Vault Compensation Plan v2

The first Vault Compensation Plan failed to pass by a vote of 25k against to 17k in favor. This revised compensation plan incorporates feedback received in the forums since the first vote, and seeks to make the process simpler while avoiding impacting DAI liquidity. A google docs version of the revised plan v2 is available here. I and the rest of the working group would greatly appreciate any feedback!


Compensation Plan v2

Pursuant to the on-chain “Vault Compensation poll” 62k yes to 33k no that passed 4/6/2020, we hereby submit a compensation plan for approval by MKR holders.

MKR Holder Declaration

This compensation plan constitutes a proposal for a compromise and settlement of claims from vault owners in respect of the circumstances underlying the vault compensation poll referenced above. No action taken by the persons drafting or implementing this compensation plan, either previously or in connection with this compensation plan shall be deemed or construed to be: (a) an admission of the truth or falsity of any claims heretofore made or (b) an acknowledgment or admission by any such persons of any fault or liability whatsoever to vault owners or to any third party. The present loss compensation plan is without prejudice to any rights or interests of the individuals drafting or executing this compensation plan and any position that any such person may take in legal or administrative proceedings and addresses certain specific circumstances where MKR holders may volunteer to provide certain compensation to vault holders. The persons drafting or implementing this compensation plan maintain that MKR holders have no obligation to issue any compensation for losses, and volunteer to implement this plan solely at their discretion.

Goals and principles

  • Completeness - Ideally this process should address all outstanding claims for compensation.
  • Fairness - Similarly situated users should receive similar levels of compensation.
  • Consensus - Giving MKR voters the ability to choose key parameters will help ensure that the final executive vote has the support necessary to pass.
  • Simplicity - Minimizing the governance overhead and time involved in agreeing on and executing the plan.
  • Economic Soundness - Using MKR for compensation reduces pressure on DAI liquidity and the peg, while potentially minimizing impact of collateral price movements for vaults receiving compensation

Overview of compensation process

Vaults that were liquidated at auction or self-liquidated their positions between 3/12/2020 12:00 UTC and 3/13/2020 13:40 UTC will be considered for compensation; liquidations falling outside of this time period do not qualify. Compensation amounts for ETH vaults that were liquidated at auction shall be determined based upon the “tab change in relation to collateral change” calculation method, and the compensation percentage chosen by MKR governance. Compensation will be distributed in the form of newly minted MKR, with the DAI to MKR conversion ratio fixed to the volume weighted average price of the FLOP auctions that took place in March. Once the compensation amount is determined, a smart contract will be created to administer the funds transfers. Vault owners not included in the initial compensation class (BAT vaults liquidated at auction and ETH or BAT vaults that self-liquidated) will have the opportunity to provide documentation to substantiate their claim, and receive compensation at the rate determined by governance. Vault holders shall be required to issue a transaction to withdraw compensation to their account, and compensation will only be made available to the address associated with the vault.

Changes from Previous Plan Proposal

The initial Vault Compensation Plan Approval Poll failed by a vote of 25k to 17k. Some of the feedback received in the forums following the vote suggested that the amount of governance overhead and number of votes involved may have been a negative for some voters. This revised plan document seeks to slip down the governance voting process in the hopes that this may gain more support. Additionally, there is concern that auctioning MKR to get the necessary DAI for compensation would have negative macroeconomic consequences. This revised plan proposes to compensate vaults in MKR. Vault owners can then hold the MKR or swap to their preferred asset without impacting DAI liquidity.

The specific changes:

  • Eliminating poll to confirm compensation method / reference price
    • Moving forward with the “Tab change in relation to collateral change” method simplifies the governance selection process
  • Combining the initial plan approval poll with the compensation percentage poll as a ranked choice vote
    • The poll structure should make it easier to differentiate between MKR voters who oppose a particular compensation plan and others who oppose any vault compensation
    • Poll also gives approval to include funding request in a subsequent executive vote
    • Scheduled for 1-2 week voting period instead of 3 days to allow greater turnout
  • Run funding for both operating expenses and compensation payments through a single executive vote
    • Drawback - some operating expenses (e.g. contract development) would need to be paid in arrears
  • Compensation will be distributed to vaults in MKR instead of DAI
    • The compensation amount will be the DAI value determined by the “tab change in relation to collateral change” compensation method, with the DAI to MKR exchange rate fixed as the volume weighted average price of the March FLOP auctions
    • Payment for project work will be issued in DAI from system surplus

Proposed Stages

Phase 0:

Deliverables:

  • Plan document posted to forum for discussion and presented on governance call
  • Need exact verbiage of on chain poll
  • Obtain quotes for work necessary for compensation disbursement:
    • Smart contract development and audit
    • Vault compensation analysis audit
    • Payment for working group time up to this point
    • Estimated cost for vault claims review and future working group time
  • Need total cost in MKR for each percentage option in poll (1), including ETH vault compensation, 50% extra for manual vault claims, and operating costs (in DAI)

Phase 1:

Governance poll:

(1) Vault Compensation Poll (ranked choice IRV voting)*

++Edit: Please refer to this post for an update to the compensation amounts listed below.

