Governance and Risk Meeting: Ep. 46 (August 1 - 2019)¶
Video | Audio | Discussion
| References | Person | Text |
|---|---|---|
| # / 00:00:00 | Rich Brown | Hello everyone. Welcome to the August 1st edition of the scientific governance and risk meeting. I was just checking the numbers and amazingly, this is our 45th scientific governance and risk meeting. That's 45 weeks of wall to wall analytical actions, so that's pretty impressive. It's quite a library of information to troll through. I think it's a bit of an achievement because keeping momentum up on these things is notoriously difficult and the fact that we've done it as a community is very heartening. It's good that we've maintained that momentum because that momentum is about to ramp up into overdrive there ... Thanks for the correction, David. This is about to ramp up into overdrive. We have a lot of things that are going to start happening in the next month or two. I keep on teasing that and it's generally my policy not to tease things because it's annoying. |
| # / 00:00:58 | Rich Brown | I was hoping to dig into some deep governance stuff today I got bumped because there's more interesting things happening that we can address immediately. We're going to have some a presentation from a new face in this call today, Primoz, who's just been working closely with our risk teams, is going to present some graphs to us, which I've been assured by Cyrus are dope, so dope. I'm looking forward to figuring out what that's all about. Let's talk about ... Well, let's do the preambles and then we'll talk about governance very quickly then we'll talk to, and we'll hear from David about what's been going on in the forums. The preamble is this. We are enormously interested in hearing what the community has to say and this is the opportunity for immediate back and forth. |
| # / 00:01:48 | Rich Brown | If you have a question, please type it into the chats. If you have access to a microphone, please pipe up and just ask your question. You can interrupt us. We don't mind. We also have a forum now. The forum is now my favorite place to be. I'm going to paste the link into the chat now. If you have a question that you feel didn't get addressed in the call or is outside of the scope of the call or just too much for the call to handle, please type it in this form thread. We're also going to be posting links to the videos after the call is over. We'll be posting summaries there. We'll be continuing the discussion. This is where the long tail governance happens. The forum is it. I think that you can refer back to the library of 46 previous episodes to figure out how important it is we consider the forum to be. I'm not going to get into that conversation today. |
| # / 00:02:44 | Rich Brown | What I am going to talk about is one of the first major items on our roadmap that's coming up on this march towards MCD launch is going to be the asset priority poll. We've talked about this in blogs. We haven't really talked about it in the call, we've talked about it in the chat, we've talked about it in the forums. The asset priority list is, our asset priority poll is a mechanism which will allow the community to surface, signal [inaudible 00:03:19] to let us know which priority, which assets they wants the community, the governance teams, the risk teams to focus on first. There's a lot of work to be done. There's seven assets that need to be evaluated with rigor. In order to get everybody on the same page, we to figure out which ones we're working on first. |
| # / 00:03:39 | Rich Brown | I've just posted a link to forum thread, cleverly entitled the asset priority poll. There's some discussions in there about different ways we can go about conducting this poll. What I'm doing here is I'd like to get some signals from the community. I want to know what the community thinks the best way of managing this poll is. There's some context about some of the limitations that we have in our existing system. We've never attempted to try to find a priority list and we don't have drag and droppable boxes and we don't have sliders and we don't have other pleasingly advanced UX mechanisms yet, but we can definitely shoehorn a priority poll into the system. Please have a look and then vote on the poll. Figure out which one you like. This is experiments and example of what I am going to continue to call an emergent process because that's what's happening where as a group figuring out how things should work. |
| # / 00:04:44 | Rich Brown | I like that way of doing things because there's two ways. Let's be binary about this. You can come up with a binder, spend six months working on that binder, drop it with a thud on a desk and then expect everybody to sort through that thing and look for edge cases or you can come up with some basic principles and basic assumptions of good ideas and then iterate over time. That's what's happening in the forums and it's like the crypto-y, open source-y. It's got us this far, so maybe it'll get us even further. What we're doing, the example of that emergent process is I referred heavily, when I was making this post, are two processes that was not created by the foundation. The processes were created by the community and this other thread called Meta governance and signal requests. |
| # / 00:05:33 | Rich Brown | We have a lot of people talking to each other, debating, what's the best way to surface a signal on the forums, and have that lead into a signal and a polling system. I actually referred to this new rules set in the forums to figure out how to do my job, which is a really cool way of moving this entire operation forward, I think. It's was super satisfying. I followed Long for Wisdom, its rules as best as I possibly could with some contributions from the rest of the group to make that the new poll for the asset priority list. That poll will be open for seven days, five days, I think. Once that's done, we will use that to craft the new poll that goes into the governance mechanism proper. |
| # / 00:06:24 | Rich Brown | That new mechanism will go live ... Oh, I'm sorry, not new mechanism. New poll will go live to determine the asset priority lists on August 12th. Before that happens, we will all be enjoying a new version of the portal as well. We're going to talk about that next Thursday. Chris is going to come in from product, will come in and give us a brief tour of what's changed in the portal and explain some of the new tools, new freedom that it presents us to do more than one thing at a time, which is going to be nice when it comes to agility, velocity and give us some insight into how the snapshotting mechanism works and how the differentiation between polls and executives. It's going to be very exciting. |
| # / 00:07:09 | Rich Brown | I think that's all I am going to get into today. Other than leaving the community with a request to hit up that forum thread, make your voice heard, read through the debates. If you have a better idea about how to conduct the poll, we would love to hear it. There's five days to give us feedback before we start working on implementing it. All right, that's it. I'm going to stop there. David, [inaudible 00:07:41] for a very very speedy. |
