SIP-112: ETH Wrappr

Author
bojan, kaleb
StatusImplemented
TypeGovernance
ImplementorTBD
ReleaseTBD
Discussions-Tohttps://research.synthetix.io/
Created2021-02-15

Simple Summary

Allow users to wrap ETH or WETH and get sETH

Abstract

We propose to deploy a new contract that accepts deposits in ETH or WETH and in return mints sETH in the user's wallet. After successful ETH or WETH deposit, users' ownership is transferred to Synthetix protocol. Wrappr is a conditionally bidirectional contract, meaning that if it holds WETH in its reserve, it will accept deposits in sETH and releases WETH or ETH back to the user. WETH or ETH can be withdrawn from the contract at any point by anyone after the corresponding amount of sETH is effectively burned.

Motivation

We are experiencing a chronic shortage of Synths which is reflected on the market where Synths are being constantly traded above their peg since mid-2020 as can be seen here on CRV. Even with the introduction of Multi-Collateral Loans in SIP-97, problems with the peg persisted. One explanation is that those loans are relatively capital inefficient since they require over-collateralization and thus are not effective tools for arbitrage and bringing synths to peg.

Specification

Overview

sETH is freshly minted whenever a user deposits ETH or WETH into the contract. The user can deposit any amount desired however, this is subject to not exceeding the maxETH configurable via SCCP. There is no duration, interest rate or collateralization ratio, as any user can at any time buy back WETH or ETH available in the contract by burning sETH. Minters benefit as minting sETH and burning sETH are subject to a mintingFeeRate and burningFeeRate, both of which are paid to the fee pool after conversion into sUSD. The contract supports both ETH and WETH, and automatically handles wrapping/unwrapping ETH for the user.

Rationale

The main purpose of this proposal rests on the following points:

  • Allowing ETH holders to easily enter and exit Kwenta with the least amount of friction possible. This venue is more efficient than trading ETH for synths on an AMM, due to the slippage. In addition, the current loan program bears collateralization as mentioned previously which cannot be lowered due to the collateralization and interest rate levied being undifferentiated across different borrowing options.
  • The contract would release sETH during times of surging demand and absorb extra synth supply when that demand recedes. Therefore it would automatically act as a peg balancing layer that allows arbers to enter and exit with the least friction possible, helping align the different pegs. One arb example, would be to mint sETH against ETH and swap the sETH for ETH on CRV.
  • The debt pool would be mostly neutral initially as the increase in sETH supply would be offset by the increase in ETH locked in the wappr contract. However, as users swap the freshly minted sETH to other synths, this would lead to a reduction in the sETH skew as ETH locked acts as a counter-weight against the long sETH position, as the total debt calculation would be reduced by the amount of ETH locked in the contract. Hence, the need to incentivize farmers to assume a short sETH positions would be decreased.
  • The minting and burning fees effectively create a new revenue stream for minters who would effectively be getting a cut from the profit generated by actors arbitraging points of inefficiency.
  • The pressure to fix the peg by reducing the collateralization ratio (increasing the solvency risk on minters) is somewhat reduced by having this contract at the spartan's council disposal, as another tool to help steer the peg towards parity (by changing the minting & burning fee as well as the maxETH parameter).
  • Supporting WETH makes the contract compatible with L2, specifically the Optimism OVM (OVM doesn’t support native ETH, instead contracts can use a native WETH deployment), as well as reduces interaction complexity for users coming from DeFi, where WETH is the norm.

Technical Specification

Contracts

Two contracts are required to be deployed:

  • EtherWrapper.sol which is able to mint sETH against WETH deposited and release WETH against sETH burned.
  • NativeEtherWrappr.sol which is a thin wrapper around EtherWrapper.sol, that supports interacting with native ether on L1.

Interfaces

The entry points for users on IEtherWrapper.sol implements the following interface.

interface IEtherWrapper.sol {
    function mint(uint amount) external;
    function burn(uint amount) external;
}

The NativeEtherWrapper has a similar interface. The contract calls into EtherWrapper for mint and burn, converting sent ETH into WETH for the mint function, and received WETH into ETH for the burn function.

interface INativeEtherWrapper {
    function mint() external payable;
    function burn(uint amount) external;
}

L2 Specification

NativeEtherWrapper is not deployed for L2, as using native ETH opcodes (callvalue, balance) in the OVM will revert.

Key Bounds

  • The upper bound on the amount of Minting is determined with a helper function, capacity, defined by max(maxETH- WETH, 0) with WETH being the amount locked in the contract. In case the user attempts to mint an amount greater than the upper bound, then capacity is minted and the residual is returned to the user (please refer to test cases for calculation specs).

