> For the complete documentation index, see [llms.txt](https://docs.tapir.money/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.tapir.money/resources/defi-investment-loss-analysis.md).

# DeFi Investment Loss Analysis 2020–2026

Most people think of DeFi risk as a stablecoin losing its peg. That happens, but it is only one way to lose money. A vault can bleed assets on a bad trade while its deposit token holds at $1. A lending pool can sit on a defaulted loan for months before anyone marks it down. An exploit can drain the reserves behind a share token in a single block.

The result is the same in every case: **the investment is worth less than it was.**

This report catalogues 29 observations between October 2020 and October 20, 2026, grouped into four categories, and counts a dollar loss for 25 of them. Each category ends with a subtotal, and the summary adds them up. This is a selected catalogue, not a complete estimate of losses across decentralized finance (DeFi).

## Summary

| Category                                  | Observations listed | Events counted | Gross losses  | Share of total |
| ----------------------------------------- | ------------------- | -------------- | ------------- | -------------- |
| Exploits that drain token or vault value  | 12                  | 12             | **$830.3M**   | 48.8%          |
| Depegs and redemption shortfalls          | 8                   | 4              | **$722.9M**   | 42.4%          |
| Trading, liquidation and execution losses | 4                   | 4              | **$97.0M**    | 5.7%           |
| Credit losses in lending pools            | 5                   | 5              | **$52.8M**    | 3.1%           |
| **Total**                                 | **29**              | **25**         | **$1,703.0M** | **100%**       |

**$1.7 billion in documented gross losses across 25 events in six years.** Every year in the period has losses, and every category contributes.

### Losses by year

| Year                      | Gross losses  | Largest events                                |
| ------------------------- | ------------- | --------------------------------------------- |
| 2020 (from October)       | $33.8M        | Harvest                                       |
| 2021                      | $11.0M        | Yearn yDAI                                    |
| 2022                      | $168.0M       | Beanstalk, Cashio, Maple                      |
| 2023                      | $245.8M       | Euler, USDC depeg liquidations, Tangible USDR |
| 2024                      | $5.9M         | Goldfinch Lend East                           |
| 2025                      | $857.2M       | Stables Labs USDX, Balancer, Stream Finance   |
| 2026 (through October 20) | $381.4M       | Kelp rsETH bridge, Neutrl NUSD, Resolv        |
| **Total**                 | **$1,703.0M** |                                               |

There are losses in every year. None of these was a one-off crisis that the industry then moved past.

### How we counted

* **Funds lost count as losses, even when later recovered.** A theft, a strategy loss or a loan default is counted at its gross amount when it happened. That holds when an attacker later returned the funds (Euler) or when users were made whole by a recovery fund (Kelp rsETH). The loss happened, and depositors carried it until someone else stepped in.
* **A depeg that recovered is not counted as a loss.** If a token's price returned to its peg, the temporary discount is left out. Only documented liquidations that the depeg caused are counted. The USDC, DAI and stETH depegs fall here. apxUSD is also uncounted because the cited evidence establishes neither a full peg recovery nor a quantified lasting shortfall or liquidation total.
* **A depeg that did not recover is counted.** For a permanent depeg, the loss is the supply before the depeg multiplied by the lasting discount or redemption shortfall.
* **Algorithmic stablecoins are excluded.** Collapses of stablecoins held up by an algorithm rather than by reserves, such as TerraUSD (UST) in May 2022 and NEAR's original USN design, are not included in any total.
* **Dollar stablecoins are counted at $1.** USDC, DAI, BUSD and USND amounts are treated as dollars.
* **ETH is converted at a historical ETH/USD quote** from DefiLlama's price feed: $2,990.93 at 23:59:54 UTC on November 30, 2025, for yETH.[\[P2\]](#supply-and-price-data)
* **Each event is counted once.** If one loss spread to several connected tokens, vaults or lenders, it appears once.
* **Totals use unrounded inputs.** Displayed amounts and percentage shares are rounded, so they may not add exactly. Years refer to the originating incident or disclosure.

