Skip to content

Ep. 29 Apr 04 - 2019

Video | Audio | Discussion

References Person Text
# / 00:00 Richard Brown All right. Well everyone, welcome to the Thursday, April the 4th edition of The Scientific Governance and Risk meeting with MakerDAO. We have a lot of people on the call today and we have a lot to go over, so I am going to speed through my preamble as quickly as I possible can.
# / 00:22 Richard Brown We're enormously interested in as much feedback as we possibly can get in these calls. We have a limited amount of time, but we have a standing room only group of extremely smart people in here and we need to hear from you guys. So, we're going to speed through stability feed. We're going to talk about the peg. We're going to talk about some analytics and we're going to talk about risk. And then, we're going to open it up to a Q and A session towards the end.
# / 00:48 Richard Brown But, my job for now is going to be to hand it over to Steven Becker and click the admit button while he's talking. So, Steven, it's all yours.
# / 00:58 Steven Becker Thank you very much. Firstly, thank you very much everyone for joining. Again, just to recap, what we try and do in this course, is cover, basically three things. I know we've heard it before, but if this is something that we can just get used to, it's a lot easier to understand the context and the objective of this course.
# / 01:20 Steven Becker The first thing to consider, and, that's something that we will do today, is, the demand and supply [inaudible 00:01:27]. The second would be collateral types, collateral portfolio and parameter adjustments, with respect to that. That will be the second theme that we'd be looking at. And the third would effectively be exogenous risk, you know, things that are outside the system but obviously affect us directly.
# / 01:43 Steven Becker These calls are centered around those themes 'cause I believe that those themes cover my entire spectrum of risks that we need to consider. So again, we remain focused today on the demand and supply imbalance. We're going to have a look at where we are with respect to that and get some sense. But please remember that your decentralized scientific governance, which also includes the stability fee, are considered and facilitated by this call. But ultimately, the decision is made by MKR Governance. This is very, very important.
# / 02:13 Steven Becker And also, keep in mind that governance is a continuous function, so please be involved as much as possible. And you know, with that being said, I will hand it back over to Rich.
# / 02:23 Richard Brown Yeah, and that was a great segue actually, because that's precisely what I want to talk about today. Governance is a continuously evolving process at MakerDAO. I'm heartened by the fact that every time we go through the process, we've improved it. One of the things that we've identified ... the community helped us to identify in the last call ... was that we need to understand community expectations about polling particularly, especially with the variability of the stability fee recently and our renewed attention towards the maintenance of the peg.
# / 03:03 Richard Brown So, what I want to do today is explain a bit about how this process has looked in the past, talk about a couple of the challenges that we're looking at, and I really want to get as much input from the people here about how they would like to see that process changed and how we can tighten up the polling in the executive vote mechanisms.
# / 03:26 Richard Brown So, here's the general picture of the way that things have been working previously. We have a community call and in these calls we talk about the data that we've seen. We look at the data and try to figure out where the market is going to be in the future, how that's affected the peg in the past and some next steps that we need to take.
# / 03:50 Richard Brown We hash that out a bit. We share graphs and then we are wrestling with this idea of coming to a consensus for signaling in these calls. And, I think that there's still some ambiguity there ... At least there is in my mind. The [inaudible 00:04:04] that we've done it previously is, we've decided as a group or the community has come to us and said, "We need a poll and we need to put a range of stability fees in front of the voting stakeholders and have them signal which one they are most interested in seeing." And, we have an executive vote, which all sounds and has worked out well so far. But, the problem that we're having internally and externally, is that there's a significant number of steps and stakeholder and overhead involved with that process.
# / 04:41 Richard Brown And, what we've been doing behind the scenes at MakerDAO for the last few weeks is, spending a great deal of time and effort trying to audit that process to figure out exactly how many steps are involved with determining that there is something that needs to be addressed by the community and ultimately having a vote and then publishing the results of those votes.
# / 05:03 Richard Brown And, I have a document here that we've been playing around with that shows that there's around 37 individual steps, handoff steps, from determining that we should do something to actually doing it and then publishing the results. There's a lot of different departments involved. There's a lot of different stakeholders and sign offs required, and that's just internal to MakerDAO. It doesn't include things like having these calls, summarizing, transcribing, uploading the audio and video, talking about it on Twitter, setting up a Reddit thread to discuss the issues, maintaining the polls, having a call about the results of that poll and then having an executive vote, then publishing the results of those executive votes.
# / 05:50 Richard Brown Anyway. So, that was a long winded explanation of some of the challenges that we're trying to address inside the organization. And, one of the things that we've begun to discover, is that for MakerDAO to remain the foundation, to remain flexible enough to facilitate these votes in a reasonable time period, it seems live seven days is probably the sweet spot for us. And so, the thing that we're thinking about is, we as a group, as a community, determine that there's a change to the stability fee that's required and we set up a poll to figure out what the community thinks that change should be. We wait for them to signal their changes over the course of two day, I think, is probably a number that we can start off with as a baseline.
# / 06:39 Richard Brown Then, we have a call to review that community signal. And then, the next day we have an executive vote and that will probably be on Friday. And, we'll wait. Traditionally, I think that we've seen, it takes like, 10 or 12 hours for the community to stake their Maker on one side or the other by an executive vote, which is really nice. And then, that gives us the opportunity to spend the next week either evaluating the effects of that change or potentially having another poll.