| # / 00:07:43 | David Utrobin | Yes, super brief. I'm not going to cover everything that is in the governance set of lines threads. If you guys don't know about it, Long for Wisdom has a consistently running thread that he updates pretty much every week with any new discussions that are going on. It's really comfortably categorized. I'm going to link the whole thread here just for you guys on the call, if you guys are interested in checking it out, but I'm only going to cover the two new threads that he added this week. If you're curious about the other threads, you can just listen to me talk about it on the last call or you could just go on the thread and check it out yourself. |
| # / 00:08:19 | David Utrobin | The two new threads that surface this week are, there's one about an explanation of continuous voting, and the peculiarities of the 726 Executive Stability Fee vote. What happened was on the 19th of July, we saw a vote passed with 32,000 MKR. But since then, somebody lifted the hat of an old proposal and this old proposal happened to have around 78,000 MKR. Now, because this old proposal is the hat, in order for the current executive vote to pass, we need to have more than 78K MKR voting. |
| # / 00:09:03 | David Utrobin | This presented some challenges to the community since it seems that MKR voters don't participate as much when it's not an emergency. Back when the peg was drifting below a dollar pretty consistently, everybody was on a high urgency mode, so we saw a lot more voter turnout in terms of raw amount of MKR that was being put up for these proposals. But since then, somebody ... If you don't withdraw your MKR from an old proposal, it's actually just going to sit there forever. |
| # / 00:09:36 | David Utrobin | Although you can't cast the spell twice on a proposal, you can lift it to the hat, which just makes it a proposal that you have to beat. The thread covers this issue and I think actually the next version of DS chief solves this issue, but nevertheless, it's an interesting thread. Actually it's not necessarily bad because it might be that if we really do need to make a rate change, we're going to see the maker governance community actually rally, they have to rally at least 78K Maker to make any real change. It actually might increase voter engagement, but again, we have yet to see it. This is just one of the current things that's happening in governance at Maker. |
| # / 00:10:23 | David Utrobin | The other thread that I want to cover is from Kwadrax. Kwadrax is one of our awesome community members. He owns the MKR Gov Twitter handle and he had the idea to make this MKR Gov Twitter handle a bulletin board for MKR holders on Twitter who want to be informed. The idea was to have a hybrid bot that automated when you votes come out, when interest to pretty much just governance stuff. New votes, interesting threads, even like a maybe timed updates of votes. |
| # / 00:11:04 | David Utrobin | If you have an executive vote going on maybe twice a day, it could tweet, hey, there's 10K MKR with 60K needed to pass. We'll have that signal going out into the public to give MKR holders one more place where they could be informed about governance stuff. I thought this was a great idea. I tagged Mark who created the Maker DAI bot that gives updates on the Maker system already. He said he's looking into it, so it'd be really cool to have a Twitter handle like that, just generally speaking. |
| # / 00:11:41 | David Utrobin | I think this also brings into attention the greater issue of the fact that MKR holders needs to have a place where they're informed about governance. That's not just dig through the forms yourself, dig through everything yourself, even though that's very important. Yeah, that's it. Those were the two new threads that happened this week since last week. Conversations have happened in all the other threads, so definitely check those up, but yeah, that's it for my section. Thank you. |
| # / 00:12:10 | Rich Brown | All right. Thanks David. We have lots to go over today, presentations from Primoz and we also have the state of the peg from Visheshalytics. Cyrus I'm going to hand it over to you to set that up and then introduce whoever you want to have to go first. |
| # / 00:12:29 | Cyrus Younessi | Yeah, sure. I think we're going to start off with Primoz who's going to introduce some data analytics and then we'll have Vishesh come after [inaudible 00:12:48]. |
| # / 00:12:48 | Primoz Kordez | Okay. |
| # / 00:12:48 | Cyrus Younessi | Primoz, you're ready? |
| # / 00:12:50 | Primoz Kordez | Yeah. Hi guys. Because my time is very limited to present to you all these interesting metrics we were working on for the past month or two together with Santiment team. I'm going to give you just a really short introduction. The company I run is called the Block Analytica. I've been working together with Vishesh and Cyrus for some time developing risk model for MakerDAO. During this time, I was also focusing on various metrics that I wanted to present to you and I came across santiment, which was able to provide me with technical support for developing this platform, which I'm going to present to you now. Please, find the link in the chat. I'm going to share my screen. Note also that the platform is still in Beta mode, and it's redesigned. |
| # / 00:13:46 | Primoz Kordez | Let me start. I'll go quickly through all the seven categories introduced. I won't cover all the graphs because there's so many of them, but then, yeah, here we go. The first category we called it network activity. It takes a look at onchain activitiy of DAI or Maker. We're going to focus on DAI now. You can also smooth the data for as many days as you want. Let's check first the address activity. This is the daily active addresses. The DAI addresses that interact daily with block chain and the number it's increasing and the trend is very positive. There's about 2000 active addresses constantly interacting, which is very positive. The orange line is the natural growth. These are the new addresses that generate the daily for DAI, which is also increasing, and very positive. |
| # / 00:14:48 | Primoz Kordez | This spike that you're seeing, this is the Coinbase effect. I'm going to call it Coinbase effect because you're going to see a bunch of times these kinds of effects from Coinbase that happened I think five days ago where they introduced the campaign. The blue line is active deposits on centralized changes and more or less it's flat because more or less everything is happening on chain with DAI, I guess. Every metric can be compared to with either the DAI price or Ether price. You can find some patterns here. I wasn't able to find anything actually, but I was able to find some patterns here, which the metric is called transaction volume, number of transactions basically on chain activity of DAI. What's perhaps is interesting is that every time the volume picks up for onchain activity for DAI and also the number of transactions, the DAI price tends to spike up. |
