[LRFC] Market Making Proposal - Arrakis PALM

Lyra Request For Comment (LRFC)
1

Simple Summary

Deploy Arrakis PALM to conduct market-making for LYRA/WETH with Lyra POL on mainnet UniV3.

Abstract

Arrakis PALM - Protocol Automated Liquidity Management - built by Arrakis Finance, is a novel liquidity bootstrapping mechanism that taps into the organic trading volume on UniV3. It is the first product built on top of the Arrakis infrastructure.

In essence, PALM helps protocols bootstrap their base asset inventory (ETH, DAI, etc.) and attain deep and sustainable liquidity. The major advantages of using PALM include:

  • Zero incentive: no LM incentive needed, liquidity bootstrapping is done solely via market making.
  • Low requirement in base asset: the initial liquidity can be made of mostly LYRA, and PALM will progressively balance it towards 50% to create an equal buy/sell support.
  • High capital efficiency: by autonomously and actively managing concentrated liquidity, especially once a sufficient amount of base asset has been bootstrapped, PALM can further reduce the slippage for large trades even if with limited overall liquidity.
  • No biased price impact: PALM conducts market making by setting up ranges / limit orders, no swaps involved.
  • Trustless & transparency: Lyra retains the ownership of the liquidity and can withdraw at all times. PALM only autonomously manages the liquidity but can never remove it. All executions of PALM can be monitored on-chain with full transparency.

Here is an example for GEL/WETH that demonstrates the overall performance of PALM and how it’s able to bootstrap and deepen the liquidity regardless of the price action.

Arrakis has been assisting Lyra with liquidity management since long ago (see LEAP-21 and LEAP-44), and we would like to strengthen this long-term relationship with more and better services that can further benefit the Lyra community.

Motivation

Currently, there are two major challenges in terms of LYRA liquidity:

  1. High liquidity rental cost: a large amount of LYRA is given away as LM incentives (5m LYRA annually as in LEAP-44) to LPs that are most probably going to leave once the reward runs low. Renting liquidity is never a sustainable solution and eventually does more harm than good to the protocol and token holders.
  2. Low capital efficiency: Bulk of LYRA liquidity resides in Velodrome, a constant function market maker that only accepts full range liquidity provision. There is also LYRA liquidity on UniV3. Although UniV3 can significantly improve capital efficiency by the concentrated liquidity feature, the complexity and lack of sophisticated management makes the capital efficiency for LYRA still far from being ideal.

All of this presents both a huge cost and a missed opportunity for Lyra. The liquidity incentive could have been reserved for better purposes that contribute to the development of the protocol, and Lyra could have consistently pocketed a handsome amount of trading fees by LPing with high capital efficiency to absorb most of the volume.

To help Lyra save the cost and capture the opportunity, Arrakis proposes to provide Lyra with the full spectrum of market-making services on UniV3 with PALM to bootstrap and create deep on-chain liquidity.

Specification

Phase 1 - Accumulate base asset

Based on the discussions with the Lyra core team, we suggest that Lyra initially deposit $100k worth of ETH and $200k worth of LYRA (roughly 67 WETH and 1845491 LYRA at the time of writing the proposal) into a PALM-managed vault. PALM will pull that ratio towards 50/50 over time.

There are three key parameters that dictate PALM’s behavior:

  • Range size (S): the number of ticks in a range
  • Allocation (L): the percentage of liquidity deployed in a range
  • Threshold (T): the boundary across which a rebalance is called

Screenshot 2023-01-19 132708
Screenshot 2023-01-19 132708905×485 33.4 KB

After running a series of simulations for LYRA/WETH, the proposed values for these parameters are:

  • Range size: 600 ticks (see official explanation of tick from Uniswap)
  • Allocation: 5% total allocation (the rest stays in the vault as reserve readily to be deployed)
    • 1.5% base range
    • 1.0% mid range
    • 2.5% asset range
  • Threshold:
    • Asset rebalance: 480 ticks from asset range limit towards mid range
    • Base rebalance: 120 ticks from base range limit towards mid range

With the values above, the simulation results indicate that PALM can effectively bootstrap WETH for LYRA over 3 months, and at the same time keep the value of the deposited liquidity close to the value of holding.

image
image1760×760 77.6 KB

image (1)
image (1)1760×760 105 KB

Phase 2 - Establish deep liquidity

Once the target ratio of 50/50 is reached, the focus is then on market-making to create deep liquidity for LYRA and to minimize the slippage on both the buy and sell side of the volume.

