Proposal: Extending time lock duration on yGov pools (voteLock)

Proposals Vaults
1

Would like to see a discussion for polling consensus on this proposal on the time lock (voteLock) for staking on yGov Pools, using yCrv BPT or YFI:

Those who have voted in the last ~3 days cannot move those staked tokens on both yGov pools. This situation inevitably creates a positive feedback loop for voter participation through governance, either via voting on or creating proposals.

Vote For:

Extending time lock from 3 days to 7-10 days

Voters Turnout
The reasoning in having a proposal for extending the timelock is that:
non-voters with smaller holdings might be abstaining their votes, which is wasteful if the minimum quorum of more than 33% of staked yGov need to be met. Similarly coupled with a longer voting window on proposals, does help in improving voters turnout rates.

Price Floor
Another way to wrap your head around this issue: unless voting participants are planning on giving up the fee rewards alpha; in the near term, voters above the threshold of 1.000 BPT will continue to stake using their YFI rewards in pool #4 or in the case of smaller holders on other pools, which increases the price floor.

Incentivizing smaller holders
Lowered entry requirements of minimum BPT staked for collecting rewards can have the same effect, see related discussion on incentivizing smaller holders to be rewarded for voting participation
(note that anyone can vote on proposals, with a balance of above 0 BPT currently):

Proposal: Lowering minimum BPT staked requirements in Governance pool to qualify for governance rewards - #5 by milkyklim

Vote Against:

To stay / reducing current 3-days time lock

Status quo
Whereas maintaining an Against vote, it is clear that both rational voters and non-voters alike will freely utilize claimed YFI rewards to stake for yield on Pool #1 - #3 (in the case of active voters staking in Pool #4, after a short period of waiting for 3 days)

While a portion of claimed YFI rewards are being sold with no restrictions at market using current Pools #1 - #3. Price will stabilize in the long run if Prop 0 is passed.

Ongoing proposals
In the case where Prop 0 is not passed or likely re-tabled, and supply remains capped @ 30k till rewards run out, that would defeat the purpose of decentralizing the protocol, although most would continue to pool their YFI tokens with Balancer pools; impermanent loss (IL) might deter even the most determined, since IL reveals itself as ā€˜permanent lossesā€™ at those higher prices and pool ratios.

Proposal 0 and integrity of the voting process

Personally I have no qualms, having the current 3-days time lock to stay, however I would like to discuss the merits of a proposed longer timelock of 7-10 days, as opposed to reducing timelock to less than 3 days.

We welcome your feedback!

3 Likes
2

Referred to these posts for ideas, when contributing to this proposal draft

a. Proposal How-To
b. How Onchain Proposals are Passed

3 Likes
3

More context on 3-day period:

Proposal votings are running for 3 days, hence, 3 day voting lockup to avoid double votes.

If voting lockup period is reduced, votes counting mechanism should be changed/revised.
If voting lockup period is extended, votes counting mechanism can stay the same.

5 Likes
4

To be clear, this proposal discussion is on the duration of the voting timelock: the voteLock applies in the following yGov pool contracts:

yGov 98/2 YFI/yCRV pool (staked BPT),

yGov YFI fee rewards pool (staked YFI) -
Presently this pool stake is not required to be locked for voting to occur, see this noteworthy rfc thread:

Proposal: voting only with YFI tokens - #22 by markpui

Another related proposal:
For vesting of YFI rewards from the Governance Pool, on a rolling 12 months schedule

(See RFC: Proposal B inset)

Request for Comment : Proposal for YFI Inflation and Supply Mechanics

Proposal: yEarn Rewards Reserve

1 Like
5

I agree with extending time lock duration. Ideally stake of voter should be locked until change, he is voting for, is made to the system

Right now, stake is locked for 3 days, but voting lasts for 5 days. This makes possible situation when bad-decision maker can do harm to the system and withdraw the stake without any harm to himself

1 Like
6

Right now, stake is locked for 3 days, but voting lasts for 5 days.

This is fundamentally wrong. Both voting and lock are ~3 days to prevent the problem you described: Proposal: Extending time lock duration on yGov pools (voteLock) - #3 by milkyklim

2 Likes
7

Agreed, current ~3 days voting timelock and voting window applies to the rest of the proposals

@cryptoseal Curious how though, the first two proposals Prop #0 and #1 had voting period of 5 days, as initially stated on the tweets.

