocols tripled in one year: from about $2.33 billion in Q2 2025 to $7.44 billion in the same quarter of 2026, a 200% jump. That money seeks the fund that protocols accept as collateral. Morpho, in fact, increased its live loans from $1.9 billion to $3 billion during 2025 (+58%; by mid-2026 they were around $3.7 billion) relying largely on tokenized collateral to become the sector's second-largest lender.

Why BUIDL remains the default reserve even if it doesn't lead in AUM?

The short answer: because other funds are built on top of it. Ondo's OUSG, one of its direct competitors in the table, has maintained its portfolio within BUIDL since a 2024 migration; buying OUSG is, in practice, a wrapped route into BlackRock's fund. And Ethena's yield-bearing stablecoin USDtb uses BUIDL as the reserve for over 90% of its backing. When a fund becomes the reserve for its own rivals, its relevance is no longer measured by its AUM: it becomes the base layer upon which the rest of the market is stacked.

Added to this is its use as real guarantee. In May 2025, Securitize wrapped BUIDL into an sToken—sBUIDL—which Euler accepts as loan collateral on Avalanche, the route through which the fund enters DeFi without giving up its institutional perimeter. And since November 2025, it has served as off-exchange margin for trading derivatives on Binance, Crypto.com, and Deribit: the balance earns yield as a Treasury bill while backing the position. A treasury looking for a cash drawer that also serves as a guarantee chooses the fund its platform already recognizes, and there BUIDL—an issuer with maximum perceived solvency, bank custody, and exchange liquidity—starts with a structural advantage that no size-based ranking measures.

This does not equate to absolute superiority. BUIDL charges between 0.20% and 0.50% for management depending on the class and requires qualified purchaser status through Securitize, with high minimums: it is infrastructure for institutions rather than a retail product. Its role as the gateway for institutional on-chain yield, along with its ETH staking product, is analyzed in the fund with which BlackRock opened institutional on-chain yield.

Why is BENJI winning the fee war with 0.15%?

BENJI plays a different game. It is the on-chain token of the Franklin OnChain U.S. Government Money Fund (FOBXX), an SEC-registered money market fund —the first to use a public blockchain as the official record of ownership, starting on Stellar in 2021— and currently deployed across nine networks: Stellar, Polygon, Arbitrum, Aptos, Avalanche, Base, Solana, Ethereum, and BNB Chain (added in September 2025). Franklin announced it surpassed $1 billion in AUM in March 2026, although the RWA.xyz on-chain registry —which only counts issued BENJI tokens— places it at approximately $735 million as of July 23; both figures leave it far from the leadership of USYC or BUIDL, and it does not need them to remain relevant.

Its weapon is the fee: 0.15%, the lowest in the category—tied with OUSG—a differential that stems from keeping the shareholder registry directly on-chain and eliminating legacy administration layers. Franklin also expanded institutional access through a partnership with MoonPay and reaches investors outside the U.S. via sister UCITS funds and in Singapore, without giving up its SEC registration. A fund can have one-fifth of the leader's AUM and still be the rational choice for those optimizing cost over lending integration: on a $100 million balance, paying 0.15% instead of up to 0.50% annually saves about $350,000 per year.

How do BUIDL, BENJI, OUSG, and USDY compare today?

The following table crosses the four products with the cutoff date for each block, and its trajectory column—three dated points per fund—reveals that they compete on different axes.

To put the scale into perspective: RWA.xyz valued the tokenized Treasuries market at approximately $15.92 billion—85 assets and about 62,900 holders—as of July 23, 2026, with an average 7-day yield of 3.30%, lower than that of several individual funds because the aggregate includes shorter-duration products. We track the full RWA sector snapshot in the tokenized asset market count, which surpassed $26 billion including credit and real estate.

Why USDY plays on a separate geographical board?

USDY is the case that breaks the direct comparison, because its competitive edge is neither AUM nor fees: it is geography and retail yield. It is a yield-bearing token backed by short-term Treasury bills and bank demand deposits, issued by Ondo USDY LLC, a bankruptcy-remote vehicle in Delaware, and each token represents a senior secured interest in that portfolio. As of July 23, 2026, it pays 3.55% annually with approximately $2.16 billion in circulation distributed across Ethereum, Solana, Mantle, Sui, and Aptos — nearly triple the 740 million from late April, the fastest growth among the four funds in this comparison.

