Crypto Report
Thursday, 19 November 2026
Stablecoins

Stablecoin Depegs: Why USDC Broke and USDT Held

Tether and Circle froze more than $2.3 billion of stablecoins tied to the $6.3 billion Binance hack on 29 October, and neither USDT nor USDC fell below $0.998. In March 2023, $3.3 billion stuck at one failed bank sent USDC to 86 cents, and in October 2025 Ethena’s USDe printed $0.65 on Binance while its redemptions cleared $2 billion on other venues.

All three reached readers as stablecoin depegs, and each one broke in a different place. USDC lost its redemption window for a weekend, USDe lost one exchange’s price feed for 40 minutes, and in the hack the issuers’ primary markets never closed. Four practitioners, the Curve Finance team, a Federal Reserve study, a New York Fed analysis and a joint BIS and IMF paper show how that difference decides whether stablecoin depegs close in hours, spread to other coins or reach currency markets.

USDC Hit 86 Cents While Circle’s Redemption Window Was Shut

Silicon Valley Bank entered receivership at 11:37 a.m. ET on Friday, 10 March 2023, and USDC holders began redeeming with Circle before the issuer said anything, according to a Federal Reserve note published in December 2025. Circle disclosed at about 10 p.m. that $3.3 billion of its reserves, around 8% of the total, was stuck at SVB. Redemptions then fell to minimal levels over the weekend, because Circle processed them during US banking hours.

The Fed’s authors note that Circle’s equity at the end of 2023 was $0.34 billion, just over a tenth of the deposit held at SVB, so without a public guarantee Circle faced the possibility of a significant shortfall. Circle pledged on Saturday to cover any gap with outside capital, and the price steadied without returning to $1.

With Circle’s primary market closed, selling moved to exchanges, where hourly USDC volume reached nearly $2 billion on 11 March and the token traded at 86 cents at its trough. Decentralized exchanges normally handle less USDC trading than centralized ones, yet they carried most of the activity that weekend. Coinbase paused conversions between USDC and dollars over the weekend.

Tim Sun, senior researcher at HashKey Group, said desks working through stablecoin depegs spend the first hour comparing executable prices and order-book depth across major exchanges and on-chain pools. His example is a stablecoin quoted at $0.98, a 2% discount, where a $1 million sell order fills at an average of $0.96 and the real loss is 4%.

“Timewise, trading desks can typically identify price discrepancies across different exchanges within minutes, but accurately determining the root cause of the spread often takes much longer.”

Trace Finance co-founder and CEO Bernardo Brites said losses in stablecoin depegs depend on where each holder can sell, and in March 2023 the dollar exits closed first.

“Where you traded determined what you lost. Coinbase had to pause USDC-to-USD conversion until Monday because its dollar exit was a bank that was closed.”

The Sunday Backstop Moved USDC Before Redemptions Reopened

Treasury, the Federal Reserve and the FDIC announced a backstop for SVB’s insured and uninsured depositors at 6:15 p.m. ET on Sunday, 12 March. USDC’s price “recovered sharply after the backstop announcement on Sunday and fully recovered once Circle began processing redemptions on Monday, March 13,” the Fed note says. Between Monday morning and Wednesday, 15 March, Circle redeemed $3.8 billion of USDC and minted $0.8 billion.

Sun, who treats reserve quality and redemption access as separate layers in stablecoin depegs, reads that sequence as two repairs carried out a day apart.

“First, the government’s deposit guarantee mitigated fears of potential losses on USDC’s reserve assets, prompting investors to reassess USDC’s value before redemptions had even resumed. Meanwhile, the reopening of redemption channels restored the mechanism enabling arbitrageurs to convert USDC into USD at close to $1 face value.”

The $2 billion of hourly volume on 11 March did not hold the price, and Sun cites that as evidence that turnover and depth are separate measures in stablecoin depegs.

“Even with hourly trading volumes approaching $2 billion, USDC still fell to $0.86. This indicates that heavy trading activity does not necessarily mean there is sufficient liquidity to absorb concentrated selling pressure.”

CoinGecko’s aggregated USDC price, which blends many venues, bottomed near $0.877 on 11 March, about two cents above the trough in the Fed’s data. Both series show the same two steps back to $1, and the two-cent gap between them reflects how stablecoin depegs print a different low on each venue.

