On-chain vault compensation poll, draft

There was pretty significant support for lower compensation on the original polls. Something like 56% of people showed some support for the < “high compensation” options. Should there be an option that captures compensation should be < “estimated remaining collateral at the time of liquidation”

Current options seem pretty binary. No compensation or 100% compensation. There is no say 50% option.

Might not even need to specify 50%. You could probably word it something like: “Compensate less than 100% of the estimated remaining collateral value at time of liquidation”. If that option were to win you could do further polls to figure out what the correct percentage should be.


My impression is that most people just wanted an accurate estimate of compensation. However, I might have misjudged sentiment. In any case, I don’t think we can ask both questions in the same poll. I’ll adjust the wording and

I think it would have to be a separate poll since we don’t have rank-choice voting yet. We can’t present too many options per poll.

1 Like

I think it would be wrong to compensate vaults liquidated at 0 unless we also compensate vaults liquidated at less than 70-80% market price, as the impact on vault holders is the same.

And we shouldn’t compensate any of the above vaults without also compensating vault holders that paid back debt or deposited additional collateral on Black Friday to avoid liquidation.

If we’re going to do this, we should be consistent, and not have tunnel vision about only looking at 0 DAI liquidations.


Agreed. I have added language to the draft in response to your comment. Does it look okay?

1 Like

No it is not. MKR token is here to heal system debt, vaults that got liquidated above the liquidation threshold did not cause system debt and should not be covered by MKR token holders. At least that was my understanding when coming on board (read: purchased MKR). Where would it stop then? Will MKR holders then become insurance providers for all sorts of Maker related incidents?

Why is no one, I repeat, not one person, arguing that, if the community decides for some kind of compensation scenario, that the Maker Foundation pays that? Multiple projects have set precedents for similar incidents and the Maker Foundation is sitting on a treasure chest of money - why is this such a taboo to talk about?


Absolutely. I can vouch for this as somebody who was forced to close my vault in a panic to avoid 0 liquidation and salvage what I was able to.

I believe if compensation is offered to those who were liquidated to 0 - those like me should be compensated in equal proportion so that the difference for all effected is made up fairly and evenly. See below for details:

Again I say this because if say Vault#XXXX gets back what they should have under a set of liquidation parameters -


Work out the ETH-USD price that is used to compensate vaults in a certain time period.

  1. 100 USD for 1 ETH + 13 % penalty
  2. 90 USD for 1 ETH + 13% penalty
  3. 80 … etc

Then those same liquidation parameters should be in fairness applied to those also effected so that everyone gets the same level of outcome.

1 Like

Non of these points are valid.

Maker is not an “experimental crypto project” targeting programmers, its trying to get mainstream adoption.

Maker was not expecting users to run keeper bots, no where on any page is it explained when taking out a cdp. If you were expecting why its it mentioned on Oasis when taking out a cdp?

My ratio was always in green or orange. I did agree to keep it collateralised and maker agreed to take my eth plus 13% to pay back my loan not all my eth.

No one is complaining about the 13%.

No one is complaining about this.

Yes, it will force Maker not to make such mistakes in the future and take some responsibility in the way you market your products going forward.

Again, as i said before the question should be how to compensate not if. Non of these points are valid or fair.


This is straying a little bit from the point of the thread. I’d love to hear people’s thoughts on proposed verbiage for on chain poll.


Personally I want to see at least 0, 50, 100% payment for exposure (maybe 0, 25, 50, 75, 100 are better valid options. As I have made clear it has always been my own personal policy to ‘split the difference’ between what people thought they should get and what I thought they should get when there was an argument/discrepancy.

One other point I am not sure if relevant. Acceptance of the offer meant they simultaneously signed off on any litigation and negative PR. The negative PR we can’t control but the litigation we can. Simply because any litigation would have to involve the wallet ID and proof of ownership. IF they accepted the payout - they accepted the terms that went along with it. If they return the payout they don’t accept the payout nor the terms and can do whatever they want.

The whole problem with this vote is that until we have real numbers to work off of we don’t know what the percents means other than 0 (no compensation).


Others have already asked for this, but it would need to be a second poll.