# / 07:08 Richard Brown And so, here's where some of the ambiguity comes in, where some of the complexity comes in. We need to determine whether we are all comfortable with these wait and see periods after these polls, or whether it would be better to do something new. And, that new thing would be the possibility of having continuous polling. And, that's an option that I want to put in front of the community right now.
# / 07:33 Richard Brown So, I know that I've kind of rambled a bit and I've been slightly all over the map, but we need to determine or we need to find out what the community wants in terms of polling mechanisms and polling cadence, I think. And, we've seen some suggestions in the sub Reddit that ... but, they vary between ... Well actually, not between. It's not binary. There's a spectrum here. So, people will see that the peg has drifted and they want an immediate change. And, other people just have a general feeling of what they think a stability fee change should look like and they want to see a poll for that. And, there are other people that are of the opinion that there should be just a weekly poll of whatever, 50 basis points of 100 basis points.
# / 08:19 Richard Brown And, what I really would like to hear today from the community is, some of their suggestions and some of the feelings and some of the reasoning behind what kind of cadence they would like to see for these polls and what kind of ranges we would present and how often.
# / 08:33 Richard Brown Maybe I'm going to leave it there because I know that was a bit of a data dump, but I'm sure there's people in this call that have strong opinions about it. So, if you want to type them in the chat and let us know what those opinions might be. Also, feel free to jump in at any time and interrupt. Because, [inaudible 00:08:51]
# / 08:51 Alex Evans Rich?
# / 08:51 Richard Brown Yeah.
# / 08:51 Alex Evans Can you define this idea of continuous polling? Is it just what you said, where there's a point five on the table every week or is it something else? I just want to make sure I understand.
# / 09:05 Richard Brown Sure, yeah. Let me give you some context there. So, the thing that we need to work on internally and externally, I guess, is agility and velocity on accepting signals from the community. And right now, there's ... I eluded to it earlier, but there's so many different steps and processes involved to get a new poll in front of the community, that it sort of involves doing the same work every week, which is onerous.
# / 09:34 Richard Brown So, there's another option on the table, and that option could be that every Monday, a poll just goes up automatically, semi-automatically, for the new term. And, I think the option is that ... some people have suggested that, 50 basis points every week and just see what the community thinks about that. And, I want to throw a bit of a curve, or add some complexity to that though, because I don't want to get into a situation where we assume that the current picture of the market is the one that we're constantly living with. So, we're not always going to be considering increasing the stability fee and we're not always going to be considering increasing it by 50 basis point.
# / 10:13 Richard Brown One of the questions I wanted to put in front of the community was this idea that, if we do have a weekly poll, it could provide a range. So, anything from negative x basis points to positive x basis points. And, judging by community interest, the state of the ecosystem and the peg, we can watch for signals that other people want it to go up, they want it to go down or they want it to stay the same.
# / 10:43 Richard Brown So, does that provide enough clarity, for this? Does that explain what I'm thinking about?
# / 10:51 Alex Evans That's clear to me. And so, the question in there ... I mean, it does get quite complex. So, if it's 50 bips up, 50 bips down, thumbs up, thumbs down, once a week or stay constant, three options, seems like enough to consider on a weekly basis. Having a sliding scale where that could be as high as ... well, we've gone as high as four and all the way to negative four percent. That changes each week and then, deciding how big that range is on a week to week basis, feels complex to me and feels like the kind of thing that we might want to do more ad hoc, as opposed to having a specific rule preset. But, that's just my initial reaction.
# / 11:43 Richard Brown Yeah, and those are some of the questions that we've been wrestling with as well. So, I want to make sure that people understand that I'm not proposing anything. I'm reiterating some options here. And, there's pros and cons to each side. One of them is engagement exhaustion, which is something that we need to seriously consider. Not everybody is going to be prepared to keep track of the markets and signal, set aside time every Monday to do the research and signal appropriately. That kind of thing has a tendency to have some significant commitment drop off.
# / 12:13 Richard Brown The pro side of this kind of an idea, I think, is that we don't need to coordinate efforts to respond every time we see the market deviating, or the peg deviating. And, there's that lag time that we've been dealing with, of, does it take a week for us to determine that something needs to be addressed? Can we turn it around in a day? Does it take two weeks? So, those are some of the things that we're worried about.
# / 12:39 Barron Gati Rich? Do you mind if I throw kind of a wrench in this whole thing at the moment? 'Cause, it seems like there's a huge, constant ... and, rightfully so. It is one of the major mechanisms, like, dependency or go to for the stability fee. And, the more I've thought about it, the more I come to the idea that we might want to ... if this is feasible, I don't know. I talked to Cyrus very briefly about it and I haven't had a chance to write it up on Reddit or in general.
# / 13:08 Barron Gati But, there seems to be just this time dependency mismatch and coupling that with irrational human behavior ... you know, humans are just terrible at rational behavior in the face of delayed punishment or gratification. Like, you know, prospect theory, risk assessment, future cost discounting stuff. And, when you couple that with this unknown information distribution across the community of CDP holders, and unknown information uptake of that community, and it might be that something new or drastic or different is needed to bring the price expectations of DAI in line with the supply changes. 'Cause, we've seen supply changes but haven't seen that follow through to order quantity and execution on the exchanges.
