[Signal Request] Increasing the Surplus Buffer (2021-11)

Overview

The last attempt of raising the Surplus Buffer by 40 MM DAI to 100 MM DAI failed onchain in late September. Since then, the growth rate of the System Surplus increased a lot - driven by a huge DAI demand from crypto-collateral and by DAI-supply slowly moving from PSM to crypto-vaults.

As the reasoning for increasing the Surplus Buffer might be a bit lengthy, the TL;DR; is

  • we’ll start burning soon and we should consider using the momentum to increase our treasury and increase portfolio risk compensation
  • MKR price might get high and it becomes less efficient to burn
  • The burning aspect doesn’t do much in terms of MKR price pumps

Data

Let’s do a quick run down of the relevant numbers first.

According to makerburn we are currently running on ~90 MM income from Stability Fees and ~40 MM expenses - leaving us with an annualized profit of ~50 MM. That’s about a million per week. As the Surplus Buffer is currently set to 60 MM and we already have a System Surplus of ~57 MM DAI we will start burning MKR in about 3 weeks. Probably a bit earlier, as

  • Core Units expenses are usually not claiming their budget as soon as available
  • Income from liquidations is not considered here
  • This weeks executive will include some SF-increases which will add another ~20 MM of annualized revenue
  • We might see some more DAI minting in the near future which will increase the revenue even more

I created a little spreadsheet showing the different options and their effect on when we would fill a higher Surplus Buffer. Feel free to clone it and play around on your own.

Taking the increased SFs into account, we would end up with something like this

Surplus Buffer increase New Surplus Buffer Burn Fraction Absolute increase per week Weeks until SB filled
$30.000.000 $90.000.000 0% $1.342.198 22
$30.000.000 $90.000.000 25% $1.006.648 30
$30.000.000 $90.000.000 50% $671.099 45
$30.000.000 $90.000.000 75% $335.549 89
$70.000.000 $130.000.000 0% $1.342.198 52
$70.000.000 $130.000.000 25% $1.006.648 70
$70.000.000 $130.000.000 50% $671.099 104
$70.000.000 $130.000.000 75% $335.549 209
$110.000.000 $170.000.000 0% $1.342.198 82
$110.000.000 $170.000.000 25% $1.006.648 109
$110.000.000 $170.000.000 50% $671.099 164
$110.000.000 $170.000.000 75% $335.549 328
$150.000.000 $210.000.000 0% $1.342.198 112
$150.000.000 $210.000.000 25% $1.006.648 149
$150.000.000 $210.000.000 50% $671.099 224
$150.000.000 $210.000.000 75% $335.549 447

Suggestions by Risk Core Unit

@Risk-Core-Unit proposes to keep Surplus Buffer in a range from minimum 2% to 3% of debt minted from volatile collateral, which is based on their new methodology of calculating Capital at Risk. This puts the proposed Surplus Buffer figure at 90m to 130m from their side.

There is also a nice chart on makerburn showing how this has evolved in the past - we have been a lot better on this ratio in the past (we even touched 2.6% very briefly, right now we are back to 1.25%) and we should consider getting to a more healthy level again.

image

In the long run we should really come up with an IAM for managing the Surplus Buffer, so we don’t have to worry about this anymore. But right now I think that’s the way forward in the short term.

The Signal Request

This Signal Request is split in two parts:

  • To which level do we want to set the Surplus Buffer?
  • Do we want to keep on burning some of the profits?

To which level shall we set the Surplus Buffer?

Pros:

  • if the system goes underwater, fewer MKR will be flopped as the higher System Surplus can cover more bad debt
  • enables us to more risk oriented behaviour - more aggressive debt ceilings, more risk taking collateral onboarding

Cons:

  • as long as the System Surplus is lower than the Surplus Buffer, we will not be burning MKR
  • DAI in the System Surplus is DAI taken out of circulation (not a big issue, the supply side is elastic due to the PSMs)

Please vote on ALL options you would support in an onchain-poll.

