What if the YFI Vault strategy was rolled into the staked governance?
That would alleviate the incentive to put any in the vault and entice users to stake rather than vault.
So instead of getting Ycurve as a reward you’d get YFI that was market bought.
Anything that changes the voting power of a YFI based upon the identity of the holder, whether it is blacklisting certain holders from voting or diluting the voting power of a YFI by apportioning it between the YFI itself and derivative products (e.g., aYFI), may seriously harm its value as collateral.
If a creditor takes full possession of the collateral (such as the case with Aave), that creditor likely expects to have full control over the bundle of property rights associated with the collateral. Furthermore, any arrangement that introduces uncertainty into the equation so that a creditor cannot be certain precisely what it is receiving from YFI to YFI will introduce all sorts of issues.
Before moving forward with any proposal that would divest YFI from bestowing the holder with full voting rights, I would strongly encourage the team to have discussions with other stakeholders in the DeFi field to learn their views on what effect, if any, implementing such proposal would have on YFI as collateral.
Perhaps this entire thread needs to be reiterated.
First off, CEX are not smart contracts, there is zero audit-ability, and users deposits WILL be used to divert governance rewards to the CEX themselves, period. There is no benefit to allowing CEX participation using funds intended for trading (no user intended delegation by depositing), this can only be remedied via international court imo.
Either you have a snapshot mechanism, which by definition lends itself to funds being shuffled in order to vote in the snapshot, or a vesting mechanism which requires locked funds and guarantees legitimacy.
If smart contracts are allowed to vote via a snapshot, then vesting comes with an opportunity cost. So the real discussion is around how to implement snapshot voting via contracts with the understanding that flash loans open an attack vector to divert gov rewards within one block.
No user has intentionally delegated governance via lending yfi under a vesting regime. If there is a technical implementation that prevents entities like AAVE from diverting gov rewards, then that is the solution. If there isnt one, then the snapshot idea is flawed.
There is no basis of reasoning behind delegation, because no user has ever delegated yfi before. CEX involvement is truly unacceptable regardless of implementation.
For the CEXes, there’s nothing stopping them from writing a smart contract that aggregates their users’ funds and uses that vote. We’d be able to see it on-chain if it emerged. If Coinbase, for example, does that, how should we respond?
I’m unclear about your flash loans point. I think most people are still in favor of giving rewards only to those who stake in the governance contract. Can you explain the attack vector a bit more?
In the case of staking, there is a limit to how an exchange can utilize funds on chain because their actions are monitored, large enough to consistently require OTC, and users will have throttled withdrawals.
The issue with a snapshot being you just sign with the token, no lockup, a contract can borrow yfi to sign and repay the loan with gov rewards within the same transaction.
I think you can set snapshot to use a block number from before the poll starts. So if this is chosen to be some random but small number of blocks before the start of the poll, actors can’t necessarily predict when to do a flash loan attack. Would this work to mitigate things?
Agreed. Are we thinking of removing the 3 day vote lock? If not, how would a flash loan work in this instance? Even without a lock, would the flash loan borrower not need to know the exact block in which to borrow?
Even a flash loan is just an example. In truth, the cost to borrow YFI in advance of the snapshot is potentially cheaper than holding YFI for the snapshot outright. The gov rewards at a certain size pay for the interest on the loan; in that case CEX have deeply ingrained edge to divert gov rewards given the bonding curve model of AMM.
This behavior has been commonplace in AAVE, govtoken deposited in advance of being borrowed for use in a snapshot - resulting in spiking lending rates, the entire process burns gas and CEX just have to transfer user funds.
A merkle root in the new Governance v3 contract sounds great.
Thank you for helping writing up YIPs.
I cannot wait to vote on the upcoming active proposal.