gm, after reflecting on the basket proposal and discussing with key stakeholders I feel this proposal doesn’t hit the mark and fails to balance yield and risk.
Firstly the yield data presented is from multiple sources, Register app + DeFiLlama and appears to be mixed between 30d ma and current APYs. While unintentional, this has the potential to mislead governors. A standardised data set will be used going forward, likely DeFiLlama 30d ma’s.
Secondly, dgnETH has grown to such a size that liquidity considerations are now at play especially with longer tail LSTs and LP positions. After some preliminary work it looks like pxETH redemptions bottleneck around 750 pxETH and with a 50% or 1500 ETH allocation are certainly above the level of risk i’d be comfortable with even for dgnETH which is generally comfortable with a higher risk profile than a yDTF like ETHplus.
Finally, we have had 2 collateral basket rebalances in the recent past which have both eaten into the backing buffer which still hasn’t re-filled despite these rebalances being roughly one month ago. While the backing buffer is deficient there is no yield streaming to RSR stakers. If we were to go ahead with this rebalance and another we have planned in the future to update the ETHplus / ETH LP to the new Convex LP that would be 4 rebalances in this quarter. These frequent rebalances risk governor fatigue and also deplete the backing buffer meaning loss of yield for RSR stakers, both things which I’d like to avoid.
Instead I propose we do NOT proceed to on-chian voting for this proposal for the reasons mentioned above and instead re-visit the collateral basket rebalance with a closer eye on liquidity constraints once the ETHplus / ETH Convex LP plug-in is ready.