Governance and Risk Meeting: Ep. 51 (September 5 - 2019)¶
Video | Audio | Discussion
| References | Person | Text |
|---|---|---|
| # / 00:00:00 | Richard Brown | Hello, everyone. Welcome to the September 51th edition of the scientific governance and risk meeting. My name is Richard Brown. I'm the head of community development at MakerDAO. In these calls, we are enormously interested in everything that the community has to say, particularly people that are willing to come up to these calls. So if you have a question or a comment, please, please, please, please, I'm begging you, type it in the chat at the side or wherever your UI happens to have placed it. Or if you have access to a microphone, please do not feel shy about getting on the mic and interrupting us. We interrupt each other all the time. There's no reason why the community should not do it, as well. |
| # / 00:00:42 | Richard Brown | Today, we have a new spin on the agenda which is possible going to be a sign of things to come. Historically, these meetings have been about governance and then they were about governance and risk. And soon, they are going to be about governance, risk, and oracles. And soon after that, they're going to be other things, as well. We are going to be having some conversations about how to optimize these calls, how to break them up, how to organize them. I'm hesitant to do that up front, because I'm sure that we've all had the same experience where trying to pre-plan your Slack meeting room organization long before you have enough reason to fill up those rooms is fraught with peril. I don't want to fragment these calls before we know that it's time to fragment them. But some time in the future, we might be splitting these [inaudible 00:01:37]. In the meantime, though, it's going to be action packed. And these calls are going to have a lot more variety than we had in the future, which might be nice. It might be overwhelming. We'll see how it goes. |
| # / 00:01:47 | Richard Brown | I am going to talk a bit about governance. LongForWisdom is going to tell us a bit about what's been happening on the ground in governance, which is the forums for us. All governance happens in the forums, lots of activity in there. We're going to hear from Nik, our head of back-end services. I believe that's correct, Nik. Is that your title? |
| # / 00:02:11 | Niklas Kunkel | Yeah, that's correct. |
| # / 00:02:12 | Richard Brown | Okay, cool. I pulled it out of a spreadsheet. I wasn't sure it's entirely up to date. So Nik, our head of back-end services, is going to tell us about how we're going to deal with oracles in the future in sort of open-sourcy, governance world that we live in now. Looking forward to that. We're going to hear from Vishesh. He's going to tell us about the state of the peg. And then depending on interest and time, we'll stick around for a half hour or so and just answer some general Q&A or do some general Q&A. |
| # / 00:02:41 | Richard Brown | One of the things that I've been falling down on, ironically ... I don't know if I'm using that correctly, but this is a scientific governance and risk call. Very rarely do we actually talk about the polls and the results in the polls in governance, and we should probably start doing that. Recently, the community has signaled its intention and actually acted on its intention to lower the stability fee, which is a very interesting and encouraging behavior that Maker holders are not viciously self-interested people that are trying to scrape every last fee of the system. We're actually all enormously interested in maintaining stability of the peg. And if that means the stability fee needs to go down, it needs to go down. And the community jumped on that on our last vote. |
| # / 00:03:33 | Richard Brown | There's another stability fee poll that ended six minutes ago, which signaled the community's intention to lower another 2%, which is very interesting because ... I'm not a financial analyst. When it comes to crypto, I barely know what I'm doing, at least when it comes to watching charts. But it seems to have been presient because the peg was over a dollar and got even more of a deviation there in the last couple days. And who knows why? Maybe Vishesh will tell us. And so I won't steal his thunder and he can explain all of that to us. |
| # / 00:04:11 | Richard Brown | We also have two interesting polls in the system right now. These are ... I sort of dramatically refer to them as watershed moments in the history of MakerDAO governance. Maybe that was a bit too much. But we are on a long road to decentralization. And the way that that works is we need to bootstrap systems and positions, rolls, responsibilities, processes in the ecosystem, and fund those positions and rolls and processes and the creation of those things. And then eventually, hand these things off to the community. The first role that we had with that model were the risk teams, obviously. The second role with that model is the governance facilitator, which is essentially what I've been doing in these calls for the last 51 meetings. And the next one up to bat is the oracle teams. We need to expose some levers there and some responsibilities to the governance ecosystem for eventual decentralization. |
| # / 00:05:15 | Richard Brown | Each one of these roles are prototypes that we're using to bootstrap the DAO ecosystem. Over time, once these roles have been nailed down and processes are in place, actors from the ecosystem hopefully will arrive and they will begin to assume some of these responsibilities. For example, the governance facilitator essentially is me. In a very near future ... I shouldn't say very. Who knows what that actually even looks like. But post MCD, the next six to twelve months, who knows? It would be absolutely stellar if there was two or three other governance facilitators that could work with me and more governance facilitators, and then I can back away from the system completely and then the community could take over. There's all kind of information about that in the governance facilitator mandate in the forums. Oh, I'm sure everybody in this call are reading through those forums daily, so I'm sure I don't need to remind you. But if you're interested in the mandate, please have a look at the forum threads. |
| # / 00:06:24 | Richard Brown | As far as the poll goes, the community has signaled overwhelming support for these roles, which is nice. Cyrus' risk team mandate has 100% support from the community from 29 voters. I am not doing as well, unfortunately. I'm a 99.something% because somebody, and whoever this is I absolutely adore them, somebody spent $20 in Maker to vote no for my mandate. And whoever you are, I like your style. I think that might be it for what I wanted to talk about today. LongForWisdom, I'm going to hand it off to you and you can give us a recap of what's been happening in the governance ecosystem. |
| # / 00:07:04 | LongForWisdom | Yeah, absolutely. |
| # / 00:07:04 | Richard Brown | Thanks, man. |
| # / 00:07:05 | LongForWisdom | You can hear me all right, yeah? |
| # / 00:07:12 | Richard Brown | Yeah. |
| # / 00:07:14 | LongForWisdom | Cool. So we've got a couple of fairly big discussion threads pop up this week, specifically around vote delegation and what looks like the soon introduction of MKR as Compound collateral. So these two threads specifically, I've had quite a lot of discussion on. I've heard a lot of good points back on forth on the vote delegation. A lot of people are obviously very pro, but we had a couple of people coming out with out against it, as well, which is good. So that was [inaudible 00:07:51]. |
| # / 00:07:52 | Richard Brown | What's the general sentiment about the MKR Compound? Staking thing |
| # / 00:07:58 | LongForWisdom | So I don't think there's any sort of big consensus about it. Some people are worried. Some people are sort of claiming that it's not such a big deal as it appears. I haven't really thought about it to decide. |