| # / 00:15:51 | Primoz Kordez | This happened also last year and this year. Most probably, this thing is connected with the fact that when there's increased activity or increased trading of DAI, people are mostly bidding up DAI, so mostly buying DAI. That's probably because when Ether prices falling, people want either to hedge and to go into DAI or what's even more possible is that CDPs they just want to protect their from getting liquidated and they start to repay their debts. That's why we always see a spike in volume and a spike in DAI price because there's obviously a bid. We also have mean and median transaction values. This metric went down from this year probably because there's much more users of DAI transacting. The mean is about few thousand, 5000, the median about 2000. You can see, again, this is the Coinbase effect when there was so many interactions, but really small amounts where we're actually traded to give ... not really traded, but transferred. |
| # / 00:17:03 | Primoz Kordez | The dailies of token circulation. Again, this is amount of DAI tokens that circulate every day, so the amount of tokens that actually are transferred. Again, it's picking up. This tells you about the adoption itself. It's connected to it spikes as well, like we saw before. What's interesting to me is that about three to four or maybe five million tokens circulate each day, which is, I don't know, about 5%. Velocity being about 5%, that's really high for a token. I think Ether has about 2% or something. DAI is trying to be medium of exchange token, that's pretty cool statistic that we see here. We have some top of transactions, but I'm not going to show you. There's nothing really relevant. The next category of CDP stats. Basically, we want it to show the activity of CDPs from locks, frees, wipes and so on, you can compare it with Ether price and what you can also do is you recalculate all the event types to DAI. |
| # / 00:18:15 | Primoz Kordez | I prefer to use DAI because if you're looking at what's happening with locks and frees, which are denominated in Ether, it's much better to look through it, I mean, to show the dollar well, that's why I chose DAI. I won't spend much time here, but again, you can see the Coinbase effect, a lot of ether, I mean, the amount of locks was pretty high, but if you look at the value itself, nothing relevant. Right. These CDPs were just opened, not much collateral was posted or DAI drawn. Of course, locks are above frees because system was growning much more collateral was provided than freed. |
| # / 00:19:02 | Primoz Kordez | Similarly, nominal terms in values. We could see some frees recently, which is connected with larger CDPs free in collateral which we're going to show later why they were doing this or which were doing this. Probably connect it with refinancing with secondary platforms. Similar story with draws and wipes, Coinbase effect here. A lot of DAI. A lot of events of DAI being drawn, but in the nominal terms, basically zero. Again, similar story, some wipes recently, but mostly we were seeing draws. I won't go into details here. Open CDPs. Again, Coinbase effect. What's interesting is that daily they're about 40 CDPs opened and it's pretty constant. They're trying to syncr..., I mean, the amount of CDPs is increasing. This is one outlier here. CDPs users, these are basically new CDPs that issue debt right. |
| # / 00:20:05 | Primoz Kordez | You could see a lot of CDPs at Coinbase issued that, but the amount is really small. Then we have CDP distribution. The amount of Ether locked in a CDP, mean and median. You could see the mean was going down since New Year. That's probably because there are more CDPs introduced with smaller amounts of Ether locked so more Ether using CDPs and probably also because larger CDPs, the whales were removing a lot of collateral, especially in the last few months, so that's why this was dropping. That's why the overall collateral was dropping. |
| # / 00:20:47 | Primoz Kordez | We also can have CDP distribution by locked Ether value based on different type of CDP groups. You could say these are the retail which have zero to 10 Ether locked in the collateral and this would be some whales above 10,000 Ether locked in collateral or 100,000. |
| # / 00:21:07 | Primoz Kordez | What's interesting is that the recent drop in collateral was mostly performed by whales you could call it, especially these two groups, the purple and the brown line. These were the guys who were actually freeing Ether. Of course, they were doing this because Ether price went up and they just kept too much collateral, also because they wanted to monetize on the profits because they wanted to repay some debt. It's interesting because if you look at a smaller CPDs, they weren't really freeing much Ether. It's more or less constant. The whales were the ones actually causing this collateral to drop in the last few months. We have pooled Ether to wrapped Ether ratio. I'm going to stop here. |
| # / 00:21:58 | Primoz Kordez | Stability fee, that's the one that's actually paid, not the one accrued and it's calculated in dollars. We could see there was actually a record happening five days ago. This was connected with one CDP I'm going to show you later, which wiped $10 million of debt in 10 days and it removed 100,000 collateral. This was one of the biggest CDPs. What's interesting is that he actually, in one chart, he paid so much Maker. What's also interesting is that he was paying the fee with the DAI actually. This event might be causing Maker price going up. I'm not sure what happened on that day. We have bites as well. This is connected with liquidations. You could see recent bites. This the volume of them. It wasn't as high as in last year, but if you look at the amount of bites, it was quite high, right. I would say that most of the liquidations that happened recently at the last Ether price drop was affecting smaller CDPs, whereas in the history, last year, bigger CDPs were liquidated. I was checking the collateralization ratios of bigger CDPs is now much higher than it was last year. |
| # / 00:23:23 | Primoz Kordez | It seems that like they learned their lesson or something. They are much more cautious than they were last year when smaller CDPs mostly got liquidated. This is a graph of liquidations at what price of Ether liquidations for the quarter, how much debt would be brought back and number of CDPs. I think there's a bug here. I'm pretty sure. We need to fix this. We also have liquidation distribution again classified by how big these CDPs are. Obviously, the purple line, the biggest CDPs. The curve is the steepest because it's every increment prize drop of Ether that would be most of the liquidations happening from their side. This is something to be aware of looking at them, what are their price points to getting liquidated? |
| # / 00:24:23 | Primoz Kordez | There's only one CDP over 100K I guess because there's only one point here, but he's well collateralized at the moment. This is pretty much a similar graph except that it looks at the share within each group. I'm going to stop here. Mid liquidation price. We'll just fix this one. Oh, no, I don't have it now. Basically, what I wanted to show was the mean liquidation price. If you don't consider the distribution, just how much would the Ether price needs to drop to have total portfolio liquidated. You could just say, "Okay, there's the DAI supply and we want 150 liquidation ratio, how much would the Ether price need to drop to get liquidated?" It's not really relevant but it's good to see distance among them or between them, but the distance cannot be shown now because they're in two different scale axis now. I'll need to repair this graph. |