  1. 24.67% (theoretical maximum auction yield) 2880324DAI 11522MKR
  2. 21% 2402973DAI 9612MKR
  3. 18% 2023954DAI 8096MKR
  4. 15% 1652134DAI 6609MKR
  5. 12% 1298983DAI 5196MKR
  6. 9% 621152DAI 2485MKR
  7. 6% 305187DAI 1221MKR
  8. 3% 101715DAI 407MKR
  9. 0% (Reject Vault Compensation)
  10. Reject Plan (Support Vault Compensation)
  • DAI/MKR estimates above assume 66% participation and a conversion rate of 250DAI/MKR… Actual compensation may be .5 to 1.5x the above current numbers (when one includes BAT vaults and self liquidators and an unknown participation rate) but should not be outside of these limits unless participation is unusually high or low and the numbers of vaults seeking to be included is significantly higher than expected. The probability the actual amount or compensation needed is outside of these estimates is estimated to be less than 1%.

Poll 1 selects the compensation percentage to be applied to vaults. For options (a) through (h), the poll vote also approves including a request for MKR funding into a subsequent executive vote to cover compensation amounts as well as DAI disbursement to pay for predefined operational expenses. Those who support vault compensation in principle but object to the specifics of the plan can vote for “Reject Plan”. Those who oppose compensation entirely can vote for “0% (Reject Vault Compensation)”. The poll will be a ranked choice vote

Deliverables:

  • Work on compensation contract development
  • Finalize manual claims processing procedure
  • Set up multisig to control contract admin functions
  • Audit contracts and compensation data
  • Prepare executive vote

Phase 2

Executive vote:

(2) Approve MKR funding for compensation claims and predefined DAI operating expenses

Deliverables:

  • Ensure broad communication of compensation package and claims procedures to affected vaults and wider Maker community, including any terms of accepting offer, deadlines, or other restrictions
    • Create and distribute a press release explaining the claims process for vault owners: vaults with collateral available to claim in Oasis, Black Thursday ETH auctions, Black Thursday BAT auctions and self-liquidated vaults
    • Coordinate comms with Maker community messaging team
    • Work with Maker UI providers to communicate claims process (Oasis, DefiSaver, etc)
  • Disburse payment for outstanding development and working group expenses
  • Receive manual claims for 30 days, and review within 60 days of receipt
    • Target for all claims to be completed within 90 days
    • Governance can retrieve unclaimed funds from contract after this period has elapsed

Manual compensation:

Claimants outside the initial compensation class can submit data to FORUM THREAD to substantiate their claim. For example, these claimants include

  • Vaults that self-liquidated
  • BAT vaults

The compensation review team (i.e., Maker_man, felix, jpritikin, monetsupply, befitsandpiper, and any additional members) will review these data. If the claim checks out then the compensation team will add the claimant to the compensation smart contract to permit funds disbursement. The compensation team reserves the right to unilaterally decide whether a claim is valid or invalid. Claims posted more than 30 days from the beginning of the compensation period will not be considered. No appeals shall be entertained. All decisions are final and at the discretion of the compensation team.

Phase 3:

Executive vote:

(3) Self-destruct the compensation contract and burn remaining unclaimed MKR

Compensation should be claimed within approximately 90 days. Once this period has elapsed, Maker governance is free to remove any unclaimed MKR from circulation.

END.

Disclaimer / Limit of liability:

The contents of this compensation plan document and any supporting information (“the Plan”) do not constitute financial, investment, legal, or tax advice. Information contained in the Plan is provided for general informational purposes only. Individuals involved in drafting or executing the Plan (“We”, “Us”) make no representation or warranty of any kind, express or implied, regarding the accuracy, reliability, validity, availability, or completeness of any information contained in the Plan.

To the maximum extent permitted by applicable law, We shall not be liable for damages of any kind resulting from any party’s use of or reliance on the Plan, or any actions or omissions of the Maker Ecosystem Growth Foundation and affiliates, the Dai Foundation and affiliates, the MakerDAO governance community, or any other third parties. By accepting the Plan, MKR holders agree to indemnify Us against any legal claims or torts resulting from actions taken to develop or execute the Plan such as assessment of individual vault claims.

