Skip to content

Latest commit

 

History

History
87 lines (67 loc) · 8.38 KB

risk_16.md

File metadata and controls

87 lines (67 loc) · 8.38 KB
Raw

Risk #15 - 12/13/2018

Introductions

  • Rich, Steven

Recap, Carry over

  • Ecosystem has an ideal state we’d like to reach and we’re bootstrapping and the complexities will be come ever more intertwined

How large is the ecosystem? Are Dai holders mostly looking to support their risk taking behaviors? Other use cases established? [2:31]

  • We have a general idea, but there is a growing subset that we’re not sure how they’re using it
  • We’ll need to flywheel the economy in a sense

Big Idea: We want to reduce the risk in the voting process and minimize contention. How do we do that? [8:08]

  • Aligning on perfect governance will be difficult, but we can iterate towards the best for us
  • As much as we can, we need to examine and rigorously vet the proposals
  • Need to have discussion around proposals on social media to analyze and dissect what is being proposed
  • The risk team proposal needs to show what the collateral is, why we want to include it and how we’re going to extract stability from this collateral
  • It will be important for the voting community to understand what tools the risk team is using and why they are using them

What sorts of information, templates, and procedures will be available from the internal risk team to new potential risk teams? [13:44]

  • As we bootstrap the internal team will provide a paper laying out the analysis provided and what tools are available
  • Eventually a spreadsheet could be the next iteration and ideally a tool or website would be available to play around with.

Since we’ll be applying a weighting system to the risk teams, will that weighting criteria provide a benchmark for teams to form against? [14:31]

  • The groups of risk teams are support teams and submitting teams. Support teams submit cleaned data or handle other tasks within the risk assessment process while submitting teams might handle support as well will be the contributors of actual proposals to the governance structure
  • Assessment of risk teams comes down to "how accurate have you been" and "how long have you been around"?

Is the first risk team internal to the Foundation or a separate entity? [17:45]

  • The first risk team is in place in the Foundation
  • The role is to provide a template and initial tooling to work with when forming new teams
  • External risk teams would likely create as separate legal entities. The construct they are working from shows to the longevity expected from the team

How do the Risk teams make money? [19:12]

  • As MakerDAO is running on its own in the wild, there will be a series of things that MKR holders will need in the form of services and those service providers will need to be paid to provide services
  • MKR holders can provide a portion of the revenue from Stability Fees to service providers through voted allocations

Is that something that will be facilitated through smart contracts? [20:44]

  • As we’re bootstrapping it would probably be facilitated off chain.
  • The future is unclear on whether it could be handled in smart contracts

External funding makes sense for services, could potential bad actors influence risk teams through that funding mechanism? What are the potential conflicts of interest there? [21:25]

  • Within the MKR distribution in the ecosystem there will be a series of bias and incentives, risk teams will not be absent of that bias
  • Proposals need to be scientifically vetted as an initial barrier. Proposals that are scammy should report that way after being put through the series of models available.
  • If a risk team starts manipulating it’s model to arrive at a better conclusion for the collateral type it will be revealed that the model is different than the one proposed when the risk team was accepted
  • Reproducibility is not just about the model but also using reliably sourced and clean data
  • Data from the likes of Bloomberg and Reuters will contain some inconsistencies. Data cleaning will be of paramount importance to ensure accurate modeling and decision making

Would we ever want to allow risk teams that are not using transparent models? [29:00]

  • The level of transparency is effectively the level of bias
  • If the risk team is not transparent then that will be a weight against them
  • Reputation is huge in providing risk services and not being transparent would likely lead to being found out. The incentive remains to be transparent as possible

Should risk teams stake MKR to better align their incentives with MKR holders? [30:43]

  • Risk teams will need some MKR to put a proposal forward into the voting construct
  • It’s been suggested and could be considered in the long term

What will the debt ceilings look like for real estate collateral? Could real estate have a higher debt ceiling? [32:08]

  • The derived the risk parameters originate from Liquidity-adjusted Value at Risk.
  • Liquidity and volatility have to be closely tied as they influence the other
  • The debt ceiling is derived from what is the maximum amount that we could liquidate over a specified time period without causing endogenous effects on the price
    • Endogenous effects are when selling a coin and the price doesn’t rebound because of a lack of liquidity in the market. If the market is liquid it will generally rebound on buys to nearly the price purchased.
  • Time frame is very important when considering a liquidity event as panic selling during a panic induces more panic
  • When considering real estate, and gold, we have to anticipate what the final recourse is through the organization facilitating the collateralized token.
  • The blockchain and tokenization is bringing significant liquidity potential to asset classes that have remained illiquid so far in the economy

Is there going to be a walled garden within the Dai Stablecoin System for tokenized securities? [40:55]

  • From a legislative point of view it may be necessary to facilitate the KYC/AML required for those securities
  • It’s likely that walled gardens would be built around specific asset-types that require their own regulatory needs

Would the recourse available on tokenized securities be limited then? [41:56]

  • The network effects being created in the ecosystem around Maker, should drive innovations around security tokens, etc and allow the recourse mechanisms to grow

Does the Foundation have a legal team working around the management of tokenized security risks? [45:42]

  • We’re not focusing on tokenized securities yet, but have begun considering the best partner to work with in the space and ensure its success
  • Ideally teams would package these securities and present them as collateral-types with their risks assessed and presented (including regulatory risks)
  • There will be opportunities that will present themselves for risk teams to assemble around very niche pieces of the risk process (like regulatory risk)

Is approving new collateral types a mechanism available to MKR holders to control the price of Dai? (Allowing new collateral types to incentivize creating more Dai) [49:10]

  • That is part of the calculation, but not used to control the price directly

We’re anticipating the votes for collateral types to be quarterly? [49:54]

  • Could be monthly initially as we get the system started and up and running
  • Will depend on how many collateral types are getting on board to start

What incentive does a commodity company have coming on if the stability fee is higher? [51:00]

  • We’ll have to be competitive in the market to encourage any action