[FIP - 142] Frax Gauges on KyberSwap Elastic Ethereum, Polygon, Arbitrum and Optimism

[Kenkyber]

1. Summary:
The outcome of this proposal is to make a FRAX pool one of the base pools on Kyber AMM’s stable swap on 4 chains Ethereum, Polygon, Arbitrum and Optimism.

In addition, Kyber has also had a successful campaign supporting Lido wstETH and stMATIC liquidity on 4 chains, and would like to run a similar campaign for Frax sfrxETH.

Through a combination of diverse pairings as well as our custom fee tiers, we can drive more volume through FRAX pools as well as sfrxETH pools.

Background and motivation:

KyberSwap Elastic AMM is a new tick-based AMM with high efficiency pools and liquidity settings which has ranked among the top DEXs across all chains in just four months since launch. With concentrated liquidity, Kyber performs especially well with stable pairs, as proven through the performance of its QiDAO $MAI pools and Lido $wstETH and $stMATIC pools.

Currently, there is a significant amount of TVL locked in both UniSwap and Curve, but this TVL is not being put in an efficient use where it can generate a decent amount of volume. In addition, Frax just launched wstETH-sfrxETH-rETH pool on Balancer with 1M of TVL but the pool does not have any volume till date. With FRAX pools as one of the stable base pools on Kyber, this can be fixed.

2. Kyber Recommendation:

We recommend gauges for the following pairs, which are built to optimize for and increase FRAX and sfrxETH volume:

Ethereum:

Add these following pools:

• FRAX - USDC (Fee tiers 0.01% or 0.04%)
• FRAX - frxETH (Fee tiers 0.04%)

And add one of the following pools:

• frxETH - ETH (Fee tiers 0.01% or 0.04%)
• sfrxETH - wstETH (Fee tiers 0.04%)
• sfrxETH - FRAX (Fee tiers 0.3%)

Polygon
Add this pool:

• FRAX - USDC (Fee tiers 0.04%)

Arbitrum
Add this pool:

• FRAX - USDC (Fee tiers 0.04%)

Optimism
Add this pool:

• FRAX - USDC (Fee tiers 0.01%)

3. Benefits:

• Increased liquidity efficiency: Liquidity is not only optimized for volume across more diverse pairs and custom fee settings, but also benefits from custom routing through KyberSwap’s DEX Aggregator.
• Partnership with a reputable and long term AMM protocol, which currently ranks 5 to 6 out of all DEXs across all chains (rank 2 on both Arbitrum and Optimism), in terms of daily volume. Current daily volume is at around $70m.
• Co-marketing with a project which has a strong community and good traction across the four proposed ecosystems.

4. About Kyber

Our product KyberSwap - an AMM model - also shares several product features as UniSwap V3 in terms of concentrated liquidity, utilizing the tick-based method and using NFTs to represent liquidity positions.

The key differences are:

a. Re-investment curves: Whereas the LP fees on UniV3 are only claimable separately, KyberSwap Elastic has a reinvestment curve: LP fees on KyberSwap Elastic are claimable in a separate compounding pool [full range 50/50], meaning more returns for LPs.

Compounding comparison:

• UniV2/KyberSwap Classic: Fees compound into original position, most earnings
• KyberSwap Elastic: Fees compound separately, middle earnings
• UniV3: Fees don’t compound, claimable separately, least earnings

b. Just-in-time (snipping attack) protection: KyberSwap Elastic also offers snipe/just-in-time protection. When a swap is executed, one can supply liquidity, swap into his own liquidity at a narrow range and skim the pool of fees, so the pool LP makes less:

• More IL because less fees.
• Some aggregators might use JIT as a way to find the best rates.
• Somewhat similar to a front running attack.

Our performance with LM has proven to be very strong on volume/tvl ratio: with only 54M in TVL we can facilitate almost 100M in daily trading volume.

5. Voting instruction:

• For: Add gauge
• Against: Do nothing

[tan]

Vote no, immediately. Frax protocol has its FraxSwap why should use Kyber and Gauges???
it doesn’t make any sense to me.
If shared fee with fraxswap, Could be done for some pairs.
“ranks 5 to 6 out of all DEXs” lol, fraxSwap wants to be ranks 4 out of all DEXs

[amirnader]

This proposal is up for voting here: Snapshot

[Oxlorentz]

How does Kyberswap Elastic differ significantly from Uni v3 which also allows ranged LP-ing and variable fees? And what is the reason for Elastic’s higher volume/tvl? Shouldn’t UniV3 be able to offer this as well, just a function of what range LP’s decide to provide for.

A majority of FRAX liquidity already exists on Uniswap, Curve and Fraxswap. I worry adding Kyber would further fragment liquidity.

Is there any more data on the expected benefit to frax for further frax liquidity in Kyberswap as a result of this gauge?

[sashamai]

How does Kyberswap Elastic differ significantly from Uni v3 which also allows ranged LP-ing and variable fees?

