LEAP-51: Lyra Governance V2
Table of contents
- Simple Summary
- Configurable Values
The next iteration of Lyra Governance.
This LEAP proposes migrating Lyra Governance from v1, a representative council model with off-chain voting, to v2, a fully autonomous on-chain system. This allows token holders to directly govern the components of the DAO which require human input, specifically:
- Protocol Administration
- Treasury Management
- Staked LYRA
In this system, proposals are fully on-chain, binding and auto-execute without reliance on any third party. It also works cross-chain, allowing token holders to govern deployments and treasuries on multiple networks. If approved, the following actions would occur:
- Disband the Lyra Council.
- Deploy Lyra Governance V2 - an on-chain system described in section 3.
- Transfer ownership of Lyra DAO’s treasuries and protocol contracts to Lyra Governance V2.
Governance and Tokens
Governance is required for the aspects of a protocol that require human input (i.e. things that cannot be automated). There are two main areas of any project that require such input:
- Protocol Administration
- Deployment of smart contracts
- Setting system parameters
- Treasury Management
- Funding the maintenance and improvement of the protocol
The need for a governance framework that can manage a protocol and treasury has led to several experiments, with governance tokens proving the most effective. Tokens align the different stakeholders in a project toward a common goal (although not without significant drawbacks, i.e. plutocracy).
Governance V1 and Tradeoffs
Lyra Governance V1 is a system for governing the DAO via a governance token known as LYRA. It was introduced in LEAP-15 and is centred around a 5 seat representative council known as the Lyra Council. Token holders elect the council, which administers the LEAP framework. A LEAP is a proposal to change part of the protocol and can be created by anyone in the community.
The council model offers great speed and flexibility, allowing the project to evolve rapidly. In 18 months, we approved over 40 LEAPs, launching the protocol, token, two major upgrades (Avalon and Newport) and many others. However, the speed afforded by the council model comes at a cost. Proposals are off-chain, which introduces implementation risk that affects both the protocol and the treasury. This serves to undermine the dependability of the project.
With the recent launch of Newport, we have a protocol that can be deployed on any EVM-compatible network and integrated with any perpetual venue for hedging. This is a foundation on which we can attract legitimate institutions to build on the protocol, provide liquidity and trade. However, they need a governance process that they can depend on. We intend this iteration of Lyra Governance to be the final step on our journey of progressive decentralisation.
Lyra Governance V2 allows LYRA holders to directly govern the protocol and treasury via an on-chain system. By staking the LYRA token, holders receive a governance token, stkLYRA, with governance rights that enable them to make proposals, vote or delegate to other token holders. Proposals are called LEAPS, which are fully on-chain, binding and execute automatically, without any reliance on a third party. Cross-chain governance is supported, allowing the management of multiple protocol deployments and treasuries on different networks.
In this overview, we provide a non-technical description of the governance framework.
LYRA token holders can stake their LYRA to receive stkLYRA, which grants governance powers proportionally to their balance. There are initially two powers associated with the stkLYRA token:
- Proposal right - gives access to creating a proposal.
- Voting right - enables voting on existing proposals.
Users can choose to delegate one or both of these rights. This enables Lyra Governance to retain the benefits of a council-based model, where less informed participants can delegate to others, without introducing the implementation risks currently associated with the off-chain council model.
Lyra Enhancement Action Proposals (LEAPs)
A LEAP is any proposal that impacts the protocol, treasury or governance framework. It has been adapted from industry standard frameworks, such as EIP and BIP. Examples may include protocol changes (listing new markets and changing parameters), treasury actions (allocating LYRA tokens and remunerating contributors) as well as meta-governance updates (changing the governance process itself).
Each LEAP must include the following:
- Simple Summary - A simplified and layman-accessible explanation of the LEAP.
- Abstract - A short description of the LEAP.
- Motivation - Why the existing system is inadequate to address the problem that the LEAP solves.
- Specification - The specification should describe the process and structure of the proposal.
- Rationale - Why were particular design decisions made? Describe alternate designs that were considered and related work.
- Test Cases - Test cases are required for code-related proposals (i.e. non treasury actions).
- Copyright Waiver - All LEAPs must be in the public domain. See the bottom of this LEAP for an example copyright waiver.
