Abracadabra wind-down passes with two wallets under a quorum the team set: MIM to pay about 4 cents

Abracadabra Money, the DeFi lender behind the Magic Internet Money (MIM) dollar token, is being wound down on the strength of votes from two wallets. Its refined wind-down proposal on Snapshot closed on 15 October 2026 at 10:09 UTC with 396.16 million SPELL for and none against. It cleared a 30% quorum that the team wrote into the proposal itself, because the 3 billion threshold built into its Snapshot space is larger than the entire SPELL vote supply. Etherscan labels the bigger of the two voting wallets as the protocol’s own MIM treasury.

The plan pays MIM holders about $0.04 per token, in wrapped ether (WETH), once the remaining collateral has been recovered and swapped. SPELL, the governance token, gets nothing. The balance snapshot that fixes every claim will be taken no earlier than 20 October 2026.

MIM traded at $0.0460 at 08:22 UTC on 16 October and SPELL at $0.0000873, according to DefiLlama price data. In January 2022, the protocol held $6.42 billion in deposits.

Abracadabra Money wind-down vote passed with two wallets

The Snapshot proposal, titled “Refined proposal for an orderly wind down of MIM and Abracadabra”, opened on 13 October at 22:09 UTC and ran for 36 hours. Two addresses cast every vote.

The first, 0xE843…A59A58, wrote the proposal and voted with 100,047,420 SPELL on 14 October at 09:01 UTC. Two minutes later, 0xDF2C…126B added 296,115,861 SPELL. That second address is a Safe multisig that carries the public Etherscan name tag “Abracadabra.Money: MIM Treasury Wallet”. On those numbers, the treasury supplied about 75% of the votes that approved the shutdown.

That voting power was hours old. Etherscan shows the Safe moving its 296.1 million SPELL into Abracadabra’s mSpell staking contract on 13 October at 13:32 UTC. Snapshot counted votes at block 26,186,459, mined at 22:07 UTC the same day. The proposer wallet had staked its own SPELL in the same contract on 28 September, the day before the first wind-down vote.

The proposer wallet has no tag. Its Etherscan history shows it was first funded on 29 July 2026 with 0.1 ETH through RAILGUN, a privacy tool, that the treasury Safe itself sent it ETH on 8 August and 28 September, and that most of its recent transactions, between 1 and 13 October, are “Exec Transaction” calls on the treasury Safe. That pattern is typical of a multisig signer. Etherscan does not say who controls either wallet, and neither does the proposal.

The treasury Safe held about 1,279.64 WETH (around $3.14 million at Etherscan’s 16 October valuation), 1.19 million MIM and 673,690 SPELL when we viewed it on the morning of 16 October. Its history for 13 October is busy. That afternoon the Safe received Yearn vault, Convex and Curve tokens from contracts Etherscan does not label and redeemed them. Between 21:21 and 21:45 UTC it swapped stablecoins, WBTC, APE, CVX and stETH into WETH through CoW Protocol and wrapped about 112 ETH. All of that came before the refined vote opened at 22:09 UTC. The proposal does not say whether this WETH is the collateral to be paid out, and we could not confirm from Etherscan that the incoming tokens were cauldron collateral.

A 30% quorum set by the team, not the 3 billion in the rules

The result stands only because the team rewrote the threshold. The Snapshot space for Abracadabra has a quorum hardcoded at 3 billion votes, while the total vote supply is about 1.03 billion. The quorum is roughly three times every vote that exists.

According to the proposal, the admin keys for the Snapshot space sit with former team members who “are not reachable”, so nobody can change the setting. The team therefore declared its own quorum at 30% of vote supply: 309,046,858 votes. The 396.16 million cast clear that bar. Measured against the hardcoded figure, they are 13.2% of quorum. Snapshot’s own record of the proposal still carries the 3 billion setting; the pass rests on the number written into the proposal text.

This is the second attempt. The first wind-down proposal, which ran from 29 to 30 September, drew 100,047,420 SPELL for and 523,446 against, again from two wallets. That total was about 3.4% of the 3 billion quorum, and the first text set no other threshold. The refined version names the 3 billion setting as the problem and adds the 30% rule.

Thin participation is not new here. In June 2026, the proposal that handed Abracadabra to a new steward passed on 4 June with 5.51 billion SPELL for and none against. It also had two voters, and it was submitted from an address Etherscan labels “Abracadabra.money: Deployer 3”.

