[Agenda/Discussion] Scientific Governance and Risk #158 - Thursday, September 9 17:00 UTC

Agenda

The zoom waiting room will be on, and a password is set to: 748478, please ping us in the #governance-and-risk chat if you aren’t let in from the waiting room. This Agenda will be updated over the following week leading up to the call as people make me aware they want segments.

2021-09-09T17:00:00Z

We’re testing out a new format this week in an attempt to curate a more engaging and deliberative meeting. This week will be light on updates and heavy on discussion topics!

Introduction

  • @prose11 : Hosting, Introduction, Agenda.
  • @Davidutro : Slides for presenters, meeting outline

Governance Round-Up

Selected Updates/Discussions

Other Discussions and General Q&A

  • An anonymous Question Box is now available! Leave your question here and we’ll try to address it. Comments on new meeting structure welcomed and encouraged!
3 Likes

G&R #158 Snippet

This snippet includes Governance, MIPs, forum updates, and Core Unit team discussions from the MakerDAO Governance and Risk Call #158.

General Updates

Votes

Executive:

  • Executive - Parameter Changes, September Core Unit Budgets, DssVest Streams - September 3, 2021.
    • Passed Sept. 4th.
    • Decreasing Flash Mint Fee (toll), increasing PSM-PAX-A Debt Ceiling, distributing Core Unit Budgets, and establishing DssVest streams for budgets and MKR vesting.
  • No planned Executive Proposal for tomorrow.

Polls:

MIPs

Weekly MIPs Update #52

  • The first Monday of the month is here, and the Formal Submission Window for September is now open. The following proposals are eligible to make it into this month’s Governance Cycle.

Forum at a Glance

Forum at a Glance: September 3rd - September 9th

Team-led Discussions

Growth

Institutional Vaults

Growth CU Weekly Update

  • We have a new integrations leader, @zxMori.
  • We posted a question for the community concerning us sponsoring the ETHOnline hackathon. Please leave feedback!
  • After the Nexo Institutional Vault, we will be looking at ETH and BTC.
    • Currently working with legal and compliance teams to help find a solution to open Institutional Vaults with traditional companies.
    • Main concern is changing the current proposal because Nexo will have to decide to agree on them.
  • Pros, cons, and clarifications:
    • Nexo Vault risk parameters can be improved; how do we renegotiate the loan terms after its 6-month period?
    • We may have to potentially increase the SF based on market conditions, which is a concern.
    • Potential to someday have high portfolio exposure to fixed-rate vaults, which can promote high SFs.
    • Implementation of DssCharter is at 200%, and LR is at 120% for the Nexo Vault, which can cause complications to borrowers unwinding.
    • Institutional Vaults is a way to increase the total size of our Vaults by having professionally committed users minting DAI.
    • Nexo will be using the Fireblocks interface to manage their Vaults.
    • Institutional Vaults can help promote increased DAI supply and allow smaller participants to use the MCD again, which is currently expensive.
  • Tokens using Maker Vaults:
    • Loopring wants to provide a service to their customers by offering Maker Vaults. This can potentially extend to other Institutional Vault companies as well.
    • We should essentially set specific agreements with these quality projects that benefit both parties and helps mitigate risk.
      • Nik responds that we had an agreement with Polygon, who fell through those agreements and wasted DAI.
    • The Impossible Trinity Issue poses a potential long-term threat for fixed-rate Vaults. We need to think carefully going forward, especially with rates, which should move with the market to manage the capital flow. However, PaperImperium thinks otherwise because there is not enough pressure on the DAI peg.

Risk

Offboarding Vaults

  • Thinking of changing auction parameters for top-tier Vaults.
    • Propose a faster execution with a set of parameters to lower the duration and thereby increase auction throughput and amount.
  • Creating a risk-scoring methodology for Vaults on the Risk Dashboard.
    • Also, creating an additional risk score on Vault safety.
  • There have been some on and off decisions regarding offboarding low-tier collaterals.
    • Risk believes that these low-tier collaterals are risky because they have very high slippage.
    • Difficult to offer more than a 10M DC, which may not even cover oracle costs.
    • Data on Compound and Aave shows that low-tier collaterals promote very little use; a few million DAI generated.
    • Maker is not well-positioned for such risky assets due to our inability to cross-collateralize.

