One grave takes Tab; the arrow keys walk the rest: left and right through a section by date of death, up and down to the nearest date in the next section, Home and End to a section's first and last. Enter or Space pins a grave's record; Escape closes it.
Each stone is one coin. Sections show why; rows show when; bigger monuments mean bigger recorded peaks. Select a cause sign to look closer.
Shape = cause of death
Abandoned
Counterparty Failure
Liquidity Drain
Algorithmic Failure
Regulatory
Stone shape is cause of death, not what holders recovered. Sections run from Abandoned at the front to Regulatory under the sea wall.Columns run from Abandoned on the left to Regulatory on the right.
Plinth steps = peak market cap
< $10M$10M$100M$1B$10B+
Peak not recorded= hatched face, outlined plinth, neutral height
Each step is ×10. At $10B and above the architecture carries the class: TerraUSD's snapped column is the tallest monument, Binance USD's sealed mausoleum the largest. Peak market cap is not what holders lost.
Year blocks, weathering and other marks
Position and marks
Year blocks= year of death, newest at the gate, oldest under the lighthousenewest at the top, each row's count under its year
Weathering= years since death
Fresh soil= died within 90 days of the latest recorded death
Bronze plaque= Pharos holds a frozen data page
Footstone glyph= peg other than USD
Cypress height is deaths that year. Hover a stone to read it; select it to pin its record. Keyboard: Tab into the map, then the arrow keys move from stone to stone; Enter pins, Esc closes.Tap a stone to open its record. Keyboard: the arrow keys move from stone to stone; Enter opens, Esc closes.
Key facts
12 months to Oct 2026
37
21 of these were coins Pharos tracked live (2 in the prior 12 months). Curated records: 16 vs 19.
Held by two coins
71.0%
BUSD and UST: $42.2B of $59.4B combined peak (104 of 114 recorded).
Median peak
$49.5M
Half of the recorded deaths peaked below this. 5 peaked at $1B or more.
Latest record added Oct 19, 2026 · Latest recorded death Oct 5, 2026 · Dataset schema 1.1 · checksum 16bcc87e
Causes of death
How stablecoins die
Each record carries one primary cause. Most deaths are small; most of the recorded peak market cap sits in a few large failures.
Abandonment is the most common cause
44 of 114 records (39%) ended with the issuer or protocol no longer maintaining the coin. Their median recorded peak market cap was $30.5M. Show these records in the register
USDKG announced on October 5, 2026 that issuance, circulation and blockchain operations would cease under Kyrgyz Cabinet Order 639-t dated August 20. The order directed liquidation of its issuer, EVA. Holders were directed to [email protected] to arrange fiat or USDT exchange, with no published payout duration. The issuer announced plans to consolidate and burn tokens and suspend the Ethereum and Tron contracts; completed burns and full holder repayment have not been established. Pharos preserves the token's historical market record as a frozen archive.
cdxUSD was Cod3x's GHO-style credit dollar on Base, where a 3-of-9 facilitator Safe held the only mint and burn keys and ordinary holders were never given a redemption route. The peg broke on August 27, 2026, when the hourly series first printed $0.8724 and both CoinGecko's daily close and the Aerodrome cdxUSD/WETH pool ended below $0.90; the price fell through $0.54 on August 28 and bottomed at $0.2457 on September 17. Supply never moved, holding at exactly 6,150,000 cdxUSD since January 2026 with the facilitator bucket level equal to total supply, so no token was burned at par during the collapse. Cod3x had already stopped supporting it: cod3x.org, its documentation index, and the technical article Pharos cites now describe only AI perp-trading agents, and that article returns 404. With roughly $23 of daily volume left across pools whose deepest executable side holds about $1,600, and DefiLlama flagging the project dead, Pharos freezes cdxUSD as an archived abandoned asset rather than an active stablecoin.
XT.com's official xtusd_usdt market stopped updating on July 18, 2026: the exchange ticker froze at that timestamp, the order book emptied, and the pair no longer appears in the exchange's symbol list. CoinGecko ended the feed on July 20 with an implausible tail near $0.04, and CoinMarketCap lists the asset inactive. XTUSD circulates on XSC, which Pharos does not index, so no contracts are configured and no identity-safe DEX route can be expressed. With the last admissible price at 0.93205 on July 20 and roughly $1.25M still circulating, Pharos freezes XTUSD as an archived market-ended asset rather than an active stablecoin.
Mu Digital announced an orderly wind-down of AZND, loAZND, and muBOND on July 10, 2026, and opened its redemption portal on August 7, 2026. Secondary trading had already collapsed: Pharos accepted its last admissible price, 0.19444 from the thin Curve pool, on July 20, 2026, and the residual pool now holds roughly $767 of liquidity near $0.03. CoinGecko stopped updating the feed on July 10 and DefiLlama delisted the project, leaving no admissible market or provider path. Pharos freezes AZND as an archived wind-down asset rather than an active stablecoin.
Once one of DeFi's largest CDP stablecoins, MIM peaked above $3.6B before repeated cauldron exploits and years of thinning liquidity hollowed out its market. The final break began on June 8, 2026. Abracadabra injected liquidity, offered SPELL incentives, raised rates across every cauldron, and suspended Curve bribes to force debt repayment, but none restored the peg. By July 26 MIM traded near $0.09 with roughly $3M circulating after seven weeks below peg, leaving no credible liquidity or arbitrage path back to one dollar.
Ebisu offered fixed-rate, open-term credit through ebUSD, a Liquity v2 fork minting against weETH, sUSDe, WBTC, LBTC, and stcUSD branches natively on Ethereum and Plasma. Launched in June 2025, it never found scale: supply peaked near $815K and sat below $25K by July 2026. The team explored just-in-time liquidations, multichain issuance, structured looping and leverage vaults, and acquisition options, but could not grow ebUSD liquidity enough to support meaningful borrowing activity. On July 29, 2026 Ebisu announced it would wind down: minting was paused, users were asked to close Troves and withdraw Stability Pool deposits and DEX liquidity, and the frontend will be deprecated on October 30. An orderly, solvent sunset. Collateral redemption remained open throughout.
