[TEMP CHECK] GHO Cross-Chain Strategy

Governance
22

Thanks @Cross-chain-Alliance, @Jommi and @bgdlabs for your valuable input here. The utilization of native bridges and the Aave Delivery Infrastructure (a.DI) was thoroughly examined during our assessment and research phase, and we would like to share some insights on our recommendation of using CCIP for GHO.

Given the GHO token contract design and the proposed cross-chain strategy, which involves all new issuance exclusively on Ethereum, the bridging of GHO must adopt a hybrid model. This entails a burn-and-mint approach on remote networks and a lock-and-release mechanism on Ethereum. Although the GHO token contract facilitates direct calls for mint-and-burn, the lock-and-release functionality lacks support within any GHO contract. Consequently, a new suite of smart contracts, operating as GHO Facilitators, is imperative to uphold this functionality. Additionally, these contracts can integrate various risk and security levers for the DAO’s governance.

Historically, the Aave DAO has favored the use of native bridges, leveraging their inherent security. While this may seem an appropriate choice initially, it does entail drawbacks in terms of latency and technical complexity that need careful consideration for this particular use case, where usability is paramount. In the context of L2 solutions, employing native bridges for transfers from L2 to Ethereum is suboptimal, especially considering factors like withdrawal windows (seven-day dispute periods). In the opposite direction, native bridges offer improved latency, but it might not be sufficient considering that fast bridging is essential to stimulate activity and growth.

The a.DI is a robust infrastructure technology, well-suited for governance and deemed acceptable for GHO cross-chain. Its reusable code allows for the deployment and configuration of a new infrastructure instance tailored to the requirements of each network and lane, aligning with the strategy proposed by @bgdlabs. However, this infrastructure primarily addresses the bridging aspect, necessitating a concurrent set of contracts to support the lock-and-release model for GHO.

Utilizing CCIP token transfers offers a significant advantage in terms of technical simplicity, requiring minimal to zero additional effort. The contracts provided by CCIP encompass all the technical necessities for GHO cross-chain operations. The TokenPool contract, integral to this infrastructure, supports the lock-and-release functionality required for Ethereum and the mint-and-burn mechanism for remote networks. Additionally, CCIP introduces risk mitigation, security and UX levers that empower the DAO and improve usability for cross-chain GHO transfers:

  • Rate Limits: An essential security feature that allows for configurable restrictions on both the quantity of tokens and the value transferred per lane over a period of time, ensuring that the bridge remains protected, as well as allowing for the modulation of bridging activity that can be managed by the DAO.
  • Built-in Relayers: CCIP’s decentralized Executing DONs improve the user experience by automatically relaying transactions to destination chains and attempting to re-execute transactions under extreme network conditions (such as spikes in gas prices). Additionally, Chainlink offers a portal through which users can permissionlessly execute the transaction again (overriding the gas limit to use) on the destination chain.
  • Flexible Billing Mechanism: the costs CCIP incurs for a transaction to be executed on the destination chain are covered by users when they make a transaction on the source chain. CCIP reduces friction for users by supporting various tokens to pay this fee, including the native currency of the source chain (e.g. gas tokens) and in LINK. Moreover, discussions are ongoing with the team to support GHO as a fee payment option in the future, which would place GHO as a leader in user experience within the stablecoin domain.
  • Programmable Token Transfers: CCIP token transfers can include additional instructions about their intended use to a receiving smart contract on a different blockchain, such as swapping or staking GHO once tokens arrive at the destination chain. The transfer of tokens and data take place in one atomic cross-chain transaction, meaning the tokens can always be assumed available when the instructions passed are executed at the destination.
  • Highly-Audited Codebase: Both the CCIP onchain and offchain code, including the token pools, have been extensively audited by 14 leading security firms—meaning Aave benefits from the resources that have already been invested in CCIP security.
  • Risk Management Network: There is a separate, independent network that continuously monitors and validates the behavior of the primary CCIP network, providing an additional layer of security by independently verifying cross-chain operations. For token transfers specifically, the network has access to additional transaction information (e.g. token, amount) that enables it to perform additional risk management policies that cannot be implemented for arbitrary messaging.

By adopting CCIP, the DAO can significantly reduce the technical effort required to initiate its cross-chain strategy as soon as possible while also minimizing development and maintenance expenses.These savings are particularly notable when compared to the potential costs associated with creating a full-fledged solution utilizing native bridges or a.DI.

The recommendation to use CCIP as the bridging solution for GHO is based on its capacity to enable a swift implementation of the cross-chain strategy with minimal technical complexity, access to additional risk mitigation tools, and an improved user experience for GHO users. The streamlined approach offered by CCIP, coupled with its inherent advantages, positions the DAO to proactively drive the expansion and growth of GHO into other networks. This choice notably solves key UX challenges, such as those associated with non-fungible token wrapping on different chains due to different combinations of chain-specific bridges.

Crucially, adopting CCIP does not preclude the utilization of other bridges like a.DI. in the future because the Aave DAO retains complete ownership of the token contracts and the token pool contracts. To ensure a secure and seamless expansion of GHO to other networks, CCIP is deemed the optimal choice at this time. A snapshot vote will be initiated shortly to gauge the community sentiment on this matter.

2 Likes
23

We appreciate the valuable insights provided by @ChaosLabs. We encourage Risk providers to actively participate in the ongoing discussions, recognizing the significant impact that this initiative may have on GHO liquidity, volume, and peg across various chains.

Our proposal involves a carefully planned rollout strategy, advancing gradually while meticulously monitoring market data. The initial phase includes expansion to a single remote network to assess performance before contemplating additional networks. Our ongoing research focuses on:

  • The initial step of the rollout plan, encompassing technical deployment and configuration of risk parameters.
  • Identifying preferred networks for GHO expansion based on metrics such as network health, liquidity, security, technical feasibility, and efforts.
  • Assessing the necessity for additional Facilitators, such as GSM or Aave Pools, to bolster liquidity provision on remote networks.
  • Formulating economic incentives to stimulate demand on remote networks, including the potential listing of GHO as a reserve on remote Aave Pools and other integrations.

We extend an invitation to Risk providers and the Aave community to collaborate in our research efforts. Your insights and expertise are pivotal in navigating the complexities of this initiative and ensuring its success.

2 Likes
24

The only concern I have is that it seems like CCIP currently only supports EVM chains.

25

Which is fine because Aave is on EVM chains and only one chain by one will be rolled out. If Aave will be on non-EVM in the future, there are other solutions or CCIP will be there already.

2 Likes
26

Yeah, but thinking about the fierce competition in the stablecoin arena, I think GHO has more opportunities in non-EVM chains like Sui.

2 Likes
28

One cross chain strategy doesn’t mean there can’t be an other strategy next to it. The good think with Aave and DeFi in general is the abilities to be composable.

Yet, non-EVM deployment aren’t impossible but require important investment in terme of dev energy and security (all audit must be remake). So not the first concerne now and if it’s appear to be it will require analysis before engaging a costly development for this DAO.

2 Likes
29

Technologically, it may not even be possible to be completely decentralized to go cross chain to non-EVM chains, so I think the focus should be to compete on the EVM front

30

Deploying GHO cross-chain is critical to its growth, utility, and stability across DeFi. The question up for debate here is not whether Aave should deploy GHO cross-chain; the community is overwhelmingly in favour of doing so. The main question is how to do so - whether to use CCIP or native bridges of the destination chains. We think that the initial design proposed by Aave Labs, after thoroughly reviewing the forum post and ensuing discussions, makes the most sense. We acknowledge the safety benefits that stem from using native bridges, but the benefits of CCIP, especially as a starting point, outweigh these considerations, namely because of the following reasons:

  • Speed and user experience are very important considerations and using native bridges sometimes compromise on these, especially in the destination chain → source chain direction. In the current market conditions, which will likely engender increased usage of GHO quickly across chains, it is important to use the most user-friendly solution.
  • Using Chainlink’s DON network does not add any additional trust assumptions upon Aave - notably, if Chainlink infrastructure fails at some point, a large chunk of DeFi will be affected. Obviously, this reflects a need for DeFi to diversify and therefore diversifying across other oracle networks and messaging infrastructures is important for the space. However, using CCIP does not limit the DAO from using a.DI or native bridges at any point - these integrations can be built in the near future and provide better trust assumptions for cross-chain GHO while initially providing a good starting point for GHO expansion now.
  • The technical complexity of deploying CCIP is lower and can be implemented more quickly with the support of the Chainlink team - this means that the DAO can also choose to implement native bridges in the near future if it chooses to do so.
  • The other benefits of CCIP - risk mitigation tools, flexible billing, programmable token transfers - are also added positives.

Therefore, we will vote YES in favour of this TEMP CHECK and welcome the opportunity to expand GHO cross-chain. We will also be in support of introducing native bridge implementations in the near future, and look forward to discussing the optimal chains to deploy on first.

2 Likes
32

This topic was automatically closed 30 days after the last reply. New replies are no longer allowed.