Enzyme Yield Source for PoolTogether

[teawaterwire]

• Proposal

  • Build a yield source for PoolTogether that integrates with specially curated Enzyme vaults.

• Motivation

  • PoolTogether is at ~200M in TVL (https://defillama.com/protocol/pooltogether)
  • PoolTogether made it possible to integrate any protocol that can serve as a yield source (https://github.com/pooltogether/yield-source-interface)
  • Each yield source requires a new contract that has to be built, audited, deployed and maintained (one for Compound, one for Aave etc) -- making it costly and time consuming
  • Hence the case for meta-strategies as yield sources: where yield sources can be actively managed to achieve the best result
  • Yearn has been building a yield source for their v2 vaults (https://github.com/pooltogether/pooltogether-yearnv2-yield-source)
  • Enzyme’s vaults can be a flexible and more profitable alternative as a meta-strategy for PoolTogether depositors:
    • From WETH to Lido Curve staking
    • From DAI to aDAI Curve staking
    • From one token on Aave to Compound or Idle according to the best yield available
  • If Enzyme’s vaults become popular as a yield source this would increase the total AUM of the Enzyme protocol, ultimately benefiting MLN holders

• Implementation

• Why me?

  • Already working at the intersection of Enzyme and PoolTogether (thanks to https://github.com/enzymefinance/MFP/issues/9)
  • Ex CTO of multis.co (crypto wallet for companies, eFounders, Y Combinator S19)
    • https://multis.co/
  • Fullstack engineer from Solidity to CSS with 10 year experience
    • https://www.linkedin.com/in/teawaterwire/
  • User and evangelist of PoolTogether since v1 (now v3)
    • https://twitter.com/teawaterwire/status/1232794151495897090
  • Active user of Enzyme through the fund "La Crypto C'est Rigolo"
    • https://teawaterwire.medium.com/la-crypto-cest-rigolo-d3e8f224044a

• Code wise

  • Making PoolTogether able to use Enzyme vaults as yield sources boils down to deploying contracts supporting the YieldSource interface (provided by PoolTogether).
  • Two contracts have to be built:
    • The factory contract that deploys new instances of yield sources (one token, one vault)
    • The EnzymeVaultYieldSource.sol that includes the logic and methods from the interface
  • The three main methods of this contract are “deposit”, “balance” and “redeem”.
  • Deposit
    • The PoolTogether prize pool will send tokens to the Enzyme yield source contract
    • The contract will buy shares of the corresponding vault and do some internal accounting to keep track of this deposit from that particular prize pool
  • Balance
    • The contract needs to compute and return the balance in underlying tokens of a particular prize pool
    • This will require for a vault to get the share value in denominated token
  • Redeem
    • The contract will burn shares of the Vault and get corresponding assets
    • The contract will have to swap these assets to the vault’s denominated token
    • And send to the prize pool the amount
  • The objective is to make a github repository similar to the one made by Yearn or Aave for the specific Yield Source implementations. With the usual tests and deploy scripts.

• Operations

  • The vaults that will be used by these yield source contracts require specific parameters:
    • No management fees (to avoid edge cases with negative yields)
    • Only yield positions (going up, or way up)
    • Could be operated by an admin, a multisig or a DAO

• Compensation

  • I estimate a complete solution requiring between 8 to 16 days
    • Including the first step that is a poc to evaluate feasibility (~ 4 days)
  • With necessary back and forth these won’t be consecutive days
  • I suggest invoicing on a half day basis for about $400
  • And sending monthly invoices in case it takes longer (no consecutive days)
  • Payment in MLN tokens at the time of transaction would be perfect

[Deepcryptodive]

I quite like the idea.

Some considerations:

• Who would be running the strategy, and what incentive do they have to do a good job? Maybe a performance fee?

• Who would pay for these potentially costly transactions? deposit, balance, redeem)

• You might want to implement some strict trading policies so the manager cannot enter positions which are not related to the denomination asset (e.g. only allow trading into stablecoins or stablecoin pools, for a USDC based fund). This will require some fine-tuning as to still allow for sufficient yieldfarming options, but avoid (im)permanent loss.

[teawaterwire]

• Who would be running the strategy, and what incentive do they have to do a good job? Maybe a performance fee?

could be someone "elected" by the PoolTogether DAO for instance -- and yes the performance fee sounds like a good incentive (at no more than 2% i think)

• Who would pay for these potentially costly transactions? deposit, balance, redeem)

that's handled by the PoolTogether Prize pool contracts when a user deposit or redeem on a pool

• You might want to implement some strict trading policies so the manager cannot enter positions which are not related to the denomination asset (e.g. only allow trading into stablecoins or stablecoin pools, for a USDC based fund). This will require some fine-tuning as to still allow for sufficient yieldfarming options, but avoid (im)permanent loss.

👍 anything that falls into the "yield" category is a good start, and then also the "provide liquidity" on Curve or stable pairs