Orki USDK was deployed only on Swellchain. Swellchain shut down its bridge flow in June 2026 and moved to halt chain operations on June 30, leaving remaining positions dependent on direct contract access or unrecoverable as the network wound down. Because USDK had no verified deployment outside Swellchain, Pharos freezes it as a discontinued single-chain CDP stablecoin.
Main Street's msUSD collapsed after Accountable terminated its verification relationship on June 20, 2026, saying the protocol failed verification standards. msUSD fell far below the dollar and remained around $0.29-$0.33 in the following week while independent on-chain analysts reported minimal immediately redeemable reserves against tens of millions of tokens. Main Street disputed the characterization and said assets remained backed, but the combination of failed third-party verification, unresolved redemption capacity, and persistent market pricing made msUSD unsuitable for active stablecoin tracking.
msY is a strategy-vault wrapper over msUSD, so the June 2026 msUSD collapse impaired the wrapper's durable dollar backing. As msUSD depegged after Accountable terminated reserve verification, msY also traded far below its prior net-asset value and could not recover independently from the parent asset. Pharos freezes msY with msUSD as a derivative collapse rather than treating the vault share as an active stablecoin-like dollar instrument.
Tether began winding down Alloy by Tether and aUSDt on June 17, 2026 after reviewing activity, demand, and product priorities. New positions and minting were closed, while existing users were given until September 17, 2026 to return aUSDt and recover XAUt collateral. The protocol remains available only to unwind existing exposure, so Pharos freezes it as an archived wind-down asset rather than an active issuance venue.
Launched in 2018, sUSD was Synthetix's flagship synthetic dollar, an overcollateralized stablecoin minted against staked SNX. Its peg held for years on a single reflex: whenever sUSD slipped below $1, stakers bought it back at a discount to repay their debt at par. SIP-420 (April 2025) dismantled that reflex, cutting the required collateral ratio from 750% to 200% and pooling debt collectively. Minting got cheap, but no individual was left with a reason to defend the dollar. sUSD fell to $0.68 within weeks and never recovered, drifting below peg for over a year as supply bled from a ~$329M peak toward a functionally insolvent tail near $0.23. In June 2026 Synthetix governance stopped repairing it and passed SIP-423: the contract was frozen, a holder snapshot taken, and sUSD made permanently non-transferable, with remaining holders entitled to 4 SNX per sUSD, locked for one year and then vested over a second year from the freeze. Unclaimed receipts expire worthless after six months. sUSD held in LP pools, vaults or other deposit contracts cannot be recovered directly; those holders go through a separate Treasury claims process using their deposit or LP receipt tokens, evaluated case by case.
Native Markets and Bridge announced USDH's sunset in May 2026 as Hyperliquid markets migrated to USDC. New issuance ended except to support migration liquidity; the USDH dashboard was scheduled to close on July 17, while Bridge continued USDH-to-USDC and fiat redemption for onboarded holders. USDH remained fully backed during the transition, but its role as an active stablecoin product ended, so Pharos freezes it as an archived migration asset.
GYEN was GMO-Z.com Trust's NYDFS-chartered yen stablecoin, backed by FDIC-insured cash and short-dated U.S. Treasury bills, with monthly Network Firm attestations and one of the most tightly regulated issuer stacks of any JPY stablecoin. Its weakness was never custody but demand: thin secondary liquidity and a history of depegs (most infamously a squeeze above 1.5x peg in 2021 that triggered the Donovan v. Coinbase class action) left it trading more like a neglected FX derivative than a stablecoin. On May 15, 2026 GMO-Z.com Trust Company announced an orderly wind-down of GYEN and ZUSD, disabling purchases and opening a redemption-only window through November 11, 2026; parent GMO Internet Group moved to dissolve the U.S. subsidiary and recorded a ¥2.2bn write-down. With issuance ended and only redemption remaining, Pharos freezes GYEN as an archived wind-down asset rather than an active redeemable stablecoin.
Hubble Protocol launched USDH in 2022 as one of the earliest Solana CDP stablecoins, peaking near $14M. By May 2026 CoinGecko reports a $0 market cap, on-DEX volume has collapsed, and the redemption pool no longer offers a credible exit. The asset is treated as wound down; future coverage scans should skip it. Symbol collides with Native Markets USDH (tracked) and Hermetica USDH (tracked).
Paxos completed its acquisition of Membrane Finance in February 2025 to absorb its Finnish EMI license and EU stablecoin issuance rails. As Paxos consolidated its product lineup, the EUROe business line was decommissioned and the platform transitioned to redemption-only mode in May 2026, with no new minting or trading. Existing holders were instructed to complete KYC and redeem outstanding EUROe 1:1 for euros via SEPA through Paxos Issuance Europe.
Quiet exit on Arbitrum: protocol site offline, no on-chain activity
Autopsy · The Fedz FUSD (FUSD)
Quiet exit on Arbitrum: protocol site offline, no on-chain activity
The Fedz launched FUSD as a synthetic dollar on Arbitrum (contract 0x894341be568eae3697408c420f1d0acfce6e55f9) and traded near peg through late 2025. By May 2026 the protocol's homepage no longer resolved, no DefiLlama TVL was indexed, and 24h on-DEX volume had collapsed to ~$100 against ~$1.5M circulating. No exploit or formal wind-down announcement was made. Pharos categorizes this as abandonment based on the loss of project infrastructure and the ~7% sustained depeg with no redemption path.
StablR suspended EURR minting and redemption after a May 24, 2026 cybersecurity incident let an attacker mint 4.5M unbacked EURR as part of a coordinated EURR/USDR exploit. StablR disclosed that the affected tokens were temporarily not backed 1:1 under MiCAR while recovery work continued. With no public restoration of minting, redemption, or full backing confirmed, Pharos freezes EURR as an archived incident asset rather than treating it as an active redeemable stablecoin.
