Governance and Risk Meeting: Ep. 52 (September 12 - 2019)¶
Video | Audio | Discussion
| References | Person | Text |
|---|---|---|
| # / 00:00:03 | Richard Brown | Hello, everyone. Welcome to the September 12th edition of the Scientific Governance and Risk Meeting. My name is Richard Brown. I am the head of Community Development at MakerDAO. Today we are going to talk about governance. We are going to talk about risk, and we are going to continue what I believe is going to be a new trend, talking about Oracles. As the governance ecosystem becomes more and more complicated, or sophisticated maybe I should say, we are going to be bringing new actors into the system that are empowered by the community to do things on behalf of the community. Hopefully in some near term but glorious future, we will begin to engage with community members who are acting on behalf of the community as well post MCD. |
| # / 00:00:47 | Richard Brown | But, in order to get there, we have a lot of work to do, and that's the work we're doing in these calls. So, today we have governance. We have oracles, and then we have risk. And, the governance... I'm sorry. I forgot my preambles again. I am going to continue to beg and grovel. If people have questions, please feel free to ask those questions at any moment, so type them in the chat if you don't have access to a microphone. If you do have access to a microphone, please interrupt us. We love it when that happens. And, just ask away. We will deal with questions as they arise. |
| # / 00:01:22 | Richard Brown | If you want to formulate your questions and ask later, at the top of the hour... at the end of the hour, sorry, we'll have just a general Q & A session, so we'll stick around for 30 minutes and talk about anything that people might have on their minds. I think that's... That's enough for the preambles today. I'm going to try and keep this as tight as I possibly can. |
| # / 00:01:42 | Richard Brown | So, in the governance world, we have a couple of interesting things happened, good news and more good news. The good news being that the community shows overwhelmingly not to fire Cyrus and Rich from the governance ecosystem. That was nice to hear, so we still have public jobs. The governance facilitator role was ratified with a 99.9% approval rating. I'm still daily appreciating whoever it was that bought $20 in MKR to vote no against me. I love you. I want to send you a T-shirt if you're willing to send me your address. |
| # / 00:02:20 | Richard Brown | Cyrus, on the other hand, was 100% approval, because I guess he's like .5% more approachable than I am, I suppose. The good news there, though, is that we can begin to start doing some work in earnest. I believe it was Cyrus that voted no. I actually planned to buy $21 in MKR and vote no against Cyrus just to make my stats look better, but I forgot. |
| # / 00:02:46 | Richard Brown | So, yeah, we have a lot of work to do, and so we are going to start doing that work. And, part of that work is evidenced by the activity in the forums recently. I know that LongFor... I'm not going to steal his thunder. LongForWisdom can tell us a bit about some of that, but we've had these slow burn issues for awhile now around cadence of the votes of the stability fees, questions around do we really need to be doing this every single week? Community has gotten together and debated that issue, come to some conclusions, and we'll be hearing about that soon. The same process has occurred around the rate stepping, so when we have these stability fee votes, we have semi... Well, basically... I'll come out and say it. Arbitrary number of options in these votes. |
| # / 00:03:32 | Richard Brown | We chose, previously, just to pick four above and four below the current stability fee and that's what people can vote on. Why? It sounds good. It turns out that there's a bit more sophisticated and elegant ways of doing that, and the community has been talking about exponential rate stepping. And, so we'll dig into that a bit. |
| # / 00:03:51 | Richard Brown | Other news on the governance front, particularly around stability fees, is that interestingly enough... And, this is something I find particularly heartening is that the community... Heartening, not because I fully endorse it, because don't think that this is a beneficial activity. So, I need to walk this fine line. The community keeps on lowering the stability fee, which is good. That's a very good thing, because it shows the ecosystem that the MKR holders, the voters, are not self-interested to the point of self-destruction. They're not going to raise the stability fee endlessly in some hopes of burning all the MKR in existence. They will actually lower the stability fee when it's appropriate. This shows that the community and the MKR holders hold the stability of DAI paramount over everything else, and that's precisely what this ecosystem... or what this governance ecosystem is designed ultimately to do. |
| # / 00:04:49 | Richard Brown | We need to support the peg, so we can support the protocol, so the protocol can support DeFi, and DeFi can take over the world. And, that's precisely what we're all about. So, lots of good things happening there. The most recent stability fee poll advocates for... I just looked, and I forgot already... 12.5. Ninety-one percent of voters came out for... of the Maker staked, came out for 12.5 percent. So, that's going to be interesting whether that gets ratified. |
| # / 00:05:22 | Richard Brown | I think I might just wrap it up there. Governance is going to be... There's going to be a lot more to talk about in governance over the next few months, but let's not drown out the interesting things that are happening the Oracles world, which we'll get to in a second. Before we do that, though, I want to hand it over to you LongForWisdom, if you could give us a recap of the cool things that are happening in the forums. |
| # / 00:05:47 | LongForWisdom | Yeah. Yeah, will do. |
| # / 00:05:49 | Richard Brown | Thanks. |
| # / 00:05:50 | LongForWisdom | Cool. So, I'll start with sort of the active discussions we've had this week that are sort of spinning up. So, I'll just... Sorry, it's not a video. I just recently posted a blog post on the DSR, which, I think, from the people who've read it, seemed to find it pretty good. So, that discussion's been kicking off over that, which is good. [inaudible 00:06:14] |
| # / 00:06:14 | LongForWisdom | So, I made a thread specifically about the current signaling process, as I currently... as I see what it is currently, I guess, sort of a description of what I view it to be. So, there's been some discussion around that as well. This is fairly important, because it sort of interacts with how the signal request have been kind of pushing or going, so if you think those are going wrong in some way, then let me know in that foot. |
| # / 00:06:43 | Richard Brown | I have it in these recaps you sort of give us just a list of the interesting things people should look at, but this is a good one. Actually, can you give us like a one minute or two minute recap of what's going on in that currently signaling process there? Those things that you're advocating? |
| # / 00:06:55 | LongForWisdom | Yeah. Yeah, I can do. Yeah, so I'm essentially, so just that we split it up into two stages. So, I set three stages. I think it's good to start with a sort of more general sort of question of whether there is appetite for change in some form, not specifically but generally around an issue. So, that led to a sort of early exit if everyone is kind of fine with something, like if someone's just picking out some personal bugaboo with the system, then everyone can sort of say it's fine and it can end early. So, while we do that staging, we can also discuss more specific things as we're signaling for that general change. So, sort of do some parallelism in there to try and increase efficiency. |
| # / 00:07:39 | LongForWisdom | Stage two, we... So, if there is appetite for change, then we move to stage two. We create a multi-choice poll that lets everyone signal "yes" or "no" to any of the suggested specific options that have been discussed or proposed by members of the community. I suggest that'll last for two weeks, because I think you need two weeks to give everyone enough time to vote and look at and vote on things and talk about things. |
| # / 00:08:08 | LongForWisdom | And, on stage three, if we've got... if one of those specific proposals has a majority backing, so if over half of the people think that one of the specific proposals is good, then we can move it to stage three, which is moving the poll onchain and trying to get it ratified by the MKR token holders. |
| # / 00:08:29 | Richard Brown | Cool. So, I encourage anybody who's interested in getting involved with governance to check this out, because this is an example of emergent process that's forming in the community, which is going to be big for us. We have... There's two ways we could... Well, let's divide the world up into arbitrary pieces, so let's look at things from a binary perspective. There's two ways we can do this. We can either come up with a 600 page manual about how governance is going to work and attempt to predict how the future will shake out. I think that might be fraught with peril. The other option on the other end of the spectrum is let's just not figure out anything, just see what happens as we go. And, then there's a happy middle ground between those two ends, where we make... we set targets for ourselves for the future. We know what we hope governance will be, and then we plan out how we're going to get there in increments, reasonable increments. |
