[AIP-124] Deprecate alUSD and alETH Bridging + Wind Down v3 on Arbitrum

Description

# Background . Now that the Alchemix team has had the chance to monitor v3 in production for multiple months, we can make reasonable conclusions about the system and apply what we’ve learned to help better service our users and goals. Among these learnings: . **1. Bridging is one of the highest risk activities in DeFi.** We saw this with the rsETH issue that happened around launch time. When assets are bridgeable it has wide-reaching implications across all chains that support the same asset. **2. Bridging creates cross pollination of v3 systems, and creates a more rigid version of v3 than what we’d like to see.** When we first imagined v3, we discussed creating two Mix-Yield Tokens per chain (ETH and USDC) with paired Vaults and Transmuters. This would allow v3 users to isolate their risk by picking a basket of yield assets they trust. But because alAssets are bridgeable, their risk is not isolated. In the worst case of a strategy becoming exploited or having an issue on one chain, underbacked debt from the affected chain could potentially cause issues on another chain. Bridging of alAssets also makes Transmuters less effective by reducing the variety of terms we can offer. Different MYTs may work best with different term lengths. In a world of chain-isolated alAssets, we would be able to experiment with multiple parallel term lengths, but since alAssets are bridgeable, we can't. Users will always just bridge over their alAssets to the most advantageous Transmuter, making our parameter tuning less effective and less impactful per-vault. With bridging enabled, Transmuter term users are also exposed to every MYT, instead of only the one on the chain they’re redeeming against. In the worst case where a strategy was exploited on one chain, and caused reduced backing for Vault and Transmuter holders, this would naturally affect the value of alAssets on that chain. # Solution . The solution is clear, we should move to deprecate bridging. The inflexibility it brings, expanded risk surface for users, and the very apparent risk of bridging in DeFi, is not worth the benefit of homogeneous liquidity across chains. If our community votes to take this route, here are the benefits to expect: - The ability to offer a wider variety of Transmuter term lengths - Reduced risk surface area for all users, and the ability to choose that risk more effectively - Peace of mind that we will not be exposed to bridging risk, which is far too widespread throughout the DeFi landscape - alETH and alUSD assets that are unique per chain, with different floating values, distinct backing and risk, independent implied Transmuter terms, and unique opportunities to trade them These changes would move us closer to a more sustainable way to operate Alchemix, which is to support only chains that are profitable. Every Alchemix deployment costs funds to maintain, and dev resources to be able to create strategies for MYTs. Additionally, with distinct liquidity profiles, liquidity per-chain becomes more important, affecting how much we can support periods of vault expansion, and the amount that fixed yield users can purchase for the same rate. With that being said, this would mean the winding down of v3 on Arbitrum. Although we have great respect and love for Arbitrum, the available strategies for our MYTs are limited, and uptake and TVL has been small relative to other deployments. This is counter to Mainnet for example, where the bulk of alAsset liquidity lives, we have a plethora of liquidity-driving assets, and more strategy possibilities. Most importantly, this would also allow us to reallocate resources improving yield on current chains, and launch on new chains. # Vote . A vote yes is a vote to deprecate bridging and v3 on Arbitrum, and reallocating resources to improve yield options on existing chains. If approved the Alchemix team will determine the accounting and necessary steps to deprecate Bridging and the wind down the Arb deployment.

Results