StablR suspended USDR minting and redemption after a May 24, 2026 cybersecurity incident let an attacker mint unbacked USDR as part of a coordinated USDR/EURR exploit. StablR disclosed that the affected tokens were temporarily not backed 1:1 under MiCAR, announced on May 29 that its Board had activated its Recovery Plan, and on August 25 put unauthorised USDR in circulation at about 14.33M while stating that properly issued tokens remained fully backed. With no public restoration of minting, redemption, or full backing confirmed, Pharos freezes USDR as an archived incident asset rather than treating it as an active redeemable stablecoin.
The game-chain dollar stepped aside for bridged USDC
Autopsy · WEMIX Dollar (WEMIX$)
The game-chain dollar stepped aside for bridged USDC
WEMIX transitioned its USD stablecoin activity from WEMIX$ to USDC.e, withdrawing Foundation liquidity from WEMIX$ pools and ending swap support across WEMIX.Fi and WEMIX PLAY in April 2026. Holders retained a conversion route through the WEMIX$ Module and a longer claim window for discontinued liquidity pools, but the ecosystem's active dollar rail became USDC.e. Pharos freezes WEMIX$ as the retired native stablecoin while keeping its residual supply and history available in the archive.
By April 2026, Buck's public site had become a shutdown notice saying the project was closed and funds were returned to token holders, while the former transparency page no longer served a reserve disclosure. Pharos freezes BUCK as a discontinued archive rather than an active yield-bearing stablecoin: the last live cache showed only about $4.6K of residual supply, no reliable price, and no working issuer reserve feed. The failure mode was an orderly wind-down rather than an oracle or collateral break, but live monitoring no longer has an active issuer disclosure surface to follow.
Nerite built USND on immutable Liquity V2-style contracts, but one of its approved collateral branches depended on rsETH. After the rsETH hack impaired that collateral, Nerite could not patch the market away or rewrite the risk controls. Supply collapsed from a $2.7M peak to under $100K, leaving USND as an archived case study in how immutable CDP systems can still inherit external collateral failure when a supported asset breaks upstream.
On 22 March 2026, an attacker exploited Resolv's single-EOA minter: no oracle, no amount cap, no max-mint guard, depositing 100K USDC and receiving 50M USR in return, then iterating until 80M unbacked USR existed for ~$25M of extracted ETH. USR crashed to $0.025 in 17 minutes; the RLP insurance layer absorbed the bad debt and was wiped out, dragging Stream Finance's $17M position down with it. Resolv burned 46M of the unbacked supply via a blacklist freeze, but USR never recovered; by April it traded at $0.16 with $95M of collateral against $173M of liabilities. Delta-neutral collateral and Apostro-attested reserves were no defense against an operator-level privilege failure.
Launched in late 2021 as agEUR, the first fully decentralized euro stablecoin, it briefly held the #1 spot among euro stablecoins with over 180 million tokens in circulation (~$200M). Tether's EURT overtook it in January 2022, and four years of slow erosion followed. A rebrand to EURA in March 2024 barely registered. By the time the team called it, EURA had fallen 98% to under $4M, ranking 10th among euro stablecoins. AIP-112 passed with just four voters (one casting 98% of the votes), a final measure of how few were still watching. The Angle team has moved on to Merkl, their DeFi incentive platform. Holders can redeem 1:1 for EURC until March 1, 2027.
Launched in April 2024 as Angle's yield-bearing USD stablecoin, backed by T-bills and DeFi lending, USDA arrived just as yield-bearing stablecoins became a commodity. It never found meaningful adoption. Wound down alongside EURA via AIP-112 in March 2026; holders can redeem 1:1 for USDC through the Angle Transmuter until March 1, 2027.
Overcollateralized on paper, unredeemable in practice
Autopsy · dForce USD (USX)
Overcollateralized on paper, unredeemable in practice
USX was dForce's CDP-style stablecoin, launched in 2021 and peaking above $205M in mid-2022. It never fully recovered the peg after the 2022 bear market: legacy debt positions on the dForce/Unitus lending protocol left circulating USX without a credible par-redemption route, and the discount widened from cents to a chasm. The terminal slide came in early 2026: below $0.75 in February, below $0.45 by April, and roughly $0.39 by July, with on-chain liquidity measured in tens of thousands of dollars and daily volume under $10. With the DF governance token near zero, no published remediation plan, and no redemption at par, Pharos froze USX as a collapsed archive asset.
GYD Gyroscope GYD · Counterparty Failure · Feb 2026
Gyro stopped spinning
Gyroscope's cross-chain GYD contract was exploited at the end of January 2026. The team told users not to interact with GYD, paused Gyro liquidity pools containing the token, and later offered a settlement to recover enough ETH to make users whole. The recovery never materialized publicly: Ethereum transfers stopped on January 31, Etherscan now reports 0 circulating GYD and $0 circulating market cap, CoinGecko no longer resolves the token through its API, and the remaining Balancer pools are one-sided GYD with zero volume.
Tether's offshore yuan experiment never became more than a footnote beside USD₮. On February 20, 2026, Tether stopped all new CNH₮ issuance and put the token on a one-year redemption clock, citing low interest, limited sustained demand, and the operational burden of keeping it alive. CNH₮ was supposed to give crypto a yuan rail; instead it spent most of its life as a thinly used side product. Holders can still redeem until February 20, 2027, but the decision itself was the death certificate.
MUST Mustang Finance · Counterparty Failure · Jan 2026
Wiped out alongside Saga Dollar in the IBC exploit
Mustang Finance was a stablecoin on the SagaEVM chainlet. On January 21, 2026, an attacker abused IBC mechanics to drain Saga Dollar (sister stablecoin); the same exploit wiped out Mustang and Colt. Mustang's residual ~$77M reported circulating is legacy debt with no live redemption path.
Palm USD promised to bridge Islamic finance and crypto with a Shariah-compliant, un-freezeable stablecoin backed by Gulf currencies. Despite announcing a $2.8B purchase agreement, actual circulating supply briefly touched $26M before collapsing to $81K. The billions sit perpetually 'unreleased.'
