Ep. 34 May 09 - 2019¶
Video | Audio | Discussion
References | Person | Text |
---|---|---|
# / 00:00:02 | Richard Brown | Hello everyone, welcome to the May 9th edition of the Scientific Governance and Risk call. We have some things to go over today, some interesting developments in the world of MakerDAO. As always, we're going to have a chat. |
# / 00:00:16 | Richard Brown | I'm just going to, well should I do the preambles, I'll do the preamble super fast. We're super excited to have everybody in the call, we have some deep thinkers and some people with some really great opinions and it's extremely important to me and the rest of the organization that everybody gets heard. |
# / 00:00:32 | Richard Brown | We want the additional insight, the additional color and commentary, so if you have something to say, please feel free to just jump right in and start speaking and start asking questions. If you don't have a microphone or if you're supposed to be working right now, just chat in the window and we will trawl for interesting questions. |
# / 00:00:50 | Richard Brown | We're going to do something else today, maybe even formalize something today that we've been doing for a while and that's continuing the call after the top of the hour. Because there's a lot of really interesting freeform risk type questions that emerged last two or three calls and I want to keep that momentum going. |
# / 00:01:07 | Richard Brown | So here's what's going to happen, Steven will give us his thoughts of the week and then I'm going to talk a bit about what's been happening in the governance world and MakerDAO in general, that affects risk and governance. Then, we'll jump into Cyrus with the risk teams and we'll see what's going on with the peg. We will have a summary from Steven and then we will stick around afterwards and just chat about money and crypto and stuff. |
# / 00:01:33 | Richard Brown | That is my preamble. All right, Steven, I'm going to hand it off to you. |
# / 00:01:38 | Steven Becker | Thank you, well, firstly, I'm not on mute. Good. Firstly, hello everyone, obviously welcome to... thank you for joining us and most of all, by joining us, you're contributing towards the governance of the MakerDAO protocol. |
# / 00:01:54 | Steven Becker | Something I can't emphasize enough that is terribly important to do on a continuous basis as well. Being engaged gives you not only line of sight, but it gives you a sense of exactly what it is that you need to contribute to and where that value may come from. |
# / 00:02:15 | Steven Becker | So where to start? Well, over the last couple of weeks, there's obviously been a lot of news in the crypto space and obviously, MakerDAO being on the forefront as one of the main components in the crypto space, we were obviously in that line of sight as well. |
# / 00:02:30 | Steven Becker | And so what we basically have clearly established is the MakerDAO protocol is important in developing the crypto space and the DeFi space and understanding clearly how it is integral to it and how we need to establish a firm sense of, not only the DeFi space and our operation, because also the traditional space as well. But in that light, we need to be very cognizant of how stakeholders use the system, organizations integrate with the system and how constantly the system is always under review. |
# / 00:03:13 | Steven Becker | And during this process, you find extraordinary gems that give you a new use case you never saw before, but also at the same time, you identify issues. And the one item that we established over the last week or so was a security issue, one that was identified and quickly fixed in the course of business. |
# / 00:03:35 | Steven Becker | Now, to that end, what I need to sort of emphasize is the Maker Foundation, which is a super user of the system, has the responsibility of facilitating the bootstrapping of the system. And in that, is the responsibility of looking after the safety as well. |
# / 00:03:55 | Steven Becker | So through that process, we've obviously established this particular issue, the issue has been fixed. And soon enough and something that I think Rich can talk to, a post-mortem will come out where everyone can get a sense of what that issue was. |
# / 00:04:11 | Steven Becker | But in terms of this particular meeting, once again, please keep in mind that the themes we like to cover are essentially the dissipation of supply and demand, where it affects the price of DAI, but as well, the effect on the stability fee as well. And there, I think there's been quite a bit of improvement. |
# / 00:04:37 | Steven Becker | To that end, there's also the second theme of looking at collateral types, the collateral portfolio and the risk parameters that are attributable to those two items. And then lastly, again, just remember the third theme of this particular call is exogenous risks, things outside of the immediate system that we need to be cognizant of, in order to make sure it operates appropriately. |
# / 00:05:02 | Steven Becker | I think I pretty much sort of covered what I needed to cover, but again, thank you very much for joining and I'll hand over to Rich. |
# / 00:05:10 | Richard Brown | All right. Thank you, Steven. I am going to post a link in the chat here and this is going to be a common occurrence as we go forward in these calls. The reason I'm doing that is because... and this is a topic I've touched on over and over again. I'll continue to do it. |
# / 00:05:28 | Richard Brown | We have a very small window of engagement in these calls, it's one hour a week to discuss some issues and kick off some debates and we need a longer tail of a discussion to occur. That discussion right now, until we find some better tooling, is going to happen in our MakerDAO subreddit. I just posted a link to the discussion thread for this call. |
# / 00:05:49 | Richard Brown | If there's questions or comments or you get particularly riled up about any one issue and you want to continue to talk about it, please hit up that thread and ask your questions there. And then we'll continue through the rest of the course of the week, to dig into some of the topics that we are introduced to in these calls. |
# / 00:06:10 | Richard Brown | So here's what's been going on. Last week, should I break down the timelines? We were alerted to an exploit in the Maker governance contract from our friends and partners at Coinbase and Zeppelin and we released an update to the contract and a notification to our subreddit, that explained what the mitigation steps were. |
# / 00:06:36 | Richard Brown | This, unfortunately, interrupted the cadence of the voting system that we had worked together so diligently to get kicked off. There's a new update that has just been posted as of 12 minutes ago, which provides some additional color. People are free to read that at their convenience. |
# / 00:06:59 | Richard Brown | What this new update does, says in a nutshell though, is that here's the nature of what happened, the exploit, here's some information about the scope of that exploit. But what's most important to us in this call is the next steps. |
# / 00:07:15 | Richard Brown | We've moved to the new contract, the new voting portal is pointing to the secure version of the contract. That contract behaves exactly the same way as the old one did, minus some functionality that was extraneous. So we're in a position now, where we can restart the governance process, because in this most recent notice we've published the source code and this is part of larger sort of crypto anarchic principles that we all sort of adhere to or at least we're building upon. |
# / 00:07:44 | Richard Brown | And that is, don't trust, verify and so now, our community is in a position to verify. The fact that the new voting portal contract is secure, they can take a look at the audit by Zeppelin, they can take a look at the audit from another team called TechShield, they can view the source code themselves on Etherscan and make sure that it satisfies their own particular requirements. |
# / 00:08:09 | Richard Brown | And now, we're comfortable, MakerDAO is comfortable with inviting people into the governance process once again. It's time to kick that thing off. |
# / 00:08:19 | Richard Brown | I'm not known generally for my overwhelming optimism, but this does provide us with an opportunity to give some thoughts to some things that we've been... or the way that the cadence has been working up until now. And here's some things that I want to put on the table, because we're in this position where we have kind of a blank slate, when it comes to kicking off the scheduling and the cadence and the rate at which we do the polling and the voting cycle. |
# / 00:08:50 | Richard Brown | In this call, I'd like to get a sense of whether everybody is sort of comfortable with the way things have been happening right now or whether we need to seize upon this opportunity to provide some improvements to the system. I want to remain as agnostic as I can when describing the possibilities ahead of us and leave it up to the community to decide, but here are some of the things that we've been thinking about internally. |
# / 00:09:10 | Richard Brown | For context, we put a significant amount of work into ensuring that we can continue this seven-day governance cadence, where on Mondays we initiate a poll, that poll provides a default range of