Swift's ledger coordinated FAB and Citi's tokenized dollars while existing systems settled them

Circle runs one issuance and redemption point. Thirty-five separate networks decide whether a USDC transfer happened, because the token is natively available across 35 blockchains and validation belongs to whichever chain carries the payment ^[3][1]. Raj Kamal, founder and chief executive of TransFi, told Crowdfund Insider: "A stablecoin tells us what value the token represents, but not who validates transactions, how consensus is reached or who controls the network." ^[6][16] He describes two layers of trust. One covers the monetary claim, the other the integrity of the ledger ^[4].

The one live transaction in the record is the FAB-Citi test. The two banks completed live US dollar transactions using tokenized deposits through Swift's blockchain-based ledger, with payment activity able to run around the clock ^[7]. Swift says final settlement continues through existing systems, and the banks retain control of their own assets and funding ^[9]. Kamal said: "But the ledger is still an orchestration layer." ^[8]

Liquidity is harder. Kamal said a shared ledger "can validate and coordinate a payment, but it does not create the liquidity needed to exchange one currency for another or connect every digital asset to the recipient's local payment rail" ^[10]. Each additional currency, issuer and network can open another liquidity pool that has to be connected, and the infrastructure question, he says, is how money moves ^[11].

Then comes the reserve income. Singapore requires full backing and bars yield. Kamal said: "I think MAS is right to be firm on reserves." ^[12] He would restrict yield to income earned on permitted reserve assets such as short-term government securities, while prohibiting leverage, lending and excessive duration ^[13]. His reason: "Once trillions of dollars can potentially sit in fully backed digital cash, the question of who receives the income generated by those reserves becomes economically significant." ^[15] Banks are already concerned that stablecoins pull money away from conventional deposits, and 21 major institutions are preparing their own stablecoin venture ^[14]. That venture adds one token to the issuer count, shared across 21 institutions ^[2].

I would expect the binding constraint through 2027 to be FX liquidity and local payment rails, not consensus design, because the only live payment described here already routes finality back into existing systems ^[7][9]. Kamal makes the counter-case. He puts it directly: "If each institution or consortium operates within its own closed network, digital money could reproduce the same fragmentation it is meant to reduce." ^[5] Two results would decide between them. A tokenized deposit that moves from one bank's ledger to another's with no leg through existing systems would show a shared ledger doing settlement instead of orchestration. A 2027 launch that publishes its access and membership terms would show whether the network is open; Kamal did not say which ledger the 21 institutions plan to use.