Investor Takeaway

Dai’s Peg Stability Module Carried the USDC Run Into Three More Stablecoins

MakerDAO’s Peg Stability Module swapped USDC for newly minted Dai one for one, and once Circle’s window shut and exchanges stopped taking USDC at par, traders used it as an exit, the Fed found. Around 1 billion USDC went into the module on each of 10 and 11 March, and its 950 million daily cap was reached both days. Dai’s price fell in step with USDC while the module was open, and a gap opened on 11 March after the cap was hit and selling moved to other channels.

“The USDC-PSM worked as expected, just not as intended,” the Fed’s authors wrote. Dai also ran modules for Paxos’s USDP and Gemini’s GUSD, and holders used them to swap weakened Dai for those coins, so the USDC run became four stablecoin depegs. More than 400 million USDP left its module, over half of USDP’s supply, and USDP fell to about 91 cents while GUSD dropped to about 96 cents.

USDT and Binance USD traded marginally above $1 over the same weekend, as money left the coins caught in the four stablecoin depegs for stablecoins with no SVB exposure. Connor Sullivan, chief strategy officer at Firelight Finance, said stablecoin depegs travel along three channels, and the last of them describes the Dai chain.

“A depeg spreads along three channels. In DeFi, the stablecoin is posted as collateral, so a price drop triggers liquidations and pulls liquidity out of lending markets. On exchanges, it hits margin systems and any oracle that reads the exchange’s own price. In fiat markets, it becomes redemption demand on the issuer’s banking partners. Holders who rotate into another stablecoin pass the pressure on to it.”

Dai’s slide into USDP and GUSD in March 2023 is a documented case of stablecoin depegs passing from coin to coin, with each one handing the selling to the next through a contract built to defend a peg.

USDe Fell to $0.65 on Binance While Its Redemptions Cleared $2 Billion

Ethena’s USDe holds its peg with staked Ethereum and short perpetual futures positions, and the funding-rate reversal on 10 October 2025 cut its market cap by more than 13%, according to a June 2026 New York Fed analysis. On Binance, USDe printed $0.65 that night, and Binance later compensated users whose USDe collateral was liquidated between 21:36 and 22:16 UTC, a 40-minute window.

Ethena founder Guy Young said $2 billion of USDe was redeemed within 24 hours across venues including Curve, Fluid and Uniswap, with price deviations of 30 basis points or less, according to Cointelegraph. Young blamed Binance’s use of its own order book as the price oracle and its deposit and withdrawal problems, which “did not allow market makers to close the loop.” The New York Fed authors, citing CoinDesk, describe the same oracle problem.

Chainlink’s USDe/USD feed, which draws on several venues, moved only about 70 basis points, LlamaRisk wrote on the Aave governance forum, and Ethena’s on-chain redemptions held around $0.9990 after a 10-basis-point fee. Aave priced USDe on par with the USDT feed, which shielded its borrowers from the Binance print and from the liquidations that follow stablecoin depegs on venues that mark collateral to their own books.

The Curve Finance team, which supplied written comments and trade data on stablecoin depegs for this piece, said its USDe/USDC pool traded near par through Binance’s window, though one later swap in the same pool cleared well outside Young’s 30-basis-point range.

“On 10 October, USDe fell to $0.65 on Binance. In the same 40-minute window, six swaps in a Curve USDe/USDC pool executed at around $0.9999, although a later Curve trade cleared at $0.927. The price on one venue was not the stablecoin’s price everywhere.”

Carlos Vendrell, founder and lead smart contract auditor at Zealynx Security, said lending protocols can separate venue errors from genuine stablecoin depegs while the price is still moving.

“Which price a protocol reads is one of the first things we check in a smart contract audit, because it decides whether a depeg is real for the system. You can tell in real time. If the mint and redeem path works and the deep onchain pools agree, the print is the venue.”

Aave’s par pricing carried its own exposure, since LlamaRisk warned in the same post that it would delay liquidations and leave Aave with bad debt if USDe’s backing ever failed. Vendrell framed the trade-off as a choice every lending market makes in advance of stablecoin depegs.

