Collection of changes for v2

Last edit: 2/10/2021

Overview

Mirror protocol v1 has resoundingly proven market demand for synthetic price exposure to real world assets, and has managed to self-organize an amazing community around itself. But it is not without its problems. In the short existence of the protocol, the community has identified many areas of improvement around the protocol & suggested potential fixes on and off this forum.

The goal for this document is to succinctly summarize the key issues of Mirror v1, and to collect the best solution to these problems that would culminate in a v2 protocol upgrade proposal. This is intended to be a living document to address problems & solutions as they arise, so if a problem area you think its important has not been included, please do leave links to relevant forum posts in comments below.

As the community forms a consensus around each of the problem areas, I will start to tag each problem with the tag “consensus”. Other states will be “unsolved”, and “debating”.

Mirror v1 key problems

  1. Premiums persist: mAssets consistently trade at a premium to underlying stocks. There is little incentive for users to try to bring them down
  2. New assets dilute liquidity rewards: new assets being added dilute liquidity incentives of existing pools
  3. Low incentive for gov participation: Rewards for gov staking is lower than for liquidity participation. For users already staked, voting actually results in a penalty, as it locks up tokens until the vote has expired.
  4. Listing process is too slow: GME, AMC were attractive listing targets when the proposals were made, they are no longer as attractive now. How can we list mAssets faster?
  5. Limit orders: Automated trading requires limit orders. How?
  6. Margin trading: It would be attractive for mAssets to trade at margin
  7. Pre-IPO contracts: This would be a cool-to-have.
  8. Shuttle is centralized

Mirror v2 solutions

  1. Premiums: 1, 2, 3. Most suggested solutions attempt to create incentives for inverse tokens, to absorb mAsset farming demand & allow users to hedge positions better.
  2. Reward dilution: This is actually an ok problem in my opinion - we just need to make sure the rewards on ethereum fall proportionally with the pool incentives on Terra to maintain parity.
  3. Low incentive for gov participation: No outstanding proposals. We should 1) only give out rewards to users that vote on at least [1/2] outstanding proposals.
  4. Listing process is too slow: 1 - reduce the effectiveDelay parameter to 1 day from 1 week
  5. Limit orders: No outstanding proposals.
  6. Margin trading: No outstanding proposals.
  7. Pre-IPO contracts: 1 (@Sihyeok please write up and link)
  8. Migrate Shuttle to Wormhole: Decentralized PoA bridge

Looking forward to thinking through to a rock-solid v2 together, folks!

Comments below →

14 Likes

Solid assessment all around. Instead of voting to whitelist new assets, why not allow for folks to come together and pool with some minimal amount. I could see a situation where some unattractive asset would attract negative votes based on emotion, but if a sizable amount of people wanted to put together a LP for it that was say… $1MM in total, then it would auto qualify.

Perhaps the margin trading and limit orders should be a completely different product or project. Keep the mAssets functionality stand alone and let other opinions compete for the best system for each.

4 Likes

Makes sense!

I’m worried this might lead to situations where small pools use faulty “rugpull oracles” that disadvantage people that invest later …

What do you think?

Also not immediately clear whether these pools should be incentivised by MIR liquidity rewards.

1 Like

Would it be possible to manually adjust the liquidity incentives per asset? If we take a look at the current assets, they are less hype sensitive, but generally in high demand (seen the volume growth).

Assets such as GME were in high demand for a really short period (a few days). It wouldn’t make sense IMO to stream liq incentives equally to for example Twitter or Apple.

IMO it would make sense to create a certain incentive to hold MIR and be an active governance participant. The current engagement on the proposals are really low. This could be something like a multiplier on their rewards based on how much they vote / engage.

4 Likes

This is true in term of MIR but doesn’t take into account volatility in MIR price and incentive to accumulation of MIR. I think that more pools means more rewards for MIR staked in governance, more accumulation and therefore it could happen that doubling the pools will not affect in an extreme way lp rewards in term of ust.

1 Like

In terms of the M-asset values being at a premium.
The main issue is supply. We need to increase supply of m-assets to meet demand. That means more minters. A Minter only really benefits from minting an asset if that asset decreases in value as they can use minting as a short. Other than that their opportunity cost of minting is very high at over 150% of the underlying assets oracle price.
I propose we actually reward minters as they are just as important as stakers. I feel like a certain portion of LP rewards should go to minters. We definitely do not want to add more tokens as inflation is already incredibly high but taking away rewards from liquidity pools and giving it to minters makes sense.

I like the proposal for Governance participation, I have been pushing for this in the telegram for a while now. I think having to vote for half of active proposals is a little too stringent. I think rewards should accumulate if you have voted for one current proposal. This will at least make longer term holders come back and read the proposals to vote on continuously.

2 Likes

What about this?

  1. only give out rewards to users that vote on at least [1/2] outstanding proposals.
4 Likes

Minting in and of itself doesn’t solve problems - what if the user simply holds the tokens he has minted? Then the masset price doesnt return to the peg, you’ve just given rewards to folks that just parked stablecoins (take no risk).

The short / masset token structure incentivises the yield maximizing minter to sell either the masset or the inverse asset to place tokens in a pool with higher yields (will likely be on average the inverse token pools as they are more risky)

1 Like

I misunderstood your answer I thought you meant half, as in (1/2), proposals.

Minters do not take risk but there is a cost, they could earn interest on a number of platforms with that stable coin.
It is not a bad thing to have stable coins staked, especially if it is your own stable coin effectively taking it out of circulation (as well as the luna that was burnt to create it). It is also not necessarily a bad thing for people who want to invest in M-assets to mint their own m-assets. the cost to them is 150-200% of the value of the asset BUT the reward is possibly some MIR tokens.

