[ARC] Gauntlet Borrow/Supply Cap Updates for Aave V3 OP and Aave V3 AVAX (2023-01-27)

[Pauljlei]

Simple Summary

After conducting research, we provide 2 options for borrow and supply cap changes for the Aave community.

These options depend on different assumptions that would be valuable for the community to align on.

Motivation

This set of borrow and supply cap recommendations will use the methodology outlined in our borrow and supply cap methodology.

One major consideration in our conservative recommendations is the frequency of liquidators using local-chain liquidity for liquidations. For Avalanche, in the last 6 months, 69.4% of liquidations involved a swap using on-chain liquidity. For Optimism, over the previous 6 months, 96.2% of liquidations involved a swap using on-chain liquidity. These indicate that on-chain liquidity will play an important role in the success of liquidations and insolvency mitigation. However, cross-chain liquidity considerations would allow for more aggressive caps. The table below will present both the conservative and aggressive values.

We highlight below the cap recommendations that are below the current levels of market supply and borrow.

TL;DR - if the community believes in the assumption that local-chain liquidity is necessary for risk mitigation in the Aave protocol, then the levels of current supply and borrow in the Avalanche and Optimism V3 markets pose outsized market risk. The empirical data suggests that local-chain liquidity is indeed important for risk mitigation. However, there are tradeoffs here as well. There is a case to be made by community members focused on growth that the presence of Aave itself on these chains will, in the long term, lead to more activity and DEX liquidity on Avalanche and Optimism. We hope that this initial post clarifies the risk tradeoffs.

Tradeoffs

Conservative Recs:

• Holistically mitigate the insolvency risk by factoring a wide array of market risks, including:
  • Liquidity concentration risk (total and per chain). If protocols allow excessive amounts of a token as either supply or borrow, market makers and users may not have enough access to the token to buoy its liquidity in times of market stress.
  • Overall DEX liquidity and volume. While the previous point addresses liquidity across each chain and all chains, maintaining robust DEX liquidity is also essential for liquidators.
  • Insolvency risk as estimated by our simulations. We can limit insolvency risk by setting caps that limit the Value at Risk (VaR), as calculated in our simulations, in relation to total reserves held in a protocol.
• Utilize conservative liquidity assumptions, which, as shown above, is what liquidator behavior empirically suggests is appropriate.
• However, conservative recommendations would freeze supply for nearly all markets.

Aggressive Recs:

• Increase insolvency risk levels from the market risks listed above (liquidity concentration risk, DEX liquidity, etc.).
• Open the protocol to greater risk of price manipulation attacks, especially on illiquid tokens. These attacks rely on a minimum number of tokens to supply or borrow in order to be profitable.
• Rely on the assumption that liquidators are able and willing to use cross-chain liquidity. The protocol faces insolvency risk if this assumption does not hold.
• However, aggressive recommendations would allow for more growth on the protocol.

Alternative Options:

• For assets that are not highly utilized, freeze the supply (by setting the conservative supply caps). For assets that generate revenue (majors like ETH and WBTC), allow for a level of organic growth (take on greater risk).
• Delist the higher-risk assets from the protocol.

Other considerations:

• Migrating user positions from V2 to V3.
• If we observe more empirical evidence for cross-chain liquidity, we can relax the assumption that liquidators heavily rely on local-chain liquidity.