## 1. Exploits that drain token or vault value

An attacker manipulates prices, mints unbacked tokens, compromises a bridge or takes over governance, and the assets behind a token, vault share or pool position disappear. This is the largest category.

| Event                    | Date     | What happened                                                                                                                                                                                                                                 | Gross loss  |
| ------------------------ | -------- | --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | ----------- |
| Kelp rsETH bridge        | Apr 2026 | An attacker drained 116,500 rsETH from Kelp's cross-chain bridge, breaking the token's backing. rsETH backing was restored about five weeks later with funds from Aave and other protocols.[\[S80\]](#event-sources)[\[S81\]](#event-sources) | $292.0M     |
| Euler                    | Mar 2023 | Lending-protocol exploit removed the assets behind deposits and connected vault strategies. All stolen assets were later returned.[\[S33\]](#event-sources)                                                                                   | $197.0M     |
| Balancer V2              | Nov 2025 | A flaw in legacy pools drained the reserves behind liquidity-provider shares.[\[S19\]](#event-sources)                                                                                                                                        | $94.8M      |
| Beanstalk                | Apr 2022 | A governance attack removed the external liquidity backing BEAN.[\[S63\]](#event-sources)                                                                                                                                                     | $77.0M      |
| Cashio CASH              | Mar 2022 | Fake collateral let the attacker mint unbacked CASH, which fell from $1 to $0.[\[S65\]](#event-sources)                                                                                                                                       | $52.0M      |
| Harvest                  | Oct 2020 | Manipulated Curve prices let the attacker trade against USDC and USDT vaults on bad terms.[\[S16\]](#event-sources)                                                                                                                           | $33.8M      |
| Resolv USR               | Mar 2026 | 80M USR minted without authorization and about $25M extracted in ETH. This is extracted value, not an audited total of holder losses.[\[S30\]](#event-sources)                                                                                | $25.0M      |
| Ostium Liquidity Pool    | Jul 2026 | A compromised off-chain price feed was used to trade against the USDC liquidity vault.[\[S26\]](#event-sources) The reported loss was 23,752,746 USDC.[\[S27\]](#event-sources)                                                               | $23.8M      |
| Yearn yDAI v1            | Feb 2021 | Forced deposits and withdrawals through manipulated Curve prices cut vault deposits from 35M to 24M DAI.[\[S17\]](#event-sources)                                                                                                             | $11.0M      |
| Yearn yETH               | Nov 2025 | An exploit hit the ETH basket backing yETH, a loss of 3,157.4 ETH.[\[S21\]](#event-sources) Valued at $2,990.93 per ETH.[\[P2\]](#supply-and-price-data)                                                                                      | $9.4M       |
| Term ETH and USDC vaults | Aug 2026 | A governance compromise removed the assets backing vault shares.[\[S29\]](#event-sources)                                                                                                                                                     | $8.5M       |
| Lazy Summer USDC vaults  | Jul 2026 | Stale-priced receipt tokens distorted share pricing, and the vaults lost their USDC backing.[\[S23\]](#event-sources)                                                                                                                         | $6.0M       |
| **Subtotal (12 events)** |          |                                                                                                                                                                                                                                               | **$830.3M** |

These attacks went in through very different places: a bridge at Kelp, contract logic at Balancer, share accounting at Lazy Summer, an off-chain price feed at Ostium, governance at Beanstalk and Term. Checking the deposit token's price alone would have missed losses in several of these investments. The two largest, Kelp and Euler, were later recovered in full, and they still count: depositors had their funds taken and waited weeks or months to get them back.

## 2. Depegs and redemption shortfalls

A token that is supposed to hold a fixed value, such as $1 or 1 ETH, trades below it, or can only be redeemed for less. When the price comes back, the discount is not counted, only the liquidations it caused. When it does not come back, the lasting shortfall is counted in full.

