Governance and Risk Meeting: Ep. 50 (August 29 - 2019)¶
Video | Audio | Discussion
| References | Person | Text |
|---|---|---|
| # / 00:00:03 | Richard Brown | Hello everyone, and welcome to the August 29th edition of the scientific governance and risk meeting. My name is Richard Brown. I am the 50th, oh my God, somebody should've gotten me a cake. What's going on? All right, the 50th edition of the scientific governance and risk meeting. My name's Richard Brown. I am the head of community development, I lost that for a second. But I'm back. |
| # / 00:00:28 | Richard Brown | I'm going to keep things short and sweet because I think that we are trying to come up with ways to make these calls more dynamic than they have been. We're interating over time and things have been going well. I'm happy with these calls. Enormously happy. But, there's still room for improvement and I want to try and raise the signal to noise slightely. And, also mix things up a bit. I'm trying to convince Cyrus that starting with governance every week is unfair because of all the influence there and emphasis we put on the government stuff, and risk kind of gets squeezed sometimes. |
| # / 00:01:01 | Richard Brown | We might be alternating on each one of these calls. Today, we're going to start off with governance, next week, potentially, we'll start off with risk. So everybody has an opportunity to put some emphasis on a part [inaudible 00:01:15] Excuse me. That they find most compelling. Oh yeah, the preamble, I almost forgot. We want to hear from our community, and this is a really great way to get immediate feedback, and it's also a great way to ping people in the foundation. If you have questions or comments, please type them into the chat at the side window. |
| # / 00:01:35 | Richard Brown | If you have a microphone, please get on the mic and interrupt us. We don't mind. I think that's it for preambles. Do we have LongForWisdom, he is here? So I'm not going to dig too deeply into all the interesting things that are happening in the forum, although those are mostly some of the things that I want to talk about today. I'll allow him to do the color commentary. But, I am going to remind people that we have two forum polls happening right now that are water shed moments in the epic history of MakerDAO governance. This is the first time that we've had a forum poll that will provide us signal to initiate a governance poll. |
| # / 00:02:21 | Richard Brown | This is a long conversation we probably might not get into today, unless people have questions about that, but we needed to figure out how do we initiate a governance poll, and then in order to gather signals from the wider community, the Maker holders sorry. And the question arose, “Well, how do we achieve enough signals to initiate a governance poll in the portal?”. The answer, after a great deal of discussion and back and forth, was the forum is the place to do that. This is the first example of that activity occurring. |
| # / 00:02:54 | Richard Brown | After I talk for a bit, I'm going to post some links to those forum threads. Please, familiarize yourselves with the conversations leading up to that poll. Make your voice heard, and then we'll see how governance actually works in practice, which is going to be exciting. The other thing I want to talk about... Actually, yeah, I'm going to keep things philosophical. I want to put some questions out there, and this is going to be an experiment, because we... I keep on saying that I'm super interested in hearing from the community, we want to get all this feedback. |
| # / 00:03:29 | Richard Brown | Then I basically say, “go”. And everybody is surprised at that, and we don't get a lot of questions. So, I want to try and mix it up a bit. What I'm going to do this call for the governance section, is I want to throw some questions out there. Some things for people to think about. And then, pick them up at the end of the hour during the Q & A session, and see if anything had percolated by the time that the calls over. And then we'll dig into some things. |
| # / 00:03:59 | Richard Brown | Here is what I want to talk about, or I want the community to think about, because they're kind of hot button topics. Actually, I don't know if I'm using that phrase correctly. They're popular topics in the general conversation room. Governance, and they are these: is delegation a thing that we should do? And it's an interesting conversation because somebody pointed this out. I think it was Kwadrax in the forums, and I've had the same...probably, I had a different reaction, but I had the same amount of interest in the fact that this has happened. |
| # / 00:04:35 | Richard Brown | Because we're in a space where, depending on how crypto-anarchic your roots are, the general intention here is that it's self-sovereignty and it's taking control of your destiny. And it's having a say in the decision-making process. And one of the interesting things we've seen, and I understand completely, is that that kind of a world is tremendously complicated and involves a lot of work. And maybe not everybody wants to do that. And so, if I had to guess, and this is anecdotal, I haven't taken any polls, but the number one requested feature when it comes to the government's ecosystem is some kind of delegation mechanism so people, and kind of be relieved of this enormous responsibility of researching MakerDAO for hours a week every week to figure out what's going on. Lots of people want to identify experts or people that align with their general interests, and have some kind of mechanism that would let them transfer their staked Maker to a third party. |
| # / 00:05:41 | Richard Brown | That's a very interesting conversation to have, and the questions that we've been sort of wrestling with is "how does that happen?" And I personally-I don't speak for the foundation on this-but I personally am extremely reluctant to add additional layers of complexity of making those into another foundation created governance ecosystem. I think a great way to move forward would be to follow the same principles that the crypto space and open-source in general has is that MakerDAO generates these financial primitives, it generates contracts, and then the third party, the community, comes in and builds the tooling on those primitives that they'd like to see and I think that we can look at the governance ecosystem in the same way that we've looked at the protocol itself. The DeFi protocol that we're building itself. |
| # / 00:06:31 | Richard Brown | So, that is leading me up to posting this link in the chat, and this is a link to HackMD that Chris Hopcroft has put together. He has been working with Aragon, or on Aragon, I should say, with DAOs for a while and he has some really interesting ideas about how to extend Aragon with connectors to allow it to do other things, so a group of people could get together and self-organize and put some assets into Compound, or they could do anything with an external Dapp, but organize through Aragon. I think that's a very compelling idea. I pinged him after I read about his Nest application and suggested, “how about you do something like that for governance?” Because it seems like a really great way for the community to begin to self-organize. Groups of people could say, “hey, I like the way you think. I'm also enormously conservative.” Or “I'm also frustrated with the lack of aggression in the system, there's all kinds of caps.” |
| # / 00:07:40 | Richard Brown | These people can get together, they can fire off an Aragon DAO. They can use the mechanisms outlined in that link that I just posted to somehow safely pool their maker together and elect a voter that would keep up with activities in the governance system, and behave in a way that the group finds acceptable. That's a long and interesting conversation, and we're looking for interest from the community people to offer up some suggestions about use cases and mechanisms if there's technical people in the community that understand how Aragon works and/or have deep insight into proxies, DS proxy, and the chief works at MakerDAO. Please hop into that HackMD and offer up some suggestions, or feel free to reach out to Chris. I can put some contact information at the bottom of that HackMD chat. I think I did. We're really looking for the community to solve problems for the community and the foundation is here... |
| # / 00:08:52 | Richard Brown | Excuse me, that was very- the foundation is here to facilitate that activity so we would certainly love to give Chris and his team, a grant to solve a problem that makes life for the community easier. So, if you're interested. Have a look at that HackMD. |
| # / 00:09:11 | Richard Brown | And at the beginning of my speech I threatened to throw out some questions that I'd like for people to ponder, and then we can pick them up potentially at the end of the hour. And these are going to sound silly, but I think that they're important, and they're going to be hard questions. These are capital H hard questions. But what I would like people to start thinking about is "what is consensus," and that sounds, that sounds dumb but it's something that's worth thinking about. Because a lot of the times, or frequently, when we're thinking about governance, we have a habit of falling into shorthand and euphemisms without realizing it, and consensus is one of those shorthands, at least when it comes to social consensus. |
| # / 00:09:52 | Richard Brown | Consensus mechanisms there's no ambiguity about it but when it comes to people talking about things in a group, figuring out when consensus has been reached and how much consensus is enough, and all the other ambiguity around that thing are enormously difficult. And so when it comes to the forums, when it comes to these debates, when it comes to the governance facilitator mandate thread, or the risk team mandate threads, these are questions we need to start answering. At what point was consensus reached? Is there a consensus? How much consensus is good enough consensus? When do we start to poll? Do we- sorry, I immediately lucked into the second question I want to ask, but here we go. |
| # / 00:10:36 | Richard Brown | We have this debate: a mandate has appeared in the forum. Does that mandate come with a poll immediately? Do people signal their support or dissent for that mandate of before the debate happens? What happens if that mandate changes materially based on the debate, yet people have already voted on that poll? Can we expect them to come back and update their votes? |
