| # / 00:00 | Richard Brown | I think we have a lot to go over today. We always have a lot to go over, but ... Welcome everybody, to the February 21st edition of the scientific governance and risk meeting, with Steven Becker, Rune Christensen, and an array of industry luminaries. Every meeting, I kick off with the same brief speech, but I'm gonna keep on doing it because it's super important to me. We love debates, we love questions, and we love interacting with people. We have things to learn, and we have things to teach as well, so let's get this circle of sharing going. |
| # / 00:33 | Richard Brown | If anybody has any questions, any comments, any thoughts, please feel free to type those in the chat in the side, if you don't have access to a microphone. If you do have a microphone, please feel free to jump in whenever you need to and ask questions. We're here to talk, and some of the questions we get are amazing. It's a highlight of my calls. I think I'm gonna hand that over to you directly, Steven and let you set the agenda for us. |
| # / 01:02 | Steven Becker | Okay, so, what I'd like to do today, is talk firstly about two things. Number one, the metrics we are looking at right now, that may inform us to consider another change in the Stability Fee. I don't think we've had a whole lot of time to go by, in order to access if the first raise was impactful. But, we are certainly cognizant of it, and we need to be very much aware of where we are right now. |
| # / 01:36 | Steven Becker | Then, the second aspect is, and I was asked this as well, is let's revisit the whole idea of how to be a risk team. We can go back and say, "Well, we don't have the formal tools arranged, but we can at least use this particular space to discuss elements about becoming a risk team, how you contribute, and what you need to basically consider. Thirdly, I would also like to maybe address some of the questions, or apparently many considerations, that have been put towards us. I think Richard's got a few of them that we can actually look at. |
| # / 02:22 | Steven Becker | To jump off on the Stability Fee, right now, the long and short of it is, we didn't see too much of an effect on the inventory levels, but we have seen a kind of a kick up in the ETH price, and we're trying to figure out if there's some sort of determination between that happening, and the fact that we've got a large provision of DAI that hasn't decreased at the moment. Second to that, and this is what I'd like to put to the folks in this call ... Since the Stability Fee raised, do you think there has been an effect on the exchange rates on these sort of collective or aggregate idea that the DAI exchange has gone to, or has settled around one? |
| # / 03:14 | Joe Quintillian | I would say it's a small effect. Sorry, Steven, but I think it wasn't this is Joe Q. I don't think there's been that much of an effect, just a little bit. That's about it. |
| # / 03:26 | Rune Christensen | My impression is, actually things have gotten worse now. We're seeing multiple Reddit posts, with people just complaining and being like, "DAI's so weak," and this kinda stuff. I think we should just try to get a consensus today to continue with the 0.5 percent raises. Just in general, there are some people out there saying, "We need to go to 4 percent," or, "We need to go to 10 percent," and so on. I actually have that ... of course it's really hard to say in advance, but I think there's a good chance that we are quite unbalanced considering that ETH is looking four ways in a lot of stuff. I think we should just get on with it. This is what I was hoping for right, right? Lev sending us some pictures that probably is a bunch of order books. |
| # / 04:25 | Rune/Lev | Yeah, maybe [crosstalk 00:04:26] I can explain it, yeah. |
| # / 04:27 | Lev Livnev | Yeah. |
| - | Male | Hey- |
| # / 04:28 | Joe Quintillian | One question, with regard to the Debt ceiling, how hard-coded is that? Is it possible to reduce that as another alternative? |
| # / 04:37 | Rune Christensen | Oh, that's actually a good other thing to bring up, but I also think we should start this discussion now with a balance. I thought, we should just learn how to use the interest rates, because we need to really be able to do that for the long run stability of the system. |
| # / 04:53 | Joe Quintillian | Okay. |
| # / 04:54 | Rune Christensen | But, bringing up a debt ceiling is a good thing. I think we should talk about that a little bit today as well. |
| # / 04:59 | David Hoffman | If we're talking about the DAI balancing around 1 dollar, as a function of the marketcap of Ether, and how much DAI CDP holders can mint, aren't we also kind of by proxy talking about changing the stability fee every time Ether changes in price to any significant degree? |
| # / 05:18 | Rune Christensen | Yeah, absolutely. This is just how it has always been explained, right? We're always going to have to very, very reactively adjust the stability fees, and later the DAI Savings rate continuously, every single week, and constantly responding to changes in market dynamics if they change. Let's kick it over to Lev, now so we can have a look at some data, the stuff we all come for. |
| # / 05:45 | Lev Livnev | Hi guys, just thought I would share some data, like Rune was saying. I put this link in the chat so you can open it. What you're looking at here, is a plot of some data that I collected. The dots are individual trades. The size of the dot is proportional to the volume of the trade. All these trades were gathered from a couple of Dexes. If you're wondering which ones, it's the Dexes where most DAI volume happens. It's Oasis, Radar Relay, Paradex, DDEX, that sort of thing. |
| # / 06:22 | Lev Livnev | Starting from about 50 days ago, every trade on the ETH to DAI market, on those dexes is listed here. What I did, was in order to get an implied DAI/USD price, I scaled the price of every trade, which is known in ETH/DAI market, with the Coinbase ETH/USD price at that moment, to get an implied valuation of the DAI that was traded in that trade. |
| # / 06:51 | Lev Livnev | That's why this is a little bit more interesting than just looking at the trades on those markets, is because, we've rescaled it by the price to get an idea of what it's saying about the price of DAI. And, is there any pattern here around whether DAI trading has a traded discount or premium? There, you can see pretty clearly why we've had to have all these discussions the past couple weeks, because there's definitely been sustained periods when most trading was clustering somewhere meaningfully below $1. |
| # / 07:29 | Lev Livnev | Obviously, you see a lot of dispersion here. That's a result of the fact that a, this aggregates a bunch of different order books, and trades are happening on different markets, but also, that indeed large bursts of trading activity often have a lot of market impact, which results in these vertical streaks that you see, which is essentially people taking out the order book. |
