BNY and Citadel Securities say they settled a Treasury repo in minutes using a tokenized note

The Bank of New York Mellon (BNY) and Citadel Securities said on October 13 that they completed an overnight triparty Treasury repurchase agreement that settled "in minutes" against the standard T+1 cycle, using a tokenized representation of a U.S. 2-year Treasury note as collateral. The trade is reported by The Block and Journal du Coin. Both pieces are secondary — BNY itself has not published a dedicated newsroom release at time of writing, and the ledger platform used is identified only as BNY's "private digital ledger." Both outlets reproduce the "first use of tokenized assets on a major triparty platform" framing as BNY's own claim.

What was settled

• Structure: overnight triparty repo. Citadel Securities posted the collateral, BNY was the cash provider and the triparty agent, and BNY ran the ledger that recorded the tokenized collateral.
• Collateral: a tokenized representation of a 2-year U.S. Treasury note. The underlying Treasury was purchased by Citadel on Tradeweb.
• Platform: BNY's "private digital ledger." Neither outlet names a product (not Canton, not Fnality, not Hashnote); the token's issuer or minter is not identified either.
• Size: not disclosed.
• Settlement claim: minutes to settle, versus the T+1 cycle on a standard triparty repo book.

The 2-year note was tokenized on BNY's ledger and treated as the collateral leg of a repo contract whose confirmation and lifecycle events — pledge, release, substitution — ran on the same ledger. The cash leg sits with BNY.

What this does and does not add

The U.S. repo market is where tokenization pilots have piled up for two years. Several reference points on the same base case have already been placed:

• Canton Network, with Bank of America, Citadel Securities, DTCC, SocGen and others, settled an on-chain intraday Treasury repo in August 2025 using USDC — a comparable "repo-with-tokenized-collateral" claim on a different stack.
• DTCC piloted a conversion of DTC-held U.S. Treasuries into blockchain tokens through its ComposerX platform on Canton. Canton has reportedly processed more than $350 billion in daily U.S. Treasury repo volume.
• Yesterday on this site: DRW and Marex booked a derivative — a total return swap on a tokenized S&P iBoxx index — on Canton via Derivnex.

The BNY trade differs from the Canton work on three axes:

1. Platform ownership. The ledger is BNY's; the Canton work is on a shared permissioned network. One is operator-controlled; the other, consortium-governed.
2. Role of the triparty agent. BNY is the dominant U.S. triparty collateral agent for cash-equivalent repo. A triparty repo settling on a ledger run by the triparty agent itself is a different claim from a bilateral repo on a shared network.
3. Cash leg. The Canton pilot used USDC. The BNY leg appears to be central-bank-settled cash — the triparty standard — with no public stablecoin component.

Where it does not help the reader

The absence of a few facts limits what a reader can do with the news today:

• No platform name means the trade cannot be traced on a public explorer; the ledger is private by design, and no counterparty has published a reference identifier.
• No size means we cannot compare it against normal triparty volumes.
• No token-issuer identity means the regulatory treatment of the tokenized Treasury — security vs. electronic deed of entitlement vs. digital asset under internal BNY policy — is not resolvable from secondary coverage.
• No BNY press release at writing means every attribution flows from one wire-adjacent BNY communication relayed by two outlets.

Readers who need a confirmed primary should wait for the BNY newsroom release or a filing-level disclosure through the counterparties' own channels before building anything on the specific mechanics.

What to watch

1. The platform name. BNY disclosing — or declining to disclose — which product its "private digital ledger" refers to, and whether it interoperates with Canton or any of the public settlement networks, is the next piece of the picture.
2. A second counterparty. A triparty repo between BNY and Citadel on a BNY-owned ledger is thin on independence. A second cash provider or a second collateral poster signals the ledger is a market venue rather than a single-counterparty workflow.
3. A disclosed size. Repo pilot volumes have run from tens of millions to the low billions in parallel programmes. Where this trade lands shapes how fast it scales.
4. Interaction with CCP or Fed settlement. BNY's triparty book sits inside Fed-settled cash. Any extension of the ledger into DVP flows that touch Fedwire — or into the Fixed Income Clearing Corporation's sponsored-repo book — is the step that would move the pilot from infrastructure demonstration to live plumbing.

Context

The BNY–Citadel claim takes the shortest path from existing triparty mechanics to tokenization: keep the agent (BNY), keep the cash (central-bank-settled), keep the counterparty roles, and replace one leg — the collateral representation — with a token on an internal ledger. That is a smaller change than the public-chain repo pilots this site has tracked, and a larger-balance-sheet one. Whether tokenization lands first as a BNY-ledger product sold to existing triparty clients, or as a shared-network product sold to a widening set of banks, is the market structure question these two tracks are now actively testing against each other.