Signal Request: Should we move to exponential rate stepping for stability fee polls?

Stability Fee increments in the polls should not be constant, they should be exponential.

Advantages

  • Lets us react more strongly to changing market conditions if necessary, without having to trigger an exceptional poll in an emergency.
  • Lets us do small scale tweaking if we want to try to find the lowest stability fee setting that maintains the peg, as some MKR holders seem to have voted for recently.
  • Lets MKR holders fix the .5% precision ‘problem’, if they want to.

Disadvantages

  • It prevents us from making more precise changes in a single poll. If we wanted to make a specific change for say 3%, this would take two polls under the proposed system.

The following options would capture doveish/hawkish sentiment in both directions with a minimum of polling options:

+8%
+4%
+2%
+1%
+0.5%

No change

-0.5%
-1%
-2%
-4%
-8%

  • Yes, we should do this.
  • No, we should not do this.

0 voters

Anyone signaling against this want to summarize their views? It seems like an easy win from my perspective.

1 Like

8% +/- options seems like too big a hammer to have on offer while there are only a couple of unchecked whales taking part in voting.

With eight options its one less poll option then the current setup and gives the possibility of finer control, whats not to like?

I’m in favor. I like math. :grinning:

Okay, not hearing a wide range of objections here, one thing that was mentioned was the +8%, -8% being too high. Is this a commonly held opinion by those voting against? New signalling poll to capture this:

Should we move to a exponential rate setting for stability fee polls?

  • Yes, limited within a range of -4% to +4% is best.
  • Yes, limited within a range of -8% to +8% is best.
  • No, I have a different objection (please consider replying if this is the case.)

0 voters

Didn’t @Vishesh suggest 0.25% changes might be enough this last call, maybe we can remove the -8%/+8% options in favor of adding -0.25%/+0.25% options?

3 Likes

So there are 2 dimensions within this signal request (and 1 poll):

  • number of choices
  • initial step (lowest fee change)

I think current poll is ok and can discard 2nd parameter for now. I see lowest fee of 0.5 percent as low enough and we could lower it later if necessary. Historically mkr holders raised fee too slow (from 0.5 percent to 19.5 percent) and did not use 4 percent increases (even when they could).

Regarding t-bone’s argument about the limiting max range because of the danger of “unchecked whales”.
It sounds like “censoring” options for mkr holder to vote on. I am not saying i am against it, but it does open some interesting questions about governance (which were likely already discussed, so i probably should not say open).
Maybe giving largest mkr holders more opportunity to game the system, would give new information about the robustness of the system or maybe also force other holders to also vote? Anyway, this questions are for other topic.

We haven’t really moved any closer to consensus in the last week here. To be honest I don’t see the whale argument as a particularly valid reason not to have +/-8%. Whales own MKR, they are valid governance participants.

As Jernejml said, there are multiple dimensions to this. Can we try to come to a consensus on an ‘acceptable’ outcome rather than a perfect outcome? Remember that we can poll for a second change in stepping at a later date if we decide that the options are not fit for purpose.

With the new polling dashboard, having multiple options is less invasive. Can we come to consensus on one of the following?

You can vote for multiple options in this poll. Please vote for all options you would prefer to the current linear steps we are using.

  • 0.25%, 0.5%, 1%, 2%, 4%, 8% (plus and minus)
  • 0.5%, 1%, 2%, 4%, 8% (plus and minus)
  • 0.5%, 1%, 2%, 4% (plus and minus)
  • 0.25%, 0.5%, 1%, 2%, 4% (plus and minus)
  • All of these are unacceptable, we should stick with the current system

0 voters

Lets try to find a consensus and hopefully get to a point where we can move this to an on-chain vote.

4 Likes

I think we need to impose some timelines for this stuff. Stating the duration of the poll and then proposing a week in which the poll should hit the governance portal would probably spur some activity.

Calendaring is something we need to address soon.

Campaigning as well. Interested parties should consider hitting up the socials and chats to generate interest in their chosen option.

Maybe some ‘Active Poll’ flag or meta thread?

2 Likes

I’m hesitant to support fixed timelines, as they will lead to on-chain polls happening without solid consensus backing them (in our current state of 63% consensus on two options, I don’t think this should move on-chain yet, we can do better). I think we need to look at other ways to spur activity: Keeping the thread active? Bringing it up more in meetings? General recruiting for governance bodies?

I was hoping to see more compromise on this thread, to me this feels like a fairly non-controversial QoL change. It doesn’t matter hugely to me which version of this we take, but others seem to disagree.

There are currently six people that are voting for one of the top two options but not the other. Are any of you willing to compromise further? The difference is having an 8% option in the poll. This poll does not commit us to ever use the 8%, ultimately MKR vote will decide the stability fee, as it does now. Not having an 8% option doesn’t secure the system in any appreciable way. If whales want to jack up the SF, they are going to do it under whichever stepping we end up with.

I have trouble seeing the 8% option as a sticking point. Do any of those six want to speak to their views for or against?