| # / 07:52 | Lev Livnev | A really extreme example of that, by the way, is earlier today. A lot of people in Maker Chat, were discussing a large CDP that got liquidated. Usually, around those times, you see very heavy DAI trading activity, 'cause the people who are liquidating the CDP are looking to get this DAI from the market. So, this created this short-term burst of DAI demand, which is actually why you see this column on the far right of the graph, which is quite interesting. That was huge volume because this was a huge CDP. |
| # / 08:26 | Lev Livnev | Other than that, I'd just like to draw your attention to the dates at the bottom, and how basically for the last couple weeks, you see a very clear center to this dispersed DAI picture, clustering somewhere around .98. It's pretty clear by that, the vast majority of the trading is happening at implied price of .98, at least if you believe Coinbase ETH/USD to be a good baseline. |
| # / 08:59 | Richard Brown | How far back did you go on these graphs? It would be interesting to see how this shifted the last time we've changed the Stability Fees. Did you see any positive correlations there? |
| # / 09:09 | Lev Livnev | This graph is for 50 days from now. I made the graph two minutes ago, so it's very fresh, but do you mean can you go further than this? |
| # / 09:16 | Richard Brown | Yeah, back to when we raised it, in our last one. |
| # / 09:18 | Lev Livnev | Yeah, I would really like to do that, actually. I would need to download more data. Unfortunately, I can't make one for you right now, because I didn't go back that far, but it's totally possible. Everyone should try to do this. |
| # / 09:31 | Joe Quintillian | Rich, while Madison's on the call right here, we'll have Madison go do that, so- |
| # / 09:37 | Rune Christensen | But, this clearly must cover when we raised it, right? 'cause we only raised it two weeks ago. |
| # / 09:41 | Lev Livnev | I think Rich was asking the previous times that we raised it [crosstalk 00:09:45]. |
| # / 09:48 | Rune Christensen | I just think this shows that a 0.5 raise, really has no effect whatsoever to some extent, right? At least if we only go from this data, and of course, we should ... There's also a bunch of other things, like inventory levels, and there's also other ways to at least get a sense of the direction, and which direction and balance pass, right? But- |
| # / 10:15 | Richard Brown | But, we can talk about a few questions though. Is the reason we haven't seen it move the needle because the change was too incremental to actually change people's behavior, or was it not large enough psychologically to make people think twice about opening a coin? |
| # / 10:33 | Rune Christensen | There's been multiple people on Reddit, for instance, making the argument that ultimately, market dynamics themselves drive changes at such a stronger degree than something like a tiny Stability Fee change. Because, a lot of people, they own CDPs for fun, right? They don't even look at the Stability Fee, or whatever. They just go do it all. That 0.5 percent adjustment currently is kinda meaningless, so- |
| # / 11:05 | Lev Livnev | If I can throw in a couple more thoughts that are useful when looking at this graph? One thing, is as Rune was pointing out, market sentiment and credit demand is very sensitive to what's going on in the market. Possibly much more so than something small like a Stability Fee move. Indeed, what we've seen in the past month, is possibly increased foolishness around ET/ USD, and not that many downward moves. I think today was pretty much an exception, but we saw remarkably few liquidations over this time, which also means fewer people who were forced to cover their positions preemptively. That's usually a source of DAI demand. Once in a while, someone decides that they got two others, but it's unlikely that that would have happened in the past month or so because of the relative scarcity of those downward moves. |
| # / 11:58 | Lev Livnev | Another thing that I wanna point out, and this has never been brought up in these calls as far as I remember, but what's going on in the rest of DeFi, is I think pretty relevant. Right now, and I'm not an expert on how stuff like Compound works, but if you go on the Compound homepage, you'll see that they're offering 2.25 interest for lending your DAI. This rate, which I to be honest, don't even know where it comes from, where it's set, or who's paying it, clearly must have some effect on the rates that you can get ... sorry, on the demand for CDPs. One obvious reason, is that right now, it's a viable strategy to open a CDP and borrow DAI from it at one percent, and then lend it into Compound for 2.25 percent. |
| # / 12:42 | Lev Livnev | I'm not actually commenting on whether that's a good strategy, but presumably someone could be doing this, and if you imagine that Compound went and doubled their rate, then this strategy becomes twice as appealing. So, clearly more people are gonna do it, if provided someone was gonna do it in the first place. Then, likewise, if it goes down, so those rates actually have an effect on DAI supply and demand as well. There's these very clear kind of exogenous things going on, but they're still in DeFi, that are worth paying attention to. In fact, I think the DAI rate on Compound was much higher. Earlier, I remember seeing it at something like 6 or 8 percent just a couple weeks ago. I think that's just- |
| # / 13:22 | Richard Brown | The part about the conversation that's becoming more popular in the DeFi space, the interesting shift from using ETH as sort of a staking or collateralization method in order to acquire DAI, but there's this aftermarket of lending organizations that want a stable coin, in order to provide collateral as well. So, we're actually feeding into Compound. We're feeding into Nuo. We're feeding into Uniswap as well, so one of the major drivers for large amounts of DAI inventory, seems to be people lending from other platforms. |
| # / 14:00 | Rune Christensen | I saw that [Vishes 00:14:01] also posted some cool stuff. Maybe if you want to jump on and explain it, that would be really cool. |
| # / 14:12 | Vishes Choudhry | Sure, yeah. I just pulled from some of the Dexes, Oasis, V2 of 0x, and a couple others like ETH2DAI one, and ETH2DAI two, what the ETH/DAI trades were, and pulled the ETH/USD price pairs from Coinbase at a five minute resolution, and then just used that altogether ... a sort of DAI implied [Vwap 00:14:44] over time. So, in the Google doc I shared, the first two graphs are kind of cut up by each of the exchange sources. Then, the second two are sort of combined altogether, volume weighted average price. |