Saga Dollar was the native stablecoin of the SagaEVM chainlet, a Cosmos SDK-based L1 app-chain. On January 21, 2026, an attacker deployed malicious contracts that abused IBC mechanisms to mint tokens out of thin air, draining nearly $7M in USDC, yUSD, ETH, and tBTC to Ethereum mainnet. The chain was halted at block 6,593,800. Saga Dollar depegged to $0.75, and the sibling stablecoins Mustang and Colt were also wiped out. The exploit did not involve a consensus breach, just a creative reading of IBC semantics.
syUSD Synnax Stablecoin · Counterparty Failure · Jan 2026
Flash-loaned into irrelevance
Synnax launched as the first SEI-backed stablecoin on the Sei network, allowing users to mint syUSD against SEI collateral. On January 9, 2026, an attacker exploited the protocol via a flash loan, borrowing approximately 1.96 million WSEI (~$240K) from the Synnax contract without repayment. The protocol migrated to a new contract but never recovered; trading volume dropped to $15/day and supply continued bleeding from its $3.9M peak. DefiLlama marks it dead from January 15, 2026.
Paxos International launched Lift Dollar in 2024 as a yield-bearing stablecoin regulated by ADGM (Abu Dhabi), distributing daily yield via rebase from T-bill reserves. It peaked at $128.7M on Ethereum before Paxos decided to consolidate around USDG and the Global Dollar Network. Minting ceased October 8, 2025; rebasing stopped December 8, 2025; remaining balances were auto-converted to USDG. A clean, regulated wind-down: no depeg, no drama, just strategic prioritization.
USPD US Permissionless Dollar · Counterparty Failure · Dec 2025
Proxy had one too many hands
Over-collateralized stETH-backed CDP exploited via a CPIMP (Clandestine Proxy In the Middle of Proxy) attack on December 4, 2025, allowing the attacker to mint ~98M fraudulent USPD tokens and drain ~237 stETH. Legitimate supply never exceeded $500K; V2 was announced but never launched.
xUSD Stream Finance xUSD · Counterparty Failure · Nov 2025
$93M loss broke three coins
Yield-bearing stablecoin that collapsed when an external fund manager disclosed a $93M loss. xUSD plunged 77% from $1 to $0.26, freezing ~$160M in user deposits. The collapse exposed $285M in interconnected DeFi debt and caused cascading contagion that also toppled deUSD and Stables Labs USDX.
Yala's BTC-backed stablecoin survived its first death in September 2025 when a security breach minted 120 million unauthorized YU tokens, draining $7.6M. The team injected $5.5M and clawed back the peg. Two months later, runaway borrowing drained all liquidity pools and YU collapsed again to $0.44. This time, nobody came to resuscitate.
USDX Stables Labs USDX · Counterparty Failure · Nov 2025
Founder drained the pool
Synthetic USD backed by delta-neutral positions across exchanges. A Balancer V2 security flaw drained $1M, then contagion from Stream Finance's $93M loss accelerated a full collapse. USDX crashed 64% to $0.35 as founder Flex Yang's wallet was linked to addresses draining liquidity. Now trades at $0.01 to $0.04 with Discord shut down.
The Level team was acqui-hired by another DeFi protocol and announced the sunsetting of lvlUSD. Users could unstake and redeem with cooldown periods reduced to 2 seconds.
csUSDL Coinshift csUSDL · Counterparty Failure · Oct 2025
Coinshift shifted away
A yield-bearing wrapper around Paxos's USDL via Morpho Blue, killed by Paxos's decision to wind down USDL. Coinshift stopped accepting new deposits and raised borrow rates to close positions.
Beanstalk's first fork tried to give the credit-based stablecoin model a clean shot on Base, free from the governance exploit that killed the original. Without collateral backing and with insufficient demand for protocol debt, the price slid relentlessly from $1 to $0.11, proving the flaw was in the design, not the implementation.
USDM Mountain Protocol USDM · Abandoned · Aug 2025
Acquired into oblivion
Mountain Protocol USDM was a regulated, yield-bearing stablecoin backed by U.S. Treasuries, earning an S&P stability assessment and peaking at $154.6M across Ethereum and Optimism. Anchorage Digital acquired Mountain Protocol in May 2025 and immediately began a three-phase wind-down: minting ceased May 12, yield dropped to zero June 11, and primary market redemption closed August 22. Remaining USDM was deposited into a Uniswap USDC pool for secondary market exit. The team and tech were absorbed; the stablecoin was not.
TRON's legacy decentralized stablecoin was deliberately sunset as the ecosystem shifted to USDD. CDP functions ended August 2025, and 95% of supply was redeemed at a fixed rate of 1 USDJ = 1.5532 TRX.
Float Protocol launched in 2021 with an ambitious vision: a floating-rate stablecoin backed by ETH, with a target price initialized at $1.618 (the golden ratio). Supply expanded and contracted via Dutch auctions using ETH and the BANK governance token. The protocol peaked at ~$5M market cap but never gained meaningful adoption. By 2025, the anonymous team had vanished, the website went dark, GitHub fell dormant, and supply froze at ~$654K. The contracts remain on Ethereum, an immutable monument to the idea that stability doesn't require a peg.
In April 2025 a governance proposal to sunset Inter Protocol cited declining market appetite for decentralized stablecoins, rising operational complexity and costs, and a refocus on Agoric's Orchestration work and the BLD token. Agoric's on-chain signaling Proposal 93 passed on May 1, 2025, starting a planned 60-day wind-down: no new vault openings or IST minting, a liquidation fee set to zero so vault users could repay IST and reclaim collateral, and weekly liquidation-ratio increases to close any remaining vaults. The sunset process completed on June 26, 2025. Pharos keeps IST as a frozen archive because the former vault and PSM dollar stablecoin no longer operates as an active mintable protocol asset.
A prepaid yen rail yielded to Japan's new stablecoin regime
JPYC Inc. permanently ended new issuance and exchange support for its prepaid JPYC series on June 1, 2025 as Japan's revised Payment Services Act took effect. Existing balances remain usable for the issuer's prepaid-payment services and transfers, but cannot be redeemed for cash and no longer have a primary issuance path. The separately tracked funds-transfer-token JPYC is the successor product; Pharos freezes v1 as its historical legacy series.