| # / 00:08:13 | Richard Brown | It's an interesting conversation that the community needs to have, though. And there's an overlap here with delegation and Compound in that the DeFi ecosystem or crypto space, in general, is different than the traditional financial world, which may come as a surprise to people in this group. The thing that I find most interesting, though, is that in the traditional world, there's checks and balances and laws and regulations and rules and all kinds of guardrails and sandboxes and velvet ropes. In the tech world, if you've build something and put it on the internet, you're expected to defend that thing against the ecosystem that you've released it into. There's this interesting overlap with how DeFi, in the crypto space, is emerging. And it's coming from a technical perspective as opposed to a traditional finance perspective. And there's the assumption in the IT world that when you put something on the internet, it's your responsibility ... you assume that there's bad actors that will attempt to manipulate it and it's your responsibility to make that system robust enough to withstand those attacks. |
| # / 00:09:27 | Richard Brown | And we're seeing some of that dichotomy come to the floor with the discussions around delegation and the lending and borrowing of a governance token. And I think that these are interesting conversations that need to be had. We can't stop anybody else in the ecosystem from doing whatever they wish to do, because that's what it's all about. But it'll be our responsibility to figure out how we deal with those things and that means, do we attempt to block activities or features that we don't like? Or do we attempt to get ahead of those features and activities? Do we attempt to mitigate those damages? There's lots of different ways to approach these things and that's part of the same conversation we've had with the delegation thing and with the [inaudible 00:10:14] thing. Some people came up with ideas of how do we stop, or how do we make it hard for both delegation and lending and borrowing of governance? I think that it will be very cool to see how that shakes up. I'm hoping that we decide that we need to make our systems more robust in order to support the rest of the ecosystem that's popping up around us. |
| # / 00:10:39 | LongForWisdom | Yeah, it's definitely going to be interesting to see how the Compound thing pans out, for sure. |
| # / 00:10:46 | Richard Brown | Anything else of note in governance this week? |
| # / 00:10:49 | LongForWisdom | So I've been pushing the signal request to [inaudible 00:10:52] rate stepping. Again, I've put what is hopefully going to be the final forum poll up, which is a discussion of what the onchain poll should look like or what form that should take. So we've had [inaudible 00:11:04] on that, contrary to the previous polls, sets a closed date for this one. That was seven days from when I posted it, [inaudible 00:11:10]. So I suggest to people to [inaudible 00:11:14] if they're interested. |
| # / 00:11:18 | LongForWisdom | The other thing that came up, which is kind of interesting, is Chris Blec asked us about whether our goal is to attract more individual voters to vote or that our goal is to attract a greater amount of MKR token, which is [inaudible 00:11:36]. Some contention the poll [inaudible 00:11:41] is around 50/50. |
| # / 00:11:45 | Chris Blec | I can jump in real quick on that. So that was sort of spurred because I'm working on some video content. Our overarching goal is to bring more people in and to obviously feel better about the consensus that we're getting. It really occurred to me that it's like a fork in the road a little bit, as far as strategy goes, about whether we're focusing on the total amount of Maker versus the total amount of people. And I think we should continue that conversation, but I also think we should note that it's fine if we say both strategies are important. The thing is, they're two diverging sets of tactics. So the same tactics wouldn't necessarily apply to getting the top 50 whales versus to getting a thousand Maker holders, period. And that's really what I was trying to get at there. So if we do continue in sort of this deadlock fashion, I think that's actually great because it means that we realize that both of those sides are important and we just need different sets of tactics for each side, which I'm going to continue to work on as far as video content and social media goes. |
| # / 00:12:48 | Richard Brown | Yeah, I agree completely. I think that we have to be cautious about thinking and [inaudible 00:12:54] that we're living in a binary world. The way that the system is designed is to get people as much skin in the game, incentive to manage the system in order to protect them. So with that model in mind, we should be trying to get more whales into the system, but we're up against human nature here. It's actually human nature, but everybody wants underdog to win or they're looking at the voting system and they say, "Voting equals representative democracy," and so there's this dichotomy that people need to work through. But you're right. Personally, I like to see both. I like to see thousands of curls and minnows and ... I can't remember what the scale is ... to attempt to offset the whales in the system to spice things up a bit, I guess. Get a bit of variety in there. But it also shows us an interesting metric that I'm super keen on, which is do we see different voting behaviors based on the amounts of Maker that people possess. So are large holders voting differently than small holders? And that's something that we should understand if that's the case. |
| # / 00:13:59 | LongForWisdom | Yeah, all true. So I think that's the extent of what's been going on, or a large portion of it. |
| # / 00:14:14 | Richard Brown | All right, cool. There's a lot of really interesting things happening. I keep on saying this, but it's fine. The forum has been a fantastic success. So successful that I don't have time to keep up with all of the successes that are in there. But through the course of the call, if you have a few minutes of idle browsing available to you, take a look at that forum post that LFW just put in the chat. See if there's anything in there that you want some clarity on. At the end of the hour, we can spend some time walking through some of these outstanding questions, because there's a lot of them. |
| # / 00:14:53 | Richard Brown | I am going to shift gears now and introduce everybody to what I referred to earlier as the new segment, the oracle segment of the call. So Nik, I'm going to hand it over to you. Do you want to introduce yourself, let everybody know what you're up to, and then take us on a little tour of the oracle system? |
| # / 00:15:11 | Niklas Kunkel | Sure. My name is Nikolas Kunkel. I work at the foundation and my official title is the Head of Back-End Services. And I've been at Maker for a little under three years now, so quite some time. And actually, I've come a long way. In that time, I've worked on [inaudible 00:15:42]. I've worked on [inaudible 00:15:45]. I've worked on [inaudible 00:15:45] contracts. I've worked on [inaudible 00:15:48] API. And I've worked on oracles. And so hopefully, many of you will have seen the oracle's blog post that we published a few days back. But for those of you who haven't, I'm going to post the link in the chat. And so if you have some time at some point, you should definitely go check that out because it's going to be a very good supplement to what I'm going to talk about today. |
| # / 00:16:27 | Niklas Kunkel | So we have a limited amount of time today and a lot, and I mean a lot, to talk about. So some of this is going to be a little bit overwhelming in terms of the volume of information and in terms of the pace at which we're addressing at. I wouldn't worry too much about that. This is more supposed to be an introduction, but you can't really know what you don't know. So I'm kind of here today to introduce you to this whole bigger world. And then next week, I can post the actual proposals to the forum and then we can have a discussion in the forum about the actual proposals and the content inside the proposals themselves. |
| # / 00:17:25 | Niklas Kunkel | So I've prepared a little presentation here. So I'm going to switch gears real quick. Can everyone see my screen? |
| # / 00:17:39 | David Utrobin | Yes. |
| # / 00:17:40 | Richard Brown | Yeah, it looks good. |
| # / 00:17:41 | Niklas Kunkel | Cool. All right. Let's get this started here. All right. The magical world of oracles. So for those of you who don't know, oracles are the way in which the DAI system is able to get prices of the collateral assets. So why is this useful? We need to know if positions are sufficiently collateralized. And if they're not, then they can be liquidated. We need to know currently in the system, when we liquidate collateral, we liquidate it at a 3% discount to the spot price. The spot price has to come from somewhere and it comes from these oracles. |
