Using LUNA-UST swap system for MIR-mAssets

The LUNA-UST is working great for creating and fixing the value of stablecoins. What do you think about using the same system for MIR tokens and mAssets? With this swap system users can mint mAssets by the value of the asset in MIR at the oracle price.

Here are some advantages that I see with this new minting system.

  1. Provide a new core use case for MIR tokens
  2. Reduce the premium price of mAssets because arbitrageurs can take UST, buy MIR, mint new mAssets at the oracle price, and sell it on the market for premium price for UST
  3. Can replace the short farm mechanism and reduce the amount of new MIR tokens created

The main risk is that mAssets will be backed by the value of MIR in addition to over-collateralized crypto-assets.
Has this idea be discussed before?

1 Like

I am for accepting spreads as a natural premium which speculators can use at high risk. I assume that the spread stays wether I sell or buy therefore is irrelevant until reasonable. When unreasonable arbitrageurs step in.
The current mecanism is well integrated in luna when you accept the spread

The spread right now is not unreasonable, but the new mechanism has other advantages. There is talk about how the short farm system is not working and that the value of the MIR token is not high enough. This can help with those two issues.

If a person thinks mir is not high enough he can just buy it :wink:
In my opinion mirror is exceptionally thought. Has a clear domain it is fully decentralized and upgradeable and waits only for third parties using the assets. They are clearly coming. I really suggest to discuss a lot, as we do, but to be patient and assess the impact of the various Mars, levana and so on. Demand of massets will come and mir staked captures a lot of value by design.


I agree that looks like a solution to MIR dilution and premium price!


The price of the mAssets are already malipulated by speculators. Speculators are good for mirror protocol, because they pay a lot of fees. Spreads is bad for trading, nobody benefits from this phenomenom (from what I understand).

I like this idea quite a lot. Burn MIR, mint mAsset at oracle price. MIR marketcap directly scales in line with market cap of mAssets. Narrowing spreads/premiums is just a good side effect.

Once MIR has a treasury/POL/reserve that establishes a floor price, no matter how small, then I don’t think there will be that much of a risk.

1 Like

Why someone would want to mint mAssets using MIR rather than aUST as collateral?

Edit: And since mAssets are CDPs, every mAsset needs to be over collateralized. Not sure why that would not be required under what you are proposing.

@Hank-O I see what you’re saying: the only kinds of mAssets that exist now are CDPs, which are inherently short positions. I think the OP is asking, what -if- we added the other kind, an inherently long position that is made by burning MIR? Could they coexist? Suppose you had a CDP-based mAsset and a MIR-based mAsset that both tracked a single real asset by oracle. Do they even need to be separate tokens?

This is a very intriguing idea and one that gives a lot more utility to the MIR token.

Exactly. CDP+minting != Burning+minting. OP is referring to #2.

I don’t think so. MIR value accrual is going to come especially during times where mAssets trade at a premium, a user could buy MIR, burn and mint mAsset at oracle price, sell on the market for profit.

Yes, I am bringing up the idea that mAsset can be created by either CDP-based and MIR-burn based.

I agree that both the CDP and MIR-burn based mAsset will be the same mAsset, which allows for arbitrators to take profits and reduce the price premium.

Maybe I’m stupid and I need an ELI5 on this, but how can you have an mAsset that isn’t a CDP?

Suppose the oracle price for a mAsset is $100 and the price of MIR is $2.
You burn 50 MIR and mint 1 share of mAsset and hold it in your wallet.
The oracle price of the mAsset goes to $200. MIR is still $2.
You decide to sell the mAsset. What happens now?

You can burn the mAsset and get $200 (100 MIR) back?
You swap the mAsset into Terraswap for ~$200?
Where did the extra value come from and what is backing the mAsset that you sold? The original 50 MIR you burnt which is only worth $100?

The reason MIR uses CDP is that you need the overcollateralization to cover the potential increase in the value of the mAsset (as set through the oracle). The fact that the collateral is backing the mAsset is what makes buyers willing to buy.

That is a very good point. This problem is different from the how LUNA backs UST because UST doesn’t increase in value.

If MIR-based mAssets is a feature, there is a risk of mAssets losing backing if the total value of mAssets dramatically increase.

