[ARFC] Adjust Risk Parameters for Aave V2 and V3 on Polygon

[RocZacharias]

Roc Zacharias here, QuickSwap Co-Founder, Polygon CTB (Community Treasury Board) board member, and long time Aave user, holder, and fan. Good to see your comment Emilio, long time no see my friend :)

I really appreciate your thoughtful and calm headed take on all of this. I am in the same boat as you in that I have such a strong love for Aave, it’s leadership, and our amazing community. Our communities have been instrumental in the growth of Ethereum and it’s scaling ecosystem.

I understand people’s concerns with the bridge pre-PIP, and I am not a fan of the Pre-PIP either. I don’t think it’s the craziest idea, and if implemented a different way, for example only using Eth to validate Ethereum, or having insurance fund, or insurance by Polygon treasury or other creative ways, it could have been more interesting. With that being said, the idea has been squashed, and I highly doubt it will be brought up again. The community on both sides voiced loudly that it just isn’t something we want which is a sign of good governance on both sides.

Keep in mind, a Pre-PIP in Polygon is more similar to a Temp Check on Aave. This wasn’t an actual PIP, never got to that stage, and never will. To me, this should be basically a finished issue. Aave Polygon has significant revenue, history, and huge user-base. Polygon Aave deployment is one of the largest by users as Emilio mentioned above. It just doesn’t make sense for the communities to “break up” over a simple discussion. If anyone disagrees, could you please explain why we would “break up” and hurt all of the users who have actively voted with their wallets to use Aave on Polygon? Many of these users are exclusively Aave on Polygon, or started their journey on Polygon, and have since expanded to more chains after first trying Aave on Polygon.

If you look at this data from Dune, Polygon Aave users are continuing to grow at a strong pace. https://dune.com/queries/4573060/7624783

Here is another interesting data point. Polygon is 2nd in cumulative fees behind only Ethereum, having 100x more than Base and BSC, 4x more than Arbitrum. I love Base and Arbitrum, no competition here, just giving stats for everyone to see. https://defillama.com/protocol/aave#fees-revenue*

I have a lot of respect for everything Aave has accomplished, and many of the incredible milestones happened with our communities working together! We literally made history TOGETHER!

Polygon is fighting hard to unite all Ethereum L2s, and will continue to be a major contributor to the future of scaling Ethereum through rollups, and I know damn well Aave will continue to innovate and provide finance to the underbanked of the world. We have bigger battles with external enemies like banks, governments, the incumbent financial industry, and overly zealous regulators who would like to see our whole industry burn a fiery death. I know strong words were had on both sides. Words among a handful of individuals don’t change our history or our mission. I hope we can come back together and keep fighting the good-fight with our communities united!

With so much love,
Roc

[ACI]

The current proposal has been escalated to ARFC Snapshot.

Vote will start tomorrow, we encourage everyone to participate.

[EzR3aL]

All active loans won’t be affected.
What you cannot do will be to take out another loan with those assets.

[MarcZeller]

the proposal was edited following community & service providers feedback.

Aave stays on Polygon, users are mostly unaffected, Aave DAO only reduce risk exposure to this network.

[MarcZeller]

that is correct. please refer to the proposal first post that has been updated or go to discord if you have more question.

[Michigan_Blockchain]

As Michigan Blockchain, we agree that cross-chain bridge vulnerabilities have been a significant security concern in the DeFi space. According to Chainalysis, approximately $2 billion in crypto was stolen across 13 cross-chain bridge hacks as of August 2022. This accounted for 69% of funds stolen that year. It is important that Aave proactively reduces risk and ensures user safety as Polygon governance is evaluating changes to its bridged asset framework. Setting the LTVs to 0% and increasing Reserve Factors will discourage new borrowing and minimize exposure to potentially unsafe assets. This will also strengthen Aave’s long term stability. We also support the user migration initiative, which will help Aave users move their positions to safer networks with incentives and support.

As Michigan Blockchain, we support adjusting the risk parameters for Aave V2 and V3 on Polygon, specifically by setting the LTV for key stablecoins—USDT, USDC, USDC.e, and DAI—to 0%. This change would prevent these assets from being used as collateral for borrowing. This is a precaution we believe is necessary given the unique risks associated with bridged assets and the deployment of large capital amounts into yield-bearing DeFi protocols.

