[ARC]: Risk Parameter Updates for Aave V3 Optimism 2022-12-29

[jmo]

Simple Summary

A proposal to adjust one (1) total risk parameter - Supply Cap - across one (1) Aave V3 Optimism asset.

This should reduce risk in the protocol, though it could limit the growth of AAVE token deposits on Optimism.

Abstract

There is currently not a supply cap for the AAVE asset on Aave’s v3 deployment on Optimism. We are going to set this quickly and provide a deeper analysis of all of the current supply caps next week. We believe these could be more conservative, but setting a supply cap for AAVE sooner rather than later will reduce the chance of new positions creating a dangerous spike in VaR.

Beyond this, the cap greatly reduces the viability of market downturn and price manipulation attacks involving the AAVE token.

Motivation

Any market without supply caps could experience an outsize risk by someone opening a very large position quickly. Maintaining supply caps that ensure that one user cannot subject the protocol to outsize risk is a key part of risk management for Aave v3 markets.

Gauntlet’s parameter recommendations are driven by an optimization function that balances 3 core metrics: insolvencies, liquidations, and borrow usage. This change was driven by analysis focusing on insolvency, as estimated by our simulations. Our agent-based simulations use a wide array of varied input data that changes on a daily basis (including but not limited to asset volatility, asset correlation, asset collateral usage, DEX / CEX liquidity, trading volume, expected market impact of trades, and liquidator behavior). The input metrics we show below can help understand why some of the param recs have been made but should not be taken as the only reason for recommendation. The individual collateral pages on the Gauntlet Risk Dashboard cover other key statistics and outputs from our simulations that can help with understanding interesting inputs and results related to our simulations.

For more details, please see Gauntlet’s Parameter Recommendation Methodology and Gauntlet’s Model Methodology.

Supporting Data on Aave V3 Optimism

Top 10 borrowers’ aggregate positions

Top 10 borrowers’ entire supply

Top 10 borrowers’ entire borrows

Aave V3 Optimism Parameter Changes Specification

Parameter	Current Value	Recommended Value	Current AAVE Supply
AAVE Supply Cap	None	100,000	87,980

Why would we limit the AAVE supply?

There’s limited AAVE liquidity on Optimism on Uniswap, with $121k AAVE/ETH being the largest AAVE pool. The current AAVE supply on Optimism Aave v3 is $4.7M, so a liquidator will not be able to arbitrage a large AAVE position solely on Optimism DEXs. However, the liquidation bonus at 10% is high enough that we anticipate liquidators should be willing to hold or bridge AAVE. As the markets on Optimism grow, AAVE will rely more and more on the asset bridge and available cross-chain liquidity. This is why we mentioned that all of these caps could be more conservative, and look forward to sharing more info there next week. If the Aave community wants to take a conservative stance on risk, we should consider lowering these across the board.

Aave Token Balances Since Oct 30th

---

---

Next Steps

• Initiate Snapshot vote immediately since the community has recently weighed in on changes of this nature.
• Targeting an AIP on [2023-01-02]

Quick Links

Analytics Dashboard
Risk Dashboard
Aave Arc Risk Dashboard
Gauntlet Parameter Recommendation Methodology
Gauntlet Model Methodology

Notes

These parameter updates are a continuation of Gauntlet’s regular parameter recommendations. Our simulation engine has ingested the latest market data to recalibrate parameters for the Aave protocol. The community has aligned on a Risk Off Framework regarding lowering liquidation thresholds.

^{By approving this proposal, you agree that any services provided by Gauntlet shall be governed by the terms of service available at gauntlet.network/tos.}

[ChaosLabs]

After reviewing the post, we wanted to share our reservations about the proposed recommendation and share our analysis before Gauntlet moves forward with an AIP.
Moving forward, we’d appreciate if Gauntlet could share their analysis before posting to avoid community confusion, as already discussed on several threads.

After looking into the state of AAVE token supply on Optimism, we have identified that the current supply of AAVE tokens on the v3 Optimism market (~88K) significantly exceeds the available liquidity on Optimism.
The high exposure is mainly due to a single position that accounts for 87% of the supply (~79K of ~88K)

Following our Supply Cap Methodology, we recommend setting the AAVE token on Optimism at 14K, based on the current liquidity levels on Optimism (detailed analysis below)

We are aware that setting a supply cap below the current supply is not the best practice. However, the current state is a market distortion that was caused as a result of there being no cap for AAVE on Optimism. Practically, if this single position highlighted above were to be closed, the recommended supply cap would be around 40% above the current supply, which is very reasonable from a usability perspective.

Analysis of AAVE token supply on Optimism

When observing the accounts that supply AAVE tokens to AAVE V3 on Optimism we have identified a single account supplied 79k AAVE position, 87% of total supply, at current health of ~1.6.

The approach to setting the supply cap for AAVE tokens on Optimism

This amount cannot be liquidated with the current AAVE liquidity on Optimism. While this account is an active, healthy account, it is still a risk for the protocol that should have been avoided had supply caps been in place. Therefore, the approach we took was to set the supply cap using the methodology according to the distribution of the other accounts, excluding the account with the biggest position. The reasoning behind this is that while this position is open, there is risk exposure that we cannot mitigate without forcing liquidation of the position. Therefore, the best approach would be to set appropriate risk parameters that will secure the protocol if this position is closed.

The recommended Supply Cap

Based on the approach described above, our simulations have produced r=0.278, which yields: (currentSupply - supply of 0x36..a8a)/r*marketCapFactor = 10k / 0.278 * 0.39 = 14K.

Our recommendation is to set the Supply Cap for AAVE tokens on Optimism at 14K.

