[Signal Request] Increase the Auction Surplus Buffer (System Surplus) - September, 2020

With fees finally accumulating, the auction surplus buffer (system surplus) is quickly (nice!) approaching the point where FLAP auctions will begin to occur. Therefore I would like to gather signal from the community as to whether they would like to increase the auction surplus buffer, hold off and allow FLAP auctions to occur, or a combination of both.

Pros of increasing buffer

  • More resources to spend on domain teams, treasury, system insurance, and other expenses.

Pros of allowing FLAP auctions

  • MKR burn will increase MKR price
  • Possibly boost morale for MKR holders having to endure since Black Thursday

If you have other pros please add in the discussion below.

Should we increase the Auction Surplus Buffer?

  • Yes, before FLAP auctions occur
  • Yes, after some FLAP auctions occur
  • No
  • Abstain

0 voters

If either "Yes" wins

  • Increase Surplus to 3 Million
  • Increase Surplus to 4 Million
  • Increase Surplus to 5 Million
  • Abstain

0 voters

If "Yes, after some FLAP auctions occur"

  • After 1 week of FLAP auctions
  • After 2 weeks of FLAP auctions
  • After 3 weeks of FLAP auctions
  • After 1 month of FLAP auctions
  • After more than 1 month of FLAP auctions
  • Abstain

0 voters

Polling will run until 2020-10-05T04:00:00Z or 40 participants have chosen a clear winner.

If a “Yes” is chosen, a pull request will be created to be added to on chain governance on Monday, October 12, 2020 and if passed, will proceed to an executive on October 16th, 2020 to be implemented.

If clear consensus is reached before Thursday, October 2nd, a pull request can be implemented to go on chain for Monday, October 5th instead.


I like it. I might offer a different approach to the “weeks of FLAP auctions” section. We should think about the relationship of DAI to MKR market cap (since MKR speculation is supporting this whole operation here). We could set a target ratio, eg 1:1 DAI minted to MKR market cap, and increase the buffer as soon as that cap is reached.

MKR Market cap: ~520 million
DAI minted: ~900 million

Burn MKR until the ratio is 1:1?


I like it. If “delay surplus buffer increase” wins we can have a separate signal for that idea, or you can do it in parallel if you wish.


Might be nice to get MKR supply back to 1 million to match initial distribution meme and break a psychological setback. Haven’t done the math on how long that would take.


Little over a month as per current rates

1 Like

The system accumulates more than 85,000 DAI per day, reaching 2 million by mid-October.

1 Like

Your pros and cons here are completely misleading. We do not need to increase the buffer to have more money to spend on domain teams, treasure, etc. MKR holders can simply vote to fund these activities directly by minting new MKR while the burn of MKR from the protocol income continues as designed. We don’t need to create a slush fund before we fund important priorities - all we need is for MKR to be valuable so that we can mint it and sell it for what the community decides is a priority. The hard part is getting consensus on exactly how to spend the money- I don’t think having a slush fund make this any easier.


You assume MKR holders want to dilute themselves to fund projects. I agree that these things aren’t completely contingent on the surplus but instead active revenues. The surplus ensures we have funds available in case of a failure of the system and stores active revenues which can be decided on by the DAO on how to be used. Having a “slush fund” definitely makes this easier as its predictable and tangible.


What am I missing here… We can easily mint new MKR and sell it for covering expenses while in parallel flap happily. Right?

1 Like

One on the key issues on why we should at some point increase the SB instead of burning maker is because of the CR of the stablecoins at 1.01.Please someone correct me if I’m wrong on this since it was discussed in the rocket chat.

At some point ( provided the peg does not drop below 1 so vault owners can liquidate and get the profit) we will stop making revenue on part of the stablecoin portfolio.

Since liquidations aren’t enabled all vaults at 1.01 that have been for more than 3 months will be undercollateralized therefore if the revenue were to be burnt we would have to mint maker to back up the undercollateralized debt.

That’s why it was also proposed to increase the SB because once we reach a certain threshold of CR of stablecoins any loss would be offset from the SB and we would still accrue income from new vaults.

