[Discussion] Increasing Surplus Buffer

thanks @Primoz for clarifying - so the main reason to raise SB is the 10xed debt exposure and not the risk of 101% LR vaults going undercollateralized. will vote for yes given that :slight_smile:

is there any thread i haven’t found yet where the latter problem is dicusses, especially the crazy idea of @monet-supply (which i like a lot)?


I have another potential solution to the stablecoin Vault issue.

Basic idea: Autoliquidation of Vaults after a certain time.

Let’s face it, the current setup we have going with stablecoin Vaults is a temporary measure enacted to keep the peg following Black Thursday events and the DeFi craze. It is far from optimal, possibly not even sustainable, so let us change how it works.

The proposed action sequence:

  1. For existing stablecoin Vault types (USDC A etc etc) we introduce upwards creeping liquidation ratios, the suggested pace is 0.25% increase per week until approximately 103% is reached. The increase is to ensure that if liquidations happen they do not eat into Maker buffers. Exact figures will of course be as per the Risk team’s recommendation. This slow pace of increase will allow Vault owners time to close maxed-out Vaults.
  2. We reduce the Debt Ceiling to 0 for the existing stablecoin Vault types, effectively ensuring no new Vaults of this type.
  3. At the same time we open a new Vault type especially for stablecoins. It works as before but with the added feature that the Vault will be auto-liquidated after a certain amount of time (blocks) has passed. How long? Some months to a year to start, but these are suggestions. This fully eliminates the potential issue of unbacked DAI in buffer/burner. As the new Vaults will be auto-liquidated anyway we could have a spectrum of Vault types, some with extremely low liquidation ratio or other experimentation to help with the peg.
  4. After the new Vault type is introduced we keep raising both the stability fee and the liquidation ratio for the older stablecoin Vaults, creating clear incentive to migrate to the new auto-liquidating Vault type.

Not everyone is going to like this proposal. It is tough love on the stablecoin Vault holders and it forces them to be more proactive with their Vaults. But for Maker it solves the potential issue of unbacked Dai in the buffers and it allows for more creativity with regards to Vault types as they will be closed after some time anyway.

Feedback appreciated. EDIT: grammar

it feels a bit too harsh from my PoV. while i would love to see DAI working (especially w.r.t keeping the peg) without having any centralized collateral, we would be in a far worse position if we wouldn’t have onboarded USDC and others

i don’t think we need to grind more fees out of the 101% vaults then necessary.

what is the advantage of your proposal vs just

  • set DC + SF to 0 for e.g. USDC-A as soon as some vaults hit 100% CR
  • create USDC-B with the same parameters as the old USDC-A
  • repeat

the latter approach is just some governance, your approach needs to be implemented (and we know of the scarce resource of dev-hours)

regarding “making sure we offboard all stablecoin vaults” (assuming this is even something we want to achieve): we can force this later too without any special code (just activating liquidation and setting up the SF again)

Not sure I get the idea. Most USDC-A could be liquidate already but we can’t because DAI is too expensive. It’s not that we are kind to our customers, is that we can’t do otherwise. Maybe I missed something.


Well I do not say it out loud but I do not expect the peg to be fixed as long as the DeFi craze makes people pay a premium for Dai. This proposal is not about the peg.

so this proposal is to make the stablecoin-vaults more and more irrelevant? don’t you think this will actually hurt the peg? wasn’t that the whole idea we onboarded them? :wink:

1 Like

The issue from what I can see is that Dai is over $1.01 and it needs to be lower in price so that these stablecoins can be removed from the system. The system currently is in the process of creating unbacked Dai.

Unbacked Dai lowers the Dai price in the market… Since this is what the desired outcome is, I see no problem with letting it persist and then inevitably self correct as the Dai price falls and then the coins get removed from the system. Once that occurs the price will naturally rise above $1.00 again because the system at that point will be backed properly again rinse and repeat.

1 Like

Where will keepers get the Dai to buy 400M worth of stablecoins if we liquidate them like you are suggesting?
I disagree with you on fixing the peg in this defi craze - it is possible. Actually, it is essential. Integrations have made it very clear that we need to be much closer to 1 than we have been the past 7 months.

1 Like

This is a slow process. Will take several months.

Even if it takes several months, liquidating the stablecoin vaults will put unnecessary upwards pressure on the peg. To get back to 1, we need people to sell between 1 - 1.01. If we liquidate the stablecoin vaults according to your strategy, then who will be motivated to sell below 1.01? Liquidating these vaults is the same as saying 1.01 is the new peg (like it has been +/- .15% for 4 weeks now).


I would be afraid to short Dai if I could reasonably be liquidated, because chances are I will, seeing how much trouble it was managing the peg last few months. Turning on liquidations is not too far off from removing USDC vault types altogether and will only make the peg problem worse. I still think we should explore QE so we can bring the peg below 1.01 and actually give stablecoin vault holders a chance to unwind.

1 Like

No this is not the case as we need to protect Maker from selling at a loss. There is always some slippage involved so there must be a safety factor built in.

I prefer the future where the protocol realizes gains instead of being concerned with selling at a loss. We can realize gains and get Dai to 1 if we leave some meat on the bone for the stablecoin vault owners.

1 Like

Best that these stablecoins are liquidated (just transferred really) into a MakerDao controlled account and then sold off when the Dai price is under $1.00.

1 Like

There is no need to clean the stablecoin library, let it slowly grow old until DAI <1. At a critical moment, we can let the old stable currency library SF = 0, and at the same time gradually reduce the DC, thereby forcing people to switch to the newly created new stable currency library B\D\C, etc., and reset the parameters.

1 Like

It’s quite similar then to the Quantitative Easing solution that governance voted against: [Signal Request] Quantitative easing to fix the peg (aka Manual PSM)

I suppose you are correct. I was unable to vote on that particular poll because I was away that week, but I would be in support of something being done along those lines should it come up in a future poll.