Risk-Off Framework for the Aave Protocol

Rate of Change Consensus

To achieve sustainability, Aave must combine value creation under favorable market conditions with protection of itself and its users when conditions become unfavorable [1].

Aave must be able to lower the liquidation threshold of collateral assets to reduce the risk it and its users face under those adverse conditions. While loan-to-value and liquidation bonus are also useful means of reducing the risk of insolvency, they are not effective substitutes.

Over the last several weeks, Gauntlet has collaborated with BGD Labs, Llama, Governance House, and other community stakeholders to align on how to most effectively serve the protocol regarding the tradeoffs between value creation and protection.

Liquidation threshold reductions can impact users in a number of ways. In particular, if a user’s account health is low, this type of parameter change can render the account eligible for liquidation.

If a user rebalances their position ahead of a liquidation threshold reduction, they will avoid liquidation. For this reason, Gauntlet will provide notice to the community via communication channels that may include the forums and Twitter, among others.

That said, there is always some chance that a user will not monitor their health factor or fail to supply collateral in time, resulting in “risk-off liquidations”.

BGD, Llama, Governance House, and other community members have emphasized that these liquidations impinge on the borrower user experience. It is therefore important that the community explicitly aligns on a desired maximum reduction to liquidation thresholds per AIP.

Community input is also valuable because limits on LT reduction impose an important trade-off:

  1. A lower max LT reduction will minimize the impact to borrowers but at the cost of reducing the rate at which risk can be taken off. This, in turn, limits the extent to which leverage can be increased during favorable conditions, making Aave less capital efficient and coming at the cost of borrow volume.
  2. A higher max LT reduction will strengthen Aave’s ability to maximize capital efficiency under favorable market conditions but increase the potential for risk-off liquidations.

A Snapshot vote will be published with the options being the following: Per AIP, each collateral asset’s liquidation threshold may be reduced by an absolute percentage of up to:

  • 3% (see point (2) above)
  • 2%
  • 1% (see point (1) above)


  • This limit would not apply under severe conditions, such as a contract vulnerability.
  • It would apply to all collateral assets Gauntlet supports, including WETH, WBTC, DAI, &c. [2].

Next Steps:

Gauntlet will create a Snapshot vote immediately. Voting ends on Wednesday, July 27.

[1] Please find a detailed explanation of Gauntlet’s Risk Management methodology here.

[2] A complete list of currently supported assets can be found here.


Snapshot vote here: Snapshot

Hello @Pauljlei & Gauntlet.

Thanks for this proposal, liquidation is the center of gravity of Aave protocol resilience, that being said most DeFi stills inherit conservative parameters as an artifact of a time where DeFi was a below 1B$ market and liquidity was 1% of existing now.

high liquidation penalties & large buffer threshold are in most volatility events probably a too good deal for liquidators but as crypto experience a black swan every other week, this proposal makes sense to allow increased buffers in dire times.

That’s why I’m supportive of this proposal.

That being said,

This proposal should have stayed a a few days in the governance thread period to allow the community to discuss it, starts the snapshot vote at least 24h after snapshot publication, and should have at least an NAE and abstain option to reflect variety of community sentiment.

Current options are a bad precedent as the current snapshot community vote only allows voters to agree with the proposal with different parameter values.

Decentralized protocols governance should always leave room for community rejection or abstention no matter how beneficial the proposal is.

I would advise Gauntlet to consider community guidelines from now on. The proposed snapshot is removed and I’ll invite you to repost it in a standard way.


Hi @MarcZeller, thanks for the feedback. We considered adding zero, but it’s equivalent to not managing risk at all - as such, it is not an option if we want Aave to be sustainable over the long term. We care about giving the community enough time to discuss and comment - this is why we made the Snapshot voting period 2 weeks, even though it was put up immediately after the forum post.

Following this feedback, we will allow more time for forum discussion and put up another Snapshot vote on Thursday, July 21, with the additional options of “0%” and “Abstain.” The Snapshot will begin voting on Friday. We reiterate that 0% is not a productive step forward.

Additionally, it would be good to revisit the community guidelines, as the community and protocol have changed significantly since those were originally established.

1 Like


You don’t get to decide on behalf of the Aave governance what’s good or not for the protocol.

The Aave governance is not soviet Russia, the AAVE tokens holders are the owner of the Aave protocol and they will decide what’s best for the protocol and our duty is to leave room to the diversity of different opinions on any topics.

It’s perfectly fair and square to explain why a proposal should be supported but I welcome that the new snapshot allows expression of the spectrum of possible opinions on the proposal.


Completely agree here.

Snapshot vote below. Voting period is from July 22, 12 PM PT to July 29, 9 PM PT.


1 Like

The Snapshot vote has now concluded. A total of 2,575 accounts have voted. Results below:

  • 3%: 466K AAVE (99.64% of votes)
  • 1%: 1.1K AAVE (0.24% of votes)
  • Abstain: 249 AAVE (0.05% of votes)
  • 2%: 210 AAVE (0.04% of votes)
  • 0%: 135 AAVE (0.03% of votes)