During the deployment period, the Lyra community has complete visibility into the execution and performance of PALM via a custom dashboard, and retains full custody of the liquidity in the vault, which means that the Lyra community can withdraw from the vault or revoke managing access from PALM at any time. PALM can only conduct market-making with the liquidity deposited in the vault and will never be able to remove the fund.

For the services provided, Arrakis charges fees on two fronts:

  • Management fee: 1% AUM fee on a yearly basis
  • Performance fee: 50% of trading fees generated.

References

Arrakis Finance
Documentation
Audits
Twitter
Discord

Copyright Waiver

All LEAPs must be in the public domain. See the bottom of this LEAP for an example copyright waiver.

2 Likes
2

@barbarossa_Arrakis thanks for putting up this proposal! Moving away from incentives and towards more POL would be of great benefit to the Lyra DAO and treasury. I think this program works best alongside the existing program until we are able to move toward 50/50 liquidity. I would add that Kwenta and Gelato have both successfully launched Arrakis PALM programs of their own. https://twitter.com/Kwenta_io/status/1630564383909883905

3 Likes
3

Interested in this proposal and the way to go long term (protocol owned liquidity).

What do you think about deploying it first into Arbitrum? There is relatively low trading activity there but more than 1M $LYRA are distributed per epoch. Also could be interesting to see how Arrakis PALM works working in parallel to CAMELOT exchange (more liquidity, more arbitrage opportunity between both…). Could create the ignition of LYRA trading activity on Arbitrum.

1 Like
4

Given that the majority of the LYRA liquidity is on mainnet, it makes sense to begin there IMO. Absolutely agree that a path to POL and fewer incentives is the right direction!

3 Likes
5

Appreciate the suggestion :pray:

From liquidity bootstrapping point of view, the more volume and existing liquidity there is, the more effective PALM can be. For LYRA, the trading activity level on Arbitrum is not there yet. Therefore we believe it’s not the optimal place to start PALM with. Of course we can easily spin up another vault on Arbitrum once the circumstances have changed.

3 Likes
6

Thanks for writing this proposal - it looks like a good idea to me. A couple of questions:

  1. Which network would this be happening on? The current Arrakis managed pool is on L1 - I assume it would be the same?
  2. What are the actual transactions needed to be called to create and initialise the pool? With governance V2 these actions are fully automated so they need to be specified in advance.
1 Like
7

Hello ser,
So to answer your questions:

  1. The proposal is to deploy on the mainnet, especially since that’s where LYRA has its biggest on-chain volume, which is the fuel for PALM. We also want to help Lyra reduce or even stop the incentives for LPs there.

  2. There will be 3 transactions needed:
    1st: approving PALM to spend the liquidity to be deposited
    2nd: depositing into the vault
    3rd: filling the gas tank

So far, projects have be depositing with multisig, whose address is also needed for us to build the vault. Once the vault is in place, we will send deposit instructions for the team to follow up on.

I’m not familiar with the automation of executing passed proposals. However, one project we work with seems to have the same setup, and what they did was first transfer the fund to a multisig after the proposal had been passed, and then do the rest as it normally would.

1 Like
8

Have a few comments too:

  1. Into which %fee pool will it be deployed?
    Would recommend to launch it with the 0.3% fee pool. Reason is it will allow higher arbitrage opportunities and will make Lyra’s trade cheaper versus the 1% pool we are actually using.

  2. Arrakis fees. 1% AUM fee on yearly basis seems Ok, but 50% of the accrued trading fees seems quite high considering the liquidity is provided by Lyra DAO and all the process is automated. Also with the deprecation of the LP program a lot of the liquidity would be in disadvantage and could exit/migrate to other programs on other chains (velodrome/OP, camelot/ARB), letting mainnet LP mainly monopolised by Arrakis PALM. I would propose a performance fee of the 25% trading fees, that seems more suitable.

  3. If PALM proposal is deployed, consider migrating the actual Arrakis/Uni program to Arbitrum with a 50% rewards reduction. We need better liquidity on Arbitrum!

1 Like
9

Thanks for the comments ser :pray:
.

We will build the vault and run PALM on the existing 1% fee tiered pool.
The liquidity to be deposited is mostly LYRA, and PALM will bootstrap more WETH over time to pull that liquidity composition ratio to 50/50. For the bootstrapping to be more effective, volume is crucial, especially non-fragmented volume. Besides, our experience indicates that low slippage with high efficiency is a way more important factor for swappers.
.