  • 60 MM (no change)
  • 90 MM (+30 MM)
  • 130 MM (+70 MM)
  • 170 MM (+110 MM)
  • 210 MM (+150 MM)
  • Abstain

0 voters

Shall we increase the Surplus Buffer slowly so we keep burning a fraction of the profits?

This is a bit tricky. With the current setup we cannot have something like “use 25% of burning and put 75% into the System Surplus”. What we can do right now is to slowly linearly increase the Surplus Buffer over time. We would need to set the angle of this slow increase according to the forecasted profit/time.

As the lerp will be configured on an absolute number over time I rather ask for an absolute number for increasing over time instead of asking for a percentage.

However, assuming the next SF-increase passes onchain, we will likely have a profit of 1.3 MM per week (source). If we want to go for a slow increase of the Surplus Buffer, those options might make sense

  • set SF to new level directly (no burn)
  • increase the Surplus Buffer by 1 MM per week (leaving ~25% for burning)
  • increase the Surplus Buffer by 0.67 MM per week (leaving ~50% for burning)
  • increase the Surplus Buffer by 0.33 MM per week (leaving ~75% for burning)

based on this calculation.

Please note that the actual % cannot get guaranteed, as we would basically just calculate the end-date before adding it to the exec - if we accrue a lot of more fees than expected the burn-ratio will go up, if we accrue less, the ratio will go down (maybe even to 0).

Also noteworthy, as it happened during Wild Wednesday: if we want to keep on burning a bit of the profit, a large wave of good running liquidation will result in a lot of liquidation penalties directly going into burning as it overflows the Surplus Buffer.

Pros:

  • we will keep burning some MKR while we also increase the Surplus Buffer

Cons:

  • it takes longer for the Surplus Buffer to get filled

Please vote on ALL options you would support in an onchain-poll.

  • set target level directly - no burning
  • increase Surplus Buffer by 1 MM per week
  • increase Surplus Buffer by 0.67 MM per week
  • increase Surplus Buffer by 0.33 MM per week
  • Abstain

0 voters

Next Steps

The Poll will run until 2021-11-19T14:00:00Z; its outcome will result in an on-chain-poll assuming the outcome of the poll deems it necessary.

9 Likes

I voted 60MM (no change).

But regardless… what is the point of this on-chain poll

(https://vote.makerdao.com/polling/QmXcVy9R?network=mainnet#poll-detail)

voted ~45 days ago?

Has something critical happened?

Is the Risk team suggesting to rush for another vote?

It’s just that it feels a bit of a waste of governance energy and gas to reiterate polls with such short intervals.

4 Likes

That was a very close poll and the initial votes here suggest there’s interest in revisiting this. I think @ultraschuppi took the initiative to put together a good SR and also got input from others beforehand to better inform himself and others.

3 Likes

I guess @Primoz might chime in here as well, but this SR has been created in collaboration with him. also just looking at this here

should show that the situation changed a lot since the last on chain poll failed

Edit: not a good screenshot - we are at 1.24% now, on the day the poll failed we were at 2.25%

2 Likes

darn, i forgot that part:

I am thinking about sketching out some pre-MIP proposal for having a IAM on the SB so we don’t need to waste so much governance effort on that topic on the lon run. if somebody is interested in bouncing around ideas, please DM me or schuppi#6905 on discord

5 Likes

Some general thoughts on burning:

  • Burning is good PR
    • EIP-1559 has basically done a lot of marketing on our behalf. Most of the cryptosphere now understands the value of burning.
  • Not burning is bad PR
    • The non-burning of MKR seems to be brought up a lot in the community outside of this forum (i.e. subreddit).
    • If we keep pausing burning for long stretches of time, people will lose confidence in our intent or ability to burn.
  • Burning is investing in ourselves
    • If we burn MKR but need the capital later, we can just mint and sell MKR at that time (at presumably higher prices).
    • While minting MKR to raise capital might be bad PR, at least we can communicate at that time why we need the capital.
18 Likes

Yes so we propose increasing Surplus Buffer by a minimum of 30m based on this:

Let me be clear that I am not against partial burning of MKR, but for obvious reasons (I represent Risk Core Unit after all) we prefer a more appropriate sized Surplus Buffer. I do think that burning MKR can impact the price of MKR positively, but mostly because of potentially positive PR.