Aave V3 Parameter Changes Specification

Gauntlet Recommendations vs. Current Parameters

Gauntlet Recommendations with Liquidity Inputs

Notes

• Values highlighted in red denote our borrow and supply cap recommendations that are less than the current borrow or supply, respectively. While these recommendations would effectively freeze certain markets, they would also mitigate risks we have highlighted in our methodology. Rows highlighted in blue are stablecoins.
• Bridged assets: There are slight modifications to the supply cap methodology for specific bridged assets. For BTC.b’s aggressive supply cap, given our current methodology, the suggested supply cap would have been less than 2800 BTC.b. Given the liquidity and trading volume of the underlying asset BTC, we updated BTC.b’s aggressive cap higher to encourage protocol usage and growth.
• Interplay with other risk parameters: The focus of our supply cap methodology is around liquidity and on-chain supply. Asset volatility, asset correlations, and user behaviors are the focus of our LTV, LT, and LB recommendation methodology. Therefore simulated insolvency risk is not a focus of our supply cap methodology but is a consideration. We will also take isolation mode and potential debt ceiling values into account, but for consistency, the supply caps for USDT, FRAX, and MAI will be shown.
• Stablecoins: We do not think borrow caps on stablecoins will significantly impact the risk profile of the protocol. In addition, there is a significant capital efficiency consideration with allowing for high utilization of stablecoins for suppliers.
• Staking derivative tokens: For AVAX, the borrow caps that our methodology suggests are lower than the current borrow on the protocol. For staking assets in general, there is a distinct use case for borrowing the asset. However, this increased borrow interest does not offset potential liquidity risks for the underlying asset nor the staked asset (recall stETH-ETH price deviations and the potential impact on those reserve pools). We note that given liquidity considerations, both the aggressive and conservative borrow cap recommendations for AVAX will be lower than the current borrow.

Next Steps

• Welcome community feedback.

^{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.}

[Pauljlei]

Given the importance of aligning on assumptions and risk preference, we plan on putting up a Snapshot vote on Thursday, 2/9/2023, with the below options. Please note that Conservative would effectively freeze supply for many markets (highlighted in red above).

1. Conservative
2. Aggressive
3. Freeze supply only for smaller markets (allow markets for majors like BTC room to grow)
4. Abstain
5. Nay

[WintermuteGovernance]

Thanks for this Paul, we appreciate Gauntlet being proactive in setting risk parameters.

I’ve attached some tables so that readers can use to better understand the magnitude of the proposed changes:

Blue and red remain consistent with your table above, and green simply highlights an increase in parameters for certain markets.

Importantly, you can see the difference ETH mainnet liquidity makes for certain assets under the aggressive supply cap recommendation.

In general, we support:

• The aggressive supply caps for stablecoins across both Avalanche and Optimism. Given the negligible carry risk for arbitrage (locally and cross-chain incl. liquidations) for stablecoins, they’re likely more robust to potential long attacks.

• Freezing supply for smaller markets through conservative parameter recommendations, and for assets that generate revenue allow for a level of organic growth. This seems to make the most sense as it takes a conservative risk approach but still ensures we aren’t cutting off important revenue generators.

One question I have is surrounding how we determine a ‘small’ market and for revenue generation, would we take into account cross-market borrower activity and not solely supply/borrow apy? E.g. users who deposit WBTC and borrow stablecoins.

Thanks!

[ChaosLabs]

Thanks @Pauljlei for this proposal.

Chaos Labs has published an updated methodology for supply and borrow caps, which we used to provide an alternative option we wish to incorporate in the coming Snapshot vote.

Before presenting our recommendations, we would like to address a few general points regarding Gauntlet’s proposal:

1. We do not support the aggressive recommendations option presented.

@Pauljlei

Rely on the assumption that liquidators are able and willing to use cross-chain liquidity. The protocol faces insolvency risk if this assumption does not hold.

As long as no data to support the assumption above is provided, the community should avoid choosing parameters that rely on it.

2. As a general guideline, we view supply and borrow caps as dynamic parameters that should be constantly monitored and updated, given changes in market conditions and utilization of the caps. Therefore, we suggest limiting the increase of caps in a single proposal to double the amount of the current caps. This will allow us to analyze and accommodate borrower and market behaviors.

   • If Gauntlet and the community agree on this guideline, we suggest updating the “Aggressive” option recommendations accordingly.

3. We support the recommendations for the supply caps on stablecoins and will not provide a separate one.

4. We are working on a separate analysis for sAVAX outside of this proposal, which we plan to share soon. We would appreciate taking it out of this proposal and having a separate forum discussion when the analysis is ready.

Chaos Labs Recommendations

In short, our methodology assumes that a given token’s new supply and borrow positions will be distributed similarly to current supply and borrowing activity. We analyze the ability to liquidate current positions given an extreme price drop and then calculate an increased cap that will still allow for healthy liquidations.