| # / 00:11:02 | Richard Brown | If not, I'm not just asking leading question. It's not going to be hard to figure out which way I understand this, but if we wait X number of days after a mandate has been posted for a debate to occur and then uphold posted, how long does that poll exist in the forum before we consider that done? |
| # / 00:11:23 | Richard Brown | Yeah, so these are interesting questions. And happily we have LongForWisdom in our back pocket. And he's set up a really great thread in the forums as well and I want people to keep on revisiting, and that is the signaling guidelines thread. David, can you dig that up for me? I had it in another window but I lost it somehow in the last ten minutes. |
| # / 00:11:46 | Richard Brown | And this is another great example of immersion process that's happening in the forum where, as we go along, we're figuring out how things should work, or how they work for us, at least, and we're refining that. And I would love for people to take a look at the signaling guidelines thread and help us refine this, because this is going to be the kind of documentation that comes out of the forum and ends up going into our portal, and ends up being canonical. We're expecting people to follow these social conventions or these governance conventions as we come up. |
| # / 00:12:24 | Richard Brown | I should be, maybe I'll leave that for philosophical questions. Let's think about consensus in the second debate for today at the end of the hour. If anyone has questions about governance facilitation, the way that that conversation occurred, anything about the hard questions and governance, or signaling, let's chat about it at the end of the hour will have governance Q&A. But at this point, my 15 minutes are up. I'm going to hand it over to you, LongForWisdom, if you want to give us a recap of the form, assuming I haven't stolen all the good parts already. |
| # / 00:13:00 | LongForWisdom | Yeah, can do. So I guess the thread with the most engagement this week was the "Why aren't MKR holders voting?" thread. |
| # / 00:13:11 | LongForWisdom | So, to sort of give a bit of background on this, I posted a thread on Reddit sort of encouraging people to go vote and get involved more, which has sort of mixed results, some people. There were a couple of people who did appear, which is awesome. |
| # / 00:13:26 | LongForWisdom | We also got a big list of reasons that people don't vote. |
| # / 00:13:31 | LongForWisdom | And so, I obviously summarized or correlated those in that, in that thread in the original post. And everyone's been discussing, which has been good, because we've come up with a couple of things, couple of ways to mitigate some of them, and recognize that some of them are really beyond the scope at the moment. |
| # / 00:13:50 | LongForWisdom | So my hope is that we, in that thread we can look at the others and then sort out which one is the things we can actually do something about as community and which ones we can't at the moment. |
| # / 00:14:03 | LongForWisdom | So that's sort of current affairs. The other current affairs thing that's probably worth mentioning is that [inaudible 00:14:07] not so much in the forum and chats. |
| # / 00:14:12 | LongForWisdom | So, we'll be on to that later obviously, but, yep. In terms of the [inaudible 00:14:17] polls up on the mandate for risk teams and interim governance facilitators. |
| # / 00:14:23 | LongForWisdom | So, suggest signaling on those discussing any points that you think are still open. |
| # / 00:14:29 | LongForWisdom | Another thing I'm going to sort of bang the drum for again is this signal request to moving towards exponential rate stepping for stability fee polls. |
| # / 00:14:40 | LongForWisdom | So, I brought this up last week, and I'll bring it up again. We're now at 62% consensus, I believe, with 24 voters last time I checked, so definitely got a couple more people voting since last week, but we still sort of lack a really strong, sort of, obviously 60% is a majority, but we don't have a really strong mandate for influencing a specific option yet. The general consensus seems to be that we want to move to exponential stepping. With the exception of Cyrus everyone's in favor. But there's sort of disagreement over what the best way to do those. |
| # / 00:15:19 | LongForWisdom | That would be the main threads. It's not really [inaudible 00:14:25] there if anyone has question? |
| # / 00:15:27 | Matthew Rabinowitz | To ask just one quick question about that, maybe it's better to save for the end of the call, but what is the harm if we have basically a fundamental idea of going to exponential rate or not? If we have a system now that allows for multiple votes that can all happen in parallel, where do we reach the point where we just say "put it to a vote and move on?" |
| # / 00:15:48 | Richard Brown | Yeah, that's one of the harder questions that I'm hoping people can talk about at the end because that's "When does consensus happen?" And I think that if we wait for overwhelming consensus or we wait for enough consensus- these are all very, very murky questions to ask. And so, potentially one of the solutions I want to- or one of possibilities out there is that we have this debates, with the debate goes on for this, we can make it arbitrary. Exponential rates stepping and we're going to talk about this for two weeks. That's the window. After two weeks are over, automatic automatically a forum poll will start. That forum post last for seven days and that's the window for people to get in there and signal. And after that... |
| # / 00:16:32 | Richard Brown | Based on that, those results, either something happens or something doesn't. I think that not having a calendar, and that's why calendaring is built in so heavily into the governance facilitator mandate, is that as we're mapping out this role that became increasingly obvious one of the most important things [inaudible 00:16:51]. Somebody needs to be able to establish and start and stop date, and make sure that things happen on time. We're kind of seeing the evidence of that with the exponential rates stepping, and the stability fee vote thing. Because this has been going on forever. We need to, well, potentially we need to have a better rate stepping and we need to consider not voting on the stability fee every single week, and set a facilitator to set up these arbitrary deadlines. I think that the chances of it moving ahead are pretty slim. |
| # / 00:17:22 | Matthew Rabinowitz | To add just one extra commentary, and we'll get, probably, to the end of it. But I also don't think we should set with the mindset that a decision is a perpetual decision forever and that it could never be readdressed. I mean, it's not the Supreme Court. If we implement exponential rate stepping and we don't like it after six months we could always put together another vote to undo it. It doesn't perpetually push it off the table for discussion. |
| # / 00:17:45 | Richard Brown | Yeah, and that's great. Thank you for raising that point. And this is something I think that we might not be stressing enough in these calls and in the governance ecosystem in general. We're not releasing proclamations here. You know what I'm talking about, this is all iterative. This is why I keep on saying that this is open-source governance, and we can look at governance as being code, because we come up with our best efforts based on the information available to us. We release that to the community, the governments facilitate our mandates, the rate, the rate stepping, the risk team mandate, anything that comes out of this ecosystem is designed to get us two, or thee or four, six months down the road. |
| # / 00:18:29 | Richard Brown | If it turns out that we need to course correct, then it's trivial to have another vote, and then update the governance facilitator mandate, update the rate stepping, update the cadence of the stability fee. So everything is up for debate. And I think that was addressed in one of the forum threads where, I think it was a crisitcism of Cyrus's doc, where in the mandate, the risk team mandate said "this is a living document." And I think that it was interpreted as the forum thread is a living document, but it's not. These mandates aren't living documents. They evolve constantly. That's our expectation. If these were written in stone it would be disastrous because we can't predict the future. We need to respond to new pressures as they arise. And so everything here is hopefully going to be iterated every time we find a flaw in the system. All right, LongForWisdom, was that the end of governance at a glance? |
| # / 00:19:33 | LongForWisdom | Yeah, pretty much. I have fairly strong feelings on things we are talking about, but I'll save them for the QA. |
| # / 00:19:39 | Richard Brown | Yeah, I think that's probably going to be a good opportunity. We should, at the end of the hour, let's resurrect rate stepping. Let's resurrect stability fee cadence and everything else that's actually in these threads as opposed to pointing out which ones are great. I think at this point, let's hand it off to risk. Cyrus, do you want to take it away? Or actually you're onto some kind of jet train- |
| # / 00:20:06 | Cyrus Younessi | Can you guys hear me? [crosstalk 00:20:08]. |
| # / 00:20:09 | Richard Brown | Yeah. |
| # / 00:20:09 | David Utrobin | Yes. |
| # / 00:20:12 | Cyrus Younessi | Okay, cool. Yeah, a couple interesting things to talk about today in risk. Before we jump into the DAI price I'm going to do a quick review of last week's risk team mandate and see if there are any offstanding questions. And then I'm going to address a point that was brought up in the forum thread which I thought was really good. |
| # / 00:20:33 | Cyrus Younessi | Okay, so, essentially, the risk team mandate is a proposal for a structure for risk teams and how they are onboarded into the governance system and what kind of documentation we can expect from them. And when I say "we" I mean what the MKR holders, can and/or should expect from them. Basically it's our attempt to deal with a little bit of the chaos that should result in just completely decentralized risk process. That being said, certain areas were kept, maybe, intentionally big because, as Rich said, a lot of governance is hard to predict, and we don't want to be over-committal in certain aspects of it. At the same time, though, we wanted to set a high bar for quality in terms of the professionalism and diligence that we can expect from the risk [inaudible 00:21:40]. |