options, depending on the state of the peg. If the peg is below, then the options available to people for the stability fee are in the positive integers. If the peg is above one, then we would flip that and we'd offer a poll with the range of zero to -4. That's the system that we have right now. |
# / 00:09:51 | Richard Brown | On Thursdays, the poll ends on Thursday morning, we take Thursday to have a call and discuss things, chat about it in subreddit, then on Friday, we have an executive vote. That executive vote replaces the state of the system and we continue this cycle on Mondays. |
# / 00:10:08 | Richard Brown | We've run into some situations though, where... some situations. We ran into a situation where we were having trouble wrestling with the idea of what the current state of the system is or what the preferred state of the system is. We use a continuous polling mechanism and that means that there's traditionally two pictures of what MakerDAO could look like in the executive. One is the existing and/or old picture and one is the new and/or proposed picture. |
# / 00:10:40 | Richard Brown | Happily in the past, we've seen sort of an overwhelming rush to the newest option that's been presented in the executive, which creates some confidence in the system and sort of lulled us into this belief that every time we have an executive, because of the weight of the poll, that signaled the will of the people. And then that got ratified. |
# / 00:11:03 | Richard Brown | A couple weeks ago, we ran into a situation where it didn't get ratified until late Sunday. And so it created this ambiguity between what is the preferred picture of the system, is the old state preferable or are people just getting on holidays or something? And just haven't gotten around to ratifying the new one. |
# / 00:11:23 | Richard Brown | So crisis ambiguity. Ambiguity is fine in principle, it's not fine if we need to have a poll the next day that's based on the picture of the system that everybody seems to agree on. So there's ambiguity here that we still haven't quite dealt with and it's kind of, in over the last few governance calls, it's faded into the background because we didn't run into this issue again. People have been happy to move the majority of the stake weights into the new executive, generally on Friday. Like I said, we haven't addressed this. |
# / 00:12:01 | Richard Brown | That's a long-winded description, to create some context around where we are right now. We've stopped the governance cadence, in order to update the contracts, we're in a position to restart them now and so there's options on the table. |
# / 00:12:13 | Richard Brown | Those options come down to what the cadence looks like, the length of the cadence and also, what occurs during the cadence and on what date. I'm going to read out two options... sorry, Cyrus, did you want to jump in? |
# / 00:12:29 | Cyrus Younessi | I also just realized that because the executive vote doesn't have a defined or has an unpredictable end to it, no matter which day we started on, even if we started on Monday morning, for any week that you have a drop-off in voter turnout from the old executive vote, you can never pre-plan for having the executive vote. And in time. So that also throws another wrench into it. |
# / 00:12:58 | Cyrus Younessi | In fact, by definition, you can't ever have a defined cadence for when the governance polls will run. |
# / 00:13:04 | Richard Brown | I agree and there's definite ambiguity here and like I said, we were lulled into this kind of idea that, oh, there's a new executive, new is better than old, so therefore everybody's going to jump on new, but that's not what continuous approval voting means. |
# / 00:13:18 | Richard Brown | So this is an unsolved problem that we still need to address. There's been some ideas about how do we decouple the ranges that are in the polls from the executive altogether? So in a poll, you could just have -100 stability fee to +100 stability fee and then presumably people would just arbitrarily pick some kind of range or that range would be... |
# / 00:13:43 | Richard Brown | Anyways, I don't want to dig into all of the possible variations of that issue that we need to address. More surface the fact that we still need to figure it out. |
# / 00:13:54 | Richard Brown | What I want to do in this call before I monopolize the entire thing is figure out if we're happy with the cadence and if we're happy with the day in which these polls happen. So the two options that we've been tossing around, in no particular order of importance, is we continue with the seven day cadence, where we... there's always a vote happening, there's always a vote or at least a cycle happening. |
# / 00:14:19 | Richard Brown | And this allows us to work in this model, where the assumption is that there's no magical stability fee that will be found that immediately solves the peg problems that we're creating forward. Guidance, that the community is ahead of this curve and as the rest of the ecosystem catches up and the peg begins to stabilize, we can begin to wind these things down and this follows the model, it is better to overshoot than to undershoot. |
# / 00:14:48 | Richard Brown | And part of that cadence is, we have in the seven day cadence, we have like I said, a poll on Monday, then it lasts for three days, then we take a day off and then we have an executive on Friday. The question here is, are we losing engagement with the executive on Friday, because people have sophisticated custodial solutions, their hardware wallets are locked in a safe in their office or some filing cabinet somewhere? |
# / 00:15:13 | Richard Brown | Or people just have things to do on the weekend, they don't necessarily want to be spending all their time thinking about Maker like I do. So there's some options on the table or we could have that seven day cadence and potentially start a poll on Tuesday and then have it run for three days, start the executive on Friday. |
# / 00:15:31 | Richard Brown | We could have a poll on Monday and then have it run for three days, but on Thursday morning immediately begin the executive and we can talk about the executive in these calls, instead of taking this sort of arbitrary 24 hour window, to think about things. We could also have polls for two days, so we just start a poll on Tuesday, have it go to Wednesday, Thursday, on Thursday, have an executive and let that run to the next Tuesday. |
# / 00:15:59 | Richard Brown | I'm not expecting people to come up with a lot of options right now, but I want to put these things on the table. We can continue it in the reddit discussion thread that I've posted. |
# / 00:16:07 | Richard Brown | The second option we've been thinking about or at least something that could be considered is, what if we had a 14 day cadence? What if we polled for a week and then had an executive for a week? That Cyrus, touches on the issue that you were just mentioning where we don't eliminate ambiguity, but potentially over the course of seven solid days, we have a much clearer picture of whether the executive was a clear winner or not or whether there was a lack of interest. |
# / 00:16:35 | Richard Brown | A 14 day cadence would also help some of the doves in the system, in the ecosystem that want to wait and see and what the effect of the peg is, stability fees have had on the peg, as opposed to the hawks, who would prefer to aggressively move forward and overshoot. Those are the two different camps that need to be addressed. |
# / 00:16:56 | Richard Brown | The other thing that I want to put on people's radar is another idea where right now, we have this range of polling options from 0 to 4%. There's a possibility that there's additional confusion created by this 0% option, because it's weird when you think about it. Presumably, if somebody votes for a 0% option, they are signaling that they think the system is fine the way it is right now and we shouldn't change it. |
# / 00:17:24 | Richard Brown | If that is overwhelmingly selected at any point in the future, what we've ultimately decided though as a group is that we don't need to have an executive at all. So it's worth considering that the options that we have in these votes, in the polls, are no change, as in no executive as the first option and then the second option is 1% increase, 2% increase, 3% etc. That would eliminate the need to have these extraneous executive votes that may be weak votes, that don't necessarily reflect the will of the MKR holders. |
# / 00:18:03 | Richard Brown | All right, that was a data dump and I'm sorry if that was a little boring. I will recap these thoughts in that reddit thread. Take away this though or this, we've updated the release of the source code for the updated governance contract. There's two audits that will be publicly accessible sometime in the next 10 or 15 minutes, depending on how many people DM me while this call is going on. |
# / 00:18:29 | Richard Brown | We have an opportunity to, we're taking this opportunity to invite people back into the governance system. We feel it's safe, we're voting with our MKR... people are capable of... have the opportunity now to view those audits and look at the source code themselves and rejoin the governance process. |
# / 00:18:51 | Richard Brown | Third or fourth, I'm not sure what number I'm on, items to think about, do we need to change the