Brian Reed: Eliza, long week — but I've been sitting with this stablecoin research and honestly it made the week feel longer.
Eliza Ward: Heavier subject than it looks from the outside.
Brian Reed: Much heavier. Because on the surface you have this headline number — twenty-seven trillion dollars settled by stablecoins in 2024, Tether and USDC leading that — and it's supposed to land as proof that this infrastructure arrived. Bigger than Visa and Mastercard together.
Eliza Ward: Hold on — combined volume of both? That's the comparison being made?
Brian Reed: That's the comparison. And I want it to be true, but let me see — the research doesn't actually separate genuine settlement from circular flows. Internal exchange movements. Money that isn't going anywhere except back to where it started.
Eliza Ward: Which means the twenty-seven trillion is real as a transaction count. Whether it maps to real economic activity is — that's unresolved.
Brian Reed: Exactly the gap. And that's what I want to actually work through — because if even half of that is wash-adjacent, the case for USDT and USDC as the new settlement layer looks different.
Eliza Ward: It looks like infrastructure waiting to be proven rather than infrastructure that arrived.
Brian Reed: But even if the volume number is murky, something real is happening underneath it. Because how does it actually work? Like, strip away the dollar figures.
Eliza Ward: It's a universal adapter plug. You have money in one venue, you need it in another — you convert into USDC or USDT, move it instantly, convert out. No bank wire. No two-day settlement window. No correspondent bank in the middle.
Brian Reed: That's actually the clearest I've heard it.
Eliza Ward: And the concrete case — imagine a derivatives trader in Lagos. Price gap opens between two exchanges. She has maybe sixty seconds before it closes. A bank wire takes two days. A stablecoin moves in under a minute. That's not theoretical, that's the rail being used right now.
Brian Reed: So the settlement function is real independent of whether the volume headline survives scrutiny.
Eliza Ward: Right — and that's actually what the correlation research confirms. Not the volume. There's peer-reviewed data showing Bitcoin price correlation of ρ = 0.9992 across markets with capital controls, in places where governments are actively trying to fragment those markets. That number is — wait, I mean, that's not noise. That's stablecoins functioning as the actual shared numeraire pulling fragmented prices into line.
Brian Reed: 0.9992 across capital-controlled markets? That's the part that actually does the work, isn't it — because it shows the function persisting even where the system is designed to resist it.
Eliza Ward: And there's a mechanism underneath it. Whitelisted entities can redeem a stablecoin trading below a dollar for a full dollar from the issuer — that pulls supply down, restores the peg. Arbitrage and settlement are self-reinforcing. The structure repairs itself.
Brian Reed: So the correlation number isn't just a data point — it's actually evidence the self-repair is working at scale, even where the headwinds are real.
Eliza Ward: That self-repair holds — until it doesn't. And that's actually the take I want to push back on, because I keep hearing it framed as design-agnostic. Like, the function persists, so it doesn't matter which stablecoin you're using.
Brian Reed: That's the wrong take. That's exactly the one I want to name.
Brian Reed: TerraUSD. May 2022. Someone attacked the liquidity pool, the algorithmic peg mechanism broke, and forty billion dollars in value was gone in a week. That's not a quiet failure where users migrate to USDC and the function carries on. There was a re-pegging scramble across the whole market. Contagion. That's not design-agnostic — that's a specific design getting destroyed and dragging the room down with it.
Eliza Ward: Okay, but — wait, I actually think the complication is that USDC and USDT held. The fiat-backed designs held. So isn't that exactly what design-agnostic means? The function survived in other assets.
Brian Reed: That's survivorship framing. You're looking at what's still standing and calling it proof of durability. But the function didn't migrate cleanly — there was a gap, there was chaos, and you had three distinct design families — fiat-backed like USDC and USDT, crypto-collateralized like DAI, and algorithmic like UST — and one of them just ceased to function as a settlement asset entirely. That's not a minor footnote.
Eliza Ward: No, I'll give you that. Design determines tail risk. The function might be durable in aggregate, but any single design can be wiped out — and no current design fully resolves the trilemma between decentralization, stability, and compliance. None of them.
Brian Reed: And if trust collapses across designs simultaneously — which May 2022 came closer to than people admit — the function doesn't migrate, it freezes. That's the scenario the design-agnostic claim can't actually survive.
Eliza Ward: The issuer concentration piece is where this gets worse, and we'll get there — because Circle's CCTP extends the function across chains, but that's also a single point of failure if Circle loses a license.
Brian Reed: And that's — okay, so let me actually say what that means concretely. Because USDT and USDC together basically are the settlement layer. That's not a feature, that's the exposure. You have two centralized issuers backed by bank deposits and short-term Treasuries, and the whole function runs on trusting their reserve adequacy and their regulatory standing. Neither of those conditions lives on the blockchain.
Eliza Ward: Which is just a traditional financial concentration problem wearing a different coat.
Brian Reed: Right. And Circle's CCTP is actually a good way to see that clearly — it burns USDC on one chain and mints it on another, no liquidity fragmentation. Elegant. But if Circle loses a single license in a key jurisdiction, that specific mechanism goes with it. The burn-and-mint stops. The cross-chain settlement stops.
Eliza Ward: And settlement doesn't migrate instantly. There's a gap — traders are asking what the new numeraire is while the function is frozen.
Brian Reed: Picture a treasury desk in Singapore — they're routing USDC cross-chain through CCTP at eleven at night. Circle's license gets pulled by morning. They're not seamlessly on DAI. They're on the phone.
Eliza Ward: And the African corridor research actually — wait, this is the part I find most clarifying. Because even where stablecoins are supposed to win, they don't dominate every dimension. The research found no single rail, stablecoins included, performs optimally across access, cost, speed, compliance, and settlement risk simultaneously. Not one.
Brian Reed: So the durable settlement layer is partly just — durable for the people already running on it?
Eliza Ward: That's a sharper version than I'd go, but — yeah, it's not far off. And the ACM 2025 cross-chain DEX arbitrage research, the year-long measurement study, that's framed as the next frontier of MEV. But most price gaps are still closing against centralized exchanges, not on-chain. The multi-chain durability thesis is forward-looking. It's not confirmed at scale yet.
Brian Reed: So the thing to watch is whether Circle or Tether face a regulatory action in a major jurisdiction before that on-chain infrastructure matures enough to absorb the shock. Because right now those two events arriving in the wrong order is the actual risk.
Eliza Ward: And that order-of-events problem — that's actually the thing I can't resolve. Because if Circle loses a major regulatory license in 2026 or 2027, does the twenty-seven trillion in settlement just... drift toward Tether over three weeks? Or do we get six months of re-pegging paralysis while every treasury desk in Singapore figures out what they're even holding?
Brian Reed: That's the actual question, yeah. And I don't — I mean, UST felt like it should answer it, but it doesn't, not cleanly. Different design. Algorithmic peg failing is not the same experiment as a fiat-backed issuer losing regulatory standing. The shock mechanics are completely different.
Eliza Ward: Nobody actually knows. That's the honest landing. The durability thesis might be structural — or it might just be contingent on Circle and Tether staying solvent and licensed. We genuinely can't tell from the data we have.
Brian Reed: I keep sitting with that. Whether "durable" means the function survives, or just that whatever's standing after the shock gets called durable in retrospect.