LEAP Life Cycle
- Creation - The author(s) write the LEAP using the specified template.
- Community - The LEAP is published to the community.
- Off-chain signalling (optional) - The LEAP is proposed on snapshot (current system) for stkLYRA holders to signal approval or dismissal.
- On-chain proposal - The LEAP is submitted to the Lyra Governance system, to be voted on by stkLYRA holders.
- Approval - If the LEAP is approved, the payload is sent to the relevant timelock for automatic execution.
- Rejected - If the LEAP is rejected, there is no further action.
In this system, Lyra Governance will take ownership of all treasury assets currently held in the multisig safes on Ethereum, Optimism and Arbitrum. This means that any movement of DAO funds will require a proposal to be passed. Any group can make a proposal to governance for funding. This may include the current core contributors and/or any other third party who wishes to work on the protocol.
To date, the majority of protocol development has been handled by the current core contributors. As Lyra continues to decentralise, we must make it easy for multiple independent groups to contribute to the protocol and receive funding from the DAO. Decentralising the development process is crucial, allowing the protocol to improve without relying on one group. When done right, it can increase development speed, improve overall quality and foster innovation. However, before governance V2, it took a lot of work to coordinate this process.
With on-chain governance, it will be easy for any team ("service provider") to apply to the DAO for funding directly. Token holders will be responsible for assessing applications from service providers and setting the project direction, high-level goals and objectives in collaboration with the community. Any service provider can work to implement these goals and propose their solution to the DAO. Here we visualise a multi-team structure:
To ensure protocol development continues smoothly, we propose that the current core contributors receive an upfront allocation of 2,000,000 USDC from the DAO. With a current burn rate of ~$2,700,000, it will allow development to continue for 9 months before they need to seek new funding. This should be sufficient for the decentralised development structure to take hold and for new independent service providers to spin up. The current core contributors' annual expenses breakdown is as follows:
The DAO treasury as of 1 March 2023 is worth approximately $36,710,083.22 and the breakdown of assets is as follows:
|Opt MMV Liquidity||300,000||1||$300,000.00||0.82%|
|Arb MMV Liquidity||300,000||1||$300,000.00||0.82%|
Lyra Governance V2 makes the protocol significantly more dependable, in part because it increases the time taken for a change to be made. This is desirable when it comes to decisions such as deploying new smart contracts, launching a new market or introducing a new incentive program. However, with respect to protocol performance and safety, responsiveness is critical.
Consider the following scenario - a bug discovered in the smart contract code that results in an active exploit. The community needs to be able to act fast to ensure that any loss of assets is kept to a minimum. In the governance system, a proposal would need to be voted on, approved and executed, which would take days at best. In addition, there are many complex parameters that occasionally require swift updates to respond to market conditions.
For safety and performance reasons, we are proposing the idea of a risk council and governance wrapper. Token holders will elect the council and set the bounds of the parameters which can be changed. This means this council will not be able to set parameters to a level where exploitation is possible, but they will be able to:
- Pause the protocol in an emergency, with only token holders able to resume it.
- Modify parameters within bounds, allowing for better performance of the protocol.
This idea will be expanded upon in a subsequent LEAP and until implemented, protocol contracts will not be transferred to Lyra Governance.
On-chain vs Off-chain Governance
Off-chain governance allows faster decision-making and improves the speed of iteration. However, it relies on third parties to implement proposals, which undermines the dependability of the project. Early-stage projects often choose off-chain governance to begin with and as the project matures, they transition to on-chain governance.
The two most popular types of on-chain governance frameworks are:
- Compound Governance - most battle-tested but needs a special ERC20Votes token
- Open Zeppelin Governor - most generic but doesn’t work easily with stkLYRA
Aave governance has a very similar structure and workflow to both of these systems but has the added benefit of working natively with the staked LYRA token (which was forked from Aave’s staked token). While it hasn’t been deployed across as many projects, it has secured Aave’s multi-billion dollar TVL for over 2 years, providing a considerable degree of confidence in it’s implementation.