| # / 15:03 | Vishes Choudhry | Then after that, there's three that follow, which is essentially, what's the DAI price for DAI/USDC? Then, direct from Coinbase, what the implied DAI/USD price would be going through USDC. |
| # / 15:23 | Rune Christensen | That's looking at the USDC market price. Shouldn't you just take it one-to-one, considering you can just convert it- |
| - | Male | Yeah, well that's- |
| # / 15:30 | Rune Christensen | ... directly in Coinbase? |
| - | Male | ... what I'm thinking. |
| # / 15:31 | Richard Brown | That's the traded premium though, right? So, how sometimes on Coinbase, USDC's trading above a dollar now. We can agree that maybe it shouldn't, right? But, it sometimes does. |
| # / 15:43 | Rune Christensen | Isn't there a page where you just click a button, and you convert the USDC to USD one-to-one always? |
| # / 15:48 | Lev Livnev | There is that page. My hypothesis would be, that it appears to be the premium for discount, just because of the thin liquidity on the USDC pairs, but I don't think anyone's actually intentionally pricing USDC distinctly- |
| # / 16:02 | Lev Livnev | Right. |
| # / 16:02 | Lev Livnev | ... from [inaudible 00:16:03]. |
| # / 16:03 | Alex Evans | It doesn't really matter, I think. I was just off the call, so I missed what you said, Vishesh, but even adjusting for that illiquidity, where sometimes USDC trades above a dollar, that is still trading below the dollar peg. Now, that is a very short ... I think the data set starts when Coinbase Pro lists DAI on December 19th, but there is practically no period since December 19th where it's traded above. |
| # / 16:30 | Rune Christensen | And Vishesh, what's the last ... the ETX one? |
| # / 16:32 | Vishes Choudhry | That was- |
| # / 16:32 | Alex Evans | ETX is TradeBlocks, isn't it? |
| - | Male | Yeah. |
| # / 16:41 | Richard Brown | Yeah, ETX is TradeBlocks, which has, essentially, an index for ETH price as it's trading volume weighted across a Kraken and Bitstamp, Coinbase, a couple other exchanges, whether it's a little bit more, certainly around whether that volume's legitimate or not. |
| # / 16:57 | Vishes Choudhry | Yeah, it's the same thing, just a slightly different ETH price source. |
| # / 17:01 | Rune Christensen | I feel like this one paints a little bit more stable picture than the other ones. The other ones clearly has that, it's going down all the time. Whereas, this one seems like it's fluctuating, but it's mostly fluctuating right at $1. |
| # / 17:14 | Alex Evans | Yeah, that would be something to double click on, in terms of what estimate of ETH price do you take to be the correct one. I would probably argue that since this is all Coinbase, we should just take the Coinbase rate, but I think that's debatable. There's also some loss in resolution, in that those are reported at daily frequency. So, really I think the right way to probably do this, would be to recompute the TradeBlocks index. I think I could only get it for days, but do it, let's say hourly or within a minute, where the variance in the price is relatively low, and then adjust it based on that. |
| # / 17:52 | Alex Evans | Maybe that would be the best one, or maybe doing something equivalent for Coinbase. I think it would be worthwhile to do, so I might spend a little bit more time next week doing that. But yeah, you're right. I think when you use that price, you get a little bit more stability around the dollar peg. I would have to double click and see why that's happening. |
| # / 18:13 | Rune Christensen | Especially looking at this stuff, I think this really speaks to the fact that a 0.5 percent Stability Fee increase has ... It doesn't even work apparently. At the same time, we're actually trending down, so things are getting worse and worse over time. I think the right thing to do now, would be ... so, right now, there's steps we take ... |
| # / 18:42 | Rune Christensen | I don't think we could go out today, and be like, "Hey, let's increase the Stability Fee with some higher amount," or something. That would be too sudden, but I think what we could do, is we could get consensus today of doing a 0.5 Stability Fee increase. Then, set the expectation that if we wont again, see absolutely zero effect of this. In fact, if we see a continuous negative trend, then we're gonna start with ... or maybe we should go back to 2 percent adjustment or something. We can have that discussion about that maybe on Reddit or something? |
| # / 19:17 | Rune Christensen | But, this is kinda looking bad, right? The thing is also, a sort of death spiral starts happening at a certain point, where you get less and less DAI demand as you can see here, because people start thinking it actually is a terrible stable point. So, it almost becomes a self-fulfilling prophesy, right? I think now we have to say, "It's time for us to more drastic the action." Then, for now, let's just see right now if there's anyone who disagrees, or has data against a 0.5 percent Stability Fee increase. If not, then we can have the ... Ultimately, I would say it's up to Steven, but we can see if based on our consensus today, see if we can then have the Foundations Risk Team propose and endorse a vote for a 0.5 percent Stability Fee increase. |
| # / 20:20 | Richard Brown | That's an interesting question. Steven, can we push that over to you? Do you want to comment on that? Is it important that we maintain a certain cadence? We can't eliminate the variables in the system, obviously. There's market forces we don't quite understand. There's crypto courses that nobody understands, but should we figure out every two weeks, we're going to make a change until we see something that's a satisfactory result? Maybe another 1.5, and if that doesn't change, move the needle, then we would go to one. Is there a standard for this stuff? Is exponential increase over steps, or do we do linear? Are there methods for this kind of thing? |
| # / 20:55 | Steven Becker | Yeah, there are methods for this kind of thing. It's generally linear, because you've gotta think about the fact that if you started at 0.5, and you increase by 0.5, it's a 100 percent increase. Then, if you increased by 0.5 again, it's a 33 percent increase. If you're talking about an impactful increase, you gotta keep a consistent percentage, so we might have to consider increasing by more than 50 basis points. That's one thing we need to put on the table for next time, not now, because we're going against what we initially agreed upon. But, outside one thing that Lev brought up, and I completely agree with this, outside of the demand function, which is something that we have to constantly be working on, is the consideration of the DeFi space, and actually use the DeFi space to our advantage in order to manage the extreme rate of corporate leak. |