# / 13:59 Barron Gati So like, something that randomly came to my head was having an auction for people who hold MKR to quote end quote, "trade" some of their MKR as an incentive to give CDP holders-
# / 14:12 Chris That seems very complicated. Doesn't that seem overly complicated though?
# / 14:16 Barron Gati Oh, yeah, yeah. No. This is just something ... There's tons of different things though, that I'm saying. Like, yes. That specific example certainly seems complicated, 'cause there's a lot of infrastructure we'd have to build and think about. But, the general point is, the stability fee might not be sufficient to drive the conversion and thus, the price appreciation and tracking that, I think, the community is going for.
# / 14:40 Richard Brown Yeah. I love, actually, everything you said. So, I think that we probably share some ways of thinking. I think we all know that crypto is not rational. The crypto market is not rational. People's responses to it are not rational and so there's definitely something to be considered with, do we just expose this lever that any sort of, panicked group of people can just start signaling for, based on whatever the market did that day, right?
# / 15:07 Richard Brown And, you also touched on this idea that we're, I think as a group, becoming more and more familiar with, is that there's no immediate cause and effect between the stability fee change and the way that the market reacts or the way that the inventory responds to those things. I think that's ... Matthew? Let me see. Is he in the call? There's been some really good posts in the MKRgov sub Reddit about the elasticity of demand that we're seeing here.
# / 15:32 Barron Gati Yeah, I read that one from last week, and that's definitely part of the picture of what I was talking about, for sure.
# / 15:38 Cyrus Younessi Can you guys ... or Rich, can you specify if you're thinking of this more as a long term problem and solution, or something that you want to implement right now? 'Cause it just seems like for now, whatever extra work we have to do to bootstrap or get things going, not really a big deal. I don't see voter apathy as a huge issue right now, 'cause everybody's kind of engaged. You thinking more of trying to find like ... I mean, obviously there's a lot of talk about like, why can't we just do an automated change to the stability every week, by whatever. That's not ... I think we should be clear on if we're thinking long term or we perceive it to be an issue today, and we need to cut down those 37 steps today, because we're running into issues with actual managing things.
# / 16:28 Richard Brown Those are all great questions, so let me clarify one thing. I'm not proposing anything. I'm trying to condense and refine the ideas that I've been seeing in chats and socials and Twitters. I think there's something to be discussed here about a continuous poll. I'm not sure whether that's a good idea or not. That's why I want this block of time that we've allocated in the call to start exploring. I'm not sure.
# / 16:53 Richard Brown Those are great points, like this picture that we see ourselves in right now is not the picture that we're always going to be in. The market shifts, [inaudible 00:17:02] comes out. All kinds of things change, so maybe be cognitive overhead cognitive load, of what we recall, is potentially worse than the lack of flexibility with having to wait two weeks before the next poll comes along, or week and a half. I don't know.
# / 17:20 Barron Gati But it seems like this is also ... you're always going to run into this problem of you might be using the old FED analogy, you know, pushing on a rope. Where you can do all of these things excellently, and you can create this whole infrastructure for the stability fee changes, but then the stability fee changes don't have the impact that you want them to have on the target price. So, I think if you're going to spend a lot of resources right now, the allocation of those resources shouldn't necessarily ... and this is just my thought. I could certainly be wrong and not convinced of this either way, but this just where I'm at right now ... the allocation of those resources right now should probably be better spent on understanding how the market is acting with respect to DAI and CDP closures, and borrowing, and wallet movements and leverage, and all the stuff so we can have better pictures to frame this discussion in.
# / 18:11 Cyrus Younessi Yeah, and that's something we're actively working on. We've put together a small but dedicated group of community volunteers to form some sort of a volunteer risk group working on sourcing data, curating it, open sourcing it for the community. We're engaged with a couple analytic firms, couple chain analysis firms, and paint the overall picture of DAI and what it's being used for, demand and supply everything like CDPs. So, we are expanding resources on kind of a base case of data that will help for anything we want to do with it. But I think even beyond that, we still have to deal with the governance aspect, as well. In my mind, it's kind of two distinct issues.
# / 19:05 Chris Padovano I have a question. Does anyone disagree that the stability fee should not be raised right now? 'Cause that seems to be something that we haven't really talked about just yet. Maybe that's coming at the end of this, but just [crosstalk 00:19:18].
# / 19:18 Richard Brown We're going to segue almost immediately, and that's [inaudible 00:19:21] into the risk presentation from Syrus, and then we'll get into that. I'm not sure that we're going to find a clear consensus on the cadence rate and utility of continuous polling right now, but what I wants to do is I'm going to post a link into the chat here. This is the discussion for this call that I'll be posting the transcripts in the audio/video of the call, as well. So, if anybody has thoughts that we don't address in the hour that we have today, please add those to that maker gov thread, and we'll continue to hash this thing out.
# / 20:02 Cyrus Younessi Yeah, I think this is a great discussion to have in MKRgov subreddit. Typically so far, it's just been about DAI price and stability fee adjustments, but I think it's time to broaden that scope out a bit, because it's important.
# / 20:19 Richard Brown I agree, and we're iterating a bit on the process, and in the future, we'll be having these discussion threads out there prior to these calls, where we'll do a better job at actually explaining exactly what the calls are going to be about, giving people an opportunity to do some research and form some opinions before we get here, and then we'll have the call and we'll continue the discussion in the thread. Hopefully, that'll give everyone an opportunity to prepare and formulate some ideas ahead of time.