In any case, partial burning should have a positive effect one way or another and it does help us with our capital backstop: if MKR burning stimulates MKR price and liquidity it helps Maker in potential debt auctions. So I’d probably prefer 30m increase and 25% burning because it both has benefits in terms of capital robustness.

10 Likes

I’m fairly new to the forum but I’ve been a long time holder of MKR. I think in general the negative reputation being gained by prolonged time between burns is underappreciated. When I first got into MKR the elevator pitch was that it was a coin that was generating revenue and ensuring payout by burning itself. I was really happy to see that pitch come into fruition this spring and disappointed when it stopped as suddenly as it started.

Now I recognize that was a function of the parameters put into the compromise of a partial burn at that time, but I think seeing the burn start and then stop again turned a lot of aspiring MKR fans away from this project and over onto something else without a tarnished reputation.

Disregarding risk-assessment seems ill advised but so does raising the surplus buffer once again just as burn seems within reach. MKR is a coin with low liquidity and thus high volatility. I fully expect such a disappointment in the eyes of many low level investors to have pronounced effect on MKR price.

MKR in general has a high bar of entry and I think we should be careful to scare of new blood once again.

9 Likes

Just a quick thought regarding this. The current annual profit estimate on makerburn is 1.4% of the current level of DAI minted from non-stable coins. If the Surplus Buffer was set to reach a target of somewhere between 2% and 3%, with no partial burning, it would take about 1.5 - 2 years before every new DAI minted started contributing to the burning. And with the current level of stability fees, as long as there is positive growth in DAI minted from non-stable coins, there would actually be no burning at all.

I therefore believe that a dynamic target (a percentage suggested by the risk team) combined with partial burning during the build up to the target would be the best way to go. This would of course lead to a higher gap between the System Surplus and the Surplus Buffer in periods with growth, which should be taken into consideration when setting the partial burning level, but at least then we would secure some level of burning during periods of growth.

Btw, I am definitely not saying that 2-3% is wrong or too conservative (I choose to trust the risk team here), but I justed wanted to point this out.

6 Likes

excellent topic.

I strongly suggest Maker 1st determine as a comunity what a reasonable capital surplus is, and then work to get towards whatever that level is. Risk-Core-Unit has already provided a minimum level which are choosing to not follow.

Can Risk-Core-Unit share how they sized the minimum 2-3% Capital Surplus range (measured versus debt minted from volatile collateral)? What are the key assumptions in this analysis? What are the downsides if we do not maintain at least these levels?

Maker has a highly leveraged balance sheet where a capital surplus and liquidity are typically used to protect the balance sheet in stressed times. And the capital surplus and liquidity are sized to deal with the risks in the balance sheet - credit risk in assets, liquidity risk in DAI…

Its great to be able to burn profits to buy back MKR, but if Maker blows up due to credit/liquidity/insolvency risk, we all know what the value of MKR will be: ZERO

My rough sense of Maker’s mission is as follows:

Maximize DAI outstanding such to the constraint that Maker doesnt blow up along the way from key risks such as liquidity, credit and solvency.