What MIM holders get: about $0.04 per token, in WETH

The payout logic follows from the numbers in the proposal. Collateral still backing MIM debt is worth about $1.2 million. Some $300,000 of it is stuck in an immutable WETH cauldron on Arbitrum that the team cannot touch, which leaves around $900,000 it can act on.

Against that, nearly $22 million of MIM sits outside protocol addresses, and bad debt stands near $21 million. The proposal describes MIM as “severely under-backed with no viable path back to parity” and puts effective backing at about $0.04 per token, meaning more than 95% of each MIM is unbacked.

One figure needs a word. DefiLlama still shows $4.78 million of total value locked in Abracadabra, split between Ethereum ($2.49 million), Arbitrum ($2.10 million), BNB Chain and Avalanche. That count includes deposits from users who have no debt. The proposal does not treat those deposits as MIM backing, which is why $4.78 million of TVL and $1.2 million of backing can both be true.

How the claim is calculated

The team plans to pull collateral out of the cauldrons with the contracts’ strategy functions, which it says can recover about 99.75% of it, blockchain, où l'on échange des jetons directement depuis son propre wallet, sans confier la garde de ses fonds à une entreprise.">swap everything to WETH and distribute it through a Merkl claim contract. The order of payment is set out in the proposal:

• Borrowers receive the value of their collateral minus their MIM debt, with that debt counted at $1 per MIM.
• MIM holders share what remains pro rata, about $0.04 per MIM on the team’s estimate.
• After six months, unclaimed funds go to MIM holders who did claim, up to $1 per MIM. Anything beyond that goes to borrowers in proportion to their debt.
• SPELL holders receive nothing.

The $0.04 is an estimate. Payment is in WETH, so the dollar value moves with the ether price between the swap and the claim, and the exact WETH-per-MIM rate has not been published.

Why SPELL gets nothing

Legal advice shaped the outcome for SPELL, the text says. Its reading is that MIM, as a debt-backed stablecoin, is a liability of the protocol and ranks ahead of SPELL, and that “SPELL token does not retain any accounting value” until MIM holders are paid in full. They will not be. The text adds that, without an orderly wind-down, users could try to force a liquidation through the courts.

A repeg was ruled out. Raising interest rates to force liquidations would, in the team’s words, give a minutes-long bump to the MIM price that “would benefit only the fastest sellers”.

What changed from the September proposal

Most of the text is the same as the first version. Two changes matter for holders:

• The balance snapshot moved from “not before October 15th, 2026” to “not before October 20th, 2026”. It will be taken after all collateral has been swapped.
• Positions on chains where the admin multisig keys “have not been transferred since the transition of stewardship” will be left untouched, like the immutable cauldrons. The proposal does not list those chains.

The second change means some borrowers may sit outside the redemption entirely and have to withdraw on-chain themselves, if their contracts allow it.

Key dates: snapshot from 20 October, Stargate exit by 15 December

After the liquidation the protocol closes. Immutable positions remain withdrawable on-chain, and the website stays online without maintenance. The proposal is blunt about it: “The team will not bear any legal or technical responsibility to maintain the protocol.”

What MIM holders can check before the snapshot

The snapshot reads on-chain balances. A holder’s claim depends on where the MIM sits when it is taken, so the practical questions come before 20 October, not after. This is information, not advice:

• MIM in a self-custody wallet is counted at that address, since the snapshot reads MIM balances. The proposal lets users set a different claim address before redemptions open.
• Borrowers with an open cauldron position are paid collateral minus debt, which is a different calculation from plain MIM holders.
• MIM held on a centralised exchange sits at the exchange’s address. Whether an exchange claims and passes the WETH on to customers is up to each exchange; we found no statement from any of them.
• MIM deposited in a liquidity pool is held by the pool contract. How such positions are treated is not spelled out in the proposal.
• Under the plan, the claim list goes on Discord and the redemption page will be linked from the main Abracadabra site. Fake claim pages are a known risk around airdrops; our guide to spotting crypto scams lists the usual signs.

From $6.4 billion to $4.8 million in four years

Abracadabra launched MIM in May 2021. Its model let users lock crypto, often yield-bearing tokens, in vaults called cauldrons and borrow freshly minted MIM against them. At its peak on 5 January 2022, DefiLlama counted $6.42 billion locked in the protocol, and MIM supply reached about 3.64 billion the same day.