# / 20:49 Richard Brown I'm going to let that go there. To recap, as a group, we need to figure out how we handle signaling. What signal do we use to determine we need a signal? What I mean by that is, how do we, as a group, as a community, figure out that it's time for us to have a governance poll and how does that governance poll get kicked off, and what's the cadence of that poll? Those are all outstanding questions right now, so if you have ideas, thoughts, or proposals, please add them to that thread that I just linked to.
# / 21:23 Richard Brown Syrus, I'm going to hand things over to you right now for the risk segments, if you want to tell us about the state of the peg. Then we can move on from there.
# / 21:29 Cyrus Younessi Sure. Give me one second.
# / 21:47 Cyrus Younessi This is the DAI price over the last week. When we were speaking last week, it was kind of hovering around 98 cents, and we had seen a bit of an uptake from the prior 4% increase. I think people were feeling pretty good when we were on the last call, but then, I think immediately after, peg kind of started to weaken again back down to the 96 range. I think a bit disappointing for sure. Then, last two-three days, we had some fairly extreme market volatility, which provided some interesting things to think about. Contrary to what we had seen in the past, we did not see DAI weaken on the rally, but in fact, we saw it strengthen back up to the 98 range, with a fairly decent volume, too. So, we don't have volume on this chart, but ... kind of hard to see, but we did see some significantly higher volume with the market volatility, and it's definitely hovering in the 98 cent range. So, let's look at the CDP stuff, then we'll come back to that in a moment.
# / 23:21 Cyrus Younessi A couple things, we saw this huge pink bar. Obviously, we saw a lot of white activity as the market rallied, as people were basically taking profits. That's essentially the component that we hadn't seen over the past few months, where typically on the rally, we'd seen speculators trying to momentum trade this, and keep trying DAI and levering up and trying to go with the market. I guess Tuesday's huge rally kind of pushed a lot of speculators to close out and take profits. That being said, we did also see a significant number of draws as well. So, the total reduction in DAI over the last few days wasn't actually anything meaningful. It kind of balanced out. Over the past week, we've seen fairly stable DAI supply. That's that yellow line.
# / 24:23 Cyrus Younessi For me, the main takeaway here is that there's definitely some conflicting components here when it comes to the relationship of DAI supply, DAI price, and the ETH price. Definitely going to require some more granular analysis.
# / 24:39 Richard Brown Syrus?
# / 24:40 Barron Gati Yeah?
# / 24:41 Ben Sparango Can you click the ETH price on that chart?
# / 24:46 Cyrus Younessi Yeah, sorry about that.
# / 24:47 Cyrus Younessi Yeah, this pink bar is where we saw the big ETH rally. Yeah, so I think the main takeaway was that speculative behavior is obviously dynamic. You can't always rely on those same assumptions.
# / 25:15 Cyrus Younessi Yeah, I think there's a lot to talk about, but for now let's just open it back up to the community, I think, and see what people think.
# / 25:24 Barron Gati Generally speaking, what would you expect people to do at different ... I don't if you've studied this or not ... at different daily or weekly price change percentages on Ether? 'Cause I can see the story going full circle, where you have one part of it basically like starting over, so price goes up, people want to close out positions, but at the same time, you have people believing the price will continue to go up, so they open positions. Have you looked at how that plays out over different percentages of changes?
# / 26:04 Cyrus Younessi That's an incredibly complex question, and I think drawing a comparison from Bitfinex and BitMEX ... I know Bitfinex definitely posts their longs versus shorts statistics, and I know those have been heavily studied over time, and people are always trying to figure out at various prices points, are people going to basically go one way or the other. As far as I know, when I looked at that stuff in the past, it has been fairly unpredictable. Trying to use speculators, the price points where they enter and exit into positions, is not very predictive, in my opinion. I think trying to guess, if ETH goes from 160 up to 200, that's going to inspire people to cover or go to lever up long more. I think it's going to be fairly difficult, if not impossible.
# / 27:13 Richard Brown Vishish brought some more graphs for us today. Do you want to do a quick screen share and have him walk us through those?
# / 27:19 Cyrus Younessi Sure.
# / 27:19 Richard Brown Vishesh, if you want to take it away for a second.
# / 27:21 Vishesh Sure.
# / 27:25 Barron Gati Who's doing the screen share, to be clear?
# / 27:27 Richard Brown Vishesh.
# / 27:29 Barron Gati Okay.
# / 27:34 Richard Brown I'm interested because I think there's going to be some interesting data in there about competing loan platforms and seeing how that affects us.
# / 27:43 Vishesh Just a quick update of what's new since last week, this is the same DAI-ETH VWAP price graph with the trading volume as well. So, just as Cyrus had mentioned at the beginning, there have been some spikes in trading volume in the last couple days, which is good to see. The other interesting thing, which I noted when I looked at ETH price as well is that that volume seems to track a little bit with fluctuation in ETH price, and this has been a long discussed topic, I think. The correlation is decent and I can look at the correlation graph next time, but not super strong. It's just the relationship with ETH price and DAI price, so you'll note that as the ETH price has come up a bit and held steady in the past month or two, what's interesting is, the supply has not changed that much, but as you can see here on the transaction volume graph, which was what Cyrus was alluding to a second ago, though the supply has not changed much, there was a huge spike in transaction volume, which is, from my understanding, from a big off-setting affect of people doubling down on draws and wipes, respectably.