| # / 00:09:25 | Richard Brown | It's always tough to predict how the future is going to look at all, ever, but more specifically with crypto. This is a very accelerated, rapidly evolving ecosystem, so the things that are happening in the forum right now are examples of how people choose to self-organize in the framework that we have to work with here. And, it's important that we remain cognizant of that evolution and that people step up and speak and have their voices heard. It's not... but, also, these are all living documents, and what also gives us enough philosophy to implement things and to do what needs to be done in order to make this protocol a success, but it also gives us enough agility to course-correct. I'm supper excited about how well things like this are turning out and the efforts of people like LongForWisdom to formalize some of the things we've been thinking about, so please have a look at this thread. If you have any suggestions about how it can be improved or any comments, please add them. |
| # / 00:10:28 | Richard Brown | LongForWisdom, one of the things I wanted to ask you about, though, and I have to admit that I haven't had a chance to pull this apart yet, but I've been increasingly interested in scheduling, which sounds like the most boring thing a person could possibly be interested in. But, there's some interesting questions to be asked right here. |
| # / 00:10:45 | Richard Brown | So, we have... somebody comes into the forum. They make a proposal about something or other. They ask an important question. How long do we talk about that question and how do we determine that that question has been answered? And, how long do we wait for a poll vote to happen, and then how long do we wait for a governance poll to occur? These are all things that I'm really having trouble wrapping my head around, because we can't predict their level of interest for this specific issue we might have. We can't predict the level of debate that it requires. We have to sort of... we have to fit these things into larger calendars of events. We need to worry about overwhelming the community with too many activities at once. Is that something that you've been considering in these process talks? |
| # / 00:11:35 | LongForWisdom | Yeah, for sure. So, just to go back a second for one of the things you just said, just about the whole figuring out everything ahead, or figuring out as you sort of go, just a sort of... I want to say, like, this sort of description I did is descriptive rather than prescriptive, right? It's literally just my view of how I'm planning to run the next couple of [inaudible 00:11:54] requests, right? So, anyone can come along and do something different or say something different or suggest something different. |
| # / 00:12:01 | LongForWisdom | In regards to scheduling, they say in the second stage, stage two, where everyone sort of discusses specific proposals. There's kind of a requirement in that step that one of those specific proposals gets a majority vote with a reasonable number of voters. So, my suggestion is if that doesn't happen, then you essentially repeat that [inaudible 00:12:23] stage, and you'd pay for another [inaudible 00:12:25], and you'd find the proposal [inaudible 00:12:27]. And, then you continue with that until, if after three iterations you've not reached any sort of consensus, then you sort of say, right, well, we can't agree on this. We'll drop it for X time and then come back to it in the future possibly. |
| # / 00:12:44 | Richard Brown | Yes, that dropping it and coming back thing is interesting to me as well. Do we need to... I see that there's value in creating a sense of urgency, so like a community needs to learn that if you don't check in at least once every two weeks, you might miss something, which I think might be a reasonable expectation for engaged stakeholders. What happens if we wanted to have a debate that's just let's just think about this, and if people care, we'll do it at some point. Like, could we have these open ended discussions that just last for months potentially, or do we have to set these like two weeks to do this, two weeks to do that? |
| # / 00:13:21 | LongForWisdom | I feel that we kind of already have that in terms of just general threads that discuss things, like you don't have to make a signal request throughout... The signal request is more sort of when something's been discussed and sort of before there's some consensus that something needs to be done, I guess. Then you set that up and try and go through the process. |
| # / 00:13:38 | Richard Brown | Are you remaining cognizant about cognitive overhead, I suppose, or the amount of work that we might be imposing on people? Do you feel that... Have you seen any evidence that some of this stuff might becoming too onerous or it's... we're in the clear there? |
| # / 00:13:56 | LongForWisdom | It's hard to judge specific evidence. I mean, no one's been sort of coming out and saying like this is way, way too much all the time. I mean, there's a couple people saying that, but no one that's engaged has really been saying that. Yeah, it's really hard, because the more time everyone can kind of commit to it, the better decisions we can make. But, obviously, that's the trade-off, right? |
| # / 00:14:17 | Richard Brown | Yeah, I have the same challenge that David has that a week can slip by very quickly and very easily, and thinking, oh, man, what have I missed? And, then having to spend an hour or two trying to catch up on the forum seems be tough. But, anyways, I think that these are obviously challenges we just need to address all the time. The fact that we have so much activity that it takes a while to catch up is probably the best problem we can conceivably have. |
| # / 00:14:46 | David Utrobin | And, so the community is credit rich, actually. Every time I catch up, I'm like really, really pleasantly surprised by the quality and the depth of the conversation that's happening. So, yeah. |
| # / 00:14:58 | Richard Brown | Yeah. Me, too. I could not be happier, and if you knew me at all, you'd know what a strange thing that is for me to say. Governance from a super high level is both a management of human risk, and we haven't bumped out into that risk yet, which I'm super happy about. There's a lot of smart people doing a lot of smart things in a lot of smart ways, so I'm very excited about how this thing is shaping up. |
| # / 00:15:26 | Richard Brown | All right. Sorry for derailing you, LongForWisdom. Were there other threads you wanted to talk about? |
| # / 00:15:30 | LongForWisdom | Yeah, there's a couple more, but that's okay. Don't worry. So, we've also seen the first collateral onboarding request, which was from the team behind TUSD. So, there's been some discussion around that, like pros and cons. Generally, I think people are a little bit skeptical of these sorts of tokens, so we'll see sort of how that pans out. Maybe we'll get some more information on that thread from them. |
| # / 00:16:03 | LongForWisdom | Let's see. What else? Oh, yeah. So, the actual signal request threads as well. So, just sort of, Rich, whatever you said, the exponential rate stepping threads. I think it's probably at a point where we want to move onchain. We kind of... A reasonable number of people seem to have voted that they want to move onchain in a specific form, so I guess that's sort of looking at you, I guess, to ask sort of how that proceeds from this point, I guess. |
| # / 00:16:32 | Richard Brown | Yeah, that's an interesting question. The plan is that once the community signals its support for something, then the wheels of bureaucracy begin to spin and then it gets scheduled for moving into the portal itself. So, one of the things I... One of the things I think I cautioned in a thread somewhere, but possibly not enough or in the right spot, is there is overhead. There is administrative overhead that is moving these things out of the forum and into the portal. |
| # / 00:17:03 | Richard Brown | And, it would be nice to add some predictability to this thing, so my hope is that the community would be fine with, at some point, consensus has been established, there's some kind of ceremony or something where the poll is closed, the thread is ended, and then at that point, a timer starts or something. And, then the next Monday of next business week, then it goes into the forum. I mean, into the governance portal. I think that would be my plan. |
| # / 00:17:35 | Richard Brown | Have you guys seen anything in the forums that'd lead you to believe that people are finding that objectionable? |
| # / 00:17:41 | LongForWisdom | No, I think that sounds good. I would just say I think that we've kind of reached that kind of end of the thread point on the exponential rate stepping thread. |