| # / 00:25:22 | Primoz Kordez | We have the events types again for draws, wipes, locks, frees and I'm going to put frees. The Ether that got freed from collateral or removed. I'm going to show you what happened recently. This is again classified into different buckets. These are the biggest CDPs, the brown and purple ones and these are the smaller ones. As I said, most of the collateral was moved from a bigger CDPs and this was actually one CDP. He removed in two days. Actually no, it was in 10 days, so there are some other larger CDPs doing the same. He removed 100,000 of Ether. It's so interesting to see that he actually traded the Ether to repay the debt. I guess this was happening from 5th in July till now and probably he was causing the price back to go up because he was buying DAI to repay the debt. |
| # / 00:26:24 | Primoz Kordez | After he was able to pay the debt, he transferred to Ether. I think one part, we are used to that as well to Compound. All the increase, I mean not all the increase, but a lot of increase at secondary lenders such as Compound that we saw in last month or last 15, 20, days a lot was caused by this CDP. We looked it on chain, it's so interesting to see if he can transfer Ether there and then he just issued additional DAI. He basically refinanced the DAI from Maker at Compound, but he didn't issue only DAI. He also issued USDC probably because the rates are much smaller. Of course, he issued DAI there because the rates are smaller as well. |
| # / 00:27:11 | Primoz Kordez | I haven't checked if he then later sold that DAI. Obviously, he had and then again built leverage, but it would be amazing to see this as well. Yeah, I think what was happening the last 20 days in relation to the peg and in relation to the secondary lenders was partially caused by the CDP. |
| # / 00:27:38 | Primoz Kordez | Next category we have collateral stats, DAI supply, nothing interesting, Ether, CDPs. We now reached a peak in April and then it went down because there was a lot of Ether removed. In USD similar story, but this includes Ether price and then the collateralization ratio which we are focusing a lot on, we know it's peaked and now it went down because Ether price went down. Also, because there are a lot of Ether removed from collateral, but what's perhaps one of the most interesting charts is this one and they call it nominal collateralization ratio. |
| # / 00:28:19 | Primoz Kordez | It basically checks the amount of Ether, but not in dollar terms, just the amount of Ether versus DAI supply. The good thing about this graph is that it isolates from Ether price, so it captures the user behavior. That's something that I wanted to see because user behavior we all know is one of the most important part at making risk assessment of total portfolio because they could protect from getting liquidated, right. You could see the inverse relation. When either price was dropping, CDPs were either posting traditional collateral to portfolio or they were repaying the debt. Once the Ether price started rising, they either issued more debt because there were more bullish and partially they also freed up some collateral because it was worth more in dollar terms, right. |
| # / 00:29:14 | Primoz Kordez | This graph is quite cool, and you can see the spike here, right. The last few weeks when Ether price drops, immediately they posted additional collateral and they protected it from getting liquidated. Something that we'll probably going to use for when we're going model the user behavior, which is one of the most important aspects of making risk assessments. We have collateralization ratio again based on different CDP sizes. Again, it's interesting to see how whales or the biggest CDPs are most collateralized. You can see here 500%, 600% and more, right. These are the biggest ones. |
| # / 00:30:00 | Primoz Kordez | Of course, when they reached the high number, they freed up some collateral. They're also refinanced recently. That's why you could see falling down, but, yeah, they are the ones which make Ether more safe in terms of how collateralized they are currently. This is one graph that should be working here. Again, this is the nominal collateralization ratio that I presented before. I wanted to see the pattern. Again, this number itself doesn't tell anything. It just the pattern that you're interested to see how they behave. You could see that they all behave similarly except again for the bigger CDPs who were a bit more cautious here and who have higher nominal collateralization ratio than the smaller ones. Let me go back. Okay. Stability fee, nothing important at the moment to analyze. |
| # / 00:31:03 | Primoz Kordez | This is one chart which is sourcing, I think basically what we wanted to show is how much of the DAI that is issued from CDPs is converted to Ether and then put back in the collateral. Like saying, re-pledging the DAI, the debt or let's say putting collateral on leverage on top of other collateral that CDPs pull from their own sources. This blue line is actually tells you the number of Ether that was brought to collateral in addition to all collateral, that was brought in by them. Or, if I correct myself now, the amount of Ether that came to collateral by trading DAI, so by leverage. We know that part of collateral in MakerDAO is purely on debt, right? Because people credit in Ether and then bring it back. I wanted to measure the share of it, so to see how much of the Ether in collateral is actually on leverage, right, was brought it back on leverage. |
| # / 00:32:12 | Primoz Kordez | You could see this ratio. This are actually ... this is percentages. It's 20% to 30%. This is 31%. You could see in the periods, when Ether price was rising people are becoming really bullish and they actually, all the DAI that they issued, they immediately converted it. By the way, this algorithm tracks it in 24 hours. They immediately converted it into Ether and brought it back into the system. It's good to detect when people are highly leveraging themselves. What's also intuitively interesting is, if this figure is like 20%, 30%, it makes sense because the total debt is around 30% of collateral. I'm not saying that all of the debt ended in collateral back, but I would say 20% for sure. |
| # / 00:33:09 | Primoz Kordez | It's good to see, again, you can see the behavior of how people actually use CDPs for leverage long Ether trading. That's quite good a metric I would say, but it's not perfect. The algorithm is not 100%. We're not doing detective onchain work. We just set the algorithm to track these events and I think it's quite okay. Let me, just a second, go to the next step. Okay. We have balances of DAI held. This is something that comes standard when you want to identify DAI demand. You can also check it for Maker. I'm doing to present DAI now. I'm not going to stop here, but I'm going to stop here to show you the number of holders that hold DAI. Again, this is for different groups. You can see again the Coinbase effect, small amount of coins, many holders. |
