Lila Soto: Hugo, okay — I was at my cousin's place, she's been doing gig work for a platform based in the US, and she got paid in USDC last month for the first time instead of a wire. And her reaction was not 'wow, crypto.' It was just — 'oh, it was already there.'
Hugo Vance: Already there. Meaning it settled instantly.
Lila Soto: Minutes. No five-day SWIFT window, no correspondent banking chain, no business-hours-only. Just — there. And that stuck with me because, mm, the number that's sitting underneath that small moment is enormous. Stablecoin transactions cleared over five trillion dollars in 2024.
Hugo Vance: Five trillion. That is not experimental volume.
Lila Soto: Not even a little. USDT alone — largest stablecoin by market cap — and USDC from Circle together are over eighty percent of all stablecoin supply. That's the market. And yet — and this is the part I can't quite resolve — to get dollars into that system in the first place, you still go through a bank. On-ramps. Off-ramps. You need an exchange or a payment provider. So the 'bank-free' story has a pretty big asterisk.
Hugo Vance: Which raises the question of what, precisely, the innovation is.
Lila Soto: Exactly — and I think the answer is something specific about the middle of a transaction. Not the entry, not the exit. What happens between those two bank-touches is genuinely different.
Hugo Vance: The settlement leg. Yes. That's where I'd want to look first.
Lila Soto: The settlement leg — yeah, that's the right place to look. Think of it like this: Bitcoin is a truck carrying nitroglycerin. Fast, borderless, but you cannot build a payment system on something that might be worth forty percent less by the time it arrives. A stablecoin is the same truck, same highway, but the cargo doesn't explode.
Hugo Vance: Separate the rail from the volatility problem.
Lila Soto: Exactly that. The blockchain is the highway — fast, programmable, no business hours. The stablecoin is what makes the cargo safe to actually ship. And because it's on-chain, you can attach logic to it. Smart contracts. Like, mm, the payment releases automatically when a shipment clears customs — no intermediary sitting in the middle waiting to confirm.
Hugo Vance: Self-executing. Versus SWIFT, which is — what, two to five business days through multiple correspondent banks?
Lila Soto: Two to five days, business hours only, fees that stack at every intermediary hop. A supplier in Vietnam getting paid by a manufacturer in Germany right now — that money is sitting in correspondent bank limbo for days. The same payment on a stablecoin rail settles in minutes, twenty-four seven, and the fee doesn't compound at each handoff.
Hugo Vance: Now — here's where I'd be cautious. If it's pegged to a dollar, how does it actually stay at a dollar? That's not a small question.
Lila Soto: No, and that's — I mean, that's genuinely the next layer. USDC holds actual dollar reserves. USDT claims to. DAI from MakerDAO does something different — locks over-collateralized crypto in smart contracts instead of fiat. And then Ethena's USDe, which hit fourteen to fifteen billion, doesn't hold reserves at all — it runs a delta-neutral derivatives position to hold the peg.
Hugo Vance: Wait — three completely different mechanisms, all claiming one dollar.
Lila Soto: Three very different bets about what 'one dollar' actually means to hold. And that's the part that gets interesting fast.
Hugo Vance: Three different bets — yes. But what strikes me is that each one relocates the trust problem rather than eliminates it. USDC holds actual dollar deposits and short-term treasuries. You're trusting Circle, and by extension, the bank holding those reserves. USDT claims something similar, though the reserve composition has never been fully audited the way Circle's has. So fiat-backing doesn't mean the risk disappears — it means the risk now lives in a custodian.
Lila Soto: Right — and DAI does something genuinely different there. MakerDAO locks over-collateralized crypto in smart contracts — users mint fewer stablecoins than the value they post. So if ETH drops sharply, the buffer absorbs it before DAI breaks. No fiat custodian at all.
Hugo Vance: But then you've just moved the trust to the volatility of the underlying crypto. In a sharp drawdown, does over-collateralization actually hold?
Lila Soto: That's the exposure, yeah. And then Ethena's USDe is — I mean, it's a completely different animal. No reserves, no collateral vault. It holds the peg by running a delta-neutral position — essentially shorting ETH in derivatives markets to offset its spot holdings. The gains and losses cancel. Fourteen to fifteen billion in supply at its 2024 peak, briefly passing DAI as third-largest. With no fiat reserves. At all.
