explains
De-pegging Events & Systemic Risk
Stablecoins de-peg. UST collapsed entirely. USDC briefly hit $0.87. Each event has a different cause, a different recovery profile, and a different lesson.
Published
Stablecoins de peg, but the instrument is not the same as the issuer and the issuer is not the same as the reserve.
Reader Brief
Stablecoins de-peg. UST collapsed entirely. USDC briefly hit $0.87. Each event has a different cause, a different recovery profile, and a different lesson.
Reading Guide
Four moves that turn de-peg events from undifferentiated panic into a calibrated risk category.
A de-peg is a price-deviation event with mechanics - not a single failure mode.
The peg-keeping mechanism is mint/burn arbitrage: when price drops below 1.00, authorized counterparties buy on secondary market and redeem at 1.00 with the issuer for arbitrage profit; when price rises above 1.00, they mint at 1.00 and sell on secondary market. The buy/sell pressure restores the peg. The peg breaks when this mechanism is impaired: redemption suspended, mint suspended, reserve quality impaired, or panic exceeds arbitrage capital. Each impairment has a different signature, a different recovery profile, and a different lesson. Treating all de-pegs the same is the most expensive risk-policy mistake.
UST 2022 was algorithmic collapse - the failure mode does not generalize to reserve-backed stablecoins.
UST was algorithmic: it was not backed 1:1 by reserves. Its peg was maintained by an arbitrage relationship with LUNA, the Terra blockchain native token. When UST traded below 1.00, arbitrageurs were supposed to burn UST, mint LUNA, and sell. The collapse mechanism: large UST sales drained Curve liquidity -> UST de-pegged -> holders rushed to redeem UST -> LUNA -> LUNA supply hyperinflated -> LUNA price collapsed -> arbitrage stopped working -> UST collapsed. $60B destroyed in 72 hours. USDC, USDT, EURC, and most production stablecoins are reserve-backed - they do not have UST failure mode, LUNA hyperinflation, because they are not backed by another volatile token. "A stablecoin de-pegged catastrophically" is true; the failure mode does not transfer.
USDC 2023 was a liquidity de-peg, not solvency - reserves intact, peg restored within hours of federal action.
Friday March 10, 2023: SVB enters FDIC receivership. $3.3B of USDC reserves were held at SVB. USDC traded down to about $0.87 over the weekend. Sunday March 12: Treasury, Fed, and FDIC announced SVB depositors would have access to all funds on Monday. Circle separately said the SVB reserve deposit would be fully available when banks opened [1][7]. Duration of de-peg: roughly a weekend. Maximum deviation: about 13%. Holders who retained redeemable USDC saw par convertibility return; holders forced to sell or liquidate during the window could still suffer permanent position-level losses. The de-peg was operational and confidence-driven, not the UST-style structural failure. Circle could not access part of the cash reserve over the weekend. Once federal action confirmed deposit access, redemption confidence returned and the peg restored. Liquidity de-pegs recover fast when the underlying assets are intact; structural de-pegs do not recover at all.
Four severity factors govern recovery - operators control three of four.
**Reserve quality and concentration:** high-quality, diversified reserves recover faster. Single-custodian concentration creates de-peg risk independent of asset quality. **Redemption mechanism robustness:** 24/7 redemption with diverse banking partners holds the peg under stress; single-banking-partner access fails. **Arbitrageur capital depth:** deep market makers such as Cumberland, Wintermute, and Jane Street absorb hundreds of millions in deviation; shallow market makers cannot. **Issuer communication:** Circle Friday-night SVB disclosure was a model - explicit reserve breakdown, named exposure, immediate communication. Issuers that withhold during stress experience deeper, longer de-pegs. Operators control reserve quality, redemption mechanism, and communication directly; arbitrageur depth is downstream of issuer maturity. Three of four factors are governance choices, not market conditions.
What a De-Peg Actually Is
A secondary-market price deviation with specific mechanics, not a single generic failure mode.
