This post proposes a possible framework with which to evaluate staking opportunities from mobilized collateral deposited with Vesta. Since releasing our first governance post describing our GMX staking strategy, we have been focused on discovering more strategies to deliver greater yield to our users. Thus far, we have executed on our GMX strategy and have begun sharing revenue to users. However, we have run into roadblocks regarding the safety and risk of some of the strategies proposed amongst contributors, especially with our most powerful idle assets such as ETH.
This post will outline the necessary process for implementing a staking strategy involving deposited collateral as well as specifications we will provide in each case for full transparency. It sets the standards through which we can accurately understand the risks that a staking opportunity can pose to the system and assets belonging to the users. A successful risk framework should streamline the mobilization process, allowing us to unlock the full utility of all our assets. It should remove blockers and unknowns around the process, which also opens the door to community-sourced staking opportunities.
As a result of a recently passed vote in favor of partnering with RiskDAO, we expect this process to be both speedy and robust. Through a combination of in-house analysis and the quantitative perspective of RiskDAO, proposed strategies will maximize yield and continue to provide users with high capital efficiency. Furthermore, in the wake of our recently proposed collateral risk framework, we have a powerful model to draw upon in forming this staking risk framework. Having similar processes in place for different types of risk should overall increase efficiency due to operational similarities.
We aim to mobilize all remaining collateral types in the near future. VST is now poised to become a stablecoin designed to leverage both the yield and capital efficiency of each token type. To that purpose, we must construct a robust framework through which to analyze risk to minimize possibilities of damage to our system and to users. By doing so, we expect to substantially increase Vesta’s revenue and usership by discovering staking strategies that supply users with substantial yield.
At a high level, our risk analysis process will have this general flow, in chronological order:
- Curia Proposal
- As a generally risky activity, all staking strategies will be posted directly to Curia before any further action or risk assessment. Staking can often have unforeseen implications, so a quick check with the community and preliminary description of intended actions should substantially increase efficiency
- Risk Checklist
- A comprehensive checklist of relevant qualitative and quantitative data is provided and evaluated, as described in the sections below.
- Community Verification
- A proposal including the team-verified checklist is submitted to the community for final read.
- Community Temperature Check and Poll
- The community is invited to share general opinions on the involved staking strategy and any relevant details, followed by a poll of community sentiment.
- RiskDAO assessment and report
- RiskDAO is engaged to provide a purely data-driven approach to supplement Vesta’s analysis. We expect RiskDAO to take on research into the staking strategy’s safety, leverage, and impernant loss potential among other key indicators of risk.
- Parameter Finalization
- Work continues with RiskDAO to finalize the final staking strategy in relation to current and forecasted market conditions.
- Testnet Deployment
- Contracts are deployed for testing and quality assurance
- Final Community Vote
- A governance vote on final deployment on the relevant network is conducted.
Please see below a table of checklist items, organized by categorization, see below for justification of various items:
Use of the checklist
We have designed the checklist to be used in a grade-based fashion, with quantitative items receiving both a grade and the associated number. Furthermore, we’re explicitly including a risk-reward consideration in evaluating the risk of strategies. Just as individual actors do for their own yield strategies, we believe it is within the protocol’s and the user’s best interest to carefully balance safety and actual upside. We’ll explain our process here in further sections.
Smart contract risk
It is Vesta’s immediate goal to ensure that all activities contribute to both the utility and stability of VST. To that end, we recognize that many instances of downside in staking and earning on idle assets are a result of issues in smart contracts. As such, we are asking all proposed strategies to provide qualifying information such as a valid GitHub repository or the number of unique stakers in the contracts associated with a stratefy. From this, both ourselves and RiskDAO can conduct the needed analyses for execution. We believe that while properly audited code is a good start, our own due diligence is the needed safeguard here to prevent any unfortunate oversights.
We also recognize that not all actors in the ecosystem have perfectly aligned incentives or flawless governance. As such, we believe that we can take the proper steps towards protecting the protocol and its users through a thorough assessment of counter-party risk. Generally, this phrase describes the risk involved in engaging another party in an agreement and the possibility of their inability to fulfill the partnership terms, either through malice or inaction. For us, this looks something like a governance attack on a partner protocol’s platform directed at us, rapid purposeful action on a staked token type to impact the security of our assets, or any other event directed to harm Vesta, be it malicious or not. We believe that looking into factors such as thequality of a partner’s multi-sig can reasonably reduce the risk of these events.
This category describes the likelihood of significant impermanent loss events, which generally describes the loss of value on one or more tokens in a single or multi-sided staking strategy. While this term is more common in relation to multi-sided strategies, we believe it still applies to single-sided ones, especially if volatility is high on the related assets. In particular, we are concerned about events in which large numbers of Vaults reach liquidation thresholds while tokens are staked, which may cause an incident similar to a “bank run,” where not enough assets can be quickly liquidated and directed to stability pools. To minimize the possibility of these events, we will be looking carefully at liquidity depth and associated token volatilities to ensure the security of the system.
This rather self-explanatory category simply attempts to describe the possibility of asset loss when staking strategies are taken. This is an incredibly important part of the risk assessment, especially when attaining a sufficient yield requires recursion or leverage. As losing collateral assets would harm the system and users to a large degree, we will be looking very closely at the risk of liquidation. We also realize that some strategies include harsher liquidations than others, that is, sometimes partial liquidation can be assured rather than a more catastrophic loss. Thus, we also take these factors into account when grading risk. Generally, if we cannot assure that the system will retain a similar collateralization ratio if liquidation is likely to occur, a strategy will not be taken.
While a staking strategy’s riskiness is first and foremost in the minds of collateral managers, it would be disingenous to propose a framework that only takes this one factor into account. In truth, while a strategy must maintain a sufficient yield to justify its risk, this also means that sometimes a higher yield will justify a higher risk. Therefore, we are committed to both the safety of collateral assets as well as their productivity. The community ultimately governs all strategies, but in proposing a certain strategy, we implore that a proper image of reality is given when describing the possible upside of a strategy. In our own assessments, we will ultimately take into account what is left on the table if action is not taken, despite certain risks.