| # / 00:34:16 | Primoz Kordez | What you could also see is that trend is picking up for retail, which is good, some sign of adoption. Whereas for the whales, for other holders, probably market makers, it's more or less flat. It doesn't changes. Holders of one million and above of DAI, it's only 14, 15 of them. But then we also have total holders balance. This one is really interesting because it basically shows you within the category how much DAI those addresses hold. It was interesting when I saw this. The black line is total DAI supply. Whenever the total DAI supply falls, the balance that falls is of course from larger addresses and that's because these are probably market makers, right. You can see this here happening and happening here and lately as well. |
| # / 00:35:17 | Primoz Kordez | I'm not sure, but I guess this could be a good proxy to measuring inventories of market makers because, I mean, obviously not all of them are market makers with such high amounts, but large part they should be. I think you can track this by just comparing the DAI supply and/or just looking at their holdings and what's happening with them. This is interesting to observe. Another very interesting thing we've been working on is holders breakdown. Now, our job would be to detect all the DAI holders and classify them who they are. Of course, that's impossible, but we were quite successful. We were able to identify, I think this is one-third of addresses and we classify them into seven categories. I'm sure other categories will follow such as market makers if you find them and so on. |
| # / 00:36:20 | Primoz Kordez | Point is, we want to fill this blank gap that's here as much as possible because the other part is pretty much unknown. Interestingly, you can just see the balances below. You can just select, let's say you want to see how much DAI is held on long term margin platforms and you could see it here. Please be aware, these are the unutilized DAI on lending platforms. These are not the issued DAI, but the ones which are just sitting there, which is good because if this part is increasing, that's a sign of demand. It's sign of unutilized DAI, which is positive for the demand. You can also look some interesting stuff such as, I don't know, treasury. You could see Aragon was pilling up DAI in their treasuries, which is great. |
| # / 00:37:17 | Primoz Kordez | I checked centralized exchanges. We follow quite a lot of them, like HitBTC is ... It has three million of DAI then than I think it's kuCoin. What's interesting is that FatBTC, which supposedly has the largest volumes, what we were able to find in their wallets was 20,000 DAI. Now, I'm not sure, I mean, it could be that we missed some wallets. It's possible, but I'm pretty sure those volumes, it just doesn't ... they're pretty suspicious with the wallets and the activity we found with their addresses. Yeah, that explains it. Yeah. That's basically it. We're working on this one heavily and it's going to need a lot of work throughout the time because addresses changes and so on. We have top DAI holders. |
| # / 00:38:12 | Primoz Kordez | Okay. Three more types. It's going to go quicker from now on. Yeah, this category, we basically want to see how much of spare capital CDP owners have. You could imagine that CDP has an address, an owner has an address on which he issues DAI, but the Ether that he locks in collateral is locked into a smart contract. He might also have some other spare assets and that's something you're actually very interested in because it tells you about their capitalization. How good are they? What's the risk assessment of CDP holders, so to say in certain terms. We want to actually measure amount of Ether and wrapped Ether they hold because probably those numbers are the biggest, but we're still working on this one. So far, we were able to include DAI and Maker and you can basically check between different groups. |
| # / 00:39:12 | Primoz Kordez | I'm going to select all CDP owners and you can check how much DAI or Maker they hold currently on their addresses and this is in dollar terms, right? It's $1.1 million of Maker and 200,000 of DAI. Now, you would expect they don't hold much DAI because when they issue it, they of course sell it. They wouldn't wouldnt holding be it if they are paying interests, unless the price is above one and they speculate, but, yeah, mostly this number is low. Here it was a bit higher obviously because they were leveraging when they had this DAI inventory they needed to sell for ether. What's interesting is the value of Maker. On the addresses they have, they hold only $1 million of Maker. I think that's a low number considering how much interest they needs to pay. |
| # / 00:40:05 | Primoz Kordez | If they don't hold Maker on some other addresses, this would mean that if there's a debt repayment they will need to buy Maker. This would obviously mean a positive sign for Maker price, I guess and for Maker token holders. |
| # / 00:40:25 | Primoz Kordez | Okay. Social stats. Here you can actually follow what was happening on social media in the world of Maker and DAI. This context window, you can select any given day you want and you can select the context size and ... let's go there. And of course, with what keywords you want this context associated, either Maker, either DAI. You can also measure sentiment on various sources and these are gif charts of both. I'm going to select DAI and Reddit because Reddit gives you most of inputs. |
| # / 00:41:12 | Primoz Kordez | This algorithm actually takes these inputs from Reddit and then makes some adjustments, makes outputs if the comment is positive or negative. You could see that this algorithm is pretty good at its job because this is the balance between positive and negative DAI sentiment. You could see when USD ... when DAI peg was struggling a bit, the sentiment was much lower. The negative balance, you could see negative balance here, as soon as the peg situation improved again, the sentiment is again better. You could play with this. You could select the Maker or Maker plus DAI, whatever you want, sources, telegram, tweets, discourse, professional traders' chat rooms, whatever. |
| # / 00:42:07 | Primoz Kordez | Okay. The last chart, I'm still working on it. Maker active supply. We have the annual reporting rates. This is again, this is record repayment of debt and Maker that was burned. This is the rate analyzed. It doesn't really make sense because this would be the rate of return if like every day of the year this will be happening. I think currently it's more about 2% here because this chart doesn't include accrued interest. Yeah, this chart could be deceiving just from that standpoint. That's basically all from my side. I think I spent all the time Rich gave me. If you have any questions, we can talk on the rocket chat or you can find me on Twitter, but we are going to probably present some other metrics on some other governance calls. Yeah, that's it so far from my end. |
| # / 00:43:11 | Cyrus Younessi | That was [crosstalk 00:43:13]. |
| # / 00:43:13 | Rich Brown | Thanks Primoz. We're going to ... sorry to interrupt you, Cyrus. We're going to stick around after the hour though. Let's hope people have questions for Primoz. wait till the end of the top of the hour and we can dig into it. Sorry, back to you Cyrus. |
