Signal Request / Poll: How should we structure monetary policy polls in MCD?

Hey all,

As we discussed on the governance call yesterday, we have to start figuring out as a community how we want to structure polls in this new multi-lever world. Gone are the days when we can just look at the Sai price, check if it’s above/below $1, and vote accordingly. We have to dip our toes into the world of active governance.

Of course, we want to crawl/walk, not run, and we have to be cognizant of not introducing too much overhead with respect to voting. I’ve put some ideas out for how these polls might be structured, but please feel free to comment on other ideas.

Option A: 2 polls. This option is the most restrictive, but carries the least overhead. One poll would be for the DSR. Another would be to manage 3 stability fees per option - SCD, MCD-ETH, MCD-BAT. Given that SCD is likely to go to 4% soon, all 3 would be the same. So then all 3 stability fees would move up and down in tandem. This obviously doesnt work if we need to manage the two pegs separately, but potentially we could create a contingency where we only make the polls more granular as needed. If the pegs are all healthy, or off slightly in the same direction then we can change all 3 at once.

Option B: 3 polls. Slightly more flexible. One poll for DSR, one for SCD, and one for MCD-ETH and MCD-BAT which will stay identical. The reason for keeping BAT and ETH the same in MCD is because we haven’t quite started diving into risk premiums yet (and even if we had, we still might choose to ignore them for the sake of stability, user growth, or migration incentives). This option allows us to manage SCD and MCD independently from one another.

Option C: 4 polls. One for each parameter: DSR, SCD stability fee, MCD-ETH stability fee, MCD-BAT stability fee. This one allows us to manage the BAT stability fee separately if necessary.

Separately, we should maybe create a separate discussion for debt ceilings. That one needs some brainstorming before we jump into polling. Thankfully, I think we have a bit of leeway in terms of timing as it doesn’t look like any debt ceilings will be hit anytime soon.

  • 2 polls (DSR and Stability Fee Package)
  • 3 polls (DSR, SCD Stability Fee, MCD Stability Fees)
  • 4 polls (DSR, SCD, MCD-ETH, MCD-BAT)
  • Other (please comment)

0 voters


Could someone please explain what, if any, are the gas implications of the different options?
Is it as simple as fewer polls = lower cost?

Pretty much, I think the gas cost to vote on any individual poll is pretty much identical. Having more polls is slightly more gas, but honestly, the gas costs to vote are negligible. If you vote early you can get away with 1 Gwei and it’ll usually go through before the poll ends.

Sorry I have been away from MakerDAO for some time so were a bit confused by referring to governance votes as polls. It is governance votes and not forum polls/signaling that is being discussed here right?

Yeah, this is a poll to determine how we might structure the governance vote/governance poll (interchangeable names I guess).

I think option B makes sense for the present moment since migration is a still big question mark. Seems like we might have to adjust the SCD stability fee at some point to incentive people to do something. BAT is little so shouldn’t need own adjustment. However for MCD I’d say sticking to option A with the contingency that we could make the poll granular makes the most sense to me. Maximizes time that voters dont need to think as much, minimizes gas costs, yet gives us a chance for respond to an isolated circumstance. We would just need to clearly outline what circumstances qualify as needed means

One of the things we haven’t discussed in this thread is “when do we start this new MCD polling?”

The Current Signalling Process dictates a 7 day discussion period in the forums, then 7 days of a governance poll, then the executive.

If people think that is too slow a process we should start a discussion about whether there is a need to implement the above polling solution sooner.


I’m going to quibble with the wording here.

The current process doesn’t really dictate anything, I specifically wrote in the first sentence that it should be seen as descriptive rather than prescriptive. It was always meant to be a write up of what I was doing, and a framework for others to follow if they needed guidance rather than me insisting we use that process.

Ideally, we would have started the process before the launch of MCD so that it could be finishing properly now (obviously hindsight is 20/20.) Since that didn’t happen, and since we’re going to need to be able to vote asap, we should either:

  • Taking the current result (3 polls with 72% share of 18 total voters) as a fairly strong signal and move on-chain with ‘3 Poll’ yes/no.
  • Move the poll on-chain immediately with all three options and implement the winning option.
  • Just take the forum result as given and start doing 3 polls each week on the usual cadence.

Personally I’d be fine with any of these, but we should be trying to expedite the process given the slightly rocky migration so far.


Hi Cyrus. I have two alternative options. I am only making these suggestions since we have no empirical data, and in the spirit of “training wheels” and they should be eliminated as soon as possible. At least initially, we make the system resemble the old system as much as possible just to get DAI going.

option 1: (2 de-coupled parameters)
We fix the Stability Fee for the two DAI collaterals. We can look at the avg historical rates for eth and BAT from lenders (Compound, Fulcrum Nuo, DxDy, etc) and figure out the numbers. I am guessing somewhere between 4 to 8, and lower for ETH. We then just do 2 Polls, DSR and SCD. ( This is very similar to option A), but different in the sense that we need DAI and SAI policy de-coupled. Prelim market data is pointing to divergence between the two pegs, so to keep them both stable they are likely going to require separate even opposing policies.

option 2: (3 Parameters )
Initially fix DSR to zero, so the two systems at least initially are alike, get the pegs stable, collect some data, then we can start experimenting with DSR. It maybe too late for this option, but thats how I would have started.


1 Like

I would go with this one. 20 votes is significant number of voters (historically in this forum) and 14/20, while other votes are distributed evenly, is solid consensus. We are early in migration period and i think it’s prudent to compromise on signalling process.

1 Like

Hey Patrick.

Maybe in some other context these would be applicable, but neither are relevant to our current situation.

Option 1 implies that the historical rates from secondary markets are meaningful. In fact, they are just derivative of the stability fees maker governance has chosen. There is 0 justification for using these historical rates as a path forward.

For option 2, the DSR needs to be going up right now, not down. We can reset DSR and get the pegs stable, collect data, etc. What we need to do is fairly clear. DSR should be going up to incentivize migration and SCD stability fee needs to go down to incentivize liquidity provision.