SIMP-3: Change Short-Farm Maximum Reward Allocation from 40% to 100%

Proposed

Simp-3 aims to provide a temporary fix to the excessive premiums by adjusting the ‘short-farm’ reward allocation from a maximum of 40% to a full 100% of rewards.

Analysis

While this proposal may not solve the excessively high premiums on its own, it will provide much needed data to the Mirror community and potentially facilitate some reduction of the current premium problem. Upon implementing this change, the Mirror community will be able to assess its effects on both premiums and liquidity to better understand what drives premiums.

Mirror has six main trader types to consider when evaluating the impacts of protocol changes. Below, each trader type is analyzed and broken down to understand their motives and net impact on the protocol.

If you don’t fall into any of these categories, reply below so that your individual case can be considered.

Directional Traders & Investors

For Mirror to succeed beyond the expiry of MIR emissions, directional traders and investors must become the target audience. Building Mirror to be better for these groups is the top priority.

Traders need accurate pricing, low slippage trading, and readily available markets.

Presently, mAsset swap (pool) prices are loosely correlated to oracle prices; it is assumed that directional traders are choosing to use Mirror for the temporary benefit of its yield farms while waiting for their trades to play out. Reallocating farming rewards to either long/short side, depending on premiums, means several things. Active traders might choose Mirror over a Tradfi (Traditional Finance) brokerage for long positions if mAssets trade at a discount (recently mKO, mSPY, mVIXY), and will choose Mirror for short positions if mAssets trade at a premium (all other mAssets). Some trading volume will be lost as trades on the wrong side of premium/discount are taken using tradfi instead; however, if mAsset to Oracle correlation tightens up, their volume will return.

LP Yield Farmers (Long-Only)

Long-only yield farmers are a unique breed. Informed LP’s have a risk framework that evaluates whether the LP rewards are adequate to compensate for potential risks of divergent loss (Impermanent Loss) in the movement of their LP tokens. Mirror had been popular among long-only yield farmers because, in the past, yields were particularly attractive relative to the volatility of the mAssets held. As TradFi market volatility has picked up, and MIR emissions were slashed after the completion of its first year, many of these LPs have moved elsewhere and Mirror TVL has fallen.

Throughout this post, an mNIO LP position will be used as an example, with a collateral ratio of 2.0, or 200%, when minting is involved. At the time of writing, mNIO has 20.48% long yield, 18.59% short yield, and price premium to Oracle of 16%. You can replicate this work with any mAsset with high premiums and see that the principles still apply.

The aforementioned long-only LP with a $10k value would have the following allocation:

  • UST $5000
  • mNIO $5000

On this $10k portfolio, the mNIO LP position will earn $2048 in MIR emissions annually, even though mNIO trades at a 16% premium to oracle price! These attractive returns are one of many possible explanations for why mAssets depegged when traditional assets crashed. LPs are exposed to price risk (but are they really? If mAssets depeg when the underlying crashes, LPs get paid to reduce their price risk!)

The proposed change in the short farm rewards allocation will cause many more of them to leave. Since they are agnostic to Mirror, this isn’t particularly devastating. A mass exodus would push down premiums but leave enough long only LPs to provide decent liquidity for the current level of trading volume. (Especially since we will still have our dedicated DNs…)

Delta Neutral - Mirror Only

Delta-Neutral yield farmers are attracted to Mirror because they are able to take advantage of aUST collateral and stack MIR yields on top of their Anchor yields with minimal additional risk. After following the requisite steps, a classic Delta Neutral position allocation using the mNIO example would look something like this:

  • UST $3333 (held in LP)
  • mNIO $3333 (held in LP)
  • Short mNIO $3333
  • aUST collateral $6666
  • $1 left in wallet to cover whatever fees or because they didn’t calibrate with 100% precision

The simulated position earns long-farm yield on the $6666 UST+mAsset LP, Anchor yield on the aUST collateral, and short-farm yield on the $3333 short farm:

  • $6666 worth of aUST generates $1299.87 annually (assuming 19.5% APY)
  • Short Farm Position ($3333 worth of mNIO) generates $619.60 annually
  • Long-Farm Position ($3333 UST + $3333 mAsset) generates $1365.20 annually

In total, this earns the position $3294.67 per year, or 32.85% APR. This 32.85% APR compares very favorably to the 19.5% APR they would just holding aUST, and the comparison will improve even more when Anchor starts to decrease its yield!

Even with a price premium of 16%, mNIO’s MIR emissions are split between long and short, with 60% going to long and 40% going to short. We can calculate the asset split between long short by using the known yields and MIR emissions split. (0.60/0.2048 = 2.9297 long per 0.40/0.2256 = 2.1517 short, or a ratio $1.36 in the LP for every $1 in the short farms.