USDL Liquid Loans USDL · Liquidity Drain · Jun 2025
PulseChain sank
A Liquity V1 fork on PulseChain where users minted USDL against PLS collateral at 110% minimum ratio. As PulseChain's ecosystem collapsed with PLS price down over 90%, USDL supply followed, from $33M peak to $3.4M. The immutable contracts still technically work, but with PulseChain abandoned by mainstream DeFi, nobody is using them.
DYAD's novel Kerosene-token model for overcollateralized CDPs failed to sustain interest. Supply declined from $20M peak to near zero despite continued GitHub activity.
Thala Labs' stablecoin on Aptos declined from $19M to near zero. Despite a $25.5M exploit in Nov 2024 (funds recovered), MOD simply failed to compete as USDT/USDC expanded on Aptos.
GMO Trust's NYDFS-regulated stablecoin failed to gain market share despite expansions to Solana and Bitstamp listing. Supply dwindled to essentially zero while GMO's JPY stablecoin (GYEN) continues.
Gravita Protocol's interest-free borrowing model against LSTs failed to maintain adoption as competitors offered better terms. Supply dropped from $15M to near zero.
An RToken built on Reserve Protocol for yield-bearing exposure on Base. The Reserve platform thrives, but this individual RToken lost adoption as users moved to higher-yield alternatives.
The first casualty of Liquity's own immutability principle. Legacy BOLD launched as the flagship of Liquity V2: overcollateralized, decentralized, and permanently unchangeable. A Stability Pool vulnerability surfaced weeks after launch. The contracts couldn't be patched, so the team redeployed everything from scratch.
The Verified USD Foundation ceased support on December 31, 2024 and began removing all tokens from circulation. Backed by tokenized T-bills via Matrixport's STBT, the omnichain stablecoin never gained sufficient traction.
Overnight Finance's yield-bearing DAI wrapper declined from $15M to near zero as the broader protocol wound down. CoinGecko reports circulating supply at zero.
After an $11.6M exploit in March 2024, Prisma Finance never recovered. Governance approved PIP-46 to decommission the protocol, reducing the debt ceiling to zero and introducing a PSM for orderly wind-down.
ULTRA PrismaLRT ULTRA · Liquidity Drain · Nov 2024
Fell with Prisma
Prisma's LRT-backed stablecoin was decommissioned alongside mkUSD after the March 2024 exploit. A stability pool bug discovered during wind-down allowed one user to drain ~14 ETH.
zkBob's privacy-focused stablecoin was deliberately sunset as the protocol switched to USDC, USDT, and ETH. The team noted BOB "did not receive enough consumer usage to justify ongoing support."
Born from the ashes of TenX, Mimo Protocol launched PAR as one of the first decentralized Euro stablecoins in early 2021, peaking at $25M during DeFi summer. But the euro-pegged CDP model never found product-market fit, and the team quietly pivoted to KUMA Protocol, tokenizing Swiss government bonds instead of minting synthetic euros. By late 2024, PAR's supply had cratered 90%, the governance forum went dark, and the MIMO token flatlined at $0.003 with zero daily volume.
UXD Protocol's delta-neutral stablecoin on Solana survived the Mango Markets exploit but couldn't attract users. The team admitted the model "isn't exciting enough for DeFi users" and the DAO voted to sunset.
Synthetix deprecated all non-USD spot synths on Ethereum mainnet as part of the V3 migration. sEUR holders who didn't swap in time could only redeem via the L1 redeemer contract at a discounted rate.
A solvent euro was retired by its license boundary
Anchored Coins stopped issuing new AEUR after MiCAR made an EU issuer license necessary for the Swiss company to continue a EUR stablecoin business. The issuer moved reserves to Swissquote and kept a high-minimum redemption route for existing holders, but announced an orderly withdrawal from stablecoins rather than a relaunch. With issuance permanently closed and the product in a regulatory wind-down, Pharos freezes AEUR as an archived legacy asset while preserving its historical market record.
Nostra's grand ambition to build Starknet's first native stablecoin launched with $472K of hope in April 2024. It proceeded to lose 96% of its supply in a slow, silent hemorrhage. Initially pitched as an overcollateralized crypto stablecoin, then quietly rebranded to fiat-backed (1:1 USDC), UNO found itself answering a question nobody on Starknet was asking: why hold a wrapper when you can just hold USDC? Nostra pivoted to its NSTR governance token and scrubbed UNO from the main site.
Lybra Finance V2's LST-backed yield-bearing stablecoin peaked at $128M. After mining incentives ended in June 2024 and development stalled, supply collapsed. The protocol is technically live but abandoned by its team.
Native overcollateralized stablecoin of the Canto L1 blockchain. Canto launched with 'free public infrastructure' DeFi primitives in late 2022, briefly attracting $300M TVL. As the chain's ecosystem collapsed, NOTE's supply fell 97% from $49M to $1.3M. The chain itself is essentially abandoned.
Preon Finance's stablecoin quietly died as the Sphere ecosystem faded. No shutdown announcement was made; supply simply declined from $32M to near zero.
JPEG'd's stablecoin minted against NFT collateral declined alongside the broader NFT market collapse. The protocol token stopped trading and the project was effectively abandoned.
Despite $3.7M in seed funding from Pantera and Solana Ventures, Hedge Protocol's stablecoin on Solana failed to gain traction in a competitive DeFi landscape. Supply dwindled from $11M to near zero.
Stabl.fi's CASH was the native stablecoin of an ambitious Polygon DeFi empire spanning Satin Exchange, Retro DEX, and more. When Satin's launch imploded on day one, the ecosystem was rebranded around Retro, a Thena fork that briefly hit $18.8M TVL before bleeding out. The team's last blog post in February 2024 acknowledged 'adoption has been fairly low and TVL has bled,' then went silent. The website, docs, and dreams of a cross-chain yield-bearing indexcoin now all return connection refused.
The original DYAD stablecoin launched in December 2023 as an overcollateralized CDP with an immutable contract design. When the team needed to add flash-loan protection and the XP/Kerosene incentive mechanism, v1's immutability made upgrades impossible. In June 2024 they deployed an entirely new v2 contract set, a bold-style migration that left v1 holders to manually migrate their collateral. The 600K DYAD minted on v1 was never formally redeemed; the contracts simply went silent.