| # / 00:17:49 | Richard Brown | Yeah, and you and I have spoken about this that you have visibility on the ground and acting sort of as advocate for the community, or at least an administrator, an ad hoc administrator. We need to just tighten up our communication a bit, so we're going to have these weekly calls of you just letting me know what's going on and reminding me what needs to be done next, which would be swell in the short term. |
| # / 00:18:09 | LongForWisdom | Yeah, sounds good. |
| # / 00:18:10 | Richard Brown | All right. Cool. Anything else of interest we need to talk about? |
| # / 00:18:16 | LongForWisdom | Yeah. So, the only other... The other one was the signal request around the cadence, like the one we've got is just a week versus potentially a longer cadence. So, it's sort of coming back to the sort of steps, stage one, general appetite for change, seems to be kind of mixed. I think it's about 50/50 in terms of whether we're happy with the one week cadence, or whether we want to change to longer. |
| # / 00:18:39 | Richard Brown | This is something else that I'm super interested in, too, is this notion that we need to have sort of... I shouldn't say overwhelming. But, we need to have a visible or a comfortable level of agreement in the forum before things can go to the portal, and I think that we need to keep in mind that we're potentially talking about two different demographics here. So, we have... In the forum, we have more of a representative democracy in the loosest, more easily sybil attacked forum imaginable, but this is just people talking about things. We can aggregate some numbers about them, or some data, the amount of people that show up and the amount of people that say yes. |
| # / 00:19:23 | Richard Brown | But, frequently... Well, I don't know. What I want the community to consider, though, is that even if there isn't overwhelming support in the forums, there still might be value in putting something in the portal for Maker holders at large to vote on, I suppose. As I started saying, this sounded like sort of... I realized it was half thought out, but it's something that I want consider. Like, is this a filter for anything to get into the governance portal? Does it have to have a certain level of support in the forums. It seems like that's the general consensus that the community's arrived at, but it's something to be considered, I suppose. |
| # / 00:20:12 | LongForWisdom | Yeah, so actually you mentioned this quotes out specifically and the exponential threads... |
| # / 00:20:18 | LongForWisdom | Yeah, yeah. So, that was the question. The last sort of part in that was about what the form the poll should take onchain, right, and whether to have it just as we vote on the consensus option, "yes" or "no", or to MKR token holders vote on the consensus option "yes" or "no"? Or, do we set up multiple polls that we let them vote on every option that was suggested rather than just like the top rated? |
| # / 00:20:40 | LongForWisdom | And, in general, the forum seems to think that it's vote on like the top rated one on the forum, but of course, then again, that's the representative democracy saying that we should listen to the representative democracy. |
| # / 00:20:53 | Richard Brown | Yeah. That's interesting. We'll figure it out. I don't think it's an existential threat at this point. Okay. Anything else on your list? |
| # / 00:21:01 | LongForWisdom | No, I think that's most of the good stuff. |
| # / 00:21:04 | Richard Brown | All right. Cool. Thanks for the recap. It's great to surface those things here. Nik, I'm going to turn this over to you now, and refresh my memory and/or clarify whether the plan has changed. Are we going to talk about sustainability in the Oracles system, or did you want to recap some of the things that happened the last week? |
| # / 00:21:22 | Niklas Kunkel | I want to go over a little bit of a recap first. So, I'm not... I think there's a little bit of... a lot of carryover between who was attending the governance call last week and this week. But, if you didn't, last week we talked about kind of the Oracle ecosystem within the Maker protocol, and kind of the different facets. And, I put forth a bunch of different proposals for the community to consider. |
| # / 00:21:59 | Niklas Kunkel | So, I'm still on track to publish a bunch of these proposals in the forum later this week, either end of today or by early tomorrow. It'll probably end up being in the governance section of the forum, and then I'll make a cross post in the new Oracles section of the forum. |
| # / 00:22:23 | Niklas Kunkel | I also wanted to give a shout out to Mariano Conti. He is the original architect of the Oracles that Maker uses in the protocol currently, and he's also designed the smart contracts for Oracle's V2. So, Mariano, I don't know if you want to say hi and introduce yourself? I wonder if everybody already knows who you are. |
| # / 00:22:54 | Mariano Conti | I'll just say a couple of words. I did the easy part, which was sitting alone with my computer, programming some of the technology that goes behind it. Now you get the fun part of actually turning this... turning it into a community and expanding it to governance, so I wish you the best of luck, and I'll be helping as well with whatever I can. But, excited that we're finally... This is what we always wanted to do, right? So, we're at the steps where the Oracles are going to become even more interesting, so thank you, Nik. |
| # / 00:23:36 | Niklas Kunkel | Awesome. And, thank you, again. So, one of the things... I don't want to go on for too long, but one of the things I wanted to touch on, I got a few questions last week from people about why are we doing this and why now? The Oracles were always meant to be handed over to governance, and it's part of this drive of gradual decentralization that the foundation has committed to that we're starting to do this now. And, in particular, the proposal that I think was the most confusing to people was the one talking about the incentives, and why do we need them? Why are we doing this? |
| # / 00:24:29 | Niklas Kunkel | So, I wanted to share my screen real quick and show you guys the same diagram I showed you last week, but to expand on it a little bit more. Okay. So, this diagram is showing the different actors in the Oracle ecosystem. So, on the bottom left here we have you guys who are the MKR token holders. On the top here, we have the feeds, and I think this maybe wasn't as clear last week about what a feed is. But, feeds are kind of humans or organizations that are running a bot, and all this bot does is it sources the prices for an asset and then it signs with your identity saying I think... you know, Feed A thinks the price of bitcoin is this. And, essentially feeds are the inputs for the Oracle smart contracts, where people will actually get the canonical Oracle price from. |
| # / 00:25:51 | Niklas Kunkel | So, if you're on this diagram, you can see the feeds are pushing the prices to the Oracle smart contracts. And, down on the bottom here you have kind of data consumers, so you can think of them as Dapps or other smart contracts. And, other smart contracts can then read the prices from the Oracle smart contracts, and they're meant to then enumerate MKR holders in some way. |
| # / 00:26:21 | Niklas Kunkel | And, so why did we set it up this way? We didn't used to regulate access on the Oracle prices to external partners. You know, why are we starting to do this now, I think, is the biggest question that we got. So, there's a couple reasons, but the main two that I would like to focus on is decentralization and sustainability. |
| # / 00:26:51 | Niklas Kunkel | Currently, the dev fund is paying feeds for the price service that they're providing. It is the intention that eventually MKR holders will take over that responsibility through the use of stability fees, because the Maker foundation may not be around forever, may not have buckets of money forever, and so if you want to make a resilient Oracle system, you need to decentralize the funding of it as well. And, the stability fee mechanism provides a nice way to do that. |
| # / 00:27:32 | Niklas Kunkel | But, if MKR holders are going to be spending their money on feeds, they're going to probably want some way to then monetize that feed infrastructure, right? If they're paying the feeds and they're administrating this entire Oracle ecosystem, it only makes sense that they have the ability to monetize it as well. And, so that's why we've added this white list to the Oracle smart contracts, and essentially, the white list is a list of smart contract addresses that governance controls. So, you guys get to decide who is allowed to read prices from the Oracles and who is not, and in doing so, you can charge a fee for that. Right? So, you can go to token company X, right, and they want to read the BTC price, and you guys can say, okay, good, well that's going to cost you 5,000 DAI annually or something like that. Or, maybe you say some amount monthly, right? |