Sure. I don’t understand how this solves the mAsset premium problem though.

So some of the people that would have just bought on the market would now mint. That will have an impact on the price.

Another possible way to incentivize minters is to possibly allow them to get an airdrop of a future token.
I feel minters are the people that make the whole platform work and therefore should be incentivized to mint. Instead they not only have to lock up 150-200% capital during the biggest bull run ever seen, they also get charged a 1.5% fee when burning their m-assets. (don’t get me wrong I think the burning cost is good as it stops people burning m-assets)

In an old topic I was suggesting that the premium is mainly an effect of high farming rewards. It can go away just with an increase of pools (new massets) faster than mir value increase.
My suggestion is to start to register parameters of the whitelisted new assets in a pool using community funds (to propose the param registration poll in governance) and release them when looks ‘necessary’ (example when the spread seems to be too high we could decide very fast to add 5 massets: they would be ready and tested in the pool).
We have a lot of whitelisted assets but it looks we have somehow forgotten about them …

2 Likes

Mirror v1 key problems

  1. Premiums persist : mAssets consistently trade at a premium to underlying stocks. There is little incentive for users to try to bring them down
  2. New assets dilute liquidity rewards : new assets being added dilute liquidity incentives of existing pools

I think listing more mAssets will fix this issue. Currently there is no incentive to sell or mint mAssets for a three reasons: (1) The overall trajectory of the markets is upward, (2) large cap stocks are not volatile and (3) the LP rewards are too high. In this situation there is no risk. Even if the mAsset fell by 20% which is extremely unlikely in this environment the LP rewards would more than compensate for that loss. As a result, it is capital inefficient to mint and provide LP with your minted token and nobody will sell an mAsset when there is no real risk. In this environment premiums will persist. If we list more mAssets the LP reward will fall, there will be risks to providing LP and people will start trading mAssets as intended.

I would also suggest modifying the LP rewards structure. LP rewards should be more weighted to trading volume of a specific pool. It makes sense to have some base MIR rewarded to each pool as we have now just to maintain pools for the assets but we should be incentivizing people to provide LP to pools that are actually used. This way as we list more mAssets LP rewards can still be high enough to attract providers to the platform for the right mAssets. The LP farming dynamic will be important overall to maintain interest in the project in the near term and this will maintain that feature as we add mAssets.

Adding mAssets is essential to the long term success of the platform and in terms of competitive advantage to other platforms that offer similar services. We need to think of a way to do that efficiently and while also maintaining LP incentives. I think the above is a step in the right direction.

5 Likes

I made a separate post here on how we could reduce premiums:

If we allow staked LP tokens to be used as collateral thann people could supply liquidity without pushing the price above peg, because they could just mint the asset they put in the liquidity pool and used that staked liquidity as collateral for the minting. These market makers would be market neutral, but be on the hook for impermanent loss.

1 Like

The one thing that is missing from a trading perspective is the ability to short assets. This would be a deal breaker for many traders unfortunately but perhaps not such a big deal for hodlers and stock value investors. Inverse assets was mentioned as one possible method to fix premiums.

I’m not sure how possible it is but ideally you want a smart contract that works like a perpetual futures contract. So an oracle will provide a price and if the asset is priced above the oracle price then margin longs pay shorts a funding payment and visa versa. The larger the imbalance between net position volume short and long the larger the hourly funding payment for anyone with a margin position providing an incentive for the synthetic asset price to return to the oracle price.

I expect this would be quite difficult to implement unfortunately.

1 Like

@dokwon

For newest spec of Pre-IPO assets, please see here

2 Likes

I have suggestion about Premium problem.
The real problem is no real chance for arbitrage when the Terra price is too premium. Example the arbitrager will sell at Terra price and buy at real world so they hope that the Terra price will come down after that they will buy in Terra and sell in real world. These will finish the strategy.

In fact that, It’s not gonna happen in Terra world because the valuation of stocks in terra are more higher than the real world. Ex. BABA in real world have a little dividend, it makes valuation BABA in long term about 240-260$ depend on opinions but BABA in Terra have a lot dividend from staking so the valuation is about like 1000$ maybe more. So the terra price is undervalue like now so no chance for arbitrager.

Solution
We need to give benefit in reasonable. We need to add staking reward for minter side 1.5 time of buyer side. Example we buy BABA in terra price for staking assume at that time APR is 100%, as the same time if we mint BABA for staking, the APR for minter should be 150%. If we can do this fundamental staking reward for minter side and buyer side, the arbitrager could take benefit when the price is premium.

Why should APR staking of minter side be 1.5 time of buyer side?
Example we buy BABA value 100$ and pool with UST value 100$ for staking APR 100%. It’s mean we invest in BABA-UST(LP) 200$ for reward 200$ per year.
In the same benefit for minter side, we mint BABA value 100$ so give collateral 200$ (assume collateral ratio 200%) after that we give BABA value 100$ to stake. Totally we invest 300$ for staking BABA-UST(LP) value 200$ so The APR of minter for staking should be 150% what it’s mean we will get reward 300$ per year for total invest 300$. These make fare for minter side and buyer side.
If we do like this, the terra price will diff from Oracle price around 1.5% in long run. Because if it premium, nobody wants to buy, they will come to mint for staking. That mean more supply and no demand so Terra price will come down in long run economic. So the arbitrager can happen because they can sure that the terra price will come to no premium in nature of these fundamental.
These just my opinion. It’s a concept. I hope it can help team. If it happens, many people will come to arbitrage a lot. But if it’s not change, A lot of people will come to buy a lot whatever premium +30% or 50%.

Isn’t minting and selling equivalent to shorting?