[Poll] Creation of a MakerDAO strategic reserves (solving the peg)

I want to poll opinion for an idea to create strategic reserves (a Maker wallet for discretionary usage at the hand of the governance) to fix the peg. This can move to a MIP13 declaration of intent if there is a bit of support and with the new insights that this discussion can bring.


What we’ve done/explore already (not exhaustive):

  • USDC-A vault with no liquidation SF 0% and LR 110% => was used for COMP farming and a whale used it for a day before using Compound only again.
  • Increasing DC everywhere again and again and SF 0% => did help to bring exchange rate towards 1.015
  • Discussing to print unbacked DAI (here and a poll here). Not strong support for the idea.
  • negative rates. Not strong governance support.
  • The PSM that is no longer worked by the foundation (no community work that I’m aware of).
  • Forward guidance that is under discussion.

Here we are, needing more DAI supply because the peg is rock solid @ 1.015. We have room for that in the vaults, but no one cares or use it for farming which doesn’t help much. That was a bet that WBTC-A would be filled in less than 8 hours … well …

Arbitrage doesn’t happen most likely because 1.015 is becoming the new normal and the outside spread to provide arbitrage is currently 1 (where we can raise rates) and 1.10 (where USDC-A can be used for arbitrage). A bit too wide.

All our tools are passive, meaning we teak some parameters and hope third parties will change their behavior. Maybe we should explore way to do it more directly.

Proposal : MakerDAO strategic reserves

The idea come from the discussion about the surplus buffer which is from a discussion about system reserves that have good gems in it. It’s not different than what was discussed at the end of the thread but with 3 more months and more details. I already touched that I think MakerDAO need some equity and a real balance sheet.

The proposal would be having a Maker wallet (the strategic reserves) outside of the DAI system with a multisig for MakerDAO governance (not sure the details). Those reserves would be used for:

  1. Arbitrage the peg
  2. Develop the DAI ecosystem (providing DAI to Aave to decrease the borrow rate or to the Uniswap DAI-#### pairs).
  3. Generate earnings to Maker that can be used to pay for stuff or burn MKR tokens. That would provide earning when SF are @ 0%, a diversification of earnings.

There will be a need for a something similar to MIP14 to manage this wallet.

Strategic reserves will have to be funded by a MKR issuance. At some point we could channel the surplus of the surplus buffer to it but let’s keep thing simple.

Strategic reserves can use Maker to borrow DAI but doesn’t have any privilege. It stays completely separate from the DAI subsystem.

It remains to be discussed if strategic reserves are another insurance for DAI holders. It will probably be used to buy MKR if needed avoiding MKR holder dilution. In such case, it’s also a safety buffer for MKR holders.

An example can be to raise 10M USDC for x MKR tokens. Use them in a USDC-A vault to generate 9M DAI. Sell 100k DAI per day for USDC (sell order on Oasis @ 1.015 to start). Leverage if needed. We can add liquidity to Uniswap DAI-USDC or lend on Aave/Compound while we are at it or stay using only Maker tools (which seems a safe start if not the best for earnings).

We can use TUSD and/or PAX as well of course.

Strategic reserves are not the best yielding opportunity (because we will have to be careful) but they will increase (hopefully) the value of the Maker ecosystem a lot hence the MKR valuation.


  • Provide power to MKR governance to act directly in the markets instead of waiting for third parties to act.
  • Quickly accretive on MKR earning per token (starting from almost 0 that’s easy)
  • Should be technically easy to setup (I assume, not sure).
  • We have 100% of control over it.


  • Such reserve can prove to be not enough to fix the peg.
  • It requires to issue MKR tokens, diluting existing token holders (but increasing the earnings per token)
  • Using other protocols can induce a risk (not that we have to use them, I assume we are okay to use our own protocols)
  • It has some management complexities

The poll

  • Yes, we need strategic reserves but it should be limited to the peg issue and dissolved
  • Yes, we need strategic reserves as a strategic tool for the Maker ecosystem
  • No, we don’t need strategic reserves. Ii will not help
  • No, we don’t need strategic reserves. it is too risky
  • Abstain, I don’t want to influence
  • Abstain, not sure what to think

0 voters

If you think we need strategic reserves, what amount should be raised in $? Notice the market capitalization of MKR is around $600M

  • 10M
  • 25M
  • 50M
  • 100M
  • More
  • Abstain

0 voters


I like this idea, but I think it’s important to have boundaries set for what the protocol can and cannot hold in reserves. For example, uniswap tokens are low volatility, provide a yield, and some tokens such as dai-eth do not have custodial risk. Personally I think these should comprise the bulk of the strategic reserves.

I think it may be okay for the protocol to hold custodial stablecoins in reserves, but only if they are widely diversified, rather than simply holding all of the reserves for example in usdc. Volatile base cryptocurrency assets like ETH do not seem like a good fit as they have way too high volatility at the moment for what should be an effort to reduce the risk of the protocol.


