Episode 145: June 10th, 2021
01:11: Governance Update
03:33: GovAlpha Core Unit update
04:18: Forum at a Glance
06:00: Protocol Engineering Team Update
14:27: Oracles Team Update
26:32: Risk Team Update
30:18: Real World Finance Update
35:02: Growth Core Unit Update
43:18: Sustainable Ecosystem Scaling Update
51:20: Content Production Update
53:16: MIPs Update
1:06:31: Formal Submission Review
1:07:13: UNIV2DAIUSDC-A Changes
1:10:43: Anonymous Question Box
Agenda and Preamble
- Hello everyone, and welcome to the MakerDAO Scientific Governance and Risk meeting number 145, taking place on Thursday, June 10th at 17:00 UTC. My name is Payton, and @prose11 on the Forum. LongForWisdom is on holiday for this week.
- If you want to ask questions or have comments, please do speak up. Everyone here enjoys answering questions.
- Our latest Executive Vote has passed on Monday, which granted seven Core Units their monthly budgets. The Executive also added the remaining Vault types to the Liq 2.0 framework and lowered the
tout fee, which is the fee out when using the PSM USDC, from 0.04 to 0.00%. In addition, the vote increased the Global DC by 50 million Dai to help prevent us from reaching the maximum DC.
Polls and Executive
- This week, all except one poll had passed, which was the Investigation of the YFI C-type Vault. The winning option was to defer. However, it will still be added to the collateralization frame as dictated in the outcome section.
- The polls that had passed are:
GovAlpha Core Unit update
GovAlpha Weekly Update
- GovAlpha has been busy during the long holiday. Check out our weekly update post on the forums for all the details.
Forum at a Glance
Forum at a Glance for June 4th - 10th
- Thanks to @blimpa for setting up and posting this week’s Forum at a Glance.
- Dai has crossed the 5 billion mark!
Global DC was increased by 500 million Dai.
- In the same vote, Core Unit Budget Distribution for June was approved; a collective sum of 1,543,127 Dai will be distributed to seven Core Units.
- USDC-A, USDC-B, TUSD-A, PAXUSD-A, GUSD-A, PSM-USDC-A, UNIV2DAIUSDC-A, USDT-A vault types will be added to the Liquidations 2.0 framework, though liquidations will be disabled.
Protocol Engineering Team Update
- In regards to schedule, this week will involve Rate Group changes and the stuff Payton mentioned. Next week we’re planning to do the Centrifuge assets. They will be on Kovan on Friday, which may be pushed back because they are in the middle of an audit, and some of the DC increases were beyond the scope of what we had originally recommended. They will be set until the second audit. We may not be going to do RWA next week.
- On the collateral onboarding side, we had to make changes to the DSSProxy actions to support the Sushi LP tokens and PAX-G adapter. This requires some UI changes. We have also been working on the SushiJoin adapter, stalled until the Masterchef v2 gets updated.
- For modules, there have been many discussions and code clean-up around the DSSVest, which is how we will be distributing MKR and paying teams DAI. We talked about the D3M direct deposit module for Aave. We began thinking about different test cases like emergency shutdown and more.
- Vote delegation is completed; it is currently passing its RPC tests, and ABDK finished the audit. We may end up having much Governance action here. Governance may have questions around this, but there are no core changes to the protocol.
- There’s a flash DSSMint module which is flash minting Dai. We’re finalizing some code changes and aligning the best practices there.
- We’ve been helping a lot on the auction demo keeper, which is a backstop keeper for liquidations. We put some work in the
clipperCallee to support LP tokens there. If anyone wants to run that, you can grab that keeper, and it should have LP token support. We will try to keep it up to date as we add different collateral types like SushiJoin. It’s easy to run; you need a private key and some ETH to start and run all the authorizations. There will be some implementation change to the clipper; The
attack function will be added to the SushiJoin adapter to collect rewards, which won’t change interfaces.
- We’ve been transferring the auditor, tooling, and legal agreements from the foundation to the DAO. I want to remind everyone that Protocol Engineering has office hours every Tuesday at 3:00 PM UTC
- Also, we’ve been talking about staked ETH and evaluating the implications; what it means to MakerDAO. @hexonaut has been looking at the UniV3 fixed spread token that can get the standard equivalent of LP tokens on a fixed spread for like-kind of assets with Uniswap. that way, we can create a Vault that possibly sucks some of the stuff such as USDC from the PSM. The idea is that people will naturally create Vaults. There may be an angle where we can this kind of stuff in place of the collateral backing of the PSM, which is attractive from a security perspective because the backlist risk would be one step remote in this more generic token that many more members would be participating in.
- Derek: I will clarify a little bit about the fixed spread UniV3. It’s still in the investigative phase, but the idea is to roll out this starting with DAI-USDC at some fixed spread just around $1. we could potentially offer this with high leverage because it’s not much different than the current USDC Vault besides the fact that it would be a regular Vault. It’s no longer on Maker’s book. As Chris mentioned, we get the blacklist risk bonus as well, as we can start charging interest for this mountain of USDC we’ve on our books right now.