Below are our recommendations, calculated using our methodology. As seen in the table, our recommendations allow for controlled growth both in supply and borrow across all assets (besides AAVE token on Optimism)

• Our recommendation for the AAVE token on Optimism is derived from the analysis provided here.
• For WBTC on Optimism, the recommended parameters are very close to the current caps and we therefore recommend not updating them.

Optimism

Avalanche

[WintermuteGovernance]

Thanks for the post @ChaosLabs!

Just one question:

2. We support the recommendations for the supply caps on stablecoins and will not provide a separate one.

Is this under conservative caps?

I like the idea of supply caps that are derived from ensuring liquidations are profitable with a high probability. But I see the methodology only takes into account downward price movements and not upward price movements, does this not expose the protocol to potential long attacks?

Thanks!

[Pauljlei]

Thanks for your feedback, @ChaosLabs. We’d like to address several points below:

@ChaosLabs

As long as no data to support the assumption above is provided, the community should avoid choosing parameters that rely on it.

We agree. As we mentioned above, for the aggressive option, the community would be betting on growth with the tradeoffs of exposure to the risk vectors described in our post above.

@ChaosLabs

We are working on a separate analysis for sAVAX outside of this proposal, which we plan to share soon. We would appreciate taking it out of this proposal and having a separate forum discussion when the analysis is ready.

We are happy to take sAVAX out of this proposal so that the Chaos team can have more time for analysis, because the current supply of sAVAX is close to the existing supply cap.

Differences in Gauntlet and Chaos Methodologies:

Chaos’s recommendations above are largely less conservative (e.g., higher supply cap recommendations) than Gauntlet’s conservative option. To avoid community confusion, we provide a summary of the differences below.

@ChaosLabs

In short, our methodology assumes that a given token’s new supply and borrow positions will be distributed similarly to current supply and borrowing activity. We analyze the ability to liquidate current positions given an extreme price drop and then calculate an increased cap that will still allow for healthy liquidations.

Below are our recommendations, calculated using our methodology. As seen in the table, our recommendations allow for controlled growth both in supply and borrow across all assets (besides AAVE token on Optimism)

From what we understand, Chaos’s methodology:

1. Analyzes the ability to liquidate current positions given an extreme price drop
2. Assumes that new supply and borrow positions will be distributed similarly to current supply and borrow activity
3. Chaos’s methodology for supply and borrow caps depends on other protocol parameters, including liquidation bonus and Interest Rate Curves

Gauntlet’s methodology differs in several ways.

1. Our methodology takes a holistic, conservative approach instead of focusing only on price drops. As we outline in our methodology, there are numerous other market risks, including Long Attacks, Short Attacks, and Liquidity Concentration Risks that our methodology considers.
2. Gauntlet’s methodology does not assume that new supply and borrow positions follow the current market distribution. This is a deliberate choice. Importantly, there are many major risk vectors that are risk vectors precisely because they do not follow the current market distribution. For example, outsized long positions, outsized short positions, or economic attacks. These positions do not follow current borrow/supply distributions. As such, making the assumption that borrow/supply distribution follows existing market distribution underestimates market risk and does not account for these risky types of user position scenarios.
3. Our methodology for supply and borrow caps is upstream of setting other parameters like liquidation bonus and IR Curves. The logic is as follows: we first set bounds on the size of the markets to protect broadly against the risks outlined above. After we set bounds on the market size, we ingest data on actual protocol usage into our risk simulations to determine other parameters like LT, liquidation bonus, etc. In contrast, if recommendations on caps were calculated depending on liquidation bonus etc., then this would result in circular logic that leads to too many degrees of freedom (e.g., the caps depend on parameters that depend on the size of the market that depends on caps).

To summarize: Gauntlet’s borrow and supply cap methodology are made without assumptions of expectation and labeling of price drops (i.e., what an extreme price drop will look like given a market cap). In addition, we are agnostic of positions on the protocol. This is because our liquidation threshold, LTV, and Liquidation bonus recs utilize simulation and current positions to analyze these critical price points for a protocol. We want our analysis of supply and borrow caps to be easily applicable to all chains and liquidity conditions.

Analyzing the borrow and supply caps as a function of liquidation bonus, LT assumptions, and IR potentially leads to parameter suggestions that end up with too many degrees of freedom. I.e., the logic could become circular and convoluted.