# / 29:16 Vishesh So, I'm not sure the reasoning behind that, in terms of perhaps people who are long on ETH or just sort of upping their amount of leverage, and then people who are worried about the stability fee or starting to feel some of the pain of keeping large, outstanding debts, decided to take some of those profits and pay down their debts.
# / 29:38 Vishesh Then, what Rich was mentioning a second ago, in terms of the interest rates, so what's interesting is we can see the Compound interest rate had come up a little bit in the last few days, but really not as much of an increase as I would have expected. Let me zoom in here for second. See, you can see, it had upticked a little bit, but really not the level of increase that I would have expected to see, which is interesting. I know there was some discussion on the chat in this call about looking at this Compound interest rate as somewhat of a guiding factor for how much we should be increasing or decreasing the stability fee by, and I know there was some discussions last week around the DIPOR rapport and all that, and that's a topic that I'm sure we'll explore a little bit further, but it is interesting that unlike in the past, this time when we jacked up the stability fee, there was not a huge change in the Compound DAI interest rate.
# / 30:41 Vishesh The other thing was the loan volume was really interesting in the past week or so. There was a sort of slow and steady increase in the amount of DAI being supplied, which I think does tend to make sense with the increase in ETH price, but the borrow volume also spiked before we saw a tremendous increase, which was kind of interesting, in the sense that there was an impact on demand I think that this is an indicator of. So, even though I expected the increase in the stability fee to have somewhat of a dampening affect on demand, as well as a dampening affect on supply, both had seemed to run up, which was another unexpected data point.
# / 31:33 Vishesh Not much to say in terms of the distribution of draws and wipes, respectively. Actually, it's been fairly well distributed for the past couple months. Not very many bites, as you can expect, but that's mainly as a function of ETH price.
# / 31:47 Vishesh Then, the sizes of the CDPs that we've been seeing some of these wipe transactions on have actually been fairly well distributed. There was a little bit of clustering of some big CDPs that had wiped down some debt, but really not enough of a skew to really make any conclusions off of.
# / 32:08 Richard Brown There's a lot to digest there. That was really interesting, though. I hadn't considered using CDP size as a weighting of potential sophistication of the user. Is that what you're doing?
# / 32:22 Vishesh I had looked at that as one factor. I had looked at just pure wipes and then I tried to see if maybe there was some sort of skew across whether large CDPs were getting wiped out or small ones, but there wasn't a huge conclusion for that. It seems pretty evenly distributed.
# / 32:41 Richard Brown It's something to watch, though. That's really interesting.
# / 32:46 Richard Brown Alright, thanks, Vishesh. That was great.
# / 32:49 Richard Brown There was a lot of chat in the sidebar here. Is there anything that should be bought to the community's attention? Is there questions that could be asked?
# / 32:59 David Utrobin Not particularly relevant to all the data and stuff, actually, I had an interesting suggestion about [inaudible 00:33:06].
# / 33:07 Cyrus Younessi Cool, no worries. Basically, his suggestion was, given the flack DAI supplied this last week, we sort of maybe equilibrated, in terms of the DAI supply, especially in he face of the market rally. So, if there is an assumption that with the current stability fee, we're not going to see any significant deviation from this rough 88 million DAI supply, then potentially, all we need to do is just soak up the static excess reserve. So basically, trying to soak up whatever is needed to get from 98 cents to a dollar. Potentially, that's a relatively fixed number, so under that context, he was recommending very small and steady, I think .5% stability fee increases up until every week, or every cadence, until we see that excess DAI soaked up. So that's one option.
# / 34:22 Chris Padovano I'm curious if you guys have reached out to some of the competing lending platforms and seeing the rates, like for example, with Genesis, or Cumberland does lending, and if you're looking to get some size, I would say we are still extremely cheap, compared to some of our competitors, and my personal opinion is that making sure that there's a gradual increase and not screwing the pooch, so to speak, with too high of a stability fee is definitely appropriate. I think we should all come to consensus on this call, hopefully, that we may need to be a little more aggressive with rates. Just looking at who the competition is and what they're currently ... what the borrowing costs are for people using those platforms.
# / 35:28 Cyrus Younessi That's a great point, and that's definitely something we've talked about looking at both the OTC lining desks, but also margin rates across various exchanges. That's actually stuff ... we have data for some of that already, just haven't found a way to output it yet.
# / 35:46 Cyrus Younessi Yeah, that's a great point. We definitely should be looking at stuff like that, though, on a weekly basis. To be clear, in your opinion, we need a fairly large increase in your mind, if there were to be a governance poll?
# / 36:04 Chris Padovano It depends on what the cadence is, right? If we can all come to agreement that for the next four weeks, we are going to raise the stability fee by 1% each week, or whatever you guys think is more appropriate, then I think that that's probably fine, but if we're not going to make a commitment to doing that, then I would much rather prefer being super aggressive in this regard, to try to get the peg back to parity, because there has been very little activity at a dollar and above a dollar, other than some errant trades, it would seem, on some of the non-custodial exchanges. [crosstalk 00:36:47]
# / 36:46 Richard Brown I want to cut in super fast, and just comment on that, Chris. One of the things that we've been wrestling with for a while now is this concept of forward guidance, and if we set out a regular schedule of increases over a period of x weeks, or we let the community determine that this is going to keep on happening, that dose serve as a signal to the rest of the market that we're all serious about this thing and it's going to continue to be increased, which might have an affect. That's also something to consider when we decide on a schedule for these things.
