SWIFT sits there in the middle. Three to five days. Six seconds on blockchain. But nobody uses it. And that gap... that's where the interesting numbers are hiding.
Actually that's not quite right. Most companies still do use SWIFT. They hate it though. Half to two percent in fees. Money's gone from your account, not in theirs yet. Three days of just sitting somewhere. But it's what banks know. Safe. Regulated. Boring.
Something shifted about 18 months ago. Not the technology. The economics flipped. USD to EUR with 100K on it now works better on stablecoins than through correspondent networks. USDC liquidity on Arbitrum is stupid deep. You push the money, it's there. Cost is lower. Speed is incomparable.
Wasn't like this in 2023. Is now.
Not every corridor yet. Some routes still need traditional banking. But the ones that work are expanding month by month and the trend is obvious if you're paying attention. Which most treasurers finally are.
Correspondent banking is a horror show when you really look at it. Australian company needs to pay a supplier in France. Doesn't route directly. Australian bank talks to a hub bank in London or New York. That hub bank has an account somewhere that talks to the French bank. But the Australian bank and French bank never meet. They have no relationship.
Each step: 15 to 30 dollars. Maybe 24 hours per step. Three legs, three days, call it 45 to 90 dollars gone just to move the money. Million dollar transfer across correspondents. Three to five thousand dollars vanishes. Just for the privilege of moving electronically through four banks that mostly hate each other.
And the money's stuck between. Australian company lost it. French supplier doesn't have it. Nobody can count on it. For smaller companies especially, that float uncertainty kills cash flow.
Banks don't speed this up because they don't want to. They're making money on float. Three days of holding it. Settlement today? That revenue's gone. Incentive structure is completely broken.
Stablecoins work differently. Australian company buys USDC on a Sydney exchange. Literally that. Sends USDC on-chain to the French supplier's wallet. French company converts USDC to EUR on a Paris exchange. Total elapsed time. 15 minutes.
Money breakdown. AUD to USDC on-ramp costs 0.5 to 1 percent. On-chain move is basically free. USDC to EUR off-ramp costs 0.5 to 1 percent. Total. One to two percent. 100K transfer costs 1 to 2K. Traditional correspondent cost is 1.5 to 3K plus three days of timing risk. Stablecoins win on price and speed both.
But only if the liquidity exists on both sides. USD to EUR? Deep as an ocean. Slippage is almost zero. AUD to ZAR? Basically nothing. The stablecoin win only happens where both endpoints have real markets.
Circle publishes quarterly USDC data. Tether's less chatty about USDT, but you can see the on-chain numbers yourself if you run an indexer. The data tells a specific story. Certain corridors just hit critical mass.
USD to EUR is the heavyweight. Biggest volume by far. USDC and USDT between the US and Europe is now faster and cheaper than correspondent banking if it's over 50K. Slippage basically doesn't exist. Rate you see is the rate you get. Multinationals with European operations are routing the big USD-EUR transfers through stablecoins. We're estimating 15 to 20 percent penetration in that corridor now. 2022 it was under 3 percent.
USD to GBP is the same pattern. US to UK stablecoin settlement is normal for corporate treasurers now. British banks actually prefer it. Less operational load than correspondent transfers. London's still a financial center so the liquidity is excellent.
USD to AUD is different. Slower. On-ramp costs higher in Australia because the market's smaller. Maybe 1.5 to 2 percent. Off-ramp is worse. Only makes sense if you're moving over 200K.
But Singapore and Hong Kong are interesting. Regulatory clarity, competitive exchanges, depth of liquidity. USD to SGD or HKD. 0.8 to 1.2 percent all in. Traditional banking is 1.5 to 2.5 percent. 15 minutes versus three days. Not even close.
Emerging market corridors are where this is actually winning hardest. USD to Mexico, Brazil, India. But the win isn't about beating the bank on price. It's about not having transfers vanish into a black hole. Traditional correspondent banking to emerging markets is a mess. Transfers get stuck. Exchange rate locked in for three days so both sides are taking risk. Stablecoin instead. Lock the on-chain rate. Six seconds to settle. Receiving company converts to pesos or reais at their own pace. Same day probably. Not a timing game anymore.
Tech companies especially are moving their contractor payouts to stablecoins. 30 to 40 percent of USD payouts to Latin America now go that way. Not theoretical. Actually happening.
Circle's euro stablecoin landed late 2023. EURC. Reshaping how Europe settles itself. Before that, sending EUR meant settling in USDC or USDT and converting both ends. Slippage. Extra cost. EURC cuts that out. German company to Polish supplier. 15 seconds on Arbitrum. No currency conversion touching the money mid-flight.
EURC liquidity's still building though. Available on Arbitrum and Polygon but the pools are smaller than USD pairs. Big transfers, 200K EUR and up, slippage still bites. More EURC issuers and more exchanges listing it changes that. Matter of months probably.
On and off-ramp capacity is a constraint. Company wants AUD to USDC. Needs a Coinbase account or a local exchange, maybe eating 2 percent fees. Banks have been slow to add stablecoin on-ramps even though some in Singapore and the UK finally are.
When banks realize stablecoin settlement can't be stopped—when they get that it's not a fad—things will shift. JP Morgan already opened USDC rails to corporate clients. Others will follow.
Regulation actually helps. EU's MiCA allows qualified issuers. FCA in the UK blessed stablecoins for settlement. Singapore's MAS got it. US is lagging like always. But companies operating with EU counterparts are operating legally and cleanly.
It's not dominant flows yet. But the trajectory is impossible to miss. Quarter after quarter the percentage climbs. Banks haven't cut cross-border fees in response. Suggests they're still pretending this isn't happening. But it is.
A multinational moving 100 percent of cross-border settlement to stablecoins saves millions annually. That math is obvious to CFOs now. Velocity's going to accelerate.
Emerging market senders win hardest. Correspondent banking's slowest and most expensive there. Companies with constant small transfers win too. Those fees compound. Treasury teams needing 24-7 settlement without bank holidays. SMEs that can't negotiate good rates with traditional banks. Correspondent banks that built business models on float and fees are getting pressure. Not disappearing yet. But the structural advantage is eroding.
You're building cross-border infrastructure right now? You have maybe 18 months before someone else captures this space. After that, execution matters more than pioneer status.