| # / 21:53 | Steven Becker | That needs to be considered, and to the extent that we need to consider it, I think it's open to everyone to suggest. But, further to that, I think we need to agree right now that if we don't see an impact from the increase in Stability Fee that we moved two weeks ago, we need to firmly agree what could be other considerations. For instance, the increase in ETH, could have just promoted people to draw out more DAI. That is a factor that is confounding to the Stability Fee. Considering we're sort of bootstrapping this, we then need to be very pragmatic in terms of how these variables ... very much like the DeFi space that we're in, will contribute to the impact in the DAI exchange rate. That is the pragmatic discussion we need to have now. |
| # / 22:54 | Steven Becker | Now, I think we need to use the Stability Fee, and increase it until we find a point where it does have an impact, but it does so exclusively and not muddled up with any other confounding variables, so- |
| # / 23:07 | Richard Brown | Are you saying that we ... Sorry. Are you saying that we need to get a clearer picture of what people do with DAI, like we need to figure out where this DAI is landing, what's- |
| # / 23:17 | Steven Becker | No, what I'm saying, is outside of the demand function, 'cause if you ... The one critical thing, the one foundational piece that you have to have, is trying to understand what the demand function is and improving it. So, what are the used cases? Where can you apply it appropriately? How do you spare that? and so on and so forth. Very much like the core community development that's focused on DAI demand development. We've got all these things in place, but we've gotta give it enough time to get the momentum back. That's a given, and I think everyone can agree that that is a given. |
| # / 23:49 | Steven Becker | Over and above that, we need to get extraordinarily pragmatic and say, "Well, what are we dealing with outside of that?" What we're dealing with, is the Stability Fee that tries to disincentivize people through the idea of costs. If we see that that's not having an impact, we need to say, "Okay, well, there's a possibility we need to increase that more on a percentage basis than on just a normal basis." |
| # / 24:14 | Steven Becker | Outside of that, so this is to Lev's point, let's look at the DeFi space. Let's look at what competes. Let's see what's driving the sort of supply and demand that we need to have out there. What sort of pragmatic choices do we need to have against that? This is, again, outside of the demand function, because if the demand function ... not if, when the demand function kicks in, then that is gonna become your pragmatic point of reference. That has to be then, really the full extent of the economy that we need to deal with, 'cause right now, with this bootstrapping, we have to be very realistic with what we can do in the space that we have. So yes, I get the Stability Fee's not working. There's two weeks worth of data, so it's statistically speaking, not really impactful, not really telling. At the same time, we would have expected some kind of response, and we're not getting it, so I think we need to keep on increasing this in a linear fashion, and as some sort of exponential fashion. |
| - | Male | Yeah, I think what we should- |
| # / 25:28 | Richard Brown | Should we [crosstalk 00:25:28] at this point? |
| # / 25:30 | Rune Christensen | Yeah. I think the most important thing that we should get out of this meeting today, is that we should ... I think the expectation has been set enough that we can definitely sign off on a 0.5 increase today. So, then according to plan, I'm like, how the scientific governance framework is ... how this is supposed to work. In the long run, that means we have a scientific debate. We've seen a bunch of evidence that says, let's increase the Stability Fee. Let's see if there's any evidence that says, let's not increase the Stability Fee. |
| # / 26:02 | Rune Christensen | Then, ultimately see if we can come to a scientific consensus. Then, based on how ... and you were just talking about this, Steven, right? We should discuss this later, based on how the risk teams are supposed to function in the long run. Then, what needs to happen, is a risk team, which right now is the Internal Foundation Risk Team, which is represented by Steven here ... signs off on this saying there's a legit scientific debate that has happened here, and there's actually a legit scientific consensus. |
| # / 26:38 | Rune Christensen | Based on that, we believe it's possible that we can actually get this. We can create a vote for this, we can recommend people vote for it, and ultimately, it's likely to pass. I think now, kind of like next step, is we see if there's any arguments against a 0.5 percent Stability Fee increase. |
| # / 27:00 | Steven Becker | The only thing I would like to pose as a question, is we generally use governance polling as the place where we get the rough consensus, as it were. I just wanna be clear so that everyone understands that what we're doing now substitutes that, and- |
| # / 27:19 | Richard Brown | Yeah. |
| # / 27:20 | Steven Becker | ... if everyone understands that, then it's important 'cause we've laid down this premise of, we're gonna do governance polling. Then, we'll have an executive vote. What we're saying right now, is that this governance polling is not going to be ... so we're gonna substitute this for the governance polling. We need to be very clear on this, because it means we are to a certain extent, going against what we've laid down in general. Would it be wise to have a governance polling that's maybe open for a couple days? |
| # / 27:48 | Rune Christensen | This is actually super important, but the fact is that right now, we shouldn't really ... The thing is, what we're doing right now is emulating the DAI Savings rate adjustment process. We're not following the long-run risk governance. This is actually another thing I wanna talk about, is I think it's time to split up short-term DAI stabilization monetary policy, decision making, then long-term risk governance, and collateral onboarding decision making, and so on, but, ... and let's get to that after, and also discuss how the risk teams ... their involvement in the DAI Savings rate versus like risk parameters, and so on. |
| # / 28:31 | Rune Christensen | Yeah, I think regardless [inaudible 00:28:32] ... Let's make sure this is not taking as a precedent for how things should be done in the long-run, 'cause really this is kind of like a mix of different things altogether. So, like Steven is saying, what we're doing now, is actually an executive vote in the system, which is not something we'd really wanna do in this fashion in the long run. Doing a governance poll, is also not appropriate because this is not a decision that's meant to be a big large community thing ... like really bulky, unwieldy thing. |