| # / 00:18:42 | Niklas Kunkel | And so up until now, the oracles have been decentralized, in a sense, but they've been mostly managed and administrated by us. We kind of think that it's time that that's changed. Governance has definitely come a long way over the past year or so. You guys have shown me successfully that you've been able to hold the peg and kind of had this drive to organize yourselves. And it seems to be working really well. So it's time to throw some more stuff at you. And so part of making the oracles more decentralized is to make sure that, instead of us being kind of in the driver's seat and making all the decisions, that it's actually you guys who make all the decisions. So this is really about handing over the keys to you guys and just making you aware of what these new responsibilities are and what you're in for. |
| # / 00:19:54 | Niklas Kunkel | So by that measure, I want to introduce the first proposal, which is an oracle governance framework. So it's similar to the general governance framework in that it defines these set of processes for how governance should run. Well, the oracle governance framework is this blueprint of the responsibilities, with respect to oracles, that have been placed on governance and kind of the processes around how we go to fill those responsibilities. |
| # / 00:20:35 | Niklas Kunkel | So what are these? Well, first of all, I want to say that the oracle governance framework is a living document. These processes not only can be changed, but they will be changed. We're going to see what works, what doesn't work. We're going to realize that there's entire new areas of responsibility that we didn't even think about, that we need to address at some point. So this is supposed to be a starting point. We need to start somewhere. So there's going to be more responsibilities and more fine-grained stuff in the actual proposal when I publish it to the forum next week, but I just want to take a high-level overview of it real quick. |
| # / 00:21:28 | Niklas Kunkel | One of the main jobs for governance, with respect to oracles, is going to be voting in feeds and voting in oracles. Now that terminology might seem a bit foreign to some of you. So what is the difference between a feed and an oracle? Well, feed is an individual who is running a bot that is pushing price updates for an asset, versus the oracle is the actual smart contract onchain that other smart contracts can query for the price. So when you basically group together a set of feeds, the onchain price that you see in the oracle is actually the median of all of those feed inputs. That means that we're placing a lot of trust in feeds to be good actors and not ... They have the ability to launch an oracle attack, which would be catastrophic. I'm sure I don't have to tell you. So we need to have strict criteria for defining what makes a good feed. Who would be a good feed? |
| # / 00:22:58 | Niklas Kunkel | The same for an oracle, right? What makes a good oracle? And what I mean by what makes a good oracle, if an oracle is just a smart contract, well, an oracle is specific to one asset, right? So ETH, we have an ETH oracle right now. We have an MKR oracle right now. When we need new assets for MCD, let's say ... just throwing it out there. Let's say REP becomes a token, a collateral in MCD. We're going to need an oracle for REP and we're going to need to define the rule set around REP, right? What exchanges are we querying? What asset pairs from those exchanges are we querying? How do we kind of filter that data to then have a ... what is the rule set around how we determine this price? |
| # / 00:23:56 | Niklas Kunkel | So the oracle governance framework is not just going to define these criteria, but it's also then going to contain the actual governance processes for voting in a feed or even voting out of feed, right? If it's not living up to the performance metrics that we should come to expect. Same with kind of an oracle. What are the processes for creating a new oracle? Who deploys this stuff? Who ensures that what is deployed is correct and is ready to be integrated? How do we then create a vote, to vote in that oracle infrastructure? |
| # / 00:24:43 | Niklas Kunkel | The oracles themselves have a bunch of parameters. So for example, one of them is the sensitivity parameter. And this is basically, what percent spread do you want to have before you update the oracle? If you, say, have a 2% sensitivity, then if the price changes 1%, then the oracle price will not change. It's going to wait for that threshold. The lower you have the sensitivity, the more, quote-on-quote, "sensitive," your oracle will be to prices changes. However, it also gets very expensive because you're doing updates more often, more often, more often, more often and each of those are consuming gas costs, it can add up very quick, especially when the network is congested and all of a sudden you have 60 Gwei transactions. |
| # / 00:25:48 | Niklas Kunkel | Another angle here is expiration. So the price expires after a certain set of time, right? Another one we have is the delay. So in MCD, the oracle value is fed into a smart contract called the oracle security module. And the oracle security module is there to protect MCD from oracle attacks. So it acts as kind of the buffer in between the oracle and the MCD system. And the way that it acts as this buffer is that it has a delay. So let's say the delay is one hour, right? If there's a new oracle price pushed, then the MCD system waits one hour before it actually uses that price and that gives it time to react to an oracle attack. |
| # / 00:26:57 | Niklas Kunkel | Another thing we're going to have here is the quorum. So if you have a certain number of feeds, how many feeds together is enough to achieve quorum, is enough to achieve a price change? Is it one feed? It is ten feeds? Is it 50 feeds? Is it all of the feeds? Is it 70% of the feeds? Is it 51% of the feeds? There are other parameters as well, but these are all kind of the main parameters that governance is going to be responsible for. It's going to have to think real hard about. |
| # / 00:27:38 | Niklas Kunkel | Okay. So that's a lot, but for the sake of time, let's keep moving on here. So the second thing I want to talk about is that you're not alone. So before we had this governance handover of oracles, I said that the foundation had been doing all of these things behind the scenes. So it's been myself, maybe you guys have heard of Mariano, Mariano Conti, and a handful of other people who have been really instrumental in making sure that this all runs to perfection and the house doesn't burn down. And so the oracle team mandate is a way to formalize this role. So just like how we have risk teams that governance can delegate certain responsibilities to, we can have oracle teams that governance can delegate responsibilities, with respect to oracles. |
| # / 00:28:51 | Niklas Kunkel | So it's not just an oracle team. It's oracle team [inaudible 00:28:57], because all of these responsibilities can be split as granular as governance wants them. Maybe particular individuals or groups have specialties that make them more suited to certain responsibilities than others. But essentially, what is the oracle team going to do? Well, there's this delegated party, by you guys, to facilitate the administration and the technical development of the oracles. So what do I mean by administration and what do I mean by technical development? So let's start with the administration. |
| # / 00:29:42 | Niklas Kunkel | The administration, there is a lot of stuff that goes on behind the scenes. Feeds will crash and they need to be notified that they need to bring the feed back up. We release new updates to the bots that the feeds are running all the time, and the feeds need to be notified to upgrade. We release bug fixes for bugs that we find, occasionally. If there is a new collateral asset that is added, the tools need to be updated to support that new asset and then that update that needs to be pushed out to all the different feeds. If we wanted to have a new asset added, we need to deploy the oracle smart contracts. So that is kind of infrastructure deployment and then someone needs to verify that running a bunch of tests to make sure that not only are the smart contracts correct as published, but in the way that they're configured and deployed, that everything is correct there as well. So there's a lot of work that goes on here. |