| # / 00:21:42 | Cyrus Younessi | So, that being said, does anyone have any outstanding questions from the risk team mandate? |
| # / 00:21:54 | David Utrobin | I had a question, basically, about, is there going to be a hiring- |
| # / 00:21:59 | Cyrus Younessi | Okay, cool, so there- |
| # / 00:22:00 | David Utrobin | Oh, sorry Cyrus. Did you hear me? Can you hear me Cyrus? |
| # / 00:22:08 | Cyrus Younessi | No, sorry. Can you repeat the question? |
| # / 00:22:10 | David Utrobin | Yeah, sure. I just had one question. |
| # / 00:22:12 | Cyrus Younessi | Yeah, can you- |
| # / 00:22:13 | David Utrobin | Yeah, I'm repeating it. So the one question I had was around how we hire individual people within these risk teams. I know in the mandate it kind of envisioned people posting up their resumes and credentials, and what they'd be focusing on and stuff. But do you people having an issue with doing this publicly on the individual level? |
| # / 00:22:46 | Cyrus Younessi | That's a good question. I don't think so. But actually that kind of ties into the question I wanted to address anyways, so I'll just jump into that. I think it was Kwadrax or somebody. On the forum it mentioned, "Are there risk teams behind the scenes, or what's the actual status update on that?" And I responded that there isn't some hidden behind-the-scenes slew of risk teams waiting in line, and how we envision that to happen is essentially, I think we're going to try to lean on community support, community outreach, not necessarily from our community but communities of the collateral partners. |
| # / 00:23:44 | Cyrus Younessi | So, for example if some collateral partner, whether it's a crypto-asset or not, wants to onboard their asset into MCD, basically, currently what's happening is they're just generally coming to us on the interim risk team and saying "We love you guys, we want to become part of the project." And I think on their end, they don't realize that there's essentially 20, 30 independent projects trying to work their way into MCD. And at a certain level it just becomes completely infeasible for us to handle the entire process for all of the project. So what we're doing is we're trying to- I kind of view my role as a facilitator to help coordinate with the collateral partner, this search for an independent risk team who can come along and work on the risk evaluation for these collateral assets. |
| # / 00:24:56 | Cyrus Younessi | So for example some of our collateral partners are a Danish company looking to make small business loans in Copenhagen. Other is South American logistics company, and so forth. And it's very globally diverse, and it doesn't really make sense for us to try to become experts in everything out there. At the same time, it's not entirely responsible to have the collateral partners propose their own risk parameters, even though in many cases they have done a lot of the work themselves. |
| # / 00:25:40 | Cyrus Younessi | So essentially we are working with our collateral partners to help source independent risk analysts. And that could be either community members, or it could be professional firms, professional consultants, accounting firms, whatever. So there's a few different ways to do it. But I think, essentially, it will be driven by the community of the collateral partner. To be clear, if we do get around to it we're happy to do it ourselves. I'm just truly speaking from a priority perspective. So we did have the priority poll a few weeks ago, right. And we have to, at some point, estimate our pace of work and how long it would take us to get through the list. |
| # / 00:26:34 | Cyrus Younessi | If the community of a particular project says, "you know what? This is gonna take too long." Then they're more than happy to push forward an independent risk team and go from there. In terms of the public credentials question that you mentioned, [inaudible 00:26:55] fine if it's some professional firm that handles this stuff on the daily, then that kind of makes total sense. I think in all cases, it should be fine. I'd have to hear some edge cases to think about why that wouldn't work. But the broader question should really be is, "to what extent can all of MakerDAO's processes be public. And that's an issue that I do think we'll run into eventually, especially in terms of data sets that, potentially sensitive data, that risk teams might use to [inaudible 00:27:40] these collateral partners. That would be an interesting obstacle. Does that makes sense? |
| # / 00:27:53 | David Utrobin | Thanks, yeah. |
| # / 00:27:54 | Cyrus Younessi | I hope I didn't cut out. |
| # / 00:27:55 | David Utrobin | Yeah, I think that made sense. You didn't cut off too badly throughout the answer. |
| # / 00:28:03 | Cyrus Younessi | Okay. |
| # / 00:28:03 | David Utrobin | I had the additional question- |
| # / 00:28:04 | Cyrus Younessi | Uh, |
| # / 00:28:06 | David Utrobin | I had the additional question of, do you have a sense of who or what will be the real first example of this? I know that the priority list results listed ETH first, REP second, BAT third, so it might be a bit premature to even ask my question. But do you expect people from the REP community, for example, to have somebody who's really technically knowledgeable about REP and all of the tech, for example, to cover the technical risks? Do you see that as going to be the first example of what you're talking about? |
| # / 00:28:41 | Cyrus Younessi | Yeah. That's probably going to be the first example. The current plan is for the interim risk team to evaluate ETH plus BAT and REP, and so I totally expect us to make some contacts on that front. But that's just kind of part of our information gathering. But I could also see, so for example we've had some other crypto assets that weren't part of the first priority list, they've already completed the onboarding collateral application, and they are essentially just getting ready to post it to the forums. And when they do that we might bring up a discussion of, well first they would get added to the next priority poll, and in the meantime we could encourage their community to put forth someone who can do an objective analysis of the collateral asset, or aid the interim risk team in doing their evaluation. |
| # / 00:29:56 | Cyrus Younessi | But the specific answer is we don't know exactly how it will look. |
| # / 00:30:06 | David Utrobin | Thanks for the answer. |
| # / 00:30:09 | Cyrus Younessi | Cool, okay. That being said, I think the risk team mandate, just like the government facilitator, will be an ongoing dialogue over the coming weeks and months. So, I'm going to leave it there for now if there are no other questions, and then we can jump into the monetary policy, and I think there are a few interesting things to talk about there as well. Vishesh, if you're on the call, do you mind taking over? |
| # / 00:30:46 | Vishesh Choudhry | Sure. I'll just share my screen. Okay. Can you guys see the graph? |
| # / 00:31:04 | David Utrobin | Yes. |
| # / 00:31:05 | Vishesh Choudhry | Okay, cool. So, just in the last 24 hours, pretty significant amount of DAI traded, the volume weighted average price, just a center, so above a dollar. And then if we look at the last seven days, The DAI price was trading roughly a cent to two cents above dollar with, interestingly, fairly average seven day trading volume. So it was pretty quiet, and then front loaded the last two or three days or so. But, yeah, you see kind of classic micro-structures here, just pretty well centered around, like I said, approximately a cent above a dollar. |
| # / 00:31:58 | Vishesh Choudhry | I think some of this was due to a little bit of re-leveraging. So that's something that we can touch on. |
| # / 00:32:10 | Vishesh Choudhry | So what's interesting is, that in the short term, what we had seen is a lot of draws and wipes canceling each other out. But actually, in the last day or two, We did see a little bit more wipe than minting. And then in collateral terms, we saw, things were pretty quiet. But you know, collateral does periodically get added as the price falls. And then we can see here is in terms of liquidations, there were some liquidations recently, but not as much as when ETH dipped about two weeks ago. So that price has been kind of slowly coming down a bit, which I know a lot of people are not happy about. But what's interesting is maybe we can look a little bit at the leverage to kind of understand that. |
| # / 00:33:08 | Vishesh Choudhry | So let me just go over here. So some of the cumulative graphs. Essentially, this ETH price has been coming down pretty consistently for the first part of 2019. There was sort of a run up and now it's come back down, again, to basically just end of 2018 levels. Or slightly, slightly below. I mean, it was roughly November, December when the price was last at this level. So what's interesting is the amount of collateral that had been locked in the system was consistently growing, the amount of collateral that was being removed from the system, voluntarily, was actually being outpaced initially in the first part of 2019, and picked up pace in mid-2019, and has now sort of flat lined. And this makes total sense, because essentially, as that ETH price was running up a little bit, There's some excitement, a little bit of re-leveraging, a little bit of taking some collateral out because it had risen in dollar terms. And now with the drop, this is pretty flat, or even somewhat, slightly increasing but more or less outpaced by the amount of collateral being added. |
| # / 00:34:36 | Vishesh Choudhry | So this is a behavior that I think we now pretty consistently say is expected. So essentially, as the ETH price takes a hit, people will stop pulling collateral out and they'll actually start putting some more in. And as the ETH price rises, and people are feeling bullish, they'll feel more comfortable pulling collateral out. So this is pretty well explained phenomena at this point. What's been happening in terms of debt and supply: so essentially, over time, we can have seen this consistent increase in the DAI supply, and so what you saw was essentially this draw line was growing, growing and growing and growing pretty rapidly. But the white line in terms of voluntary shuts and wipes had essentially kept pace initially and then sort of gotten outpaced by the amount of debt that was being minted, but just in the last few months, had picked up and accelerated and started to close the gap. |