| # / 00:43:24 | Cyrus Younessi | Yeah. I was just saying that was awesome. Incredible work. Let's be sure to post this on the forum and in the rocket chat. I think there's an enormous amount of information here to digest and I feel like we'll all be taking a look at these for the, honestly weeks to come, and have plenty of questions. But this was truly amazing to watch. |
| # / 00:43:52 | Primoz Kordez | Thanks. |
| # / 00:44:00 | Cyrus Younessi | I guess, let's finish up our agenda for the day and then we can maybe revisit this after the hour and get some questions in. Vishesh, do you want to present some stuff about the DAI price? |
| # / 00:44:20 | Vishesh Choudhry | Yeah. I'll go quick. I mean, we just wanted to make sure Primoz got a good chance to go over all those great graphs. I'm just clicking to share. All right. I mean, I think people got a good sense of a big chunk of the metrics this week, so I'll just go quickly through DAI price. It's been fairly stable over the last 24 hours. The spread was evenly distributed around the dollars, slightly skewed down. Then what's interesting is actually if you look at the last seven days, it has pretty much been hovering just above and below the dollar. Over the last seven days, it's netted out to actually slightly above a dollar, even though in the last 24 hours it's slightly below. I would say pretty darn balanced right now. Supply had come down to about 77, 78 million last we checked on it. It's pretty much sat there. One thing that I did want to look at that was new this week and I know we're going over a lot of new metrics, a lot of new topics, especially as we go into risk modeling. |
| # / 00:45:41 | Vishesh | One of the things that I wanted to start to think about and get away from that simple collateralization ratio graph that we've been looking at every week is a more granular breakdown of risks. I tweeted out this graph here, some of you may have seen it, but essentially was just stratifying the DAI supply by collateralization ratio buckets. All of the CDPs, all the DAI that lives in them counted up and stratified into different effectively bands of risks and then graph that visually here is a stack graph to add back up to the DAI supply. Then you can also see here smoothed out a little bit over a seven day mean average. It's really interesting to see, and this goes in tandem with what Primoz was showing, as that DAI supply grows, even though collateralization ratio shot up from 300% to almost 500%, and then back down to around 450% and has now come down a little bit lower, it doesn't tell the whole story because it's really not the greatest measure of risk. |
| # / 00:46:56 | Vishesh Choudhry | Because risk actually more is weighted more heavily on the lower collateralization ratios and much less on the higher collateralization ratios. Above 500% collateralization, there's very little to no risk, right. It would take a tremendous drop in ETH price to create a negative outcomes for Maker holders. Then what we really want to be visualizing is how much of the debt is really living in this red zone? That's roughly what this strives to do is to say, how much of the debt is below, say 200% collateralization or 300% collateralization? That is a much better tracking of the aggregate amount of risk in the system because as Primoz had shown, there's a bunch of large CDPs that are very heavily collateralized and so those actually carry much less risk with them. |
| # / 00:47:51 | Vishesh Choudhry | It's really interesting to see is actually in tandem with what Primoz was showing again, where his graphs shows spikes in "leveraging behavior" around September and December, that actually corresponds very well with the amount of debt that lives below, say 300% collateralization. Those values were at their highest around September and December and have come down fairly significantly since then. You can see just this huge growth funny enough aligned with the explosion of secondary lending platforms, a growth in the percentage of CPDs or rather the percentage of debt that lives at an extremely high collateralization ratio. It's possible and this would take a lot more investigation that those secondary lending platforms actually disproportionately took on riskier loans for Maker, right. |
| # / 00:48:47 | Vishesh Choudhry | We know, we've seen and shown that there was a lot of refinancing behavior, but it would be really interesting then to understand if it was disproportionate refinancing where some of the more risk-seeking CDP owners were the ones that refinance to secondary lending platforms, which would make sense with lower collateral requirements, liquidation penalties, ratios, things like that. It does make sense that Maker has some of the more conservative risk parameters out there right now so it would actually attract some of the less risky lenders and borrowers. That's a really interesting thing to double click on. Just to touch on one other point here, I'm sorry there's a little two day break in the graph there. The fees, a lot of people checked out Maker burn, Primoz touched on this, but the aggregate amount of unpaid fees has continued to grow as we all know. |
| # / 00:49:46 | Vishesh Choudhry | That actually dipped a bit and there was a huge spike in the amount of paid fees just in mid-July, which corresponds to a lot of the volatility, the liquidations, the repayments that we saw and the supply coming down pretty significantly in that same time period. Again, apologies for the break in the graph, but the conclusion here is that the paid fees have actually, they shot up right in the middle of July there to re-pace with the unpaid fees. You could see the unpaid fees were increasing at a much higher rate before, but with this rapid increase and a leveling off of unpaid fees due to a significant decrease in supply, and as I had shown previously with the circulation graphs, a lot of that supply was with older debt before. |
| # / 00:50:45 | Vishesh Choudhry | Those older CDPs had a much higher proportion of the unpaid fees because they've been accrued for quite a while. Somebody had recently shared in the Maker chat their breakdown of unpaid fees versus potential liquidation penalties and they'd shown that the unpaid fees don't in any cases come close to exceeding the liquidation penalties, so that's not a realistic concern at the moment. This is I think extremely positive for Maker holders because they're realizing more of those revenues and there's less risk of unrealistic revenues. Those are positive stats. |
| # / 00:51:30 | Vishesh Choudhry | To touch on the secondary lending platforms again, real quick, because this is, I think Primoz touched on a lot of the Maker stats. For the secondary lending, there was this run up in the rates and the utilization on the secondary lending platforms just at the end of July there. That has come down a little bit, so that's pretty interesting. The DAI supply overall has come down and the 20.5% stability fee has held. There wasn't a change there that could have caused an increase in supply and potentially an increase in supply in secondary lending platforms, but what's interesting is it was not moved. This is just stabilization from what's been going on with ETH price and I think some degree of delayed effects, which was something that I was stressing last week is given some time, some of these metrics do tend to like go out, which is I think basically what we've been seeing happening because the DAI price, which was sitting above had come below and then gone back above. Given a little bit of time to oscillate, it does even out to about a dollar. |