Oracles

Offboarding Vaults

  • Continued discussion from Risk (above):
    • Onboarding low-tier DeFi tokens is sustainable on L1 only if the collateral generates a significant amount of DAI.
    • There has not been enough experimentation, and the ones that were onboarded did not have much traction.
    • Being favorable by other tokens and teams does not compensate for the risk they provide.
    • We need to be more liberal when considering onboarding and offboarding tokens.
    • We can offboard these low-tier collaterals for now. When they become profitable, we can then onboard them again.
      • These tokens are currently not helping to increase DAI supply; Time is better spent elsewhere.
  • MakerMan: We need to focus on the big fish and stop wasting time with the small fries.
    • Maker’s competition is tough, and we need to match the competitive strategy at least if not do better. These mistakes can leave us in the dust.
    • We need to set strict requirements for onboarding and offboarding.

Episode 158: September 9th, 2021

Agenda

Video

https://www.youtube.com/watch?v=0rChBliTRYE

Introduction to the New G&R Call

David Utrobin

00:59

  • I am David from the GovComms Core Unit. We focus on making information more accessible and digestible.
  • We partnered with GovAlpha to do brainstorming to improve the G&R call. The problem during the last year has been this general town hall for status updates, sharing, and coordinating. With more CUs coming along, it is impossible to scale the format of the call in the format we have now. Therefore, we are making changes. We attempted to brainstorm various ideas that included CU rotation and team updates but concluded that it would eventually be futile with increasing CUs.

  • Our primary question is to define the G&R call. What is it? Once we define this call and understand the audience, we can optimize the way it is presented.
  • We will be reducing the call length to 1h15 or less, and we will be doing smarter team updates.
  • Written updates allow completeness for archivability, reference material, and complex reports. CUs and teams should post their weekly team updates on the Forums before the G&R meetings.
  • Spoken updates are better for problems and blockers, seeking feedback and engagement. We want CUs to focus on the spoken update during this call.
  • You can share in the forums what your team is doing and be reserving our spoken updates for the most pertinent topics.
  • The definition of this call needs to be sharpened around the G&R of the Protocol.
  • The idea is to share updates and discuss ideas that you think need to be discussed with the other parts of the organization, de-siloing and prioritizing information.
  • The remaining 30 minutes of the call will be open for special presentations and open discussion.
  • The goal is that if you are a stakeholder and want to know a specific thing, you should be able to find that through the written publications of the different teams in a precise search. Then, you would join these calls to talk about current events and issues.

Questions

  • Juan Guillén: What is the objective or scope of the G&R call?
    • David Utrobin: We need to sharpen the definition of this call to have a concrete objective. We will be working on this in the coming weeks. Generally speaking, this call discusses and shares important updates. However, we will need to prioritize certain teams and topics on a week-to-week basis.
    • Juan Guillén: I understand. I believe this would be better and easier for everyone if we cleared up the definition and objective.
    • David Utrobin: The move that we are making is trying to step away from this meeting, focusing on updates, which have been a chunk of the call for the past year. We want to incentivize more cross-team and in-depth discussion on current events and issues.
    • Chris Mooney: We can quickly highlight our team’s top update. We do not need to mention every line item, only the important ones.
    • Juan Guillén: We would like the audience to be explicit in who they are and what they want.
    • David Utrobin: The purpose of this call is to discuss protocol management and move through issues or events. The purpose of this call is not to update MKR holders.
    • Payton Rose: Our mission here with this rearrangement is to promote more discussion.

Introduction

Agenda and Preamble

Payton

12:41

Governance

Payton

13:21

Votes:

  • Executive:

  • Polls:

MIPs

Pablo

16:50

Weekly MIPs Update #52

  • The first Monday of the month is here, and the Formal Submission Window for September is now open. The following proposals are eligible to make it into this month’s Governance Cycle.