| # / 00:35:47 | Vishesh Choudhry | And so that's where you started to see this DAI supply start to come down. There were also some liquidations, but then we essentially see this precipice around mid-July where just you know, that draw amount is still somewhat consistently growing, slowing down a little bit, but then that wipe line is just accelerated rapidly. What's interesting is, since you know, maybe the last couple of weeks, that wipe line has also slowed down. So the supply has actually been in a relatively steady state. So even though in the in the week time frame, it does look like there's a bit more wipe than draw, in the grand scheme of things it's still pretty comparatively flat. And then we see in the longer time scale as well. So essentially, that ETH prices it takes those hits, there's some slight stepwise jumps in the amount of debt that's liquidated. |
| # / 00:36:46 | Vishesh Choudhry | And so you know, you saw huge jumps at the end of 2018. Still pretty unparalleled. And then some slight jumps around some of these smaller dips in the beginning of 2019 in April, turn of the month, and then somewhat significant one, just like in that same mid-July timeframe. So a couple of smaller ones two weeks ago, and then very small one very recently. But by and large people have been, I guess you could say managing their risk better, and collateralizing their CDPs to counteract some of these risks. |
| # / 00:37:29 | Vishesh Choudhry | So yeah, we can see again, the collateral amount in nominal terms versus the ETH price, essentially just a converse relationship. And that collateral amount has come down a little bit during the bull market, and then sort of recovered and keeping comparatively flat in the last couple of weeks. And again, this is just ETH price versus the debt. |
| # / 00:38:01 | Vishesh Choudhry | Not a super strong relationship between these two. |
| # / 00:38:05 | Vishesh Choudhry | So one thing that I did want to touch on was, so we saw a lot of these trades happening in the past week or so reduce the- the variability of DAI prices been comparatively lower, I think, than some of what we'd seen in the past. So if we pull up an older graph here, you'll recall that we had seen some pretty significant variability in the DAI price in the beginning of 2019. And essentially the standard deviation of that, just go back to a couple weeks ago, the standard deviation of that had sort of come down a bit. And then this price, though it is hovered, sometimes above, sometimes below a dollar, it's actually decreased that spread quite a bit. And so what's interesting is, even this is a pretty significant spread for the past couple of months. |
| # / 00:39:07 | Vishesh Choudhry | But pretty tight just hovering either slightly above or slightly below one dollar. So DAI is comparatively pretty stable, is I think, the takeaway there, though, at the moment trading slightly above. And what's interesting is this is even post, the most recent, stability fee change, it's had a few days to normalize, and we haven't seen a huge change in what's been going on with DAI price. So I'm fairly confident, whether it's correct or not, there's going to be, and I've seen this in the forums and the chats, some people pushing to lower that stability fee few again. Listen, in terms of addressing that, that potential proposal: the collateral amount is something that I want to look at. |
| # / 00:39:59 | Vishesh Choudhry | So this collateral value in the past two weeks or so has been comparatively stable. Collateral has actually come up slightly since the 18.5% stability fee decrease. So what's interesting is in there were small amount of liquidation. So that's really where that comes from. So stability fee drop, this change really doesn't have much to do the stability fee. Supplies actually come down slightly, it's come back up in the last few hours, just a hair. But basically, it was sitting around 78 million for the past couple of weeks, tick down very slightly tick back up very slightly. So not a ton of movement there despite the stability fee change. |
| # / 00:40:48 | Vishesh Choudhry | So there's not a huge impact on leveraging behavior, which I know has been one of the worries that we have is when we lower the stability fee, the potential that you just can incentivize significant amounts of re-leveraging, and crash the DAI price, that hasn't really happened in the past couple of days. Although the DAI price, I would say, is also probably propped up by a little bit of purchasing behavior. This is the most likely explanation of this. So yeah, ironically, though, changes are being made somewhat quiet on the Maker front. Yeah, I guess maybe we can open it up to questions, and hopefully I won't eat up all the time this time. |
| # / 00:41:38 | David Utrobin | So Shesh, I have a quick question and I want to like echo the sentiment that I saw on the chat earlier today. I think there was a bit of like urgency in how consistently DAI has been trading above the peg. And I know that a lot of people are like should we- there's I just see the conversations that there's a big tendency to have like an emotional response that like triggers, like overshooting or something. And also, there's like concerns about whether or not we can make the change quickly enough in response to these kinds of situations. So, so yeah, I don't know if you want to comment on that, because I feel it's a bit of an [inaudible 00:42:24] factor. |
| # / 00:42:26 | Vishesh Choudhry | Sure. |
| # / 00:42:27 | Vishesh Choudhry | So I want to just flash back to here, maybe. Okay, so, Early July was probably one of the more quote unquote, urgent moments of, "hey DAI is taking consistent hits. We should increase the stability fee." This was also probably prior to having some visibility on what's been going on with secondary markets. So even then, we weren't really talking as much about utilization rates and things like that, which I've ticked up slightly since the 18.5% stability fee increase, which is actually pretty interesting. |
| # / 00:43:06 | Vishesh Choudhry | So DAI has gotten slightly cheaper. Utilization has held pretty steady or actually increased. So what that means is, from a scientific standpoint, those are actually like metrics that you could use to justify making a case for a stability fee decrease. |
| # / 00:43:22 | Vishesh Choudhry | Is there anything that I see that is sort of prompting worry or concern that the price is too high? No. It's, again, maybe a cent or two to above by large. Of course, there's a spread always. But in general, the bulk of the trading over the past week, is just a cent or two above. So there's nothing urgent about it in my eyes. I don't see any of that in any of these graphs. |
| # / 00:43:56 | Vishesh Choudhry | That being said, could you make the case for "okay, the fee was decreased, and everything was more or less fine, so you could potentially decrease it again?" |
| # / 00:44:05 | Vishesh Choudhry | Yes, that case could certainly be made. I would again, prompt maybe thinking about small changes half a percent, a percent. And I know we've been having this other thread about "how should pulling be structured?" And all that, what should the options be, but just in terms of making well founded, careful non-rash decisions, dropping by half percent, to check and make sure that this will continue to stay steady, and that you can offer the best rates possible, does not seem unfounded at this point. But no, I don't see in either direction at the moment anything for concern, like I said, it's actually been comparatively quiet. |
| # / 00:44:51 | David Utrobin | So to respond to that. I know the governance poll just concluded with the community agreeing to lower the fee 16 and a half percent, or at least not to lower the fee, but to put an executive vote to lower the fee to 16 and a half percent, which is 2% down rather than a smaller half a percent move. And I think that the main concern is that DAI is trading so consistently for so long above a dollar. Because it hasn't just been seven days that this has been going on. I think it's been almost two weeks if I'm not mistaken. And so does the duration of DAI trading above the peg, at even if it's just a cent, or a cent and a half or something, does that warrant a bigger move down? I mean we can always readjust next week, right, if it's an out-sized move. |
| # / 00:45:40 | Matthew Rabinowitz | I wouldn't say that I wouldn't say it's a larger amount would be warranted. The real question in the info stack, supply and demand and when they're in equilibrium. If both of those two items are static, demand, we will figure out what the supply should be. The issue is that we don't know what demand is. We can control supply, and we're using a stability fee to try and find where demand is. It's just elusive. We can't quantify it yet. The question really becomes, if we can quantify it, and if it's growing, and if demand is in the driver's seat, right now, we're just on a unicycle, and we have pushed on one side, now we've passed over an equilibrium, and now we have to accelerate to catch ourselves. |
| # / 00:46:25 | Matthew Rabinowitz | That's the challenge we have in single collateral. And that's really the question. Is there a need to do more than that? No, I don't think either. I agree that we should be at a smaller number, the communities agreed on 2% fine, but we should do something on a reoccurring basis to get back to an issue. We don't cross back the unicycle so far that we over adjust. |
| # / 00:46:52 | Vishesh Choudhry | Yeah. So I think there were a lot of things that you just brought up in quick hit, and David, I want to address your initial question as well. So the one thing is, I do think from a process management standpoint, consistent small changes, when supported by fact, is a good way of approaching things rather than large, sudden stochastic changes in rapid succession. |
| # / 00:47:21 | Vishesh Choudhry | As a process management principle, I do think that that holds. The other thing, addressing, half a percent versus 2%, et cetera. You've heard me many, many times advocate for these smaller percentage changes relatively regularly, rather than sudden 2% changes infrequently. So, if you were to draw like a two by two grid, in terms of frequency and severity of changes, I do think that the smaller frequent changes is going to yield superior results than large, infrequent changes. Which is more or less how things have been going, I mean, I'm calling 2%. large. And again, that's a discussion point. But yes, I will advocate for the smaller changes, because the bigger the change, the greater the likelihood that you're going to over or under shoot. That's a process management principle point. Now as far as like, what's actually going on, so the cache is a really interesting thing to look at. |
