- Rich, Jess, Steven, Soren
- What is scientific governance? Making arguments in risk management with data
- History of probability theory, tools that can be applied to risk management systematically
- Think of risk in layers, not categories (economic, financial, and operational)
Are we reinventing the wheel (when it comes to governance)? Could we leverage other solutions in the space, like Aragon, for governance management?
- Governance was built in to the process and adapted from bitshares. Due to specific needs of MKR governance, it’s been established from the ground up specifically for MKR asset decisions.
- Other teams are working on it, but not far enough along to validate using.
- Security is the utmost importance, so it would be additional burden to audit outside software
Should Maker holders add weight to risk teams that have long positions in the asset classes they are profiling?
- Initially no, could lead us to cartels through bias
- Need independent teams, bringing orthogonal views to propose constructs that are reproducible
- Scientific governance will not stop plutocracy, but will heavily point out bad actors. If discourse goes one way then voting goes another, it will be a clear signal that there is something malicious occurring
Has research occurred into what sorts of defense mechanisms are available to protect against plutocracies?
- Been outside the scope of our initial research as we lay the ground work for voting. 1 MKR, 1 vote
- Lots of discussion has taken place already. Can MKR influence people to vote without directly incentivizing them?
- Two votes: Governance vote, to vote on things, and Executive, vote to implement the changes
- Emphasize the importance of voting on the success for the project as a whole and continue engagement
- Incentivizing votes can be manipulated quickly
- Corporate shareholder voting sees significant apathy as well
Jumping back into: How can we implement scientific governance?
Entire risk function is horizontal
- Qualitative approval:
- What are the fundamentals behind a token (organization running the token)?
- What is the likelihood of changing their outlook?
- Quantitative implementation
- Qualitative rating
- Want to have many different points of view on an assessment. More views, the better
- Catch risks from more angles
- Quantitative assessment
- Ideally the constructs will be different, so weighted average is more likely to be on target
If the competitive teams have thorough discourse then you can be assured that the decisions formed have been well considered.
What is a construct? Who creates the first one?
- A construct is a collection of models, techniques, and ideas. Risk teams provide their constructs for MKR holders to assess.
- The first step includes providing an assessment (through each construct) of the collateral type for acceptance at any level
- The second step is where the actual implementation parameters are weighted per accepted constructs
To start the models will likely be in the same ball park, but down the road could provide significantly different positions on an asset. You might include them to make a compromise between the different views. Ideally you have several models, so you’re encompassing as many points of view as possible.
- Gov vote + Exec vote to implement asset types
- All the teams will be auditing each others models
- Ideally the models are transparent so all aspects can be critiqued
How will the community interact with risk teams or come to understand proposed constructs and profiles?
- Social media will host many debates of collateral in general and when the types become more granular we could see a breakout to specific groups analyzing different types of potential collateral (utility, collateral-backed, security, etc.)
- If it takes a good amount of time for MKR holders to understand the recommendations from teams then the business intelligence is failing somewhere
- Some sort of application to see: risk teams think, what their conclusions are, what their overall evaluations are
- Should be able to dive into their ‘work’ if necessary, but there will be clear conclusions delineated
- Internal risk team will provide the first construct for the community to engage with and token holders to vote in
- Eventually it will be abstracted to higher levels
- Hard core members will debate at first principles levels to idealize what should be employed and why
- Provide the basis for constructs and models
- Show the longevity of teams acceptance
- Weighting may decrease with lack of longevity prospective
- Two types of Gov: Proactive and reactive
- Proactive: implementing specific parameters through careful moderation
- Reactive: take a debt ceiling for example. When the profile of the underlying asset changes over time we would want to update based on the impact on the portfolio as a whole
- Two types of votes:
- Governance vote: Resolution on a conclusion about collateral
- Executive vote: Vote to implement the accepted changes
- Schedule of votes:
- Once-off votes - accepting/removing teams will be one time generally
- Intermittent votes - oracle inclusion/exclusion could be more regular as security information processor is lacking in crypto currently
- Regular votes - collateral votes, constantly monitoring change options
- Initially would be equally weighted. As time goes on and more data becomes available we can asses the validity of constructs and adjust the weightings
- Assess teams against their probabilities provided
- It could become its own vertical where teams are assessing the risk teams themselves
- It’s a gradual process, we’re very early
- We’re establishing these discussions for future decision makers (<-pun intended)
- We’re laying the ground work for all transactions need Dai