| # / 29:00 | Rune Christensen | Stability Fee changes will have to happen, or rather, Savings rate changes will have to happen very often, basically all the time in response to market dynamics. But, one thing I think is necessary no matter what, is that we at least do a ceremonial little pause to see if there's anyone who wants to basically argue for the opposing viewpoints. Then, we have a risk team. Steven will sign off on the fact that we have followed the process. Then, we can go out and do this, and present this to the whole voter community, and get them to see if they will support it. |
| # / 29:45 | Steven Becker | Well this bit, does anyone disagree with that construct? I think we've got 33 people online. If we've got sufficient people here that disagree with this current construct, in terms of adjusting the Stability Fee and then the DAI Savings rate going forward, I think it would be a good sample in understanding what a broader view would be like. If there's no one here that disagrees with this construct, then I reckon we have a decent enough sample to justify the fact that we can use this process for these kind of adjustments. I'm gonna give it 10 seconds for any disagreements, and then if there isn't, I- |
| # / 30:25 | Rune Christensen | Yeah- |
| # / 30:27 | Richard Brown | I might wanna clarify that a bit. I'm not sure I disagree, but I think establishing a signal or rough consensus ... I have a lot of strong opinions about the term "rough consensus," so we need to be careful about what that means, and when it's been established. But, I think there's a good middle ground here. |
| # / 30:44 | Richard Brown | This is one of the primary venues for people to be heard, present data, or we can come up with a plan, but I wanna make sure we don't use these meetings to pose a question and then come up with an answer, and that at the end of the meeting, the ship has sailed. So, there should be an opportunity for people who still contribute to the conversation if they can't make these calls. |
| # / 31:07 | Rune Christensen | I think though, that we are following a process that has already worked, right? It already went through with the community before ... That is last time. For a while, we presented evidence and so on, and then we did a 0.5 percent raise. Once again, we have no data against the 0.5 percent raise. I think we can do it. Now, let's have a little pause for 10 seconds and see if anyone wants to present data or arguments against the Stability Fee raise. |
| # / 31:36 | Chris Burniske | I don't have anything against the raise. I think I'd like to discuss whether it should be 50 basis points or a 100 basis points. David [Perry 00:31:46] just posted something to this end as well, and I don't have the expertise to say whether it should be 50 or a 100. I think I'm more concerned with making sure we appear rational, and not erratic to the community with how we decide between the 50 basis points and the 100 basis points, because Steven just gave the example of, if we raise 50 basis points right now from where we are, it's a 50 percent increase. So, that is less of an increase than the last one we went through, which was basically doubling. So, when I think about the current market dynamics, we had Ether Rally, so that makes it easier to create more DAI. And, we have a DeFi space that's showing some strong traction, which is also creating more demand for DAI. I think we have these two drivers for a DAI supply glut. |
| # / 32:49 | Chris Burniske | Personally, I think once Ethereum forks ... I personally don't see ETH continuing to rally super strongly here, and so I don't think we will face that "supply glut" headwind in a really strong way in the coming few weeks. That would for me, argue for a 50 basis point increase, 'cause then it's really the DeFi demand and supply glut that we have to take into consideration. |
| # / 33:19 | Richard Brown | Chris, that's a great point. So, I wanna make sure we're capturing the variables that we need to address here, so whether it's 50 or a 100, is something we need to figure out. The other thing I'd like to make sure we keep cognizant of, what's the cadence of the changes that we're prepared to make until we see a positive solution here? Does that mean that every two weeks, this sweet spot can be stretched out to four weeks to see whether things work or not? I think that cadence is gonna be determined by how quickly ... and correct me if I'm wrong here, everyone, how quickly we approach the danger zones, the debt ceiling. Is that the cap we're looking at here? |
| # / 33:59 | Rune Christensen | Let's discuss the cadence, but I think the main goal, is still ... I don't think we can possibly do a 1 percent increase today for instance. If we did that, there would be no kind of warning on that, and something on that, definitely would have to. There's no expectation on the cadence, so theoretically, we could do a 0.5 increase weekly, currently. And, we haven't really set up an expectation that that's not going to happen, but we have set expectation that changes will be 0.5 percent. |
| # / 34:33 | Rune Christensen | Today, we have to do what we can. Basically, the data's looking pretty critical. We kind of feel like we have to do something to deal with it, but I think today, the most we can do, is a 0.5 percent increase. Then, let's discuss how we can then set ourselves up to be more effective at the next couple meetings. |
| # / 34:54 | Rune Christensen | I would just like to go back to, basically, there's no data against it. With that, Steven, I think I'd like you to sign off and say, "Here, we have a consensus, and as a result, the foundation risk team is gonna create a proposal for 0.5 percent Stability Fee increase. |
| # / 35:17 | Steven Becker | Yeah, that's ... I agree with. We will go ahead with the proposal, but to David, Terry, and Chris's point of view, you guys are correct. To Rune's point, this is more about signaling and sticking to what you said. So, we gave ourselves the ambit of 0.5 percent increases. That's available to us today, but we can lay the foundation for something to adjust going forward, and obviously we can use that foundation next week to maybe increase it more so. |
| # / 35:52 | Steven Becker | Given the fact that this is recorded, and given the fact that most folks out there that are not online right now, would wish to see this, it would be great for them to be prepared for the fact that we've created a new foundation from which we can actually upgrade. It's just more signaling thing. I do understand, at the same time. Considering that it's really a weekly basis, and we don't really have a statistical telling of the impact yet, I think we can safely say that the foundation can be built now for a further and larger increase next week. To that point, I will agree, and I'm going to put forward a proposal to increase the Stability Fee by 0.5 percent, that being 50 basis points. |