cadence? And if, so which one? If not, do we kick off on Monday with just the same old poll strategy that we've had up until now? |
# / 00:19:08 | Richard Brown | All right. I'm going to leave it there. |
# / 00:19:11 | Cyrus Younessi | Are we going to do any community discussion right now? |
# / 00:19:16 | Richard Brown | I was just about to do that. I would love some comments. Akiva, what do you have to say? |
# / 00:19:24 | Akiva Dubrofs.. | On the point of extending it to a 14 day cadence to satisfy the doves, the thing is there were doves out there on the calls like months back, who were saying that we were already going to be overshooting it by doing these 4% or 3% hikes. And they've actually been proven wrong. |
# / 00:19:41 | Akiva Dubrofs.. | So I think the doves actually have a poor track record in this whole thing and because of their poor track record, we shouldn't be really looking to help them out with a 14 day cadence. |
# / 00:19:52 | Richard Brown | That's an interesting point to make. I guess we can avoid saying that the doves are wrong, wrong, wrong, but we could definitely look at the numbers and once risk and Vishesh come in with some input, come up with some kind of a rough consensus about whether the stability fee movements have been successful or not. |
# / 00:20:13 | Richard Brown | And one of the things we could potentially do, if we move to a 14 day cadence is double the range, because one of the concerns will obviously be that 14 days will not give us enough opportunity to approach the stability fee as quickly as we need to. We could have the options of 2%, 4%, 6%, 8% increases every two weeks or 1% increments, all the way up to eight or 16 or who knows? There's a lot of options on the table. |
# / 00:20:43 | Richard Brown | I don't have any strong feelings one way or the other, so I think that it's going to be up to the level of interest and engagement from the community and people on this call right now. And the level of engagement that happens in the MakerDAO subreddit. If there isn't an outpouring of interest and/or suggestions about what we should change, we will go back to the polling mechanism on Monday and just sort of continue to do things the way we have been. |
# / 00:21:12 | Richard Brown | Louis, did you want to say something? |
# / 00:21:13 | Louis Aboud-Hogben | I just wanted to make the comment, you said that if DAI is trading below the peg, then the only options that will be presented for voting are to raise the stability fee. I just want to present the scenario where the stability fee is already quite high and the trend of a DAI recovery is really strong and obvious and we may want to lower the stability fee, even when it's below the peg, just to get in front of the curve. |
# / 00:21:45 | Louis Aboud-Hogben | Like if the stability fee is up at 50% or 60% and we're lowering it in increments, it could take a long time to get back down to a reasonable level. And if we can see a chart that DAI is trading up over a series of weeks, then we may want to get in front of that. |
# / 00:22:03 | Louis Aboud-Hogben | I think it's, eventually, you'll want to be less reactive and more kind of proactive with the changes in the fees. |
# / 00:22:13 | Richard Brown | I agree. Thanks for bringing that up, because this is a risk that I failed to identify. I thought I was getting ahead of this. When we presented that 0 to 4% increase, it felt like an arbitrary imposal of a range from the foundation and I can assure you that is not what happened. It was just the numbers that were presented in a reddit thread, nobody questioned them, it encapsulated the range of all previous rate increases that we've considered in the past and it just stuck. And also, largely it's a UX issue, because having 8, 12, 16 different options in the polling interface is kind of a nightmare to manage and display. |
# / 00:23:00 | Richard Brown | But you've raised a good point, it might be time to just simply use this opportunity to figure out, is the range that we should have been using, is this a good time to just do -4 to +4 always? |
# / 00:23:13 | Louis Aboud-Hogben | I think it'd be good to always have one small decrease option in the polling, like whether it's down 2% or something, just to gauge whether anybody's thinking that that could be appropriate. |
# / 00:23:29 | Richard Brown | Someone had raised this issue in the chat or the subreddit, that perhaps there's some unconscious biases being applied to voters when they see, perhaps somebody's a moderate and they see a range of numbers and because they're hardwired as being a moderate, they're going to pick whatever is in the middle. So the actual display of that data is somehow influencing the way people voted. |
# / 00:23:54 | Richard Brown | I'm not sure I completely agree, I like to think that after people have run the gauntlets to actually get into this portal and they have a significant amount of MKR that they're going to stake, that they're not bamboozled by ranges of numbers. But that's something to consider, that perhaps if we did have -4 to +4 and then we found that the votes trended to 0, whereas right now, they're trending to the midpoint of 0 and 4, then there might be some mechanism at work that's just sort of confusing people there. |
# / 00:24:26 | Cyrus Younessi | We still have to figure out if in the future we want this to be automated or not. I mean by allowing negative options or lowering the stability fee option, that's essentially saying "we don't ever want this to be automated". |
# / 00:24:42 | Cyrus Younessi | Because I think an automated process would look at the DAI price and then just decide if the fee needs to be increased or decreased. |
# / 00:24:51 | Richard Brown | That's a huge can of worms. I'm not sure that we're in a place to dig into it yet, but I think that the general assumption is that over time, as we aggregate more data, as we have some more sophisticated models and the risk function has been fleshed out a bit more, that we will be at this situation where either we have an algorithm that gives us a temperature reading and then we use that algorithm to influence what the voting process- |
# / 00:25:17 | Cyrus Younessi | I'm not advocating one way or another, I'm just saying that there's still this open discussion of, are we humanizing or manually doing a process that we expect to be automated one day? And we're doing that because it's kind of the prudent and cautious way to move forward. Or is this a kind of governance process that we want to be, have our human fingerprints all over, long term? |
# / 00:25:43 | Richard Brown | I think the answer to those questions are TBD. I think as we have these calls, we're figuring out what governance looks like for us and what I would hope the most conservative and cautious way humanly possible, until we get the confidence in the algorithms and implications of having, relying on computers to achieve homeostasis for a peg for us. I think that's probably where we're going to get eventually, I just don't know how soon or how much effort it's going to take to get there. |
# / 00:26:18 | Richard Brown | Casper, "is the point of Maker..." do you have access to mic, do you want to ask that question? |
# / 00:26:30 | Kacper Wiekit | I'm not quite sure what the kind of governance process may be able to maximize the cash flow for Maker Burn, because there is one goal and there's another goal and I'm not quite sure how both goals can be achieved in the same time. |
# / 00:26:54 | Richard Brown | That's a super interesting question. I don't know, but here's what I think I do know, that it's my general impression that the MKR voters understand that the number one priority for the ecosystem is a stable peg. And that all things sort of flow from that. |
# / 00:27:13 | Richard Brown | One of the most encouraging things I've seen recently with our governance process is the fact that MKR holders consistently are making these hard decisions, obviously, raising the stability fee is not popular. But they understand that if sentiment hits and whatever impact that might have on whatever is the publicly trading price of MKR is, is worth accepting now in hopes or the understanding that this is strengthening the entire system for the future. |
# / 00:27:45 | Richard Brown | I get a great deal of satisfaction of seeing that people are still willing to make those hard decisions. I would love to hear some other opinions though. Is the assumption generally that strong DAI equals strong MKR eventually, but they're not tightly coupled in the short term? |
# / 00:28:11 | Richard Brown | Lawson, did you have a suggestion or comment? |
# / 00:28:18 | Lawson Baker | Yeah, my comment was more towards like trending towards the automation and then above that, I kind of gave an idea over treating it similar to a difficulty adjustment, where you're looking at price oracle relative to what else is happening in the system. Obviously, you need like more modeling to be able to do that, so you couldn't do it today, but you could see that being something you'd trend towards. |
# / 00:28:42 | Richard Brown | I think that's the general assumption, as we become more sophisticated, we have more data, trustful data and more insight into the system, more understanding of what the confounding variables are. Then, algorithms become more viable than they are right now. |
# / 00:29:03 | Richard Brown | Sorry, we're at the halfway mark and we haven't gotten into risk yet. And I feel like that's where a lot of the value is, so to summarize, we have some interesting things happening in the governance world. The interesting questions are being asked about the cadence. If there is significant interest and/or input provided, we can revisit it. Otherwise, on Monday, we'll kick off another poll, that looks exactly the same as the ones we've had in the past. |
# / 00:29:29 | Richard Brown | All right, Cyrus, I'm going to hand this off to you for risk. |
# / 00:29:31 | Cyrus Younessi | To Kacper's question, I guess we can revisit that after the call. I'm happy to talk about that. |
# / 00:29:39 | Richard Brown | Yeah, that'd be a great post-call discussion about what's more important, DAI or MKR? |
# / 00:29:47 | Cyrus Younessi | I think today's risk agenda is going to be somewhat similar to last week's. We're going to hear from Vishesh, who is going to run through the state of the peg, then afterwards, it's going to be primarily open discussion. A few topics we can talk about, kind of sourced from the various social media forums over the past week, but no strict defined agenda unless we need one. |
# / 00:30:15 | Cyrus Younessi | Vishesh, if you're ready. |
# / 00:30:18 | Vishesh Choudhry | Sure. Share my screen. So just the broad-strokes to start, I think I have tried to hold a fairly conservative view up to now. I think it is fair to say that like in the past week or two, we've started to see some really positive indicators and I'm okay with coming to that as a conclusion. |
# / 00:30:56 | Vishesh Choudhry | I think the other thing is, broadly, I'm working towards getting to a good, observed model that describes what's going on with DAI price and supply and then once we feel confident that we can describe what's been going on, then we can start to work toward some of the goals that people have mentioned on this call, in terms of getting to more automated management of the system. |
# / 00:31:22 | Vishesh Choudhry | So as far as what's been going on with DAI price, stability fee increases have been very strong from around mid-April, up till a few days ago and the DAI price peg has been holding relatively steady through that point. There were some sort of periods of up and down, Tether news etc, but broadly, in terms of volume weighted average, it's been holding pretty steady. |
# / 00:31:49 | Vishesh Choudhry | And then in the past week or so, I would say it's trended up a bit. And then sort of in between in the variance, it's hit the peg at certain points and even shot above it, like the day of the Tether news. So for me, what this says is at least what's been going on with the DAI supply and combined with the stability fees has had a measurable impact and it seems that DAI is at least trading close to the peg and it's unclear if we continue to increase the stability fee, if it would necessarily reach the peg. But at the very least, it's definitely hovering close to. |
# / 00:32:35 | Vishesh Choudhry | So I had started to look, as I mentioned, into the DAI price and the ETH price. We've seen this relationship I think hold fairly consistent and we've mentioned this a few times, ETH price being a big driver for DAI makes sense, in the sense that what's going on with ETH drives the sentiment and desire for leverage, which is a big use case for Maker. |
# / 00:33:02 | Vishesh Choudhry | Now, one thing here is, this is just the simple moving average, one of the other components that I'd looked at was the ratio of the simple moving average to, in the short-term, to the simple moving average in the long-term. So let me just show that here. |
# / 00:33:21 | Vishesh Choudhry | And I had started to run some regressions on this, but I know I'm getting a little bit ahead of ourselves here. That had shown an even stronger relationship. The reason that this to me is an important metric, is because it starts to speak to effectively in the short term, how cheap is ETH trading to what the long-term trend is. |
# / 00:33:42 | Vishesh Choudhry | And so that makes sense in that as ETH is trading cheaper, you would expect more bullish sentiment and so more folks could be selling DAI to get leverage on ETH. What's interesting is I think this sort of relationship has held steady, but has been a bit broken recently and so DAI price, it's hard to say whether it's due to Tether news or whether it's due to stability fee increases, has maintained kind of its strength in the face of what's been going on with increasing ETH price. |
# / 00:34:16 | Vishesh Choudhry | So just go back to here, so ETH price has been consistently increasing over the past week or so, everyone's been very excited. But at the same time, DAI prices held relatively steady or even increasing in that time period. So whether that's due to stability fees or whether that's due to Tether news and other factors is very hard to say, but it is definitely a positive indicator. |
# / 00:34:39 | Vishesh Choudhry | So DAI supply has been consistently coming down with the recent stability fee increases. I think this is the most clear and measurable impact of the stability fee changes and then we can sort of talk about what's the relationship between DAI supply and DAI price? And I think that's a little bit more complex, but it is to me, fairly clear that the increasing stability fees has had an impact on DAI supply. |
# / 00:35:06 | Vishesh Choudhry | There have been very few large draws recently and there have been some significant wipes about two weeks ago and then three weeks ago. So I think this is a fairly clear trend. |
# / 00:35:22 | Vishesh Choudhry | I had mentioned briefly last week, sort of what's been going on with the age of debt and I think this pairs well with what's been going on with supply. Effectively, I think the circulation of debt has improved with the recent increases in stability fee. |
# / 00:35:37 | Vishesh Choudhry | So this is another to me clear impact of the increasing stability fee, as opposed to DAI price, which is a little bit harder to draw a strong correlation with. But essentially, as the stability fee has come up, for the first time in basically ever, the average age of debt that is open has been decreasing and the average age of debt that's being closed is increasing. |
# / 00:36:03 | Vishesh Choudhry | So circulation of debt has now hit a point that it never has in the history of Maker. So to me, this is interesting. |
# / 00:36:12 | Cyrus Younessi | What's average age of debt being closed, what does that mean? |
# / 00:36:18 | Vishesh Choudhry | So effectively, as DAI is either liquidated or paid back, what the age of that DAI was at the time of being closed out. So effectively that means for the first time, people are okay with... there have not been very many liquidations recently, so it's mainly driven by wipes. People have been more okay and increasingly okay with paying back old debt than they ever have been before. |
# / 00:36:47 | Vishesh Choudhry | Up till now, it's kind of been silently accruing older and older. I can show you in the long-term. |
# / 00:36:53 | Vishesh Choudhry | So basically, the trend up to this point has been that DAI has been getting older and older, so people have just been sitting with that DAI and not paying it back, but recently, that trend has started to reverse itself. So that to me is a very important health indicator for the system, because I think it will be important to maintain a strong circulation to achieve homeostasis as Rich mentioned. |
# / 00:37:18 | Vishesh Choudhry | So there's been a lot of discussion about these secondary lending platforms, their role with Maker. I think there's not a lot new to report here. The rates and volumes have held fairly steady on these secondary lending platforms, kind of hovering around that 14, 14.5 level. So effectively, what this says to me is the Dharma, Compound, NEO was mentioned. A few of these kinds of platforms have really nestled into that secondary lending relationship with Maker, where people would tend to refinance their loans onto those platforms or would instead try to purchase DAI and lend it out on those platforms. |
# / 00:38:04 | Vishesh Choudhry | There's been a lot of discussion about whether that actually positively impacts DAI price, I don't feel the use case of people lending that DAI out on those secondary platforms, necessarily improves the DAI price, because it doesn't overall impact the supply. But what is interesting is the amount of DAI that gets locked up in those platforms, that is not necessarily lent out is generally a positive for the circulating supply. And then a positive in the sense that it's good, but it's a negative. And then the other factor is, if people are taking out loans and refinancing them on those secondary platforms, that does not have an overall effect on DAI supply, but could be a positive for DAI price. |
# / 00:38:51 | Vishesh Choudhry | And then I don't want to eat up too much time, draws have been low, so there's not much to say about new versus old. The collateralization ratio, so this has been interesting, where as I mentioned, DAI supply has been consistently going down, but the collateralization ratio has been kind of consistently increasing over the past week or two. |
# / 00:39:14 | Vishesh Choudhry | So the total amount of DAI has been going, down the denominator has been going down, but