Selling surplus DAI for other assets creates risk that would need to be managed. Unless the thing could be automated or delegated, which seems doubtful, governance overhead is going to create too much fatigue on voters. I might support narrower, less open-ended approaches.


Regarding the farming ( compound) there is two variables the interest and the token distribution.
The token distribution remain constant however the number of dai. So more dai you put into compound less the reward is. Recently, the reward has been cut significantly by adding more dai and also die to the drop of the token price.
It is just an opportunity that will end as soon as the gap is filled. Probably around 200m or 400m more inside compound. Just wanted to say that dai are not necessarily mint to go to compound but they end up there due to the opportunity and it is not useless to fill compound as it has to be filled.

Then I think the main problem is technical. It must to pass by the protocol as it cannot be humain manipulated and probably need some contract dev.

I like the Uniswap poll, but I am not too sure about the V2 and liquidity fees if you put 20m fees can be very high.
V1 doesn’t manage pair.

For me ideally I would see our own polls same type of poll as Uniswap ( the contract itself is very easy). 50% dai, 50% collateral for each of our collaterals. The poll can have no transaction fees or little. Need to work out the security issue of not having fees.
Specificity, you inject only dai into the poll and you withdraw only dai. That will sort of buy the collateral or sell it.
Only maker protocol can inject liquidity ( no other parties into the poll), the injection will be done via governance.
The token generated is placed into vault with 50% to allow governance to free the remaining dai inside the token eventually. The dai freed can only be reinjected into the same or another poll. You are 100% safe because your token contain 50% dai.

I am not too sure about the fluctuations of the reserve.

Benefit : you can use the poll for the liquidation.

are you suggesting we rebuild an internal uniswap? That would involve bootstrapping an entire new network effect. The fees are no higher on uniswap v2 than v1, there’s just the potential for some of those fees to go toward a governance team. Currently this option is not active. Also by holding uniswap tokens the fees would go to the maker protocol, we wouldn’t be paying them.

I also think it’s strange to say that uniswap is very simple. There was a lot of work put in to building out v2. This is like saying the maker protocol is very simple so some other community should just recreate maker.

There is some new fees for the liquidity provider on V2.
We can add fees but the fee income is depending of the liquidity inside the poll.
With the number we are talking more than 10m, the fee will be irrelevant anyway. And it is the same for uniswap the fees we will get are going to be near the floor.

The fee works because you have low liquidity provider vs volume.

The contract itself is simple as the opposite of the all environment, website etc … the contract itself is smart but simple 100 lines or something like that. And most of it is to work out the fees.

Another point, you can’t inject only dai as far as I am aware of. It is also public, so the second part about the vault would not work.

I like this idea but I don’t think we necessarily need to print more MKR to fund it. I’d rather start building out the infrastructure for these reserves and use the system surplus to fund. We could think of these as governance managed savings funds. Even something similar to YFI’s yvaults maybe.


The fee is a flat 0.3% of transaction size, which is given to liquidity providers. Historically for many pools it is something like 10%+ yield annually: https://medium.com/@pintail/understanding-uniswap-returns-cc593f3499ef.

We would inject dai to buy eth, then use this along with additional printed dai for the dai-eth pool. This gives the protocol square root exposure to eth volatility, and injects 50% of the total value of the uniswap token into the market.

I’m sorry this is getting off topic so I won’t address it again here, but I just feel you’re mis-characterizing uniswap, or there’s some misunderstanding here, although I can’t quite pinpoint what it is.

1 Like

it looks, there is some new fees during the minting for liquidity provider :

Regarding the 0.3% fees you get, they are depend of the volume. for example if you have one poll dai/usdc with let say 10m you need 1m in volume by day to get 10%. today volume on oasis is 100,000 which is if you extrapole ( I doubt there is this volume everyday) an equivalent of 1% yearly.

I do believe lower fees will bring more people into markerdao.

Also I am not too sure about the safety of the new contract. It hasn’t been used enough for me.

Thanks for those contributions.

For what can go inside thoses reserves. I think we should start small to have a consensus. DAI, USDC, TUSD and PAX (the last 2 will probably be onboarded when those reserves will be launched) and allow usage of Oasis and Maker Vaults. Except DAI, all assets have a 50% cap of the reserves. It seems we are talking about $10M, nothing that will impact the risk profile we already have. A $10M reserve would be a 1.6% of the MKR market cap. At the worst, I doesn’t impact much. I would love to have Uniswap and Aave/Compound as well (again no more than 50% per protocol), we will see.

Regarding the Uniswap discussion, I think we need to have respectful partnerships and provide value to DAI users. Uniswap provide an easy way to provide liquidity, is well integrated in other tools and easy to use. The DAI-ETH pair has enough liquidity and it’s not even $10M that will make a real difference. I was aiming at the DAI-USDC pair because here $1M can make a real difference and boost usage (comparing with the USDC-USDT pair). It would reduce the cost to convert DAI in USDC (I know there are other better exchanges for that). I would see the same for TUSD, PAX. With $1M, we would be the reference pair for those crypto dollars (I would add PAXG as well).