| # / 00:28:41 | Niklas Kunkel | There is a couple things here, though. So, what this proposal does is it's not actually enforcing this particular combination of relationships between all the actors, right? So, I said earlier that the Maker development fund is paying the fees right now, so really this piece does not... this is not applicable to this proposal. Right? I've also said last week that for business development purposes, and really for just being a good actor in the ecosystem. We shouldn't be charging anything to be on the white list for the first year. We have a lot, and I mean a lot of projects using our Oracles. You know, for better or worst, we don't even know the extent of how many of them there really are, and this is just by nature of us being the first to release any Oracles and maintain them |
| # / 00:29:51 | Niklas Kunkel | So, the ecosystem kind of really grew very, very quickly by just integrating our Oracles in everything, so we don't want to risk some kind of chain reaction by kind of pulling the rug out from everyone and saying, okay, now you're cut off from Oracles and nobody has a plan B. Right? So, even by adding this white list, we don't have to actually use it right away. It is another tool for MKR token holders to kind of use down the road. |
| # / 00:30:38 | Niklas Kunkel | And, another thing I wanted to bring up is that there are ways to utilize the white list to extract value from the Oracle ecosystem that is not monetary. For example, you could have some kind of quid pro quo relationship between someone who wants to do something for us, but also gain access to the Oracles. So, let's say there's a DEX, and this DEX wants to start offering margin trading. Well, if you want to offer margin trading, you're going to need Oracles, right, to know when to liquidate the positions, when to margin call. So, what if we could have a deal with this DEX where we say, hey, we'll give you the Oracle services, but in exchange, we want you to list DAI as a base token trading pair. Right. |
| # / 00:31:48 | Niklas Kunkel | So, these are the kinds of things that are possible, so I kind of just want to get everyone's minds thinking more in that direction, and so to just really drill this point home, the acceptance of this proposal does not mean that this is what's going to happen on day one when we launch Oracles V2, right? The Maker development fund will still be paying for the feeds. Ideally, they're not going to charge people for the first year to actually use the Oracles, but it is an explicit approval that this is the model and relationship of incentives and actors that we're moving towards. |
| # / 00:32:38 | Niklas Kunkel | So, this is going to be one of the proposals that I will be posting tomorrow on the forum, so if you have any feedback or concerns, comments, or if you even have some better ideas of how we should structure this, I am all ears. So, that is it from my end. |
| # / 00:33:09 | Richard Brown | All right. You surprised me. I was not expecting this, so if anybody has any questions, you want to simmer on that for a second, then type them in the chat on the side or ask us now. We will also have, like I said earlier, the 30 minutes at the end of the hour to do a random Q & A. |
| # / 00:33:30 | Richard Brown | Thanks for that, Nik. That was great. Cyrus, I'm going to hand it over to risk now if you want to take it away. Did you want to talk about the DSR today, or do you want to talk about the peg first? |
| # / 00:33:41 | Cyrus Younessi | You know, I'm thinking of maybe holding off for a week on the DSR talks, because it seems like there's some good discussion kind of getting started in the forums. And, we just... I think the blog just went out yesterday or the day before, so I'm thinking of kind of letting that stew for a bit and then circling back when we've heard some more thoughts. But, if there are any questions, or if anybody else wants to talk about it, I'm certainly open to as well. |
| # / 00:34:18 | David Utrobin | I guess, maybe covering the gist of what it is that's being talked about. |
| # / 00:34:23 | Cyrus Younessi | Sure. Give me one second to pull it up on my... I'm going to skip the mechanics portion of it and just kind of focus on the governance and risk portion that was... that I highlighted in the forum thread. |
| # / 00:34:56 | Cyrus Younessi | So, I guess the first major... It's not really a decision but rather an educational topic, is where the DSR even comes from and how it's created, and so forth. So, essentially, the DSR is... when you put DAI into the DSR contract and then you retrieve it, obviously, you are taking out more DAI than you put in. So, natural question, where does additional DAI come from? The DSR contract has the functionality to essentially just print that DAI and give it to you regardless of any exogenous factors or just collateral or what not. |
| # / 00:35:50 | Cyrus Younessi | And, at first glance, this seems concerning, right, because kind of the main tagline has always been collateral backed DAI and this appears to be an instance where DAI is being minted without any collateral being posted against it. But, to kind of remedy this little issue, the amount of DSR that is created gets weighted... not weighted, but gets counted against the collected stability fees or the internal surplus of the system such that if, for whatever reason, there wasn't enough collected stability fees to pay out the DSR, then the DSR essentially gets recorded as bad debt. Right? And, so would be subject to an MKR dilution. |
| # / 00:36:54 | Cyrus Younessi | So, the point being that on the technical side of things, DSR is paid for by MKR holders out of the crude stability fees, so now the questions that I pose to the community, which is kind of the discretion that will need to be had probably over several months, to be honest, is what to do about that cost? Should the MKR holders just absorb it, or should they kind of redirect a DSR fee onto the total stability fee for CDP owners, or some hybrid, some blend? |
| # / 00:37:52 | Cyrus Younessi | And, there are a number of economic or business considerations to be had here, so kind of in the naïve sense, if you were to just pass on the 100% of the cost to the CDP owner, then you could end up in a situation where certain collateral types, which are depending on generally low stability fees in order to be buyable, would immediately get priced out. And, so the question is to what extend do we allow for... I don't know if subsidy is the correct word here, but, essentially, do we allow for leeway for certain collateral types, and then either MKR pay for it out of the total stability fees collected or, potentially, there are options of collateral specific DSR adjustments. |
| # / 00:38:59 | Cyrus Younessi | So, for example, if there's a particular collateral type that is extremely popular, is generating a lot of DAI, and is, for whatever reason, contributing to... is proportionally contributing more to a DAI price decline, then would it make sense to add extra stability fee to the collateral type loan separate from its already calculated risk premium rate? So, that's kind of one big question that I'd pose here. |
| # / 00:39:39 | Cyrus Younessi | Other questions are how do we go about calculating or evaluating these phenomenon? Phenomena. In the sense that will we even be able to kind of assess which collateral types are contributing more or less than other collateral type storeds or particular DAI price decline, and stuff like that. You all should give it a read and think about some of these topics and post some thoughts. There's already a bit of a good discussion going on, so I'd like to kind of see that take hold over the next week or so and then maybe we can revisit. |
| # / 00:40:24 | Cyrus Younessi | That being said, if no one has any... or, looks like there's one question. Okay, yeah, so if anyone has any... no one has any questions, we can move on to the monetary policy aspect. Okay. One question. Allen says, "Why not start the DSR at zero and then increase it?" Yeah, that's an interesting idea that was brought up a few other times, so I think one answer to that is under the current governance cycle that we are proposing, it's typically a one month governance cycle. We're not sure with what cadence the DSR adjustments would come in yet. Would they be kind of every week versus the monthly schedule we're planning right now? So, we're going to have to kind of decide that from a governance perspective. |
| # / 00:41:34 | Cyrus Younessi | But, that being said, if we could have frequent cadence adjustments, then sure, we could start it at zero and then go from there, maybe start it at some very low, safe number as well. But, if you have any specific reasons why you think it should start at zero, once again, please post your thoughts in the forum. |
| # / 00:41:58 | Cyrus Younessi | Another question. How will users hedge DSR exposure if they have a collateral specific DSR? I'm not sure if I understand the question, but interesting, nonetheless, about how people will go about hedging various exposures. All right. Maybe we can talk a little bit more about this at the end of the call. I want to make sure we get through the monetary policy stuff, and then we can circle back, for sure. |
| # / 00:42:44 | Cyrus Younessi | All right, Vishesh, if you are on the line... |
| # / 00:42:50 | Vishesh Choudhry | Sure. Okay. So, I'll just start sharing my screen here. Cool. Hopefully everyone can see everything all right. So, just to kind of set the stage. What's been going on in the last week or so? ETH price has been comparatively stable. There's not been that much activity. It did tick up slightly, but more or less, pretty stable at the end of the day. |