| # / 00:31:06 | Niklas Kunkel | On the technical development side, all of these tools need to be programmed. So I mentioned earlier, what if we added REP, right? Well, in code, in these tools, you're going to need to define, well, okay, what is this REP token? What exchanges should it get the prices from? What pairs on those exchanges should it grab the price from? Once it has this set of prices, what is the filtering, modeling mechanism that it uses to arrive at a final price? So these are all questions that we have to think about, whether we want to have a generalized framework for every asset to follow or whether we kind of define subclasses and say, well, these types of tokens should be handled this way and these types of tokens should be handled this way, right? So maybe something that's low volume that has very low liquidity needs to be handled very differently from something that has very high liquidity. |
| # / 00:32:20 | Niklas Kunkel | So Rich, I can't actually see the chat. So if you see anything that you think is important to kind of put an attention on real quick, feel free to chime in. Okay? |
| # / 00:32:34 | Richard Brown | There should be a pile of questions. I've seen one question, one comment. Do you want me to wait until the end or just interrupt you? |
| # / 00:32:42 | Niklas Kunkel | I'm going to leave that up to you. I have a lot of stuff to go through, but if I'm obviously missing something, then ... |
| # / 00:32:51 | Richard Brown | All right. Well, let's consider yourself interrupted at this point. So let me ask one question from Johann. And also, I want to remind people, if you have access to a microphone, please raise your hand. There's functionality to do that in the UI here. Or just jump in and ask a question. But Johann asks, "Have any details been released on how new oracles can be added to the feed?" |
| # / 00:33:17 | Niklas Kunkel | Okay. So I think it's actually the other way around. The feeds are the ones that are feeding the prices into the oracles. So to add a new feed, we have not published anything as of yet, but we will publish something next week. Inside of the oracle governance framework is going to be a section defining the criteria for what it takes to be a feed. And then how we vote that in. And the same goes, actually, for oracles, like your original question stated. If we want to pursue an oracle for some token, how do we do that and what are the requirements the criteria first said that we should take a look at? |
| # / 00:34:10 | Richard Brown | Johann, do you want to do a follow-up question you wanted to ask? |
| # / 00:34:13 | Johann Eid | Yeah, no. Actually, it answers my question. So basically, I'm from Chainlink and currently, we have a ETH who has the reference contracts, which has 14 feeds you could call them, which are feeding ETH, who has ETH price and our incentivize with LINK token. And basically, this oracle would be a good add, I think, to what you guys are building. So it was just to get this information. So I can wait until next week for more details. |
| # / 00:34:43 | Niklas Kunkel | Okay, let's talk more once the info is all public. Okay? |
| # / 00:34:46 | Johann Eid | All right, awesome. Thank you. |
| # / 00:34:48 | Niklas Kunkel | Thank you. |
| # / 00:34:51 | Richard Brown | The next natural question to ask is incentivization mechanisms. Is that something that your team, Nik, is prepared to talk about yet? Or is that going to be in the future? |
| # / 00:35:01 | Niklas Kunkel | So it's going to be talked about later in this presentation. |
| # / 00:35:06 | Richard Brown | I won't spoil it, then. |
| # / 00:35:08 | Niklas Kunkel | Okay. |
| # / 00:35:08 | Richard Brown | I think that's it for the questions that I can see so far. Continue on. |
| # / 00:35:12 | Niklas Kunkel | Okay. So just a couple quick last notes on the oracle team before we move on. The oracle team itself does not have special privileges to enact any changes. It's not an executive of government. |
| # / 00:35:34 | Richard Brown | Actually, let me interrupt you then, because that's a big one, I think. This is a large conversation that we're having. When we introduced these mandates, the community ... there's two different things happening. We're exposing levers and we're allowing the community to begin the process of taking over responsibility for a lot of these activities. But the opposite end of that coin is that we're, as a community, empowering a certain group of people to make decisions that they are best qualified to make and do things that we need them to do in order to maintain the health of the protocol. |
| # / 00:36:11 | Richard Brown | So there are going to be situations or a group of responsibilities that the oracle teams just ... that's part of their day-to-day gig. And so if something needs to happen, they're going to go ahead and do it, like reacting to emergencies or adding new feeds or ejecting somebody immediately, if they need to. |
| # / 00:36:29 | Niklas Kunkel | So I agree with Rich. I just think that those things should be doled out with the consent of governance, right? |
| # / 00:36:40 | Richard Brown | Oh, yeah. I agree completely. |
| # / 00:36:42 | Niklas Kunkel | On the base layer, you have this oracle team mandate, which all oracle teams follow. And there is a minimum level of executive privilege given to those oracle teams. And then for specific oracle teams, certain privileges can be granted by governance. |
| # / 00:36:59 | Richard Brown | Yeah, I agree completely. I think the area of attack, though, in oracles is significant. And governance is expensive in time. So for governance to do something, you're looking at basically a two-week minimum turnaround time from this is a problem, this is the debate, this is the vote and best case scenario. So there are going to be situations where the oracles need to be more agile than that. |
| # / 00:37:24 | Niklas Kunkel | Mm-hmm (affirmative). I agree. So in the spirit of gradual decentralization, there are going to be situations where the oracle team is going to have to take more of a hands-on role. I just want to make sure that the mandate doesn't come out of the box with this broad executive privilege. |
| # / 00:37:51 | Niklas Kunkel | So the job of the oracle team is really to make proposals and then kind of give context and guidance around those proposals and why they were drafted in a certain way and give this perspective. And then governance works on those proposals. So really, the job of an oracle team, besides the administration and the technical development, is really to actually as this governance liaison to kind of be the bridge between everything in the oracle world and governance. |
| # / 00:38:28 | Niklas Kunkel | So one of those main things that an oracle team does is they're going to have to publish a roadmap of what they're interested in doing. And then that roadmap needs to be ratified by governance and they need to then give regular updates on the progress of that roadmap. So be that, being on the governance call and giving an update or posting on the forum and giving an update on the progress on the roadmap. And the oracle team will then be evaluated based on how they do against the roadmap, how it's ratified. |
| # / 00:39:10 | Richard Brown | All right, Nik. The questions are coming in now. Chris, I know you do have a mic. Do you want to ask your question? |
| # / 00:39:18 | Chris Blec | Sure. Yeah, I was just wondering if the new model requires trust of any sort in the partners. Because as I was reading through the blog post, I noticed we talked about how individuals had to stay pseudonymous pretty much if they were running feeds to avoid bad things happening to them. And so what's to say that these other projects like dYdX and Set couldn't face the same types of situations? |
| # / 00:39:49 | Niklas Kunkel | Okay. So actually, feeds was the next thing I'm going to talk about. And one of the first things I'm going to talk about is why individuals in organizations kind of have different game theory play out. And so why one is desirable versus the other, and especially in terms of when it comes to identity. So just to kind of go back a little bit, the feeds are the ones that push the asset prices. Currently, we have 15 of them. 14 of those are for the ETH/USD oracle. |