Appendix: Icebox

Potential options for Maker governance not included in proposed governance plan.

In the event that the initial MKR poll to approve the compensation plan is rejected, the Maker community can signal which parts of the plan were contentious and potentially revisit some of the questions below.

  • Poll to confirm payment of compensation in MKR instead of DAI or ETH/BAT
    • A poll was developed to decide this question. However, during the drafting of this poll, it became clear that it is much simpler to compensate in DAI.
    • Proposing to move forward with compensation in MKR because Maker does not currently have the ability to acquire collateral assets. Maker voters can instead use the tab compensation calculation method, and a DAI/MKR price based on the average FLOP auction clearing price from March 2020
    • Deciding to compensate with ETH/BAT will likely add substantial delays to the compensation process
    • Compensating in DAI may negatively impact DAI liquidity and the peg
  • Poll to confirm compensation method and reference price
    • In the interest of simplicity, eliminating this poll and choosing to move forward with “tab change in relation to collateral change method” and conversion to MKR
  • Poll to determine/confirm the time bounds (start and end time) for vault auctions or self liquidations that will be covered by compensation plan
    • Proposing to move forward with vault inclusion criteria and start/end times from original Liquidation Report as discussed in the forums
    • Based on initial analysis, it looks like the Black Thursday period (3/12/2020 12:00 to 3/13/2020 13:40, ETH auctions 763-4324) covers the vast majority of users who lost funds, and covers substantially all of the zero bid auctions as well as the worst of the Black Thursday price falls and gas spikes
    • Possible counterargument: first and last 0% collateral return auctions near Black Thursday occurred at 3/12/20 8:00 UTC and 3/14/20 0:00 UTC, and last 0% collateral return auctions before auction freeze module and USDC were added occured by 3/16/20 13:00 UTC
    • Alternative time bounds would require additional analytical work, delaying issuance of compensation
    • Additional auction selection criterion:
      • < N% tab recovered,
      • Auction fetched < % some market price (only have OSM, uniswap markets, anything else need data which will delay compensation)
      • Collateral returned was < some chosen nominal %
  • Some process to select or approve the compensation review team
    • To speed up the process, we propose that the team will be formed by the same people who drafted the plan.
    • Some process can be added to the plan if the community feels like this is needed.
  • Appeal process for compensation review team decisions
    • To keep things simple, compensation review decisions shall be final and non-negotiable.
    • If the community feels like there is need for an appeal process, this can be added but will increase complexity, time, and cost required to administer the plan.
  • Poll to approve compensating vaults that self-liquidated
    • Proposing to move forward with “yes”, based on principles of fairness and completeness above
    • Users who took action to avoid liquidation likely saved MKR holders from additional bad debt, so it would seem unfair if these users ended up worse of than similar vault owners who allowed their positions to be liquidated
  • Poll to confirm using the same time bounds, reference price parameters, and compensation framework for self-liquidating vaults as for vaults that were liquidated by the system
    • Proposing to move forward with “yes” based on principle of equality
    • If poll resolves to “no”, follow up work/polling may be required to set reference price and time bounds for these vaults or identify alternative compensation heuristics
  • Poll to confirm mechanics of compensation for vaults that were liquidated at auction
    • Option 1: Uniform minimum collateral return percentage for vaults within time bounds (1) - topping up any vaults that received less than agreed standard (current compensation analysis covers this)
    • Option 2: Uniform premium/discount percentage versus VWAP of ETH/DAI or BAT/DAI
      • Need to determine the time used for valuing vault collateral (auctions don’t settle instantly)
      • Need to perform additional vault compensation analysis to determine overall and vault-by-vault compensation levels
      • If we do any compensation that has a time element to determine some value then someone will need to choose what this means for compensation. Example liquidations happen in phases (tend, dent, deal) and these phases usually happen in different blocks. Someone will have to define what time we use to define a market price for this case so a time based value choice can be applied to existing data.
    • Option 3: Tab based compensation. Apply the nominal collateral return to determine a collateral return deficit to multiply the tab in DAI to determine DAI compensation. This option is by far easiest to calculate over all possible claimants but it fails to compensate directly for collateral value (option 1)
    • Options 1 and 3 from this section have been incorporated into poll (2) to determine compensation method and price
  • Whether to suck the maximum possible compensation in a single executive or to spread it out over multiple executives
    • Spreading out the fund raising over multiple executives puts less pressure on the price of MKR from FLOP auctions. However, this approach has the danger that later tranches of funding may not be approved. Compensation funds could be exhausted and some otherwise eligible vault owners might not be compensated.
    • Therefore, we will suck enough DAI in a single executive to pay all potential claims. Surplus DAI will be returned to the protocol after a suitable waiting period.
  • Whether to wait until the surplus account can fully fund this compensation plan to avoid FLOP auctions.