Next Steps

Adding an additional set of options to the community is skipping the most important and fundamental step of this proposal - which is for the community to align on assumptions and strategy in order to have a principled approach to risk management. Therefore, we propose the below:

• We will revise the Snapshot vote to be the below options instead:
  • Conservative Assumption
  • Aggressive Assumption
  • Conservative Assumption only for smaller markets (while allowing majors room to grow)
  • Abstain
  • Nay

The assumptions depend on the following question: should the community consider overall liquidity (aggressive) or only consider native-chain liquidity (conservative)? This question depends on the community’s preference: does the community want the Aave protocols on Avalanche and Optimism to be safe under current liquidity conditions (in which case, Conservative is the choice)? OR, do we want these protocols to grow to increase usage and over time hope that liquidity conditions improve as a result of Aave’s presence on these chains?

The strategy and preference of the community greatly impact the actual parameter recommendations.

Whatever option the community decides on, Gauntlet and Chaos can then follow up with specific recommendations incorporating that assumption. We are happy to collaborate with Chaos here. We hope that the numbers in the original post provide some context on what the parameter recommendations would look like under conservative vs. aggressive assumptions.

@WintermuteGovernance - For Option 3, Chaos and Gauntlet may come up with different definitions as to how this exactly will look like. But a preview of what this could look like is that for the top [ X ] assets (defined by metrics like revenue generation, market cap, etc), the cap would allow room for organic growth. Before any on-chain vote, we are happy to align with the community via Snapshot. But first, the community would align on broad preference so that our recommendations can be relevant to the community’s risk/strategy preference.

To give more time to the community to review the above, we plan on putting up this revised Snapshot on Tuesday, February 14th. To clarify - this Snapshot would be a temperature check to help the community align on assumptions (which are strategic in nature), and then Gauntlet and Chaos can follow up with specific recommendations.

[WintermuteGovernance]

Thanks Paul,

To make voting simpler from our end, we prefer the recommendations to be based on native-chain liquidity including stablecoins (conservative).

However, we would still like to see option 3 which allows for relaxed parameters in revenue-generating markets.

[ChaosLabs]

We support Gauntlet’s proposal to postpone the vote on the supply and borrow cap amendments to understand the community’s stance on the assumptions and preferences. We want to emphasize a few points regarding our approach:

1. One of the main differences between Gauntlet’s and Chaos’ approach is that we assume that new supply and borrow positions follow the current market distribution position distribution while Gauntlet does not. It is important to emphasize that as this is an assumption in our current methodology, we limit the increase of the supply cap in each proposal to 2X the current supply to confirm our assumptions of user behavior while limiting the exposure.

2. This assumption is made for each asset individually because ignoring it would require a generalized assumption on asset usage, which could be overly conservative (for instance, assuming that 100% of the supply of assets serves as collateral).

3. @WintermuteGovernance We consider other risks as constraints, such as price manipulation attacks, utilizing our asset protection tool. When making our recommendations, we view all risk vectors holistically and will update the community if we need to deviate from our public methodology.

[eboado]

Really nice to see the risk providers of the community providing valuable data points and high/low-level arguments regarding so an important topic as the protocol caps.

From my perspective, a fully conservative approach is not really good for the growth of the ecosystem. Yes, it would allow almost perfect protection, but the protocol should be focused more on 1) accepting slightly more risk to support growth and users holding a different types of assets, always with limits obviously, 2) speed on adjustment of the parameters when required, by potentially introducing even some kind of automation based on some algorithmic considerations (maybe on-chain even).

Especially important at the moment with the growth of Liquid Staking Tokens is to really adapt the risk considerations to the nature of each one of them. For example, it is not the same as an LST on which the secondary market plays (for now) a huge role like stETH, compared with others like sAVAX on which basically there is perfect redeemability, only boiling down to security risk.

Additionally, something that I think should be avoided as a big priority is to create blockers of liquidity due to caps, as we are seeing on Polygon, with liquidity having difficulties migrating from v2 to v3. Apart from causing extremely bad user experience, those kind of blockers are basically like shooting ourselves in the foot, given that I think everybody agrees from all types of perspectives that 1) liquidity on Aave ideally should always grow over time 2) respecting 1 (or even not always), it is way preferable to have as many migrations to v3 as possible.