Hugo Vance: Fourteen billion. Without a dollar in a bank somewhere.
Lila Soto: Not one. The yield it generates actually comes from the funding rates on those short positions. But — and this is the part that keeps me curious — we haven't seen USDe under real market stress. Delta-neutral works beautifully until derivatives liquidity dries up or funding rates go negative.
Hugo Vance: Now — I find genuinely odd about all of this that USDT was created in 2014. Originally just for traders who wanted to avoid Bitcoin's volatility between positions. It was not designed as settlement infrastructure. A container shipping firm using it to pay a port operator in Rotterdam — that use case wasn't in anyone's plans in 2014. The market just... pulled it there.
Lila Soto: Oh — that's actually kind of perfect as a case. Because imagine a treasury manager at a freight company, Thursday afternoon, trying to pay demurrage fees on a delayed container before the port charges compound overnight. SWIFT won't clear until Monday. She routes it through USDT in twenty minutes. The question of whether Tether's reserves are fully audited is — mm, it's real, but it's not her problem in that moment. Urgency chose the dominant rail.
Hugo Vance: Which is how systemic dependencies form before anyone's examined them carefully. And that reserve transparency question becomes everyone's problem the moment something breaks — though whether the GENIUS Act actually forces that examination, or just creates the appearance of one, that's the part we should get into next.
Lila Soto: And the GENIUS Act is exactly where that examination either happens or doesn't. But — I want to slow down on the ramp problem first, because I think it's weirder than it sounds. Like, take a freelance developer in Lagos. Client in Toronto pays her in USDC. The blockchain leg — instant, done. But she needed to buy into USDC through a local exchange first, and that exchange has its own KYC queue, its own fee structure, operates in a regulatory grey zone. So the middle is fast. The edges are not.
Hugo Vance: The endpoints still run through intermediaries.
Lila Soto: Yeah — and counterparty risk doesn't disappear, it just moves to the exchange. Custody risk, corridor-specific compliance, exchange solvency. Those are still real failure modes. Which is why what's actually happening in practice is — I mean, it's not pure stablecoin rails. It's dual-rail. Companies are combining stablecoin settlement for the middle leg with traditional banking for the entry and exit.
Hugo Vance: Dual-rail. So not a replacement — a patch over the seams.
Lila Soto: A pragmatic one. Chainalysis has flagged this — stablecoin liquidity is also fragmented across chains, so a single rail might not even cover the whole corridor. You might need two blockchains and a traditional leg just to get Lagos to Toronto cleanly.
Hugo Vance: Which is where the GENIUS Act lands, yes. If it passes, those local exchanges — the grey-zone intermediaries — suddenly face federal standards. That clarifies things. But it also raises the question of whether 'clarifying' just means the exchange in Lagos now needs a U.S.-licensed partner to touch USDC. That's not liberalization. That's a different bottleneck.
Lila Soto: Oh — that's the tension exactly. Regulatory clarity could pull institutions in because the rules are legible. Or it could impose bank-licensing requirements that strip out the permissionless quality that made the developer in Lagos viable in the first place.
Hugo Vance: Does it make stablecoins safer infrastructure — or just a surveilled payment processor that happens to settle faster?
Lila Soto: That's — yeah. That's the question that doesn't have an answer yet.
Hugo Vance: You know what I keep returning to — not the regulation question. It's the simpler thing. Trust didn't go away. It relocated. From 'will this asset hold its value' to 'who's holding the reserves, and what are the rules around them.' USDC made one bet on that. USDT made a different one. The market, so far, has not cared which bet is cleaner.
Lila Soto: Mm — and that's what makes the GENIUS Act feel like a real hinge, not just a policy footnote. Because regulators don't engage until something has scaled enough to matter. Five trillion in 2024 transactions. That's not regulators being proactive. That's them catching up. Whether USDT's dominance or USDC's positioning wins after that — I genuinely don't know. But I think about my cousin, already-there in minutes, and I wonder if she'd notice either way.
Hugo Vance: She probably wouldn't. Which is, I think, the quietest version of the answer. The infrastructure question — blockchain-native SWIFT, or just a faster PayPal with a longer brand name — that's still open. But for her, the settlement worked. That's all the trust she needed in that moment.