Appendix: Smart Contract Specification

See https://github.com/jpritikin/BlackThurs

Appendix: Q&A

Q: Can we determine the asset reference price/compensation method and the auction refund percentage simultaneously?

A: This would be unwise. These polls are mutually dependent. Intermediate vote results on one poll would change the probabilities on the other poll.

Q: Could the compensation amount polls be run in the opposite order?

A: Yes, but the proposed order seems easier for MKR holders to decide. It may be hard to decide the auction refund percentage without knowing the reference price.

Q: Why wasn’t the asset reference price poll initiated on May 11?

A: Rich pointed out that signaling guidelines were not followed. Rich requested that more attention be put toward building consensus. For example, we should map out a complete plan before we run more polls.

Q: Why do we need a smart contract to convey compensation?

A: How else can it be done? Compensation will involve 100s of token transfers. It is not possible to batch all of these transfers into a single executive spell. A natural person could be nominated to issue the transactions, but what assurance do we have that the tokens will be transferred correctly? The total amount could be large. Who do we trust to handle that kind of cash?

Q: Is requiring users accept a release to claim compensation useful or legally valid?

A: We will draft the language as if it has legal force, but whether or not it has legal force is not important. The agreement is important because it educates that vault holders can either claim compensation or join as plaintiff in a legal action, but not both.

Q: Will vault holders be able to interact directly with the smart contract as well or will interactions need to take place through the GUI?

A: It is always possible to interact directly with a smart contract, but we will encourage invoking claim() from our static webpage that informs the claimant that claiming compensation causes them to waive the right to sue.

Q: What benefits do we obtain from using a smart contract and asking vault holders to claim compensation?

A: By requiring the vault holder to claim compensation, we ensure that they currently have an active wallet (that is, private key is not lost). In addition, we can stipulate that the act of claiming compensation waives the possibility of future litigation.

Q: How do we handle the final executive vote to fund compensation contract if MKR voters decide to use a reference price based on when the compensation is issued? Depending on how long vote takes to pass, there is the potential for the market price drifting away from the intended reference price.

A: The last opportunity to set the reference price is when the contract is created and address recorded for approval in the executive spell. That’s why particular language is used in the reference price poll.

Q: Where will the GUI to claim compensation be hosted?

A: On github as a static web page. If we can get a domainname like march13claim.makerdao.com to point to our static web page, that will be great, but it’s not required.

Q: What is the Tab compensation plan (option 2d)?

A: While going through claimants both within the BT plan proper and those who might be outside of the plan (self liquidators like vault 2288) we found that for most cases looking at the collateral change vs. the borrowed debt (TAB) change that it was easiest to value the damage in terms of the debt change as the Liquidation Ratio set by the borrower effectively determined the collateral value. In the loosest sense it makes more sense to value the compensation based on the amount borrowed since liquidation would occur based on the changing value of the collateral. The only change here is that we use the collateral return deficit to multiply the TAB change to determine compensation in DAI.

Q: Define the collateral return deficit?

A: If governance decides that compensation should be determined based on how much collateral was returned then one can uniformly apply this ‘nominal collateral return’ to the ‘actual collateral returned’ to create:

Collateral return deficit = nominal collateral return (chosen by governance) - collateral returned

Typically we look at this as a percentage because the theoretical return percentage is defined by the LR for the collateral type.

Q: Can you give examples of application of compensation calculation method

A: Using a nominal collateral return of 18% as a working example and value for poll option 2a at 100DAI/ETH, 2b at 240DAI/ETH, 2c = 170DAI/ETH

Example 1: Self liquidator Example 2: 0-bid liquidated

Vault 2288 3834

Before 5005ETH 100.1

After 585.18ETH 0

Collret 11.69% 0%

Tab 423835.88 9049.466

Colldef 18-11.69%=6.31% 18%

(.18-.1169)5005=315.8155ETH .18100.1=18.018ETH

(DAI) (.18-1169)423835.88DAI=26744.04DAI .189049.466DAI=1628.90DAI

Conversion to MKR at volume weighted average FLOP auction price from March

(using ~250 DAI per MKR as a placeholder value for this example)

Example 1 continued: Example 2 continued:

26744.04 / 250 = 106.976 MKR 1628.90 / 250 = 6.516 MKR

Appendix: Working Group Payment

Working Group rate of pay: 50 DAI per hour

Hours worked to date:

  • Joshua Pritikin – 40
  • Monet Supply – 40
  • Maker Man – 30
  • Felix – 4
  • Befitsandpiper –