The defining condition is who it can be sold to: only non-U.S. investors. This restriction, which would be a limitation for a traditional product, is the value proposition here: it allows offering a higher yield to international retail within a more tax-efficient structure, without the burdens of a U.S.-registered fund. While BUIDL and OUSG target qualified U.S. institutions and BENJI covers both sides, USDY captures the non-U.S. saver who wants dollar yield without opening a brokerage account in New York. Ondo also covers the institutional end with OUSG, and we develop their full Treasury strategy in the analysis of their bet on tokenized public debt.

What does each fund really measure when it says it's "winning"?

Placed at the same table, the four products win on different variables, and confusing them is the mistake made by those who only read the AUM ranking. The chronology helps show why the dominant metric lost its explanatory power as the market matured:

• March 2024 — BlackRock launches BUIDL with a $100 million seed; within six weeks it becomes the largest fund in a category that already existed (FOBXX since 2021, OUSG since January 2023).
• May 2024 — BUIDL hits its peak market share: ~46% of the entire market. AUM and leadership are one and the same.
• January 2025 — Circle acquires Hashnote and inherits USYC, the fund that will eventually dethrone BUIDL.
• March 2026 — USYC surpasses BUIDL in AUM by leveraging convertibility against USDC and its use as margin collateral. Franklin announces that BENJI crosses $1 billion.
• April–June 2026 — The tokenized Treasuries market hits $15.07 billion at the end of April and $14.79 billion as of June 10; BUIDL's share drops to around 17%, but it remains the reserve for OUSG and Ethena and collateral in Euler and Binance.
• July 2026 — The market reaches $15.92 billion as of July 23. USDY nearly triples its size in three months (from $740 million to $2.16 billion) and becomes the third-largest fund in the category, ahead of BENJI and OUSG combined; BUIDL's share remains around 16%.

From that sequence emerge two wars that the ranking hides. The integration and collateral war: who becomes the piece that lending protocols and stablecoin issuers accept as guarantee, where BUIDL starts ahead even if it does not lead in size. And the cost war: who offers the same yield while discounting less commission, where BENJI matches OUSG at 0.15% without being the largest. USDY plays on a separate board —the geographic one—, targeting the retail investor outside the US. The AUM ranking only scores the issuer's commercial acquisition, and that is why it crowns USYC with about 2,960 million while the other two wars are decided on axes it does not measure.

What to look for before treating a tokenized fund as "on-chain cash"?

For a treasury evaluating where to park an on-chain balance, AUM matters less than five prior questions the table doesn't fully answer: Is the token accepted as collateral in the lending protocol I already use? How long does a redemption take and against which asset is it settled? What annual fee eats into my balance? Can I legally buy it from my jurisdiction and with what minimum? Who custodies the assets and which regulator oversees them? We detail an organized framework for answering these—issuer solvency, redemption mechanics, reserve transparency, contract risk—in the guide to evaluating a tokenized fund.

A nuance the market is beginning to incorporate: not all tokenized collateral is demand cash. The funds in this comparison are open-ended products whose value moves little, but the frontier of DeFi collateral is stretching toward instruments with maturity dates, such as the principal tokens we analyze in the Pendle maturity collateral standard. The lesson left by BUIDL, BENJI, OUSG, and USDY is that the indispensable fund for on-chain treasuries is rarely the one at the top of the ranking: it is the one a protocol already recognizes, the one that charges less, or the one that can be sold where others cannot reach. Three advantages that no size-based classification knows how to measure.

Sources and links: RWA.xyz — Tokenized U.S. Treasuries · Yahoo Finance — Circle USYC overtakes BUIDL · CoinDesk — Circle overtakes BlackRock · CoinCentral — USYC overtakes BUIDL as largest tokenized fund · Franklin Templeton — BENJI and MoonPay · Ondo Finance — USDY · CryptoBriefing — RWA deposits in DeFi +200% · CryptoSlate — the mechanical reason for the flippening