| # / 00:40:33 | Niklas Kunkel | Let me see if I can ... okay. If you go to MakerDAO.com/feeds, you'll see a very, very sexy dashboard that shows what the oracle price is and it'll show the address and price and timestamp of every single feed that went into constructing that price. So the oracle price that we have right now, the model is that it is the median of all of the feeds. Feeds are compensated. They're paid monthly. And currently, the foundation is footing the bill for that. This is the first time that we're kind of publicly talking about this, but the feeds have been paid about a thousand DAI per month and that's to kind of cover their expenses in terms of infrastructure and to cover their time for promptly addressing any kind of situations that pop up or promptly installing updates and upgrades. |
| # / 00:41:47 | Niklas Kunkel | But eventually, it's not sustainable that this comes from the foundation. It's actually not very decentralized either, if you think about it. And so eventually, the feeds should be compensated from the stability fees. And when that transition happens is up to governance. And along with that, governance should be deciding, how much should feeds get paid? They're providing this valuable service but, apart from that too, there's the risk of bad behavior. So the compensation, in a sense, also has to match the risk of or what is the incentive to be a bad actor, right? But that's a whole can of worms that I don't want to go into right now. |
| # / 00:42:49 | Niklas Kunkel | So kind of getting to your question ... so up until now, all the feeds that we have are individuals and they're pseudonymous, which is essentially saying that we don't know who they are. There are two individuals within the foundation who actually know the identities of all of the feeds. But as more feeds get added, they will not be made aware of who those feeds are. And so even that will become diluted, eventually, to the point where no one in the world actually knows all the feeds or the majority of the feeds. |
| # / 00:43:33 | Niklas Kunkel | So why are these individuals ... why do they have to kind of stay anonymous? The big part is what I call the coordination problem. If everyone's identity is public, then you know exactly who you need to kind of bribe, who you need to extort, and you make them a giant target. So this presents a lot of risk to the individual and a risk to the system, as well. If these people are way too easy to communicate and start to group up, that presents a giant risk. |
| # / 00:44:23 | Niklas Kunkel | So in the blog post, I mentioned [Raul 00:44:26]. Here is an alternative, right? Here are these organizations, these groups of people. And I made the case that these groups of people should be public. And I think there is a lot of good reasons why. I think groups of people are much harder to coerce, extort, and bribe. They're much more resistant. And another thing that comes into play here is their reputation. So you have this idea of reputation of individuals versus reputation of organizations. |
| # / 00:45:10 | Niklas Kunkel | So let's take Mariano, for example. Mariano has been very public that he is a feed. I think he's the only person who has been public about it. If Mariano decides to launch an oracle attack one day, which is that his feed just starts pushing BS prices, Mariano is just some dude in Argentina who's going to be laughing at a bunch of people on the internet being mad at him. When you have something, say 0x, which is an organization based out of San Francisco with a reputation that they need to protect, that they value, because their livelihood depends on their product being successful, their reputation is worth a lot more to them than Mariano's reputation is worth to him. So for an organization to run a feed, it's actually more advantageous to force that organization to be public exactly because they start to put their reputation at stake and their reputation has real value. |
| # / 00:46:27 | Richard Brown | Can I come in for a second with a question? With the public organizations, that also opens up the door to the traditional legal system, right? So is there an expectation that there'll be contracts that are going to be signed and SOAs? |
| # / 00:46:43 | Niklas Kunkel | Yeah. That's a good point, Rich. so there's an official agreement with them for a service that they provide. And when an organization has more surface area, let's say in terms of legal stuff, than an individual in some random part of the world. |
| # / 00:47:04 | Richard Brown | I suppose you can also impose some rules, too. Like, they need to be bonded or notarized or ... |
| # / 00:47:13 | Niklas Kunkel | There's a lot more flexibility to work with here that you can use to reduce risk, legal surface area being one of them. So another way that reputation factors in is trust by users. So one of the biggest complaints about our oracles has been, "Oh, well, there's this group of 14 or 15 Maker insiders," or whatever, "and they could kill the system at any time they wanted. Why won't they tell us who they are?" To an extent, I understand that criticism but, because of the reasons I said earlier for why they have to stay anonymous, that's basically my response. But since organizations have reputation, their users will inherently trust our oracles more if they know that 0x is running one of the feeds, if they know the dYdX is running one of the feeds, if they know that Dharma is running one of the feeds, if they know that Chainlink is running one of the feeds. So there is an argument to be made here that reputation is hugely beneficial, a public reputation is hugely beneficial when it comes to organizations. Does that answer your question? |
| # / 00:48:41 | Richard Brown | If you mean me, yes. I think so. I don't know if there was other questions. |
| # / 00:48:50 | Niklas Kunkel | [inaudible 00:48:50] question earlier in regards to why adding organizations as feeds was a good idea. |
| # / 00:49:00 | LongForWisdom | I had a different question, if I could chime in. |
| # / 00:49:03 | Richard Brown | Sure. |
| # / 00:49:06 | LongForWisdom | Just on the point of, even if you assign a company or an organization to run these feeds, there are still going to be individuals within that organization that can control them if they are bribed or extorted or whatever. Are there any sort of guidelines you're providing to those organizations to handle this in a safe way? Or is there any sort of thought on that, I guess? |
| # / 00:49:28 | Niklas Kunkel | Yeah. So we'll essentially be handing guidance on how that needs to happen. One of them, I think the obvious one here being kind of a multisig. So one person does not have the executive power to just control the feed. It needs to be a set of individuals within that organization who all need to sign any kind of ... as you have turn in an organization and people leave, how you go replacing the signatories on that multisig. So, yeah. We will be providing guidance, especially on, for example, how do multisig keys need to be stored, et cetera, et cetera, et cetera. |
| # / 00:50:16 | Chris Blec | One more question on this. So one of the benefits of pseudonymous individuals is that the individual feed owners don't know who each other are either, right? So in this new world, the public organizations are going to know who each other are and will have opportunities to all get together in a room and have a meeting about their price feeds, right? So is there any concern here or anything that addresses the possibility of any sort of collusion between the different organizations, with regard to prices? |
| # / 00:50:54 | Niklas Kunkel | Sure. So I think that's a valid concern and it's up to the governance and the oracle team to be aware of that. So one of the things you can look at is, what is the quorum for an oracle attack to even happen? Is it 51% of the feeds? Is it 70%? Is it 20%? And what is the proportion of public feeds you have to anonymous feeds. And so you can't ... |
| # / 00:51:26 | Richard Brown | [crosstalk 00:51:26] thinking about with that weighting looks right now? Are you going to try and emphasize public over private or private over public? |
| # / 00:51:34 | Niklas Kunkel | At the moment, I would say I would like to add more public organizations, but there's a lot of variables at play here, right? I'll get into this a little bit more in a second, but the entire oracle stack is essentially going to turn into a product. And you want to rope people and you want to rope organizations and you want to rope entire ecosystems into our oracle ecosystem. It provides a certain kind of stickiness to it. And so there's all these different factors that all kind of pull and tug on each other in different directions that you have to juggle here, right? So what is an appropriate level of risk to the system and what is an appropriate level of optimization for a better product, optimization for better access? |