If the parameters are adjusted so that a 16% premium causes up to 100% of the MIR emissions to go to the short side, the DN farmer would earn:

  • $6666 worth of aUST generates $1299.87 annually (assuming 19.5% APY)
  • Short-farm position $1111.07
    • $619.60 they earned before
    • $491.47 in additional short-farm rewards
      • Aggregate 1.36x long short ratio means that this 2:1 ratio farmer captures only 36% of their long-only rewards when they become short-side rewards
  • Long-farm position $0 MIR emissions

This $1111.07 reward on a $3333 short position represents an APR of 33.33%, which will be used in a later example.

The DN farmer is still earning 24.11% APR, which outperforms Anchor, but now we’re paying more to short-only farmers to incentivize their reduction of premiums.

Delta Neutral - Opportunistic

The opportunistic DN farmer’s goal is to profit while pushing Mirror Prices to align with oracle. When the mAsset trades at discount (recently mKO, mSPY, mVIXY), they long-LP on Mirror and hedge at a traditional brokerage account.

When the mAsset trades at premium (all other mAssets), they short-farm on Mirror and buy long in their brokerage account.

For the same mNIO, their $10000 portfolio mix looks like this:

  • UST $0 (no LP at TradFi brokerage)
  • NIO at brokerage → $5000
  • Short mNIO worth $5000
  • aUST collateral $10000

They earn:

  • $1950 from Anchor on their aUST
  • $929.50 on their short farm
  • nothing on their TradFi NIO because it pays no dividend.

That’s a combined yield of 28.79%. It pays to be opportunistic, but the strategy is strictly dominated by the 32.85% the Mirror only DN farmer is getting, and it’s a lot more of a hassle to manage (fiat gateways, ugh!)

What happens when 100% of rewards shift to the short side? The opportunistic DN farmer earns:

  • $1950 from Anchor on their aUST
  • $1666.60 on their short farm
    • Using 33.33% short APR we calculated previously
  • Still nothing on their TradFi NIO

That’s a whopping combined APR of 36.17%

As long as we have MIR emissions, we will have our Mirror only DN farmers, but now the rewards for pushing mAsset prices back to Oracle will strictly outperform them, encouraging more opportunistic trading and giving Mirror a chance to restore mAsset price integrity.

When assets are at parity, the opportunistic DN farmers will step aside and leave the yields to the DN farmers.

Delta Neutral - Mirror Only, no LP

There is a second class of Mirror Only DN farmers that strives to improve their capital efficiency on high premium farms by holding a long position, but not pairing with UST in the LP. Their $10000 portfolio mix would look like this (same as the opportunistic DN farmers, except that their activity does not push prices toward parity):

  • UST $0
  • mNIO $5000 (re-bought)
  • Short mNIO $5000
  • aUST collateral $10000

Since they have the same asset blend, their results will mostly be the same as for Opportunistic farmers. For dividend paying stocks, the opportunistic farmers have a slight edge (which means that Mirror is actually BETTER for dividend paying stocks than it is for non-dividend paying stocks, instead of worse).

Opportunistic farmers also pay lower trading fees on their TradFi accounts than these “no LP” DN farmers pay using Mirror only, and they can get better trade execution for large order sizes (an advantage that would be amplified if liquidity pools shrink in size too much in response to the change). Opportunistic farmers can boost their efficiency even further by using options and margin in their TradFi accounts.

These are small advantages for opportunistic farmers, but enough to at least give a shot at outcompeting the “no LP” DN farmers. The activities of Mirror Only “no LP” DN farmers are otherwise not directly addressed or resolved by this proposal.

Degenbox Loopers

This crew uses aUST to mint the lowest MCR assets (using borrow instead of short farm), and sells them to leverage up their aUST positions. Since they’re not LPs and not short farming, they won’t be impacted by the change, and no position calculations are necessary.

Summary/Commentary

In this proposal, we have focused on understanding short positions (and intentionally chose an example with long APR higher than short APR even though the premium is high) because we perceive premiums as being a more urgent problem than discounts, but the same principles apply. On a discounted mAsset, we want to see long LPs (and opportunistic DN farmers) have greater rewards per $ invested than Mirror only DN farmers. It doesn’t matter if we have DN farmers dumping MIR, as long as their strategy is always strictly outperformed by strategies that help drive mAsset prices to align with Oracle prices.

Rebalancing MIR emissions is a short-term solution to an immediate problem. We will still need a long-term solution, but let’s get this minimal effort change done to take some pressure off while we work on the long-term solutions.

We would like to acknowledge the work on this topic that has already been submitted by other Mirror contributors:

@Acidistheanswer -

@Cyptofan

MSC has opted for our own post to stay in alignment with a standard SIMP naming schema, but will be posting a governance poll for this proposal within a few days.