A stablecoin de-pegs when its secondary market price diverges meaningfully from the 1.00 reference. "Meaningfully" varies by counterparty: 0.5% deviation may be tolerable for retail; 0.1% deviation may break institutional models.
The peg-keeping mechanism
If price is below 1.00, authorized counterparties buy on secondary market and redeem at 1.00 with the issuer for arbitrage profit. This buying pressure pushes price up. If price is above 1.00, authorized counterparties mint new tokens at 1.00 and sell on secondary market for arbitrage profit. This selling pressure pushes price down. The peg holds when the arbitrage mechanism functions. It breaks when the mechanism is impaired: redemption suspended, mint suspended, reserve quality impaired, or panic exceeds arbitrage capital.
UST 2022: Algorithmic Collapse
The largest stablecoin failure was a structural algorithmic failure, not a reserve-backed issuer liquidity event.
The largest and most consequential stablecoin failure in history happened in May 2022. Roughly $60B in market value evaporated within 72 hours [2][3].
What UST was
UST, TerraUSD, was an algorithmic stablecoin: it was not backed 1:1 by reserves. Instead, its peg was maintained by an arbitrage relationship with LUNA, the Terra blockchain native token. Burning 1 UST minted $1 worth of LUNA; burning $1 worth of LUNA minted 1 UST. When UST traded below 1.00, arbitrageurs were supposed to buy UST cheap, burn it, mint LUNA, and sell. This mechanism worked when LUNA had reliable market value.
The collapse mechanism
May 7-9, 2022: large UST sales drained Curve pool liquidity. UST de-pegged to 0.98 then 0.95. Holders rushed to redeem UST into LUNA via the burn mechanism. LUNA supply exploded through hyperinflation. LUNA price collapsed from $80 to $0.0001 within 72 hours. The arbitrage no longer worked: arbitrageurs would receive worthless LUNA for UST. UST collapsed to below $0.10 and never recovered. Final outcome: ~$60B in combined UST/LUNA market cap destroyed. Multiple lender bankruptcies, including Celsius, Three Arrows Capital, and Voyager, followed within weeks.
Why UST does not generalize
UST was algorithmic. USDC, USDT, EURC, and most production stablecoins are reserve-backed. Reserve-backed stablecoins do not have UST failure mode, LUNA hyperinflation, because they are not backed by another volatile token. The lesson: "a stablecoin de-pegged catastrophically" is true and important, but the failure mode does not transfer to reserve-backed stablecoins. Risk policy should distinguish.
USDC 2023: Reserve-Location Risk
The most consequential test of reserve-backed stablecoin resilience was a liquidity event, not a solvency failure.
The March 2023 USDC de-peg was the most consequential test of reserve-backed stablecoin resilience. Circle had $3.3B of exposure at Silicon Valley Bank [1].
The sequence of events
Friday March 10, 2023: SVB enters FDIC receivership. Friday evening: Circle discloses $3.3B of USDC reserves held at SVB. USDC trades down to about $0.87 over the weekend. Sunday March 12: US Treasury, Fed, and FDIC announce all SVB depositors will have access to all funds on Monday [7]. Circle states that the $3.3B reserve deposit will be fully available when banks open, and secondary-market confidence returns [1]. Duration of de-peg: roughly a weekend. Maximum deviation: about 13%. Holders who retained redeemable USDC saw par convertibility return; holders forced to sell at the lows or liquidated by venue/oracle mechanics could still suffer permanent losses.
Why this was different from UST
The reserves were always there. SVB held the cash; the cash was deposit-insured plus protected by federal action. The de-peg was a liquidity issue, not a solvency issue. Circle could not move funds out of SVB over the weekend, so could not honor redemptions, so secondary market priced uncertainty. Once federal action confirmed deposit access, redemption capability returned and the peg restored within hours. The lesson: reserve-backed stablecoins can de-peg from operational issues, including custody concentration, without losing solvency. Recovery is fast when the underlying assets are intact.
