Introducing @GammaSwapLabs . I hate IL, you hate IL, LPs hate IL, IL sucks. Now with GammaSwap, you have a way to hedge IL!

1/ IL stands for impermanent loss, the amount you have lost by providing liquidity due to the changing ratio of assets as price moves instead of just holding the two assets. IL was pretty bad when AMM's first came around, but Uni V3 and concentrated liquidity made it even worse.

2/ If you are LPing, you are essentially short volatility of the pair of tokens. This is because you do not want one token's price moving too much relative to the other, as this would cause IL. The larger this price move, the larger the IL you experience.

3/ So if you want to hedge your IL, you want to be long volatility, meaning you make money as the relative price moves. The question becomes, what type of instrument will allow you to make money as the relative price moves?

4/ Well we want whatever the opposite of LPing is right! Cause if he's losing money as the relative price moves due to IL, the opposite thing is gaining money as the relative price moves due to IL.

5/ So how do you create this "opposite" position? In general if you own something, you are long it. So LPs are long a liquidity position. And the opposite of a long is ... a short! And so we want to find a way to be short a liquidity position.

6/ You can't exactly just short a LP token right now. However, through lending & borrowing markets, you can short something. If you borrow something and then sell it, you are basically short whatever you borrowed!

7/ Back to GammaSwap and this "opposite" position. We have figured out we want the ability to "borrow" a LP position. How could we do that? Someone has to lend it for us to be able to borrow it. Who will lend it to us?

8/ Maybe when someone wants to add liquidity on Uniswap, they could lend us their LP token. Say a user wants to add liquidity in the ETH/USDC pool. Rather than doing it through Uniswap, he would add liquidity through GammaSwap.

9/ GammaSwap would then deposit that ETH & USDC into Uniswap and keep the Uni LP token. They will also issue a GammaSwap LP token to the depositor so that he can claim it back when he wants to.

10/ Now GammaSwap has a Uni LP token that someone can borrow. Like any other time you want to borrow something, you would have to post collateral. In this case, a user would have to deposit ETH & USDC as collateral.

11/ Once the user adds collateral, to open his borrow position, GammaSwap redeems the Uni LP token for the underlying tokens. So now GammaSwap has a position of ETH & USDC, and the original lender has a synthetic LP position in the Uni ETH/USDC pool.

12/ Know what the difference between a ETH & USDC position and a synthetic ETH/USDC LP position is? Wait for ittttt. IMPERMANENT LOSS.

13/ The lender that originally created a Uni LP position has the payoff of a LP position that includes IL, but the actual tokens are not in a liquidity pool, and so do not experience IL. GammaSwap takes that difference and pays this impermanent loss to the borrower.

14/ GammaSwap calls this impermanent gain. The borrower receives the impermanent gain. What's in it for the lender that originally deposited ETH & USDC for a Uni LP position?

15/ First of all, GammaSwap uses an index to track the fees assuming the original Uni LP position still existed. So the borrower has to pay what the trading fees would have been to the lender.

16/ In addition, the borrower has to pay an interest rate that increases during periods of higher volatility and IL to the borrower. So you get the payoff of a Uni LP position + an additional interest rate. More fees!

17/ All of this can be summarised by this diagram.

18/ As usual, if the borrower doesn't have enough collateral, he will get liquidated. The liquidation threshold seems to be 90% LTV. Now the cool part is that because the collateral you post and the LP position you borrow are in the same assets, it's hard to get liquidated.

19/ The only way to get liquidated is from trading fees and interest on their loan. Assuming you have two tokens, the value of the LP position of the two tokens will always be less than/equal the non-LP position (collateral) due to IL. So hard to get liquidated from price moves!

20/ A few observations. You can't provide liquidity while hedging IL for a IL free LP position because you'll be paying the trading fees to the user you are borrowing from anyway so it nets out.

