Rewarding Minters with UST from overpriced LP pool

I think we are all paying attention to the mAsset premium.
After observing the economy of the Mirror in recent weeks, I have come to one conclusion.

Think about it. All mAssets are minted by minter by creating CDP. In order to close the CDP, one must eventually buy a mAsset, which means that the first seller will always be Minter and also final buyer is.

The CDP does not generate any profit by itself. On the other hand, the holder of a mAsset can make a profit by staking it.

And importantly, even if the Oracle Price falls below the price of the mAsset, the holder of the mAsset does not have to sell the mAsset. Acutually, there is no need to sell at a lower price because the premium will not cause any loss to the holder and they know that Minter will eventually buy it back to close the CDP. If the holder doesn’t sell, the only factor driving the price down is minter’s selling, but in order to sell mAsset, they need to mint another, and eventually they will have to buy back. Who in the world would mint in that situation? It will only cost you and benefit other minter

The exception to this is when the premium exceeds the minimum collateralization rate , which is 150%. In this case, you can make a profit by Minting and selling the mAsset in the market and intentionally liquidating the CDP.

In other words, a equilibrium can be created by adopting a strategy where no holder will sell until the MCR limit is exceeded. In this situation, as mentioned before, the holder side will be rewarded for staking, but the Minter side will not be rewarded in any way. When this equilibrium occurs, the supply of mAssets will hit a ceiling because no one will Mint, and the Mirror Protocol will fade away.
And that is where the Mirror Protocol is actually headed right now.

I know there are several proposals to prevent this.

  1. lowering the MCR
  2. reward minting with MIRs

1 will certainly reduce the premium, but it will eventually lead to equilibrium. Lowering the collateral ratio is also undesirable as it makes the vulnerable to default risk in case of sudden market changes.

2 is the directly way to reward Minter for absorbing the premium. IMO it is essential to incentivize minters to peg price, but I don’t think it is sustainable because MIR are not inexhaustible.

So I propose another way.
Rewarding the Minter directly, using UST of the LP pool.
By removing USTs from the LP pool, we can directly control the premium and the supply of rewards is sustainable.

Thank you for reading, any comment?

1 Like

Interesting proposal. I’ve been thinking along similar lines, although more about redistributing the MIR rewards. Using the LP pool UST should be more sustainable in the long run, however.

When this equilibrium occurs, the supply of mAssets will hit a ceiling because no one will Mint

Since minting is essentially shorting I guess those interested in shorting assets would still do it? mAssets would be priced at 150 % of the underlying, though, and there wouldn’t be much arbitrage.

One negative aspect of using the LP pool UST instead of MIR rewards to incentivize minters is that liquidity providers will see the value of their provided liquidity go down over time, assuming constant asset prices and no re-invested rewards. So it may require a bit more re-investment management to see the value of your pooled liquidity go up.

While I like the idea, the current Uniswap-forked mechanism does not allow the protocol to be able to take UST from LP pools.

This may be possible in the next upgrade of Mirror *wink wink*

2 Likes

Hello, I am new here. Discovered the protocol a week ago and love the idea. However, the issue of the assets not staying pegged to the related stock is a big problem. I have a long experience in finance (trader for 10 years) and I hope I can help solve this problem.

I think OP’s idea is on the right track. However, you don’t want to want to reward Minters. You want to reward people who sell the tokens they minted token and lower the price of the mAsset, not people who just mint the tokens and just stick it into the LP pool.

What I suggest is that we should split the staking reward between LP token holders and people who are short the mAsset (we can give anyone who minted a token and also sell it a “short mAsset” token for staking). When the mAsset price is higher than the oracle price, give more MIR tokens to shorts and less to longs. And when the mAsset price is too low, then longs get more rewards (like they do now, they get all) and less to shorts.