Structural vs Liquidity De-Pegs
Not all de-pegs are equally severe. The category determines the recovery path.
The category is not just terminology. It determines whether holding through is rational, whether liquidation is urgent, and whether the issuer has a recovery path.
| Type | Cause | Recovery profile | Examples |
|---|---|---|---|
| Structural | Reserve insolvency or impaired arbitrage mechanism | Often no recovery; total loss possible | UST 2022; Iron Finance 2021 |
| Liquidity | Reserve operationally inaccessible; redemption suspended | Fast recovery once access restored | USDC 2023, SVB; Tether 2018 brief deviations |
| Sentiment | Market doubts despite intact reserves | Recovery as confirmation of reserves; days to weeks | Tether 2022, post-Terra; USDD periods |
| Issuer-specific | Operational issue at issuer, such as legal action, key compromise, or freeze of issuer accounts | Variable; depends on resolution | Less common; risk-managed by issuer governance |
Why the category matters for risk policy
Structural de-peg -> zero recovery. Position should be liquidated immediately at any price. Liquidity de-peg -> high recovery. Holding through is rational; selling at the bottom is the worst outcome. Sentiment de-peg -> recovery likely if reserves verified. Hold-or-sell decision depends on issuer transparency. Issuer-specific -> case by case. Diversify exposure across issuers. A risk policy that treats all de-pegs the same is structurally wrong. The de-peg category drives the response.
Four Determinants of De-Peg Severity
Why some de-pegs recover in hours and others destroy the issuer.
Four factors explain why one de-peg can restore quickly while another collapses permanently. Three of the four are governance choices operators can influence directly.
Factor 1: Reserve quality and concentration
Stablecoins backed by high-quality, diversified reserves recover faster. USDC reserves were almost entirely T-bills and bank deposits in March 2023 - the only issue was concentration of cash at SVB. Tether reserves in 2022 were more diverse but less transparent, which created sentiment exposure. A stablecoin with concentrated reserves at a single custodian carries de-peg risk independent of asset quality.
Factor 2: Redemption mechanism robustness
Redemption that is backed by diversified banking access holds the peg better under stress. Redemption that depends on a single banking partner, or on banking-hour windows that cannot meet weekend panic, is exposed to liquidity gaps. Post-SVB, Circle announced new automated minting and redemption arrangements through additional banking partners [1].
Factor 3: Arbitrageur capital depth
The peg holds when arbitrageurs can absorb selling pressure with their own capital while waiting for redemption. Deep-pocketed market makers, including Cumberland, Wintermute, and Jane Street, absorb hundreds of millions in deviation. Shallow market makers cannot. This is why mature stablecoins have stronger pegs: the market maker ecosystem around them is larger and better capitalized.
Factor 4: Issuer communication
Circle Friday-night SVB disclosure was a model: explicit reserve breakdown, named exposure, immediate communication. The de-peg deepened until federal action; then it reversed quickly because there was no information vacuum. Issuers that withhold information during stress experience deeper and longer de-pegs. Tether 2018 disclosure issues are case in point.
What This Means for Risk Officer Policy
Three practical recommendations follow from the de-peg taxonomy.
The taxonomy is useful only if it changes policy. For risk officers, the practical move is to separate stablecoin exposure by mechanism, issuer concentration, and reserve transparency.
Evidence And Sources
This raw HTML export preserves source visibility for crawler and contractor review. Indexing decision: index, follow.
- $3.3 Billion of USDC Reserve Risk Removed, Dollar De-peg Closes - Circle
- The Trades That Triggered TerraUSD's Collapse - Chainalysis
- Public information and stablecoin runs - BIS
- High-level Recommendations for Stablecoin Arrangements - Financial Stability Board
- Policy Recommendations for Crypto and Digital Asset Markets - IOSCO
- The Anatomy of UST - Stanford Blockchain Lab
- Joint Statement by Treasury, Federal Reserve, and FDIC - Federal Reserve