P.S. pricing of assets on non-Ethereum networks based on Chainlink is not using secondary market liquidity on the specific chain, but “aggregating” prices on different venues (centralized and decentralized) and assumes perfect communication, even if that is not precise.
Risk parameters should use this both on the “plus” side (without it, will be simply reckless to list the big majority of assets) and the “minus” (there is bridging time risk).

[Pauljlei]

Thanks, @eboado @WintermuteGovernance, for your feedback on this important topic.

We have published the Snapshot vote here. Voting begins tomorrow (Wednesday, 2/15/2023) and will last until Tuesday, 2/21/2023.

The Snapshot has the below options:

1. Conservative Assumption

2. Aggressive Assumption

3. Conservative Assumption only for smaller markets (while allowing majors and strategic assets room to grow)

4. Abstain

5. Nay

Option 3 will utilize different assumptions for different assets - it is ultimately up to the community to decide the risk preference of the protocol. For example, more aggressive assumptions may potentially be used for LSTs on their native chain, as @eboado outlines above.

Following this Snapshot vote, we will present specific recommendations to the community before putting it up for vote.

[fig]

Hi guys - thanks for putting this forward.

A good way to greater align the community and risk managers.

We are leaning towards “conservative only for small markets” and “aggressive” to encourage more growth and reclaim market share as users renter the market & incumbents grow in size.

What does the “Nay” vote signify? Notice some folks leaning this way.

[Michigan_Blockchain]

Thanks @Pauljlei and @ChaosLabs for putting forward this important proposal.

@Michigan_Blockchain is leaning towards option 3: Conservative Assumption only for smaller markets (while allowing majors and strategic assets room to grow). We do have some questions / concerns:

“One major consideration in our conservative recommendations is the frequency of liquidators using local-chain liquidity for liquidations. For Avalanche, in the last 6 months, 69.4% of liquidations involved a swap using on-chain liquidity. For Optimism, over the previous 6 months, 96.2% of liquidations involved a swap using on-chain liquidity.”

1. Does it mean 69.4% of liquidations involved a swap using local-chain liquidity? Or is it on-chain without decomposition into local-chain and cross-chain?

2. If there’s data on this, have you observed any correlation between the overall size of of an asset and the percent of liquidations involving a swap using local-chain liquidity? That is, are big assets like ETH more likely to see cross-chain liquidity used for liquidations? Or is there no pattern at all?

3. We feel the community is being asked to choose an assumption without given enough information on the consequences of the choice, especially if our choice ends up being wrong. For example, if we choose the aggressive assumption, then what is the potential magnitude of losses if liquidators actually do NOT use cross-chain liquidity? Would it be possible to use a specific asset as example, and estimate the amount of insolvency or VaR given a certain shock / attack when we make the wrong assumption?

We are leaning towards option 3 as a safe middle ground. The Conservative Assumption seems likely to stunt growth. The Aggressive Assumption might lead to large losses if it turns out to be wrong, although we don’t actually have a sense of how large the losses could be.

[jmo]

1. Local chain
2. We investigated what percentage of liquidations would fail without local chain liquidity. This was not broken down by asset. Unless you have reason to believe that the more popular assets are safer, I would go with the conservative option here.
3. The potential magnitude of losses would be roughly commensurate to the total borrow in each deployment of the protocol. It depends how much the collateral assets drop in price. Let’s use a very simplified example Right now there is $100mm in BTC.b collateral, with 75% liquidation threshold, leaving 25% of buffer for Aave. An insolvency could occur if BTC drops in price by more than 25%, as the limited on-chain liquidity could result in liquidations failing and Aave having to hold the BTC collateral. Let’s say BTC drops 40% quickly. Liquidations would very likely work in the mainnet Ethereum market but it’s hard to be so certain on Avalanche. This would result in Aave having $60mm in collateral against up to $75mm in liabilities, so a shortfall of up to $15mm.

