Mirror Steering Committee member here: We see that premiums are way out of control, we’re working on it. Unfortunately, not much can be done until proper liquidation mechanisms are implemented (ongoing). In my personal downtime I’ve been working on a solution for premiums for a while and people have expressed the idea of creating a Mirror based/owned DEX as a way to address premiums, whether through a dynamic fee structure, concentrated liquidity, or otherwise. Below you’ll find my recommendation as to what I think would work super well to not only improve capital efficiency, but also provide Mirror more options or tools to fix said premiums:
A huge problem with current Mirror is that each premium has to be adjusted relative to individual mAsset conditions. Minimum Collateral Ratio plays an enormous role in how individual pools behave relative to Oracle prices, evident to the recent mKO and mSPY flippening to huge discounts from prior premiums. Attempting to figure out the most optimal MCR relative to each mAsset, or by implementing an entirely external vAMM to a normal AMM to allow new methods of minting, as I’ve discussed previously in two of my articles, can work, but are extremlely complex, risky, and are difficult to implement on-chain.
Enter my creation of a new pooling method using a vAMM. If you haven’t looked over the paper, do so now. By pooling UST into a single, large pool and then by breaking the allocation of said pool down into chunks relative to mAsset pool size according to Oracle Price, premiums become a more solvable problem. By addressing premiums as a global function, reward allocation becomes much more effective at combatting premiums. Within the conclusion section, you’ll notice my reward structure recommendation. In regards to each of the types, I have this to say:
Type 1: This would be helpful in keeping rewards trackable and liquidity available to frequently traded mAssets; there would be very few surprises as a result of the lack of moving parts.
Type 2: This would be extremely impactful in swaying users to provide more UST or mAsset relative to the global premium. The issue with this is that it’s a very crude solution for small premiums. Individual mAssets that are heavily traded would not be prioritized as a result of blanket reward rate adjustments.
Type 3: A combination of the first two types, perhaps with caps, would allow for a more dynamic environment where causal relationships as to what exactly is influencing premiums would be difficult to determine. Liquidity wars would certainly be fought under this method as rewards would dynamically adjust on both UST and mAsset sides alongside the war on premiums. Highly complex, but potentially worth it.
To note, rewards would not need to be allocated based upon premium, but based upon what users vote or decide given that premiums would be global. This way, liquidity is partitioned to pairs that need it. A veMIR could be implemented alongside this pool to allow for such gauging/voting.
Let me know your thoughts.