This kind of idea has already be implemented by others. On sites like Deribit, there is a perpetual future that is very similar to the mAssets here (except you only put down margin instead of the full amount to open a position). When the perpetual future price is too low compared to the oracle price, people who are long (own) the futures get paid while the people who are short (sold) the futures has to pay interest, and vice versa. This way it create an incentive for the perpetual future to be close to the oracle price. The deribit mechanics is here Deribit Perpetual | Deribit.com for anyone interested

In the stock markets, shorts also receive interest from the cash they get by selling the securities short (well, if you are big and back when interest rate is not 0 anyways), so there is a parallel there as well for shorts to get something for shorting securities as well

Some more details of an example of how this idea can work

A) Create a shorting mechanism that is able to receive reward
You can allow minting of a “Short mAsset Token”. For example, when you mint a “short mAAPL” token, the following takes place

  • You post sufficient collateral (say 200% of the market price of AAPL)*
  • 1 mAAPL token got minted with the collateral locked
  • The mAAPL token is sold to the mAAPL/UST liquidity pool
  • You receive the UST from selling the mAAPL token and the “short mAAPL” token

When you burn the “short mAAPL” token, you will need to supply the cash* to buy back the mAAPL token from the liquidity pool, and then burn both the mAAPL token and the “short mAAPL” token together. Margin call liquidation also proceed in a similar fashion (the current mechanism can be adapted to do pretty much the same)

*instead of supply cash at these steps, one can apply a flash-swap type concept and reduce the amount of UST supplied, if UST is used as collateral

B) Anyone who own these “short mAAPL” token has a short exposure to the price of the mAAPL token. One can stake these “short mAAPL” token in order to potentially receive reward.

C) The MIR token award mechanism need to be changed into:

  • If the mAsset price is in line with the oracle price (for exmple, if the difference is < 1%), equal amount MIR rewards are distributed equally to the staked LP token holders and the staked “short mAAPL” tokens holders
  • If the mAsset price is higher than the Oracle price, the MIR awards will start to shift to more for the “short AAPL” staked than the LP staked. In order to keep a tight pricing relationship, the slope should be fairly steep. My proposal is that if the mAsset price is higher than the oracle price by 2%, then 80% of the MIR rewards should go to short tokens staked and 10% goes to LP tokens staked. If the price is higher by 10%, then 99% goes to short and 1% goes to LP. This is reversed if the mAsset price is too low
  • Therefore, if the mAsset price persistently stay high, one can collect most of the staking rewards by being short, therefore compensating for having to post collateral and taking on the risk of the real world price of AAPL goes up. There is also going to be low incentive for anyone to buy mAAPL at that point to become a LP holder, and thus creating a force to push the mAAPL price to be close to the oracle price. The same works if the mAsset price is too low

D) It is possible that current LP token holders would want to also hold and stake “short mAAPL” token to guarantee that they will continuously receive reward. That way, they will be able to contribute to the liquidity pool and earn rewards, but also have a very small exposure to the price of AAPL

E) This reward system is not suspect to manipulation. Given the current fee structure (0.3%) and reward paying frequency (~every hour?), the cost to manipulate who gets MIR rewards is going to be too high

1 Like

Yes, I totally agree with your opinions.

I think we already have “short mAAPL token” which is actually the CDP.

From doc:

Collateralized Debt Position
New tokens for a listed mAsset can be minted by creating a collateralized debt position (CDP) with either TerraUSD (UST) or other mAsset tokens as collateral. 
The CDP is essentially a short position against the price movement of the reflected asset,
\i.e. if the stock price of AAPL rises, minters of mAAPL would be pressured to deposit more collateral to maintain the same collateral ratio.

So all we need is some mechanism that paying interest to/from longer/shorter to peg price

I think you are right too! Great!

This also means someone can mint a coin, put the coin in the LP and stake the CDC, and achieve stock-neutral (for now) position with providing liquidity

Hi guys,

Just learned about this project. Wanted to drop my 2 cents here. In Fantom they have the opposite problem, users can mint fUSD which is a USD derivative from Fantom tokens, collatorized at 300% by default. If the C ratio is above 500%, then the users will get rewarded with Fantom tokens. Currently there’s no liquidations, so this caused the fUSD value to be 30% bellow actual USD, which is due to the oversupply of fUSD and it being minted.

So yes, rewarding minters will work and I also think KC_crypto’s funding fee idea will work as this is how current crypto perpetual futures exchanges keep the price not higher than the asset price.

Thanks to @Cryptozyme,
Proposal is formalized here: [Proposal] Dynamic reward distribution