Again regarding the third question - these sort of price events are rare. As you rightfully observe, the tradeoff here is complex as is the uncertainty. There is a small chance of losing a lot of money, versus a high chance of limiting growth in these deployments. This is the sort of thing that Gauntlet focuses on, but understand why some participants might find this decision difficult. It might be fine to just defer to the risk managers and other stakeholders here. If you look at our delegate platform for Maker, we call out that we don’t participate in every vote. I think some specialization is probably a normal part of governance participation.

[Pauljlei]

TLDR: Please see our Specification Summary below for a synopsis of recommended changes.

Motivation

Following the Snapshot vote, we are updating and providing more clarity on Gauntlet’s borrow and supply cap recommendations for Aave V3 markets on Avalanche and Optimism chains.

Our initial recommendation for the same markets can be found here and a brief description of the methodology used can be found here.

The above methodology has been fine-tuned based on the feedback we received from the community in our previous post.

More specifically:

• For small or “non-strategic” assets, we set conservative borrow and supply caps.
• For large or “strategic” assets, we set aggressive borrow and supply caps.
• For assets listed on V2 Avalanche, we aim to accomodate the migration from V2 to V3. Generally, we considered it safer for the protocol to deviate from our initial recommendations to allow migration to V3 to continue rather than leave assets on Aave V2.
• For liquid staked tokens (LSTs) that are easily redeemable (i.e., sAVAX), we readjusted borrow and supply caps upwards by considering liquidity of the underlying token.
• For very aggressive supply cap increases for assets that were near the supply cap, we limit the change to 150% of increase relative to current supply cap. We may follow up with additional recommendations for further supply increases in the future.
• For supply cap decreases where the recommended values are below current supply, we limit the change to a 75% decrease relative to current supply. We may follow up with additional recommendations for further supply cap decreases in the future.
• Our approach is to update borrow cap and supply cap recommendations once borrows or supplies reach 75% of their respective caps or liquidity conditions change for each respective token.
• Further, we provide bridge information for assets on V3 Avalanche and V3 Optimism. We hope this information around bridges provides greater transparency to the community for their decision-making.

Updated Parameter Recommendations

Bridged and liquid staking tokens

The majority of bridged tokens on the Avalanche chain used on Aave V3 (WETH.e, AAVE.e, LINK.e, WBTC.e, DAI.e, BTC.b) are issued by the Avalanche bridge and proof of reserves are provided on their website. FRAX is issued by frax.finance, MAI is issued by mai.finance, USDT is issued by Tether, and USDC is issued by Circle.

sAVAX is the only liquid staking token provided by Aave V3 on Avalanche, and the staking services are provided by BENQi.

For Optimism, bridging from Ethereum happens through the official bridge for all Aave V3 assets except for sUSD where the Synthetix bridge can be used. Bridging back to Ethereum through the native bridge introduces a 7-day delay, which can force users to use 3rd party bridges.

Chain	Asset	Issuing Entity	Official Issuance/Bridging	Estimated Days to Bridge in	Estimated Days to Bridge out
Avalanche	DAI.e	Avalanche	Avalanche Bridge | Core	<1	<1
Avalanche	FRAX	Frax Finance	https://app.frax.finance	<1	<1
Avalanche	MAI	MAI Finance	https://www.mai.finance/	<1	<1
Avalanche	USDC	Circle	Circle | USDC Stablecoin, Payments & Treasury Infrastructure	KYC restrictions/TradFi Speeds	<1
Avalanche	USDt	Tether	https://tether.to/en/	KYC restrictions/TradFi Speeds	<1
Avalanche	Aave.e	Avalanche	Avalanche Bridge | Core	<1	<1
Avalanche	BTC.b	Avalanche	Avalanche Bridge | Core	<1	<1
Avalanche	LINK.e	Avalanche	Avalanche Bridge | Core	<1	<1
Avalanche	sAVAX	Avalanche	Avalanche Bridge | Core	<1	<1
Avalanche	WBTC.e	Avalanche	Avalanche Bridge | Core	<1	<1
Avalanche	WETH.e	Avalanche	Avalanche Bridge | Core	<1	<1
Optimism	DAI	Optimism	Optimism Gateway	<1	7
Optimism	sUSD	Synthetix.io	Synthetix Staking	<1	<1
Optimism	USDC	Optimism	Optimism Gateway	<1	7
Optimism	USDT	Optimism	Optimism Gateway	<1	7
Optimism	AAVE	Optimism	Optimism Gateway	<1	7
Optimism	LINK	Optimism	Optimism Gateway	<1	7
Optimism	WBTC	Optimism	Optimism Gateway	<1	7
Optimism	ETH	Optimism	Optimism Gateway	<1	7
Optimism	WSTETH	Optimism	Optimism Gateway	<1	7

Specification

We provide a tailored set of recommendations below, aiming to account for the community’s strategic preferences while limiting governance fatigue resulting from other ongoing parameter proposals. We hope that the below provides transparency into the risk analysis rationale.

