Aave (AAVE) AAVE

AAVE is the governance token of Aave, the decentralised lending protocol, and its supply is capped at 16 million. Anyone searching for Aave crypto should start with what the token does not do: it pays no automatic dividend, because the protocol’s fees go to a DAO treasury. Buybacks funded from that treasury stopped on 19 April 2026, and the records we could check showed no restart as of 2 November.

This page covers the token: its supply, its votes, staking, the fights of 2026 and what can go wrong for a holder. The mechanics of lending are on our Aave protocol page. Statuses below are dated as of November 2026 and come from Aave’s documentation, its governance forum and Snapshot records. Price and market data sit in the live panel on this page.

What is Aave crypto? One name, two things

The protocol is Aave; the token is AAVE. You do not need the token to use the protocol: anyone can supply assets or borrow against them without holding a single AAVE.

Suppliers deposit assets into pools and receive aTokens that track their balance plus interest. Borrowers post collateral and pay interest, and a position that falls below its safety margin gets liquidated. Our guide to crypto lending and borrowing walks through it. Aave V4 went live on Ethereum on 30 March 2026 with a hub-and-spoke design, and V3 still runs on many networks.

DefiLlama showed about $16.5 billion of value locked in Aave and about $11.9 billion of active loans at about 14:00 UTC on 2 November 2026. Over the previous 12 months it counted about $634 million in fees, most of it paid to suppliers, and about $90 million as the protocol’s own revenue.

From ETHLend to AAVE: the token’s history

The token started under another name. Stani Kulechov’s ETHLend, a peer-to-peer lending platform on Ethereum, sold its LEND token in a sale that closed on 30 November 2017. The team’s own post put the raise at $16.2 million, while other trackers show different totals. Etherscan records about 1.3 billion LEND.

By 8 January 2020, when Aave’s changelog dates the launch of V1 on Ethereum mainnet, the project was called Aave and had moved to shared liquidity pools. The token followed that year. AIP-1, the first Aave Improvement Proposal, was written by Marc Zeller on 21 September 2020 and did two jobs: swap LEND for AAVE and start the Safety Module. Decrypt reported on 26 September that the four-day vote “appears largely ceremonial”.

The migration went live on 2 October 2020 at 100 LEND for 1 AAVE. Glassnode counted over 70.5% of LEND migrated in under two weeks. The stragglers never fully arrived: a proposal to wind down the migration contract passed Snapshot in November 2025 and reached the on-chain stage in May 2026, over the objection of Zeller, who said it had moved without discussion. TokenLogic had counted about 305,000 AAVE in that contract in September 2025. It held none when we read it on 2 November 2026, which points to the wind-down having gone through, though we did not find the vote record itself.

AAVE tokenomics: a 16 million cap, a reserve and buybacks on hold

Etherscan shows a maximum total supply of 16,000,000 AAVE. Glassnode’s 2020 breakdown gives the split: 13 million minted for LEND holders and 3 million sent to the Aave Ecosystem Reserve, the fund that pays staking rewards and other governance-approved grants.

That cap is a rule governance could in theory touch. The contract is an upgradeable proxy, according to Etherscan, so nobody should call the supply immutable. Aave’s Governance Framework v2 routes any upgrade of the AAVE token through the Long Executor, its strictest path, with a 10-day vote and a 7-day timelock.

Buybacks began in April 2025. TokenLogic, a DAO service provider, reported in February 2026 that the programme had acquired more than 205,000 AAVE, over 1.28% of total supply, in under a year. The purchases were not burned. The 2025 mandate sends them to the Ecosystem Reserve, and TokenLogic describes them as a way to fund staking emissions and service providers, so bought tokens can flow out again.

The budget was already being cut before the crisis. A Snapshot vote in March 2026 backed a cut from about $50 million to about $30 million a year by 363,253 votes to 3,199, after TokenLogic warned of a structural deficit: an optimistic 2026 budget of $190 million against 2025 revenue of $142 million.

Then came the rsETH exploit. TokenLogic’s pause proposal reads: “AAVE buybacks are paused effective April 19, 2026, and will remain paused until the situation surrounding the rsETH incident becomes clearer.” Snapshot voters ratified it between 28 April and 1 May, with 617,169 for, 4,893 against and 158,478 abstaining. DefiLlama’s holder-revenue series, which counts buybacks, shows its last entry on 19 April.

Aavenomics 3.0: announced, not shown to be running

The next tokenomics plan carries the name Aavenomics 3.0, and coverage of it conflicts. On 25 June 2026 Kulechov wrote on X that it “will have immutable and automated buybacks of AAVE”. The Defiant, in an undated article, and Whale Alert on 27 June then said it was live.

The governance record does not support that. Forum members were still asking for a buyback roadmap on 29 September. That same day Crypto Briefing reported that Kulechov had floated a burn, without saying how it would work. Reports dated 23 October describe a livestream in which he set out buybacks, staking incentives and a burn together, with repurchased tokens split between staking pools and a burn address.