| Event                   | Date         | What happened                                                                                                                                                                                                                                           | Status              | Counted loss                                                                                              |
| ----------------------- | ------------ | ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | ------------------- | --------------------------------------------------------------------------------------------------------- |
| Stables Labs USDX       | Nov 2025     | The synthetic dollar fell below $0.60 within a day and never recovered; it traded at about $0.04 a month later.[\[S82\]](#event-sources) Includes a Lazy Summer USDC vault that held about 16% in sUSDx and could not exit.[\[S47\]](#event-sources)    | Permanent           | 683.46M USDX × (1 − 0.0403)[\[P7\]](#supply-and-price-data)[\[P8\]](#supply-and-price-data) = **$655.9M** |
| Neutrl NUSD             | Aug–Sep 2026 | After withdrawals were suspended, a September report described a redemption rate of 0.51 USDC per NUSD; this is not a verified final whole-book recovery rate.[\[S45\]](#event-sources)                                                                 | Permanent           | 53.39M NUSD × 49%[\[P6\]](#supply-and-price-data) = **$26.2M**                                            |
| Circle USDC             | Mar 2023     | Access to $3.3B of reserves at Silicon Valley Bank was frozen, and USDC traded as low as $0.86 before recovering on March 13.[\[S68\]](#event-sources)[\[S36\]](#event-sources) About 3,400 positions on Aave were liquidated.[\[S83\]](#event-sources) | Recovered           | Liquidations: **$24.0M**                                                                                  |
| Tangible USDR           | Oct 2023     | Liquid reserves ran out while backing stayed illiquid. The later redemption design targets 0.532 USDC per USDR on 35,909,552 accessible tokens, conditional on funding; the target can rise.[\[S41\]](#event-sources)                                   | Permanent           | 35.91M USDR × 46.8% = **$16.8M**                                                                          |
| Maker DAI               | Mar 2023     | DAI fell below $0.90 through its USDC collateral and recovered with USDC.[\[S69\]](#event-sources)                                                                                                                                                      | Recovered           | Within USDC                                                                                               |
| Lido stETH              | Jun 2022     | Forced selling by large holders pushed stETH below ETH; Coinbase reported a 2.2% discount on June 8.[\[S70\]](#event-sources) The gap later closed.                                                                                                     | Recovered           | No documented liquidation total                                                                           |
| Apyx apxUSD             | Jun 2026     | Preferred-equity collateral fell below par, and apxUSD traded around $0.90. The issuer's post-mortem reported stabilization at $0.95, not a full peg recovery.[\[S44\]](#event-sources)                                                                 | Recovery unverified | No quantified lasting shortfall or liquidation total                                                      |
| NEAR USN                | Oct 2022     | The original algorithmic design and double-minting left a $40M collateral gap, which the NEAR Foundation covered.[\[S39\]](#event-sources)                                                                                                              | Algorithmic         | Excluded                                                                                                  |
| **Subtotal (4 events)** |              |                                                                                                                                                                                                                                                         |                     | **$722.9M**                                                                                               |

The USDX collapse is the largest single loss in this report. A dollar token with about $683M in circulation lost more than 95% of its value within a month, and holders had no route back to par. The USDC depeg was larger in reach, pulling the second-largest stablecoin 14% below par over a weekend, but its price recovered, so only the forced liquidations it triggered are counted here.

## 3. Trading, liquidation and execution losses

Here the reported loss starts in the strategy: a liquidation goes wrong, an exit costs too much in slippage, or an outside manager reports a loss. The deposited coin can stay at $1 while the vault share is worth less; a strategy loss can also cause its own token to depeg, as with xUSD.

| Event                   | Date     | What happened                                                                                                                                     | Gross loss |
| ----------------------- | -------- | ------------------------------------------------------------------------------------------------------------------------------------------------- | ---------- |
| Stream Finance xUSD     | Nov 2025 | An external fund manager running the strategy behind xUSD disclosed a loss, and deposits and withdrawals were suspended.[\[S14\]](#event-sources) | $93.0M     |
| Hyperliquid HLP vault   | Mar 2025 | The USDC vault took a loss closing a large ETH position it inherited from a liquidated trader.[\[S03\]](#event-sources)                           | $4.0M      |
| Yearn yUSND             | Sep 2025 | Not enough USND liquidity for a rewards auction; selling collateral cost the vault 5.8% of its assets.[\[S02\]](#event-sources)                   | $25.5K     |
| Yearn Compound WMATIC   | Jan 2024 | A leveraged strategy was liquidated before it could rebalance.[\[S01\]](#event-sources)                                                           | $18K       |
| **Subtotal (4 events)** |          |                                                                                                                                                   | **$97.0M** |

Size varies a lot within this category. Stream's $93M is a headline event, while the two Yearn cases are small. The small ones still matter, because they show a vault can lose principal through liquidation or poor execution without requiring a depeg of the deposited coin. Strategists absorbed the WMATIC loss, and Yearn covered yUSND principal; the yUSND disclosure says realized yield potential was still reduced.

