MIP 84 - Morpho Optimizers Deprecation Phase 3

Morpho Optimizer
1

A recent upgrade of Aave to version 3.2 introduced breaking changes which affected several external protocols, including the Morpho Aave v3 Optimizer. The upgrade included modification of the emode feature and deprecation of stable rate borrowing on Aave. After the update, the function Pool.getReserveData() returned address(0) for the stableDebtToken address, causing calls to this token to revert, and making liquidations and loan repayments impossible on the Aave v3 Optimizer. The Morpho DAO Multisig took emergency action to pause new borrows on the Optimizer to mitigate risk while a fix was worked on. Fortunately, BGD proposed rolling back the breaking changes due to the effect on integrators, which was executed by the Aave DAO, mitigating the need for a change to the Optimizer code.

These events highlight both the merits of the immutable Morpho model, which gives integrators confidence that they do not have to monitor and respond to changes in the underlying code, and the ongoing risks presented by the Aave v2 and v3 Optimizers.

Morpho Blue now has sufficient liquidity to welcome all optimizer users, and there is a dedicated interface for migration. In light of these facts, the Morpho Association proposes that the DAO take further steps to deprecate the Optimizers, which began in MIP-52 and continued in MIP-71. All MORPHO rewards for Optimizer usage were discontinued in July, and usage of several longer tail assets paused.

We propose that the DAO set a target date for a full pause of new supply and borrow on the optimizers of December 20 to provide ample notice to users and integrators. The DAO would vote on this proposal in advance, and then on that date would take the following actions:

  1. Pause supply and borrow of remaining unpaused assets on the Aave v2 optimizer (WETH, USDC, and WBTC)
  2. Pause supply and borrow of remaining unpaused assets on the Aave v3 optimizer (wstETH collateral, supply/borrow of ETH)

This will prevent all new supply and borrow across all optimizers. Existing positions will not be impacted. After allowing time for suppliers and borrowers to close their positions, the DAO can consider final deprecation steps in a future proposal, such as adjusting parameters to encourage closing any remaining positions, and removing optimizer support from the frontend.

6 Likes
2

Gauntlet supports the pause and deprecation of the optimizer integrations with Aave v2 and V3 and the eventual deprecation of the optimizers from the front end. Morpho’s immutability is a distinct advantage somewhat undermined if there is significant dependence on Aave governance via the v2 and v3 optimizers.

3

In general I have nothing against this proposal, but why kill the optimiser product? It was a nice one, and it doesn’t look like it required a lot of maintainance and costs, am I wrong?
Perhaps we could have just put a fee on it, directing revenues to MORPHO holders in some way?

4

As noted in the OP, one of the reasonings for moving forward with the deprecation is that maintaining the optimizers requires the Morpho DAO to monitor for any breaking changes or dangerous risk management decisions from the underlying protocol. In the recent incident, a breaking change by Aave resulted in an inability to liquidate positions, which could have caused bad debt.

1 Like