- Christopher Mooney: I forgot to mention a lot is going on L2 work that is going on. We are constantly in touch with L2 providers in the space.
Oracles Team Update
- Earlier this week, we moved the Oracle Core Unit to MIPs into formal submission, which will be included in the June Governance Cycle. There was a lot of work going on behind the scenes.
- We had a bunch of polls this week related to whitelisting for yEarn Vaults. All those passed and will be going to the Executive vote tomorrow. All of those are for newly delegated Vaults that yEarn is doing. I will begin submitting the MIP to delegate a lot of authority for nondangerous oracle functions to the Oracle Core Unit; for example, The whitelisting function can have no harm to the protocol; this will help us prevent long delays from the Governance Executive votes and improve onboarding by preventing spam votes in the Governance.
- We’ve been talking a lot about staked ETH, and there is a lot of interest coming in. eventually, all ETH will become staked ETH at some point. Right now, there is just liquidity. There is no direct redeemability, but this feature will come up soon to redeem staked ETH. We’re pushing and prioritizing it. We looked into this in the past, but the Oracle Team blocked this because creating a staked ETH oracle is difficult to safely create due to all the liquidity exclusively in Curve with very little volume in other platforms. Exploits like flash loans might manipulate the prices, but there are some workarounds around this. We’re working with the Lido team to figure this out, but we should find a way to create our oracle.
- A signaling poll passed this week to change the DAI-USDC LP parameters by lowering the SF to 0.5% and CR to 101%. The lowest we can do is 102% because the LP token oracles math is based on approximations. There are things that we did to optimize the gas which doesn’t compute precisely. Lastly, if you compare the oracle price to the real price, the drift is not that bad. in extreme cases, the delta may go up to 1%, making Dai minting profitable and dangerous.
- Christopher Mooney: We’re going to leave liquidations off as we had discussed with the mandated actors. We want to get a better set of liquidation parameters.
- Brian McMichael: An idea on the instant access module is that we can work together to develop technical implementation by the time you put up your MIP.
- Nick: Sure!
Risk Team Update
- Regarding collateral evaluations, we’re finishing the last two evaluations for Sushi which would be SUSHI-ETH, SUSHI-LP pair, and xSUSHI. This will be posted within the following days.
- We started discussing the next potential implementations or collateral onboarding between other facilitators. We’re likely to focus on three major implementations and evaluations, which will keep us busy the next few weeks. We will be starting with staked ETH collateral onboarding, D3M model implementation, and Polygon collateral onboarding. These three could lead to a more meaningful Dai supply. Staked ETH and D3M could be more demanding right now where we need to evaluate Aave protocol risk instead of just collateral risk.
- We’re adding more metrics to the risk dashboard. We should soon have the auction statistics page live. Also, we’re making simulators where community members can simulate losses for Maker loans based on the parameters picked. Parallelly, we’re building a separate page to put all the recommendations, research, and evaluations based on segments. This will be helpful for the community and future analysts.
Real-World Finance Team Update
- There are not many updates this week. I recently came to know that I am discussed to be removed from the organization. To clarify, I work for Maker with 100% dedication and do not involve any politics. I would appreciate constructive feedback on this.
- Payton: LongForWisdom will be back next week. If there is any Governance-related information, he will help you out.
Growth Core Unit Update
Growth CU Weekly Updates
- We are in Miami! We had a lot of meetings, and one of them was with Aave Pro. We tried to understand what it takes to have an institutional product. I suggest everyone read the forum post to explain institutional Vaults and the opportunities available. In the case of Aave, it’s easier because they need to create a separate pool that they are using for the institutions. They are including USDC, ETH, BTC, and AAVE. We’re also trying to integrate here but are planning to have permission DAI, which would be possible after the D3M module.
- We also spoke to Celcius. They have a similar thing in mind but want their CEL token integrated as collateral in Maker. We’re aware of the costs for incorporating new collateral, and of course, we check different things like understanding the users of those Vaults. We should start thinking about shutting down old collateral types that are not working and preparing the questionnaire before onboarding new collaterals.
- Opera integrated Dai in their beta android app, which is famous in emerging markets like Nigeria. We’re in touch with them to understand if they can integrate Optimism DAI. We are also talking with BlockFi, but more details are coming soon on that.
- Dai was listed in Bithumb in Korea. Dai is the first stable coin there, and the hardware wallet listed Dai. We see a lot more adoption of DAI in Klaytn among other networks.
@Lozadauis12 organized a tournament with Axie in Latin. it’s very interesting. please participate. Next week, we will have this tournament streaming live on Twitch and YouTube which will help Dai to be named on these popular sites.
- Christopher Mooney: Did we reach out to someone in El Salvador to see if there are any opportunities that we can leverage on BTC and Dai Vaults? BTC and USD are legal tender there.
- Nadia: Surprisingly, they were not much into crypto. We’re trying to reach out to crypto ATMs there.
Sustainable Ecosystem Scaling
- Mariano: What is allowing to onboard service providers into the ecosystem?