## 4. Credit losses in lending pools

The pool lends to a borrower, and the borrower does not pay back in full. This kind of loss tends to show up slowly, through missed payments, restructurings and write-downs, and the final number can take years to settle.

| Event                      | Date     | What happened                                                                                                                                                                                                                                                               | Loss       |
| -------------------------- | -------- | --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | ---------- |
| Maple / Orthogonal Trading | Dec 2022 | The borrower defaulted after misrepresenting its finances, leaving USDC and wETH pool lenders with unpaid loans.[\[S05\]](#event-sources)[\[S06\]](#event-sources)                                                                                                          | $36.0M     |
| Goldfinch / Lend East      | Apr 2024 | At maturity the borrower was expected to repay $4.25M of a $10.15M facility, a 58% principal shortfall.[\[S12\]](#event-sources)                                                                                                                                            | $5.9M      |
| Goldfinch / Tugende        | 2023     | A $5M loan defaulted. The original schedule estimated a 6.31% credit write-down, offset by 2.36% other interest to a 3.95% net pool-value decline. The later exit included $250K loan recovery, $210K closing fees and separate community support.[\[S09\]](#event-sources) | $5.0M      |
| TrueFi / Blockwater        | Oct 2022 | The borrower defaulted on a BUSD loan from the tfBUSD pool. Contemporaneous reporting quoting TrueFi put the amount still due at $2,967,458, rather than the original $3,419,190 facility.[\[S07\]](#event-sources)[\[S07a\]](#event-sources)                               | $3.0M      |
| Goldfinch / Stratos        | Oct 2023 | The December plan implied a $2.96M Senior Pool gap: $16M principal less $13.04M expected borrower recovery, before sponsor support. Junior claims are excluded.[\[S10\]](#event-sources)[\[S11\]](#event-sources)                                                           | $3.0M      |
| **Subtotal (5 events)**    |          |                                                                                                                                                                                                                                                                             | **$52.8M** |

Credit losses are easy to miss because a pool's quoted value can stay flat while the loan behind it is already in trouble. In the Goldfinch cases, recovery depended on years of restructuring and on sponsors deciding to step in.

## Conclusion

**Investment losses in DeFi are serious, they recur, and they come from many directions.**

Over six years, the events in this report add up to **$1.70 billion in gross losses across 25 events**. Temporary depegs that recovered are left out, and so are algorithmic stablecoins. What remains is money that was actually stolen, traded away, defaulted on, forcibly liquidated or permanently lost to a broken peg:

| Category                           | Gross losses  |
| ---------------------------------- | ------------- |
| Exploits                           | $830.3M       |
| Depegs and redemption shortfalls   | $722.9M       |
| Trading, liquidation and execution | $97.0M        |
| Credit losses                      | $52.8M        |
| **Total**                          | **$1,703.0M** |

Three conclusions follow from these numbers.

1. **This is a recurring problem, not an occasional crisis.** The sources of loss change, from flash-loan attacks in 2020 to bridge, price-feed and governance exploits in 2026, but some form of loss shows up every year. Fixing one attack vector has not made the next one go away, and the two most recent years, 2025 and 2026, are the largest in the report.
2. **The peg is only part of the picture.** Exploits, strategy losses and bad loans add up to $980.2M, more than half the total, and many of them hit while the deposit token held at exactly $1. Watching only a deposit token's price would miss losses in the investments behind it.
3. **Depositors cannot count on being made whole.** Some depositors got their money back, through Euler's recovery, the Kelp rsETH recovery fund, Yearn's support or Goldfinch's sponsors. Each of those recoveries depended on someone else acting: an attacker returning funds, other protocols pooling capital or a sponsor choosing to pay. Outcomes took weeks in some cases and years in others, and the cited sources do not establish full recovery for many of the events in this report.

The common thread is simple: **DeFi investments lose value from many causes, repeatedly, and depositors cannot rely on getting that value back.** Any serious approach to DeFi risk has to deal with falls in investment value from any cause, not just the peg.

## Sources

### Supply and price data

* **P2** — [DefiLlama: ETH price near midnight UTC, November 30–December 1, 2025](https://coins.llama.fi/prices/historical/1764547200/coingecko:ethereum). The request timestamp is December 1 at 00:00 UTC; the returned quote is November 30 at 23:59:54 UTC. ETH $2,990.93.
* **P6** — [DefiLlama: NUSD circulating supply](https://stablecoins.llama.fi/stablecoin/346). 53,390,062 NUSD from August 27 to September 17, 2026.
* **P7** — [DefiLlama: USDX circulating supply](https://stablecoins.llama.fi/stablecoin/214). 683,455,832 USDX on November 5, 2025, before the depeg.
* **P8** — [DefiLlama: USDX price, December 5, 2025](https://coins.llama.fi/prices/historical/1764892800/coingecko:usdx-money-usdx). $0.0403. DefiLlama's price series stays near or below that level through 2026.

