[Discussion] Greenbacks


There has been a lot of talk about The Peg™ and how to manage it and get it back on track. Some people have suggested protocol-minted unbacked Dai, others have suggested changing how the levers work, I myself suggested giving those levers to the system in exchange for a few more, higher-level levers. But I've done more thinking of my own, and I came up with this, which I hope may appeal to more of the community. Unbacked Dai itself is just not a good idea, I think, for multiple reasons, chiefly that all Dai must be backed by more than its value in collateral. However, I also think that the money supply must increase to reduce the upwards pressure on the price of Dai.


Enter Greenback Dai. These tokens are a special subset of Dai, equal in trade value but not redeemable for collateral in an emergency shutdown event, that are minted by the protocol in limited numbers following a vote by governance, and they are for situations like where we are now, where The Peg™ simply will not get back to where it's supposed to be no matter which levers we pull. I haven't an idea on how they'd be distributed, or to whom, but one way to burn them would be to have a contract that people can return them to, and offer something (maybe a small amount of MKR, there are lots of interesting ways to handle it that I can think of) for taking on the increased risk of using this unbacked currency. This would essentially mint what are effectively unbacked Dai, without actually minting unbacked Dai. All Dai remains backed, the overall supply of Greenback Dai can be easily monitored, they will be permanently destroyed upon payback thus decreasing the money supply, and users get rewarded for taking them and turning them in over time. The idea is that these Dai-value tokens will temporarily reduce upwards pressure on The Peg™, but will disappear over time. My only worry about this idea is that the amount of Greenback Dai we'd have to mint in order to fix The Peg™ would be absurd and would cause problems for the system. Thoughts?
1 Like

I think you are proposing issuing a separate stablecoin in the hopes that it will decrease the demand for Dai and bring its price back to parity.

The fact that competing stablecoins have been able to hold the peg while Dai has not shows that there is a demand imbalance for Dai specifically, not stablecoins in general.

Therefore, unless the manner in which Greenback Dai derives its value is the same as standard Dai (redeemable for collateral during ES), I don’t think it will help alleviate the demand issue. If it is redeemable during ES, then it has the same problems as minting unbacked Dai, namely that there is not enough collateral to satisfy everyone in case of settlement.

Anyway, the QE/OMO proposals don’t involve keeping unbacked Dai in circulation. The minted Dai is immediately used to purchase collateral, which is held on behalf of the DAO (rather than held by the DAO, on behalf of a vault owner). This has the same effect of users opening vaults and shorting Dai, except the DAO can do it more directly and without capital constraints.

1 Like

DAI is backed not only by its value in collateral but also by its properties (intangible) that centralized stablecoins don’t have:

  • the inability for governments to freeze it
  • its governance
  • does not rely on banks (EDIT)

When the demand is greater than the supply, the market thinks that DAI is backed by more than $1 in value (collateral+intangible).

We have to decrease the collateral backing because the intangible value of DAI is increasing and the sum is greater than $1.

I don’t see the need to introduce a new type of DAI - less complexity is better in 99% cases.

1 Like

How would you ensure that these tokens have the same value as DAI?


I can’t argue that.

I probably should have specified (and I’m going to edit my OP accordingly upon finishing this one), Greenback Dai would not be redeemable for collateral during an ES, which makes the currency inherently riskier, which is why I suggested rewarding users for taking it on.

I now understand why people have been suggesting lowering the LR. Makes sense, thanks!

That is a very good question and, somewhat embarrassingly, I do not have an answer. I am aware that once it got picked up by anything outside Maker, all bets are off because by design there would be nothing to make it hold its value.

Overall, I’ve disappointed myself with this post, and I should have taken more time to flesh it out. I should not write late in the day. :laughing:

Hah, no need to feel bad about it. I’ve been here for over a year, I’ve had most of these conversations three times over with three different people.

The same themes tend to reoccur as new people join (which is completely fine.) Just means we have to work on education and informing people.

I honestly think the best way to learn everything is just to post stuff like this and be willing to be wrong in front of people. Ask questions, figure out where you’ve missed things. It’s the fastest way to absorb knowledge from the rest of the community.

1 Like

I like the concept of creating additional “stablecoins” with different properties. Unfortunately we can keep talking about all our great ideas but as long as we have no domain team/dev bandwidth nothing is going to actually get done.

Releasing something like a non-pegged stablecoin (similar to the TRFM or “rai” which another project is working on) would be interesting. This would be more of a pseudo stable asset that may be a better fit for some of these yield farming protocols, and could reduce the demand on the usd-pegged dai. Depending on the application I can see pseudo-stable assets with different properties being better fits.


I don’t see how it can work technically.
You need to have the contract to be interchangeable so let say we can flag it which should not be possible with a ERC20.
Then how would you exchange it against a normal Dai.

Had an epiphany on account of that post. Thank you.

You wouldn’t need to. We could allow it to be deposited into the DSR (which people seem to have forgotten about, for some reason), and then for each Dai earned in savings, an equal amount of Greenback Dai would be burned. The system would effectively be borrowing against future stability fee income. We’d have to make changes to the DSR contract, but the idea is that when you deposit Greenback Dai into the DSR, instead of the interest adding to the overall balance, the Greenback Dai balance is slowly converted to Dai using the savings rate. The Dai earns interest as normal, regardless of the presence of Greenback Dai. The effective savings rate is therefore:

S = Id - Ig


S is the effective savings rate,
Id is the rate on the Dai balance,
Ig is the rate on the Greenback Dai balance.

This defeats the purpose of the greenback in that you now have to lock up the excess supply you created, which is only helpful while in circulation.

Correct, but governance can vote to mint these tokens–in theory, this would be a self-limiting process, since I can’t imagine that the majority of MKR holders would continuously vote to take on more unbacked debt than necessary–and use them to, for example, pay domain teams, when emergency action is required and there is insufficient Dai in the buffer.

To clarify, the system would not ever hold greenbacks itself, and would instead mint these tokens and as part of that process immediately send them to their proper recipients. I think MKR holders will be prevented from simply voting to mint endless amounts of greenbacks for themselves to either run out the system and inflate their MKR holdings or to create ever-growing streams of Dai savings effectively from nothing, since they’d have to agree to help each other collectively screw the system.

The recipients could then either put their greenbacks in the DSR to be converted, or trade them for assets or services–really no way to prevent that, but upon further consideration there’s really no need to, since the fact that they can be converted to Dai (and that this is pretty much their sole reason for existing) should, in theory help peg them to the Dai. Personally, I would not pay more for such a token than I could potentially redeem it for.

This would allow the system to take on some debt and temporarily increase the money supply by borrowing against future stability fees, all without diluting MKR, which would then be diluted only in the event that:
a) governance could not agree to mint greenbacks; or
b) stability fee revenue was insufficient to pay back the debt.