While this cap effectively pauses AAVE supply, until the position opened by 0x36c4bd54d54dd898c242f5f634f5d0cef3be2a8a is closed, we believe it is the appropriate cap that will secure AAVE V3 on Optimism.

[fig]

There are a lot of variances between these two recommendations - 100,000 vs. 14,000

Before heading to a vote, I’d love to see further discussions between these two managers to help the community distill which direction is best.

Another option is to initiate a modified Snapshot with both options:

• Adjust supply cap to 100,000 AAVE

• Adjust supply cap to 14,000 AAVE

• NAY - do not adjust

• Abstain

We want to ensure the proper risk management efforts are achieved promptly and hope to help facilitate this.

[ChaosLabs]

Thanks for this response @fig. Effective communication between risk managers is crucial to increase community clarity and a better understanding of proposals. We have been working on a proposed framework for risk manager collaboration which we plan to post later this week for community feedback.

Regarding the proposals:
The main reason for the conservative figure on our proposal is that we did not assume liquidity bridging in our methodology and recommendation - it was based solely on the current liquidity levels in Optimism.

Given the assumption of bridging liquidity, the recommendation made by Gauntlet is reasonable. However, the proposal needs to include on-chain data or analyses to support it, leaving us uncomfortable relying on theoretical actor rationality. We urge Gauntlet to share any supporting data used in this analysis, as this will help inform the community about this decision.
Additionally, if other chains are indeed considered for liquidations, what methodology or heuristic was used to determine the 100k figure? Aave is a relatively liquid asset and could be more significant than 100k if that is the approach we’re taking as a community.

[jmo]

@fig I’m open to publishing a snapshot with both options, I think that will be a reasonable way to let the community weigh in here.

However - as I mention in the first post - we do believe that the supply caps on Optimism could be more conservative. We’d be happy to set a lower supply cap here - our goal with this parameter update was to just limit the risk and we can fast follow with more detailed recommendations. I don’t think 14k is crazy - though it does freeze the growth of the market as @ChaosLabs points out.

The issue here is that if we were to apply the same logic across the board, it would likely more or less freeze all assets on Optimism as there is not great on-chain liquidity for the other assets as well. That is to say - following the logic you delineate here, we would likely have to set lower caps for all of the assets on Optimism.

Our goal is to propose a set of caps to the community next week that would allow the community to decide “Do we want to bet on cross-chain liquidity here?” If the answer is yes, then we can likely move forward with caps that look more like the current ones, as opposed to caps that instant freeze new deposits for assets. However, if the answer is no, then this amounts to winding down the Optimism market quite a bit, which we know was a market the community was betting on from a growth perspective.

I’d prefer that we make this decision for the community as easy as possible. Instead of setting a conservative AAVE supply cap as a one-off, our goal was to set one quickly to limit risk increasing in the market and then follow up with conservative supply caps for all assets next week.

This path should make the decision making for the community as clear as possible - it’s sometimes hard for people to weigh in on individual parameters like the supply cap for one asset on one market. Hopefully if we frame the decision more broadly it will help the community make an informed decision here.

[ChaosLabs]

@jmo

The issue here is that if we were to apply the same logic across the board, it would likely more or less freeze all assets on Optimism as there is not great on-chain liquidity for the other assets as well. That is to say - following the logic you delineate here, we would likely have to set lower caps for all of the assets on Optimism.

The supply cap recommendations adopted thus far have not frozen any assets on Optimism. Most supply cap recommendations have room for growth, allowing healthy usage and market participation. Additionally, as stated in the methodology post, we can make several relaxations to enable higher caps, given the community’s desired risk appetite.

@jmo

I’d prefer that we make this decision for the community as easy as possible. Instead of setting a conservative AAVE supply cap as a one-off, our goal was to set one quickly to limit risk increasing in the market and then follow up with conservative supply caps for all assets next week.

If you plan to recommend conservative caps next week, wouldn’t it make more sense to start with a risk-off “temporary cap” rather than one that relies on the assumption of bridging liquidity? We agree that a cap should be implemented ASAP. However, since Gauntlet has yet to provide a clear framework or supporting data for this proposal, we recommend implementing the 14K supply cap we proposed for the immediate term while the community deliberates on this methodology for higher caps.

@jmo

Our goal is to propose a set of caps to the community next week that would allow the community to decide “Do we want to bet on cross-chain liquidity here?”

We look forward to your proposal next week. The discussion around setting caps and defining risk parameters should always be data-driven, and we hope to see this incorporated into a defined framework or methodology. We are eager to see your analysis on liquidity bridging, and if the community decides to adopt this concept, we will incorporate it into future versions of our methodology.

Chaos Labs has been operating transparently and sharing the considerations, constraints, and assumptions of our recommendations and methodologies. We expect the same from Gauntlet, as this approach will improve overall protocol risk management while allowing the community to make an apples-to-apples comparison when assessing methodologies and determining risk appetite.

[Pauljlei]

We have moved forward with a AIP to add a supply cap to the AAVE token on V3 Optimism in order to immediately reduce tail risk without damaging user experience in the short term. The rationale is straightforward - immediately reduce long tail risk without any impact on user experience. No cap is significantly riskier than having a cap .

The forum thread brought up valuable points about assumptions and methodology. We look forward to iterating on that with the community moving forward, which of course, may result in adjusting the 100K cap. But in the meantime, we are biasing towards action to set a cap on a market that currently lacks a cap, while having longer discussions about assumptions in parallel.

[PennBlockchain]

We’re reading the reasoning here and agree with the point of having SOME cap vs none. Whether that’s 100k or even less at 14k, whatever it is, starting with ensuring a high cap is immediately passed is critical in our opinion. We believe the 100k high cap that has been proposed is a good start and further decreasing/changes can be discussed with a “hedge” in place.