| # / 00:43:26 | Vishesh Choudhry | That's reflected in what's been going on with collateral, so not much movement in terms of the total amount of collateral locked. Not much added. Not much removed. Not since basically the end of August. So, actually, again comparatively... so, the... just to touch on the stability fee real quick... so, you know we saw that there were a few drops to the stability fee over the last couple weeks. You know, first 18.5, then 16, then 14.5 on September 8, so what's really interesting is during these first couple of drops, there was actually a dropping DAI supply at the end of the day, which is very odd. But, around the 4th this started to tick up a bit, and then with the 14.5% drop on the 8th, it ticked up a little bit more, so there were a decent amount of draws. There was a little bit of circulation, some wipes as well, which is particularly interesting. |
| # / 00:44:34 | Vishesh Choudhry | So, a slight increase in the DAI supply, a total of a million or a couple million over those few days, which actually basically more or less just balanced out some of the drops just at the end of the month. So, all in all, in the one month timescale, pretty levelled, but there was a dip before and then after this [inaudible 00:44:56] was increased... or decreased, a little bit of a recovery. |
| # / 00:45:00 | Vishesh Choudhry | So, interesting to watch this trajectory. It will potentially keep increasing, but... it should, but hard to say given that things seem comparatively kind of quiet right now. |
| # / 00:45:17 | Vishesh Choudhry | And, as far as just some of the longterm activity, you know, what we saw was in collateral terms, basically, the evaluation of the community on the amount of risk that they need to manage in their CDPs was sort of, over the beginning of the year, that was pretty significant, then it had slowed down a bit, and so the amount of locked collateral had started coming down again and getting outpaced by the amount of collateral that was being moved. And, then relatively recently, just at the beginning of August, the amount being removed started to slow down pretty significantly, and the amount that was being added started recovering again. |
| # / 00:46:01 | Vishesh Choudhry | So, now these are kind of on pace, but in nominal terms, that curves have not been moving that much. And, then in terms of debt, so the amount of debt that's been being pulled out had kind of consistently been increasing throughout the history of Maker, and then again, somewhere around mid-end of July, that slowed down a little bit, but then again, so did the amount of collateral that was being wiped. Just the amount that was being wiped was slightly outpacing the amount that had been being drawn. Although, that trend seems to have slowed down over the last few weeks. |
| # / 00:46:44 | Vishesh Choudhry | So, potentially, a little bit more of an increase in DAI supply coming, but we'll have to continue to watch that. As far as just longterm relationships between ETH price and what's been going on, so there were some bites over the last month or two, nothing relatively recently, but that makes sense because ETH price had come down a little bit at mid-end of July. |
| # / 00:47:09 | Vishesh Choudhry | Again, relationship between ETH price and collateral has held pretty steady, pretty solid. In relationship everyone is pretty familiar with this one at this point. ETH price had sort of in the longer run been decreasing a bit since July, and so collateral, which was initially dropping due to the nominal drop in ETH price, was sort of recovered and reversing with that trend. So, again, people perceive that as ETH price drops, there's increased risk, and so they add more collateral. That is a pretty well-known phenomenon at this point, at least in this community. |
| # / 00:47:48 | Vishesh Choudhry | Relationship between debt and ETH price. So, actually kind of comparatively not very well related lately, though, obviously, as the amount of collateralization moves... and, I think the likelihood here is that people are sort of controlling or constrained in their capital, and so what happened was as that price moved and you saw those strong relationships between ETH price and the amount of collateral that was locked or freed, instead as a lever, people started to manipulate the amount of debt that they had. So, basically, as that collateralization ratio fell, one of the more common activities, rather than adding back collateral, was actually to remove some debt, clear it out, refinance it, whatever. |
| # / 00:48:36 | Vishesh Choudhry | And, so that's, I think, part of the reason why you saw some of this drop in debt, and that's pretty well explicable by a bit of decreased demand for leverage, given the way ETH prices have been. But, it does seem that there is a little bit of change on the horizon. It sort of depends on what ETH does next. So, I think it's a little bit of wait and see right now, just kind of watch and see what continues to happen. |
| # / 00:49:03 | Vishesh Choudhry | As far as the DAI peg itself, over the last 24 hours, pretty stable, slightly above $1.00, very low trading volume, just about a million, which is half of what we usually see on like a quiet day. But, the price at $1.01 roughly, and then if we look at... so, yeah, I mean I think was- |
| # / 00:49:29 | David Utrobin | Vishesh, I have a quick question for you, if you don't mind. |
| # / 00:49:32 | Vishesh Choudhry | Yep. Yep. |
| # / 00:49:33 | David Utrobin | I know that a lot of times when we look at the 24 hour DAI volume weighted average price chart, it's usually a lot more... it follows that kind of bell curve distribution. Whereas, lately in the last day or two, I've noticed that it's been a lot more fragmented and weird. Do you have an idea about that? Do you think it's just... |
| # / 00:49:48 | Vishesh Choudhry | Volume, I mean- |
| # / 00:49:50 | David Utrobin | ... a result of lower volume? |
| # / 00:49:51 | Vishesh Choudhry | Yeah, I mean lower volume will do that. The prices, I think, follow a better curve when there's stronger volumes, which makes sense. When there's lower volumes, easier to be a victim of slippage, easier to sort of have price impact major trades, so just naturally, you do tend to see that sort of spread out a little bit more. So, I really do think that's the explanation. It is interesting... |
| # / 00:50:18 | David Utrobin | Thank you. |
| # / 00:50:18 | Vishesh Choudhry | ... to see on low volume days... sorry, I do nerd out a little bit. On low volume days, it is interesting to see some of the microstructure patterns, so what's cool is you see Kyber here on the sort of low and high ends of these prices. Coinbase is almost always like dead center. Oasis is generally dead center with maybe a little bit of a wider spread. Uniswap is pretty flat throughout and tends to get pushed up either to a higher end or to a lower end usually, so it's pretty interesting to just watch these microstructures. At some point, someone will write a paper on that, but... |
| # / 00:50:51 | Vishesh Choudhry | So, to talk about some of the secondary lending platforms and what's been going on there, so I think with some of the changes in the fees and some of the lower supply, it is easier for some of these platforms to run into a bit of a crunch, so the rates can jump up or drop down a little bit more than they had, say, three months ago. And, so that's where you start to see some of these spikes in values that you can explain around stability fee changes, though that makes sense.. |
| # / 00:51:25 | Vishesh Choudhry | What's interesting is with the 14.5% drop, there was actually not a ton of responsiveness you some in some of those other metrics and in some of the secondary lending metrics. So, that's a pretty interesting phenomenon to observe. Additionally, you see in the borrow volumes, so borrow volume was for... we'll just look at dYdX and Compound separately, and then we can talk about all together. But, for dYdX that borrow volume had ticked up maybe 20%, 30% just in the last, say, three weeks from 6 million to actually up to about 9.5 million, so a little bit more. |
| # / 00:52:09 | Vishesh Choudhry | Compound, you saw from the end of August, there was a little bit of a jump, and then starting to dip down very slightly, about 3 million, which is not a huge percent of change for them, but it is a little bit of a quieting down. And, you do see this sort of like couple month long trend of about 27, 26 million. Whereas, dYdX was very low around end of July at around 3 million, and you know, has basically tripled its borrow volume since then, but nominally smaller numbers on dYdX. |
| # / 00:52:46 | Vishesh Choudhry | So, all said and done, the total secondary lending borrow volume on at least these two platforms around 30 million in end of July and around 34 million, so a slight increase, but overall, comparatively pretty level. What's interesting is that increase was primarily happening during a time of a decrease in stability fee, so decreasing stability fee, theoretically, having these positions on Maker itself, is a bit more attractive. So, that's a bit contrary to what you might expect. |