Many stablecoins on Polygon are bridged from other chains such as Ethereum, meaning their 1:1 backing depends on assets held externally. In December 2024, a proposal was put forth on the Polygon forum suggesting that approximately $1.3 billion in stablecoins (DAI, USDC, and USDT) currently idle on the Polygon PoS Bridge be deployed into yield-bearing liquidity pools via ERC‑4626 vaults. While deploying this idle capital can unlock significant yield and bolster ecosystem growth, it also introduces additional risks. Idle funds in the bridge are typically held in straightforward contracts, limiting exposure to the complex vulnerabilities inherent in active DeFi protocols. Once deployed, however, these assets interact with more intricate smart contracts, liquidity pools, and yield strategies, thereby increasing their exposure to risks such as smart contract bugs, liquidity crunches, and potential exploits.

A critical concern is that a hack or exploit at the bridge could result in the loss or manipulation of the assets backing these stablecoins. If such an event were to occur, the stablecoins could lose their 1:1 peg—a depeg scenario that would have devastating consequences for the Aave markets on Polygon. Given that a substantial portion of the stablecoins on Aave V3 is used as collateral (with figures like $63 million USDT supplied, $55 million USDC supplied, $10 million USDC.e, and $9 million DAI supplied), any significant devaluation could trigger massive liquidations and widespread financial losses for users.

It is important to note that while idle capital on the bridge might seem “safe” due to its minimal exposure to complex DeFi interactions, it also comes with the drawback of opportunity cost—it is not being productively deployed to generate yield or support liquidity. However, deploying the $1.3 billion into yield-bearing strategies increases its exposure to market dynamics and interconnected protocol risks. In such a scenario, if a vulnerability in any layer—be it the bridge or the underlying DeFi application—were exploited, the impact would be amplified by the large capital involved, further threatening the peg of the stablecoins and the stability of the Aave markets.

In conclusion, by adjusting the risk parameters and setting the LTV of these bridged stablecoins to 0%, Aave should mitigate the compounded risks arising from both potential bridge exploits and the increased exposure due to active deployment of large capital in yield-generating protocols. This proactive measure is essential to protect users from cascading liquidation events and to maintain the overall health and stability of the Aave V2 and V3 markets on Polygon.

-Kerem Dillice and nsks

[ACI]

After Snapshot monitoring, the current ARFC Snapshot ended recently, reaching out both Quorum and YAE as winning option, with 590K votes.

Therefore, [ARFC] Adjust Risk Parameters for Aave V2 and V3 on Polygon has PASSED.

Next step will be the publication of an AIP for final confirmation and enforcement of the proposal.

[eboado]

Some feedback on the risk proposal on my side:

• In its current shape, it merges different changes that are difficult to vote on. For example, I’m against some of the LTV0, but supportive of other modifications like RF.
  This is really far from ideal because voting becomes some type of weighted decision: if a voter agrees with 80%, but not with 20%, the policy can be “Well, I vote FOR because generally I agree more than disagree”, or “I simply don’t want that 20% to be applied, no matter the positives, so AGAINST”, or “I can’t make a decision, so I don’t vote”.
  This decision-making becomes even more relevant to highly critical voters, so it is very negative.

• In terms of my arguments FOR/AGAINST by components:

  • RF increases. Kind of agree with all, but for different reasons. On “legacy” assets without native minting alternative (USDC.e), RF should just be maximized because protocol should just promote its offboarding, same as in all networks.
    On “healthy” stablecoins that don’t require off-boarding, RF becomes a monetization lever for a network that definitely has created operational overhead (this whole discussion). So definitely the vote on it is legitimate, and I personally support the changes, as they are not even so radical (20 and 25%).
  • LTV0. AGAINST on “healthy” stablecoins (USDC and USDT (PoS)), FOR on “legacy” ones (DAI, USDC.e), and generally, pretty misdirected.
    On USDC the argument is clear: it is natively minting, and there is no reason to completely flag it as not suitable as collateral because it is objectively fine.
    On USDT (PoS), I probably agree that being a bridged asset the risk profile is slightly worse than before, but LTV0 doesn’t make much sense, because it is not realistic for Polygon to do radical movement of its backing, considering how the whole community backslashed the previous.
  • Missing additions. There are aspects that to be consistent should have been proposed instead of others:
    • Reducing LT (hard) of stablecoins eMode on v3. This matters, because there is no rationale to keep so high params on an eMode which factually has no good usage associated, only negative for Aave when a stablecoin depegs.
      It is true that LTV0 indirectly helps, but it is simply the wrong tool to use.
    • Reduce the LT of all stablecoins across the board. If the risk profile is lower and usage as collateral is not to be promoted, the solution is to progressively reduce LT.
    • Reduce LT on v2 assets. LTV0 on v2 is just a very soft measure on borrow openings, not as effective as v3 (there, it removes factually collateral power). So arguably that change has zero effect almost, and what should have been done is just to reduce LT progressively.