The original Beanstalk was drained of $182M in a flash-loan governance attack in April 2022. Undeterred, the team relaunched on Arbitrum with a credit-based model where soil demand and adjustable interest rates (Temperature) were supposed to keep BEAN at $1. It didn't. BEAN slipped below peg in mid-2024 and never came back, proving the design flaw was in the algorithm, not the security. Even Pinto, the community fork that tried the same model on Base, met the same fate.
Fluid Finance SA was dissolved by bankruptcy and renamed in liquidation effective 2024-04-22. Its DUSD and issuer websites no longer resolve, so Pharos freezes DUSD as an archived legacy asset rather than an active issuance venue.
SiloDAO deprecated XAI after the stablecoin failed to scale DEX liquidity and repeatedly traded below its peg. Governance raised borrowing rates to force repayment, burned XAI credit lines across 17 silos, and retired the mint module so no new XAI could be created. The token nevertheless retains 5,358,653.23 XAI of residual supply, verified on Ethereum at block 25,627,256, representing legacy borrows and stranded wind-down liquidity rather than a live stablecoin system.
The idealist's stablecoin: RAI rejected the dollar peg, embraced 'ungovernance,' and floated freely on pure math and ETH collateral. The team achieved their goal of removing all human governance, then did what the protocol couldn't: moved on to build HAI on Optimism.
Yeti Finance's cross-margin lending protocol on Avalanche allowed borrowing YUSD at 0% interest against diverse collateral. After declining TVL and concentration among few users, the team announced dissolution in December 2023, raising interest rates to encourage position closure and returning 90% of treasury to YETI holders.
Allowed minting R stablecoin against wstETH and rETH collateral. A smart contract exploit drained $3.3M in ETH, though the hacker lost money due to a slippage error, netting only ~$1.6M. Raft halted minting, offered a 42% recovery plan, and was abandoned.
Tangible's grand experiment in real estate-backed stablecoins ended exactly how skeptics predicted: with a classic bank run on illiquid assets. USDR promised 16% yields backed by UK rental properties, but when the $6M DAI cushion was drained in a single week, holders discovered that tokenized houses cannot be liquidated at 3 AM on a Wednesday. The team candidly admitted there were 'too many attack vectors in the design.'
Vesta Finance dissolved after founders proposed to exit citing "cultural differences." Community voted 96% for dissolution; VSTA holders received pro-rata treasury redemptions in USDC and ARB.
Hector Network's stablecoin collapsed after the Multichain bridge exploit wiped out $8M of treasury assets. The DAO voted to liquidate the remaining $16M treasury, ending a project already plagued by allegations that the team squandered $100M+ in funds.
Bridge-wrapped USDC on Fantom became worthless overnight when the Multichain bridge collapsed after the CEO was arrested by Chinese authorities. ~$180M in 'USDC' was revealed to be backed by nothing as the team lost access to MPC servers. Circle froze $63M, and Fantom's entire DeFi ecosystem was devastated.
Parrot Protocol raised $85M during Solana's DeFi summer, backed by Alameda Research. Two years and one FTX collapse later, the team held a governance vote to liquidate the $74M treasury and go tokenless. The vote passed 99.8%, allegedly stuffed with insider wallets, and IDO investors got back a dime on the dollar.
Acala's aUSD lost its peg after a catastrophic exploit in August 2022 minted 1.28 billion unauthorized tokens. Converted to aSEED in July 2023, ceasing to function as a stablecoin.
FIAT DAO created a clever primitive: mint a dollar-pegged stablecoin against fixed-income assets like Notional fCash and Sense Principal Tokens. But the protocol barely reached $2M before its parent entity merged with BarnBridge, right before the SEC came knocking. When BarnBridge halted all operations in July 2023, FIAT was left orphaned with $7,600 of permanently stuck tokens and a domain that no longer resolves.
Unit Protocol's USDP accepted everything from Yearn vaults to SushiSwap LP tokens as collateral, a collateral buffet that attracted nearly $280M. But quantity of options couldn't substitute for quality of demand, and when Paxos swooped in with its own 'USDP' ticker, the identity crisis accelerated the decline. Supply withered to under $10K.
USDK did everything right on paper: regulated US trust company, ERC-20 standard, clean audits. But its custodian Prime Trust was secretly gambling client funds on algorithmic stablecoins. When Prime Trust collapsed into receivership (revealed to be $82M short on customer fiat), USDK's 'fully backed' promise became just another line in a bankruptcy filing.
GFX Labs' fractional reserve experiment promised to be Ethereum's answer to traditional banking: mint USDi by depositing USDC, earn rebasing yield from borrowers. Launched with Chainlink oracles and a $2.5M seed round, USDi briefly touched $8.4M before the bear market crushed demand for yet another USDC wrapper. The team quietly pivoted to DAO governance consulting and DEX development, leaving USDi to wither to under $20K.
mStable dreamed of unifying the fragmented stablecoin landscape: one meta-stablecoin to swap them all, with built-in yield. It touched $171M, but frictionless stablecoin swaps couldn't compete with DEX aggregators and simple USDC. When the co-founder walked away in early 2023, the DAO chose an honorable death: acquisition by dHedge.
Reserve's prototype stablecoin was deprecated when the protocol launched its RToken system. eUSD (Electronic Dollar) replaced RSV in the LATAM-focused Reserve app.
Once the third-largest stablecoin at $23.5B. The NYDFS ordered issuer Paxos to stop minting, while the SEC signaled intent to sue. Binance ceased support in Dec 2023, auto-converting remaining balances to FDUSD.
USP USP Stablecoin · Counterparty Failure · Feb 2023
Flash loan, lights out
Platypus Finance's stablecoin was backed by LP tokens from its stableswap pools. A flash loan exploit in February 2023 drained $8.5M, collapsing USP's peg. The protocol never recovered and ceased operations.
Sperax promised auto-yield just from holding its stablecoin on Arbitrum, no staking required. Then a rebasing bug let someone mint 9.7 billion tokens out of thin air. The funds came back, but user trust didn't. Backed by Alameda at peak hubris, USDs survived its exploit only to bleed out slowly.