| # / 00:52:42 | Vishesh Choudhry | Then, on the secondary lending platforms, even though this utilization was peaking, as the DAI supply came down a bit, the utilization on these secondary lending platforms came down with it as well. The excess supply here shot up relatively recently, which is pretty interesting. Without significantly more DAI being minted, suddenly a decent amount of DAI was supplied to secondary lending platforms. The borrow volume has been pretty perfectly flat. It came up slightly and then came back down a bit. Overall, it has been flat over the last 10 days. |
| # / 00:53:31 | David Utrobin | Vishesh, I've a quick question. |
| # / 00:53:33 | Vishesh Choudhry | What that means is even though there isn't significantly more demand for DAI, there's isn't significantly more DAI being borrowed, the overall float of DAI is actually less. More DAI was added into secondary lending platforms, which means presumably there was some large purchase of DAI and someone is holding the bag on that 20.5% stability fee and someone else just obtained that large quantity and supplied it to secondary lending platforms, is the most reasonable explanation. There is your condensed version this week. Again, same as Primoz, open to questions afterwards, happy to double click on anything that ... I know it was a fair amount pretty quickly. |
| # / 00:54:22 | Rich Brown | We're approaching the top of the hour here. I'm not sure if we have any other formal presentations. Please, everyone we're happy to stick around for another half hour so stick around. Matthew, we still have to get into your weekly narrative as well. Do you want to start that now or did you want to ... are you ready to kick that off? |
| # / 00:54:46 | Matthew Rabinowitz | I mean, basically that's the main point, which we talked about a little bit earlier was related to the voting. There are really three or four points to talk about related to the voting spell and the desire to have some semblance of an expiration or a lifespan as it relates to when a spell is submitted and if it doesn't get implemented within X period of days, weeks, months, whatever the number is, we all agree on, that it's somehow find a way to expire. In the scenario that we have right now, we're at some 78,000 Maker committed to one given spell. Today, that's something the community can overcome. Right now, 3000 some odd Maker on the burner wallet, not that big of a deal. When that number, you add a zero to that, as time goes on, it becomes a larger and ever present order issue to get rid of. |
| # / 00:55:37 | Matthew Rabinowitz | The ability to muster the number of Maker to get rid of it, the easier solution is just to have spells expire. There was a forum post, people went back and forth and apparently there is a way to implement it technically. I don't understand the nuances of how to do it, but it's something that the apathy related to voting is one thing. The apathy ability to be even able to vote is one that, without being too dramatic, it can present an existential threat unless we address it. That and in concert with the fact, right now we have the pegs, it's more or less holding and the stability fee is still elevated but functional. Question really however now is after an enormous amount of refinancing, which is a very healthy sign in an economic system, we've never been this size, never had the amount of DAI issues that we have right now. |
| # / 00:56:29 | Matthew Rabinowitz | If we find ourselves in a scenario where demand is driving the shift, not really our ability to restrain the outstanding supply, if the peg starts to break upward, by default, we will be that the natural response would be to lower the stability fee to cause the creation of more DAI to then re-harmonize the price. But by doing that, and we have no data to support what the elasticity would be in terms of how much DAI would be needed to be minted or created to keep that peg, I'm the broken record on saying the debt ceiling is that subject matter that really needs to be addressed sooner than later, especially as demand goes, if demand really is driving the ship. Now, you've mentioned, Rich, a number of times that the portal is coming and if it comes relatively soon, it's not an issue. |
| # / 00:57:21 | Matthew Rabinowitz | If we get ourselves where that portal gets delayed, the ability to vote gets delayed and our ability to get signal from the community becomes a much greater issue to have what priority of what vote needs to happen each week because if we're locked into one vote. One signal for a poll, and then one signal to actually make a decision and put it completely in serial, then there's a much more pressing issue, I think there. It doesn't look like it's an issue, but I think it rapidly could become one. Any other points I can bring up in the chat session after the fact. |
| # / 00:57:58 | Rich Brown | Yeah. I agree with you, Matthew, and it's a good ... possibly, it's an interesting segue we can circle back to Primoz and Vishesh questions. But as far as governance goes, I'm always hesitant to offer promises about launch dates, but let's call this like a 98% confidence level that next week the new portal will be launched. Once that new portal comes out, the gloves basically come off on what it is we're capable of doing with governance as a community and we're going to have to start talking about some things. Then we've identified issues with the ecosystem, we've identified improvements. Some of the examples are, let's talk about exponential rates stepping, let's talk about changing the cadence of the votes, let's talk about X, Y and Z, let's talking about the debt ceiling. |
| # / 00:58:54 | Rich Brown | Therefore we've been comfortably living in the world of theoreticals at this point. That's going to change because once the new portal is there, we're going need to start digging into these things. I think that that activity is going to happen in the forum and we need to make a concerted effort to start digging up some of these potential polls that we've been discussing over the last couple of months and figuring out how we get them into a cleaned up states and refined and then starts seriously considering putting them into the polling system. |
| # / 00:59:30 | Matthew Rabinowitz | With those polling systems, do we know if we're going to be potentially exposed to the same type of spell, lack of expiration issue with that new system? |
| # / 00:59:40 | Rich Brown | Well, that's our ... Sorry, that's another great point because this is something that needs to be talked about as well because as far as ... I'm trying to pay as much attention as I possibly can. It's been a busy couple of weeks to the chat and into the forum. I've seen some discussions that there's different options that work here and this is something that we've identified for a while as a problem and I've seen it myself when I've tried to run scripts against the chain to figure out who did what when. The first time I did that I was horrified to learn that there's still somebody who is staking their Maker against the 0.5% stability fee increase which was the first one we ever did. |