the ratio has overall been going up. So the amount of frees has been outweighing the amount of locks recently. So effectively, people have been taking more collateral out, but not as much as the DAI supply has been decreasing, which kind of makes sense, in that people are going to hold their collateralization ratios within a relative range. But they've been okay with deleveraging the amount of DAI that they have relative to that ETH. |
# / 00:39:49 | Vishesh Choudhry | So again, that is to me a positive indicator because as people seek more and more leverage on Maker, that depresses the DAI price. So this is potentially a sort of intermediate variable that we can observe, that is not the independent driver, but definitely, I think has a relationship to what's going on with DAI price. And that people are sort of maybe letting their foot off the gas a little bit, in terms of how much leverage they're seeking. |
# / 00:40:16 | Vishesh Choudhry | Not much to say on the individual transactions. So if anybody has specific questions, I'm happy to double-click into something otherwise, otherwise I'll hand it back. |
# / 00:40:44 | Cyrus Younessi | Lawson had a question about the age of DAI calculation. And is that equivalent to the age of a CDP? So maybe you could talk a little bit about your methodology on how you calculate that. |
# / 00:40:58 | Vishesh Choudhry | Sure, yes, so the age of DAI is not fungible with the age of CDPs. So a CDP, like often CDPs could be opened early on and have very little activity and so in that sense, that CDP would be very old, but if a large quantity of debt was drawn recently, the debt, the age of the DAI would be very low. So they're not the same at all. |
# / 00:41:26 | Vishesh Choudhry | And I actually tend to think that the number of CDPs, the age of CDPs, the amount of activity in CDPs is not a super important metric. I think it's more interesting to track when people are kind of actually transacting and putting their money where their mouth is, but the pure just CDP stats, I clicked a few buttons and didn't do anything with it. It's not super important to me. |
# / 00:42:02 | Cyrus Younessi | Cool, well, thanks for that. That was actually a really good presentation, probably my favorite one so far. Awesome. |
# / 00:42:12 | Cyrus Younessi | So I guess the question is, what does this mean for the stability fee? And I have some thoughts about... I mean I know there's been a lot of open discussion on how high the stability fee has to go before DAI hits a dollar. And first question is, can we glean that information from the data so far? And if not, how can we start thinking about it? |
# / 00:42:44 | Cyrus Younessi | The point I'm trying to make is kind of this notion of the elasticity, where no matter how high the stability fee goes, DAI just won't hit $1 for various reasons. So I think that might be a good place to kick off conversation. Anyone have thoughts on that? |
# / 00:43:14 | Vishesh Choudhry | So I can chime in real quick. I think time is an important variable to look at here too, because like there is a very real consideration that in the long term or over a reasonable time scale, DAI would not sit at $1, but we know like in short time bursts, DAI does hit $1 or go above. |
# / 00:43:39 | Vishesh Choudhry | It's really a question of like, in a longer time scale, I think there is a very real potential. We haven't seen it happen up to this point, that given the transaction costs that are involved, that DAI or the risks involved, DAI might not necessarily be ever equal to $1. |
# / 00:43:59 | Cyrus Younessi | Right. And I think that's definitely a very real possibility. So how do we deal with our reliance on backward-looking data? And we've noticed this trend in the reducing, the reduction of the DAI supply, stability fees are going up, seems to be having a measurable impact there. |
# / 00:44:25 | Cyrus Younessi | I know there's a lot of people that have concerns that no matter how low the DAI supply goes, it won't get DAI equal to $1 exactly. |
# / 00:44:36 | Richard Brown | Can we talk a bit about what that means though, because there could be a significant number of reasons behind that. And here's where I like to tell everybody, please explain it to me like I'm five. So is this primarily the friction of getting the premium associated with the friction of getting in and out of DAI, is the reason why perhaps 98.9 or whatever is the real price of DAI? |
# / 00:45:00 | Cyrus Younessi | No, I think it's not the transaction cost friction, but rather the nonequivalence of DAI and an actual US dollar, in compositional terms. I mean one DAI is not quite literally $1, it's designed to track $1 and if there's always going to be that lapse in confidence, that little gap where these two things aren't identical in makeup, then there might be some risk premium there that needs to be filled up. |
# / 00:45:34 | Richard Brown | Isn't that sort of a function of what people are using? It has to do with demand though, so if people are using DAI as collateral or as a transaction or a settlements token or something, then yeah, there's a lot of friction there. But if somebody wanted to get paid in DAI and pay their subscriptions in DAI, then that becomes a significantly different picture, doesn't it? |
# / 00:45:57 | Cyrus Younessi | Right, so the open question is, if we've maxed out how high the demand can go, without kind of alternative monetary policy tools, such as the DSR, is it worth considering either... I don't know, doing something with the stability fee, maybe some would argue to not increase it anymore or maybe some people think we're still on the right direction. |
# / 00:46:24 | Cyrus Younessi | To me, it's kind of an open question. I was hoping some people had some interesting thoughts on that. |
# / 00:46:37 | Vishesh Choudhry | I mean from a statistical standpoint, one of the challenges I think to do regressions and draw strong correlations on some of this data, it's just been that time is an enemy and that we've made a very large quantity of changes in a very short time period. And so like the more time that we have to observe what the impact is of the stability fee, versus ETH price. |
# / 00:47:05 | Vishesh Choudhry | So effectively like I'm thinking of stability fee as like an interacting variable between ETH price and DAI price and so I think it just fundamentally modifies relationship between those two variables. But there's a lot more investigation that's necessary to confirm things like that. |
# / 00:47:22 | Vishesh Choudhry | And the more time that we have with this data, the better I think we can get to conclusions. So there is probably a very real question of how much pressure do we feel to need to make changes, versus taking a beat and observing what's going on? |
# / 00:47:41 | Cyrus Younessi | Yeah, that's a great point. I don't think we can know that before the fact. I mean we have been moving at a pretty steady, but maybe I mean it's been quite a rapid increase over the past three, four months. I don't think it's obvious to me that we should either slow it down or keep it at the same pace. |
# / 00:48:07 | Cyrus Younessi | But I'm interested in hearing if there are any ways we can at least reason about it and at least come to some sort of rough ideas on whether the stability fees should just keep going up until DAI hits $1 or not. |
# / 00:48:25 | Louis Aboud-Hogben | I have some thoughts. I think the trend is really strong at the moment, we've sold a bunch of our DAI at some good prices and obviously, DAI supply is coming down. I think to keep raising it from this point, when the trend is this strong, but I would look at the trend and use that as the guide, the DAI price has been coming up, things seem to be working at current levels. |
# / 00:48:52 | Louis Aboud-Hogben | If we go too far up, in terms of stability fees, it's going to take us a long time to get back down to reasonable levels. I think we have good reason to just be more steady at this stage and see if the trend holds, rather than constantly ratcheting them up, which is going to have some other effects on the whole ecosystem. |
# / 00:49:19 | Louis Aboud-Hogben | I just say look at the trend and take some comfort in that, because I think you can overshoot with these things. |
# / 00:49:29 | Vishesh Choudhry | So just to double-click on that, because I'm really curious about what people think the risks are associated with overshooting and how, like maybe some kind of reasons for why we think that would be difficult to unwind. Because I'm not personally seeing where there is significantly large risk in overshooting the stability fee. And then ratcheting it back down, if we do determine that we've overshot. |
# / 00:49:53 | Louis Aboud-Hogben | I think the one thing is like just knowing where the liquidity provider's inventories are, because once that all dries up, then there's not going to be a lot of offers in order books. So I think, yeah, it would be more about just like managing the liquidity of the system at this early stage, that like everybody, Maker, we and a bunch of other people are still sitting on a bit of DAI, but as soon as that's all gone, I think you will start seeing it trade above $1 or at least trade on really thin books, which is probably not what we want at this stage. |
# / 00:50:39 | Louis Aboud-Hogben | I would agree with your sentiment, generally, but