| # / 00:48:35 | Vishesh Choudhry | So I just filtered down to sort of that same August 20th range that we were looking at for the price here. So the point, David, that you made that I want to address that DAI has consistently been trading well above dollar, I don't think that that's necessarily true. DAI has been trading a bit above a dollar, and a little bit more above a dollar than it has below a dollar. But really, the bulk of that DAI has been trading maybe a cent or two above, and then even occasionally a cent or half a cent below. |
| # / 00:49:15 | Vishesh Choudhry | So I don't think there's some huge imbalance going on right now. But it is obvious that it is slightly above. I think, when slightly above maybe make slight considerations. And I think we've long tried to get to a rule or a mathematical formula, and like, how much should you change how often. I think generally, that's hard to do. If you hold one variable constant, you say, okay, the frequency is going to be a relatively consistent or constant variable, then you can isolate the other variable of magnitude, and start to think about that more formulaically. |
| # / 00:49:59 | Vishesh Choudhry | One really simple way would be, okay, if you keep fairly consistent changes on a scale of, say, half a percent to a percent, with some grades between. Or zero to one and a half percent with some grades in between. And then you determine how far on that scale you want to go, based on how above or below a dollar DAI has been trading, that would be a really simple way of doing it. |
| # / 00:50:26 | Vishesh Choudhry | A little bit more refined way would be to start to look at, hey, we have X amount of excess DAI on secondary lending platforms for X amount of time. And coming up with a really simple formula of saying, hey, that should translate to Y percent change in stability fee, would maybe be a little bit more refined. Or maybe some combination of the two. So as we start to think about better ways of thinking about these things, I think those are things to take into consideration. |
| # / 00:50:57 | Vishesh Choudhry | But again, it does not appear to me that in either direction, there's any sense of urgency, emergency, or extreme conditions going on. With regards to DAI specifically, when it comes to ETH, yes, there's definitely some reasons to worry about what's been going on lately. And there have been pretty significant movements in terms of price. So that long run average, yes? |
| # / 00:51:29 | LongForWisdom | Before you get into the next but can I just comment on the frequency thing? |
| # / 00:51:34 | Vishesh Choudhry | Yeah. |
| # / 00:51:35 | LongForWisdom | Yeah. So we're talking about frequency like it's fairly consistent. And the way the polls work, we put the polls out consistently, but because the executives run until they pass, right, the actual changes to the system don't come consistently. So as you saw, talking about like isolating variables, would it be easier to isolate the other variables to isolate the change variable to like 2%, which seems to be the favorite of like whales currently. |
| # / 00:52:03 | Vishesh Choudhry | I- |
| # / 00:52:04 | LongForWisdom | And then try to figure out the frequency, how quickly or like, how much time it takes it takes to take into effect, I guess? |
| # / 00:52:12 | Vishesh Choudhry | Yeah. |
| # / 00:52:12 | LongForWisdom | You get what I'm talking about. |
| # / 00:52:14 | Vishesh Choudhry | Fair question. I don't think it would be easier to try to isolate magnitude over frequency, I understand that there's some structural reasons why frequency can be a bit unpredictable. But it's sort of like, I don't know if I'm going to buy or sell something. But I know when I'm going to buy or sell, it seems a little bit more sensible than, I know when I'm going to make a transaction, I have no idea what kind of transaction I'm going to make. |
| # / 00:52:48 | Vishesh Choudhry | It does not seem to me to make a whole lot of sense of like, okay, we're going to make a 2% change, no matter what. And then it's either going to be positive or negative, depending on what everything looks like on, say, August 1st, and then August 19th, and so on, so forth. That does seem a little bit tough. I don't know. These are not easy problems. But- |
| # / 00:53:17 | Matthew Rabinowitz | Sorry, just to carry on to that thinking. It's like, imagine right now you're viewing it as a frequency of what happens every other week, or 10 days versus 14 days. We just view it as an entire calendar year. How many votes do we actually get implemented on an entire calendar year basis, the frequency makes if it's plus or minus a day or a week, one extra vote or one less vote really should smooth out over the course of an entire calendar year. |
| # / 00:53:46 | Vishesh Choudhry | Right. Yeah, I mean, if you're two days ahead, two days behind, here and there, as long as you're making sensible moves that are based on, say, at least a week's worth of data, it shouldn't be a huge deal. I think that becomes significantly more problematic when you're trying to make changes on two, three day timescales. Which, again, I didn't address frequent versus infrequent, but I've always been a proponent of making slower changes. |
| # / 00:54:24 | Vishesh Choudhry | They can be regular, they can be somewhat predictable, et cetera. But making changes within a three day timescale, or based on three days of data, for example, just never seems to be a good idea. And I think if you look at cases where multiple- I was just sort of looking at the history of Maker here. If you look at cases where multiple changes were made in quick succession, I think more or less they were then reversed, undone, or had to be otherwise addressed. So I think, from a scientific standpoint, as we start to build this history of data points, there's not a lot of data points right now, which is why historical analysis is tough. |
| # / 00:55:05 | Vishesh Choudhry | But my guess is the trend on this is those were cases of over and under shooting. And so it does seem to be that the system is not managed as tightly when changes are made in rapid succession. So it will be interesting to see what comes out of the next change or so. But I do think if you're making 2% changes within four days and four days, you tend to get into some interesting scenarios where it's maybe not the best management system. Sorry, I know it's eating up a lot of time. |
| # / 00:55:51 | Richard Brown | Actually, I want to jump in there. Because you just said something that's interesting, because we've talked about, "let's change the exponential rate to exponential rate stepping." "Let's change the cadence to two weeks, because that just makes sense to humans." But what you just pointed out, though, is that we we can look back at this empirical data that we have and see that all things being equal, which they never are in this space, we can see that if we bunch up a bunch of pile of votes together in a three week period, then traditionally, we end up on doing that change in X number of weeks afterwards. Can we come up with sort of this algorithmically same cadence on voting and rate stepping based on past data, or they're just too many confounding variables in the market for us to be able to just look at the graph to see what we should be doing? |
| # / 00:56:44 | Vishesh Choudhry | I unfortunately, think it's a lack of history problem. There's just not enough data points to make statistically significant conclusions on questions like that. But- |
| # / 00:56:54 | Richard Brown | Some point though? |
| # / 00:56:55 | Vishesh Choudhry | I do think from common sense- well maybe yeah, from a certain point. But I do think from a common sense standpoint, we can evaluate this question of saying, if you have less than a week of market data to make a decision, you're probably not making the right decision. Or you should have a very low confidence in whatever decision making and it's more or less coin toss of whether or not you're doing the right thing. |
| # / 00:57:18 | Vishesh Choudhry | That seems pretty reasonable. But you look at these changes, like they're kind of clustered, right? And back at this point, you consider February to end of March, these were spaced out a little bit more. And I think these changes: there was a very high degree of confidence that they were in the right direction. Now it's a lot easier when we're talking about one to seven and a half percent. But then there's this sort of odd spacing between seven and a half to 1.5. And then from 11, to 19.5 was in rapid succession. And then maybe less than a month later? Yeah, less than a month later, than it was unwound and the fee was dropped, again- |
| # / 00:58:02 | Richard Brown | We can interpret it a different way too because we can't necessarily attribute this to rational activity as well, though. It could be that people are just simply chasing a penny above or below and hoping that what they just did had and affect no. Right? |
| # / 00:58:20 | Vishesh Choudhry | Well isn't that what's happening right now? |
| # / 00:58:23 | Richard Brown | It doesn't matter. It's open for discussion I guess. |
| # / 00:58:27 | Vishesh Choudhry | It does seem right. So there was a appetite for a drop for a pretty long time. And then that drop was recently made. I mean, I don't know if I have a longer time scale on this graph, I don't. But this was sitting at 20 and a half percent for quite a while. And then the 18.5% drop was made. And now we're already talking about another drop. But then why was it sitting at 20.5% for so long as another question of- this is not a very responsive system. And I know, Cyrus doesn't want this to be like a day-to-day change kind of thing, or truly dynamic, really algorithmic. And I think that makes through a lot of reasons that make a lot of sense for not having that kind of system. |
| # / 00:59:13 | Vishesh Choudhry | But maybe it is a bit on the other extreme of kind of choppy, where it's not about it being too slow or too fast, I think it's maybe about the lack of predictability, the lack of seemingly logic behind when the changes are made. |