A fiat-backed stablecoin tied to the Huobi ecosystem. When Justin Sun acquired Huobi and replaced it with USDD, HUSD was delisted with no redemption path. Crashed to $0.28.
NEAR Protocol's native algorithmic stablecoin launched in April 2022. After UST's collapse, it was hastily converted to USDT-backed, but a double-minting bug left it $40M undercollateralized. The NEAR Foundation formally shut it down in October 2022 with a $40M 'USN Protection Programme.' Remaining ~$38M sits unredeemed.
A partially algorithmic stablecoin on Avalanche from the Spice Trade DEX ecosystem. Launched in May 2022 into the post-UST wreckage, it never gained meaningful adoption. The peg broke almost immediately, sliding to $0.07, and the anonymous team abandoned the project. $18M in ghost supply sits unredeemable on Avalanche.
Raised $1.3B in ETH at launch, but Tribe DAO voted to wind down citing mounting technical, financial, and regulatory risks. An $80M hack of merged Rari/Fuse lending markets sealed its fate. Holders redeemed 1:1 for DAI.
VOLT VOLT Protocol · Counterparty Failure · Aug 2022
Inflation-proof wasn't
The first CPI-pegged stablecoin promised inflation protection via yields from Rari Fuse lending markets. When the $80M Fuse exploit wiped out those markets, VOLT lost its yield engine. The Tribe DAO dissolution four months later removed the last collateral backstop; the team pivoted to Ethereum Credit Guild.
CoinFLEX's interest-bearing stablecoin promised yields up to 114% APY from its futures lending market. When Roger Ver's account blew an $84M hole in the exchange, flexUSD holders discovered 'interest-bearing' also meant 'loss-bearing.' The exchange froze all redemptions, rebranded into OPNX with Three Arrows Capital's founders, and that venture also collapsed. 166M flexUSD tokens sit unredeemable at $0.08.
Kava's flagship stablecoin was the crown jewel of the first cross-chain DeFi hub: mint USDX by depositing BTC, XRP, BNB, or KAVA into CDPs on Cosmos. It peaked at $176M in April 2022, then UST's implosion exposed a fatal flaw: USDX had accepted UST as collateral. The peg cracked to $0.55 and never healed. Kava Labs quietly pivoted to AI infrastructure and 'DeCloud' ambitions, leaving USDX to bleed over three years to $11M, trading at $0.44.
Fantom's native overcollateralized stablecoin required 300-500% FTM collateral ratios, but collapsed alongside the broader DeFi crisis in 2022. Traders exploited the gap between its market price and collateral value, devastating protocols like Scream. The peg broke to $0.50-$0.70 and never recovered, sliding to under $0.10. With the Fantom network itself migrating to Sonic, fUSD became a zombie token, still in wallets but functionally dead.
The largest stablecoin collapse in history. UST's algorithmic peg relied on minting/burning LUNA, but a coordinated sell-off triggered a death spiral that vaporized $40B in combined value within days. Anchor's unsustainable 20% yield had concentrated 70% of supply in a single venue.
The most-used non-USD stablecoin in the Terra ecosystem, powering real Korean e-commerce payments via the CHAI app with millions of transactions. Destroyed alongside UST and LUNA in the May 2022 death spiral. All Terra stablecoins died simultaneously, but KRT was the only alternative peg with meaningful market cap.
An algorithmic stablecoin on Fantom hit by flash loan exploits totaling $16M, then finished off by contagion from UST's collapse. A further hack in May 2023 delivered the final blow.
The original Beanstalk deployment grew to $100M and $144M in liquidity before a flash-loan governance attack drained $182M. The attacker flash-loaned $1B+ to gain 79% governance power and passed a malicious proposal in a single transaction. BEAN crashed from $1 to $0.09. The protocol was 'Replanted' with new contracts four months later.
CASH Cashio Dollar · Algorithmic Failure · Mar 2022
Infinite mint, instant death
Missing validation code on Solana allowed an attacker to mint 2 billion CASH using a fake worthless token as collateral, draining $52M. CASH crashed from $1 to $0.00005 in minutes. The unaudited contract was a sitting duck. The hacker, Shakeeb Ahmed, was later caught and sentenced by the DOJ.
Venus Protocol's synthetic stablecoin on BSC was minted freely during DeFi summer 2021, too freely. A $77M bad-debt crisis cratered confidence, VAI shed its peg to $0.74, and minting was paused. Supply collapsed 99% from $300M+ to under $3M.
Dubbed crypto's "first large-scale bank run." IRON was partially collateralized (75% USDC, 25% TITAN token). When whales dumped TITAN at its peak, a flawed redemption mechanism sent TITAN from $65 to zero in hours, dragging IRON down with it.
ESD Empty Set Dollar · Algorithmic Failure · Jan 2021
The coupon experiment
Pioneered the "seigniorage shares" model in DeFi. When ESD traded below $1, users could buy coupons (burning ESD) in hopes of future redemption at a profit. The mechanism worked during expansion but collapsed when confidence evaporated: coupons expired worthless, and ESD fell to $0.01.
DSD Dynamic Set Dollar · Algorithmic Failure · Jan 2021
Forked ESD. Same fate
A fork of ESD with faster epoch cycles (2 hours vs 8 hours), designed to stabilize more quickly. Instead, the shorter cycles amplified volatility. DSD spiked to $3 during expansion then collapsed to $0.24 in the same month, proving that faster reflexivity cuts both ways.
An anonymous fork of the Basis design, BAC lost its peg within weeks of launch. Later revealed to be co-founded by Do Kwon under a pseudonym -- who learned nothing before building the even more catastrophic TerraUSD.
One of the first stablecoins ever created (2014), NuBits held its peg for two years before collapsing when holders dumped it to chase Bitcoin gains. A pioneering cautionary tale about algorithmic pegs backed by volatile assets. (Excluded from timeline view due to early date.)
Counts include every record. Peak market cap is each coin's approximate peak, summed only where recorded; it measures size at the top, not holder losses.
Timeline
Documented deaths per year, by cause
Algorithmic failures belong to the early record
Counterparty failures are a larger share of recent deaths
Bars count records by death year. The row under the axis is the median peak market cap of that year's deaths.