Also on the voting interface which now features disabled buttons on yGov pool; but not for Fee rewards pool, am I missing something?
Does this mean that the staked YFI can be withdrawn at any point regardless of voting?

yGov Pool UI: (near the footer)

Unstaking tokens only allowed once all your pending votes have closed at Block:

voteLock[msg.sender] = lock.add(block.number);

8

Are you sure it was not applied to 0 and 1, too? And what are the disabled buttons your are talking about?

We are effectively voting with BPT, therefore, YFI in 4th pool are not locked.

1 Like
9

Sure that clears it up, might have to amend the post to reflect this!
ty

However, some proposal drafts have advocated to YFI as stake for voting, and ability to collect fee rewards in pool #4

10

Since I have voted in the past ~3 days, there is a nice notice displaying, with those two unstaking buttons greyed out at the bottom

Unstaking 2020-07
Unstaking 2020-071044Ɨ390 23.6 KB

1 Like
11

This is nice update to UI! I donā€™t see that one cause I am not locked.

2 Likes
12

@zirs3d thanks for setting out the proposal & some thoughts around it.

The concern: Lack of voter turnout, leading to lack of quorum on proposals calling for a vote. For a proposal to pass (either for/against), quorum must be met. Not enough holders are voting as measured by staked yGov.

The (hypothesized) sources of the issue (and how they could be solved):

  1. Individual voters with smaller staked yGov feel that their voting ā€œvoiceā€ will not count against other voters with larger staked yGov.

In the conventional (non-crypto) space, this does not exist in a legal sense as most democracies are a 1-person, 1-vote system. The closest fix is to cap amount of staked yGov that would count for a particular address.

  1. Voters do not feel strongly about an issue enough to vote, or they do not have any information to feel they can make an informed vote. This could be lack of information about the poll itself.

In the conventional (non-crypto) space world, this is fixed by:
2.1. Having a longer time period for a proposal to be considered. This is like the reading of bills in parliament or senate. This allows sufficient time for publicity around the vote. In theory, letting a proposal be open longer for an on-chain vote should improve its ability to meet quorum.
2.2. Commissioning/applying detailed research on the cost-benefit impact of the proposal outcomes. This is similar to the independent economic agencies producing impact analysis. This allows voters to understand assumptions & modelled outcomes in greater detail.

  1. YFI tokens play a dual role in the eyes of the holders: They have a monetary value in terms of crystallizing rewards, yet they also have value in terms of their ability to perform governance.

This is evident in terms of the impact of time-locking. YFI token holders need to consider if they would like their liquidity to be locked (which may represent an increasing opportunity cost of capital if there is a more profitable opportunity to deploy the locked (& converted) YFI elsewhere).

In the conventional (non-crypto) world, this is solved by the fact that you canā€™t (legally) sell your vote. You canā€™t monetize your vote in the short term - and only hope that your vote leads to a better economic outcome for the economy in the long term.

A straightforward example of a solution that I have seen is the one proposed by @cp287

I believe his/her proposal is worthy of debate & consideration.

Taking the above issues & solutions together, there could be a combination of different methods to improve quorum for proposals. For example, I would propose considering a combination of 1 or more of the following:

  1. Vote only with YFI tokens, not staked yGov.

  2. Implement a minimum 2-4 week period for any proposal to pass.

  3. (related to 2 above). Designate an advocate/s for the FOR argument, as well as for the AGAINST argument.
    In the private equity space, this is similar to having a devilā€™s advocate on the investment committee as transactions are evaluated. This helps expand the debate. The advocatesā€™ help generate thinking & debate around their side of the argument that then better informs the entire community.

  4. Assuming staked yGov remains, then, when counting quorums, exclude smaller staked yGov.

  • All votes count.
  • But when it comes to quorum counting, exclude smaller staked yGov. E.g. quorum is determined by those above 10,000 BPT. However, this leads to another issue where quorums could pass, with not enough votes cast, simply based on large holders.

Relating to the proposal to expand the time-lock period, I am concerned that it may discourage voting if the entire system remains based on staked yGov (since a lot of staked yGov is chasing yield). If people see the time-locking as an opportunity cost, then I think it may not have the intended impact to improve votes. However, I do like having a longer time-lock period if YFI tokens are applied as the sole voting token.

4 Likes
13

Totally support this. Time locks are good for defending against sophisticated governance attacks involving short-selling.

2 Likes