| # / 36:41 | Rune Christensen | Okay, time to- |
| # / 36:42 | Richard Brown | # / [crosstalk 00:36:42]. |
| # / 36:43 | Lev Livnev | Can I add a technical point, since we were talking about magnitudes of changes and stuff like that? I would advise against thinking about Stability Fee, or interest rate. In general, changes in terms of doubling it, so going from 0.5 to one being doubling, increasing it by 30 percent, halving it, or anything like that. It's actually kind of meaningless to use those relative change terms. It's kind of like saying double the temperature. Increase the temperature by 30 percent. It's a quantity that doesn't actually have an absolute zero, so it doesn't make sense to even measure it like that. Just to steer this quantitative part- |
| # / 37:25 | Rune Christensen | Yeah. |
| # / 37:25 | Lev Livnev | ... of the discussion. |
| # / 37:25 | Rune Christensen | No, that- |
| # / 37:26 | Steven Becker | Yeah, that's agreed, but at the same time, we don't know the effect of the Stability Fee. So, getting that it's sort of exponential effect working for us now, might get us to a point quicker in understanding that the Stability Fee does have an impact. I wouldn't say it's a principle that you would apply going forward, but it's almost like a principle you would apply getting to a point where you find you have an effect- |
| - | Male | Yeah, I mean I- |
| # / 37:54 | Steven Becker | ... and really neutralizing it from that point of view. |
| # / 37:57 | Rune Christensen | I agree with the notion that doubling interest rates doesn't make too much sense in that regard, but interest rates are super arbitrary. And, like Steven was saying, the thing is, that our arbitrary change level of 0.5 percent, considering the current conditions, is probably set too low. We should make it bigger. I just think a funny ... well fun, or whatever, data point on this whole approach, is kinda like having an arbitrary number and then making changes to that. That's how central banks do it. They always change with 0.25 percent. The thing that then determines how much of a change they do over time, is how often they do these changes. They always do them ... most of the time with 25 basis points. |
| # / 37:51 | Richard Brown | Would it help us with messaging, and potentially the psychological impact of these changes, to have a message about the cadence that we're looking at? So, we say we're gonna do another 50 basis points this week, and if that doesn't work, we're gonna continue doing it until it does work. Would that help, maybe move the market in the way we're hoping? |
| # / 38:10 | Rune Christensen | I think we should, I think- |
| # / 39:14 | Steven Becker | With the stocks and point, I agree. But, at the end of the day, it might not actually be that informative, because you might find that the confounding variables while we're bootstrapping the system, might actually play a greater part. So, we could be constantly increasing the Stability Fee until it has an effect, but maybe because the confounding variables have stopped having an effect. It really is, considering what we've discussed today and the input that we've got today, is really ... I must say, it's brilliant. This is the kind of input we need on a constant basis, until such time, as we framed all of the issues we have, not only in bootstrapping to critical mass and having a robust system, but also when the system does become robust. You're gonna have a very well-defined framework as well. |
| # / 40:03 | Richard Brown | That sounds great. I think we've reached ... I'm gonna say it, rough consensus. So, we have a way to move forward, but I wanna throw a wrench in the works and ask a kind of tough question here, because it feels like we are operating with one of our hands tied behind our back until we get our DSR-type mechanism into the framework. Is there a possibility that there isn't enough room on the Stability Fee to see the kinda changes we want in inventory? |
| # / 40:30 | Steven Becker | Yes, is the short answer. |
| # / 40:32 | Rune Christensen | But, that's only possible the other way though. Right now, we're in a happy position where we can just jack up the Stability Fee however much we need. It only becomes a problem if we reach zero and we want to go further down, but there's nothing below zero, so- |
| # / 40:46 | Richard Brown | I guess the risk I'm thinking about, is in order for us to balance inventory, we need to raise the Stability Fee to the point where we've absolutely priced ourselves out of the ecosystem. We're no longer competitive, and- |
| # / 40:56 | Rune Christensen | But, that's exactly what we're trying to achieve, which is kind of funny, right? But, literally, what we're trying to achieve, is we- |
| # / 41:00 | Steven Becker | It will be- |
| # / 41:00 | Rune Christensen | ... want to scare away the customers in a sense. |
| # / 41:04 | Steven Becker | ... in a very, very controlled fashion, where it is seen as momentum behind something that is getting critical mass, as opposed to just turning off the lights. |
| # / 41:17 | Matthew Rabinowitz | Now, to give you the question I've got then, and this is kind of related to the debt ceiling component also. During the course of the last year, the price of Ether has continually gone down, so the issuance of debt has been hitting a headwind of both participants owning Ether, needing to both put collateral in and continue to borrow DAI against it. |
| # / 41:37 | Matthew Rabinowitz | When we turn that ship at some point, maybe it's in a month, two weeks, a year, whenever, and the price starts to accelerate going up, the creation of DAI can also then exponentially increase. I don't know the technical components about how it's put in for multi collaterals ... forgive that, but if there's not a throttling method that allows that to be slowed down on the issuance path, some portion of that may turn around to be sold off into the fiat world, which makes us look and feel like it's worse, but it's just the market improving. |
| # / 42:10 | Matthew Rabinowitz | A question I have is, is there a throttling method and/or kind of a ... What's the word? If you were to have the debt ceiling be right at the current debt amount, the DAI issuance amount, where no more could be created synthetically, you're stopping the creation of new debt, but you're not inhibiting people from repaying it. Your- |
| # / 42:33 | Lev Livnev | The Stability Fee is that throttle that you're talking about. So, you're describing a scenario where there's increased foolishness in the markets, causing an increase in credit demand. The Stability Fee should then come up in order to balance that credit demand against the DAI demand. |
| # / 42:47 | Matthew Rabinowitz | No dispute. That's in a completely efficient market. I 100% agree, but, we've gone through from nothing, to whatever it's at ... 80 million now. While the price is declining, that interest rate, at that point, ends up being so variable. |