| # / 00:59:31 | Vishesh Choudhry | And then magnitude is something we can talk about. But again, when you're going to limit to really small changes, or really big changes, or something in the middle of being 2%, I think more often than not people are going to end up choosing that middle ground. And I think 2% is what they've kind of landed on. If, okay, well, we're not going to sit here and monitor the system all the time, we're not going to sit here and make decisions about Maker all the time, Speaking for Maker holders. And so I think they've decided, when we do decide to make change, We're going to make a 2% change and try to make sure it's in the right direction. And that seems to me like a balancing of the cost of monitoring and management and trying to make actual changes and have the system be responsive. |
| # / 01:00:17 | Richard Brown | So is it too early to make some assumptions about why these things happen? So obviously, stability of the peg is priority number one. But it feels to me like there'd be different levels of reluctance when it comes to raising or lowering. Have you seen any evidence that the community is happier to raise it than it is to lower it or lower than it is to raise it? |
| # / 01:00:43 | Vishesh Choudhry | It's almost like there's two communities, I don't know is maybe a more cogent explanation. But it seems to be that there's like different voices that take over at different times, is kind of how it feels. And I think there is a set of people that are more likely to advocate loudly for decreases. And it seems like there's a set of people that speak up more when when we're talking about increases. And it seems at the moment the, I don't know, somebody was saying doves in the in the chat. But whoever this group of people is that tend to advocate for these drops is speaking up at the moment. |
| # / 01:01:31 | Richard Brown | I'm going to ask you what we should do. And I know how much you love that. Should the community be trying to err on the side of lowering? Or should they be erring on the side of should be focused more on stability, regardless of whether the fee is high or low? |
| # / 01:01:52 | Vishesh Choudhry | I mean, again, just basic best practices, process management, it actually doesn't really matter as much what direction you're talking about as how you're going about doing it. So if we're talking about allowing some spacing and making smaller changes, then no matter what change you make, there's a reduced likelihood that you're managing the system incorrectly, right? so you're reducing your risk of how you're managing the system. And that's just kind of like, again, basic process management best practices. So that I think should be the top priority, but then we're talking about, okay, when the system is able to withstand a drop the stability fee again, maybe give it a couple more days to confirm that there aren't further economic fallouts or that there wasn't some external action going on in the market that is covering up the actual effect of the drop in the stability fee, which is again, totally possible to allow sort of that dust to settle, I think gives you a significantly clearer eyes and a increased degree of confidence that you're making the right change. |
| # / 01:03:10 | Vishesh Choudhry | But assuming that that all happens over the next, say, three or four days, then maybe in the following week poll you do go for- and again, I know some of these options are now getting restricted, but in an ideal world, maybe half a percent or percent change. |
| # / 01:03:30 | Vishesh Choudhry | So, now I do think dropping by 2% is going to be potentially premature. It's not to say it won't work out. But it's just in terms of what options have what degree of confidence, what outcomes. Again, 2% change, I don't see data supporting anything saying that we know that's gonna work out all right. But I think we can be far more confident that if there was another half percent drop, it's gonna work out all right. So that's kind of what I'm going with here is: what outcomes do you have what degree of confidence in? And I think the greater the change the smaller degree of confidence you have that it's going to work out, right? You're not gonna have to unwind it and deal with- |
| # / 01:04:18 | Richard Brown | Degree of change and confidence in that change is important. But it's also the impetus behind the people that are changing things. That's what I'm trying to wrap my head around, because we fall into this hawks and doves category, which is an easy bucket to separate people into. I wonder whether those buckets are in fact, stability seekers versus people that are concerned about the competitiveness of the product, of Maker. |
| # / 01:04:42 | Vishesh Choudhry | I think it's both. I think it's holistic, right? So my guest system motivations, which is, again, not something I'm really going to do to sit here and say why people are doing things, and I don't think anybody really reasonably could, but it does appear to be what's being discussed, at least on the chats is a combination of DAI price peg stability, but then also competitiveness of Maker and attractiveness of rates. |
| # / 01:05:12 | Vishesh Choudhry | And so there's certainly, logically has to be an element of, of the quote unquote, doves seeking competitiveness of Maker rates, and saying, hey, in the event, and we hear this, even even on these calls, like, in the event that we can drop the stability fee, we should. Is, I think, that camp mentality. And so once there's blood in the water, maybe or, once the first drop has been made, and everything seems to be all right, then it's no surprise that there is an appetite for further drops. Because then it's okay, well, we can achieve the outcome, and we being these doves, can achieve the outcome that is desired, and the system will allow it, Maker voters will allow it, and the peg will sustain it. |
| # / 01:06:03 | Vishesh Choudhry | And so it's maybe a little bit of give an inch take a mile as well. In the event that the first change was made, and the system didn't blow up, people decide to do it again, makes sense. And the same thing happens with stability fee increases when we're in a very bullish market. And people are saying, hey, there's all this leverage, the stability fee was increased. There wasn't a significant jump and DAI price. That means that there's room to increase it again. And it's like the same mentality on both sides. Just depends on whether we're talking about a variable market and different voices take over with kind of actually really the same argument. |
| # / 01:06:44 | Matthew Rabinowitz | I mean, it's almost like there's a monetary inertia, just to come up with a phrase where you make a change, it doesn't really look like anything's happening. You're steering the ship, and you keep seeing that you're headed in one direction to arc the wheel all the way over because it becomes critical. But by the time you start to unwind it, and you start heading the direction you want it to go, the inertia is still trending the wrong direction. So doing smaller moves with a higher frequency, not frequency in terms of every day, but every other week, or whatever the term, whatever the frequency comes down to be. You're saying without a higher probability of minimizing that inertia and getting into the right place? |
| # / 01:07:20 | Vishesh Choudhry | I think there would be pretty definitive, reduced process management risk if you made a half a percent change every two weeks. And then you just decided what direction that change needed to be, or you agreed that there should be zero change, because there's data to support that, based on what I've seen so far. |
| # / 01:07:43 | LongForWisdom | So I think the inertia point is pretty accurate. Like, I think- |
| # / 01:07:46 | Vishesh Choudhry | I agree. |
| # / 01:07:47 | LongForWisdom | Main problem is that we make a 2% change, and then nothing happens. Everyone's like, that didn't work. Let's do it again. |
| # / 01:07:53 | Vishesh Choudhry | Yeah, you everybody's looking for that line. There's no visibility on the line. So you just keep pushing until you know you pushed over the line. |
| # / 01:08:02 | Richard Brown | Well, that was that question that Matthew brought up a long time ago. It's the supply and demand buffer that we don't have quite a clear picture of what that is, and how long it takes for it to come into effect. Right? |
| # / 01:08:13 | Vishesh Choudhry | I'm curious what David saying in terms of feedback. Oh, yeah. I mean, of course- |
| # / 01:08:21 | David Utrobin | Yeah. Cause Chris was like the more money they tend to make the louder they'll get, but I don't actually think that that's the case. I think people just kind of have emotional reactions to things like your graph, right? They see the- certainly for me, I mean, I've seen for the last seven days, it's like trading above makes me wanna, |
| # / 01:08:38 | Richard Brown | It's the old zoom in, zoom out and check the x axis problem. I think that it's difficult when humans are in charge of this stuff. |
| # / 01:08:47 | Vishesh Choudhry | Well, that's why I like to show like some of the historical data instead of just the more recent stuff is just to put things back in perspective. Because in the grand scheme of things, things have been comparatively quiet. Supply has been lower, which means excess supply has been lower, which means the effects of leverage are a little bit less exaggerated on things like DAI price. |
| # / 01:09:10 | Richard Brown | That's what he meant, right? Like if you look at some trusted data as a state of the peg over the last four to six months, you can basically use it as a ruler, it's fantastically stable, and so micromanagement makes me nervous. |
| # / 01:09:27 | David Utrobin | Yeah, I guess the main Visheshism of the week is the whole process management side of things. I actually really like how you framed that. |
| # / 01:09:35 | Richard Brown | I was hoping for an ATM quote, actually. |
| # / 01:09:39 | Vishesh Choudhry | Yeah, I've started to catch on. If I quote a movie or an old TV ritual, snag it. |
| # / 01:09:47 | David Utrobin | And medics are strong. [crosstalk 01:09:50] |
| # / 01:09:56 | Richard Brown | So I'm not sure where we're at right now. I promised, threatened, questions about governance at the end of the hour. There's a lot of interesting monetary policy things happening as well. So, I think that maybe we'll just roll the dice again, and see if we can have a general Q&A session today. Historically, people have been super reluctant just to start asking questions. And so much so that I'm considering actually just paying people to start doing it. First person to ask a good question gets 5 DAI or something. |