# / 37:21 Barron Gati Just out of curiosity, Chris and Rich and Syrus, or anyone, please feel free to weigh in as well, if you had to take a position, would you rather have the DAI trading above a dollar, or below a dollar, assuming the distance is small?
# / 37:41 Chris Padovano Lev, do you have any thoughts on this?
# / 37:47 Barron Gati 'Cause my thought is that the natural state of the DAI, given the system, might be slightly below a dollar. Kind of like-
# / 37:55 Joe Quintilian It's a function of interest rates, right?
# / 37:58 Barron Gati Yeah.
# / 37:58 Joe Quintilian If I'm a trading desk, I'd rather take my money out and take my 2% and put it in CDs or put it in the bank. That's what's ... until DAI savings rate comes in, we will ... it's tough for us to keep it at one, but I think that if we move the stability fee up further, Chris and I don't agree on everything, but we agree on this, in terms of going on and increasing it at a 1% or .5% rate for the next couple weeks. I think we should do that. That's what my opinion is.
# / 38:42 Cyrus Younessi Do you want to talk about any sort of inventory market making, or things you've seen on the trading site, in terms of activity? I think we skipped that part.
# / 38:50 Joe Quintilian Chris hinted at it, but we did get above one in some venues on the move up, so when people wanted to go and liquidated their ... when they wanted to get out of their ETH, they came and immediately lifted to get into DAI. That happens, and because DAI is on really 30 venues, you really have a liquid on ... there's no venue that's really won out yet, and because of that, if a person wants to get out of a million DAI, or really a quarter million DAI, they really have to lift up, and they lift up the one, and that's what cause the temporary spikes up to one. But I have seen there's plenty of bids done at 98 cents, 97 cents, so on the OTC front, but over the past week, that's what's kind of been happening.
# / 39:47 Matthew Light Hey, this is Matthew, and I've done a lot of market making over the last half year or so, mostly on ETH to DAI, and before that, Oasis. What I've seen since February, is we've never really had even moments where DAI was trading over dollar. Compared with the centralized exchanges, Coinbase pro and Kraken, the price has pretty much been ... for the last two months, it's always been below a dollar, basing that on the ETH price. That's very different than what we saw in November, December, and January. So, at this point, we know there's a really large inventory of DAI he hands of market makers, including the Maker market makers, too. There's an extremely large inventory, and I think probably all the market makers would like to reduce the inventory of DAI that they're holding in exchange with things like Ether and other stablecoins. So, I think we need to be fairly aggressive and not just do quarter point, half point here and there. I think we need to be aggressive and get the inventories down, because until those inventories are down, we're not going to have any real price discovery going on in this market. That's my thoughts.
# / 41:30 Richard Brown Okay, so, that circles us right back to kind of what I was talking about. The top of the meeting is how do we start a signal to get a signaling poll out there, and I'm still wrestling with that idea, but I think that we can maybe experiment with some of these ideas.
# / 41:46 Richard Brown I've been watching the chat intermittently, and I was seeing things from let's be slow and steady to we need an aggressive change in order to fix this thing. It's very confusing from my uneducated perspective because I'm no financial genius, and so I don't have insight into he reasoning that goes behind some of these suggestions, and that's kind of where I get confused, I think.
# / 42:11 Richard Brown So, figuring out what the bar needs to be, I think, when we do these changes, is something we need to consider as a group. I think we should raise the fee is probably not going to be a super compelling answer.
# / 42:27 David Utrobin Rich, I have a quick question for you.
# / 42:29 Richard Brown Sure, just let me finish this line.
# / 42:32 Richard Brown But if we see suggestions we need to raise the fee by x because of y and z, I think that's going to go a lot farther towards instituting some kind of a change here. Some inertia.
# / 42:46 Barron Gati This is something Volcker and Bernanke have studied extensively and Greenspan to a lesser extent in just their general studies, and you definitely see market signaling being something that is of the utmost importance to the fed. There's a lot of lessons that we could probably take from that, and one of them is, if you're trying to hit a specific target and you are worried about overshooting that target, it's more important to signal caution to the market than to signal aggression. But on the flip side, if you want to signal and you're not worried about overshooting, and you just want to get control of a system, you can go like how hard Volcker did in the early 80's and not worry about overshooting on one side, and signal aggression, and the market will be more likely to respond. So, if you want to put in place a system of we are going to do a 3% increase every single week until this happens, and you set that expectation, you will very likely get a supply significant decrease in supply. It might overshoot a dollar and you might overshoot but if that is the goal, I think there are some lessons there from the Fed.
# / 44:02 Cyrus Younessi Well, I think we have been trying to signal that for actually a month and a half now, except instead of saying as a community, we might do something every week, I think it's maybe been closer not he order of every two weeks, just because of the way our governance process works, and trying to let the market absorb our changes, but we have been fairly-
# / 44:25 Barron Gati It's ... the consistency has been there, but the aggression has not been there. For example, if you look at those rate hikes that Volcar did in the early 80's to get us out of the inflationary expectation spiral that started ramping up with commodity prices at the end of the 70's, you'll see he basically, he just went petal to the metal. He raised short term interest rates to unimaginable levels, like 24.5% on a three month treasury bill. That's a bonkers number, looking back at it, but that's what was "needed" to get the market expectation aligned with ... to get them in line. So, the consistency is definitely there in terms of the signaling, but the aggression might be able to be ramped up, and the consistency of that expected aggression.