As a result of our asset-by-asset analysis, we recommend proactively freezing new supply for several assets to mitigate outsized tail risk.

V3 AVAX:

• Stablecoins: the majority of Gauntlet’s recommendations are more aggressive than other recommendations. As such, we will not move forward with our stablecoin recommendations.
• AAVE.e: our analysis recommends an increase in the supply cap to allow for migration, but we will not move forward with this recommendation because there is an existing proposal to increase it (that is more conservative than our recommendation).
• BTC.b: we recommend decreasing the cap to below the current supply, which would effectively freeze new supply of BTC.b. We also recommend lowering the borrow cap. BTC.b is an asset that can pose potential risk given the size of the Aave market relative to circulating supply (over 50%).
• LINK: our analysis shows that the supply cap can be higher, but we will not move forward with this recommendation because ongoing proposals are increasing it. For the borrow cap - our analysis shows that the borrow cap can be higher, but because there is still room for borrow growth under the current cap, we will not move forward with this parameter change.
• sAVAX: our conservative figure is lower than other ongoing recommendations. However, if the community has an Aggressive risk preference, the ongoing recommendations may be preferred. Thus, we will not move forward with our recommendation.
• AVAX: we recommend decreasing the supply cap and implementing a borrow cap.
• WBTC.e: we recommend decreasing the supply cap and implementing a borrow cap.
• WETH.e: we recommend decreasing the supply cap and decreasing the borrow cap.

V3 Optimism

• Stablecoins: the majority of Gauntlet’s recommendations are more aggressive than ongoing recommendations. As such, we will not move forward with our stablecoin recommendations.
• AAVE: we recommend decreasing the cap to below the current supply, which would effectively freeze new supply of AAVE. AAVE is an asset that can pose potential risk given the size of the market relative to the circulating supply.
• LINK: we recommend lowering the supply and borrow caps.
• WBTC: current supply is greater than even our Aggressive recommendation (624). As such, we recommend lowering the supply cap, which would effectively freeze new supply of WBTC and also lower the borrow cap.
• ETH: our recommendations are more aggressive than the current caps. Because the current caps still allow room for growth, we will not move forward with these recommendations.

Specification Summary

For better user experience, we round the recommendations to an easily interpretable number.

Asset parameter current parameter recommended parameter notes

V3 AVAX

Asset	Parameter	Current Parameter	Recommended Parameter	Notes
BTC.b	Supply Cap	5,800	3,000	Effectively freeze supply
BTC.b	Borrow Cap	3,190	900	—
AVAX	Supply Cap	13,100,000	3,800,000	—
AVAX	Borrow Cap	NA	1,200,000	At this cap, utilization would be 80%
WBTC.e	Supply Cap	5,233	2,000	—
WBTC.e	Borrow Cap	NA	1,100	—
WETH.e	Supply Cap	113,000	38,000	—
WETH.e	Borrow Cap	62,150	33,000	—

V3 OP

Asset	Parameter	Current Parameter	Recommended Parameter	Notes
AAVE	Supply Cap	100,000	45,000	Effectively freeze supply
LINK	Supply Cap	258,000	160,000	—
LINK	Borrow Cap	141,900	84,000	—
WBTC	Supply Cap	1,100	620	Effectively freeze supply
WBTC	Borrow Cap	605	250	—

^{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.}

[Pauljlei]

For clarity, Chaos has also provided recommendations here. We are happy to move forward with 2 options to the community (in which case, we can provide a summary of differences so the community can choose), or are open to other suggestions from @ChaosLabs or the community on how to move forward swiftly.