The fall began weeks later. On 27 January 2022, blockchain investigator ZachXBT identified “0xSifu”, treasury manager of the sister project Wonderland, as Michael Patryn, a co-founder of the collapsed Canadian exchange QuadrigaCX. Both projects were led by Daniele Sestagalli. TVL halved to $3.17 billion by 1 February. The Terra collapse in May 2022 hit next, since MIM had been widely used in Terra yield strategies, and TVL was down to $664 million by 1 June 2022. No 2026 source links Sestagalli to the June handover or the wind-down.

Source: DefiLlama protocol and stablecoin data.

Three hacks in 20 months

The hole in MIM’s balance sheet comes largely from exploits. The proposal itself blames “a series of hacks”:

Each exploit created MIM debt with no collateral behind it. After the January 2024 attack the DAO said it would buy back and burn MIM, and after the October 2025 one it bought the 1.79 million MIM back from the market. Neither closed the gap. TVL dropped from $59.6 million to $43.4 million in the three days around the March 2025 hack.

The June 2026 depeg and the handover

On 2 June 2026, a Snapshot vote handed operations, the treasury and development to a group called Anubis. The proposal named Greg Dewitt for operations and Paul Parker and Glenn Kennedy as directors, with partners Nemesis Trading and Alapin Holdings (dao5). Its stated aim was to revitalise Abracadabra.

MIM lost its peg within weeks. It slipped to $0.74 in mid-June, recovered to $0.89, then fell to $0.49, Cointelegraph reported on 25 June, citing CoinMarketCap prices and thin, lopsided Curve liquidity. The team raised rates on every cauldron and said its priority was to “return MIM to a healthy (and liquid) peg”. DefiLlama’s daily prices show MIM at $0.993 on 1 June, $0.313 on 27 June, $0.101 on 1 July and $0.018 on 28 September, the day before the first wind-down proposal.

Treasury moves that critics call “looting”

The wind-down text does not mention money that left the treasury over the summer. Protos reported on 30 September, citing a post by Pharos, that the treasury sold MIM on Curve on 8 and 11 June for about $0.5 million, and that around $8.3 million moved out on 14 July 2026: some 2,142 ETH and WETH to a Binance deposit address and about $4.3 million in USDC and USDT to an unlabelled address. Mikko Ohtamaa, co-founder of Trading Strategy, described it as “treasury looting”, according to Protos.

These are accusations. No authority has said that any wrongdoing took place, and the team has not answered them in any source we found. Etherscan confirms part of the trail: on 14 July, at 15:52 UTC, the treasury Safe sent about 366.68 ETH to an address Etherscan tags as a Binance deposit, a minute after sending 44.67 WETH to the same address. We have not matched the rest of the 2,142 ETH Protos describes, and we do not know where the funds went after that. Molly White’s Web3 Is Going Just Great noted that the project promised a peg-recovery roadmap in July and then stopped posting. The funds Protos describes are not counted in the redemption plan.

Why MIM was a debt-backed stablecoin, not an algorithmic one

MIM is often lumped in with TerraUSD, but its design was different. UST relied on a mint-and-burn link with another token. MIM was a collateralised debt position (CDP) coin: each MIM was borrowed against crypto locked in a cauldron, and liquidations were meant to keep every token covered. DefiLlama classes it as crypto-backed. Our guide to how stablecoins work explains the main designs.

The weakness of a CDP coin is plain in this case. It is only as good as its collateral and its contracts. When an exploit mints or withdraws MIM without collateral, the debt stays in circulation and nothing stands behind it. Thin liquidity on Curve then turns a solvency problem into a price problem. The same risks run through most lending protocols, as our DeFi explainer sets out.

For European readers, MIM was never a regulated stablecoin under MiCA: it had no authorised EU issuer and no reserve or redemption rights of the kind the rules require. Our MiCA guide covers what that framework demands of dollar and euro tokens.

Open questions before the claim opens

Several points remain unclear on 16 October. The proposal sets no date for the Merkl claim and no exact WETH rate. It does not list the chains left out because admin keys were never transferred. It does not say whether the 1,280 WETH in the treasury Safe is collateral bound for claimants. And the team has given no public account of the July outflows.

We found no statement on MIM from the SEC, the FCA or EU regulators, and no lawsuit. The next fixed point is the balance snapshot, due on 20 October at the earliest, followed by the Stargate deadline on 15 December.