A proposal to update four (4) risk parameters, including Liquidation Threshold (LT), Loan-To-Value (LTV), Liquidation Bonus (LB), and UOptimal for FEI on Aave V2 Ethereum.
Motivation
As part of the ongoing risk mitigation strategies implemented for V2 Ethereum, this proposal suggests another series of risk parameter updates to minimize potential exposure to FEI. As a reminder, FEI has been frozen since September 2022, with the RF set at 99%.
Over the past months, we have witnessed a significant drop in market cap, and together with the recent price fluctuations, we recommend decreasing Aave’s FEI exposure to a minimum.
Lowering LT and LTV to zero - This will prevent new borrowings against FEI and potentially further protocol losses should FEI’s price fluctuate due to limited liquidity.
By reducing the LT to zero, two additional accounts will become eligible for liquidation, totaling a value of $1,051. The main account to be liquidated can be found here
Raising the LB from 6.5% to 10% to provide a stronger incentive for liquidators to address underwater positions.
Decreasing the UOptimal from 80% to 1% to increase the borrowing rate and encourage borrowers to repay their loans.
Specification
Asset
Parameter
Current
Recommendation
FEI
LT
75%
1%*
FEI
LTV
65%
0
FEI
LB
6.5%
10%
FEI
UOptimal
80%
1%
*For technical reasons related to V2 we will set the LT 1% rather than at 0
Next Steps
Following community feedback, submit the ARFC for a snapshot vote for final approval.
If consensus is reached, submit an Aave Improvement Proposal (AIP) to implement the proposed updates.
In addition to the changes proposed, we suggest including another one: update the price feed of FEI on Aave v2 to an adapter composed of the Chainlink ETH / USD with a fixed percentual discount.
Rationale
On an asset that is clearly in advanced stages of offboarding and deprecation, having a price feed based on secondary market liquidity is a liability, because liquidity is thin and will only get thinner, over time.
Thin liquidity introduces a risk of price manipulation in what concerns the Aave protocol, so even with a frozen asset, this risk could materialize through existing user positions.
What are the practical consequences of the price feed swap?
The current status is the following:
There are ~15’000 FEI supplied and ~4’600 FEI borrowed on Aave v2 Ethereum.
The asset is frozen, so there can’t be new deposits or borrowings, only repayments, withdrawals, and liquidations.
The asset is enabled as collateral.
Now, let’s assume 2 types of price manipulation on FEI, down and up:
Down. When a price is manipulated down, borrowers of the asset are benefited, and depositors are harmed, as it will potentially affect their withdrawal claims. Considering that no new borrowings can be opened on FEI, and also that a good percentage of the active borrowers will be liquidated with the proposed changes by @ChaosLabs , the potential exposure is most probably in the order of hundreds of dollars. And that is assuming a price manipulation, which is not realistic by itself.
So there is not really any risk in the down-side manipulation scenario.
Up. When a price is manipulated up, this allows for the depositors of the affected asset to borrow more of other assets, or withdraw other collateral, if Liquidation Threshold allows so. In practice, price manipulation up is equivalent to artificially inflating the Liquidation Threshold of an asset.
Even with the proposed LT change to 1%, theoretically could be possible to apply this on the FEI case, even if in practice highly improbable or even impossible, considering liquidity on the secondary market, mint/burn dynamics, or sizes of aFEI positions.
So the current scenario (again, both are only theoretical) is 1) manipulation down doesn’t affect the protocol 2) manipulation up could in a really really edge scenario, in practice maybe impossible.
By swapping the FEI/ETH price feed to an ETH/USD * (1 - discount), the consequences are:
Full theoretical protection against any manipulation.
FEI price is disconnected from its “real” price, but given the 1% LT, this doesn’t really matter for the protocol.
No requirement anymore from Aave for Chainlink to run the FEI/ETH feed, which can become more and more fragile as liquidity gets thinner.
Aggregating the previous, the net impact of the price feed swap is positive, and so our proposal.