| # / 00:53:23 | Vishesh Choudhry | So, we look at supply volumes. The total supply, again, around end of July, was around 35 million and now is up to about 48 million, so that's a pretty significant increase. And, that does sort of interestingly correspond to the time of the drop in stability fee as well, but that has been pretty steadily increasing. So, I would actually say the interesting part about the supply volume on these secondary lending platforms is how little it was correlated to any movement in stability fee. It just appeared to be a somewhat independent track. |
| # / 00:54:06 | Vishesh Choudhry | So, we look at the excess cash. So, if you recall, two or three months ago we were looking at a graph that was decreasing, and so utilization was like significantly increasing during that time. During the same time period, utilization has been significantly decreasing, and again, leveling around this sort of effective floor for these platforms around 70%-75%. You generally don't see it go much lower than that, because then the rates start responding accordingly. So, it seems to be a bit of a natural economic floor, thought it's not like a hard coded floor or anything. |
| # / 00:54:46 | Vishesh Choudhry | So, yeah, the excess supply on these secondary lending platforms has come up. It's a very interesting time right now, because there is a lower stability fee. There is a bit more excess supply on secondary lending platforms, which, as we've discussed, can often be interpreted to be related to excess supply overall. And, given that the actual total Maker supply has been pretty steady, decreased a little bit, recovered a little bit, but more or less overall pretty steady, this could be interpreted as a bit of a drop in demand. And, so I don't want to FUD, but I would just say keep an eye on this with regards to how much we're dropping that stability fee, because though the DAI price is doing well right now, if there is a sudden bull market, that would be a pretty fluid situation and you could see these rates and the prices for DAI respond pretty quickly. |
| # / 00:55:52 | Vishesh Choudhry | Though, it will be interesting to sort of... that would be kind of a third cycle for DAI, so it would be interesting to see what happens in that situation and how resilient it actually is now, because what we did observe was the second time we went through that situation, DAI was a bit more resilient than it was the first time. And, so as we come around again, it will be very interesting to watch. |
| # / 00:56:17 | Vishesh Choudhry | Yeah, and then... So, just in the longer time scale, seven days, again, slightly low volumes there, though it has been lower in the last couple days than it was overall for those seven days, but the price, again, pretty smack dab similar to the price for last 24 hours. So, this you do, as I mentioned, David, see a little bit more of that curve, primarily because it is that long a time scale. |
| # / 00:56:46 | Vishesh Choudhry | And, you do see this sort of bimodal situation going on, because, basically for a few days there, this was roughly the center, and for a few days, this was the center. So, that's what explains that is just a bit of a shift in price, though, again, within a pretty small band when you really think about things. I'll just switch a few decimal places. You... yeah. We are talking about a pretty small range of variability in the first place. |
| # / 00:57:16 | Vishesh Choudhry | Okay. So, that's the top of the hour. I will pause for any questions, and then hand it back to, I guess, Cyrus. |
| # / 00:57:25 | Richard Brown | We have a deep discussion about the state of your battery going on in the chat, if you want to comment. |
| # / 00:57:33 | Vishesh Choudhry | Yeah, I saw that. Don't worry, y'all. I have chargers. |
| # / 00:57:36 | Richard Brown | All right. There was a question from [Bartech inaudible 00:57:42], though. When you're measuring the Kyber reserve... Can you see that one? Is it specifically Kyber only, or is it the Kyber trades that are being forwarded to Oasis? |
| # / 00:57:54 | Vishesh Choudhry | So, it is my understanding that that is filtering out for double counting. That's a data set that I get from Lev's group, but yeah, that is filtering out some degree of double counting with regards to Kyber, if that was the concern. It was something someone tweeted about it at some point as well, and we addressed that a while back. |
| # / 00:58:26 | Richard Brown | Do you want me to read these out, Vishesh? Or, can you see them on your screen? |
| # / 00:58:29 | Vishesh Choudhry | Well, I see the USDC/DAI. Yeah, so we've sort of gone back and forth on USDC/DAI. |
| # / 00:58:34 | David Utrobin | Hold on. Hold on, Vishesh, just one second. Since this call... this is just a general matter. Since this call is recorded, and also since we do audio versions, it'd be good to like really clearly read out the questions, so yeah, just wanted to make sure that the people listening after the fact get a sense of the question. |
| # / 00:58:52 | Vishesh Choudhry | Yeah, good flag, David. So, the question is about incorporating USDC/DAI in the metrics. Is there a reason to do it? Is there a reason not to do it? I don't think there's a specific reason not to do it. Is it currently a big part of what I'm looking at? No, this is the DEX, ETH to DAI trades at the moment, but I know Lev had talked about potentially including that in his data set, which would make it easy for me to pull. I think it's just an operational thing. There's a limited number of hours and things to work on, and so a lot of that time does go to some, you know, like the risk modeling work. But, there's no specific reason, I would say, to not do it that I can think of. |
| # / 00:59:49 | Vishesh Choudhry | Okay. I think those were all the me questions, so I'm just going to hand it back to Cyrus. |
| # / 00:59:56 | Richard Brown | All right. |
| # / 00:59:56 | Cyrus Younessi | Okay. Cool. |
| # / 00:59:59 | Richard Brown | Let me jump in really quick and transition here, so we're at the top of the hour. If people would like to hang around, we'll do a general Q & A for the next 20 or 30 minutes, or until interest wanes. For the people that need to bounce, though, thank you to Nik for your presentation. Thank you, Vishesh. Thank you, LongForWisdom. That was a very interesting call. |
| # / 01:00:25 | Richard Brown | Standard reminders apply. Please join the discussion in the governance forum. Threads have been posted. It's not hard to find, forum.makedao.com. Lots of interesting things happening there, governance, risk, and now Oracles. The new mandate from Nik will be going up soon, so keep your eyes out for that as well. All right. Thanks, everyone. It was a really good call. Please, feel free to hang out for a general Q & A. Back to you, Cyrus. |
| # / 01:00:51 | Cyrus Younessi | Okay. So, let's talk a little bit more about the DSR. So, Allen asks, "After the DSR, when making changes to DSR stability fees, will both be moved at the same time, or will one be preferred over the other one?" That's a really good question, and I don't think we know the answer right now, because we'll have to kind of see how things shake out. |
| # / 01:01:20 | Cyrus Younessi | But, essentially, it depends what kind of behavior you're trying to incentivize. So, if we wanted to create more demand for DAI, then we could just raise the DSR without necessarily touching the stability fee, and that would essentially... look, if you're trying to raise the DAI price, then that should obviously have an affect. |
| # / 01:01:50 | Cyrus Younessi | The secondary part of that would be do you then also raise the stability fee to compensate for that extra DSR? And, I guess if you did both, then they would both have separate affects on the DAI price, so yeah, I guess we're going to have to learn through the governance community how we want to incur these parameters. I am kind of thinking that there's going to be a lot of overshooting and undershooting, and not entirely sure how it's going to shake out through governance, to be honest. |
| # / 01:02:35 | Cyrus Younessi | Kind of the goal was that stability fees would generally not be too variable, so one idea that was floated around, was that stability fees would only change... and, this is just purely hypothetical, but once a month during the standard governance cycle, so every month we can evaluate how has the DSR evolved over the past month and make any necessary stability fee changes, and then the DSR itself could be changed separately through a weekly process. So, every week we could have governance for DSR, move it up or down, but then only change the stability fee once a month. So, that was one potential idea that was floated around. |
| # / 01:03:28 | Cyrus Younessi | In terms of preference, I think there is going to be a preference, but I think that's going to probably require some more sophisticated statistical models on the demand and supply of DAI. And, I kind of think we'll cross that bridge when we get there, because if we're going to... I think we're all hoping, crossing our fingers, for huge expansion in DAI supply over the coming years. It's probably just going to be difficult one way or another to kind of decide what is preferred over the other with our lack of historical data. |