21/ More people probably want to hedge IL during periods of high volatility. Which means that more Uni LP positions will be redeemed. Which means that there is less liquidity. Which probably means more volatility in the pool.

22/ If GammaSwap becomes really big, this could be a problem. Feel free to correct me if I'm thinking about this wrong. A simple solution could be to limit concentration, e.g. GammaSwap sets a parameter that doesn't allow their protocol position to be more than 20% of a DEX pool.

23/ All in all, a super cool protocol that has utility as long as AMM models that have IL exist. Even though I believe the future is an on-chain CLOB, this is a much needed stop-gap in the mean time.

Tagging some folks for visibility
@BarryFried1
@crypto_condom
@Riley_gmi
@DAdvisoor
@ChadCaff
@Crypt0_Andrew
@ologai
@crypto_linn
@hodlKRYPTONITE

You can read the unrolled version of this thread here: https://typefully.com/purplepill3m/EIUmSgR

GammaSwap: The Ability to Hedge IL

The number of DeFi protocols that have managed a successful liquidity incentive program is few and far between. We've tried vote escrow (ve) models, Olympus DAO's infamous (3, 3) and a mix of everything in between.

I believe we have seen more than enough graphs of how these normally end both in terms of TVL and token price. The team at @ByteMasons behind @Reaper_Farm have created a new incentive system called Reliquary.

The key part of their new system is the increasing convexity of the rewards. Most liquidity rewards are currently linear, which means that if you LP for twice the time, you get twice the rewards.

For full transparency, similar systems exist out there with various use cases.  Examples include GMX multiplier points, Liquity Chicken Bonds, Osmosis bonding, etc.

An example of a convex function would be something like a square function. If you LP for twice the time, you get four times the reward, if you LP for three times the time, you get 9 times the reward.

From a technical perspective, Reliquary uses maturity tranches, similar to step functions in mathematics. This means that for each period of time LP'd, the protocol gets to decide what the incentives for that period should be.

This is cool because it adds granularity to LP incentives. For example, data analysis has shown that on average LPs leave after 15 days. Protocols can design their maturity tranch so that the liquidity incentives only start increasing after Day 15.

Importantly, this reward distribution mechanism is done in a non-dilutive manner. Meaning that the total rewards being emitted within a pool always stays constant, regardless of how many users there are or how long they have been LPing for.

Some extra cool parts, each user gest a NFT that represents their LP position. As usual, when a NFT gets issued. You can lend it, lock it or sell it. Extra functionality!

Assuming a liquid market develops around these NFTs, it would be interesting to see how ppl price them given the future reward flows from how the maturity tranches are defined and their associated rewards. Probably some fancy discounted cash flow models you can build.

Another interesting part of Reliquary. Normally emissions are baked into a masterchef contract for the overall protocol. In Reliquary's case, the emission rate is defined within a separate contract. If you wanted to change the emission rate, you could upgrade this contract only.

If for example, a user deposits more tokens into an existing position, the contract calculates a new weighted position maturity, taking into account your new deposit amount relative to your existing position.

e.g. if you deposited 10% more tokens, your position maturity tranche wouldn't change that much. I suspect this method probably made a lot more sense than issuing a new NFT for the second position. I would also assume your NFT updates whenever you change your position.

ByteMasons CEO @0xBebis_ has made a proposal to SushiSwap to adopt Reliquary's emission model, especially when SUSHI has reached a 98% distribution rate, meaning the Sushi team has to think very carefully about how to effectively use the last 2% of token emissions.


Governance proposal attached below. https://forum.sushi.com/t/sushiswap-x-reliquary-rethinking-emissions-for-the-long-term/11533

All in all, improvements in liquidity mining programs are always exciting to see given how broken they currently are. Will be interesting to see how different protocols adopt this and design different maturity tranches, hopefully in a data driven manner.

You can read the unrolled version of this thread here: https://typefully.com/purplepill3m/vazdlLt

GammaSwap: The Ability to Hedge IL

Reliquary: A new token emissions model