• Generally, and even if knowing that the usage of stablecoins as collateral is not relevant on Polygon, I don’t particularly agree with cutting a use case (shorting “volatile” asset) for little reason, caused by the lack of trust in the team directing the chain.
  As Aave, we should not care too much about that, and almost exclusively about users. At least until there are some warning symptoms, and then the off-boarding should just be complete: all assets, and incentives in another network interested in supporting.

My recommendation is or 1) a follow-up proposal even if not really optimal or 2) re-creation.

[fulltawch]

Seeing as the pre-PIP by Polygon didn’t get anywhere, are these measures being considered not too drastic? Moving away from Polygon over a pre-PIP that had no steam in the first place anyways sounds more punitive than preemptive

[Terminal]

The CEO of Polygon Labs publicly claimed AAVE is an inferior protocol full of bugs, without providing any technical evidence, undermining the hard work of all of AAVE’s technical teams.

If Polygon wants this actions reverted, please start by providing specifically in code what are this bugs you are referring.

Even if this statement was true, the correct and professional move of a CEO would be to point them out and try to fix them with the team, considering how big AAVE is in their L2. This alone pin point very weak and unprofessional leadership.

[MarcZeller]

I echo part of this feedback, we suggest a follow-up proposal after this one to adjust some things.

LTV0 on USDC and then USDT0 when implemented is not justified we should offer a lower LT on these assets than other networks to promote alternative with without the need for a LTV0, we stand firm on other stables

LT reduction on stablecoin emode (more or less deprecation) should be added.

[cryptonft]

As I understand it the proposal is not happening on Polygon, and was voted down, specifically in response to justified AAVE concerns about it. My understanding in polygon, is that it simply will not happen period.

I can understand concerns and conflict, but would hope AAVE would continue on Polygon. Continued partnership is mutually beneficial for both. Both protocols and teams have been around a long time, and have done a lot for crypto in general.

If there is some signaling desired, perhaps a reduction in % instead of going to zero. I would call for fellow AAVE stakers to vote to continue relationship with Polygon, and have a modified proposal if signaling is desired.

There are so many outright scams and failures in crypto, and this is a long term relationship, between two excellent stable projects, and short term hiccups and clashes are to be expected.

Voting no on this proposal, reconsidering, perhaps reproposing with different weights, would be beneficial for AAVE medium to long term.

[EzR3aL]

I will support the vote in its current form but also support the change on stablecoins (see comment from eboado) and will vote accordingly.

[MarcZeller]

---

Title: [ARFC-ADDENDUM] Adjust Risk Parameters for Aave V3 on Polygon

Date: 2025/02/28

Author: Marc Zeller - @ACI

---

Summary

This proposal addendum seeks governance approval for adjustment of the AIP 254 to reintroduce LTV for non-bridged stablecoins in Aave Polygon V3.

Motivation

Based on service provider and community feedback, this proposal aims to adjust AIP-254 to reintroduce LTV for non-bridged stablecoins in Aave Polygon V3, as they carry no rehypothecation risk. It also reduces stablecoin Emode LT in Aave Polygon V3 due to insufficient stablecoin/stablecoin trading volume to justify the current risk level.

USDT is scheduled to migrate to the USDT0 standard soon, as recently implemented on networks like Arbitrum. To promote stablecoin diversity and taking an optimistic view on the reduction of bridge risk for this asset, this proposal aims to restore an LTV for USDT.

Specification

Risk Parameters have been updated 2025-03-03 due to Chaos feedback.

Deployment	Asset	Current LTV	Proposed LTV
Polygon V3	USDT	0%	70%
Polygon V3	USDC	0%	70%

• Make all stablecoins in stablecoins emode as non borrowable

We recommend the following LT and LTV for the Stablecoin E-Mode on Polygon Instance.

Current LTV	Proposed LTV	Current LT	Proposed LT
93%	91.25%	95%	94.25%

Next Steps

1. Collect feedback from risk service providers for this addendum
2. If risk service providers confirm the suggested parameters, escalate to AIP

Disclaimer

The ACI has not been compensated for creating this proposal.

Copyright

Copyright and related rights waived via CC0.

[ChaosLabs]

Overview

Chaos Labs supports the proposal to adjust LTV and LT for stablecoins on Polygon. However, we propose slightly different values that will maintain the stated goals of improving the user experience for non-bridged stablecoins while eliminating loopholes that could allow continued borrowing against bridged stablecoins.