  • The upper bound on the amount of Burning is computed as WETH locked in the contract multiplied (1+burnFeeRate) (please refer to test cases for calculation specs).

  • mintFeeRate and burnFeeRate are both bounded between 0% and 100%, inclusive.

Test Cases

  • Given that a user has u amount of ETH and u WETH and the contract has c amount of ETH in spare capacity

    • Given that u is larger than or equal to c
      • When the user attempts to deposit u ETH into the contract
        • ✅ Then it succeeds and the following take place:
          • c ETH is wrapped into WETH and is locked in the contract
          • c(1-mintFeeRate) is minted into the user's wallet in sETH
          • c*mintFeeRate worth of sETH is sent to the fee pool in the form of sUSD
          • u - c worth of ETH is refunded back to the user
      • When the user attempts to deposit u WETH into the contract
        • ✅ Then it succeeds and the following take place:
        • c WETH is locked up in the contract
        • c(1-mintFeeRate) is minted into the user's wallet in sETH
          • c*mintFeeRate worth of sETH is sent to the fee pool in the form of sUSD
          • u - c worth of WETH is refunded back to the user
    • Given that u is strictly lower than c
      • When the user attempts to deposit u ETH into the contract
        • ✅ Then it succeeds and the following take place:
          • u ETH is wrapped into WETH and is locked in the contract
          • u(1-mintFeeRate) is minted into the user's wallet in sETH
          • u*mintFeeRate worth of sETH is sent to the fee pool in the form of sUSD
      • When the user attempts to deposit u WETH into the contract
        • ✅ Then it succeeds and the following take place:
          • u WETH is locked up in the contract
          • u(1-mintFeeRate) is minted into the user's wallet in sETH
          • u*mintFeeRate worth of sETH is sent to the fee pool in the form of sUSD
  • Given that the contract's capacity is zero, as maxETH is locked in the contract

    • When the user attempts deposit ETH or WETH into the contract
      • ❌ Then the transaction reverts due to max capacity being reached
  • Given that a user has u amount of sETH and the contract holds c amount of WETH

    • Given that u is larger than or equal to c(1+burnFeeRate)
      • When the user attempts to draw out ETH from the contract by burning u sETH and flagging withdrawal in ETH
        • ✅ Then it succeeds and the following take place:
          • c WETH is unwrapped to ETH and is sent to the user
          • c sETH is burned
          • c * burnFeeRate worth of sETH is swapped to sUSD and sent to the fee pool
          • u - c(1+burnFeeRate) worth of sETH is refunded back to the user
      • When the user attempts to draw out WETH from the contract by burning u sETH and flagging withdrawal in WETH
        • ✅ Then it succeeds and the following take place:
          • c WETH is sent to the user
          • c sETH is burned
          • c * burnFeeRate worth of sETH is swapped to sUSD and sent to the fee pool
          • u - c(1+burnFeeRate) worth of sETH is refunded back to the user
    • Given that u is strictly lower than c(1+burnFeeRate)
      • When the user attempts to draw out ETH from the contract by burning u sETH and flagging withdrawal in ETH
        • ✅ Then it succeeds and the following take place:
          • u (1-burnFeeRate) worth of WETH is unwrapped to ETH and is sent to the user
          • u * burnFeeRate worth of sETH is swapped to sUSD and sent to the fee pool
          • u (1 - burnFeeRate) sETH is burned
      • When the user attempts to draw out WETH from the contract by burning u sETH and flagging withdrawal in WETH
        • ✅ Then it succeeds and the following take place:
          • u (1-burnFeeRate) worth of WETH is sent to the user
          • u * burnFeeRate worth of sETH is swapped to sUSD and sent to the fee pool
          • u (1 - burnFeeRate) sETH is burned
  • Given that the contract's holds no WETH

    • When the user attempts draw out ETH or WETH from the contract
      • ❌ Then the transaction reverts due to the contract being out of WETH
  • Given that a user attemps to mint with WETH and msg.value is greater than zero then the following takes place:

    • ❌ The transaction reverts due to the user minting with both WETH and ETH

Configurable Values (Via SCCP)

For the wrappr contract, the following values must be set:

  • maxETH the maximum amount of ETH held by contract.
  • mintFeeRate the fee for depositing ETH into the contract.
  • burnFeeRate the fee for burning sETH and releasing ETH from the contract

Proposed Initial Values

The following values are proposed as an initial configuration:

  • maxETH 5,000
  • mintFeeRate 5 bp
  • burnFeeRate 5 bp

Copyright and related rights waived via CC0.