Why we invested in Cordant
Getting out of the swivel chair – why we invested in Cordant
On July 10, Circle became the first stablecoin issuer to hold a national trust charter from the OCC1. FedNow settled roughly 5 million payments in the second quarter of 2026, up 83% quarter over quarter2. Stablecoins, tokenized deposits, and real-time settlement aren't a side bet anymore; they're a new rail sitting directly on top of the one every institution already runs: SWIFT, ACH, card networks, core banking. That's the thesis we hold with real conviction in payments and digital assets: the winners will be the business models that bridge the old rail and the new one, not the ones that pick a side.
But bridging the rails only solves half the problem. The other half is what a bank, a PSP, or an issuer does the moment the money actually moves.
This is what happens across Cordant's design partners: a payment settles in stablecoins, and the instant it clears on-chain, someone opens a spreadsheet to check whether that settlement actually matches what compliance policy allows, because the ledger that recorded it and the policy that governs it have never spoken to each other. The new rail didn't remove that step. It added one more system to check by hand.
Cordant comes out of stealth today, and we are excited to co-lead its seed round. Their solution is built for that other half: a layer that reconciles what an institution's contracts, policies, and procedures say should happen against what its systems actually show. Reconciliation is the wedge. The bet is the record of what an institution owes.
The swivel chair
In our experience, banks stitch together six or seven antiquated systems that were never built to talk to each other. Nothing in that stack holds the answer, so a person does, someone turns from screen to screen and reconciles by hand. That person is the integration layer. Call it the swivel chair, a thesis we've been writing on since 2024.
A payment is a claim, not an event: several independent parties have to agree on it after the fact. The payer's bank, the acquirer, the rail, the custodian, and the compliance function each hold a partial, private view, and discrepancies between them are the normal state, not the exception. Five partial views, but no shared one.
Every new rail is one more screen
Stablecoins don't replace the chair, they scale it. Every rail an institution adds, a stablecoin rail included, is one more boundary it has to hold in agreement with all the others: one more ledger, one more settlement window, one more policy to reconcile by hand. The chair gets one more screen. Nobody takes one away.
Getting out of the chair
Cordant is designed never to touch the money, no keys, no flow of funds. The ledgers and rails already hold the canonical transaction. Cordant is built to hold the canonical expectation and act on the gap between the two. That's why a bank, a custodian, and a processor can each hand Cordant operational data they'd never hand a competitor: being one of the systems of record is exactly what disqualifies you from arbitrating between them.
Take a contract that specifies settlement at T+2 while the system quietly delivers at T+5. Nothing alerts. The ledger records what it received, the rail reports what it sent, the dashboard is green, and every system is technically behaving correctly. The break surfaces weeks later, if at all, in a spreadsheet, an audit, or a counterparty's phone call.
Cordant is designed to read both sides without requiring an upstream schema change — no rip and replace. On one side, the events an institution's systems emit across payments, cards, banking, and ledgers. On the other, the governing logic in its contracts, policies, and commission schedules. The contract term is the expectation, the settlement event is the observed fact, and the divergence between them is the alert. What it's designed to leave behind is the product: the clause, the event that contradicts it, a timestamp on both, and a record of who was told. The person who used to swivel between screens gets the answer instead of the job.
By Cordant's own estimate, a single custom integration like this costs $500K to $1M today. Replace that with a layer that gets cheaper to extend with every institution that adopts it, and the economics flip.
Why we invested
Proof is turning into a filing requirement, with dates on it. Starting in November, SWIFT will require every bank to be able to receive a new, standardized format for reporting and resolving payment disputes. At first, firms can keep handling those disputes the old way underneath the new wrapper. That grace period ends in November 2027, when every dispute has to be handled in the new format, start to finish3. The chair has one more year. For stablecoin issuers, the bar is already higher: under the OCC's proposed GENIUS Act rules, an issuer would have to publish a monthly reserve report, have an accounting firm examine it, and have its CEO and CFO personally certify it4. We expect that bar to generalize, and we don't think most institutions can clear it today.
An agent can read everything the institution did and nothing it owed. The Cambridge Centre for Alternative Finance, with the BIS, IMF and World Economic Forum, finds four of the top five AI use cases in finance are back office5. In our view, the model is not the constraint, and nothing in the stack says what the institution owed. An agent inheriting the swivel chair just swivels faster. The day an agent releases a settlement, reprices a merchant or clears an exception, somebody has to defend it against the term it was taken under. The industry is handing agents identities and spending limits, but it has not built the record that shows an agent did what policy said.
A PSP, an issuer and a bank describe the same constraint. Cordant says roughly ten institutions meet with the team weekly: PSPs, issuers, banks, stablecoin orchestrators, fintechs and payment processors. What convinced us is where they put the problem. A back office that caps commercial strategy is not a compliance complaint. It is a growth constraint.
The team
Eric Rosenthal (CEO), Juan Jose Huezo (CPO), Lior Levitt (CTO) and Sagi Ittah (founding engineer) all come out of Rapyd. They scaled it across 50+ countries, oversaw 100+ partner integrations, and carried compliance across dozens of jurisdictions. Every integration added a boundary. Every boundary produced exceptions, and the exceptions came to them.
Looking ahead
Picture the institution that gets this right. Its obligations are legible to its software, so a broken term surfaces the day it breaks, not weeks later. It adopts the next rail, stablecoin or otherwise, and the chair doesn't get another screen. Every institution that gets there becomes the reference for the next one: the same clauses, the same rails, the same exceptions, recognized faster each time. We are backing Eric, Juan, Lior, and Sagi to take the chair away.
1. Circle, "Circle Receives Final OCC Approval to Establish National Trust Bank," July 10, 2026; CNBC.
2. Federal Reserve Financial Services, "FedNow Service Volume and Value Statistics," updated July 6, 2026. Source
3. Swift, "ISO 20022: Exceptions and Investigations" and "Transforming exceptions and investigations." Source
4. Office of the Comptroller of the Currency, Notice of Proposed Rulemaking implementing the GENIUS Act, Feb. 25, 2026; see also Morgan Lewis, "Stablecoin Regulation: OCC Proposal Under the GENIUS Act." Source
5. Cambridge Centre for Alternative Finance, in partnership with the Bank for International Settlements, IMF, and World Economic Forum, "The 2026 Global AI in Financial Services Report." Source
Disclaimer
This communication reflects the views of the Motive investment team as of the date hereof and is subject to change without notice. Motive or its affiliates may hold positions in the securities of companies discussed herein. References to Cordant as a portfolio company of Motive are provided for informational purposes only and do not constitute an endorsement of Cordant's products or services. Product capability descriptions reflect Motive's understanding of Cordant's intended functionality and may not reflect current or future capabilities; Cordant makes no representations regarding those capabilities. This communication contains forward-looking statements based on Motive's current expectations and assumptions; actual results may differ materially. Nothing herein constitutes an offer to sell or solicitation of an offer to buy interests in any Motive managed investment vehicle.