First, in January 2019, we launched the Kyber APR, introducing capital efficiency/concentrated liquidity to DeFi. We later improved on the Uniswap V2 model by adding concentrated liquidity on top in our AMM - KyberSwap Classic. Uniswap then came out with UniswapV3, and we liked their tick-based approach. We built KyberSwap Elastic from original code, adapting the tick-based approach, and made additional improvements. Here is a blog: KyberSwap Elastic Vs Uniswap V3: A Comparison | by Kyber Network | Kyber Network

A few major differences outlined in the article:

• Reinvestment curve: LP fees automatically compound into a separate 50/50 pool to maximize earnings, while Uniswap V3 fees do not. This also serves as full range liquidity as support to the main pool(s) for that pair.
• Just-in-time attack (JIT) protection: People tend to agree that sandwich attacks are a significant problem. However, JIT, where attackers put in and pull out their liquidity ‘Just in time’ (before and after a huge swap happens on the pool) to earn huge trading fees without having to deal with IL risk, is even more significant:
  • The cumulative profit of the JIT attacks in the ETH/USDC 30bps pool on Uniswap, for example, is 93 times more than the cumulative profit of the Sandwich attack in the same pool ($1.212M vs. $13k). See more data here: JIT in UniV3
• Another major difference is that KyberSwap is also a DEX aggregator (in fact, we invented DEX aggregation in 2018). We have the ability to whitelist connector tokens such as FRAX or frxETH, which allows FRAX assets to support hops on trade routes like other major DeFi tokens [WBTC, ETH, major stables]. Uniswap is not as flexible on this.

And what is the reason for Elastic’s higher volume/tvl? Shouldn’t UniV3 be able to offer this as well, just a function of what range LP’s decide to provide for?

• This is a good question. It is obvious why we would have higher volume/TVL ratio than several other major DEXs, but why UniSwap V3 as well? It is related to a few other differences I did not mention above.
• We attribute our high performing capital efficiency not only to our concentrated liquidity and custom fee tiers, but also mainly to our innovative farming contract: [Link A in next reply] (allowed links limited)
• First, there is a difference between TVL and liquidity:
  • Liquidity is the amount that you can trade at a price. TVL is the total amount that you can trade (sum of liquidity at every price).
  • KyberSwap Elastic farms are meant to incentivize liquidity and concentration more than TVL, which increases volume/TVL ratio.
• Uniswap might have the ability to customize fees like KyberSwap does, but the impact on volume/TVL from fees is actually a lot less than liquidity and concentration. From our research so far, there are 3 things that affects Volume/TVL
  • Concentration level (the effect of our farming contracts on user behavior) → it may cause more than x10 difference in volume/TVL
  • Fee tier → it may cause x2, x3 difference in volume/TVL
  • Arb/MEV “bot” Adoption → x1.5, x2 difference in volume/TVL
• Lastly, it’s also about setup and operations. We pay particularly close attention to liquidity pool setup, diversification of pairs without too much fragmentation, sustainability and operations; and don’t just let the liquidity pools run or sit without attention. Sometimes, we diversify pairs to enable more volume or add different fee tiers, but this depends on what our partners want. The proposed setup is what we had already discussed and designed based on feedback from the FRAX team.

A majority of FRAX liquidity already exists on Uniswap, Curve and Fraxswap. I worry adding Kyber would further fragment liquidity.

First, as the inventor of DEX aggregation, we already solved the biggest problem of fragmented liquidity across multiple DEXs. Second, as a stablecoin, I believe FRAX would want as much adoption as possible. I understand a certain amount of deep liquidity in a pool is required to qualify as AAVE collateral, but we don’t aim to replace. The pool types in our proposal serve to increase the depth and decrease the slippage around the actively traded price. It does not serve as deep liquidity on price fluctuation.

Is there any more data on the expected benefit to frax for further frax liquidity in Kyberswap as a result of this gauge?

I don’t have specific data for FRAX itself, but it would be similar to other stable pairs. Here is a reference with Lido across the same four chains. I’ll insert some Coingecko markets links for reference.

• stMATIC markets: [Link B in next reply] (allowed links limited)

• wstETH markets: [Link C in next reply] (allowed links limited)

Here, we also have another metric analysis which is volume/rewards, adjusted for fee tier difference:

This is the comparison between two similar pools: KyberSwap wstETH/USDC vs Balancer wstETH/bb-a-USD. We can conclude that KyberSwap Elastic requires much less farming rewards for a unit of volume. (the difference is about 4 times) Considering that the 2 pools have different fee tiers → we calculate the reward per unit of fee where Kyber requires 47% more reward.

If there are further doubts or suggestions, we are flexible with our proposal.

We have about 16 hours left for the vote. Hope for your support.

If you need to reach out for a quick question, my TG is @sashamai and my twitter is @Mai_DeFi. I am the Head of Business Development at Kyber Network.

[sashamai]

Link A: Our farming mechanism as described above:

Kyberswap Docs | Kyberswap Docs

[sashamai]

Reference to similar campaigns.

Link B: Lido Staked Matic Price in USD: STMATIC Live Price Chart & News | CoinGecko
Link C: https://www.coingecko.com/en/coins/wrapped-steth#markets

[ssmccul]

Proposal needs to be atomized and each gauge should be voted on independently.