No proposal on Aave’s Snapshot space since January 2026 approves Aavenomics 3.0 or a buyback restart, and we found no specification on the forum. DefiLlama still showed zero holder revenue over the past 30 days. As of 2 November, the plan has no published burn size, no funding rule and no start date that we could verify. No AAVE burn has been confirmed, so calling the token deflationary would be wrong.

What the AAVE token gives holders, and what it does not

Votes are the core right. Aave’s documentation counts the combined balance of AAVE, stkAAVE and aAAVE (AAVE supplied to the protocol) that an address holds or has been delegated, on Ethereum mainnet. Holders can delegate voting power and proposition power separately. Our guide to governance tokens and fee switches explains the model.

A proposal starts on the forum, gets four days of feedback as an ARFC, then a binding Snapshot vote that needs 320,000 AAVE; authors outside the approved list need 80,000 AAVE of proposition power. It then goes on-chain as an AIP, with a quorum of 320,000 and a timelock before execution. The standard route takes about 13 days.

The fee right does not exist. Interest margins, liquidation fees and flash-loan fees go to the DAO’s Collector, not to holders and not to stakers. Aave Labs’ own 2026 framework restates that the DAO receives 100% of protocol fees. Value reaches a holder only if the DAO votes to spend treasury money on something that touches the token, such as buybacks or staking rewards. A “fee switch” for stakers was proposed in the 2024 Aavenomics temperature check. We found no sign that it is active.

The GHO perk has also changed. stkAAVE holders once borrowed GHO at a discount of up to 30%. The Aave V3.4 design removed that native discount, and Aavenomics planned “anti-GHO”, a non-transferable credit for stakers, in its place. We found no official confirmation that anti-GHO has launched, and no fee discount for holding AAVE elsewhere on V3.

AAVE staking: stkAAVE and Umbrella are not the same thing

Two staking systems now sit side by side, and the newer one does not take AAVE at all.

Umbrella, live since 5 June 2025, accepts aTokens (aUSDC, aUSDT, aWETH) or GHO on Ethereum. Stakers earn Safety Incentives, which can be paid in tokens such as AAVE, GHO or USDC, plus the aToken yield where there is one. In return they cover bad debt on the asset they staked: slashing fires automatically once a deficit passes a DAO-funded offset. The cooldown is 20 days, followed by a 2-day exit window, and funds stay slashable during it.

stkAAVE is the older route, and the one that does take AAVE. AIP-1 started it in 2020 at 400 AAVE a day, and the original design allowed up to 30% of the module to be used in a shortfall. That cover has since been wound down: a TokenLogic proposal of September 2025 moved to cut slashing from 10% to 0%, and the stkAAVE contract returned a maximum slashable share of 0 when we read it on 2 November 2026. Its exchange rate still stood at 1 to 1, so stakers had not been slashed. Governance can change these settings, and Umbrella’s documentation still describes the slashing shutdown as tied to Umbrella reaching sufficient scale.

Rewards come from the reserve and have been falling. TokenLogic lists a path of 385, 360, 315, 260 and then 220 AAVE a day. A May 2026 proposal, which passed Snapshot by 435,812 to 743, set 150 AAVE a day to target about 2.75% a year, and the contract confirms that rate on-chain. With about 2.48 million stkAAVE outstanding, or 15.5% of the cap, 150 a day works out at about 2.2% a year by our arithmetic, before compounding. That is a reading for 2 November, not a promise.

Leaving is quicker than it was. A March 2026 vote approved a 2-day cooldown instead of 7, and the contract now shows 2 days, followed by a 2-day exit window. Stakers who miss that window have to start the cooldown again.

AAVE voting power: no October vote drew more than 140 addresses

Etherscan listed about 191,000 AAVE holders on Ethereum on 2 November 2026. Far fewer vote. On the Snapshot vote for the Aave Foundation, Phase 1 (7 to 10 October), 77 addresses cast 572,726 AAVE, about 3.6% of the cap, all in favour. By our calculation from Snapshot’s data, the top five held 90.2% of the power cast and the largest alone 33.2%.

That is one vote, not a map of who owns AAVE. The figures describe addresses, some of them delegates voting tokens they do not own, and they say nothing about the holders who stayed away. The other October votes drew 76 to 140 addresses and roughly 360,000 to 414,000 AAVE, not far above the 320,000 threshold.

Nothing on-chain stops a conflicted vote. In February 2026 an Aave Chan Initiative (ACI) proposal for mandatory disclosures and conflict-of-interest norms failed on Snapshot, 688,619 against to 603,607 for.

2026 for AAVE: a governance war, an exploit and a freeze

Aave Labs against the DAO

The dispute opened in December 2025, when Aave Labs changed the swap integration on its front end and delegates asked why the fees did not reach the DAO. Labs answered with the “Aave Will Win” framework. The temperature check in February passed narrowly, 622,256 to 497,050 with 64,193 abstaining. The ARFC cleared Snapshot in April by 727,572 to 174,330, and the on-chain funding vote passed on 12 April with 522,780 for and 175,310 against, about 75%, according to The Block.