like we shouldn't be too afraid of those kind of outcomes, but if we can manage it and get ahead of the curve... I think it was kind of the reactionary approach that got us into this situation in the first place, where Maker's inventories went to zero in December and the stability fee was at 0.5%. It took us a very long time to get the stability fee back up to a reasonable level, even though the trend around supply and demand for DAI was pretty strong for four three months or so. |
# / 00:51:17 | Louis Aboud-Hogben | Now, we've got a really good trend that we should I think be taking advantage of, rather than constantly moving in the same direction and kind of backing ourselves into a corner. |
# / 00:51:32 | Richard Brown | As is usual, the conversation is just heating up as soon as we reach the top of the hour. So I want to remind everybody that we're going to continue the chat after Steven gives us his final thoughts. The call will continue to be recorded, so if you're interested in talking about money and crypto and nerd stuff, please hang out. |
# / 00:51:51 | Richard Brown | Steven, did you want to leave us with something? |
# / 00:51:52 | Steven Becker | I kind of wish that this could be two hours long, but I suppose if it was two hours, we'd go into two and a half hours and then so on and so fort. There's no escaping this, but I mean the level of dialogue has gotten to the point where I think Rich would agree with me, is exactly where we want to be. We want to be constantly talking about these particular issues and thinking about them in terms of models. |
# / 00:52:18 | Steven Becker | But also please keep in mind, make sure you have enough bandwidth left over to consider upcoming collateral portfolio or how we look at that, what's the risks around that. And then also being very cognizant and I'll always repeat this, that this is a decentralized governance application that we're looking at. This is really trying to figure out the best way of keeping the momentum going, so that everyone is obviously very well informed. |
# / 00:52:50 | Steven Becker | I do feel like sometimes, I'm the credits at the end of a film or something like that, but it's essential to everyone to be aware of the continuous nature of this thing, where your attention should always be on looking after the robustness and integrity of the system. And that's me. |
# / 00:53:10 | Steven Becker | So thank you very much for joining again. I certainly do appreciate a participation. |
# / 00:53:16 | Richard Brown | All right, thank you, Steven. Let's continue talking about crypto and money though. Sorry, Louis, are you capable of picking up your thoughts where we cut you off? |
# / 00:53:30 | Louis Aboud-Hogben | I pretty much said what I had to say, it's just that like with weekly governance calls and the expected increments for stability fee changes, it's going to take a long time to, for instance, affect a change back down to an 8% stability fee. And if that's what's necessary over the medium term, then rather than continually ratcheting it up, when we've got a strong trend in the DAI price improving, I think we can just sort of hold it steady and not put ourselves too far in one direction. |
# / 00:54:08 | Cyrus Younessi | I think the overshoot is easier to deal with, as long as we can reliably keep increasing the debt ceiling. I think it's a lot easier to incentivize kind of the more general non-market making retail to generate DAI to sell it above $1, than it was too incentivize the reverse. |
# / 00:54:35 | Cyrus Younessi | So I am less worried about it, but I will admit that certain days, we did see a ton of DAI being generated in a single day. And right now, we only have about roughly 20 million of debt ceiling room. That could easily get cleared up in a couple days under the right scenarios. |
# / 00:54:54 | Cyrus Younessi | And then you talk about one to two week turnaround for governance poll and executive to increase the debt ceiling or lower the stability fee. I think the governance could become a lagging factor there. But I do think in general, it's an easier problem to deal with. |
# / 00:55:20 | Cyrus Younessi | Rich, do you think there's any merit in having some sort of a mandate for kind of, I don't want to say emergency, but like a spontaneous governance poll for a situation that requires immediate attention? |
# / 00:55:38 | Richard Brown | Yeah, I think there is, but we've run squarely into this problem in crypto, which seems to be kind of unique, where we don't know who's doing what and we don't know how to talk to them. So that's a big problem. We would need to have some kind of a robust polling or push method that would allow us to contact stakeholders and put this stuff on their radar. Or else we would have to like really promote this assumption that as a MKR holder, it's your responsibility to do like daily check-ins on what Maker's up to. I'm not sure that's reasonable. |
# / 00:56:13 | Richard Brown | I don't know how we would get the word out to enough people in time to make that stick. |
# / 00:56:25 | Vishesh Choudhry | But I will argue that like from an economic standpoint, I think it is easier to walk down the DAI price than it is to try to bring it up. Like even if you made the same kind of magnitude of changes in the stability fee, because the arbitrage case is so much clearer when DAI is trading above $1, I do agree with Cyrus that it is much easier to walk it down. |
# / 00:56:55 | Vishesh Choudhry | The way I see the situation is right now, we have a choice between kind of holding out and getting a little bit more data on what's the impact of the ETH price on DAI price at current levels, versus making more changes and trying to get data on what is really the upper bound of what we're talking about here. |
# / 00:57:31 | Richard Brown | Diego raised an interesting point. Diego, do you have access to a mic or should I read this out for you? He's circled back a bit on this idea that it's one of the basics, but it's something that I get hung up on frequently and maybe it's just me. |
# / 00:57:47 | Richard Brown | But it sounds like a silly question, but how do we know when the peg is stable? And that's something that I keep on trying to wrap my head around. And we see this a lot in chats and various socials, where people that are unused to watching grass for a living will do a brief check-in, look at the last 30 minute tick and then say, okay, we're screwed or everything's fixed. |
# / 00:58:10 | Richard Brown | And I'm trying to wrap my head around what kind of a deviation over time do we consider acceptable, what kind of window are we looking at before we, as a group, kind of go, okay, things are looking better now? As Diego pointed out, is it a deviation of X around $1 for one month? Is that a success metric for us? |
# / 00:58:36 | Richard Brown | I don't know, you see where I'm going with this. So how are people thinking about this problem? |
# / 00:58:48 | Kacper Wikiet | I would say that one can measure the peg is stable if there is OTC dealer that can like exchange $1 for one DAI with some fee that is equal on both sides. So it's like if you have access to open books from some guy who is doing OTC trading in transparent way, the quantity of inventory that if he's getting too much DAI, then it's like probably the peg isn't going well. And so one can shoot for like half of the capital in DAI and half of the capital in dollars. |
# / 00:59:32 | Kacper Wikiet | So there is website showing the DAI that is sold below $1, is like $16,000. And I think it's showing that the governance fee is still a bit too low, because it's not cleared yet. And I think it would be interesting to see, for a brief period of time, what will happen if the DAI price will be significantly above $1 for some period of time. |
# / 01:00:10 | Richard Brown | That's interesting because that raises the question of whether it's not just, it looks stable according to this graph, but we add some additional weighting metrics to this. So it's stable according to this graph and then we factor in the fact that there's this X amount for less than a dollar for sale on Coinbase Pro and this is the general sense of the inventory levels across four different market making desks. |
# / 01:00:36 | Richard Brown | Does that make sense, is that reasonable or is that missing the mark? So Cyrus maybe a question for you then we have the ability to look at something like Vishesh's awesome graphs and get a general sense of where things are at, but as Kacper pointed out, we also need to take into account just the order books that are out there and the market making desks as well. |
# / 01:01:07 | Richard Brown | Are those sort of leading or trailing indicators of what we see on the graphs or are they something that would color our interpretation of those graphs? |
# / 01:01:26 | Cyrus Younessi | I think they would be, I think they would be a leading indicator of sorts. I mean the OTC desks are probably the first to respond and adjust to various policy actions. |
# / 01:01:48 | Richard Brown | So it's this sort of mysterious lack of access to business intelligence from these desks, that are valuable for everybody to know. |
# / 01:01:56 | Cyrus Younessi | So if they were also CDP holders and they might be the first to... not CDPs. If they were pure market makers, they might be the first to buy up DAI in expectation of increasing DAI