Providing liquidity to money markets is important as well. Currently DAI is expensive on Aave and the supply is only $16M, we can move the needle a bit. Decreasing the incentive to deposit DAI on Aave and increasing the incentive to borrow DAI (instead of buy) is good. We can provide $5-20M (depending on the leverage) almost doubling the supply. Getting half of the flash loans fees (without doing the work) and at least 1% APR.

The guideline shouldn’t be the yield (at this stage at least), it’s the ecosystem building. Bringing value to DAI users so some will create vaults and expand our market share. Even with a 50% APR on the strategic reserves, it would be only 0.8% over the MKR market cap (for $10M reserves). Good, but not a game changer.

Regarding the governance. If we decide for something simple like selling $100K of USDC per day on Oasis at max(min(average last day, 1.015), 1.005) it will require one vote once and some delegates to propose and verify and confirm the sell order each day (7 of 13 validators for instance).

Regarding the funding, It would take decades if we wait for the system surplus funds and it require a change in the Maker protocol. I think that having something segregated with low technical complexity is better to start and test. But I agree 100% to use the overflow of the system surplus if that scales well.

If that works and restore the peg it would mean than $10M directed is more important than the last $100M+ added to the DC. It would mean that we can restore SF to at least 1% which is $4M earning a year (almost the half of those reserves).


Adding a link to @MakerMan post on MIP14 which is related to this one as well.

He proposes 2 of 3 or 3 of 5 for validation where I proposed 7 of 13. I think more is more resilient and provide more diversity and avoid collusion. Maybe the truth is in the middle.

I also found a signal request for the creation of a treasury. It’s not far from that (in the implementation, the motives are different).

Any profit from the strategic reserves funds can be used to pay for Maker expenses starting with the oracles.

My idea to merge those two topics is the following. We will need some money to start the strategic reserves (let’s say $10M). One way is to mint 20k MKR and transfer them in the fund (which can sold them as needed on the open market to avoid moving MKR price too much) That would lead to a small dilution for MKR holder (no one like that). Another way is to ask the foundation for those 20k MKR and in exchange take a bit of their expenses. No dilution for existing MKR holders but obviously the foundation should have a fair deal. As a bonus it will move forward the decentralization of Maker. What do you think? Does that makes sense?

In the later case, we need a bit of financial planning.

MKR minting has been pretty unpopular with the overlords (MIP 14 failed to pass and Black Thursday compensation failed to pass).

Might get more buy in with leading this proposal with a new Treasury domain team or core member via MIP0.

An alternative would be to inflate the surplus via (unpopular opinion) increasing the base rate to non-zero, and taking from there

1 Like

I wrote a document following the MIP13 guidelines, anyone can comment on there if more convenient. I’m not sure a MIP13 is the way to go. @LongForWisdom if you can help here it would be welcome.

@Jtathmann , MIP0 seems to add new people, i.e. new expenses which we don’t know how to pay anyway right? I’m more in a MVP mode here. Always provide value before expenses and structural overhead. If such strategic reserves are useful, I don’t doubt we will make them permanent and it will need more permanent structure at scale. If not we can just resolve the reserves. Maybe I misread your point, but asking for money + adding expenses is too much. I agree 100% that even asking for money will be far from easy.

1 Like

Skimming it, it seems like a good use of the MIP13 framework. In what way do you feel that it’s unsuitable @SebVentures?

Happy to provide any help I can with regards to formatting or explaining the process. Feel free to get in touch on rocket chat.

The reason why I polled for “Yes, we need strategic reserves as a strategic tool for the Maker ecosystem”—I truly believe that If you’ve got an early stage company without much revenue, then it’s probably not too smart to stash away cash to cover 6 to 12 months, or more of operating expenses. But—it is definitely smart to save for the future and adopting a good approach to accumulating reserves.

With that in mind—I do not generally agree with diluting MKR because such tool was intended to be used for emergency purposes—and I’m a true believer that you should never get high from your own supply—because eventually events like those of March 12, 2020 will from time to time come home to roost.

I would rather push & introduce Real World Assets—like Real Estate—as a side component and use the generated fees from such to build the strategic reserves fund (SRF). Maybe a decent percentage at a time.

However—I do like Seb’s idea of: “Transfer of some MKR tokens from the foundation in exchange of taking some foundation expenses.” How that would work is beyond me…

This is mos def a subject that the community will have to confront one day–glad Sebastian has put the pedal to the metal…


It was more along the lines of providing structure to make these decisions, since at large MKR holders won’t be voting on granular decisions, at least that’s the end goal.

I think it would be more palatable to package these all together: protocol DAI transfer + strategic reserve/cash account + treasury oversight

1 Like