2 Likes

finally, remember that a capital surplus is typically expressed as a percentage of risky assets. And DAI and risky assets have been growing rapidly but the capital surplus has not been growing.

going forwards Maker should probably manage the surplus as a percent of the risk assets…

3 Likes

Greetings everyone, I am new to this forum, I am a holder of 10 makers, been holding them since july, I was very impressed with makers gains back in may and this is what compelled me to buy it when it crashed to 1915 usd, Maker continued thriving even in july when juxtaposed to the rest if the crypto market, however i have noticed that ever since these surplus buffer proposals started surfacing, things started heading sideways and the coin no longer became attractive to investors and whales, I understand where everyone comes from when they need to collect as many as dai possible when the market is in good health, but please reconsider, this not only tarnishes the image of the project (since it is marketed as deflationary and yet never burns), but also upsets current maker holders, makes them question the governance expenses, accuse the employees of theft and question the i lntegrity of the whole system, please reconsider this proposal or lower the surplus buffer to 30 mm again, we all have a lot at stake in this project and its not in the intrest of any of us if it fails.

1 Like

I know it is frustrating to delay potential burn again, after we saw a run to over 6k before we had a good burn, MKR seems to react pretty disproportionately, removed some supply and the price went ballistic! But the safety reasons to increase the SB are essential to “survival” and for the sustainability for the protocol.

This is 100% true!

But there was an on chain vote end of September and now we discuss just after a month the same topic I thought there should be a cooldown when you want to vote on the same topic again?!

3 Likes

The previous signal was started at the beginning of September, it’s now the beginning of November. The DAI supply has increased by ~33% since then, as we’ve gone from ~6B to 8B, further as people have noted the ratio of surplus to risk-DAI is lower. If circumstances change significantly, that is a reason to present options to vote, once again.

That said, I do think it would be preferable to approach this in a more durable way. We could vote on a percentage to try to maintain, or vote on a desired rate-of-increase for the buffer. In an ideal world, I think we’d have an instant access module to change the buffer according to some set of rules predetermined by governance.

10 Likes

Why would the value be zero instead of the market cap of MKR reduced by the surplus buffer deficit?

I am voting to up the SB to 90 or higher and to take DAI to the flapper at some rate (.33M or .66M/week)

What I think would be interesting is to alter the flapper so that all MKR purchased enters the DSSVest contract (funding for MKR compensation to CUs) and then basically to have a DSSVest SB limit with similar ‘lerp’ type function to formally burn some fraction of MKR added to the DSSVest contract.

Point here is that whilie everyone wants to burn MKR I think it would be prudent to tie DSSVest funding directly to MKR being bought by the flapper, and basically lerp burn some fraction of that MKR depending on protocol needs.

Either that or we could just accumulate all flapped MKR in DSSVest and have governance decide whether/when to burn from that contract or to simply add a lerp kind of MKR burn function there.

This kind of thinking basically eliminates MKR minting for DSSVest, and ties MKR compensation to the protocol being profitable and lerp flapping for MKR to fund DSSVest. Then DSSVest basically pays out what is has accumulated as a ratio of what it has to what is requested.

1 Like

No, that’s not what this poll means. It’s the rate at which the surplus buffer increases, not the rate at which DAI moves to the flapper. We can’t directly control that, it’s a function of profit and rate-of-increase-of-buffer.

1 Like

Ah thank you for clarifying @LongForWisdom I want some burn component but just altered my vote to 1M and .66M vs. the .33 and .66M.

2 Likes

I said blow up, meaning DAI holders lose confidence in the value of DAI and pull most of their funds out of Maker.

Thus Maker is left with little capital or value

1 Like

I also suggest that capital surplus touches on other big issues at Maker: capital allocation and overall strategy.

Every DAI of profit can be used for three purposes:

  1. Burning MKR. Similar to dividending profits.

  2. Investing in building/scaling business. Basically reinvesting profits to generate more profits later.

  3. Building capital surplus which is basically an insurance policy in stressed markets. This decision is key to the risk appetite and policy of Maker. One can argue building capital surplus is really building the business (#2).

All three are valid (and competing) uses of profits.

In a typical finance company the CEO, CFO and CRO will be very focused on capital allocation , return on invested capital and overall capital efficiency. And the board of directors will be involved overseeing these key key issues.

In general building Maker for rapid growth requires reinvesting more capital in the business (and capital surplus) and burning less MKR.

5 Likes