I like your idea, but I’m not sure how to it could be made legally binding.

1 Like

Please take a look at the paragraph I added at the end of the poll draft.

1 Like

This is the most important point, no one in Maker is able to reasonably make a decision on this issue without the numbers being completely worked out… which is complex for the reason everyone have pointed to,people paying back debt, non quite zero bids, the market volatility itself, the state of ethereum (if this was a mempool attack what will that mean, should EF compensate…?), ect…

1 Like

Its not about the risk of leverage being disclosed or not. Its that Maker markets the product, its a completely novel system that we don’t know the behavior of and hasn’t even been stress tested like SCD. Acting like we were ready to reinvent finance as soon as MCD launched is an absurd perspective in context of the systems immaturity. We need to make it clear that EVERYONE could lose EVERYTHING at anytime (including mkr holders) (this is possible in multiple ways, successful governance, oracle attack, mixture of other complicating events, failed ES ect ect). Telling people anything else is a straight up lie. That reality isn’t restricted just to Maker, but crypto in general (and the whole universe for that matter… nothing is guaranteed).

The point of scientific governance is to understand the complexities, design things to mitigate the risks involved with those complexities, and to be able to properly explain how we came to those designs, so other knowledgeable people can judge, and critique them.


Detailed P&L analysis, https://ournetwork.substack.com/p/our-network-issue-14

How much is potentially owed to vault holders? My impression was that it was 1-3 million DAI, but these numbers seem much higher. Do I understand correctly?

Hey. Those Vault losses I presented are not truly representative for compensation purposes. Vault loss that I calculated also includes a penalty fee and ETH price decrease during auction time for particular vault. For compensation calcs, you need to exclude those two effects plus of course assume how high bids would be during that time.

Finding a decent benchmark to compare to is going to involve a lot of bias though, because we haven’t had such high liquidation volume event in the past, accompanied by lowest DAI liquidity and highest ETH volatility seen during Maker existence.

1 Like

Is having an estimate for the impact a blocker for the purposes of this on-chain poll?

Can accurately quantifying the impact be punted for purposes of this poll? It seems that if we could just drop the estimate in the draft, or just link to a few sources for estimates if we deem that necessary. There is already some language in here about doing further analysis maybe we can sort this out then.

No, I don’t think so unless somebody volunteers to do the work.

It would be nice if we had a tighter upper bound though.

A couple of notes about this section:

My number one piece of feedback is that the overall tone is a bit combative. It might be better to use language with a little less implied bias.

Most important thing I would suggest would be to avoid us vs them language wherever possible.

Just looking at this point:

I think you could probably reword that as something to the effect of:

“Vaults were liquidated because they failed to meet the collateralization requirements of the protocol. By opening a vault users implicitly agreed to maintain an adequate collateralization ratio.”

I think that gets the point across without all the finger pointing.

There are several points in the “arguments against” section about not adequately understanding the risk of the system. Not sure they need to be called out separately.


I think all of those arguments ultimately distill down to vault users are expected to understand the ethereum ecosystem. You could probably say something like:

“Vault users are expected to understand the ethereum ecosystem and accept the liabilities associated with using that system. All contract code is open sourced and available to be audited. As such, by using the protocol it is understood that the users are adequately informed and accept the risk of their actions.”


For this idk that we need to be that explicit. If and when analysis is done we can figure out some way to ratify it’s findings


Primož’s numbers assume keepers paying near oracle prices which is unreasonable during market crash (he was not calculating compensation numbers, just profits/loses by ecosystem actors AFTER the event) At one point price went from 170$ to 140$ in less than an hour. Rational keepers would bid very conservatively in such environment. They would need to sell eth very quickly (pay high gas fees, find best market etc…). All this would need to be coded in advance (detection of market crash, strategy during market crash, etc…). It’s the nature of eth collateral not having some ‘fundamental value’ that complicates things for keepers.

If i would code my keeper, i would immediately shutdown (or go to special mode) in such conditions,
send notifications (possibly to wake me up). It is huge risk/reward situation for keepers.

1 Like

@Andy_McCall I have mostly adopted your suggestions. Please review the latest draft.

Thank you!