| # / 01:10:28 | Cyrus Younessi | I want to bring up something first before we jump into the Q&A [crosstalk 01:10:34] |
| # / 01:10:37 | Cyrus Younessi | I just wanted to say our own Maker [inaudible 01:10:42] was one of the winners for building a [inaudible 01:10:49]. That's pretty cool. And I was also- |
| # / 01:10:55 | Richard Brown | Yeah, and he's on the call. [crosstalk 01:10:57] |
| # / 01:10:58 | Cyrus Younessi | Yeah, I know he's on the call It'd be cool if he could come on and talk about it one of these calls. |
| # / 01:11:00 | David Utrobin | I'm gonna- |
| # / 01:11:03 | Richard Brown | I wouldn't mind hearing like a brief recap right now, actually, Andy do you have access to a mic? Can you give us like a two or three minute recap of what you built? Because the- [crosstalk 01:11:13] Jump on a camera right now and present. |
| # / 01:11:20 | Richard Brown | Maybe not. All right. But yeah, it'd be really interesting because rate swapping is a conversation we've been having the community for a long time. And presumably, it will unlock a lot of institutional, potentially non retail interactivity in the ecosystem. Let's definitely try and get him to do a presentation part. |
| # / 01:11:43 | Richard Brown | That's actually an interesting idea, Cyrus, like we could probably spice this thing up a bit if we started getting some interesting DeFi players to come in and explain how they're going to fix the space, some of their new tooling. |
| # / 01:11:58 | Cyrus Younessi | Yeah. |
| # / 01:12:01 | David Utrobin | I'm down to have people do this on the community call, because I feel like we have bit more bandwidth there than here. Here, we're usually struggling for bandwidth because we have so much to cover. |
| # / 01:12:13 | Richard Brown | Yeah, well that's up for debate. I like to talk about DeFi stuff here, and community stuff there. But yeah, we'll figure it out. Does anybody want to earn five DAI and ask a very piercing question about governance? |
| # / 01:12:29 | David Utrobin | I could do it for free. |
| # / 01:12:30 | David Utrobin | So there is a thread on parallelism that I think LongForWisdom posted, or one of you guys. |
| # / 01:12:36 | LongForWisdom | That was me, yeah. |
| # / 01:12:36 | David Utrobin | So it was, am I right in understanding that, cause I only remember, like, recently reading it. Am I right in understanding that because now we can have concurrent governance polls, why wait too long? Why don't we just throw out these things that we're polling for on the forum immediately to governance polls? |
| # / 01:13:00 | Richard Brown | I can easily answer that question, because it just started and we're too busy. But the idea is that, yes, the floodgates have been opened. And we are now prepared to do all of the things that we wanted to do in the last six months. That's why I brought it up at the top of the hour, specifically calling out the rate stepping and the cadence thing. We've had these things percolating. Now we sort of have this, at least a prototypical government's facilitator role here. And so maybe we can assume that that's a thing. And I can take the liberties of starting that stuff up, and the expectation that they'll get ratified, or we can just wait a week. |
| # / 01:13:42 | Richard Brown | But but what I would like to do, and one of reasons why I wanted to talk about in this call is that, I want to sort of formally bring those conversations back to the fore, and then apply some of the sophistication that we've created over the last few months. So let's look at LongForWisdom signaling guidelines doc. Let's review the rate stepping and stability fee cadence posts that I put in there at the christening of that forum, I think. And let's formalize them and sort of start them again. And that's I think the only kind of rate limiter we have right now, with the polling system, is going to be cognitive overhead. At some point, we're going to determine what the level of too many things happening and governance is. We might actually be testing that soon enough. But in the interim, we should probably get those things underway. |
| # / 01:14:36 | Richard Brown | I know that there's still some discussions to be had some clarity around what the exponential rate stepping looks like, and why we're doing it and which parameters that would apply to. But these are conversations we should resurrect and polish off and then see if we can get some momentum there. |
| # / 01:14:55 | Matthew Rabinowitz | I would actually like to just, just for the commentary in general. For when we deal with things that are related to monetary policy and how we have system-wide monetary inertia of moving money back and forth, the pragmatic approach of very conservative small movements, like we talked about probably a very wise course of action.When it comes to just general governance as a whole. I mean, I just, I'm a absolute advocate in violent execution, and not waiting for it to be perfect. If we end up with perfect consensus, and there really isn't a reason to have a vote in the first place. So if we get to a point, and we just have lots of holes, then we'll get to a point where we hit that limit? Fine. But until that point, just violently execute it even if it's not perfect. |
| # / 01:15:39 | Richard Brown | Yeah, well, I think that I agree completely with that model. I think that like, obviously, there's some grounds and, probably will drop the violent part of that when we publish the ideas in the forum, but aggressively moving forward with things. But let's iterate. Let's gather all the available evidence, let's have a reasonable amount of well-considered conversations, and then let's do things. And as long as we clearly understand the risks and benefits associated with whatever activity we're going to take, keep this show on the road, and iterate rapidly as we go along. I think I'm obviously not advocating that, let's all switch to liquid democracy tomorrow and see what happens to the MakerDAO ecosystem, that would probably be an ill-considered idea. But if it's, let's figure out how to get forum or governance polls happening quicker than they are already, that's a great experiment. Let's start doing it. |
| # / 01:16:37 | LongForWisdom | Yeah. So I agree that we should try and make things as [inaudible 01:16:43]. But there's one specific situation that I really want to avoid, which kind of plays into what I'm suggesting that we do. And that is that we have like, a majority for something in the form. [inaudible 01:16:55] |
| # / 01:16:57 | Richard Brown | Sorry, you're super quiet LongForWisdom. You want to do what? |
| # / 01:17:01 | LongForWisdom | Is that better? No. |
| # / 01:17:08 | Richard Brown | I think so. If you've stopped talking, then no, but- |
| # / 01:17:11 | LongForWisdom | I said a few more things. Okay, so I was saying that there's a specific situation I want to avoid, which is that we get 60% or 70%, or some majority in the forum for something. And then we move on chain thinking that that's great. And then we have one whale that comes in and the 30% that like blocks the whole thing? |
| # / 01:17:32 | Richard Brown | Yeah, and that's- |
| # / 01:17:33 | LongForWisdom | Well, yeah, sorry. While that's a valid thing to happen, I think that's going to be pretty distressing for many people that voted in the forum polls. |
| # / 01:17:43 | Richard Brown | I agree. And I think that that this is something we need to communicate over and over and over again, that when we do these forum polls, what we're doing is sort of, well, I hesitate to say democracy, but it's people voting with one person, one vote, presumably. There's no civil protection there. So we can only guess. But it's a signal that gives us enough confidence to move something into a different type of voting system. And that's basically one vote per Maker, which is very different than the type of voting that we have in the forum. |
| # / 01:18:17 | Richard Brown | So we need to understand that when we gather signals in the forums, we're testing the temperature of the room, basically. The very big room. But ultimately, it is up to those whales. That's the system that we built. And we need to accept the fact that this is not a bad thing, this is exactly what we wanted to have happen in the first place. And if we do think it's a bad thing, then we have to change it all. And we'd start talking about that. I have no idea what that would look like. But we could figure that out. If it turns out that 100% of people in the forums want to do one thing, and two whales don't, and it doesn't happen, then that's the world we live in. |
| # / 01:18:53 | Richard Brown | We need to solve that problem. And we can either revamp all of governance, or we could start exploring other options like the delegation, self-organization, how do people start collaborating? How do they campaign? This is something that I've been talking to you about to more and more, is that we need to transition away from people saying that "I think we should do some X" in a tiny little forum post, and move towards, okay, if the community, if the people in the forums want to do something, then they need to consider tweeting about it. And/or talking to other people about it, and/or creating some kind of compelling blog post about why something should happen, and then trying to convince the whales what they should be doing. And if that doesn't work, then they should organize, and they should make their own whale and move the system in the ways that they want to move it. |
| # / 01:19:56 | Godsflaw | I have a question. Is there any appetite in me starting to mock up the delegation stuff? |
| # / 01:20:04 | Richard Brown | Possibly. So here's the problem with this. We put out a help wanted post in the forums. Lots of people have already just chatted here and there about the need and/or desire to have a delegation system. A few handful of people have expressed interest in doing it. And only one team has shown any sort of long term commitment to that idea. So it'd be my preference that the people that are actually likely to pull this thing off would get into a chat with each other and try and pull resources and get on the same page. |
| # / 01:20:42 | Richard Brown | So Chris Hopcroft, I don't know if you know him, or of him. But I can put you in contact with him. And I'm reasonably sure he'd be enormously happy to have the additional help. Or at least you guys can coordinate. |