Forum at a Glance

Artem Gordon

19:50

Forum at a Glance: September 3rd - September 9th

Team-led Discussions

Institutional Vaults

Growth Core Unit

Growth CU Weekly Update

23:25

Growth Update:

  • We have a new integrations leader, @zxMori. He will help partners with technical questions when they want to integrate the Maker protocol into their projects.
  • We posted a question for the community concerning us sponsoring the ETHOnline hackathon. We can give prices. Please leave feedback in the weekly update.

Institutional Vaults:

  • After the Nexo Institutional Vault, we will be looking at ETH and BTC.
  • Currently working with legal and compliance teams to help find a solution to open Institutional Vaults with traditional companies.
    • These traditional companies are interested in having liquidity from the Maker Protocol and offer safe, professional Vault management.
  • The main concern is the KYC problem and changing the current proposal because Nexo will have to agree on these changes.
  • Traditional companies are interested in experimenting with DeFi and starting getting liquidity from the Protocol.

Discussion:

  • David Utrobin: What are the pros, cons, and clarifications?
    • Nadia: Nexo Vault risk parameters can be improved. We need to think about how we renegotiate the loan terms after its six-month period. Makerman brought a new idea in our Signal Request Poll to renew the next cycle of this institutional Vault. After six months, we have to renegotiate the terms of the loan. If the market changes drastically, we will not have a good position against the market and Nexo. If we give them a small SF, but we have to increase it because of the market conditions to the regular Vaults, that could be seen as unfair by the community in general.
    • Primoz: If exposure to fixed-rate Vaults becomes too high, there will be a risk on regular Vaults. We may then have to potentially increase the SF based on market conditions, which is a concern. Total exposure to fixed-rate Vaults needs to be related to stablecoin backing. Potential to someday have high portfolio exposure to fixed-rate vaults, which can promote high SFs. Setting a threshold when fixed-rate vaults could have at most X percent of the stable coin banking is an idea. And it is not currently an issue because if Nexo migrates and mints. There is about 700 million exposure for fixed terms and 20% of the stable coin backing. As soon as you know this ratio starts increasing, we should be cautious. Another risk is implementing DssCharter, which forbids the borrower to mint additional Dai or withdraw collateral below a certain collateralization ratio threshold. In Nexo’s case, it would be 200% where they cannot mint more, but liquidation happens at 120%. In addition, if the collateralization ratio drops below 200%, the borrower cannot make a certain unwinding procedure that some borrowers commonly use to withdraw collateral and convert it to Dai and then make repayment. This would not be possible unless the borrower uses a flash loan. The borrower needs to be aware because if they rely on this standard unwinding technique, this will not be possible due to DssCharter. We make sure to communicate this with all the borrowers to avoid unexpected issues. In summary: the implementation of DssCharter is at 200%, and LR is at 120% for the Nexo Vault, which can cause complications to borrowers unwinding.
    • Nadia: Institutional Vaults is a way to increase the total size of our Vaults by having professionally committed users minting DAI. In the current situation of the Protocol, the Vaults on both sides are smaller than 200 million, and the biggest Vault we have now is Nexo for 400 million. The Institutional Vaults will commit to mint 200 extra million, which means that one user will have 600 million Dai minted. These numbers are based on L1 parameters.
    • Wouter: What interface will Nexo be using?
    • Derek Flossman: Nexo will be using the Fireblocks interface to manage their Vaults. We had a call with them today discussing the technical aspect of things, which seemed very positive. We walked through increasing Vault size, migration, and future practices of internal systems under certain conditions.
    • Chris Mooney: We understand the philosophical approach of permissionless systems that allow anybody to participate. Creating permission systems is tying us to old-world models, which diverges from the blockchain philosophy. This is something to be aware of. Nadia brings up good trade-offs. However, we need Dai supply and desperately. Right now, small users cannot afford to participate and mint enough Dai. However, we need to find good ground between allowing protocol users to mint Dai and not seeming like a permission system protocol.
    • Brian McMichael: DssCharter can get us there. But this two-tiered collateralization ratio where you cannot withdraw Dai beyond a certain point but LR is lower. This doesn’t help borrowers but assures us that Vaults will be overcollateralized, reducing rates and attracting larger participants.
  • MakerMan: I support Institutional Vaults and RWA. I think we can do better terms. And I like the idea of having a buffer. However, my biggest issue is the impossible trinity, which we cannot yet get out of. We need to deal with rates. As soon as we lock down rates, the PSM will soak up the rate differences until they exceed. We need to think very carefully about how we work with fixed rates in the long term.
    • Chris Mooney: Interest rates cannot float up fast enough, which means we should not have a problem. Also, Nexo will not be against lowering rates. This should not put downward pressure on the peg.
    • Brian McMichael: We have been above peg for over a year, and the PE team has been working on methods to put downward pressure on the peg. This does potentially create issues that MakerMan mentioned. However, this is something we have not dealt with for a long time and thereby have been focusing on the other side of the peg for the meantime.
    • David Utrobin: The Maker protocol seems to have moved away from a rate-changing policy of changing the peg. Even though we have a more fixed-rate Dai supply, we are not using it. We are simply trying to get more Dai supply.
    • PaperImperium: Primoz mentioned earlier that if we have too much debt from Institutional fixed-rate Vaults, we would have to raise the rates aggressively on non-fixed-rate Vaults. We should establish a relationship between the amount of debt we issue at a fixed-rate to the stablecoin reserve we have as accounting for that risk.
    • David Utrobin: The more fixed-rate large Vaults, the permissionless Vaults end up subsidizing the rates.
    • MakerMan: This leaves our options to raising rates on non-fixed rate Vaults, which will eventually close or raise the DSR and end up burning through the surplus buffer. Thankfully, we have a lot of stablecoin in the PSM right now.
    • Someone: I want to bring up the idea that we will run into this same situation with RWAs. The trinity constrains them. We cannot raise rates on them. We would have to increase DSR and end up pinching our profits gradually.
    • MakerMan: Yes, the Impossible Trinity Issue poses a potential long-term threat for fixed-rate Vaults. Going forward, we need to think carefully, especially with rates, which should move with the market to manage the capital flow.