Total outstanding working group time as of 6/15/2020: 114 hours

Total outstanding working group payment as of 6/15/2020: 5700 DAI

Future tasks requiring compensation:

  • Additional writing, analysis, and project management (working group)
    • Cost estimate: 3000 DAI
  • Create compensation smart contract (outside developer)
    • Cost estimate: 5000 DAI
  • Audit compensation smart contract (outside auditor)
    • Cost estimate: 5000 DAI
  • Review and process claims from BAT auctions and self-liquidated vaults (working group)
    • Cost estimate: 5000 DAI

Estimated total operating expenses: 23700 DAI

18 Likes

I want to credit @SolKo Maker Vault Compensation Updates and Status for the idea of compensating in MKR instead of DAI.

8 Likes

Hello All–is there any way we can do a call/meeting to go over Plan v2? a lot to digest, and might shed some light on what exactly Pv2 will accomplish.

I for one voted “yes” for Comp. Plan v1–which was against my better judgement. For one, as a non-believer of frivolous lawsuits, I am against the litigation brought against the Maker Foundation. This could have been settled IMO, within the community/ecosystem. Yet, I voted to give the CDP/Vault owners the benefit of the doubt, even tho in my heart I was against it.

Another thing–do we really believe the Whales/VC’s are going to want to put their name behind any of these Compensation Plans? As an American–I would not. IMO–the American based Whales/VC’s etc–will not get involved. But I could be wrong… but if I’m not–how do you rally enough Executive Votes to get it approved?

5 Likes

@Joshua_Pritikin @monet-supply @MakerMan @befitsandpiper
Thank you for all of the long and hard work you have put into this over the last few months!

Now, let’s get this plan finalized and voted through! What happened was tragic. I hope that this helps everyone put it behind them.

*Disclaimer, I @FourthStreet am not a part of the group seeking personal compensation. I support this plan though and will vote it through if given the chance.

8 Likes

Why is MKR compensation proposed at the 250 MKRDAI rate?

3 Likes

I really like this way of breaking it down. Seems like the most fair way to do it. I’d like to share some thoughts from a keeper’s perspective so we can better arrive at compensation price that is fair for both MKR holders and btfo moonbois.

I’m going to start from the top number, and work downwards.

  1. 24.67% rarely happens. The exception being self liquidation or low number of auctions for keepers to compete. Black Thursday had a huge number of auctions, so that puts a higher premium on being a keeper. In single collateral DAI this premium was fixed at 3%. Usually, 3% was enough to clear the auctions quickly, but in times of extreme volatility (such as the USDT scare in early fall of 2018) there is historical precedent of 3% being an insufficient incentive to clear the liquidations. Even if we had hundreds of keepers, I would argue that the normal 3% incentive would have been pushed to something like 5-6% do to the insane volume of liquidations to process. So our bid is already down to 5-6% below spot.

  2. There was a supply glut of DAI at the time. As a keeper, I need to be able to instantly arbitrage my profits to get the DAI back. After all, ETH could go to 20. Unlike the moonbois, I realize that and must derisk to get back into DAI. DAI was trading at +1.07 when I needed it most, so after slippage this is more like 7-8% that the moonbois will be paying for. Add this to the “keeper incentive” I described above, that brings us to 12-14% below spot.

  3. ETH was dropping so fast. DAI was mooning harder than ever. As a keeper, I need to bid low in case ETH keeps tanking/DAI keeps mooning by the time I can recapitalize. This is the hardest account to factor in, as each keeper has their own risk tolerance when it comes to exposure on the ETH/DAI pairing. I personally put this at additional 4-9% “keeper ratio risk” which brings us to 16-23% below spot.

I hope this makes sense. Even if we had a flawless auction system, the bids still would have come in 16-23% below spot price.

As first choice, I will be voting for:

Second choice:

Third choice:

Anything higher than option 6 is completely unjustified.

6 Likes

It was meant to be an approximation of the average price of MKR in the flop auctions from March (sort of approximating if Maker has paid out at the time).

1 Like

We will have a chance to discuss this plan in the next governance and risk call.

1 Like

Why is using the average MKR price from the March price auctions appropriate for this compensation plan?

3 Likes

I didn’t choose this price but if we want to debate this I am happy to step up here because while I didn’t come up with it I agree with the choice.

The simplest way I can put this is with an example.

Lets use vault2288

Guy self liquidated 4419.82ETH saving the system from another 400K DAI of 0-bid losses if he let his vault liquidate that would have had to be made up by subsequent MKR auctions at 250DAI/MKR approximately which would have cost the protocol an additional 1600 MKR!!