| # / 00:52:44 | Niklas Kunkel | So let me kind of take this a step further here. So for those of you who didn't read the blog post, the first wave of feed candidates that we have from organizations are dYdX, 0x, Set and Gnosis. These guys have been working with us closely over the past few months in terms of this greater vision we had for the feeds to have these public-facing entities. And they are all already very integrated into our oracle ecosystem in terms of oracle v1, in terms of using our prices. Set protocol has actually been one of our earliest beta partners. They've been using our BTC oracle, which is running on the oracle's V2 stack since back in February, I believe. So that is a long time and they helped us iron out a lot of kinks in the system so it'd be ready to go for Multi-Collateral DAI. dYdX has been using our oracle's V1 from day one. 0X has been using us, as well, for some eternal stuff they've been working on. And Gnosis has been using us for their DutchX auction platform. |
| # / 00:54:22 | Niklas Kunkel | So this is as much about bringing in already existing stakeholders in the ecosystem, right? Stakeholders who are using the oracles and therefore have an incentive to, again, be a good feed, in that respect. The more connected you can make the oracle ecosystem, if you can have someone be both an oracle user as well as a feed where launching or participating in an attack actually makes their own system vulnerable to being compromised, the more you can generate those connections, the more you actually lower the overall risk. |
| # / 00:55:10 | Niklas Kunkel | Okay. So this one is a little bit controversial, I think, from what I've seen in response to the community. But I think if you think about it, it's going to make a lot of sense. So the last proposal I want to address is the oracle incentives restructuring. And what this is all about is basically creating a balanced model of aligned incentives for all the different stakeholders in the oracle ecosystem to make the oracle infrastructure self-sustainable and more decentralized. |
| # / 00:55:49 | Niklas Kunkel | So the key feature here that people will probably notice is that, in oracle's V2, our oracles will be whitelisted. And what this means is that each oracle smart contract will have a whitelist that regulates who is allowed to read the prices from that oracle. So if you want to use, say, the ETH/USD oracle, you need to be on the oracle whitelist. And so this is a tool for governance and for MKR holders, in general, to effectively monetize the oracle infrastructure. Because I think what you're going to find is that when eventually this transition of, rather than the development fund paying feeds and it going to stability fees instead, then as we up the numbers of feeds, as we up the numbers of oracles, you're going to find that the costs get pretty, pretty high. And having this tool in our tool belts, should we want to use it to recoup some of those costs, I think can be very beneficial. |
| # / 00:57:11 | Niklas Kunkel | One thing to note here is that maybe the whitelist isn't used for monetization right away. I can be also used for quid pro-quo business development purposes. So maybe there is a DEX out there that wants to do some kind of stuff using margin and they need oracle prices. And then they come to us and they're like, "You know what? We'll list all of our tokens against DAI if you give us oracle prices." So that can be another way for governance to then leverage, say, the utility of what their oracles provide and actually get something out of it, as well. So there are ways to extract value from this that don't necessarily have to be money. |
| # / 00:58:08 | Niklas Kunkel | All of these bullet points are actually from a different slide, my bad. The whitelist does not strengthen the peg or increase the utility of DAI. My bad. So one thing that I want to talk about is, how do we create an incentive, this kind of incentive system? And how would that actually empower us to do different things? So when I say that people on the whitelist may have to pay to read those prices, who are they paying? Well, they're paying the MKR holders because it's all about having this virtual alignment of all the different incentives. So if MKR holders are figuratively paying feeds who are pushing prices to oracles which are providing services for customers and the customers are then paying MKR holders, you have this kind of virtual circle. |
| # / 00:59:13 | Niklas Kunkel | And because you have this, you have something really cool. So up until now, because of the way incentives have been set up, there was never any incentive to create an oracle for an asset that was not a collateral type in Multi-Collateral DAI. But now, since access to whitelists on oracles will generate revenue, there's actually a direct incentive for MKR governance to look at oracles for assets that are not collateral types. |
| # / 00:59:52 | Niklas Kunkel | So here's kind of what I was talking about earlier. You have MKR holders who are paying feeds. So MKR holders would say, "Hey, we want your feeds to start running an oracle for, " god forbid, "XRP or TRON or Dentacoin or ..." |
| # / 01:00:10 | Richard Brown | This is all going on your permanent record there, Nik. So be careful. |
| # / 01:00:15 | Niklas Kunkel | Right? Shitcoin XYZ. Those feeds are then providing that service. The data consumers are consuming from the smart contract. Often, the oracle smart contract and for the privilege of being on that whitelist, they are [inaudible 01:00:32] MKR holders. And by the way, the way that these payments would work is it would work similar to how the buy and burn currently works. So these payments would essentially go into the stability fee buffer where they then just go through the regular pipeline of buy and burn. So effectively, data consumers would be burning MKR and that's how MKR holders see a benefit. |
| # / 01:01:02 | Richard Brown | Again, Nik, before you jump into the next slide, we're at the top of the hour. So we need to wrap this up soon. Is there still a lot to do? |
| # / 01:01:09 | Niklas Kunkel | This is actually my last slide. |
| # / 01:01:11 | Richard Brown | Oh, okay. Perfect. |
| # / 01:01:12 | Niklas Kunkel | Okay. So proposal 4.5. this is kind of a continuation of what we just talked about in terms of the whitelist. The smart contracts out of the box will come with a whitelist, but I don't actually think governance should use the whitelist right away. And there is a lot of reasons for this. The first is that our oracles, right now, we're one of the first in the Ethereum ecosystem. And they were organically adopted in a lot of different places. So much so that we don't even know all of the different infrastructure that is dependent on our oracles. We could possibly cause a chain reaction. |
| # / 01:02:10 | Niklas Kunkel | So one of the reasons for the whitelist, to begin with, is the fact that we can actually map out who is using our oracles and see these kind of chains of risk building up and being aware of them. So seeing who is using them and how are they being used. So flipping the switch and all of a sudden saying, "Nope, you don't get oracles anymore, there's a whitelist now," that could potentially have catastrophic effects. So I would exercise caution and I would like to, before this proposal, that anyone who wants to join the whitelist for any oracle, current and future, can get access for free for one year. |
| # / 01:03:06 | Niklas Kunkel | So this kind of makes sure that we don't pull the rug out from underneath anyone, we don't set off anyone scrambling to come up with some shitty oracle solution in a one-month span. This gives people time to either transition away from our oracles or gives them time to integrate our oracles and say, "Okay, this is a service that I think is worth paying for." There's a good reason to have it be free, though. That's not just having to do with being a good actor and kind of thinking of the ecosystem. We can kind of be greedy at the same time and say, "Well, we have this infrastructure and we want people to use it." So giving it away for free is a good way to bring people into our ecosystem and really bootstrap that ecosystem. |