### Event sources

* **S01** — [Yearn: WMATIC strategy liquidation](https://raw.githubusercontent.com/yearn/yearn-security/master/disclosures/2024-01-03.md), 2024-01-03.
* **S02** — [Yearn: yUSND liquidation-reward auction](https://raw.githubusercontent.com/yearn/yearn-security/master/disclosures/2025-11-26.md), 2025-11-26.
* **S03** — [Arkham: Hyperliquid whale transfers loss to HLP](https://info.arkm.com/research/hyperliquid-whale-passes-4m-loss-to-hlp-vault), 2025-03-12.
* **S05** — [M11 Credit: Orthogonal Trading default enforcement](https://medium.com/@M11credit/m11-credit-update-on-orthogonal-trading-9f896c264f8f), 2022-12-14.
* **S06** — [The Block: Orthogonal Trading $36 million default](https://www.theblock.co/news/markets/2022-12-05-maple-token-down-more-than-24-cuts-ties-with-orthogonal-trading-following-36-million-default-192166), 2022-12-05.
* **S07** — [Archblock: TrueFi special-servicing update](https://forum.truefi.io/t/archblock-special-servicing-update/1481), 2023-08-16.
* **S07a** — [Contemporaneous reporting quoting TrueFi's default notice](https://www.tradingview.com/news/cointelegraph%3A88253d01f%3A0-truefi-issues-first-default-notice-on-3-4m-busd-loan/), October 2022: $2,967,458 still due at default. The original issuer blog is unavailable; this is the reported default-date claim, not a final lender-loss ledger.
* **S09** — [Goldfinch: Tugende restructuring and exit update](https://gov.goldfinch.finance/t/update-2-warbler-labs-and-tugende-have-agreed-in-principle-to-a-restructuring-plan-that-may-result-in-a-material-recovery-for-the-goldfinch-senior-pool/1835), 2023-09-05 and 2024-12-16.
* **S10** — [Goldfinch: Stratos pool impairment](https://gov.goldfinch.finance/t/update-on-stratos-pool/1865), 2023-10-07.
* **S11** — [Goldfinch: Stratos recovery and backstop plan](https://gov.goldfinch.finance/t/stratos-technical-changes-to-support-re-finance-and-warbler-backstop/1928), 2023-12-20.
* **S12** — [Goldfinch: Lend East pool shortfall](https://gov.goldfinch.finance/t/update-on-lend-east-pool/1957), 2024-04-02.
* **S14** — [The Block: Stream Finance external-manager loss](https://www.theblock.co/newsletters/the-daily/2025-11-04-the-daily-stream-finance-external-fund-manager-discloses-93-million-loss-sequans-dumps-nearly-third-bitcoin-holdings-more-377629), 2025-11-04.
* **S16** — [Harvest: flash-loan economic attack post-mortem](https://medium.com/harvest-finance/harvest-flashloan-economic-attack-post-mortem-3cf900d65217), 2020-10-26.
* **S17** — [Yearn: yDAI vault exploit](https://raw.githubusercontent.com/yearn/yearn-security/master/disclosures/2021-02-04.md), 2021-02-04.
* **S19** — [Balancer: November 3 exploit post-mortem](https://medium.com/balancer-protocol/nov-3-exploit-post-mortem-51dcbeb6b020), 2025-11-18.
* **S21** — [Yearn: yETH exploit disclosure](https://raw.githubusercontent.com/yearn/yearn-security/master/disclosures/2025-12-01.md), 2025-12-01; 3,157.401 ETH gross loss per the [yETH recovery plan](https://gov.yearn.fi/t/yip-90-yeth-optimistic-recovery-plan/14573). Press reports put the loss at about $9 million ([ForkLog](https://forklog.com/en/yearn-finance-defi-project-hacked-for-9-million/)).
* **S23** — [BA Labs: Lazy Summer July 6 exploit retrospective](https://forum.summer.fi/t/lazy-summer-protocol-exploit-july-6-2026-ba-labs-risk-curator-retrospective/856), 2026-07-08.
* **S26** — [The Block: Ostium post-mortem](https://www.theblock.co/news/ecosystems/2026-07-30-ostium-post-mortem-exploit-410122), 2026-07-30.