Prepared by: @josephsavage, @EuphoricBadger, @0xCoolGuy

Edit History: @mentions for usernames

10 Likes

Nice thread, encourage everybody to vote it. When will be a proposal?

1 Like

In a few days. Even though it’s been widely discussed, we don’t want people to feel railroaded by this proposal.

That was a concern when we thought MSC had been widely discussed already but people were still blindsided by the narrow window between our official formation proposal in forum and our funding request in governance.

2 Likes

Nice thread, this should encourage people to short and hopefully bring the premium back down.

1 Like

Bravo! WIll be an outstanding yes from me! In the end, i know this took a while to cook up but all arguments aside I hope the previous couple of days really jumpstarted the gears to turn.

2 Likes

Maybe you can stop trying to randomly liquidate users and othewise fucking up the protocol in between calling everyone thieves you disgusting sack of rat shit?

1 Like

Can we please try to keep things civil in here?

2 Likes

jesus, i mean he cant have a fully developed frontal lobe right? sub 20 years old

This would be a great step towards the right direction! Love it!

I believe there is also the other category:
Degenbox loopers who are using short farming
It may be worth for them to wait 14 days to receive short farming rewards instead of plain borrow and sell. Or at least it will be worth for them to wait after 100% rewards go to short side.

2 Likes

This is just estimate, you can not be sure that the LPs will remain with enough liquidity. Will the DN stay in LP with impermanent loss when rewards go to zero? What if they switch to DN no LP?
Wouldn’t be less risky to make adjustment in 2 (or maybe 3 steps)? I suggest to adjust the ‘short-farm’ reward allocation from a maximum of 40% to 70% in the first step to see what is the effect and then eventually move further.
Generally, risk of unexpected results is bigger when the change in the system is bigger. So why not to do it gradually in smaller steps?

2 Likes

Correction of calculation of DN Opportunistic:

The total investment is 10000 in Anchor and 5000 in Nio Brokerage. The correct calculation of APR would divide the returns by $15000 instead of $10000:

If you divide returns by $15000, that’s a whopping combined APR of 24.11%, not 36.17%

1 Like

You get $5000 from the sale of the mNIO, so your math is not correct. Total capital is $10000.

1 Like

Oh, you are right, my mistake.

1 Like

Yes this. I’m sorry I did not make that clear. It can be either journaled or used for something else (money is fungible) but the net capital in the trade is only $10000.

1 Like

Will the DN stay in LP with impermanent loss when rewards go to zero? What if they switch to DN no LP?

Rewards will never go all the way to 0, though. I didn’t address it explicitly, but long LPs also earn the trading fees. They’re currently negligible for most pairs, but that just means that we are subsidizing way more market depth than we actually need!

All of my scenarios evaluated the impact on different strategies while the premiums are still high. If premiums successfully decrease then a normal DN strategy strictly dominates the no LP DN (because half of the time the no LP DN earns hardly any MIR yield).

If they do all switch to “no LP” until premiums are normalized, that increases the slippage costs for everyone (and therefore the competitive edge for opportunistic DN, because they only pay slippage on one side of their trade, and it’s already the side we’re they’re selling at premium anyway). Smaller pools mean more slippage, and the greater your size the more advantage you get for keeping half of the trade in traditional brokerage.

Yes. But if the APR from fees would low, informed DN investor would stay away of LP because of impermanent loss. Me personally would move to DN no LP or into DN oppourtinistic group to avoid imbalance in positions due to AMM balancing if the APR is low.

I understand that the LPs are larger than necessary now and that less MIR subsidizing may decrease the size of pools. Smaller pools will lead to higher APR from fees. That sounds good so far. But I still think that the sharp change from 40% to 100% may lead the system into big oscilations between states “premium high, everything on short side, almost nothing to long side” and " premium lowers to 0, short rewards declines rapidly, LP increases rapidly". So, just please consider implementing smaller change first to see if it works in desired way.

Volume isn’t the best measurement of requisite depth. Size you can trade v slippage is what matters.

By increasing the reward allocation to 100% for the shorts only means that the very high premiums would get close to the oracle price faster → in case of low premiums the 100% allocation would rebalance meaningfully to the long side, so if we lower the high premiums faster, we’d also get the long APR faster.

3 Likes

Is it possible to get a graph to understand the impacts of these changes? Would be nice to see:

  • Premiums → X [-20%, 20%]
  • Yield → Y [0%, 40%]
  • Graph 1 → Comparing short (old) vs short (new)
  • Graph 2 → Comparing staked long (old) vs staked long (new)

Overall, this sounds great. It gives me confidence in the app to see a proposal this well structured.

1 Like