| # / 01:04:10 | Niklas Kunkel | So someone who is building something and, right now, they're at the beginning and it's not making any money, but they need oracless well, they can say, "Okay, well I can use the Maker oracles." And then later, if their product actually turns into something that gains some traction and works, then they'll be a long-term customer. But then [inaudible 01:04:33]. They need that entry point to come into our ecosystem in the first place. And setting up that kind of whitelisted cost barrier is not the best way to bootstrap that. |
| # / 01:04:52 | Niklas Kunkel | So I think that's it for me. Like Rich said, I've already gone over my time. So I don't think we're going to have time for a lot of questions today, but like I said, I will be present on the forum. I'm going to be posting all of the actual proposals later next week and then we can all have some very in-depth discussion around feedback and adjustment to these proposals until the community feels comfortable that this is something that we can work with together. |
| # / 01:05:33 | Richard Brown | All right. Thanks for that, Nik. That was great. A lot to digest. Possibly, we might even get you back next week, too, if there's more questions that arise. We're at ten minutes after the hour. I still want to get into the state of the peg. So Vishesh, maybe you can pick it up here. And if we have time, we can all stick around afterwards. Maybe answer some questions that Nik or anybody else who has a question they'd like to ask. |
| # / 01:06:01 | Vishesh Choudhry | I'll try to go quick, with time consideration. Okay. So I'll just start off with the most interesting thing. So essentially, with the two relatively rapid drops in the stability fee, the borrow rates on Compound, dYdX on a weighted average basis ... I kind of tracked down with that. So essentially, at 20.5, which was sitting there for, say, over a month, that borrow rate started to come down a little bit on those platforms. With the upcoming drop to 18.5%, there's a little bit of an uptick in activity. And then the rates kind of tracked a little. And then again with the 16.5% drop, we have an uptick in activity. And then the rates, again, track. So this is kind of a pattern we've noticed when the stability fee moves, sort of how these borrow rates respond accordingly to basically maintain a somewhat healthy point or two-point buffer, which occasionaly drifts to, say, maybe three, three and a half when there's low utilization and the rate is very high. Or maybe a point when the rate is dropped and utilization is sort of tracking. |
| # / 01:07:24 | Vishesh Choudhry | So as far as the borrow volumes ... So during that same time period, borrow volume, if you see in the last week or two, has risen a bit more for dYdX and a very slight dip for Compound. So compound has been more or less steady with a very slight dip and then dYdX did experience a little bit of a bump with those rate drops. I think part of that has to do with the fact that there were slightly different use cases for dYdX versus Compound. And as those rates get cheaper, you may see people essentially attempting to engage in a bit more fancy footwork in terms of the derivative structures that they try to construct out of the borrowing [inaudible 01:08:08]. |
| # / 01:08:08 | Vishesh Choudhry | Hopping over to supply volumes, essentially that has been consistently rising over the past couple months. That was rising a bit more, again, for dYdX recently than for Compound. Compound basically came out one to two million and that tracked a little bit with the drop in the stability fee. It was actually a longer-term, steady increase. So again, this excess supply kind of went hand in hand with that although it was increasing a bit more dramatically. So essentially, there's been a bit lower utilisation on these platforms, recently. And so you can see that here. |
| # / 01:08:53 | Vishesh Choudhry | The utilization rates for Compound came down pretty significantly for when the rates were kind of close to touching the stability fee to now back around that sort of 70, 75% range which, in the longer term ... I wish I'd shown the longer-term scale here ... is a bit of a steady state. It seems, for Compound, it doesn't generally dip below this sort of 70% level. And dYdX, again, a little bit all over the place, a little bit more unpredictable, but still in that 70 to 80% range, which it has been pretty consistently. |
| # / 01:09:29 | Vishesh Choudhry | So at the end of the day, just a little bit more supply on Compound for basically similar borrowing volumes, which is pretty interesting considering that the stability fee rates have dropped down. So essentially, if you think about this as the stability fee being, quote-on-quote, "too high," and you think of the utilisation on compound as being not artificially propped up, but more of a short-term increase ... and again, we talked about some of the impact of InstaDapp refinancings and things like that. But as it came down to a little bit more of a steady level, then this just became a little bit more of a constant despite the drops in stability fee. So that just means the rates kind of track with that. Demand levels didn't move very much and neither did supply levels, which is kind of why those rates, coming back to here, basically just maintained that buffer because, as it's cheaper to borrow DAI on Maker, then you have fewer refinancings. So a little bit less increase in borrow volume on platforms like compound. Because if there's a one point difference in the rates, maybe some people trust the Maker contracts more or something like that, and so they decided to just keep their CDP positions on Maker. |
| # / 01:10:56 | Vishesh Choudhry | So that's what you kind of see reflected in this pie graph. So essentially, this supply has been comparatively steady for a little while now. It, interestingly enough, took a slight dip on the 28th. So essentially, there is a large quantity of DAI that was wiped on the 28th, but there wasn't sort of a commensurate refinancing, to secondary lending platforms. So it wasn't so much, I think, refinancing as just a little bit of capitulation of leverage positions, because that overall supply did come down. So essentially, this was when ETH crossed the 170 line, I believe it was. So presumably, some people just had kind of a hard line at 170 bucks for ETH and decided to close out some leverage positions. |
| # / 01:11:47 | Vishesh Choudhry | But similarly, some leverage positions were opened up with the news of the drop in stability fee and then again with the most recent drop. There was a little bit more debt taken out, but not enough to really cancel out the amount that was wiped. So essentially, part of the question earlier was, okay, why is DAI price still up despite the stability fee having been dropped? Generally, we expect more debt to be drawn out and so more DAI on the market and a bit of a drop in price. And the answer is ultimately because there was not a commensurate increase in DAI supply. So that, presumably, is an indicator that either two things happened to cancel out at the exact same time or the rates were just too high for the demand levels that were out there. |
| # / 01:12:43 | Vishesh Choudhry | As far as collateral ... In collateral terms ... In ETH terms, this collateral has been pretty steady. There was a bit of an increase in the amount of collateral at the same time that dropped below 170 ETH price. But in dollar terms, that's actually come down a little bit. So that's something I can show you, as well. In dollar terms, the collateral value has been slightly, slowly decreasing, but more or less steady. And then in nominal ETH terms, that has been pretty counter-active to the drops in ITH price. So essentially, people are just managing that buffer level of collateralization that they're comfortable with, which is something that you see time and time again. |
| # / 01:13:34 | Vishesh Choudhry | So no significant amount of bites recently. So essentially, kind of oddly quiet despite the drops in stability fee, which can, I think, be used as evidence for consistent, slow decreases. But again, as always, I would stress taking that cautiously and doing it in small increments, but higher frequency is all right. I think, generally, you should wait a week, at least, to see the effects of these changes, though. |