The fee structure is baked into the smart contract, therefore not possible to change. The product is a full spectrum of liquidity management services, so that Lyra can focus on its core business without being distracted. Automation is to ensure the precision and scalability of the executions, while Arrakis still has to maintain and monitor the vault for matters such as system upgrade, parameter adjustments when necessary / requested, provide insight / suggestion, etc.
.

It is fair to assume that some liquidity will leave once the incentive is cut off, and this is actually the whole point of PALM, i.e. with high capital efficiency, you don’t need to incentivize LPs for deep liquidity. Liquidity is only as good as the volume it facilitates, and using PALM means getting back the liquidity independence with POL.
.

As elaborated above, volume is of the utmost importance for PALM to be effective, and the mainnet volume of LYRA has been consistently higher than other networks. So PALM will first be deployed on mainnet, and later expand to other networks once Lyra community has gained confidence in it after some time of observation.

1 Like
10

Sorry, 50% fee seems too much. It should be easy to adjust as it’s just a parameter. We need to put the funds and kick the actual rewards for LPs while PALM is taking a privileged position and taking a too big cut. It would be a NO for me.

1 Like
11

Appreciate the comment :+1:

I fail to see why the “privileged position” is a bad thing. The liquidity in PALM is POL, so that privilege is directly bestowed on the Lyra Treasury. The whole point is that you only need that to support the existing volume, instead of renting excess liquidity from external LPs. There is no point in renting $1m liquidity for $100k volume.

Whether or not it’s too much, that has to come with a comparison:

  • How much does Lyra spend on LM?
  • How much of that liquidity will still stay as Lyra runs out of incentives?
  • How much of that liquidity actually facilitates the volume, i.e. capital efficiency?

LM farmers essentially take LYRA allocations out of the treasury, while PALM shares the revenue with Lyra, on a no cure no pay basis, i.e. we only get paid if we facilitate the volume.

2 Likes
12

This is NOT intended to deprecate the current rewards program on L1 but rather complement it while the Treasury bootstraps its own POL.

3 Likes
13

I think the last sentence here is a bit misleading, there is a 1% AUM fee regardless of volume facilitated, correct?

1 Like
14

So in general I would say I see more pros than cons for the proposal. The 50% fees seems high to me but open to give it a chance.

  • Would it be better to set a trial period for the program’s duration? Let’s say 4 months to see how it performs?

  • Does LYRA DAO have 100k $ in ETH? Why not use a USDC pair (DAO treasury has ‘‘plenty’’)?
    For the DAO would be great to accumulate USDC as an hedge to the market, also the core contributors budget is in USDC and MMVaults are in stable coins.
    Also if the original Arrakis program is still continuing on L1 it will allow a higher trading volume due to the arbitrage opportunities between the LYRA/USDC and LYRA/WETH pools. More trading = more fees for the POL.

Some other observations:

  • Decrease the original Arrakis rewards on L1. We will not need so much if we launch the Arrakis PALM program.
  • Launch a similar program (original arrakis liquidity mining) on Arbitrum. We need more liquidity there
  • The time to launch this is asap as there will be not much $LYRA issuance as the MMVaults rewards will be in $ARB and $OP tokens during the next 3/4 epochs. It will be a good time where the lack of new $LYRA hitting the market will help to increase the $LYRA value and absorb the ratio swap from 2/3 $LYRA to 1/2 $LYRA proposed by the PALM program.
1 Like
15
  • Main trading volume for $LYRA is still on Optimism.

It benefits for the big liquidity it has in Velodrome (LYRA/USDC) but also because of 2 other pools with different pairs allow arbitrage due to high variance in the daily prices LYRA/OP and LYRA/WETH

Note that the liquidity on Velodrome is not costing a cent to Lyra as it comes from being an original Velodrome supporter.

image
image1212×421 31.6 KB

red= pools on Optimism
blue= pool on arbitrum
black= pools on mainnet

1 Like
16

Yes, my bad :pray:
Since that’s such a small fraction (we only get 5k if a protocol deposits half a mil), I often forgot about it, lol

2 Likes
17

yeah, was only looking at volumes on Uni. my bad for not making it clear.

i think it’s still beneficial to get on mainnet first, to avoid competing for volume with velodrome for now, since it is undeniable that there is much more liquidity in velodrome for LYRA, and even though PALM is supper capital efficient, it’s still not a magic pill.

1 Like
18

What’s the general opinion about launching it as LYRA/USDC instead of LYRA/ETH?

I think would be better for the DAO the use of USDC as treasury doesn’t have WETH. Also as the original arrakis LM program will be still running (LYRA/ETH) it will benefit from the arbitrage opportunities between both pools.

1 Like