YFI can now be borrowed to participate in Yearn governance, via Aave

Well, it was never a question of “if”- just of “when.” YFI is now available for deposit and borrows on Aave. This provides a new benefit for YFI holders (to use their YFI as collateral), and may ultimately result in a more efficient market for YFI overall (with the ability to short it).

But with this may come some other consequences. This is by no means an issue yet, but it’s something I’d like folks to think about proactively:

Now that YFI is available for deposit on Aave, it also means that it can be borrowed. This means that a malicious actor (or just an actor with a relatively unpopular agenda among “real” YFI holders) could borrow YFI at the market rate and use it in YFI governance.

Current governance rules do not require any minimum staking period before someone can vote with their YFI tokens. All that exists (I believe) is a 3 day lock-up period on the back end where someone cannot withdraw their YFI after voting. That lockup on the back end is good as it will prevent participants from flash loan participating in YFI governance.

In looking at recent votes, we seem to have about 2000 to 5000 YFI participating somewhat consistently. We’ll have to see how these Aave markets evolve, but I could see it being possible (at some point) that for an especially close vote, an actor is able to borrow enough YFI to swing the outcome.

To be clear, I don’t view this as an imminent threat, but it’s something we should all monitor carefully and consider options to address if it becomes a problem. Some options might include:

  • Require a minimum time tokens are staked before they can be used in a governance vote
  • Lengthen the withdrawal period
  • Vest additional voting rights or profit rewards the longer tokens are staked in the governance contracts

Each one of these has progressively more severe drawbacks, but we may need to be proactive and consider one or more of these.

EDIT: I totally missed this, but here’s a proposal from Andre which ties to point #3 in possible solutions. I support a solution like this gYFI time weighted voting power distribution.

14 Likes

I think the first and second are good measures to add. I would like to see what the downsides are (I can think of some), and how these can be balanced.

I’m in full support of these options. I’d be happy if all 3 were implemented, but if we had to only pick one, it would be Option #1 for me.

This forces tokens to be locked in governance for long periods of time, which I would argue is good for the protocol long term, we need voters to take a long term view on YFI

Im liking the last 2 of those suggestions. The vesting additional rights one has been raised already by Andre in another thread actually.

I don’t like the first one because you’re basically guessing about what is coming up to vote. You may not be interested in voting in the next few agenda items, but when one pops up out of the blue you’ll miss it because your tokens wont have been staked long enough. Letting people in to vote immediately would be offset by the longer withdraw window though.

I actually think we should encourage people to keep their YFI staked in the governance contract and prioritize them over people who may use their YFI for other purposes, but only participate in votes opportunistically. I don’t have an issue with people using their YFI as collateral for borrows or speculation, but their needs are less important to me versus those who want to use it primarily for governance / protocol rewards.

I do think we should couple this with a formal “abstain” option for all votes (so holders can still earn rewards without being forced to cast a gainful ballot on an issue they may not understand or be interested in).

EDIT: A pre-lockup period before voting also presents other unique challenges. For example, if a hostile actor does try to sway a vote, they’d probably do it last minute, before anyone else could lock in their shares to vote.

6 Likes

Yep, totally agree with you there.

1 Like

I totally agree with this, and have mentioned before that I think we should be incentivizing governance more specifically for this reason. I’m not sure if this would look like increased fees to drive higher rewards to stakers/voters– but I do think that economic incentives are the best ones for governance participation.

For the voting process, I also think “abstain” is an important third option (for larger holders who want to claim rewards frequently) and vote delegation for smaller holders who want to participate in governance but don’t want to spend the gas– I’ve got a small group of friends who all hold less than 1 YFI and are in this situation.

Lastly, I really, really think we should implement Andre’s proposed weighting system. I still like the idea of weighting someone’s lockup equally across the entire lockup period, likely with a max of 1 year lockup (so 1 YFI locked up for 365 days is 1 gYFI, locked up for 182 is ~0.5 gYFI, etc). Perhaps we could do a poll on vote.yearn.finance for this– gauging sentiment mainly on A) whether the gYFI would be transferrable or not, and B) how the weighting should work for the lockup period (diminishing weight, static weight, etc)

3 Likes

For me the economic advantages to keep staked at the gov are crucial. e.g. if I have an official yfi vault (that can also gives access to vote) and that pays me a market APY, I would prefer to leave my YFI there.
I believe the goal has to be to have as much as possible YFI in official stake channels. Like this the community can counter attack any outside offensive (as the one explained in this thread).

2 Likes

Agree with DCInvestor. We need to be realistic about opportunity cost and incentives. Governance needs to be incentivized in a way where if I’m looking for yield/return, it is in my interest – a YFI holder – to pay attn to the votes & participate. If that’s not the case, eventually all roads lead to people using YFI outside the protocol & a possible attack vector forming, either say: borrow YFI + vote with it or degen farmers losing large market share of YFI and harder to now reach quorum or something else I haven’t yet thought of yet. Regardless, to direct this protocol to capture growth/scale/long term value we need stewards + community engagement and yes we’ll need rewards for that. Splitting the token or a synthetic token for voting seems like a bad idea with a lot of downstream effects potentially. I don’t expect gov rewards to compete with crazy 2000% flash farm meme coins, but if we can get a proper yield with much less risk (than that of degen farming) I think many (including myself) will step up and keep YFI in the protocol… let’s simply just create better incentives.

You may be interested in a similar (lengthy) thread on overlapping ideas https://gov.yearn.finance/t/gyfi-time-weighted-voting-power-distribution/

While it would be unfortunate for anyone to lose YFI, it would not affect quorum because those YFI were not part of the quorum calculation. Quorum is based on the number of YFI staked in the governance pool not the total number of YFI in existance.

We discussed the risk of a malicious actor borrowing YFI and prevailing on a system ending vote on the other thread as well. @Joey suggested that in the short-term even if a malicious actor prevailed on a vote, it would still require the actions of other to implement, and Andre or the multisig would refuse to implement. While not a good solution long-term, it makes me less fearful of such a takeover or catastrophic outcome – assuming Andre and the multisig are willing and able to go against the vote. Curious what those on this thread think about that.

Also consider what the real value of encouraging or forcing governance participation is when YFI holders choose to seek significantly higher yield elsewhere. Participating is important, but contributions can come in many forms – including work product, testing, ideas, and suggestions – and voting is just one way to participate.

2 Likes