| # / 01:04:20 | Cyrus Younessi | "Any thoughts around a pooled model for DAI locked in DSR with an underlying token that is a claim on the asset pool?" That is also an interesting question, the idea of some sort of tokenized version of the DSR. Kind of think that's certainly, technically something that's possible. I kind of imagine we'll see some community implementations of that pretty soon after MCD launches, if I had to guess. Yeah, something that I know a lot of people would be looking forward to. |
| # / 01:05:08 | Cyrus Younessi | "Stability fee changing once every six months". The question is, or the comment is "should the stability fee be changed once every six months?" I don't think you can really set any cadence, really. I think the one month is just kind of under standard operating procedures, but there could conceivably be times where you need to have mid-month changes for some sort of emergency procedures. I think it's kind of okay to re-evaluate or monitor collateral types on a monthly basis. It wouldn't require an enormous amount of work, just kind of some checking in on things. |
| # / 01:05:51 | Cyrus Younessi | I know at some point in the past, I think Rune had floated an idea about doing quarterly evaluations. We'll see. I think in the early days, the more monitoring and updating we can do, the better. Try to keep our foot on the gas as much as possible. |
| # / 01:06:12 | Cyrus Younessi | "Will the DSR stability fee be used to keep the DAI supply growing at a slow pace?" I don't know about keeping it at a slow pace, but I think the pace that will keep DAI stable and keep debt ceilings within reasonable limits is probably the way to do it. The question we should be asking is at what rate should the global debt ceiling for the entire system be growing? |
| # / 01:06:51 | LongForWisdom | I got a question, if that's all right. |
| # / 01:07:02 | Cyrus Younessi | Sure. Yeah. |
| # / 01:07:05 | LongForWisdom | So, you mentioned we could potentially increase the DSR without increasing the stability fee. |
| # / 01:07:12 | Cyrus Younessi | Right. |
| # / 01:07:12 | LongForWisdom | So, does that affect the risk of the system, right? Because, we set the risk premiums based on the risk of a specific collateral. |
| # / 01:07:20 | Cyrus Younessi | Yeah, it does. |
| # / 01:07:21 | LongForWisdom | If we increase the [crosstalk 01:07:22] it screws that right? |
| # / 01:07:24 | Cyrus Younessi | It does in a roundabout way, because if the MKR dilution potential is lowered because they're now essentially paying for the DSR, then yeah, basically there is less MKR dilution potential for the rest of the system. Now, I don't think it's going to be a huge deal. At a small scale in any sense. But, surely, if there's a situation where the DSR needs to be like 10%, and then we only want to pass off like, for whatever reason, 3% of it to the borrowers, then yeah, that's going to create a huge problem, for sure. So, it has to be within limits, but something that we're aware of. |
| # / 01:08:19 | LongForWisdom | Thanks. |
| # / 01:08:32 | Cyrus Younessi | Another interesting comment that was brought up on the DSR by [Fenny inaudible 01:08:41], actually one of our in-house counsel, was that if only a small percentage of the total float of DAI is locked up in the DSR, then you wouldn't have to pass on the full amount to the stability fee. So, for example, if the DSR is set to 5%, but only 10% of the DAI supply is locked up in the DSR, then that 5% that is being paid out is only being paid out to that fraction of the total DAI supply, right? |
| # / 01:09:24 | Cyrus Younessi | So, then you wouldn't go and assess a 5% DSR fee to the borrowers, because that 5% would be essentially on the total DAI supply. So, we'd have to equate the nominal amounts that are being paid out if we were to pass it along. It's kind of an interesting point. I guess, David says he made that point months ago. Good for him. I don't remember that. |
| # / 01:09:51 | David Utrobin | I even made an Excel sheet calculator and sent it to Rich, even before I was hired, about this exact point. |
| # / 01:09:56 | Cyrus Younessi | Did you? |
| # / 01:09:56 | David Utrobin | Yes, I did. |
| # / 01:09:57 | Cyrus Younessi | Really? Okay. Cool. Cool. Good for you. |
| # / 01:10:02 | David Utrobin | Yeah, scratching my own... itching my own- |
| # / 01:10:04 | Richard Brown | I don't remember that, actually, so it's yet to be confirmed. We'll see. |
| # / 01:10:10 | Cyrus Younessi | Yeah, I didn't- |
| # / 01:10:17 | David Utrobin | You cut out, Cyrus. |
| # / 01:10:19 | Cyrus Younessi | I think my internet is dying. |
| # / 01:10:31 | David Utrobin | You're back now. Your Internet's slow. |
| # / 01:10:32 | Cyrus Younessi | Can anyone hear me? |
| # / 01:10:33 | David Utrobin | So that... Yeah, you're here. |
| # / 01:10:35 | Cyrus Younessi | But, yeah, it is good to be reminded, because this total stability fee, it's not as simple as saying just risk premium plus the DSR fee, so we'll have to do a little bit of extra math there. |
| # / 01:10:49 | Cyrus Younessi | Question about stress testing the system in such cases described before. I'm assuming about... in terms of MKR dilution potential. Good question. A little bit out of scope for this call right now, but happy to talk about it further in the governance channel. |
| # / 01:11:16 | Niklas Kunkel | Okay. I think I see a question from Frank, and he wants to know if the feeds that are providing the prices to the Oracles if they have any requirement to own MKR, or if they have any reasonable motivation to provide valid data? |
| # / 01:11:39 | Niklas Kunkel | So, currently, the incentive for them to have integrity is the fact that they get paid, right? So, they're getting paid every month from the Maker Development fund. The idea is that in the future, governance, when they take over paying the stability fees would also gain the lever of deciding how much the feeds get paid, right? So, you can increase or decrease that incentive depending on how risky you want to be. |
| # / 01:12:14 | Niklas Kunkel | From a technology point of view, we can also aid in kind of risk mitigating here. Right. So, you mentioned is there a requirement for them to own MKR? So, I think what you're kind of alluding to here is are they going to stake MKR, or should they stake MKR? My opinion on this is yes, partly because the more you have them stake, the more they have to lose, the more integrity they are bound to have, but then we get to, okay, what is the right amount of MKR that they need to stake? |
| # / 01:12:51 | Niklas Kunkel | And, so there's a couple other things that we can do here. One thing that's been brought up is, well, if you force someone to stake MKR to get some kind of compensation stream, don't you also need to kind of increase the compensation based on the opportunity cost of giving up ownership/control of the MKR? That's something that maybe is more of a question for Cyrus, but yeah, so staking is on the roadmap or will be on the roadmap. You guys will get a chance to actually vote on that roadmap to see if you approve, and we'll have a couple other kind of tricks in there to lower that risk and provide more of an incentive for feeds to have integrity. |
| # / 01:14:07 | David Utrobin | I have of a kind of a question that relates to my little comment there about stochastic modeling. Isn't stochastic modeling basically stress testing for any financial risk stuff, generally? |
| # / 01:14:24 | Cyrus Younessi | I wouldn't call it stress testing, but I think what you're trying to say is like simulation testing. |
| # / 01:14:29 | David Utrobin | Yeah. Yeah. |
| # / 01:14:31 | Cyrus Younessi | And, yeah. I mean, not all simulations are for stress testing purposes, but yeah, I see what you're saying. |
| # / 01:14:40 | David Utrobin | Okay. Cool. Just wanted to self-check if I'm somewhat right about that. Okay. |
| # / 01:14:59 | Anjan Vinod | Hey, Cyrus, had a question regarding for MCD, the difference between REP V1 and REP V2 in terms of the launch collateral of MCD. So, how are you thinking about incorporating REP V1, and if not... oh, sorry. If so, does it make sense to incorporate REP V1 and then over time as the migration process, then kind of deploy for REP V2? |
| # / 01:15:28 | Cyrus Younessi | Right. We were actually just talking about this a couple days ago in their discord channel. Interestingly enough, I think their community themselves are a little bit kind of pushing for, or kind of suggesting, that we hold off and wait for V2 of REP, and I'm kind of onboard with that as well, just merely from the perspective of the overhead required for risk evaluations and governance and everything and pushing a collateral type through. And, to have for a short-lived collateral type may not be worth it. because, I don't think there would be any smooth transition where we could just kind of flip from one to the next. |