He recovered 11.69% collateral while also saving the system 1600MKR or 400K DAI of losses. If governance chooses 3,6,9% this guy gets no compensation.

In the compensation plan example we assumed governance chose 18% for illustration purposes and this would have given this vault holder 26744DAI compensation or if we look at this in ETH it would have been 315.8155ETH.

That ETH today would be worth approximately 126,000 US or about 124K DAI

Using the current model if we took 18% in his DAI tab and paid out in MKR at 250DAI/MKR we’d have 26744/250=106.976MKR (as in the example). Right now this would fetch about 160ETH or only about 1/2 of the ETH compensation to make 18%. So basically this person (even at 18% selection) would get ETH value return of about 15% (or about 3.5% more than the 11.69% he already has) and this is ONLY if governance chooses 18%.

IF we had done this at current MKR Market prices this compensation would have come out at 26744/580=46MKR approx 70ETH or a real collateral return of about 13.1%. a far cry from the 18% collateral return and significantly lower than the 15% using 250DAI/MKR

So the reason why we would use the 250DAI/MKR price is two fold. In the end when used against the compensation % being in the DAI tab vs. collateral we make up for:

  1. not compensating in collateral (collateral with a DAI price at the time of liquidation is a far cry from the total value these vaults believe they have lost which is a value perceived in collateral that has since liquidation appreciated at least 3-4 fold) but also
  2. added compensation for having to wait for all of this and believing they probably are going to get nothing while watching ETH price climb 4 fold.

The goal here is to at least try to fairly offer compensation based on the percieved value lost even if it is going to come up significantly short. Meaning if governance were to decide that 9% or 12% collateral value should have been returned and uses the above method to make up the shortfall NO VAULT will get back to the 9 or 12% of their vault collateral and hence also will have lost out on the opportunity cost of that collateral value increasing.

The simplest way to put an answer to your question is that even at 250DAI/MKR and current MKR/ETH ratio this model is not going to make up the whole collateral difference even if 24.67% is chosen simply because ETH price has moved up so much since this event and using market price of MKR in DAI pretty much makes compensation at the 24.67% level easily refusable by vault holders. Even at 24.67% and a MKRDAI price of 580 (.2467-.1169)*423835.88 = 55014DAI approx 95MKR = 142 ETH = 14.5%-11.69%=2.8% collateral compensation vs. 580/250 = 2.32 or 2.8% * 2.32 approx 6.5% on top of his 11.69 to give about 18% collateral return vs. the 24.67% governance selected as a compensation choice.

Lastly governance gets to choose the compensation % value and given that most people (lix is a good example below) are going to want to return 3-9% using even 250DAI/MKR means only 500-2000 MKR will be offered in TOTAL to all of these vault holders. Better than nothing to be sure, but a literal drop in the bucket here both in terms of real MKR dilution (even relative to the what approx 15,000 MKR that was sold to raise the 5.3M DAI at the time) and the real value of the ETH these vaults lost based on current ETH USD prices.

One fact I want to point out. It was the self liquidators that basically saved the system from significant additional MKR losses. Guys like vault2288 should be getting a bonus compensation based on how much DAI losses they saved the system to reward people who were in great haste and with strong emotion were clearing their collateral to pay down debt during distressed markets and high gas prices, to try not just to save something for themselves, but also to save the system from literally giving away their collateral and eating greater losses. They should not be getting punished because they salvaged collateral only to find because they didn’t lose that much they get less compensation in a compensation plan than someone who did nothing. Vault2288 I consider a Maker BT hero because of this. He alone saved Maker in total as much MKR as will be offered to ALL vaults at the 6% level by self liquidating and almost 10% of the amount if 24.67% were offered.

3 Likes

Interesting @MakerMan. It might make more sense to use a higher MKR/DAI price. For example if it was 500 MKR/DAI, mkr holders would be more comfortable choosing a higher number. Making the distribution to more people, more fair?

I’m not sure if MKR holders are willing to do this. Well, we never put to vote whether they wanted to compensate at current ETH prices or the black thursday eth price.

Well in a sense you are right that compensation spreads farther with a higher MKR/DAI price but in effect those that receive compensation will get less.

Well everything is up for discussion and debate. I totally encourage any and all feed back.

You are completely correct in that any polls regarding pricing of collateral to determine compensation was left off the plate for the sake of trying to simplify this massively into a single poll and to provide a more realistic alternative to compensate in something governance can obtain (MKR) vs. DAI (which would be terrible to markets trying to raise DAI by selling MKR) or ETH (worse selling MKR for DAI then buying ETH). The idea of compensating in MKR was brought up by vault owners and honestly makes much more sense generally than taking MKR and converting it into DAI or collateral.