(not today… but in the future)… what is the general thought on a weighted consensus model? such that if we have (10k mkr in favor of 10%… and 1k mkr in favor of 0%. that the ratified amount would be the weighted average … not just 10%… thoughts?

I can’t see a situation where we would need to make a 8% change?

  • it’s hard to imagine unexpected situations in advance
  • imo extra flexibility outweighs potential ‘larger whale attack vector’, if there is any at all (we still have executive vote)
  • if there won’t be any significant changes to the poll in near future, i will remove my support for 8p option and go with an options with wider consensus.
2 Likes

Hi.

Agree with the notion that exponential changes in SF is a good idea.

However I disagreed with both the suggested limit ranges (hence this reply). I assume that fixed limit ranges are currently the only possibility however imo it would be a good idea with an option that is double the previous change whatever that change was. So if we for example reach +/-8% then the next poll should include +/-16% options. This should then of course be reversed at some point when appropriate.

Br.

Hi Gin, great to hear from you, welcome to the community. :slight_smile:

I believe variable stepping ranges would be possible, but I think it adds a level of complexity that isn’t entirely necessary. We are already adding complexity by potentially moving from fixed steps to exponential steps.

We should try to move one step at a time here, I think. If we can get exponential stepping agreed and passing an on-chain vote and it turns out to be useful, then we can consider further improvements.

I’d like to speak to voting here quickly, as I’m not 100% sure that all participants are approaching it in the way I’d hoped they would. We are currently voting in the poll here.

This poll allows you to vote for multiple options. You should vote for any option that you would vote for in an on-chain poll. For each option pretend the choice is between that option and remaining with the current fixed rate stepping system for stability fee polls.

It feels like there should be more compromise here, no one has voted for the option that we should remain with the current system, so each of the 16 people that have voted want to see a change, however, until we can get a higher degree of consensus and more participants it’s difficult to justify moving to an on-chain poll.

I would suggest that we are looking for at least a 75% consensus around one option, and that it would be great to see at least 20 participants total to vote in the poll. If we can reach this level of agreement, I’ll start pushing for this poll to move on-chain.

  • I am against exponential, but I don’t have strong feelings about it at the moment.

  • I think for our use case, small linear adjustments works just fine. I don’t anticipate another situation like the beginning of the year where we were so far off the peg. I feel like any situation where we would need to make an 8% change at once is rare enough where it would justify a special emergency poll, and thus it would be unnecessary to have that option during normal operation.

  • Additionally, I think anything less than 1.0% is unnecessary. 1% + 1% increments is simple and effective.

Why don’t you anticipate such a situation? In my mind a similar situation would be more likely to occur than not. And even if it wasn’t I would still prefer the most powerful “tool” for handling such a situation just in case. It’s really bad if the peg is off for weeks and weeks like the last time imo.

Do you have any other reasons to be against exponential changes other than what we have now works just fine? I have heard other arguments that it would be too violent a change and therefore risk market instability. If you have already answered any off these questions anywhere please feel free to just link me there.

A few questions I see need to be answered:

  • Is there some format for Governance Polls that the community can follow and submit here on the forum to ease the burden on @rich.brown and others involved in the process of setting up proposals? Can we create some sort of guide for this perhaps?
  • What is enough consensus? It seems to me that the majority are for this change and that we merely need to poll for the specific configuration.
  • What is the duration of time for this specific Governance Poll when/if it happens?
1 Like

I don’t think we’re likely to see any more movement on the last poll. I’m still unsure whether 62% consensus is enough to get this through, but lets find out!

This poll will determine what the on-chain poll looks like, my hope is that eventually we can set enough of a precedent that this won’t be necessary every time.

  • #1 Create an on-chain poll in the form: ‘Consensus Option, yes or no?’ - We implement the consensus option if Yes > No
  • #2 Create an on-chain poll in the form: ‘Which one of these suggested options do you want implemented (includes a status quo option)?’ - We implement the highest scoring (unless status quo wins, in which case we do nothing.)
  • #3 Create multiple on-chain polls, one for each option in the last poll, each has a yes/no option. - We implement the outcome with the highest ‘score’ so long as that ‘score’ is positive.
  • #4 This is not ready to go to an on-chain poll.
  • #5 I just want to see the results.

0 voters

Here are some of my thoughts on the above:

  • On #1: The consensus option is ‘0.25, 0.5, 1, 2, 4’, This is fixed from this point, everyone has had time to signal if they were inclined to do so. I will be closing the previous polls.
  • If we go with #1 we are allowing 1-person 1-vote to control which proposal makes it on-chain. This is putting a layer of democracy over our plutocracy. I would argue this is good, but others may not feel the same way.
  • #2 is plurality voting. It can result in splitting the vote, allowing status quo to win even though there is more consensus for some change, even if the flavour of that change can’t be agreed upon. I’m very against this option.
  • #3 is similar to the asset ranking poll. Here ‘score’ = ‘yes votes’ - ‘no votes’. We would be ranking solutions. This might be overkill, we’ve essentially done this already with 1-person 1-vote, but this way will give MKR holders the final say.

Based on suggestions that we need more calendar. I’ve set this poll to expire 1 week from today.