| # / 01:16:23 | Cyrus Younessi | And, in fact, the game theory and the cryptoeconomics of REP V2, as I understand it, are significantly different, so it would have to have a fresh pair of eyes on it anyways. A couple people from the Augur, The Forecast Foundation, have also suggested potentially waiting for V2, but all that being said, if kind of a collateral application gets filled out from anyone from the community, whomever gets posted, and MKR holders, for whatever reason, want it, and they vote it up on the priority list, it's certainly... we're not in charge of saying what gets in and what doesn't get in. So, it can happen. |
| # / 01:17:15 | David Utrobin | I think that brings up a really interesting question of how often we should re-poll of the asset priority list. |
| # / 01:17:21 | Cyrus Younessi | Right. In this particular case, we obviously didn't do our own due diligence on this, so I feel like our original asset priority poll was very biased. Not biased, it's just incomplete, right? We should've specified if we were talking about V1 or V2, because I do think that changes things considerably. |
| # / 01:17:43 | Cyrus Younessi | That being said, we kind of just had this discussion with the Augur community in the last couple days, so still kind of mulling it over. That's a great point. |
| # / 01:17:54 | Anjan Vinod | Yeah, that makes sense. I guess, just to add on, REP would be essentially fungible between the V1 and V2, aside from a fork itself, correct? So, if you... if MCD would incorporate REP 1 and in the invent that there is a fork on V2, then REP holders would have to migrate, but then you have that 60 day allotment to kind of migrate as well. So, you could incorporate V1 and then over time if you have a fork, you have enough time to them migrate over to V2. |
| # / 01:18:34 | Cyrus Younessi | So, if there's a lot of V1 collateral in MakerDAO and then that 60-day window starts, what happens in terms of... Oh, let's see. What happens in terms of Oracle prices? What happens if you don't switch within 60 days, and have to think about kind of other ways for CDP users to somehow game the system? Actually, are there ways we could unintentionally hurt CDP owners, or are there ways that CDP owners could unintentionally or intentionally hurt MakerDAO? |
| # / 01:19:16 | Cyrus Younessi | So, I think there are at least a few open questions. I don't know if there are simple answers to these. If they are, then sure, it's a possibility. We can surely talk about it as a community. I don't really have too strong of thoughts one way or another. |
| # / 01:19:38 | Anjan Vinod | Got it. Understood. Thanks. |
| # / 01:19:42 | Cyrus Younessi | Actually, Nik, what would happen in that case when a token upgrades, and... does anything happen from the Oracle side, because the exchanges will just kind of behind the scenes shift from one to the other, but does the market activity stay fluid? Do they keep trading open? I guess, do they shut down trading for... |
| # / 01:20:13 | Niklas Kunkel | There's usually a window where they- |
| # / 01:20:15 | Cyrus Younessi | ... a period. It shouldn't have to an impact, right? |
| # / 01:20:19 | Niklas Kunkel | Right. So, basically, whenever this upgrade/fork/whatever occurs, there's usually, you know, we're going to be putting this pair in maintenance mode for the next six hours, and when the pair comes back up, the order book is basically clear. |
| # / 01:20:43 | Niklas Kunkel | So, from an Oracle standpoint, we would just... it's whatever the exchange thinks REP is, that's what the Oracle is going to source. So, that's actually a risk, right? We, as a community, may have a difference of opinion of which, say, REP token is the real REP token, right? So, it's not something we can passively just say, oh, it's not our problem, it's the exchange's, right? Because, individual exchanges may have difference positions, so say Coinbase and Kraken have alternating views on what the real REP token is, you know, we can't just have the Oracle keep sourcing prices for both and taking the average. |
| # / 01:21:30 | Niklas Kunkel | So, it's something that we have to be actively involved in kind of first deciding which kind of version of a token that we want to support and when that transition will happen, because it involves having to close... so, there's a couple things. You either have to force close all existing positions, which you can do by just setting the collateralization ratio super, super, super high, and then just force liquidating everyone after giving them a heads-up of, say, a month beforehand, right? You need to close your position. You need to close your position. |
| # / 01:22:19 | Niklas Kunkel | Or, an alternative is that we say as a governance community, well, we'll support both, so now you need Oracles that are sourcing prices for both. But, that's really up to you guys. |
| # / 01:22:36 | Cyrus Younessi | Yeah, I'm not a fan of bringing on collateral that we know at some point we're going to have to force close people out. That just seems like something we shouldn't want to involve ourselves at this point, especially without any sort of robust decentralized messaging plan. We'd have to spend some time figuring out how we can even get messages out to people. |
| # / 01:23:00 | Niklas Kunkel | So, Cyrus, let's say that OmiseGO has a bug in their token contract, and they immediately redeploy the token, that is just an externality we're going to have to deal with. It's not something that we can just kind of plan beforehand. Oh, we're never going to take any shitty collateral that could possible lead to this circumstance. |
| # / 01:23:31 | Cyrus Younessi | I mean, if Omise GO has a bug in their token, you think that MakerDAO should just force close everybody's positions on the spot? |
| # / 01:23:44 | Niklas Kunkel | I mean- |
| # / 01:23:48 | Cyrus Younessi | That's an interesting... it's an interesting question for sure. We should have some plans for these kinds of- |
| # / 01:24:01 | Niklas Kunkel | There's different scenarios. There's a planned, kind of- |
| # / 01:24:04 | Cyrus Younessi | Eventually, I think we need to start [inaudible 01:24:05] mitigate them. |
| # / 01:24:15 | David Utrobin | We did not hear anything of what you just said. |
| # / 01:24:20 | cmooney | Cyrus, if you turn your video off, you'll be able to probably speak better on lower bandwidth. |
| # / 01:24:28 | Cyrus Younessi | Sorry. [inaudible 01:24:33] |
| # / 01:24:36 | cmooney | Or not. |
| # / 01:24:42 | Niklas Kunkel | So, maybe I can just chime in for a second. We'll have different processes in place depending on the scenario, so there's a big difference between... I don't want to keep picking on these guys... you know, Omise GO planting a bug in their token contract and immediately deploying a new one and a kind of planned migration where, say, Augur thinks, hey, we need to expand the capabilities of our token, and so we're going to do a planned token upgrade on this date one month from now. |
| # / 01:25:22 | Niklas Kunkel | So, how we handle those two scenarios is probably very, very different, but I would implore the community to look at these primarily from a risk perspective. If you have to force close everyone's position because it presents some kind of existential risk to the system, then it doesn't matter that the optics of force closing are bad, you have to. But, that's at least my position on this. |
| # / 01:25:59 | LongForWisdom | I have to disagree with that, really.. |
| # / 01:26:14 | Richard Brown | Yeah, there's... we should point out that there's some dramatic proclamations being made, and none of that is written in stone. So, there's going to be lots of conversations about how to manage risk in edge cases in the future. We don't know what it looks like for a token to get compromised en masse yet. We are going to find out what that looks like. I guarantee it. But, there's... |
| # / 01:26:37 | Richard Brown | And, I'm not the risk analyst. I'm not a system market tech, but I am aware that there's four or five different things we can do in those situations of the top of my head, like set the debt ceiling to zero, you set the stability high, you liquidate people, you allow them redeem, you just immediately transfer their whatever existing collateral to ETH and then call day on that thing, or you liquidate them and you make them accept the penalty. So, there's all kinds of things we can do. It'll be fun to figure out exactly how that happens, though. |
| # / 01:27:13 | Richard Brown | We are, actually... we're at a half hour after the call, and I have other engagements, so I can no longer stick around. I'm not sure whether Cyrus is going to make it back or not. What do people think? Are there more questions to be asked? I can hand the hosting off to somebody else, or should we stick a fork in it and continue this next week? |
| # / 01:27:39 | David Utrobin | I think we've heard a lot. I think it's probably good we stick a fork in it. We usually do it an hour and a half anyway. |
| # / 01:27:45 | Richard Brown | We do usually do it an hour and a half. So, thanks, everybody, for joining. Lots of great calls. I've learned a lesson that if we put up with at least 35 seconds of uncomfortable silence, some of you will continue to answer questions, so we'll use that as a plan in the future, and I won't end these calls so quickly. All right. Thanks, everybody. Talk to you next week. Look for forum thread, look for YouTube, look for SoundClouds, and follow me on Twitter. Like, subscribe, and whatever else. All right. Thanks, everybody. Buh-bye. |