Non-E-Mode

Due to technical aspects of Aave’s contracts, users are able to borrow more than permitted by their collateral’s LTV while still being strictly limited by its LT values. This is possible through the use of flash loans when opening a position and is facilitated by third-party Aave UI tools such as DeFi Saver. Hence, setting the LTV of bridged assets to 1% enables their full use within the protocol. However, setting the LTV to 0% mitigates these issues and stops users from opening additional positions while ensuring that the change triggers no liquidation. As such, we recommend maintaining the LTV for DAI and USDC.e to 0%.

We do agree with the proposed LTV values for USDT and USDC, as their native integration ensures that the assets are safe from rehypothecation.

E-Mode

In the Stablecoin E-Mode, we concur with the goal of reducing the leverage possible in this market. However, it is not possible to set LTV to 0%, and as stated above, all non-zero values can be bypassed using flash loans, effectively making the asset’s LT the relevant parameter for reducing possible leverage. Thus, we recommend maintaining a 3 percentage point buffer between LTV and LT in this E-Mode and we recommend setting the LT at a point that minimizes liquidations while also meaningfully reducing the potential leverage. Based on the currently opened positions in the market, we can determine which positions will transition to a liquidation phase following the reduction of the E-Mode LT.

In the chart provided below, we plot the cumulative liquidatable value by LT reduction. Reducing LT to 94.25%, and thus setting LTV to 91.25%, minimizes liquidations while reducing leverage available to new positions.

Based on the current market, this will induce $39.6K worth of liquidations, 47.4% of these involving users leveraging USDC.e.

This LT and LTV reduction within the E-Mode represents the first step of a gradual adjustment to the E-Mode parameters in order to better align them with the intended usage and risk profile of the stablecoin assets listed.

Specifications

We recommend the following LTVs for the currently listed stablecoins within the Polygon Instance.

Deployment	Asset	Current LTV	Proposed LTV
Polygon V3	DAI	0%	0%
Polygon V3	USDC.e	0%	0%
Polygon V3	USDT	0%	70%
Polygon V3	USDC	0%	70%

We recommend the following LT and LTV for the Stablecoin E-Mode on Polygon Instance.

Current LTV	Proposed LTV	Current LT	Proposed LT
93%	91.25%	95%	94.25%

Disclaimer

Chaos Labs has not been compensated by any third party for publishing this recommendation.

Copyright

Copyright and related rights waived via CC0

[LlamaRisk]

Summary

LlamaRisk supports the changes proposed by @ChaosLabs. These modifications will serve as a further step toward mitigating exposure to Polygon’s bridge risk. At the same time, favoring non-rehypothecated assets within Aave’s Polygon market will ensure an enhanced user experience and greater flexibility.

To complete the rebalancing of the market, further changes to the parameters, as suggested by @eboado, will be necessary. The Risk Stewards could also execute this gradual process once the rationale and targets are set and announced in this governance thread.

Changes in Stablecoin Supply

On February 25th, 2025, the AIP setting the LTV of most stablecoins on Aave’s Polygon market was executed. Since then, we have seen outflows amounting to 10–20% of the supply of these stablecoins, suggesting that suppliers have begun to rotate out of the market. While some of these moves could be associated with the overall negative market sentiment, it is evident that the outflows have been significant. Nonetheless, the supply figures for USDC and USDT are expected to rebound if this ARFC Addendum is enacted.

Source: Aave V3 Polygon, March 3rd, 2025

E-Mode

At the time of writing, the loan in the Stablecoins E-Mode totals $12m. The proposed changes to E-Mode’s LT parameter might trigger minor liquidations but are not expected to impact the Polygon market aggressively. Making all stablecoins non-borrowable in this E-Mode would prevent new borrow positions from being opened. Therefore, these changes are anticipated to initiate capital migration from the Stablecoins E-Mode. Further reductions in LT and LTV would be needed to reduce the borrows completely.

Disclaimer

This review was independently prepared by LlamaRisk, a community-led, non-profit decentralized organization funded in part by the Aave DAO. LlamaRisk is not directly affiliated with the protocol(s) reviewed in this assessment and did not receive any compensation from the protocol(s) or their affiliated entities for this work.

The information provided should not be construed as legal, financial, tax, or professional advice.

[web.3]

why was the AIP even executed? Polygon canceled their idea to misuse bridge reserves until further notice, no?

[EzR3aL]

Because the DAO voted accordingly.