* **S27** — [The Crypto Times: Ostium recovery launch](https://www.cryptotimes.io/2026/10/01/ostium-23-75m-hack-recovery-3321-wallets-repaid-345-lps-face-1000-choice/), 2026-10-01. Reported gross loss: 23,752,746 USDC, used unrounded in totals.
* **S29** — [The Block: Term Finance governance exploit](https://www.theblock.co/news/defi/2026-08-23-defi-lending-protocol-term-finance-loses-an-estimated-8-5-million-to-governance-exploit-412543), 2026-08-23.
* **S30** — [Resolv: March 22 incident post-mortem](https://resolv.xyz/blog/resolv-postmortem-march-22-2026-incident), 2026-04-04.
* **S33** — [Euler: exploit and full recovery](https://www.euler.finance/blog/war-peace-behind-the-scenes-of-eulers-240m-exploit-recovery), 2024-01-10.
* **S36** — [Circle: USDC reserve risk resolved](https://www.circle.com/pressroom/3-3-billion-of-usdc-reserve-risk-removed-dollar-de-peg-closes), 2023-03-13.
* **S39** — [NEAR Foundation: USN collateral gap and backstop](https://near.org/blog/statement-in-full-near-foundation-to-fund-usn-protection-programme), 2022-10-24. Describes v1 as "the algorithmic USN".
* **S41** — [Tangible: USDR redemption-v2 implementation](https://github.com/TangibleTNFT/usdr-redemption-v2).
* **S44** — [Apyx: reserve and redemption framework after June stress](https://blog.apyx.fi/apyx-2-0-redemption-value-total-collateralization-evolution-of-the-dividend-backed-dollar/), 2026-06-15; [issuer post-mortem](https://blog.apyx.fi/post-mortem/), 2026-06-08, reported apxUSD stabilizing at $0.95.
* **S45** — [The Crypto Times: Neutrl opens NUSD redemptions at 0.51](https://www.cryptotimes.io/2026/09/18/neutrl-opens-nusd-snusd-redemptions-as-on-chain-rate-reads-0-51/), 2026-09-18.
* **S47** — [BA Labs: Lazy Summer sUSDx exposure retrospective](https://forum.summer.fi/t/arbitrum-usdc-fleet-susdx-ark-retrospective-ba-labs/529), 2025-12-01.
* **S63** — [Beanstalk: assets stolen before Barn Raise](https://bean.money/blog/thoughts-before-the-barn-raise), 2022-06-05.
* **S65** — [CertiK: Cashio incident analysis](https://www.certik.com/zh-CN/blog/cashio-app-incident-analysis), 2022-12-06.
* **S68** — [Federal Reserve: SVB, USDC and stablecoin contagion](https://www.federalreserve.gov/econres/notes/feds-notes/in-the-shadow-of-bank-run-lessons-from-the-silicon-valley-bank-failure-and-its-impact-on-stablecoins-20251217.html), 2025-12-17.
* **S69** — [Federal Reserve: primary and secondary stablecoin markets](https://www.federalreserve.gov/econres/notes/feds-notes/primary-and-secondary-markets-for-stablecoins-20240223.html), 2024-02-23.
* **S70** — [Coinbase Institutional: stETH price gap, June 8 2022](https://www.coinbase.com/en-gb/institutional/research-insights/research/monthly-outlook/liquid-staking-june-2022), 2022-06-08.
* **S80** — [The Block: Kelp DAO's rsETH bridge exploited for roughly $292 million](https://www.theblock.co/post/397988/kelp-daos-rseth-bridge-apparently-exploited-for-roughly-292-million-in-layerzero-based-attack), 2026-04-18.
* **S81** — [Cointelegraph: Kelp DAO says rsETH fully restored five weeks after hack](https://cointelegraph.com/news/kelpdao-says-rseth-restored-5-weeks-after-protocol-hack), May 2026.
* **S82** — [The Block: synthetic stablecoin USDX depegs below $0.60](https://www.theblock.co/post/377917/synthetic-stablecoin-usdx-depegs-pancakeswap-lista), 2025-11-06.
* **S83** — [Kraken: stablecoin depegging](https://www.kraken.com/learn/stablecoin-depegging). About 3,400 liquidations on Aave v2 and v3 from about $24M of collateral during the March 2023 USDC depeg. Gauntlet separately reported about $300K of realized bad debt on Aave ([Gauntlet](https://www.gauntlet.xyz/resources/aave-resilient-through-usdc-volatility)).
