Flop Debt Auction Parameter Adjustment Discussion

On the 3/17/2020 installment of the Daily Governance and Risk call, the community discussed parameter changes on upcoming scheduled flop debt auctions. At present, auctions will open at 10:28 ET (14:28 UTC) on Thursday 3/19/2020.

Background info:

The flop auctions have been well documented in the forums and here. The technical parameters are detailed here as well. One of the details that has been discussed has been the current starting price. During the MCD launch in 11/2019, the parameters were set for lot sizes of 250 MKR at 50k DAI lot which works out to 200DAI/MKR. Although DAI/MKR is largely illiquid, it would currently sit at ~215 DAI/MKR at time of this writing. Given the recent volatility of all assets (ETH, MKR, DAI), this DAI/MKR pair could be volatile and change during the run up and/or over the course of the auctions. As a result the starting bid could prove to be problematic if 200 Dai for the flop auctions were more expensive than the market rate. This would be the main issue and would interfere with smooth running auctions.

What is the goal of this adjustment:

  • to ensure that the process of debt auctions clears the market and the debt can be removed from the system
  • to keep the auction on schedule

Adjustments discussed:

Should governance adjust the flop auction parameters ahead of Thursday? Suggested adjustments would be between:

  1. Increase the lot size (lower the starting bid price for the flop auction)
  2. Decrease the flip (allows quicker auction resets)
  3. Increase the size of the discount factor such that minimum bid price drops faster in between failed auctions

General Pros and Cons


  • lower starting price welcomes more bidders to the auction and increases likelihood that market will clear
  • decreases likelihood of failed auctions that could potentially hurt confidence, especially if bad debt lingers in the protocol for a long time
  • avoids potentially delaying schedule for flop auctions if we were to experience extreme liquidity event over next day or so


  • introduces confusion of logic behind starting bid price. Selected values could seem arbitrary
  • keeper software might have to be updated on short notice, increasing technical risk, especially if the executive vote clears in the middle of the auction. Additional technical risk to implementing governance poll and executive vote on such short notice
  • could be an opportunity for speculators (would this open up opportunities to short MKR?)


This forum post is a high-level overview of adjusting the upcoming flop auction parameters. The next step could be a poll for the suggested adjustments.


Aight, I’ve discussed the following polls with Cyrus. I realise we don’t have a full breakdown of what the impact of changing each parameter is, but the general feeling seems to be that they all more or less have the same effect of (hopefully) making the FLOP auctions more efficient / effective.

Should we change the FLOP auction parameters before Thursday?

  • Yes
  • No
  • I don’t know

0 voters

Which FLOP auction parameter changes are you in favour of?

Current Values
DUMP = 250 MKR
TAU = 72 Hours
PAD = 20%

  • Increase the lot size (DUMP parameter) resulting in a lower starting price-per-MKR.
  • Decrease the maximum auction time (TAU parameter) such that the auction resets more quickly if there are no bids.
  • Increase the discount factor (PAD parameter) such that when the auction resets, the price-per-MKR decreases by a larger amount.
  • I don’t know.

0 voters

You can vote for multiple options in the second poll. Please vote for every parameter you would be happy to see change.

The outcome of these polls will either influence inclusions in the executive vote, or may determine what goes into on-chain governance polls (though I am unsure if these are going to occur given the time constraints.) If I understand correctly, the specific values of these changes will be left to the Interim Risk Team to determine.

@Aaron_Bartsch has added some thoughts and a poll for delaying the FLOP auctions below.

Can we add a poll for delaying the flop auctions entirely? If you want I can make a proposal thread but since we’re already here I figured it would be a good idea.

The reasoning behind delaying is to provide extra liquidity in these trying times. Typically unbacked dai is bad if the peg is below $1.00 but since we’re actually above, the extra 5M liquidity is actually beneficial to the system. Signalling a delay would allow the 5M that keepers are trying to accumulate to go back into the market, allowing the peg to drift back down.