“Pricing at par is a bet too. LlamaRisk warned it could delay liquidations and leave bad debt if the backing failed. Every price feed picks its failure: one venue overreacting, or a peg that never reacts.”

Sun counts USDe’s episode among the stablecoin depegs driven mostly by venue pricing and liquidity, with the funding-rate reversal as backdrop, and he warns that the line between venue and asset can move during the event.

“A transient 35% discount on a single platform is not necessarily more dangerous than a 5% discount that continues to widen and cannot be fixed via redemptions. Once a localized price anomaly is adopted by oracles in other trading or lending protocols, it can trigger cascading liquidations and concentrated redemptions.”

Sullivan applies the same venue-or-asset test to insurance written against stablecoin depegs, where the trigger in the policy wording decides what is actually being covered.

“A deviation on one exchange, with redemptions open and other venues near par, points to the venue. A deviation that shows up everywhere at once, with redemptions slowing or paused, points to the asset. Cover wording has to draw the same line. A policy that pays out on a single-venue print is insuring an oracle design choice, not the stablecoin itself.”

The New York Fed authors argued that USDe “played a material role in amplifying the decline in ether prices.” A reader comment published under the post disputes that, saying Ethena held more than $9 billion of stablecoin collateral for redemptions and that the liquidations came from USDe posted as collateral on Binance. Holders left either way, as USDe’s market cap fell from about $14.7 billion on 9 October to roughly $6.4 billion by late December.

Tether and Circle Froze $2.3 Billion and Both Pegs Held

USDT and USDC kept their pegs through the stablecoin depegs of 10 October 2025, when USDC lost $280 million of market cap in a day and USDT gained $282 million, the New York Fed authors found. The 29 October Binance hack tested the same pair against a different trigger for stablecoin depegs.

Tether said it froze more than 1.567 billion USDT and Circle more than 800 million USDC, and Binance said at 00:52 UTC on 30 October that about $3 billion had been frozen in total. On 12 November, Binance said it had covered all $6.34 billion of affected assets from its own funds, including its SAFU reserve.

USDT’s lowest price in the 18 hours after the theft began was $0.9980 and USDC’s was $0.9987, Webacy found, so the freezes set off no stablecoin depegs in either coin.

The Curve Finance team said trading in its largest stablecoin pool during the hack window ran heavily in both directions, while one of its automated defenses against stablecoin depegs ran out of room.

“During the 24-hour stress window that took place on October 29-30, 2026, Curve’s 3pool processed $402 million of swaps between DAI, USDC and USDT, more than 21 times its usual daily volume. Its net swap flows were much smaller, as traders supplied about 8.23M USDT and took out 4.29M USDC and 3.78M DAI. The pool supported substantial trading in both directions, rather than simply processing a one-way rush for the exit. That said, the same period also showed that there are limits to what automated peg-management mechanisms can accomplish. As crvUSD traded slightly below peg, the crvUSD/USDT PegKeeper withdrew 9.55M previously supplied crvUSD from the pool in 14 actions. That keeper’s debt reached zero, leaving it with no more capacity for withdrawals.”

Sullivan said the two coins came through both events on the strength of their primary markets, and he counts the freezes as a second layer of protection against stablecoin depegs that carries its own cost.

“USDT and USDC held through October 2025 because their pegs sit on the primary market: institutions that can redeem at par directly with the issuer, backed by reserves the market trusted. The reported $2.3bn freeze after the Binance hack shows a second form of insulation, which is issuer control. Freezing tainted funds protects the peg, but the loss moves onto whoever holds the frozen balance. Insulation fails when the doubt is about the reserves or the issuer itself, because then the primary market is what breaks.”

Live market data shows only part of what decides whether stablecoin depegs spread, according to Sun, because the state of an issuer’s reserves sits with banks and custodians.

“Whether underlying reserves are intact and whether assets have potential losses often involve off-chain information from banks and custodians, which cannot be definitively answered by real-time market data alone.”

The BIS Finds Stablecoin Depegs Multiply Spillovers Into Currency Markets

More than 70% of cumulative net inflows into stablecoins come from currencies other than the dollar, so most stablecoin purchases are foreign-exchange trades, according to a joint BIS and IMF working paper by Iñaki Aldasoro, Paula Beltrán and Federico Grinberg. That gives stablecoin depegs a direct route into currency markets. The authors studied four dollar stablecoins, 27 currencies and 64 exchanges from 2021 to 2025. They estimate that a 1% rise in net stablecoin inflows widens parity deviations by about 40 basis points, weakens the local currency by 5 basis points and widens the short-term dollar premium by 5 to 10 basis points.

