USDC: The Fiat-Collateralized Model, the SVB De-Peg, and Circle's Sole Control

CENTRE aims to provide the first: a fiat-collateralized approach. One unit of tokenized fiat currency is backed by one unit of reserved fiat. More so than the other approaches to stablecoin development, the fiat-collateralized approach requires meeting firm traditional regulatory requirements, requires issuing members to have strong auditable reserve capability for traditional backing assets (such as fiat banking relationships), and provides less decentralization — and it is also currently the most robust approach in terms of price stability.

CENTRE wrote that into its own whitepaper before a single unit of USD Coin existed. Circle launched the token on September 26, 2018 through the consortium it had just formed with Coinbase, and the founding document named the cost of stability in the same breath as the stability itself. It took the designers who wrote “provides less decentralization” less than five years to pay that cost in public: on March 11, 2023, USDC fell to 87 cents.

An issuer, a reserve, and a redemption desk are the three things Bitcoin’s design was built to need none of. USDC’s own whitepaper is organized around exactly those three, and by September 2026 the result circulates at roughly $75 billion. The twelve-chain design comparison puts USDC in the one column where price does not move at all — pegged, alongside USDT, where new money reaching the issuer mints another unit instead of moving the price.

What the whitepaper commits to, in its own words

CENTRE’s document lists four possible stablecoin designs and picks one by name. The three it passed over are crypto-collateralized tokens backed by reserves of crypto assets, algorithmic non-collateralized tokens that rely on software economic models with no underlying collateral, and a hybrid of the three. CENTRE took the simplest and most centralized of the four, and stated the trade-off in the open rather than burying it. The alternative it did not choose, full decentralization, gets the same directness:

CENTRE addresses the centralization tradeoff by envisioning a network of multiple token-issuing members, thus providing multiple reserves and liquidity sources for network users rather than presenting a single collateralization gateway point of failure. This approach is distributed, though it does not purport to be — or aim to be — entirely decentralized.

“Distributed, though it does not purport to be entirely decentralized” is not marketing language walked back later; it is the sentence CENTRE chose to publish about its own product on day one. Bitcoin buys censorship resistance by having no issuer, no reserve, and no promise of redemption; USDC buys price stability by having all three. Of the axes named by the six structural features behind Bitcoin’s digital-gold status, this is the part USDC gave up from the start.

The same document names a Bitcoin-specific use case for the token, and it is the one that actually grew the category:

A hypothetical investor may choose to protect himself from bitcoin’s fluctuating value by trading his bitcoin for US dollar tokens on a supporting exchange, and be certain that the value of those US dollar tokens will not fluctuate.

That sentence describes USDC’s largest real use to this day: a parking spot for capital fleeing Bitcoin’s own volatility, denominated in the currency Bitcoin was built to need no permission from.

How minting, redemption, and cross-chain transfer actually work

USDC’s supply grows and shrinks through bank wires, not mining or staking, and the 1:1 promise runs through a specific institutional gate. Circle Mint lets approved institutions deposit US dollars and receive newly minted USDC at par, with no spread and no per-transaction fee — but the minimum deposit is $100,000, which routes retail buyers to exchanges and OTC desks instead, each charging whatever spread the venue sets that day. Redemption runs the same path in reverse: an institution sends USDC to Circle, Circle’s smart contract burns it out of circulation, and a dollar wire goes back to the institution’s bank.

Two different populations therefore experience two different versions of “1:1” — one frictionless and direct for the institutions with Circle Mint access, one mediated by a market maker’s price for everyone else. The free mint and redemption at par are also what hold the price at one dollar: if USDC drifts even slightly off the peg, someone with Circle Mint access can profit from the gap, and their trades push the price back.

Whether the reserve exists in the amount claimed is left to outside accountants. Since the 2018 launch, Circle has published monthly reserve reports with attestations, from Grant Thornton at first and, from January 2023, from Deloitte, one of the Big Four. Each report checks bank balances, custody statements, and on-chain data to confirm that Circle holds at least as much in dollar-denominated assets as there is USDC in circulation. But a monthly attestation is a snapshot, not a live guarantee. On the day a bank failed, that gap became a gap in the price.

Moving USDC between blockchains used to mean locking a coin on one chain and minting a separate wrapped copy on another — two tokens claiming to be the same dollar, with a bridge contract as the single point of failure between them. Circle’s Cross-Chain Transfer Protocol, launched in 2023, replaces that with a burn-and-mint sequence: the token is destroyed on the source chain, Circle’s attestation service observes and signs off on the burn, and anyone can submit that attestation to the destination chain, which mints an equivalent amount to the recipient. No wrapped USDC, no second token to reconcile, and no bridge contract holding a pool of assets to steal — but also no step in the sequence that does not pass through Circle’s own attestation.

The protocol now reaches about thirty blockchains. Bitcoin has nothing analogous to reconcile, because it has no second ledger issuing the same asset under a different name. What holds USDC’s cross-chain transfers together is the one thing Bitcoin showed could be done without: a trusted third party.