Another factor is the fact that it currently costs a premium to obtain Dai. Keepers wanting to participate in the FLOP auctions are going to take this premium into account when calculating their bid parameters. If the Peg is closer to or lower than $1.00, keepers will be willing to accept MKR for a higher price.

I propose that we delay the flop auctions until the average dai price is at or below $1.00 for more than 24 hours.

Edit: Based on @LongForWisdom’s advice below I have created the following poll. See reasoning above:

Should we delay FLOP auctions until the overall VWAP on http://dai.descipher.io/ for 24 hours is at or below $1.00?

  • Yes
  • No

0 voters

I could add this, I think maybe in this case feel free to create your own poll in this thread? I don’t want to become the gatekeeper of polls.

I’ll be happy to add a giant: “LOOK DOWN HERE AT THE OTHER POLL” to the first reply if you do :slight_smile:

Interesting option @Aaron_Bartsch .

What if the <overall VWAP on http://dai.descipher.io/ for 24 hours doesn’t reach at or below $1.00> anytime soon? Would that in effect be nullifying the decision to hold the auctions? Should there be a limit to hold long we would delay it?

1 Like

Why? Whats the rush? The real problem with having unbacked Dai is when there’s too much of it and peg is below $1.00. I’m not following why we have to 0 out the system so quickly. What are the optics?

The biggest problem we face right now is another flash crash and liquidity crunch. Delaying the FLOPs are the best way to stop this from happening again right now.

I could see worst case scenario is that we seesaw way down past the peg now that DSR and SF are essentially 0% but I think the above scenario has a much higher probability of happening.

The rush in my mind is that if we were forced to call ES right now, Dai Holders would not get $1 worth of collateral, they would get less than that, maybe (made up numbers ->) 0.97? 0.95? 0.9? The less Dai that exists the worse the potential haircut gets.

Dai is supposed to be backed, that’s essentially our job, to deliver a stable currency that is fully backed by collateral. We are not currently doing our job, so we need to run the FLOP auctions.

The only point at which I’d say we shouldn’t run the FLOP auctions is if running them risks the entire protocol collapsing. Despite the insanity of the last week, I don’t think we are quite there yet.


The parameters were already widely publicly communicated, so I’m against changing them this close to the start and especially when auctions are running. If it comes to this, I think reducing tau is the safest option to achieve goals without causing more confusion. Another hidden advantage of the current parameters is that they can help spread out the auctions.


There are several ideas to unpack here.

First, we don’t know at what point we’re going to ES. You can see the Foundation is very hesitant to table that discussion as its the last last last resort.

A more likely scenario is that we pause auctions so that liquidity can come in and continue liquidating vaults. In order to do this we need more Dai in circulation. If we wipe out 5.3M more Dai in 2 days and we crash on Friday where does that leave us. I’d say delaying the FLOPs at the potential for a devalued ES is a safer route than removing liquidity during a liquidity crunch. It’s just too much all at the same time. I guess this boils down to philosophical differences in some ways.

Yes, we effectively failed at doing our job. That doesn’t mean FLOPs are off the table. It’s not like we’re not doing them and instead are just gonna replenish with stability fees for the next year. I’m just saying, maybe we chill until the dust settles.

Traders are in panic mode right now. In a week or two, MKR could be valued on the open market at 3-400 and by rushing we just gave keepers a massive discount based on current market sentiment. I’m not sure from a business perspective this is a good move either.

I don’t have much to argue from a techincal point of view, just from a theoretical/probabilistic one.

1 Like

If you lower the minimum bid enough, the auction should be assured to clear the bad debt just from indicative interest from paradigm dharma etc. Who’s to say that they have not already taken the dai for their bids out of circulation? I’m guessing that some of the 5MM supply contraction has effectively already taken place

I’m sure a lot/most have (which probably contributes to the current peg problem). I’d also like to argue that maybe it’s not a great idea to maximally dilute MKR governance value at this juncture. Thats just my opinion.

Most people already made up their minds that this was happening regardless so my arguments are a little late to the party but I wanted to air them out for posterity anyways. Perhaps there isn’t another crash and/or USDC is enough to solve our liquidity problems and FLOP bids come in at a reasonable price. We don’t know yet. I just want to make everyone perfectly clear whats possibly at stake here.