The hardest part of working on this plan is that while we have some amount of input regarding the plan from forum people like yourself. Given the lack of participation in the last poll I really don’t think the compensation group has any real guidance on how to proceed to lay this to rest one way or the other. We literally are making our best guess on what forum commentary we did get.

Honestly depending on what the next poll brings (if we get a lack luster no again here) I will probably have run out of options and personally will find it much easier after all of the work we have done to leave this issue as dead via lack of governance participation.

IF there is one thing I want to see from this vote it is participation.

I’d like to think I am speaking for the entire compensation group here but am speaking for myself. In this current poll I want to see a definitive YES or NO with an executive level of MKR participating in this decision one way or another. No matter how the vote turns I’d love to actually see MORE MKR participatig in this one decision than in any other executive ever to lay this to rest, or to move it forward. This isn’t about whether Maker compensates vault owners. This is about whether Maker holders feel settling this issue one way or another IS IMPORTANT not just to them but also to vault holders and the entire ecosystem once and for all.

3 Likes

Please explain this number. Not everybody is a keeper.

The principle is understandable, but the numbers are not self explaining. You have to sell your solution to the MKR holders/voters. Do not forget, that this will be also in the news. And you do not want to embarrass the Maker protocol and community.

So it needs to be understandable from the first look. As a first one - The DAI value does not matter. We users/bulls do not understand DAI/USD value. We understand ETH value. We know how many ETH we had and we do want to know how many ETH will be compensated.

Here is an easy to understand example:

The Poll should have 6 options:

1) Full compensation in MKR with a fixed exchange rate (statistical average since march 12 2020 up to the executive vote) for ETH and BAT.

(Collateral in BAT or ETH) times (24.67%) minus (stability fee);

2) 80% compensation of 1);
3) 70% compensation of 1);
4) 60% compensation of 1);
5) 50% compensation of 1);
6) do not compensate with MKR (other solution needed).

The math for maximum compensation might be boggy (i do not know - Working group numbers are different), but that is easily to fix.

The fixed exchange rate of MKR to ETH and BAT can be defined as the statistical average from march 12 up to the executive vote. Ad up all average prices from every day since march 12 and divide by the total number of days.

These numbers should not be there. The 0% vote amounts to public sabotage. The go ahead for compensation was already decided.

Self-Sabotage
https://vote.makerdao.com/polling-proposal/qmwfvvguaf8rz8xwgv2cqnzzt9t5h6epzh17qmk2ue99y4

Going back on MKR voter word is very, very, very bad publicity.

The least compensation should be 50/50. A fair split of the damages between users and MKR holders.

MakerDAO has to be accountable for the technical risks. The users have to bare the market risks.

0 bid was technical.

The market was also in a big dip.

Is this your vault?

To be clear! There are no heroes. Some did know more (like vault 2288), others did know less.

1 Like

No

I am entitled to an opinion that you are entitled to disagree with. One could easily turn this around and say everyone that was liquidated pretty much would have had no problem if they maintained their CR above the LR and liquidations were never guaranteed to return collateral. Even now this is true and even with the new liquidation system about to come on line.

I do try to stay more in the middle ground so thanks for reminding me. Ok no heros here.

Ofc which is why in the plan we explain the reasoning behind the numbers:

So the % compensation is based on a difference between how much collateral a vault retained subtracted off from whatever % is decided in the poll and this percentage difference is then applied to the DAI tab at the time of liquidation to determine the compensation value in DAI which is then used with an average MKR auction price during the week to detemine the amount of MKR they would receive.

From my liquidation report:

Collateral Return Percent = (LR – (1+LF))/LR
• LR = Liquidation Ratio = 150% = 1.5
• LF = Liquidation Fee = 13% = .13
• Collateral Return Percent for above =(1.5-(1+.13))/1.5 = 24.67%

You then kind of ‘go off’ with the following statements:

and then propose your own version.

This is the place to discuss suggestions you do not define your terms.

stastical average of what exchange value? average based on what criterion of what dataset of market prices? The compensation group decided that it was easier and in fact gave similar results in terms of compensation with the simpler ‘DAI tab change vs. vault collateral change’ method as the primary formula that could be universally applied to all vaults along with the compensation percentage as a simple easy to measure and implement method of compensation.

If you are going to come up with a different method that requires data that we don’t have on hand at the moment, will you also provide this data and come up with real estimates of compensation based on the model you chose? (i.e. produce an alternate plan to consider?)

