Stablecoins

Stablecoins

Tokens whose price is designed to stay pegged to a reference value — almost always the US dollar. "Pegged to $1" doesn't mean the price never deviates; it means a mechanism keeps deviations self-correcting. The mechanism is what matters. Every stablecoin design is a different answer to the same question: what forces the price back toward $1 when it drifts?

Stablecoins are DeFi's money market. Lending protocols denominate debt in them. AMMs use them as the reference leg of every pair. Cross-chain payments use them for settlement. When the dollar-denominated unit drifts by 5%, every protocol assuming stability takes unexpected losses. When it collapses 90%, the damage is existential.

Fiat-Backed (USDC, USDT)

The simplest model: for every stablecoin in circulation, the issuer holds $1 in bank accounts and money market funds. Peg holds as long as the custodian is solvent and the redemption channel stays open.

USDT has been in circulation since 2014 and holds $100B+ — the largest stablecoin by market cap. A 2021 CFTC settlement revealed reserves had included commercial paper and secured loans, not pure cash. Current attestations show primarily US Treasuries. Incorporated in the Cayman Islands, quarterly attestations only (no full audit). Traders use it for liquidity; institutions prefer USDC.

USDC is issued by Circle, licensed by the New York DFS, and publishes monthly reserve attestations from Grant Thornton. Holdings restricted to cash and short-duration Treasuries. The March 2023 SVB collapse exposed the model's limits: Circle disclosed $3.3B of USDC's ~$40B in reserves at Silicon Valley Bank when regulators seized it. USDC fell to $0.88 before the FDIC's deposit backstop announcement returned it to $1.00. The peg held — but only because of government intervention that wasn't guaranteed when the crisis began.

Both issuers maintain blocklists of wallet addresses that cannot transfer tokens. This is regulatory compliance, not a bug — but it contradicts the permissionless premise of crypto.

Overcollateralized (DAI / USDS)

No bank account. No custodian. Peg maintained entirely by on-chain mechanics: users lock collateral worth more than the DAI they borrow, making undercollateralization structurally impossible under normal conditions.

DAI (MakerDAO / Sky) has been running continuously since 2017. To mint DAI, you open a Vault and lock collateral at a minimum ratio — typically 150% for ETH. If collateral value falls below the liquidation threshold, the protocol automatically liquidates the position to repay the debt. No discretionary intermediary; the math enforces solvency.

The Stability Fee (interest rate on DAI borrowing) and DAI Savings Rate (yield paid to locked DAI) give MKR governance continuous control over DAI supply and demand without off-chain intervention.

Multi-collateral DAI now includes USDC via the Peg Stability Module (direct 1:1 swaps), which provides a hard peg anchor but means roughly half of DAI's collateral is now USDC — introducing the same issuer and regulatory risks the protocol was designed to avoid.

Black Thursday test (March 12, 2020): ETH dropped ~50% in hours. Liquidation bots failed under network congestion; some won collateral auctions with $0 bids. Protocol accumulated $8.3M in bad debt, covered with a dilutive MKR auction. DAI briefly touched $1.06 but never broke significantly. The protocol survived, patched its auction mechanism, and has operated continuously since.

Algorithmic (the failure case)

No reserve, no collateral. Stability maintained by minting and burning a secondary token to absorb price pressure. The appeal: a $1 token with no bank account, no custodian, no reserve to trust. The problem: the stabilization mechanism becomes the mechanism of destruction when confidence fails.

TerraUSD (UST) collapsed in May 2022. UST maintained its $1 peg through a mint/burn relationship with LUNA: burn $1 of LUNA to mint 1 UST; burn 1 UST to receive $1 of LUNA. With $18B of UST in circulation and Anchor Protocol offering 20% APY on UST deposits, when large positions began unwinding, the peg slipped to $0.98. Holders redeemed UST for LUNA at the protocol rate. LUNA's price fell as more was minted — requiring even more LUNA per redemption — which pushed it lower — requiring still more. UST went from $1.00 to $0.01 in 72 hours. LUNA went from $80 to fractions of a penny. $40 billion in combined value was destroyed.

The death spiral is not a bug in any specific implementation. It is intrinsic to the architecture: the token designed to stabilize the stablecoin is precisely the token whose value collapses in the crisis the mechanism is supposed to handle. Every seigniorage stablecoin (ESD, Basis Cash, Dynamic Set Dollar) followed the same path. None have succeeded.

Yield-Bearing Stablecoins

Stablecoins that pay yield to holders without requiring active staking. The yield is embedded in token appreciation or rebasing. The catch: yield always comes from somewhere.

sDAI — Spark Protocol wraps DAI at the DAI Savings Rate. If the DSR is 5%, sDAI appreciates 5% annually against DAI, no action required beyond the initial wrap. Yield source: stability fees paid by DAI borrowers. Risk: MakerDAO smart contract risk only.

USDe (Ethena) — delta-neutral strategy: hold long spot crypto (ETH, BTC) + short perpetual futures of equal size. Funding rate paid on the short generates yield; ranged 15–30% in the 2024 bull market. Risks: exchange counterparty risk (positions held at Binance, OKX, Bybit), and funding rates go negative in bear markets. USDe is a structured product, not a reserve stablecoin.

USDY / USDM (Ondo Finance, Mountain Protocol) — tokenized short-term US Treasuries paying approximately the risk-free rate (~5% in 2024). Regulated issuers, KYC required, not available to US retail. Classic RWA structure: TradFi yield via DeFi rails, TradFi compliance still attached.

Risks

Custodian / counterparty: fiat-backed tokens are only as good as the custodian. The SVB event demonstrated this risk concretely. DeFi wrapper provides no additional protection against custodian failure.

Bank contagion: Silvergate and Signature Bank collapses cut off 24/7 dollar settlement rails the crypto industry relied on — not because reserves were held there, but because the banking infrastructure was fragile.

Regulatory: issuers can freeze tokens (USDC, USDT maintain blocklists), regulators can require reserve changes, and KYC requirements can expand at any time.

Algorithmic death spiral: architecturally inevitable for any reserve-free seigniorage design. No algorithmic stablecoin without real reserves has survived a sustained confidence crisis.

Related

real-world-assets · dao-governance · eigenlayer · rollups · ethereum-l2s

Sources