Median peak by year is in the data table on narrow screens. * 2026 runs through Oct 5, 2026. 21 of 31 records in 2026 come from Pharos's own tracked archive. No records exist for 2019 and 2020; that is a gap in the catalog, not evidence of no failures.
Data table
Documented deaths per year by cause, 2018 to 2026, with tracked-archive records and the median recorded peak market cap.
Year
Total
Abandoned
Counterparty Failure
Liquidity Drain
Algorithmic Failure
Regulatory
Tracked archive
Median peak
2018
1
0
0
0
1
0
0
not recorded
2019
0
0
0
0
0
0
0
no records
2020
0
0
0
0
0
0
0
no records
2021
5
0
0
1
4
0
0
$318.8M
2022
14
1
3
1
9
0
0
$118.3M
2023
18
7
5
4
1
1
0
$37.7M
2024
23
13
1
6
1
2
2
$32.1M
2025
22
8
6
6
1
1
2
$34.6M
2026 (through Oct 5, 2026)
31
15
11
3
1
1
21
$26.0M
Size
Peak market cap by cause
The largest coins did not all collapse
Each dot is one death with a recorded peak, on a log scale. The vertical tick marks each cause's median. Select a dot to open its row in the register.
One death
Cause median
Top five, labelled
Abandoned
Counterparty Failure
Liquidity Drain
Algorithmic Failure
Regulatory
med $30.5M · n=40 (+4 n/r)
med $142.0M · n=21 (+5 n/r)
med $26.0M · n=21
med $91.1M · n=17 (+1 n/r)
med $81.0M · n=5
10 records have no recorded peak and are not plotted. Size buckets: Under $10M 16 · $10M to $100M 48 · $100M to $1B 35 · $1B or more 5.
Data table
Recorded peak market cap by cause: records with and without a recorded peak, the median and the largest.
Cause
Recorded peaks
Not recorded
Median peak
Largest
Abandoned
40
4
$30.5M
FEI $2.3B
Counterparty Failure
21
5
$142.0M
USR $586.6M
Liquidity Drain
21
0
$26.0M
MIM $3.6B
Algorithmic Failure
17
1
$91.1M
UST $18.8B
Regulatory
5
0
$81.0M
BUSD $23.5B
Methodology
What counts as dead
Two routes into the cemetery
Tracked, then frozen (25): coins Pharos monitored live and froze once they had effectively ended or failed. Their detail pages stay online with archived data.
Curated (89): failed or discontinued stablecoins documented from public sources.
Active depegs, which stay on the depeg tracker until they recover or are frozen.
Quarantined records, withheld after a reviewed lack of supply or market-cap coverage.
Delisted records, which fell outside listing scope. They keep a read-only profile with a dated reason.
Inclusion and primary cause
Inclusion
A stablecoin is included when it had a public market and public sources show it failed or was discontinued: an announcement, filing, governance record or press report, or, for a coin that faded without one, market data showing its collapse. There is no size floor; peak market cap is recorded when known.
Primary cause
Each record carries one primary cause: the root reason the coin failed or its issuance and operation ended, whether or not it lost its peg. Triggers such as runs, exploits or orders are described in the obituary.
Fields and limits
One primary cause per record.
Death dates are precise to the day for 13 records and to the month for 101.
Peak market cap is approximate, optional and not a loss figure: 10 of 114 not recorded, never counted as zero. BUSD holders were converted, and FEI and EURT had redemption routes.
Each record links one public source. Most document the failure or discontinuation directly; a coin that faded without an announcement links the best public context Pharos found.
The catalog is not exhaustive: nothing is recorded for 2019–2020. 21 of the 31 records in 2026 so far are coins Pharos tracked live.
New records are appended to the daily Telegram digest. The RSS feed carries the 50 most recent deaths.
How to cite
Pharos, "Stablecoin Cemetery Dataset," schema 1.1, 114 records, latest record added Oct 19, 2026. https://pharos.watch/datasets/stablecoin-cemetery.json (MIT).
Frequently asked questions
How many stablecoins have failed or been discontinued?
Pharos documents 114 stablecoins that failed or were discontinued between 2018 and 2026: 89 curated records and 25 coins Pharos tracked live before they ended. The latest recorded death is Oct 5, 2026. The catalog is not exhaustive.
What was the largest collapse?
TerraUSD (UST) is the largest recorded collapse by peak market cap: it peaked at $18.8B and failed in May 2022. Its recorded primary cause is algorithmic failure. Regulatory and abandoned exits count as discontinued, not collapsed.
What was the largest discontinued stablecoin?
Binance USD (BUSD) is the largest discontinued stablecoin by peak market cap: it peaked at $23.5B and was ended by a regulator or licensing regime in Feb 2023. Peak market cap measures its size, not what holders lost.
Is peak market cap what holders lost?
No. Peak market cap is each coin's approximate size at its peak. Some records ended with redemption or conversion routes for holders; others ended far below peg. Pharos does not publish a loss figure, and records peak market cap for 104 of 114 records.
Are stablecoin deaths becoming more common?
Not demonstrably. Pharos recorded 37 deaths in the 12 months to Oct 2026, against 21 in the 12 months before. Curated records were flat at 16 against 19; the rest of the change comes from Pharos's own tracked archive (21 against 2).
Are algorithmic stablecoins more likely to fail?
The cemetery records deaths, not launches, so it cannot measure a failure rate. Among recorded deaths, algorithmic failures were 14 of 20 records through 2022 and 4 of 94 since.
How does a stablecoin enter the cemetery?
A stablecoin is included when it had a public market and public sources show it failed or was discontinued: an announcement, filing, governance record or press report, or, for a coin that faded without one, market data showing its collapse. There is no size floor; peak market cap is recorded when known. Records arrive two ways: 25 coins Pharos tracked live and froze after they ended, and 89 curated records documented from public sources.
Can I cite this data?
Yes. The JSON and CSV exports of all 114 records are published under the MIT license. Cite the dataset URL with the record count and the date you used; each record also has a permanent link on this page.
