Should we increase the ETH-A debt ceiling by $10 million in the executive vote on 24/04/20?

As title. The global debt ceiling should also increase proportionally if this is voted in.


  • We are within 2.5 million of the debt ceiling.
  • We are making changes that specifically incentivise the creation of more DAI (though admittedly not from ETH)
  • We want the price of DAI to drop.
  • The ETH price has been moving upward which tends to drive demand for DAI minted with ETH.
  • The ETH-A SF is set to 0% which is a large incentive to mint DAI from ETH.
  • We are shutting down SCD soon, which may result in more demand for leverage using ETH on MCD.


  • Risk in the case of large ETH crash due to the OSM attack vector.
  • Yes
  • No

0 voters


As discussed in today’s governance call, the increase in debt ceiling also accommodates the draining of the migration contract. So there is a higher-than-usual likelihood that the ETH debt ceiling gets hit very quickly, potentially as early as after tomorrow’s executive vote. If you are in favor of a debt ceiling increase, what is a desired amount? 10 million has been the norm for prior increases to the debt ceiling in cases like this.


10m sounds good to me, fwiw

1 Like

Do we have any reason to strongly limit growth in dai from eth? Perhaps we should raise it by 20-25 million even to give ourselves some breathing room.

There is a reason not to leave the gap between the ceiling and the current amount too high. If the value of ETH drops by more than 33% in an hour then there is an incentive to mint as much DAI as possible and accept liquidation. This is quite bad for the protocol. The debt ceiling limits the potential severity of this.


Looks like tomorrow’s executive is gonna be packed with tons of changes! :grin:


MKR holders should choose the amount in the poll. It’s not OK to just throw the value into the executive vote.

I agree, it would be better to be able to vote on the exact amount with MKR (ideally as a repeatable ‘this is our debt ceiling policy from now on’ rather than as a one-off, and using ranked voting). But realistically, the executive is tomorrow. Is it technically possible to put a on-chain poll in tonight and have it finish tomorrow morning? Yes. Is it warranted? In this case, I don’t think it is.

I don’t think it’s warranted in this case because this is a very one-sided issue. I’m willing to go out on a limb and say that there is no valid argument for opposing a 10M raise as against no raise at all. The one con I could think of (the OSM attack vector mentioned above) is always a risk, and increasing the ceiling by 10M is a relatively small increase in risk that is entirely justified.

To be completely frank, I would have been happy if Cyrus or the Smart Contracts team just announced they were including it in the executive.

One can argue that rights or freedoms are being trampled on by doing governance in this way, but the thing to ultimately remember is that MKR Holders have the final say. Nothing we say or do here can change that. If MKR Holders HATED the idea of a 10M debt ceiling rise, then they could elect not to vote for the latest executive. I am literally the first person in line to oppose bundling, and I’ll happy spend hours trying to convince anyone who thinks it is a good idea that there are better solutions, but in this case it doesn’t matter, as this change is utterly non-controversial.

In conclusion I expected this to be a complete non-issue, the current result of the poll 21 for, and 1 against essentially shows that it is a complete non-issue. To be completely honest, I’m annoyed we spent so much time talking about it in the governance call, and I’m annoyed I had to make this poll. We are literally doing everything we can to increase the amount of DAI in circulation, there is no possible world in which we do a bunch of stuff to incentivise DAI minting and then decide not to increase the debt ceiling on ETH.

As to increasing it by more than 10M, well, sure we could have done that, but given the late hour, and the fact that everyone would rather see 10M raise than nothing at all, I think this is fine. Increasing it too heavily does meaningfully increase the OSM attack risk, so it makes sense to aim to keep this ~10-20mil over the current amount.


I agree in principle. But in practice, we need to rethink voting system/portal.
We cannot really scale much at this point.

1 Like

Pretty much agree. We really need to rethink voting and loath massive bundling going on.

I think the 10M bump from 90-100M on ETH-A is least of issues in next executive. But everyone will be scratching head if next executive doesn’t pass.

1 Like

OK, but please let’s do the poll next week for the next ETH DC. BTW, I missed the “Supplimental USDC Stability Fee Adjustment” poll which stated:

“This Governance Poll (FAQ) will be active for three days beginning on Wednesday, April 22 at 9 PM UTC.”

Well, It’s Apr 23 and the poll is not active anymore.

As discussed in the meeting earlier, we really need to decide a debt ceiling policy for ETH-A rather than having to deal with making the same decision over and over.

If you or someone else would like to start a thread about possible debt ceiling policies, that would be fantastic.

Regarding the polls, I imagine it’s just a typo, you’d have to talk to the people that put those votes into the portal (I imagine this was either @rich.brown or @cyrus)


5M would be more fitting IMHO given how last-minute this conversation is, but 10M is also not too bad. If this goes into the exec, the most important this is to communicate it clearly. Make sure it stands out in the vote description and the blog post that this is a last-minute addition that was not polled and added later, and albeit not ideal, the market situation kinda demanded this step.

1 Like

Can we vote about the next ETH DC next week so that it can go into executive vote on Friday? Nobody likes last minute decisions.

I think the reasonable values are:

  • decrease it to 90M (that seems a bit dangerous?)
  • leave it as it is (100M)
  • 110M
  • 120M
  • 130M
  • 140M

@rich.brown @cyrus


Anyone can create a signalling thread about this. You don’t need to ping Rich or Cyrus.

As said before, it would be great to have someone start polling about a repeatable debt ceiling policy.

Have been quite off recently, amazing all the work you guys are doing, ref DC and knowing the importance of increase of dai supply while reducing any potential attacks from having a greater DC than needed, couldn’t we vote on having a DC 10mill>current supply at all times and if needed vote on changing it? It would reduce continuos voting on the matter and focus on other matters like black thursday (analysis, compensation to vault owners) and onboarding new collaterals (a ton of collaterals were proposed that’s amazing!!). SF and DSR I don’t mention since it’s cero so my focus is elsewhere.

Didn’t read this, totally agree

An automated DC policy would need to be designed with care to avoid a feedback loop that would contribute to OSM attack vector:

  • Price of ETH drops > 33% in 1 hr
  • Attacker mints DAI until current supply is within 10MM DAI of DC
  • DC automatically increases to ensure there is at least 10MM of headroom
  • Attacker mints more DAI

To avoid this, the DC calculation could reference the DAI supply from 24-72 hrs in the past. If there was a black swan event, this should leave plenty of time for MKR holders to make adjustments to the DC policy.

1 Like

Why do we need a signalling thread if we already pretty much agree on the range? I agree that the DC policy is better but it will probably take some time.

The automated DC policy makes sense. Manually adjusting the DC every week or two does not make sense to me because we are not removing any risk over having it automated in the simplest possible way.