Another alternative was to keep the burning while later in time dropping the SF of stable coins to cero. That would imply burning but leaving money on the table for all stablecoin vaults that have room to grow on the interest.

In summary my vote is more related to this point than domain teams or expenses (and also taking into account a month if not more of maker burning) those could be taken out of the SB and eventually if it were not enough increase it. If we are to use the funds from the SB we better make sure that they are not contigent and the best alternative I’ve heard so far is to increase the SB (once we stop burning maker if that is decided) keep the interest on stablecoins & keep track of underwater vaults since those funds on the SB are temporary pending write off of the vault.


Correct me if I’m wrong, but I was under the impression that a larger surplus would have allowed us to avoid the minting of additional MKR after Black Thursday, as the surplus might not have run out during the auctions.

If that is true, I would imagine the surplus should be some rough percentage of the total DAI supply. Maybe somewhere between 10-20% in order to account for black swan events. A Black Thursday with 50 billion DAI in circulation would be very different than the Black Thursday we experienced.


I agree. Still, from marketing perspective, i think it would be good to burn some MKR. Since (crypto) market is very speculative at this point, this might even push MKR price higher.

In the end, i think I’ve learned 2 lessons this year:

  • unexpected (rare?) events are probably much likely then we thought
  • we need to go faster (even when things seem to calm down) -> we should not optimize (discuss) solutions too much to find perfect solutions, but focus on moving forward first

Anyway, efficient MKR burning is quite unimportant at this point.

1 Like

I don’t think it’s a good idea to continue to increase the surplus buffer. As noted here, the surplus buffer belong to DAI holders in case of an ES. Assuming the farming stop and we go back to 400M DAI, a surplus buffer at 4M DAI would mean DAI = 1.01$. If DAI is below that, that incentivize triggering an ES (especially if a good amount of the collateral is stablecoins).

Having more DAI aside is also not helping the peg. Each more DAI is basically a new USDC in a vault. But I don’t see a way to solve that.

I’m not against the overall idea but I would put the surplus of the surplus buffer in a DAI holder remote wallet (like strategic reserves).

I also like the idea to burn enough to go back to 1M MKR.

Finally, most of the fees are from stablecoin vaults, so those will likely by cut down in 2 months. It’s not quite clear that this income is real (as we have no solution to liquidate yet). It could soon become an expense. We might want to avoid to burn virtual money.


I voted to stop accumulating DAI. Unless we have a detailed plan about what to do with the surplus, who will do it and how much money is needed - I think its best to stick to the original plan and burn MKR.

Straying from the original plan should at least require a separate (mid-week) executive vote.

1 Like

Not really. Having a slush fund without a clear mandate from the community on who controls it and how to spend it can be a set up for recrimination and mistrust and breaking the social contract with MKR holders. I would start with a mandate on what and how much we should spend money on to support the protocol and then consider how to fund it. Its the same cost for MKR holders if you divert income into a slush fund or mint new MKR to pay for a cost but allow existing income to be used to burn MKR. The main issue here is getting consensus on how to spend the money before diverting income from burning MKR.

1 Like

There is no income from the USDC vaults unless the peg returns to $1.00. If they were liquidated today, all of the USDC would be used to buy back the outstanding Dai at $1.01 and there would be losses on the CDPs with an LR < 1.01 that would require minting MKR to pay off. We should burn MKR with the surplus Dai, however, because we will need to mint MKR when we liquidate these CDPs unless the peg returns to $1.00 or below.

1 Like

I also want to add that MKR price helps stir up excitement about the project, which can help bring attention and lure in new talent. I think it’ll be easier to get new domain teams onboard after a sustained MKR pump


I’m closing these polls as it’s past the communicated closing date and I haven’t seen @Aaron_Bartsch in a few days.

I don’t think there is a clear consensus here to make any changes based on these poll results. Note that @Risk-Core-Unit have posted a thread here which may have some impact on the surplus buffer.

Thank you to all for voting.

Anyone know the address that shows the surplus balance, I’d like to be able to track it.

As I understand it’s not actually ERC20 DAI at this point. In fact I’m pretty sure it’s not even one field in the smart contract. I think it’s a aggregation of a couple of internal accounting values in the vat. Some of the smart contract guys may be able to confirm.