| # / 43:03 | Matthew Rabinowitz | My question is, is it possible just to have a throttle on the DAI issuance on an ... I don't care, whatever it is ... a hundred K a day. It doesn't matter what collateral it's based on, to specifically cause the supply to not be ramped up, 'cause if the price of Ether tomorrow, spiked to a thousand, there will be people who will borrow against their Ethereum holdings to go pay off their house. That house is not inside the blockchain. It will make it look like the price will go craziness on the peg, 'cause I don't necessarily think people will be dumping money into DAI at that point. They will probably be taking the other direction. |
| # / 43:38 | Lev Livnev | Don't you think this throttle mechanism is like a sharp cliff response of the Stability Fee? Essentially, we're saying for the first hundred K of issuance today, we're gonna have the normal Stability Fee. Then, after a hundred k, our stability fee is gonna be infinite? Then, obviously that kills credit demand completely. Then, we wait until the end of the day, and then we add another hundred to the ceiling, or whatever, like you're suggesting? It's basically just creating a really- |
| - | Male | Go ahead. |
| # / 44:02 | Lev Livnev | ... it's creating this nonlinearity [crosstalk 00:44:03] kind of response, but can you explain why you think that nonlinearity is justified, as opposed to some more gradual responses to stability? I think we can all agree that no matter how bullish the market is, or how much credit demand there is, we can probably pick a Stability Fee that will be high enough to balance that against demand, like 5 percent, 10 percent, 20 percent, whatever. At some point, it becomes- |
| # / 44:29 | Matthew Rabinowitz | But, when you're that high, to inhibit the demand, those types of interest rates ... you've priced yourself completely out of the market. At that point, it would be easier just to not create it. |
| # / 44:40 | Rune Christensen | Here's the thing, there's actually a mechanism that's quite like what you're talking about, that's a part of the framework here. That's what's called "The defense of debt ceiling." So, this idea that we never wanna leave a really big empty space of tons of unused debt ceiling. Rather, we wanna make sure that the debt ceiling, like the actual parameters that's in the live system on the blockchain ... make sure that if suddenly there's someone who wants to generate like a billion DAI, he can't actually just do that in one go, but there is actually, like you reach it or get close to it. |
| # / 45:16 | Rune Christensen | Then, there is sort of an off chain mechanism. So, basically a risk team that then is like, "Okay, it's safe to keep going, and keep going with that," until we hit the point that's actually set by governance, which is called the "practical debt ceiling." So, there is actually a mechanism to some extent that's like this, but I think it's very important that that mechanism is only seen as to deal with absolute kind of disaster situations, like very weird edge case situations, and not really something that should ever kick in under normal circumstances. So, normally people should run into the issue that, "Oh, whoops, you can't do your whatever, like draw for some regular amount of money." |
| # / 45:56 | Rune Christensen | I also think there's a lot of hypothetical scenarios that can be thought of, of weird behavior and weird situations. I think in general, we should kind of be empirical about this stuff, in the sense that we ... In the early phases of DAI, we can't actually afford that the system looks a little bit wonky, because there's a ton of like a search and demand. That kind of throws it off balance a bit, and then we figure out how to deal with that. Hopefully, we'll always end up dealing with DAI Savings Rate, and be flexible enough with using it. If not, then we would be something like, "Let's be more aggressive with how we use the Defensive Debt Ceiling," for instance. |
| # / 46:44 | Rune Christensen | The main thing I wanna talk about, and I also want Steven to make his point about risk team, so I'll be very quick. I think it's time that we really split up the discussion about something like the debt ceiling, and long term theoretical considerations about how we run the system and is basically looking. So, the collateral side, the supply side of things, with the more short term DAI Savings rate focused ... short term monetary policy discussions that are very actionable. Especially currently, where they're really the only ones that right now, are based around actually doing real stuff with a real system. I think basically, the easiest way to do this, would be to have a Wednesday meeting at the same time. That's the monetary policy meeting, and then the Thursday meeting like now, is then the long-term risk governance meeting. |
| # / 47:41 | Rune Christensen | Basically next week, we could then have the Wednesday meeting be the one where we discuss ... Next Wednesday, we decide what are the new changes in Stability Fee going to be like for the short run to deal with this problem of, basically, this five doing 50 basis points changes and nothing happens. We'll talk about that on Wednesday. |
| # / 48:02 | Rune Christensen | Then on Thursday, we will be talking about the debt ceiling, both from a theoretical perspective for the long run on the system, but also talking about the fact that the current debt ceiling's about to get hit in Single Collateral DAI. A lot of the theory around this ... 'cause one of the interesting things, is that people typically think the reason why you have to increase the debt ceiling, is because supply reaches a certain level, and CDP demand reaches a certain level. What's funny is, it's actually a demand that sort of drives when you must increase the debt ceiling, and it's actually a question of when DAI demand explodes that really creates that sort of pressure to eventually ... So basically, face the fact that eventually we either have to increase the debt ceiling, or we have to shut down the system. |
| # / 48:55 | Rune Christensen | There's no option to keep the debt ceiling at a hundred million forever, so I think this way, we will be able to focus a lot more in these discussions. Then, especially make sure that on Wednesdays, it's really about just focusing on being actionable on protecting the DAI peg, for the short run. We already discussed this with you a little bit, Steven, but what do you think, and what do other people think? |