| # / 01:20:56 | Godsflaw | Right. He's on working on the Aragon proposal, right? |
| # / 01:21:00 | Richard Brown | Yeah. |
| # / 01:21:00 | Godsflaw | Okay. |
| # / 01:21:02 | Richard Brown | You guys can at least spec it out and/or figure out whether your visions align. And if they don't align, It's a big community. So you guys can compete and get two different options going. There's all kinds of stuff that can happen here. Who knows. maybe Aragon is not the greatest solution. I think it feels to me intuitively, like, all the frameworks are there, and we don't have to reinvent the wheel. |
| # / 01:21:21 | Godsflaw | Yeah, and it comes with a UI, which is nice. |
| # / 01:21:24 | Richard Brown | But who knows, it might be a blocker, or maybe somebody can just put together a better solution too. So, some competing solutions would be great. If you have something in mind, Godsflaw. Ping me, and I'll put you in contact with either Chris or we can try and get something else on the go. And I think the UI stuff is going to be important, it's the UX stuff is going to be even bigger, and then figuring out the complexities around custody, I think. For this to work it needs to it this kind of a process is ripe for exploitation. So somebody can come along and have all the best intentions. But once they have 50,000 Maker under their belts, they might realize some of the darker natures have come to the fore, or somebody could just be a huge scam artist. We need to make sure that people that are staking their Maker and allocating voting rights to somebody else or doing it in a way that's safe and recoverable and all the rest. So there's some complexities that need to be figured out there. |
| # / 01:22:26 | Richard Brown | Was there some questions here? 75%, what's going on? |
| # / 01:22:37 | LongForWisdom | I was talking about the thing I said. |
| # / 01:22:43 | Cyrus Younessi | I had a question about the polls for ratifying amendments, and if you have any- |
| # / 01:22:53 | Richard Brown | Yeah, can you talk louder than the train dude? I can't hear you very well. |
| # / 01:23:01 | Cyrus Younessi | Last week, we talked about how we'd have some arbitrary discussion time followed by the polls, but I don't think we talked about, how long the poll might be easily last like. Have we considered doing like a three-week poll or four-week poll? Just letting that run in the background during the discussion itself? So maybe starting- |
| # / 01:23:21 | Richard Brown | For the mandates, or for something else? |
| # / 01:23:22 | Cyrus Younessi | For the mandates. |
| # / 01:23:23 | Richard Brown | Yeah, And with this, I think this is where this suffers kicking the can. And so all of these dates are arbitrary. And I think that that there is a slight push-back in that it felt like we were imposing restrictive timelines on people, which I can assure you was not the case. But people can find conspirators wherever they look, I guess. |
| # / 01:23:41 | Richard Brown | But we have no idea what the community's appetite is for any of these things. And so we need to do it a couple times to figure that thing out. And so the rough guess was two weeks to talk about a mandate, and then one week to poll for it in the forums, and then one week to do it in the governance portal. Maybe that's sounds good enough. We'll see how it goes. |
| # / 01:24:02 | Richard Brown | But I think we're going to have to iterate. And that's one of the metaphysical things I wanted to talk about in this call, too, but we're running out of time, is how much debate is enough? And at what point do we determine that consensus has been reached? Because reading through those mandate threads, you kind of have to do a sentiment analysis with like, was there any major flaws determined in this mandate? Are people generally accepting of these things? And the problem is, there's no algorithms for that. So somebody has to make a decision. And making that decision is risky, because human frailties come into play. But the short answer is, I have no idea. I think the community needs to let us know like, hey, next time around, we need to be able to talk for a month in the forums, and we need to have three weeks. We need to figure it out. I have no answers. |
| # / 01:24:55 | Richard Brown | Was that LongForWisdom? Were you going to say something? |
| # / 01:24:58 | LongForWisdom | Yeah, just I think the polls can help mitigate the need for that sentiment analysis you're talking about, right? |
| # / 01:25:06 | Richard Brown | Well, yeah. So here's my issue is that I'm very, very hesitant. Very hesitant to put out a mandate and then attach a poll to it and say, like what you guys think? Because these conversations go back and forth for weeks at a time and very compelling arguments happen. And then maybe somebody comes in and says, okay, right, let's change half that mandate. And then at that point, we have a poll, and it's super stale. That's based on the people that voted initially, were basing their votes on old and outdated information. We have to guess, did anybody come back and update their votes? I have no idea. |
| # / 01:25:41 | Richard Brown | So what I'd really like to see is that we have, proposal goes in the forum, we have a debate that lasts for X amount of time. And then once the debate dies down, or some kind of consensus has been reached, or whoever submitted that proposal is now unwilling to make any further changes, they just want to get a message a signal out of it, then that poll gets tacked onto the thread. |
| # / 01:26:03 | Richard Brown | And then we can set some kind of arbitrary timelines for how long that poll lasts. I'm a much bigger fan of that idea, because at least we can have this assurance that everybody is voting on the same picture, as opposed to various versions of that picture that they might stumble on. |
| # / 01:26:22 | LongForWisdom | There are other ways of solving that problem, though, you can create a poll when it goes out the first time. And then as soon as you make any significant changes, you can close the poll and start a new one. There are other solutions to that. Other than waiting for the discussion to finish before starting a poll. |
| # / 01:26:39 | Richard Brown | I guess. So say that we had a proposal, a poll, and then three other versions of that poll as we go along? Until? I don't know. So what would we do with the iterations of those opinions over time? Would we discard the first ones? Or would the first ones color the second ones? Or can we expect the people that come back over and over again to reevaluate? |
| # / 01:27:06 | LongForWisdom | You have to hope that the latest one is the most representative, right? Like, it also lets you see that if you make a change, and you end up with less consensus than you did before, you can then see that and then go back to a previous version. |
| # / 01:27:17 | Richard Brown | That's interesting. |
| # / 01:27:21 | Richard Brown | So yeah, okay. Well, yeah, let's continue to talk about and see what works. |
| # / 01:27:27 | Richard Brown | I keep on coming back to not turning makerDAO into work for people and trying to present as much data with this little overhead and time requirements as we possibly can. So if we're talking about the rates, the exponential rates and everything, you talked about that before, where we're constantly hovering back and forth between 60% approval of one of these slight variations and the rates. |
| # / 01:27:57 | Richard Brown | And my concern is that people just get sick of this stuff after a while. And we can't expect 50, 100 or 1,000 people to be watching this forum thread to keep up to date with the last quote, storm, that reply storm that occurred. We need to figure out how to summarize and say, okay, hey, we've all talked this thing to death. This is the final version of the thing that we want to get a signal on, here's the thing where you signal, and then we can all have some confidence that that maximized the interaction, and has the cleanest signal to noise ratio. |
| # / 01:28:37 | Richard Brown | But like you said, there's arguments to be made that we can track sentiment over time, or we can, I don't know, it's going to be complicated. I think that we should figure out a plan, play with it and see what works and what doesn't. And then next time around iterate. I think the good place to start with that would be the signaling guidelines doc, which is why I posted that earlier. I'd love to people to dig into that and get internalized, offer some suggestions, and then we can put it into the portal as being like this canonical talk about how you do signaling in the forums. |
| # / 01:29:13 | LongForWisdom | Yeah, I agree it needs more [inaudible 01:29:15] really. It's mostly a summary of things people said, like a month ago. |
| # / 01:29:22 | Richard Brown | Yeah, it's sort of a testament, well at least anecdotal evidence about turning MakerDAO into work, where last night, it was like 11 at night, and I was thinking, okay, I should go into that guidelines thing and add all the arguments for and against the stuff that I just said, like, let's do debates. Or let's do, proposal, let's do debate, let's do a poll. And then let's do formal governance poll. And then I was like "aw man this is too much work, I have to go to bed." So I didn't do it. So we have to ask ourselves, people that aren't drawing a salary from the foundation, how much work are they going to do for us in these things as well. So we need to try and make things as easy as possible without making them too dumb again. |
| # / 01:30:07 | LongForWisdom | Yep. Agreed. |
| # / 01:30:10 | Richard Brown | Alright, so we're at 30 minutes past the end of the hour, yeah, it is my personal problem. And they're all now everybody else's problems to immortalized forever on YouTube, and in the transcripts. |
| # / 01:30:22 | Richard Brown | Yeah, it's half past the hour. So let's put it to bed. Thanks, everybody for the lively discussion. Thanks to Cyrus and to me, and to Matthew and to the Vishesh for the interesting discussions. Please, if you have questions that weren't addressed, or if you have other things you want to talk about, join the forum. There's almost too many really interesting discussions happening right now. Get informed. Vote on the risk-team mandate and the governance facilitator mandates. There going to be up for another four and a half days. And then we'll move it into the governance portal and see what people say there. All right, thanks, everybody. |
| # / 01:31:08 | David Utrobin | Thank you, everybody. |