To ensure that MIR is a reliable collateral for mAssets, the market cap of MIR + other collateral needs to be significant higher than the market cap of mAssets. To make MIR a safer collateral, the mirror protocol needs a significant treasury.

ok maybe I can address this. in this ‘new’ hypothetical mAsset creation mechanism. MIR does not back the mAsset, it is permanently burnt. Unretrievable.

similar to how UST has a peg price ($1) and a market price, where LUNA acts as the restorer of difference between the two prices; MIR would act as restorer between the oracle price of mAsset and market price.

So no there is no more ‘backing’, just like how UST has no backing but still is able to maintain its peg through seignorage. UST has no ‘backing’ but buyers are still willing to buy.

Example 1 (mAsset price>oracle):

  • mTSLA trading at $950, TSLA oracle price is $900.
  • User will buy $900 worth of MIR, burn that to mint 1 mTSLA
  • Sell mTSLA on open market for $50 profit.

Example 2 (mAsset price<oracle):

  • mVIXY trading at $16, VIXY oracle price is $17, MIR price is $2.50
  • User will buy 1 mVIXY for $16, burn mVIXY to ‘mint’ $17/$2.5 = 6.8 MIR
  • Sell 6.8 MIR on open market for $17 and profit $1.

it should now be clear why this mechanism is far more scalable than CDPs, which require someone to constantly lockup >1.5x value of mAsset in order to mint it, hence why OP proposed it.

1 Like

What you are essentially saying is that MIR protocol takes the losing side of every trade, the protocol absorbs all the losses as the oracle price goes up, while minting mAssets with no value/collateral backing them.

That’s not going to work.

Look at the whitepapers and work through the scenarios. Nobody would buy the mAsset if it isn’t backed by value/collateral. The value that accrues to the mAsset owner (the long) when the price goes up needs to come from somewhere—specifically another user who minted the mAsset (the short), not from the protocol. There is no free lunch.

The LUNA-UST relationship does not exists for an mAsset that is tracking some unrelated oracle price.

In the LUNA-UST relationship, there is some algorithmic monetary policy on how LUNA is minted/burned by the protocol, isn’t there? Can anyone here explain how that works? I think that’s what is missing from this proposal. I suspect there may be some rational relationship between the collateral ratio of CDP-based mAssets, the total market cap of MIR, and the total value of any mAssets created by burning MIR, and that if we get that relationship wrong then the proposed change may fail. That’s one of the reasons why I asked whether the inherently short and inherently long tokens needed to be different.

(I guess whenever a mAsset is created by CDP, a matching sLP is created with it. So we already have separate short and long tokens for each mAsset, and if mAssets were also to be minted by burning MIR, the amount that was made by MIR-burning could be computed by subtraction…)

Assuming @henryxcham is correct that a treasury would be needed to implement this change, can a treasury be raised from within the protocol? If it can’t, then this proposal might be dead.

Would like to defend the OP here for coming up with an innovative idea.

that’s like saying nobody would buy UST if it isn’t backed by value/collateral. CDP is not the only way to create a synthetic tracking token fyi… UST isn’t backed by anything (burn != backed) and yet 10B of it has been bought.

Is LUNA taking the losing side of every trade? We might all be very poor if that’s the case. MIR’s value in this mechanism would be driven by the premiums that perpetually exist for mAssets. premium = MIR burn = value accrual. More demand for mAssets thus leads to MIR becoming more valuable, which should be the case. Of course, there is downside if mAssets have no demand and start trading at a discount, but that means we have more things to worry about than tokenomics.

mAssets are supposed to track a very much related oracle price, by the way. UST also has a true oracle price, which TS calls from various sources, if you read the whitepaper. If mAssets keep their peg to the true oracle, I don’t see why we should criticise the way in which this peg is achieved. Of course, much of this depends on SC technicality which I don’t know about, but conceptually I think we can all agree that long-short CDP is not inherently a scalable way to mint mAssets.

Fully agree with @Kalirren that a treasury belonging to MIR is needed for sure, raised from fees or seignorage (similar to LUNA), to curb any price volatility and ensure liquidity. LUNA has survived for so long without a backing treasury due to its deep liquidity.