Labs got $25 million in stablecoins and 75,000 AAVE, about 0.47% of the cap, vesting over four years. The DAO got a commitment that 100% of revenue from Aave-branded products goes to its treasury. BGD Labs and Chaos Labs ended their work for the DAO around the vote, and ACI cast the largest vote against, 166,200 AAVE.

The vote did not settle the brand. An ARFC that passed Snapshot in October sets up a memberless Cayman Islands foundation meant to hold the trademark, but Phase 1 covers only incorporation and the first appointments. We did not confirm its on-chain stage. Transfers of the trademark, domains and code each return to governance.

The rsETH exploit

On 18 April 2026 an attacker obtained 116,500 rsETH through Kelp DAO’s bridge route and deposited 89,567 of them on Aave, borrowing 82,650 WETH and 821 wstETH, according to the incident report by Aave service providers. It states: “Aave’s smart contracts were not compromised at any point during this event.”

The same report modelled bad debt of about $124 million if losses were spread across all rsETH, or about $230 million if isolated on layer-2 networks. Those are scenarios. We found no settled final figure. The DAO then backed the DeFi United recovery effort with 25,000 ETH, on a Snapshot vote of 371,482 to 144, and Aave Labs said the follow-up AIP would liquidate the attacker’s position without exposing Umbrella stakers to slashing. For holders, the bill came as a smaller treasury and the buyback pause: TokenLogic’s May proposal says capital was redirected to the balance sheet after the 25,000 ETH went out.

The Binance hack spillover

On 29 October a wallet linked to the Binance hot-wallet exploiter supplied 60 million USDT to Aave V3 and borrowed 19,762 WETH against it. Guardians froze 292 reserves across V3 and V4, and LlamaRisk said the exploiter was selling tokens including AAVE on decentralised exchanges. Repayments and withdrawals stayed open, and LlamaRisk said the position did not put suppliers at risk.

Aave reopened its markets on 30 October, but not fully. Nine assets, USDT and USDC among them, kept a zero loan-to-value setting so they could not back new loans; our report covers how Aave reopened with nine assets at zero LTV. As of 2 November 2026 that limit was still in place: USDT and USDC both read 0 on Ethereum Core, and users were asking on the forum when it would lift.

A second freeze followed. Guardians froze tBTC on Aave V3 Ethereum Core at 10:01 UTC on 31 October after Threshold paused minting and redemptions, and the reserve still read frozen on 2 November. LlamaRisk has said it will propose restoring it once tBTC operations are back to normal.

AAVE risks a holder should weigh

Value capture comes first. Nothing obliges the DAO to send revenue toward the token, and the reserve keeps paying out AAVE through staking rewards and the Labs grant while buybacks are off.

Technical failures have been small so far on the core contracts, though not absent. DefiLlama lists about $56,000 lost in August 2024 through an access-control flaw, and about $862,000 in March 2026 from an oracle misconfiguration. The bigger losses came from collateral, as rsETH showed: a token Aave lists can fail outside Aave.

Staked funds can also get stuck. During the rsETH crisis, Aave Labs proposed pausing the Umbrella aWETH stake token, a state that blocks deposits, withdrawals, transfers and slashing while rewards continue. That token read as unpaused on 2 November 2026.

Regulation remains open. Kulechov announced on 16 December 2025 that the SEC had ended a four-year investigation, citing a letter dated 15 August in which staff planned not to recommend enforcement action, The Block reported. That is a staff decision, not a ruling on whether AAVE is a security. In the EU, the MiCA approval of 13 November 2025 belongs to Push, an Aave Labs company that runs stablecoin on-ramps, and we found no authorisation covering the token. Our guide to DeFi regulation under MiCA and in the US sets out the open questions.

Copies are a trap for the unwary. CoinGecko lists AAVE contracts on Base, Arbitrum, Optimism, Polygon, BNB Chain, Avalanche and Solana, among others, but voting counts balances on Ethereum mainnet, and we could not verify which of these copies Aave governance endorses. Nothing on this page is investment advice.

Where to buy AAVE

We checked all ten exchanges below on 2 November 2026, on their own pages or public market data. Check that the service is open to you, since availability varies by country.

• Euro pairs: Kraken and Bitvavo both list AAVE/EUR. Bitpanda sells AAVE through its buy page.
• US dollar pairs: Coinbase and Gemini list AAVE/USD, and Crypto.com lists AAVE_USD and AAVE_USDT.
• USDT pairs: Binance, Bitget, Bybit and OKX. Binance has no AAVE/EUR spot pair.

Anyone buying to vote needs the Ethereum token, so check the network before withdrawing. Two records will show whether anything has changed for holders: the next TokenLogic funding update and any Snapshot vote carrying the Aavenomics 3.0 name.