- Wouter: We are using a model that we are developing to get service providers onboarded into the ecosystem. Any service that may be provided through the DAO, such as a company that does e-mail marketing campaigns, which is a service provider, can be reached out to. We then explain how to work with the DAO protocol. This is difficult because these companies are used to working with centralized organizations through traditional methods. We help understand these differences and assist them in creating a business unit inside their organization that can ultimately put up their MIP to provide their service through a better-integrated business model between that company and the DAO. Part of these methods is the Incubator program, which is up and running. We will continue to document more on this.
Content Production Core Unit Update
Content Production Core Unit Biweekly Update for May 27th - June 9th
- This week’s primary focus revolved around hiring. We’re looking for a couple of writers and a video producer. Juan, Andrew, and others have sent us some candidates. We will begin interviewing and hopefully begin onboarding some people soon.
- The vote on the mascot has ended with a tie between the capybara and dragon. The best solution for this is to host an art contest and get feedback from the community. We are scheduling a planning session next week by reaching out to people interested in this.
Weekly MIPs Update
- Juan Guillén: David, did you purposely miss the GovAlpha budget modification proposed MIP?
- David Utrobin: Sorry, I did not realize I had missed it. I will go over it now:
- We saw a budget modification posted on the forum by @Deimos. This is an interesting case because it is the first time a non-MakerDAO team member has posted a budget modification on behalf of another Core Unit. He posted a modification for GovAlpha stating the team should get MKR compensation. We spoke about this during the internal Gov Domain chat channel. I was not sure how to initially respond to this; assign an SP number or move forward in another way? Ultimately, I gave my recommendations and helped it move forward by proposing that this budget modification should come from a member of that CU itself. When somebody from outside proposes, a person from the CU team should clear it with a MIP editor. However, I am open to ideas. The author of the proposal is very nice and openly allowing GovAlpha to take over from here.
- Payton Rose: This wouldn’t relate to all budget modifications because if you want to defund a CU, the CU itself won’t propose that, of course.
Formal Submission Review
- LongForWisdom will return Sunday evening and will add things into the voting portal for relevant MIPs.
- Currently, we have gone over all of the publications. The nine MIPs which David has presented are good to go and will appear on the ratification polls.
- LongForWisdom will still be the final reviewer and help maintain our standard workflow.
- There was a small typo in the section going over the copy of different options. Luckily, on the vote itself, all the options appeared correct.
- It is clear that the community preferred to lower the LR for UNIV2DAIUSDC-A. This was a ranked choice vote, and the vote to decrease LR to 101%, SF, was overwhelmingly the winning vote by the long run. We understand that there may be community intention to drop to a much lower LR. However, with Nik’s comment in mind from Oracle’s side, it is appropriate to put it in this week’s Executive as 102% LR rather than 101% stated by the winning option. Risk perspective does advise changing some of the
ilk parameters for how much we’re allowing into liquidations at once. If we were to flip the liquidation switch on, it would take some time to discuss and execute.
- Sam, @hexonaut posted in the chat that he is the author of the signal that led to this vote. He believes we should push it down to 101% if it’s okay with the mandated actors.
- Hexonaut?: It’s in line with what we have been doing; Governance signals, intent, and mandated actors. We can take a lighter or measured approach if need be.
- Payton: The plan for tomorrow is for the copy to drop to 102% rather than 101%. As Nik mentioned, we can make a further signal to push it down to 101%. We will also have to get liquidations figured out.
Anonymous Question Box
- This week’s question is: What price differences need to occur to get liquidated on a 102% ratio compared to the previous ratio of about 105%?
- Wouter: Based on the leverage levels, it’s not that much of a difference.
- Payton Rose: Yes, it’s not that much, but these are stablecoin pairs, so in theory, neither Dai nor USDC should diverge too far from their Peg. We do still have the base oracles at 1. It’s more in terms of the concern coming from the SF; if the pool itself is not earning more than the SF, that will push items over into liquidation. Also, the transfer to v3 is looking to be the case for this, resulting in positions being undercollateralized. Therefore, dropping this by 0.5% does help mitigate that as well.
- Nik Kunkel: Dai and USDC are hard pegged to 1 in the protocol. The LP token is only increasing due to trading fees generated. It is only the SF that can put you in the liquidation range.
- Nik Kunkel: This is an interesting point that is more theoretical than practical, but if Dai or USDC is hard de-pegged and we had too much impermanent loss, it drops below the fees generated by the pool is highly unlikely but theoretically can happen.
- Payton: We will have much larger problems if that were to occur.
Common Abbreviated Terms
CR: Collateralization Ratio
DC: Debt Ceiling
ES: Emergency Shutdown
SF: Stability Fee
DSR: Dai Savings Rate
MIP: Maker Improvement Proposal
OSM: Oracle Security Module
LR: Liquidation Ratio
RWA: Real-World Asset
RWF: Real-World Finance
SC: Smart Contracts
CU: Core Unit
- Artem Gordon produced this summary.
- David Utrobin produced this summary.
- Sai Teja produced this summary.
- Everyone who spoke and presented on the call, listed in the headers.