| # / 49:25 | Steven Becker | May I make a suggestion? I like the idea. Can we take this particular meeting, and first portion it into two parts to deal with those two points? The reason being, is to try and find another date in everyone's calendar, and get the same response that we've gotten today ... might be a little bit of an ask, but if we can develop it in this meeting, and it takes a life of it's own, I think we can then manage to transport it across to a new date. It's just trying to get everyone to fit in a new calendar item, might be a little bit of a push right now. |
| # / 50:04 | Rune Christensen | I just think, to be honest, the actionable ... that's the short term stabilization ... or eventually, DAI Savings rate adjustments, they are a little bit more about the people actually in the market, who actually are trading, and so on. And, who surely will fit in the time to fix the broken peg, will be able to make it. whereas, the more ... where everyone from the more broader and more theoretical, and so on, is more ... like better fit for the [inaudible 00:50:43]. I'm totally open to either way. I just think- |
| # / 50:48 | Richard Brown | We have an opportunity to find out the answer right now, so maybe if people wanna just signal in the chat whether they feel an additional meeting, or augmenting the agenda's we have here would be useful. We can do that in the next five minutes. |
| # / 51:02 | Rune Christensen | Yeah,- |
| # / 51:05 | Chris Burniske | I think it's- |
| # / 51:05 | Rune Christensen | Sorry, Chris. |
| # / 51:05 | Chris Burniske | I think it's easier as the hybrid step, to augment the schedule. Maybe, you start the first 30 minutes with the short-term rates, so then at whatever time you start ... and I guess 12:30 p.m. Eastern, you know it's intent to switch to the more theoretical longer-term. Constraining time tends to lead to more efficient conversation. |
| # / 51:23 | Rune Christensen | That is a valid point, but I still think ... Today we spent 40 minutes on some of the DAI Savings rate stuff, for instance. The thing is, right now, we haven't even gotten started with the actual point of these discussions, risk governance framework, the collateral parameters, and all that stuff. |
| # / 51:51 | Steven Becker | The exciting thing about this, is that you have started the catalyst behind this thing. We're going to have meetings where we're gonna have a lot more efficient metrics being shown. You're gonna have meetings where everyone knows what we need to discuss, and the discussion's gonna be more poignant. Right now, it really is gathering our thoughts about, what are the items we need to consider? I reckon as we go along, we will become a little bit more efficient in terms of what we are discussing, because I certainly don't think there's an infinite number of variables out there that's going to constantly confound our analysis. So, this is a great starting point. |
| # / 52:36 | Steven Becker | The fact that we've used one hour of the risk governance to discuss this, is great because I think it works part in possible with what the risk teams look like, how do we consider collateral types, because this gives an ability for us to inform how we need to consider the rest of the governance issue, when it does come back as the full point discussion- |
| # / 52:59 | Rune Christensen | Yeah, that's a great point, but I just wanna say that once we do roll out the risk parameters, we're gonna need an hour for that only. If we only have half an hour ... 'cause that's when the real engagement really starts, so maybe we can say, let's have a hybrid for now, but then keep in mind that once we roll out the risk parameters, we open them up for debate. Then, we'll move that half hour to the Wednesday. |
| # / 52:22 | Steven Becker | Yeah, I believe that we'll probably have to transition into something else, but as Chris suggested, I think if we do it more naturally and organically, it will bring everyone into the next iteration. |
| # / 53:37 | Richard Brown | Alright, that sounds like a good place to stop, because we've made some decisions. We've talked about the next proposal. Steven, you're gonna be handling the communications and release of that. People can expect to see another 50 basis points proposal coming tomorrow? What do you think the rough timeline would be, Steven? |
| # / 53:55 | Steven Becker | That's a very good question. I don't wanna commit to anything right now, but I would like to try and get something out. I'm in Omaha, and I still need to go to a few things, so I'm gonna put Tuesday as my first target. It's a soft target, but hopefully by then ... not on Tuesday, but by then, we should have something out. |
| # / 54:22 | Richard Brown | Okay, so we'll look for that next Tuesday. We had an interesting conclusion about talking about a deeper monetary policy. We'll shift the agenda around and make sure people know to expect that for the next call, and see how that evolves. Thank you everyone for the excellent questions. This was a really good meeting. |
| # / 54:45 | Chris Padovano | One question,- |
| # / 54:47 | Richard Brown | Okay. |
| # / 54:47 | Chris Padovano | ... would it be helpful if we had someone who took notes during these meetings? Not like rough ... like formal notes, but just someone from the risk team goes through, takes notes, and then circulates it with the meeting. Is that something that you guys can [crosstalk 00:55:06]. |
| - | Male | That would be perfect. |
| - | Male | Sorry? |
| - | Male | Yeah, I can ... |
| # / 55:10 | Richard Brown | Sorry, I'm getting some crazy feedback from you. |
| # / 55:11 | Chris Padovano | Sorry. |
| # / 55:12 | Richard Brown | There you go ... Yeah, that's a great idea. So, Scott here, and one of our community members was doing that for the longest time, but it's an enormous time sink. But, happily, we have a community development department, of which I am the head, so if somebody would like to step up and offer to take minutes on these meetings, I will be happy to discuss a reasonable compensation plan. We can figure out how to do that, I think. |
| # / 55:37 | Rune Christensen | Yeah, but I mean that's the [crosstalk 00:55:38] |
| # / 55:39 | Richard Brown | ... we talked about. |
| # / 55:40 | Rune Christensen | I totally agree. From now on, let's ... Okay, the meeting's over, but I just wanna say that I think in general, the risk teams are the ones that are responsible for facilitating meetings like this. So, it makes sense that we basically say something on that, that's a standard responsibility of the risk team, which then, since the risk team is now essentially the foundation, then we have risks taken care of, but I agree that that's kind of a standard department ... |
| # / 56:23 | Richard Brown | Alright. So, there's a lot of important people in this meeting. I'm sure lots of us have hard stops, so let's stick a fork in it there. Thank you everyone for joining. I'm going to stop recording now. If people wanna hang out though, that's entirely up to you, so thanks everybody for coming by. See you next week. |
| # / 56:39 | Chris Burniske | Thank you. |