Offboarding Vaults

Oracles and Risk Core Units

Risk - Primoz Kordez

49:18

  • Offboarding Collaterals in Terms of Cost & Risk
  • We are working on changing Auction Parameters for Top-Tier Vaults. In the beginning, they were conservatively set to last around 40 minutes. Longer auction duration presents more market risks. There will likely be a proposal for auction parameters to be set to 40 minutes, improving auction throughput. Changes will follow next week with an update on the pros and cons and a signal request.
  • Risk Scoring Methodology is in progress. For every price drop, we will have a good estimation of which users are safe and which are not and whether risk parameters need to be adjusted.
  • Community voted to offboard eight Vault types. Growth CU proposed to discuss with the teams behind these tokens to see their potential to make more DAI, so we would not need to offboard them. The main reason to offboard them was because of oracle costs. Nadia suggested keeping some of the collaterals because of the potential for business development benefits. It is a matter of oracle cost vs. business development benefits.
  • From a risk perspective, it’s unlikely that lower-tier collaterals such as 0x, Loopring, BAT, MANA, and so on can have high-depth ceilings. The launch liquidity is low for these collaterals, which is risky because of slippage from Liq. For example, if you want to sell two million of 0x, you will incur 45% slippage on the chain, which is very risky.
  • Maker is not well-positioned in terms of leveraged and compounded assets. If you hold BAT or 0x, you likely hold THETA or other altcoins. If you want to leverage, you would likely engage in cross-collateralization. AAVE is much better positioned for compounding.
Oracles - Nik Kunkel

56:23

  • High gas costs generally mean higher associated oracle costs. This is only sustainable for collateral types that bring in significant amounts of DAI or collateral types that have the potential to bring in significant amounts of DAI.
  • There is a need to streamline internal processes from offboard discussion to execution. Offboarding vaults should not be controversial. If the liquidity profile does not support much traction, we should be able to offboard it.
  • This is not an exception but the norm. We have successful vaults like YFI and LINK. For the most part, though, onboarding yield tokens did not have very successful results. Offboarding these tokens should not be controversial. If the point of not offboarding Vaults is to provide favors with these teams, we are essentially paying out of pocket for a potential but unrealized benefit.
  • Assets should not remain onboarded simply because we’re talking with the teams to monetize treasuries. These should be treated as two separate conversations.
  • “Onboard liberally, offboard liberally.”