From a two-company consortium to one company

CENTRE launched in 2018 as, in Circle’s own words, “an emerging consortium,” with Circle and Coinbase jointly running its governance. The whitepaper’s promise of no single collateralization gateway rested on membership: CENTRE itself was to enforce requirements on issuing members covering audits and solvency, licensing and compliance, and capitalization thresholds and limits, so that no one issuer had to be trusted to supply those controls.

On August 21, 2023, Circle and Coinbase dissolved it, saying that growing regulatory clarity for stablecoins in the US and elsewhere meant a separate governance body was no longer needed. Circle took sole control of USDC’s issuance, its governance, the smart-contract keys that mint and freeze the token, and the decision to launch it on new blockchains. Coinbase, instead of co-governing a consortium, took an equity stake in Circle. The two companies kept sharing the interest income from USDC reserves in proportion to the USDC held on each of their platforms, and began sharing equally the income generated by USDC’s broader circulation. The chief executive’s own framing of the change:

Streamline the operations and governance, and enhance the direct accountability of Circle as the issuer.

A network built to have no single collateralization gateway now has one, named in its own press release. Less than two years after that, on June 5, 2025, Circle itself went public on the New York Stock Exchange under the ticker CRCL, pricing its IPO at $31 a share — above the range it had guided to — and closing its first day of trading at $83.23, up roughly 168 percent. The company now holding sole control of USDC’s keys answers to a board of directors and an SEC filing calendar, which is the most conventionally centralized structure available to an American issuer.

The reserve, and the two days it didn’t hold

USDC’s 2018 launch described the token as backed by cash reserves; the reserve’s actual composition took three more years to become fully that. An attestation covering July 2021 reported 61% cash and cash equivalents, 13% Yankee certificates of deposit, 12% US Treasuries, 9% commercial paper, and 5% corporate bonds. Circle moved the reserve to 100% cash and cash equivalents that September. Bitcoin has no reserve composition to argue over, because nothing backs it. Until the backing itself comes into doubt, that difference only looks like a weakness.

The doubt came in March 2023. Circle disclosed that $3.3 billion of the reserve — about 8% — sat at Silicon Valley Bank, which regulators had just shut down. On March 11, USDC fell to $0.87. Circle said it would cover any shortfall from its own corporate funds if the deposit was not recovered in full, and once regulators said all Silicon Valley Bank depositors would be made whole, its chief executive named what had been at stake:

Trust, safety and 1:1 redeemability of all USDC in circulation is of paramount importance to Circle.

The peg came back on March 13, after a weekend in which federal regulators guaranteed that every depositor, Circle included, would get its money. A promise and a price are different things, and the gap between them, for those two days, was exactly as wide as the bank’s solvency. Bitcoin has no bank to depend on and nothing analogous to guarantee.

What Allaire has said about Bitcoin

The person who runs the business built on everything Bitcoin’s design refuses has repeatedly credited Bitcoin with the opposite qualities:

Bitcoin thesis is very much that we are going to see continued growth in non-sovereign money and non-sovereign is going to become more important and not less important.

And more people are going to see the value of a censorship-resistant, highly secure digital asset such as Bitcoin.

Two months later, he located the value not in a market theory but in particular people:

Clearly, a non-sovereign digital asset like bitcoin is attractive to people who are interested in moving capital into a place where they can control it themselves.

It’s the digital gold thesis, and I think a lot of both institutional accumulators of bitcoin, individuals, very specifically individuals in jurisdictions or environments where the intense concern about capital controls are there.

Bitcoin itself has become one of the largest and most important alternative investment assets on the planet.

None of it is offered as a setup for writing Bitcoin off later. USDC was never built to replace what Bitcoin does — a bank wire is what it replaces. Allaire can speak for both without contradiction because the two answer different demands.

Significance to Bitcoin

Run USDC through the six structural features behind Bitcoin’s digital-gold status and each one reads negative: an issuer sits where system decentralization would be, one company sits where no controlling organization would be, a dollar deposit sits where a fair launch would be, an active CEO sits where a departed founder would be, a reserve balance sits where a fixed supply would be, and a 2018 launch date sits where first-mover weight would be. USDC isn’t a failed attempt at those six features — it was never entered in that race. It is what a dollar looks like after every one of them has been traded away on purpose, in writing, for a currency that does not move.

The comparison of fixed and adjustable money puts fiat-pegged tokens in a group of their own because their supply is the issuer’s discretion itself, and USDC’s 87 cents is a measured example of how far that discretion can swing on one outside event, a failed bank.

What makes the record worth keeping is the speed of the trade. In under seven years, USDC went from a consortium of two founding companies promising no single collateralization gateway, to one company holding sole control of the keys, to that same company answering to a stock exchange and a board of directors. Bitcoin has spent seventeen years with no consortium to dissolve and no board to go public, because it was never built with either to begin with.

Electronic cash versus digital gold asks whether the currency used for payments and the currency used to store value can be the same thing. USDC is a design that answers no. To take the payment role, it gave up every property the storage role demands. If Bitcoin carries a lasting tension from trying to hold both roles in one design, USDC never tried to resolve it: it took the payment side and handed the rest to a regulated issuer. That both designs keep existing side by side is itself the sign that the question Bitcoin first raised still has no single answer.