At a technical level, Lyra Governance is very similar to Aave Governance v2. It shares the same high-level components and source code, with minor adjustments:
- Governance Contract
- Handles the proposal life cycle
- Governance Strategy Contract
- Determines the strategy used to calculate voting power
- Timelock Contracts
- Short - Handles execution of protocol + treasury proposals
- Long - Handles execution of stkLYRA + meta governance proposals
- Proposers - Initiate the proposal process
- Voter - Vote on active proposals
In the diagram below, we visualise the architecture of these components:
The governance contract is responsible for administering the proposal process. It has five steps:
A proposal can be created by calling the
create() function. The caller must have a proposer power greater than the threshold and they must maintain it until the proposal is executed. After the proposal is created it assumes the
PENDING state until the vote begins.
Voting begins after the
votingDelay has elapsed, a value which can be read from
getVotingDelay(). At this point, a snapshot of voting powers is taken and can no longer be delegated/transferred for the proposal being voted on. The proposal state is
ACTIVE and users can submit a vote for or against the proposal, weighted by the user's total voting power (tokens + delegated voting power), within the allotted
3. End of Voting
For a proposal to pass the voting power of for-votes needs to reach the quorum set by the
[MINIMUM_QUORUM] parameter. In addition, the difference between for-votes and against-votes (in % of total voting power) needs to exceed the vote differential threshold set by the
[VOTE_DIFFERENTIAL] parameter. If the proposal has passed, then the proposal state becomes
SUCCEEDED, otherwise, it is
4. Queuing and execution
SUCCEEDED proposal can be queued and will be executed after the execution delay and before the grace period expiration. The
delay can be fetched from
getDelay() and the
GRACE_PERIOD can be fetched from
The validation and execution of the proposal are performed by the time lock executor. A queued proposal state is
QUEUED. A successfully executed proposal state is
EXECUTED. If a queued proposal has not been executed before expiration, then the proposal state is
5. Proposal Canceling
If the proposal creator's proposal power decrease and no longer meet the
PROPOSITION_THRESHOLD, any user can cancel the proposal. In addition as an initial safeguard to the protocol, the guardian account, controlled by a community multisig, is able to cancel a proposal before a proposal is executed. A cancelled proposal state is
The Governance Strategy is responsible for determining how voting power is calculated.
The timelock contracts are responsible for automatically executing approved proposals.
Short-time lock executor
The short-time lock executor can change the following parts of Lyra:
- Protocol Deployments
Long-time lock executor
The long-time lock executor can change the following parts of the Lyra protocol:
- Staked LYRA
- Governance System
The cross-chain bridge executors extend Lyra Governance on Ethereum to other networks.
In order to migrate from V1 governance, the following actions need to occur:
Extend Current Council
- Extend the current Lyra Council until the implementation of this LEAP is complete (expected Mid March 2023)
Deploy Lyra governance V2
- Deploy the governance system on Ethereum L1
Ownership of treasury
- Transfer assets from Ethereum DAO Multisig to Short Executor
- Transfer assets from Optimism DAO Multisig to Optimism Executor
- Transfer assets from Arbitrum DAO Multisig to Arbitrum Exectuor
Ownership of protocol
- Transfer ownership of Optimism Protocol to Optimism Executor
- Transfer ownership of Arbitrum Protocol to Arbitrum Executor
- Governance Wrapper and Risk Council
There are currently four three rewards programs that require off-chain distribution:
- Lyra MMV Liquidity Providers
- Lyra Trading fee rebates and short collateral
- LYRA token liquidity programs
Transfer 3 months of rewards to the CC multisig so that these programs can continue to run.
Test cases are included with the implementation and will be available soon.
- Voting Delay: 7200 blocks (approx 1 day with 12s blocks)
- Guardian: Lyra Guardian Multisig
- Executors Whitelisted: Long and Short
- Owner: Long Executor
- Admin: Lyra Governance V2
- Voting Delay: 1 day
- Grace Period: 5 days
- Proposition Threshold: 0.5%
- Voting Duration: 3 days
- Vote differential: 0.5%
- Quorum: 4%
- Admin: Lyra Governance V2
- Voting Delay: 7 days
- Grace Period: 5 days
- Proposition Threshold: 2%
- Voting Duration: 10 days
- Vote differential: 15%
- Quorum: 20%
- Contributor funding: 2,000,000 USDC
Copyright and related rights waived via CC0.