Halving the frictions between stablecoin and traditional currency markets would cut the dollar-funding spillover by roughly half and the currency effect by about a third, the authors estimate. Their stress case covers stablecoin depegs directly. When inflow shocks coincide with redemption frictions, “as during the Terra/Luna collapse or the SVB-induced USDC depeg,” the interaction between redemption costs and depleted intermediary capital “can multiply baseline spillover effects severalfold,” they write.

Brites reads the same stablecoin depegs the other way, from the position of a firm that converts reais into stablecoins, and he starts with the intermediary at the center of the BIS model.

“The BIS model turns on an actor it never names: the balance-sheet-constrained intermediary between stablecoin demand and spot FX.”

His disagreement with the BIS authors is about the direction of the effect while the peg is broken, and both place the lasting damage from stablecoin depegs on intermediaries’ balance sheets.

“A depeg interrupts the channel rather than amplifying it. The transmission assumes the stablecoin is a dollar. When that breaks, demand does not leave dollars for reais, it rotates coins, as the flight from USDC into USDT at a premium showed in March 2023. Where a depeg bites is inventory: whoever warehouses the broken coin against local-currency liabilities has lost its hedge, which is the BIS’s own result that spillovers grow when intermediaries take losses.”

Brazil shows how much money now runs through the channel that stablecoin depegs would hit, as Brazilians bought $14.68 billion of crypto in the first half of 2026. That was up 135% from $6.24 billion a year earlier, with stablecoins making up more than 90% of identified demand, according to central bank data reported by Valor International. The IMF’s July financial system stability assessment for Brazil found stablecoin purchases two to three times more sensitive to global shocks than portfolio or direct investment flows, Cointelegraph reported, a pattern that fits stablecoins’ role as Latin America’s route to dollars.

Brazil has also narrowed which intermediaries carry the flow, and so which firms would hold the inventory in stablecoin depegs. Since 2 February, Resolution 521 has capped international transfers with virtual assets at the equivalent of US$100,000 when the counterparty lacks authorization to trade foreign exchange. From 1 October, Resolution 561 barred eFX payment providers from settling cross-border payments in stablecoins, Global Finance reported. Rules on which firms may carry stablecoin flows also differ across the US, EU, UK and Asia.

Brites, whose Trace Finance routes conversions through a bank authorized by Brazil’s central bank, said central banks watching for stablecoin depegs get more from conversion data at licensed institutions than from offshore order books.

“What a central bank can read is net conversion flow at licensed institutions and the gap between the local stablecoin price and the onshore rate.”

Investor Takeaway

Redemption Flows Show Stablecoin Depegs Before the Price Does

Primary-market redemptions of USDC rose sharply after SVB entered receivership and peaked shortly before Circle’s 10 p.m. announcement, the Fed note shows, and “some (possibly more informed) traders ran ahead of Circle’s own announcement.” Price lagged in USDC’s case, and it fell only after the window closed.

LlamaRisk treated Ethena’s on-chain redemption rate, steady around $0.9990 through the Binance print, as the better gauge of value when stablecoin depegs hit a single venue. Binance reached a similar conclusion afterward and added USDe’s redemption price to the index it uses to mark the token.

Brites said traders can sort stablecoin depegs in minutes with three checks that start with venues and end with the reserves behind the token.

“You can tell in minutes. Three checks: is the discount the same across venues and in the deepest on-chain pool; is the primary market open and clearing at par; is there reserve news that explains the size.”

Sun puts the redemption mechanism above the price print as the signal that matters most in stablecoin depegs.

“The most critical risk signal during a stablecoin depeg is not how far the price deviates from $1, but whether its underlying asset value and redemption mechanism remain reliable.”

The record on lasting losses points the same way, as solvency failures caused 72% of the $2.5 billion destroyed across 36 loss-bearing stablecoin collapses from 2022 through 2026, according to Webacy research.

Originally reported by FinanceFeeds.