I think @Aaron_Bartsch is making great points, but we are all speculating on the market at this point.

Specifically the point on:

Signalling a delay would allow the 5M that keepers are trying to accumulate to go back into the market, allowing the peg to drift back down.

I wouldn’t bet that market participants will just dump the pent up DAI back onto the market. It is more complicated than that because 1) we will have USDC collateral to factor in, 2) wouldn’t the same keepers go back to the market to hoard DAI again as the VWAP got closer to 1.01 again (when the delay would closer to be called off), and 3) if governance reneges on the MakerDAO announcement that we were holding these auctions to assuage collateral holders, would we more participants just defect from DAI altogether. My main point is that there are too many factors at this point for us to assume market behavior.

Discussing the parameters changes has been good and ultimately makes us think about plan B and Cs for what happens if the flop auctions start to fail. I am not as bullish about an option that could potentially delay the auction indefinitely and remove a key component of the system <2 days after MakerDAO publicly touted it.


Thats a great point Rich. I guess it’s too late now. We should have signalled a delay sooner so keepers wouldn’t start accumulating yet. That would have been the prescient thing to do. Damn.


So much DAI is still locked in DSR, I think this may be a key issue that will need to be addressed if an initial auction fails. How do we get it out of DSR and into liquid supply if 0% rates and a 2-10% premium to USD won’t do it?

I think this is just a matter of time. It really hasn’t been that long since the DSR was reduced to zero. If you look at the amount of Dai locked in the DSR it has been decreasing rapidly already. There are some good charts on MakerBurn that show this.

Liquidity should improve if the rate of Dai exiting the DSR continues to be high over the coming days.

1 Like

Did Maker Foundation try to find a buyer of a large quantity of MKR in the recent weeks? Did any potential buyer contact them? If someone like Dragonfly or a16z wants to buy a large quantitiy of MKR - now they can get a good price.

OK, I agree that it’s too late for the auction delay. If there is some <$120 MKR after the auction - I’ll try to get some.

Having listened to the conversation about the MKR auctions has left me frustrated. First of all I am appalled by the disregard of the best price signal on MKR that we have - which is the market price.

There are arguments back and forth about the liquidity being so low that the market price does not matter, this is complete BS. The MKR token is presently on a whole range of exchanges, although liquidity is indeed low.

Any talk about lowering the auction price to 150 or even 100 Dai is just talking down the market. So what about the range of community members willing to buy at 100 Dai? So what? All they are saying is: Please give me MKR at half the price. They gave a completely non-binding indication they are willing to buy at 100, they gave no indication they would refuse to buy at the market price.

People would do well to remember that the auctions are highly exclusive to begin.
To participate you need:

  1. coding skills to make the bot work
  2. a full ethereum node
  3. minimum 50,000 Dai ready

These requirements are the limiting factors for participation in the auctions, not the auction price of MKR.


@Planet_X Yeah, luckily defiexplore/defisaver may have a UI ready for the auction.

Also this is why I have an issue with lowering the TAU too much. We want a competitive auction, give people time to bid.


I think that @Planet_X makes a good point about the market price being a good indicator during normal times. That being said, MKR/DAI has already slipped below $200 during this past week according to some market prices, so our current starting bid wouldn’t clear.

I would consider the following 1) what is the cost of failed auction (high given this is a remedy for a previously messy auction)?, 2) there is already a higher cost of entry given that you participation prerequisites that @Planet_X mentioned (so we’ll have slippage given that everyone can’t participate), and 3) it is simply too late to technically remove the limiting factors mentioned (we’d all agree to do so if we could).

Any of these adjustments aren’t a silver bullet the larger problem; I can see how it can be frustrating. However they may mitigate any further breakdown of the collateral that we’ve tried to auction and slowly restore health in overall system.

Edit: Polls are now closed
I will be closing these polls in approximately 1 hour (on the hour: 19:00 UTC.) If you wish to signal please do so before then.