Discussion

  • Nadia: We should have further discussion regarding monetizing treasuries. We are not staying with Loopring simply because of a potential benefit. Loopring is creating an L1 solution tool to help launch Maker Vaults on L2. It will not just be for Loopring token holders but other token holders as well. The vision we have on growth is that we should form partnerships instead of putting a proposal on the forum and accepting collateral. We should set agreements with other protocols. This is part of the process behind offboarding, which can offboard the tokens while maintaining partnerships with these communities. It is important to ensure that these partners will follow through on their end in these partnerships. For reference, Polygon promised several goals that they did not achieve by their designated time frame. There is a need to separate the relationship with partnerships versus the reality and the current Vaults. Vaults can be renegotiated. Right now, there are conversations and communication between different Vaults, but there is no clear framework that establishes what we want. It is not just about offboarding collaterals, but onboarding collaterals as well.
    • David Utrobin: Frequently, we have good business logical reasons for offboarding Vaults. While we want to maintain business partnerships, offboarding Vaults do not necessarily negate relationship management. We are spending $250000 a year on Oracle fees, and we need to ensure that these collaterals fulfill the fundamental goals of Maker.
    • Kurt: We should be focusing on the next 10x Dai - are we going to get 10x, 1x, 2x? What’s our floor for our interest rate?
    • Nadia Alvarez: We need to focus on Institutional Vaults. If we focus on these tokens that can bring large Vaults to the Maker protocol, like Bitcoin and Ether, or monetizing treasuries, that will be a different Institutional Vault.
    • David Utrobin: Can we make smaller people that want to be part of Maker subsidize themselves to be here? Can we have them pay for their oracle costs?
    • MakerMan: We need to do a one-billion-dollar deal and focus on putting things in line. We need to compete with the rest of their protocols, with how they are playing it. We can bring that ten wBTC over to Compound, over to Maker, with bigger collateral types. I want to figure what happened with Polygon and Nexo. We need to focus on the bigger players. Let’s figure out how to grow the next 10 billion, not the next 10 million. We are bringing RWA players because we have to. We can’t grow this way, focusing on 10 million at a time.
    • Kurt: Regarding my earlier comment about 10x: think about the revenue that would be brought in by a 10x of DAI supply. Maybe we should be working on that and eat the cost of the smaller things. We need to grow focus on that. We are spending more time on minutia without properly sizing the importance of discussing and debating. There should be a streamlined process for offboarding, and we should do that now rather than later. We have been talking about offboarding for months. I do not understand why we are wasting so many mental operations and time overhead for offboarding operations when it should be much simpler.
    • MakerMan: Either do it or say that we are eating those costs. We should be talking about more productive things.
    • Nadia Alvarez: We received a message from Binance that they want to release the pair MKR because there is not enough liquidity. They asked if we could release funds for liquidity to maintain the pair. Perhaps that is something we can discuss regarding offboarding the pair.
    • MakerMan: Does this not just come back to a framework for onboarding/offboarding? If you maintain the minimum requirement, you’re onboarded. If you no longer meet those requirements, you’re offboarded.

Conclusion

Payton Rose

01:16:38

  • We will talk about the migration away from RocketChat in next week’s call.
    • David Utrobin: If you have requests for discussion topics, feel free to reach out to Payton or David.
    • Thank you to all those who wrote in the chat regarding their experiences with the call and its new format.

Suggestion Box

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

Liq: Liquidations

CU: Core Unit

Credits

  • Andrea Suarez produced this summary.
  • Gala produced this summary.
  • Artem Gordon produced this summary.
  • David Utrobin produced this summary.
  • Zainab Nisa produced this summary.
  • Everyone who spoke and presented on the call, listed in the headers.​

The full call recording is now available on the MakerDAO Youtube Channel: