Wormhole x Axelar Thoughts
Before the vote for this temperature check ends, LayerZero Labs would like to share thoughts on the Axelar/Wormhole joint proposal in the context of why we believe LayerZero V2 is a better solution.
Our current wstETH proposal can be found here. It outlines how Lido DAO may leverage LayerZero V2’s modular Security Stacks to extend wstETH to BNB Chain using a 2/2 or 3/3 verification setup that includes Polyhedra’s zkLightClient, a Nethermind DVN, and/or a third party bridge (like CCIP, Axelar, or Wormhole).
The argument can be broken down into two basic points:
1. Vendor Lock-In vs. Modular Security
This Axelar and Wormhole implementation risks vendor lock-in, as it is essentially a 2/2 multisig with the option for Lido DAO to “add more Endpoint providers," without instruction on how to do so or a roadmap for other bridges to be included. As shown by other multi-bridge solutions like xERC20, MMA, and Hashi, adding extra verification methods is not an easy task due to audits and development time. LayerZero Labs does not believe that Axelar and Wormhole are incentivized to push for additional bridges to be added here.
In contrast, LayerZero V2’s vendor-agnostic approach to security was designed to enable permissionless incorporation and evolution of verification methods. LayerZero V2 implicitly supports 15+ client-diverse Decentralized Verifier Networks (DVNs) that can be configured together to verify messages, including zkLight clients, native bridges, Ethereum client teams (like Nethermind), community signers (like StableLab and Gitcoin), and third-party bridges (like Axelar and CCIP). V2 also allows Lido DAO to configure its selection of message executors, what chains to deploy to, and how many block confirmations DVNs must wait for to verify a message for wstETH.
It is possible to wrap bridges as “Adapter” DVNs. At launch, Adapters for Axelar and CCIP will be live. If Lido DAO prefers to include Wormhole as a verification option, LayerZero V2 supports a Wormhole Adapter as well and LayerZero Labs will handle audit costs.
In other words, LayerZero V2 allows for Lido DAO to leverage multiple verification methods without being beholden to Axelar and Wormhole to continue working together to build third party Endpoints to tack onto their proposed architecture. Additionally, Lido DAO has the ability to change the verification methods used for wstETH messages at any time.
While biased, this is how Axelar/Wormhole versus LayerZero V2 should be imagined:
-
Axelar/Wormhole: dependency on new, untested code to conjoin two third-party bridges without a plan to expand to new bridges practically enforces vendor lock-in.
-
LayerZero V2: permissionless and configurable verification chosen by application developers ensures modularity.
2. Short-Term Collaboration vs Long-Term Solution
The multi-bridge approach posited by Axelar and Wormhole proposes an entirely new and untested codebase to accommodate a short-term BD win for two third-party bridges. Nearly all bridge hacks of the last 18 months have been a result of the introduction of smart contract upgrades or un-methodically tested (hardened over time x value secured) code. Even assuming positive collaborative intentions, the likelihood of a vulnerability in an ad-hoc implementation should be considered a very real risk. The LayerZero protocol has secured $40B+ over more than 18 months; the protocol was designed for the exact purpose of enabling multiple message verification methodologies and can support Axelar and Wormhole as options today.
Furthermore, the Axelar and Wormhole’s joint effort does not inherently guarantee future cooperation or extensibility to other chains. A few questions to consider here, based on the assumption that Axelar and Wormhole are competitors:
Are Axelar and Wormhole incentivized to…
- work together long-term?
- extend this wrapper to new chains and assets?
- focus on auditing and shipping this product on time?
- add new verification methods to this architecture?
LayerZero Labs believes the answer to each of these questions is no. And if Lido DAO members agree with us, then they should be worried about the long-term health of wstETH should this Axelar and Wormhole proposal be selected.
Conclusion
The Axelar and Wormhole proposal replicates features already present in LayerZero V2 (and other multi-bridge solutions like Hashi, MMA, and xERC20) but on a lesser scale. LayerZero V2’s architecture already encompasses the functionalities proposed by Axelar and Wormhole, rendering their solution somewhat redundant in the face of LayerZero’s more comprehensive, widely-adopted, and audited solution.
Good developer infrastructure should be both evergreen and adaptable and LayerZero V2 is good developer infrastructure. It is an immutable, permissionless, and censorship-resistant protocol built to support wstETH (and any token that uses the OFT standard) over a long time horizon.
For more information on LayerZero V2, please see our V2 Lido DAO proposal, V2 Medium post, V2 whitepaper, V2 documentation, and V2 launch video.