# / 45:19 Richard Brown We're coming up on 10 minutes left and I really want to leave this call with some kind of sense of direction, so let me recap some of the options that we're talking about right here.
# / 45:29 Richard Brown We're talking about the community sets an expectation for x percent increase over y weeks, and that is strong forward guidance for the rest of the ecosystem to understand. That's one option. The second option that we're talking about it is we present basically, a weekly poll with a range of options, and we allow people just to find what level they are interested in and deal with the pros and cons associated with that response, or that process. A third option is that we sort of continue along the path that we've been going down, where we try to identify a need, and then we get a poll up and running as soon as possible, once we get a general consensus.
# / 46:15 Richard Brown So those are the three options that we're looking for. I don't know how we do signaling to figure out which signaling method we want. Like, do we want to do plus ones in the chat? Do we want to have passion speeches, like people with microphones? Do we want to continue this discussion in the makergov thread? What are people thinking?
# / 46:37 David Utrobin There was also the suggestion of informal Twitter polls that was also in the chat.
# / 46:42 Barron Gati Is there a way to get a sense for the community's feeling about overshooting or not overshooting? Because I feel like that could actually drive how hard this should be pushed.
# / 46:53 Matthew Rabinowitz [crosstalk 00:46:53] Let me chime in on that one. This is just my sense of it, is that the stability fee, when you start seeing the DAI price ... excuse me, the DAI supply, contracting and expending, but more or less staying static, and we reasonably believe demand is materially lower, as we talked about a couple weeks ago, maybe it's 70 million. We need to raise the stability fee to the point where we have a linear reduction in that DAI supply, not exponential, like carrying a large cooking tray of water. It's really easy to tip it over if you just make too much of it shift off. We just don't want to have an exponential contraction of the supply. [crosstalk 00:47:33] It's important to just have a constant cadence of 50 basis points, 100, whatever the number is. But it's really the target of getting those two to mesh, but linearly.
# / 47:46 Barron Gati That is an excellent answer to the question. Thank you. So basically, we want to avoid the conversion of a linear to exponential change in supply decreases, so in that sense, then yes. Something where you say you're signaling to the market, that would be much more along the lines of Bernanke Powell, Yellen, where you're looking at saying, "This is what we are going to do. We are 100% committed to this set of actions until this happens." And set that expectation, and then follow through with that expectation every week kind of thing.
# / 48:20 Alex Evans From my perspective, I have no worry whatsoever, that we open like that, as far as overshooting the price of DAI and we move up or we have a significant contraction in DAI supply, we move up above a dollar. These market makers have been out there accumulating tons of DAI for months now. The worst thing that happens in my mind is that we overshoot temporarily. The one dollar mark is they get compensated for doing so, which doesn't sound like a bad outcome to me, but that's just my perspective.
# / 48:57 Richard Brown That's interesting. So, let's avoid exponential in changes and look for linear alliance with that first option that's been discussed, where we set an expectation of x percent over y weeks. Am I understanding that correctly?
# / 49:10 Barron Gati That's if you subscribe to the don't overshoot and don't risk exponential decreases of supply philosophy. If you believe ... I can't remember the name of the person who just spoke. Placeholder conference room. If you believe the placeholder conference room, then you would definitely do the opposite. You would say let's do 5% this week, 5% next week, 5% the week after, until we see x.
# / 49:36 Richard Brown But that's linear. That's not exponential, right?
# / 49:38 Steven Becker No, that would likely cause an exponential closure rate, because you're-
# / 49:42 Richard Brown Oh, I see what you're talking about. The closure rate. [crosstalk 00:49:45]
# / 49:47 Matthew Light Well, the market makers are sitting on an enormous inventory of DAI right now, and they have to not been able to offload any of that DAI for dollar for two and a half months now. Even with spikes, they haven't been able to offload it. So, they have a lot of DAI that they can supply for a dollar, a dollar one, a dollar two. I'm not really worried that we're going to have some catastrophic situation where DIA trades for $1.25. I just don't think that's likely to happen, given the amount of inventory being held right now.
# / 50:36 Barron Gati Yeah, I there's a lot of latent demand sitting out there ... sorry, a lot of latent supply at a higher price, you'll definitely see that push back in the case of an overshoot very quickly.
# / 50:55 Aviv Milner We might consider as well is just the fact that the price per week has been going up, and the overcollateralization percentage on CDPs is quite high, so people are ready to print DAI. You have 430%, so-
# / 51:08 Baron Gati Whoa, what? The overcollateralization percent increased by a third in two weeks?
# / 51:16 Aviv Milner That's what the charts say.
# / 51:17 David Utrobin It's currently at 395%.
# / 51:20 Chris That makes sense, right? ETH is [crosstalk 00:51:23].
# / 51:26 Barron Gati Sorry, I forgot that it was a cross sectional state. I thought it was dependent on the current price, more so than the weighted average past price. Got it.