price. |
# / 01:02:12 | Cyrus Younessi | But I mean I think that's a pretty big presumption that we can just get OTC data and just kind of plaster it on a website for everyone to see. I'm not sure, maybe it is, but I just don't think that's immediately obvious how we can... |
# / 01:02:29 | Vishesh Choudhry | Yeah, so just to chime in on that real quick, from a data perspective, like I think order book data particularly gives you a sense of liquidity and that is an independently measurable, helpful metric. The specific like depth of the order book at any given point in time is to me, just like an increased degree of granularity, particularly on a time scale of like what is going to happen to the price. |
# / 01:02:57 | Vishesh Choudhry | Because if there's a liquidity crunch on DAI in a particular moment, then we could anticipate that there will then in subsequent moments be an increase in price. So that kind of thing is where it could be a leading indicator, but I mean I think first and foremost, it just tells you how liquid DAI is, is the most important thing. |
# / 01:03:16 | Richard Brown | Okay, that's interesting, the other one that I forgot to list of these confounding variables, but I just posted a link to it and LoanScan has a page that they show what the borrow APR is across some of the other debt providers. The liquidity providers, debt facilities, I'm not even sure what they're called anymore. |
# / 01:03:36 | Richard Brown | But is that part of sort of this weighting algorithm, is the deviation of the stability fee against prevailing market rates? Would be able to factor into this. So if it turned out that over the course of a couple weeks or months, Maker once again was by far the most enticing option, when it came to minting DAI, would we factor that into current state of the peg? |
# / 01:04:15 | Richard Brown | I guess I kind of drifted off the at the end of that question, but it feels like there's market making inventory, there's a time window in which the peg has remained stable, plus or minus this percent. There's the competitiveness I guess index of whatever the stability fee happens to be at that moment, is what I was trying to get towards. |
# / 01:04:41 | Cyrus Younessi | I don't think we end up in a scenario where the stability fee is below the Compound rates. Again, I think we saw pretty clear demand for an arbitrage trade, where it was quite simple to borrow from MakerDAO and loan on to Compound and the reverse is not true. So I don't think we see that again. |
# / 01:05:13 | Cyrus Younessi | But I do have a somewhat related question and it kind of also circles back to another one of Diego's questions, where he says once we have DAI users and the DSR, why can't the price hover at $1? That's what I was trying to lead into earlier with, I don't know for sure, I don't think anyone knows for sure whether or not DAI can hover at $1 with DSR or without the DSR. |
# / 01:05:43 | Cyrus Younessi | But I am interested in hearing what people think a reasonable first estimate for what the DSR should be, once it is available. In the sense that, what do we expect the total stability fee for ETH to be if we had alternative policy tools at our disposal? |
# / 01:06:18 | Cyrus Younessi | Matthew, I know you just joined the call. You got any thoughts on that? |
# / 01:06:23 | Matthew Rabinowitz | I would always say, I mean I'm a general, I take the general view that the DSR is an exceptionally potent tool in gathering as much data as possible and basically, the equivalent of access to treasury yield. I mean 100 basis points is better than most financial institutions out there. If we make it anything significantly greater, we take a chance at having there be so much demand, which is a wonderful problem initially, but we need to step our way into it. |
# / 01:06:53 | Matthew Rabinowitz | So I would start at 100 and go very incrementally, say 25 basis points every other week, every month, until we find the sweet spot. |
# / 01:07:06 | Cyrus Younessi | So you think it's going to take some trial and error, we'll have to kind of tinker around a bit until we get to kind of the right range. |
# / 01:07:16 | Matthew Rabinowitz | I think it's going to take some trial and error, but also remember, every time this DAI is trading lower than $1, whatever the interest rate is divided by 0.99 or 0.98, so if we say it's 1%, but the price of DAI for whatever reason is 95 cents, it's actually the equivalent of higher than 1% yield. So the equivalent of a yield to maturity calculation. |
# / 01:07:43 | Matthew Rabinowitz | So I mean it'll have a strong effect, it'll have a lasting effect. I guess at the end of the day, the real point here is that when you're having a DSR or the stability fee, at the end of the day, you're just trying to control supply on one side of it by either incentivizing the removal of supply or penalizing the creation of too much DAI. We're just going to have to find that balance in there. |
# / 01:08:11 | Cyrus Younessi | So is it taken for granted that the total stability fee... |
# / 01:08:41 | Cyrus Younessi | silence |
# / 01:08:41 | Cyrus Younessi | Receiving different cash flows. |
# / 01:08:44 | Matthew Rabinowitz | Yeah, I mean let's step back. I think the question is not only where we're going to go in the first quarter, trying to trial and error this. I think the real, another way to look at this would be where do we want it to be in five years and back our way into how do we make the decisions that get us there? |
# / 01:09:01 | Matthew Rabinowitz | I mean one of the greatest values that I foresee of the DSR and using the stability fee in concert together is that we're going to find the market fit, where the DSR, the savings rate is in effect higher than most banks can provide for the average consumer. And the overall stability fee to borrow against your asset will ultimately one day, in five years, probably be lower than what you can get at a commercial bank. |
# / 01:09:29 | Matthew Rabinowitz | So if we kind of back our way into where those two numbers kind of aggregate into around, you pick a number, two, three, four, five area, the question really is, do we start off with a low stability fee and an elevated... excuse me, a low DSR and an elevated stability fee and then slowly ratchet up the stability fee? Or do we accelerate the decompression of the... accelerate the stability fee decrease? |
# / 01:09:56 | Matthew Rabinowitz | I don't know which one will have the most direct impact immediately. We're just going to have to try those both together, but I think the ultimate objective is to lower the stability fee back down to normal, more reasonable levels. But we don't want to destabilize the process by causing spurts in demand that causes to be more erratic. |
# / 01:10:19 | Cyrus Younessi | So I mean is it also taken for granted that the ultimate stability fee offered by MakerDAO can be competitive or even has any relation to traditional world products? I mean there's an argument that MakerDAO is a sufficiently different beast that it comes with its own set of pros and cons. And the benefits of MakerDAO system might require kind of different pricing of its services. |
# / 01:11:03 | Matthew Rabinowitz | I personally choose to believe the best market fit for us is to compete directly with financial institutions, both on lending and providing an interest-bearing product and grab market share. Because there's a direct market fit. |
# / 01:11:18 | Matthew Rabinowitz | I mean to buy and use DAI in everyday commerce, you shouldn't have to be someone who borrowed it. You can just be someone who bought it and is using it, choosing to use it in that same context. If you choose to buy it and then store it in the effect of a short-term CD that has a yield, much like a savings account, I think we should compete with that. |
# / 01:11:39 | Matthew Rabinowitz | silence |
# / 01:12:07 | Richard Brown | We're getting some dead air here. Are we running out of steam Cyrus or did you have a rebuttal? |
# / 01:12:13 | Cyrus Younessi | I'm still good to hang around, but up to everybody else. I mean I agree with the notion of obviously trying to be competitive. I'm just less certain that it's guaranteed to happen and that we should try to force the system to go a certain direction. |
# / 01:12:38 | Cyrus Younessi | A lot of people have preconceived, I mean people have ideas of where they want to see all the various parameters and it's never going to make everybody happy. And ultimately, it's not like anybody gets to decide. I think pretty much everything is market driven. So we'll see, but it would be great to kind of have some idea or estimates of where we think certain parameters will end up, at least from the governance perspective. |
# / 01:13:10 | Cyrus Younessi | I mean I'm not concerned, I'm just starting to think about how to initially set, how we're going to initially set parameters when MCD launches. |
# / 01:13:21 | Richard Brown | Matteo, do you have access to a mic? You have an interesting perspective on the DSR that I hadn't actually considered before. |
# / 01:13:29 | Matteo Leibowitz | Sure, yeah, can you hear me all right? |
# / 01:13:32 | Richard Brown | Yeah. |
# / 01:13:34 | Matteo Leibowitz | Sure, so my point was just that, in order for the DSR to actually be effective at times where DAI is trading below $1, it should really be above the lending rates offered