| # / 01:14:17 | Vishesh Choudhry | So as far as what's been going on with DAI price ... I mean, comparatively low trading volumes. So you see this, as well, transaction volume has been down a little bit this past month. So I think there is a little bit of a dip in both trading and transaction volume, general activity. And so that's really interesting that the price has stayed up a little bit despite that dip in usage. And so, in the last 24 hours, there's been 1.6 million roughly traded under these exchanges for ETH. That is relatively low and so, take this with a grain of salt, but the price has been kind of hovering in this 1.5% above a dollar range. Essentially, over the last seven days, that same trend is held with roughly 13 million, which is, again, a little bit low in terms of the amount of trading volume. But the price is pretty consistently just a little bit above a dollar. |
| # / 01:15:25 | Vishesh Choudhry | So those drops in stability fee either did not have much of an effect or were washed out by a fair amount of DAI essentially getting bought up to pay down debt positions to close out some CDPs that were outstanding. It's very hard to say, again, which of those two things is the case, but at least that's what we can see in the data. |
| # / 01:15:52 | Vishesh Choudhry | All right. I will hand it back in case there are any questions. |
| # / 01:16:00 | Richard Brown | All right, thanks Vishesh. I wasn't keeping track. Does anybody have any questions about the peg that they would like to ask? Or questions for Nik? We had a lot to go over in that presentation, as well. Actually, Chris, yeah. This is a relevant question if you wanted to ask that one. |
| # / 01:16:24 | Chris Blec | Sure. I'm wondering if we're giving people enough time to hear about the stability fee changes. I think there's an expectation that a stability fee change occurs and that people hear about it instantly. And I'm wondering also, as a second part of that, if we have look over longer periods. Is there a possibility that a stability fee change from two weeks ago might be just sort of affecting what's happening now? And is there any way to look at that? |
| # / 01:16:55 | Richard Brown | All right, there might be some technical difficulties with Vishesh, at the moment. |
| # / 01:17:19 | Matthew Rabinowitz | There's no way to determine when a stability fee impacts and when it impacts the price. There's no data to support that, but you can imagine there is an inertia that's shifting over to both, the less you do. |
| # / 01:17:33 | Richard Brown | I think there might be two issues here. One of them is that we have no idea what the delay is between cause and effect. And we also have a problem where, this is a crypto-wide problem, where it's very difficult to notify people that something has happened in the crypto space. This is a longer discussion, but this is something that we need to figure out for the entire space but also in governance in general, is how do we let people know ... engage stakeholder, how do we let them know that it's time for action or if an interesting data point has arisen. |
| # / 01:18:12 | Richard Brown | It's a constant battle with communication and transparency. And frequently, we're accused of not being transparent, which sort of blows my mind because I have trouble counting all the different outlets that we have for broadcasting information into the world. But trying to figure out some way to keep voters involved is a tricky one, whether it's [crosstalk 01:18:38]. |
| # / 01:18:38 | Chris Blec | On the flip side of that, Rich, it doesn't seem like we're doing a great job of isolating the experiment so that we can connect the cause and the effect, at least from my limited observation. It seems like we're making these changes so quickly that we're not, if there is a big effect at some point, we're not going to be able to accurately isolate the cause, potentially. So I'm wondering if the cadence of the changes in the polls should be something that we should rethink. |
| # / 01:19:06 | David Utrobin | Yeah, I think that Vishesh actually covered this topic pretty well in the last call. So if you watch his segment, that's why he's advocating for ... more frequent changes are fine, but at smaller magnitudes, right? We had a whole conversation about this idea of magnitude versus frequency. And also the fact that it's hard to isolate stuff because we operate in the real world. Vishesh has spoken about the fact that the stability fee is very much akin to a sensitivity parameter, almost, because the price action of ETH actually is one of the larger affecting variables on the day-to-day DAI price. So, yeah. |
| # / 01:19:52 | Vishesh Choudhry | Yeah. Part of it's also frequency and also predictability. So to the extent that you're making a half percent change every two weeks. And you can be recently confident that you are controlling for some of the variables that you introduced into the system by actions [inaudible 01:20:13] |
| # / 01:20:13 | David Utrobin | You're [crosstalk 01:20:18] |
| # / 01:20:17 | Richard Brown | I think we're losing you, Vishesh. |
| # / 01:20:24 | Vishesh Choudhry | Yeah. So I was just saying, if you are controlling for the way that you're [inaudible 01:20:29] more confident that what you're actually observing is endogenous changes in the system. |
| # / 01:20:36 | Vishesh Choudhry | So that's why I was suggesting something like half percent changes every two weeks. You don't need to sit with the same stability fee for a month and then suddenly make two changes over the course of a week and a half. |
| # / 01:20:50 | Richard Brown | So this ties into the rate stepping and cadence discussions that we've been having in the forums. And I think that it's heartening to see what the community is bringing back to the floor but, not only do we need to talk about exponential stepping for the stability fee raises, we also need to talk about how often we're doing this thing. There's still a lack of clarity around whether exposing this button for people to press is incentivizing them to press a button because a button has been presented to them or whether there is pros and cons with just putting a bit of a rate limit around this thing by exposing the vote every two weeks, and see how that works. That needs to get brought to the floor again. |
| # / 01:21:42 | LongForWisdom | Yeah. So my plan is to start pushing the cadence thing, the stacking thing through so I have some idea of what the [inaudible 01:21:48]. |
| # / 01:21:49 | David Utrobin | So would we want to do a forum poll first? Or do you think that it makes sense just to set it to a governance poll right away? Because the question doesn't seem too hard to formulate. It's really just a yes or no, do we move to a [crosstalk 01:22:04] cadence? |
| # / 01:22:05 | Richard Brown | Well, I want to keep the same process that's emerging. It's fairly well defined at this point. So there's ideally, I brought the cadence and the exponential thing from Reddit and re-posted it into the forums and then in the forums, it got re-posted into another forum. We're getting some murkiness on where these things arise and maintaining conversations. But ideally, I think that it'd be good to have a brand-new, clean forum post that says, "Here's an encapsulation of all the thinking up until this date about the cadence." Then specify, "Here is the seven or 20, 14 or whatever days that we are going to spend discussing it. And at the end of 14 days or seven days, there will be a poll and then that poll will last for whatever," and then after that, it goes into the governance thing. So I think that there's value in repeating those same steps for everything before it hits the governance poll so that people know what to expect and they don't go on holidays and come back and find out that there was a five-day governance poll that there weren't expecting. |
| # / 01:23:18 | Richard Brown | We are at a half hour into the Q&A session. So I think that this might be a good place to stick a pin in it. If people still have questions that they want to see addressed or they have new questions they want to add, please go to the forums, I'm begging you, and check out this thread. This is where the summary will be posted, where the videos will be posted. And [inaudible 01:23:55] |
| # / 01:24:07 | Richard Brown | ... hearing more about the oracles in these calls in the future. Thanks, Vishesh, for your presentation. I think that's ... thanks to the community for all the great questions this week. That was excellent. All right, I think that's it. Thanks, everybody. |
| # / 01:24:21 | David Utrobin | Thank you for an awesome call. |