Design Stablecoin

Problem Statement: Fiat-Backed Stablecoin Issuer

A fiat-backed stablecoin holds a 1:1 peg to the US dollar by keeping one real dollar of reserves for every token in circulation — the Circle USDC model, as opposed to crypto-collateralized DAI or algorithmic designs. You are designing a regulated issuer: the system that mints tokens when institutions wire in dollars, burns them on redemption, and continuously proves the reserves exist. The token contract is the small part; the hard system is the off-chain treasury, banking, and attestation machinery behind it.

The defining invariant — and what an interviewer is really testing — is that on-chain liabilities (`totalSupply`) must never exceed provable assets (reserves). Every token is backed by a dollar held in T-bills and cash buckets, and the mint-only-after-irrevocable-fiat-settlement rule (wire in before mint) is the cardinal safety property. Get it wrong and you have printed unbacked money.

The scale to anchor: peak transfers ~550 TPS (mostly on-chain ERC-20 transfers you only index), mint bursts ~200 wires/hour, a peg-stability band of ~30bps soft / 100bps hard halt, and a reserve ratio ≥100%. The consistency split to state up front: the issuer's liability ledger is strongly consistent (CP, RPO zero) because it is money, while the on-chain index is eventually consistent (AP) with repair jobs.

The failure stories that define this problem are specific, not generic: an SVB-style banking shock (reserves trapped in a failing bank, triggering a depeg — this actually happened to USDC in March 2023), a bank holiday blocking redemptions, an oracle fault mispricing the peg, and a double-mint from a bank retrying a webhook. The token also needs regulated hooks — pause and blacklist for sanctions — that a pure DeFi token would not have. Naming proof-of-reserves merkle roots, HSM signing ceremonies, and these treasury shocks is what separates a staff answer from an ERC-20 tutorial.