# / 51:37 Chris Sorry, do you want to continue? I think you were making a really good point.
# / 51:42 Aviv Milner Sorry. I'll continue. So anyway, what the point is, is that we know there is the ability to print DAI. If DAI was over a dollar for too long, people would start printing DAI, assuming that it would go back at a dollar, or below a dollar, but right now, from my point of view, it seems like we should be defending DAI holders, because as a maker, we have to protect them, and people holding DAI, these market makers, they don't have an exit, so I don't think we should be too worried about overshooting, given that people can print DAI, but that we haven't been able to sell above a dollar for two and a half months, like the previous gentleman was saying.
# / 52:19 Cyrus Younessi Yeah, but if you print DAI 'cause it's at 102 but the stability fee is something enormously high, then you might not see market makers printing excess DAI to push the peg back down if it's just super high.
# / 52:32 Chris Padovano Well, market makers wouldn't print DAI 'cause they're just holding DAI and supply CDP holders, and kind of people looking for credit are going to be printing DAI [crosstalk 00:52:44]
# / 52:43 Cyrus Younessi I'm assuming after market makers have all their inventory soaked up, sure. That's not an overshoot in my mind. An overshoot would be is if market makers are able to unload all their DAI, and then it goes above a dollar. Which is what would happen if we see some sort of exponential closure of CDPs. It's not ... it's the same problem in the other direction. I'm just saying we have to be somewhat careful.
# / 53:10 Aviv Milner I think part of the problem is this idea that 7.5% is really high, or even 10% is really high, when we're seeing that the market on other exchanges, other platforms, the rate is so much higher than what Maker's offering. So, I think that's the bias that we have to get out of our heads, that 7.5% is high. It might not be high, and maybe 18% is high.
# / 53:33 Matthew Rabinowitz I just think they're totally uncorrelated. I don't think we should look at them like that. Think we just need to work toward getting supply and demand to match. Right now, we don't know what the exact demand number is. We all can appreciate there's too much supply. The question is how we deal with it.
# / 53:47 Chris Padovano The only thing is, I think the demand question is pretty easy to answer. There's market makers who are looking to acquire DAI to make markets across the board, so it's really just like [inaudible 00:54:02] market makers that are creating the demand. It's not like you can go out and you can pay your rent in DAI. It's not like you can use DAI in auger. Right now, it's just ... all the demand is coming from liquidity providers, so I think that for the time being, while the peg is kind of unhinged from the target price, that there should be more of a focus on some of these folks who are holding big DAI bags. [crosstalk 00:54:35]
# / 54:34 Richard Brown I'm going to jump in really fast because we're at the top of the hour and we're going to start losing a lot of people. I think there's a passionate discussion happening here, and I'm happy to keep on going for the people that want to stick around, but let's leave some next steps here. There's two options on the table that I've managed to record. So far, it's which type of poll are we going to initiate immediately, and then there's a discussion about whether aggressive action is warranted or whether slow and steady action is warranted, so we have these two caps ...
# / 55:04 Barron Gati There's a compromise camp there, too, which is a consistent, slightly higher but not fully pedal to the metal situation, as well.
# / 55:16 Richard Brown Sure, that's a good point. So it's not everything in the world is binary, but the point I want to make is, let's continue this discussion in that Reddit thread that I just linked to, and I think the ... an interesting lower bound, or interesting time limit is that we'll finish the discussion there, and then at the end of the day, we'll figure out what the community feels the next step is for as far as a poll goes. I think that that's going to be the intermediate step and will get more complex, possibly, for the next call.
# / 55:50 Richard Brown That was my ... Steven, did you have any thoughts you'd like to leave us with?
# / 55:55 Steven Becker Yeah, just the one thought that if you guys consider the fact that everyone had really great ideas, the real challenge is, and this is also the richest point, is we're making all these points from [inaudible 00:56:07] make it a point. We're making all these points from almost a centralized point of view, but we've got to start thinking, how do we apply this from a decentralized point of view. That's the essence of this. What I'd love to do in some sort of imaginary world, is take all these data inputs that I've heard over the course of the hour, and somehow give it to some independent AI machine to come up with a conclusion for us. But in effect, that's what we've got to do from a decentralized point, is figure out the construct we present to the community, so they know which buttons to click and which ideas to implement, so that we do come up with a robust conclusion when it comes to stability fee or risk parameters and so on, so forth.
# / 56:48 Steven Becker So, that's the real challenge here. I don't think any of you guys have a problem with the demands of economic concepts. All that is actually there and it really is very subjective. It's stability to present in a decentralized framework that allows us to get a robust conclusion. The other thing to effectively keep in mind is the momentum of this discussion really gets pushed from this call, and to Rich's link that he's put into the chat, please carry on discussing this further, on any kind of medium that you can think of, and it would be really appreciated.
# / 57:29 Steven Becker On the back of that, thank you very much for joining this call.
# / 57:34 Richard Brown Alright, thank you, Steven. I am fine with hanging out for a while to see if this conversation wants to continue, but I think I am going to stop recording, because it's already a lot asking people to review an hour every week.
# / 57:50 Richard Brown Thanks, everyone, for showing up. Thanks for the amazing questions and comments. As always, I am going to be posting audio and video links soon. The transcript will be up in about 24 hours, and I'll provide a summary in that discussion thread that I've been promoting in the chat as well so we can keep this ball rolling. Alright, thanks, everyone.
# / 58:08 Cyrus Younessi I think Jordan makes a really good point in the chat.