I am pretty much going to stop there except to add a pretty basic comment regarding the agressive approach in your replies here. I can be pretty strong at times so I tend to be more forgiving of people who have strong emotions about what they are writing or speaking about. Being passionate about things is ok. So I simply make the observation and follow up with a suggestion. tone down the ‘bold’ and try to avoid ‘should not’ and replace with ‘might be better if’, “I suggest the following:” or some variation of a reply that is suggestive of positive changes to the plan (or even an alternate plan) that seem less confrontational and more helpful.

Just a suggestion.

I am all for someone, anyone proposing another version of a compensation plan. Define your terms, provide some data and another poll that has some MKR numbers. You say 25.42% - stability fee. Actually there is a Liquidation fee of 13% or are you really talking about the SF at the time? You also say:

and

Which I think based on the above previous comment means by your own admission your plan is a bit boggy mathwise.

The devil is always in the details which our plan happens to provide (I/we did our best to provide). If you don’t like us listing the DAI estimate we used to calculate the Maker estimate - ignore it and focus on the MKR estimate.

As to the 0% vote being

Given we have two conflicting polls on whether to compensate that occured over 3 months. I think having a valid Reject option in the current plan, as well as one that says ‘I want some other plan’ is important especially given the conflicting results in the two polls we have run so far. I think the compensation group also doesn’t want to limit what governance can say here. I would like to hear from others before making my own decision on whether to alter to poll or plan based on discussion here.

This is why we present these things - to get public comment. So thank you for your comments!

3 Likes

lix my friend. Thank you for chiming in with an elaborate explanation of your reasoning. I think a lot of people who would be yes to this are coming up in the 6-12% range for the same reasons.

My only point and pitch here is remember we are taking the difference in collateral between before and after whatever happened to a vault against the voting % compensation choice and applying this to the DAI tab change at the time. Choosing 6% as you have above will not make 0-bid vaults that got no collateral back whole to 6% of their collateral (ETH say) mostly because of 4x in ETH value over their DAI tab change. In the end even after the conversion from DAI to MKR @ 250DAI/MKR we come up short on the 6% returned collateral (i.e. if the vault owner took the MKR and traded for collateral/ETH).

The point with the method and these numbers was to find a consistent between all vaults ‘fair’ way of distributing however much MKR is associated with these voting choices between the vaults that is easily ‘calculable’ with on-chain data. It was not to actually bring these vaults up to the chosen collateral return % being chosen because literally this is now much more expensive with ETH at 400USD vs. the auction at the time 100-150USD. But using MKR valued at 250 vs current 560-580 does seem to help bring this closer than using any other metric (which will be significantly more complicated to calculate and require more polls) to some reasonably fair value that then is pretty well bracketed by the estimates given.

I will have to look back at this but I think I have 130K ETH liquidated for all these vaults during BT (not including the self liquidators, and BAT group). So if we think we would give back 6% ETH on that 130K liquidated we’d end up at 7800 ETH and with MKR/ETH trading at 1.5 here this would mean 5200 MKR. The 6% above will return about 1221 MKR or only 1/4 of the 6% of all ETH liquidated (so about 1.5% of the USD ETH value via MKR).

If one really wanted to give back something like that 6% but spread it over vaults you would pick the 12% option as this would pay back 5196MKR spread over all vaults that got 12% or less of their collateral.

Realize I am not trying to make you choose 12% I am just trying to explain the reality if you do choose 6% in terms of how much ETH a person would be getting back via the amount of MKR being distributed (1221 MKR or only about 1.5% give or take of their collateral auctioned depending on if they received any collateral back between the 0 and 6% range).

Even at the theoretical maximum 24.67% at 11522MKR or 17283ETH (1.5 ratio chosen) this 130K ETH would only be getting back 13.2% collateral value as MKR at current collateral valuations using the model we have chosen.

1 Like

You are right.

I calculated by hand and rounded some numbers at first.

Nice to see, that you are not calculating by hand :sweat_smile: :sweat_smile: :sweat_smile:

Take This data put it into an EXCEL table and evola you get

My intention here is to help reduce the steps needed to resolve the situation.

It would be nice if the solution would be obvious, fair and well presented.

I hold some MKR myself. So i am biased for MAKER not to get bad publicity.

To me, it makes more sense to give them the USD value of what they would have gotten back at the time of the liquidation end in DAI. That way we can just take it out of system surplus instead of having an auction of minted MKR. I don’t see how the current ETH or MKR prices are relevant at all in the discussion.

4 Likes

I agree to draw DAI compensation from system surplus instead of using MKR.

Dai would have been valid in my opinion if you had compensated immediately. I wonder what thoughts would be if price continued to plummet. Probably comp in Eth?

Compensation in dai now will be nearly as offensive as a 0% vote. I dont think this option should be included at all personally. The vote has already passed. 0% compensation is not compensation.

Waiting 6 months for that result will break people

1 Like