by these secondary lending platforms. For a DAI holder, DAI holders aren't necessarily concerned as to whether their actions are affecting the Maker system as a whole. They're just looking for the best possible yield. |
# / 01:14:12 | Matteo Leibowitz | So [inaudible 01:14:14] them, whether they're locking up DAI through the DSR or if they're lending out through all kinds of protocols. |
# / 01:14:26 | Cyrus Younessi | At some time though- |
# / 01:14:27 | Richard Brown | It's far more aggressive than I'd ever considered the DSR going. Was that something that's been talked about Cyrus, on our side? |
# / 01:14:34 | Cyrus Younessi | Sorry, let me respond to Matteo's comment. |
# / 01:14:39 | Matteo Leibowitz | One point, arguably you could say that DAI holders should care where they lock it up, because locking up through the DSR is more likely to lead to DAI returning to its peg. But at the same time, if there is a significant enough spread between the DSR and these alternative yields, it shouldn't actually matter too much. |
# / 01:15:04 | Matteo Leibowitz | And it all depends on confidence around Maker's policies, capacity to actually return DAI to peg. |
# / 01:15:14 | Cyrus Younessi | Right, my concern is the difference in counterparty risk for these different services. So locking up DAI in the DSR versus Compound versus Dharma, they all carry different counterparty risks, which may not be immediately evident in the price, in the rates that they... |
# / 01:15:39 | Matteo Leibowitz | That's true, although one would hope of the medium to long term that all these contracts will have undergone significant audits and perhaps even the kind of formal verification that they're- |
# / 01:15:53 | Cyrus Younessi | Not even just technical risk, but just actual counterparty risk. So for example, with Dharma you get collateral back, when you deposit DAI. |
# / 01:16:06 | Matthew Rabinowitz | Its it also appropriate to just compare it to the corollary scenario of a central bank, basically creating value and allocating it to a traditional bank, which then is just recirculating it? In the context here, we're talking about like the DSR is actually removing circulation of those tokens. Dharma is not really removing them, they're just allowing you to earn a yield and charge in other yields as they lend the money back out. |
# / 01:16:35 | Matthew Rabinowitz | One is a direct change in supply, the other one is not. |
# / 01:16:40 | Matteo Leibowitz | So my argument is that for the DAI holder, they don't really care what's happening on the macro side of things. They're just looking for the best yield. |
# / 01:16:50 | Matthew Rabinowitz | Respectfully, I mean I hear you say that, but there's a reason why people buy US Treasury yields and not that of another bank. I mean even if a bank may generate more yield, treasuries and themselves still have command, there's still a demand for them. |
# / 01:17:08 | Matteo Leibowitz | But I don't think we really have the equivalent of like a Federal Reserve in this new paradigm. That's just stressing that Maker is the Federal Reserve. I personally think that it comes with pretty much identical technical risk as these secondary lending platforms. |
# / 01:17:25 | Matteo Leibowitz | Really, what Cyrus was talking about is kind of technical risk, as far as I'm concerned, Cyrus's point, it was no technical risk regarding the liquidation process. So is there a margin of safety in place when you're lending through Compound, that if the borrower's collateral falls below a certain amount, can it be liquidated in time and cover the debt? |
# / 01:17:54 | Matteo Leibowitz | I imagine, again, as these markets mature and become more efficient, we will basically see the same kind of liquidation efficiencies across all these different platforms. |
# / 01:18:07 | Vishesh Choudhry | And I think it's fair to, I think they will seek just whatever's the highest yield. And then the other thing is, on platforms like Compound, yes, there is a small amount of counterparty risk if somebody does not have sufficient collateral due to price changes. But I think the risk of that, I haven't pulled the date on this just yet, but I think the risk of that has been practically relatively low. So I think it will just be about whatever is the highest yield. |
# / 01:18:35 | Vishesh Choudhry | I will say though, that it probably should be kind of adjusted for the ratio of supply volume to borrow volume. Because if I go and offer a loan on Dharma or offer a loan on Compound, it's not necessarily 100% of the amount that I'm offering, is taken up. |
# / 01:18:53 | Vishesh Choudhry | If I know for a fact that if I have 1,000 DAI and I lend it on Compound and maybe 50% of it gets borrowed or something like that, but if I know for a fact that I can get the fixed yield on the DSR, then that may be a reason that I would choose to lock it up in the DSR instead. |
# / 01:19:13 | Vishesh Choudhry | But I do think Matteo's like original point definitely holds true, which is that in order for it to be effective, it probably has to match or roughly be close to the rates that are being offered on some of these other platforms. But you can probably adjust that for the amount of volume that's actually filled, versus just offered. |
# / 01:19:36 | Matthew Rabinowitz | Here's another viewpoint on that same topic. The more that we increase the DSR, whatever we start at, fine, X. Ultimately, we're going to be able to pull the aggregate stability fee down. At the end of the day, we're going to find ourselves, I believe we're going to find, the secondary lending platforms are going to basically be boiled down to whatever stability fee/DSR that Maker ultimately sets, because it will be the trend setter, not really the trend follower. It's just a question of when those happen. |
# / 01:20:08 | Matthew Rabinowitz | Some of these platforms are subsidizing some of those rates today. How long will that last? I don't know the answer to that, but at some point if we set the DSR to X and then have to make it 2X to make it more competitive and the stability fee as an overall factor can drop from 16 or 19, whatever it is, down to say 10 and then down to five and then as we slowly pull the DSR in, the challenges with those secondary markets is that they're going to in effect be squeezed. |
# / 01:20:37 | Matthew Rabinowitz | The question will be, how do they remain competitive? Then, we bring up the point that you're mentioning of, how do we set our rates? We just don't want to get ourselves in a feedback loop. That's I guess my primary point. |
# / 01:20:50 | Vishesh Choudhry | I totally agree with that. I think there's two separate points, one is what happens given a DSR level, compared to external lending rates? And then the other is, how do those external lending rates respond to the DSR and the stability fee? |
# / 01:21:06 | Vishesh Choudhry | And I think it's a very good point, like that is completely a feedback loop. If we start to set the stability fee or the DSR based on those lending rates and those lending rates are not subsidized and are set based on the stability fee. So that's a really good point. |
# / 01:21:23 | Cyrus Younessi | I mean I think we can all agree that the rates should match the risk of the lending. It's very clear that lending DAI on various platforms does not carry the same risk in a multitude of ways. This is also one of the main issues I have surrounding the DIPOR rate and I think they have been part of some conversations about this offhand. |
# / 01:21:59 | Cyrus Younessi | But I think that if you're not taking into consideration the counterparty risk, it's difficult to even pin down with that metric means. If you lend out DAI on Dharma and accept ETH as collateral, versus accepting some other random coin as collateral, there should be different rates offered for the DAI loan. |
# / 01:22:26 | Cyrus Younessi | How to kind of combine these all together and trying to figure out what this inter protocol rate is, is quite confusing to me. |
# / 01:22:36 | Matthew Rabinowitz | I would ask another question for any of the market makers that are on the call. I am not one, but I'd be curious why in many cases, so I've been told, that there's some significant quantity of market-maker DAI that's hanging around. If any of those folks have decided to put those quantities into a secondary market lending structure to make some yield off of it and or their sensitivity/desire to use the DSR to generate an equivalent yield. |
# / 01:23:07 | Matthew Rabinowitz | Because if I were to be a market maker and I were to have I don't know, pick a number, a million DAI and I wanted to make yield on it while I was waiting for the price to be elevated, I would block it away and earn that 1% yield and be able to unblock it with whatever time horizon the contract would allow. That would make me feel more comfortable as a market maker, as opposed to putting it on a secondary platform. |
# / 01:23:30 | Matthew Rabinowitz | Are there any market makers that can comment on it? |
# / 01:23:42 | Matthew Rabinowitz | silence |
# / 01:23:50 | Richard Brown | I'm not seeing any left in the room. I think they dropped and... |
# / 01:23:59 | Richard Brown | Speaking of dropping out, I have a hard stop at 10:30. So everybody can hang out, but I need to stop the recording for fear of losing all of this great content. So I'm going to press stop. |