| # / 01:00:17 | Rich Brown | We need to figure out what to do with these presumably abandoned blocks of Maker and there's different options on the table. We can either just set a time when it's like after six months that Maker just gets sent back to the wallet, maybe after three months that Maker can be poked by a keeper that say like, "Hey, this Maker is not participating anymore let's put it back into that person's wallet." There's also the possibility that when an executive is voted in, and another executive is voted in, there's no going back anywhere, so all the Maker that's no longer applicable to come back. |
| # / 01:00:57 | Rich Brown | There' lots of options on the table here, yeah, that we need to think it about though. We can't do anything for a while arbitrarily. |
| # / 01:01:04 | Matthew Rabinowitz | The biggest point I guess that's important is that we get a vote and we get clear signal that fixing the issue is something that needs to get fixed. |
| # / 01:01:15 | Rich Brown | Yeah. I agree. And obviously, that starts in the forums. So, I think that you referenced the forum thread about this that I haven't actually stumbled on yet. |
| # / 01:01:24 | David Utrobin | I could be repost that. |
| # / 01:01:27 | Rich Brown | Cool. Yeah, we need to figure out what that looks like because there's game theory at work here, we need to think about, there is also signaling at work here. Think of it, there's more options out there. Like it doesn't need to actually be a contract change. It could be somebody throws up a webpage, it says like, "Hey, look at all this Maker." Like it just puts them visibility on stale and or abandoned stake weight, which gives the community an opportunity to do some awareness campaign type stuff as opposed to [inaudible 01:02:03]- |
| # / 01:02:02 | Matthew Rabinowitz | That article, the forum article makes a great technical argument and I'm all for helping define ... the key point I'm trying to bring up more than anything is the humanity side of this. If you develop a system that's not designed to only run a year, but really to develop the run several decades or who knows how long it'll run, the humanity of all of it starts to kick in and no one wants to be the guy who gets hit by a bus. But we really hope the guy who's got or girl who's got 78,000 Maker doesn't get hit by a bus right now. |
| # / 01:02:35 | Rich Brown | Yeah, that's another interesting, I think that the entire crypto space at all hasn't quite wrapped their heads around what happens to your abandoned tokens. It's not just a Maker problem. It's a DAI problem, one that people talk about once in a while. |
| # / 01:02:49 | Matthew Rabinowitz | In a scenario that's so essential for voting with the intent of active, vibrant voting, it becomes more acute. |
| # / 01:02:57 | David Utrobin | Yeah. There's also- |
| # / 01:02:58 | Rich Brown | I agree with you. This is ... go ahead David. |
| # / 01:03:03 | David Utrobin | Yeah, I was going to say, there is also Cmooney who put up a comment that actually looked at the UDS chief and multi collateral DAI. And I think that this case there we're experiencing of single collateral DAI is actually like it's prevented in the code by not letting previously lifted proposals to the hat be lifted again. I think that's what the code says. I'm not a coder, but like just looking at it, I'm pretty sure that that's what it means. But for those of you that are interested, I linked his comment. |
| # / 01:03:36 | Rich Brown | Yeah, that removes the surprise aspect, but it still doesn't address the stale and or abandoned problem. And Matthew raises a good point that potentially we need to move staker back ... staker, Maker back and forth in order to secure the system. And if a whale or a consortium of whales or delegate systems come into effect and then that Maker gets abandoned and/or stuck on a proposal that we can't overcome as a community, that's a huge problem. So I think that to, it's not unsolvable problem because at that point there's drastic solutions that are available. We can vote to slash that abandoned Maker, but I don't even want to consider the followup that would occur in that situation, but there's things that we can do. But in the short term, we need to get into the forum and all align and start collecting some really good ideas and then figure out what the best way forward is. |
| # / 01:04:42 | Rich Brown | I'm kind of partial to the poking stale stuff, but I'd use it to figure out whether it's a transparency issue. We just read only page somewhere, or whether it's something that resurrects this conversation that we've had over the course of last, or maybe it's been a while since we've talked about this, but we have a situation now in our ecosystem where, yeah, in crypto in general, that there is no way to notify people what's going on in the crypto world. There's no push mechanisms in this world because what if for whatever crypto anarchic reasons people are not attaching email addresses to their Ethereum addresses. So letting people know that something is up is tough, we can't push, people have to pull. And this might be another mechanism where somebody from the community or somebody from the foundation comes up with a mechanisms to notify people that something has happened with their Maker, with their CDP, with DAI in general. It's a tough problem I think. |
| # / 01:05:53 | David Utrobin | So do we want to move to Q&A for Primoz and Vishesh? |
| # / 01:06:03 | Rich Brown | Yeah, sounds good to me. David, were there some questions in the chat related to those presentations we can dig into? |
| # / 01:06:09 | David Utrobin | No, I didn't see any explicit questions in the chat, but I did have a question for Vishesh about the phenomenon of more excess DAI showing up on Compound now over time. So like, I guess I'll just ask him a question. Right. So, do you think that now since that amount of DAI, excess DAI is increasing and the borrow rate is decreasing, it's going to cause this positive feedback loop of even more people refinancing their debt from Maker, DAO to Compound and thus further decreasing the total DAI supply. Do you see the cyclical continuous cycle that happens until like an actual demand floor is reached? |
| # / 01:06:59 | Vishesh Choudhry | So, the first thing to think about I think is the pathway. I think we tend to focus on prices and supply values, but I think thinking about the physical pathways for obtaining these assets and how they wind up where they do is extremely important for understanding this stuff. So, for example, if you think about, if supply is actually contracting and at the end of the day the amount of DAI that lives on Maker has not significantly changed, and the amount of DAI being borrowed on secondary lending platforms is not significantly changed, then you have to think about what did change. And I think this goes back to some of the leverage graphs that we were showing. So the CDP distribution graphs that I had shown or some of Primoz's graphs on DAI for ETH trades. Like in the last seven days, this stat was about 28 million DAI sold for ETH. |
| # / 01:08:12 | Vishesh Choudhry | That's not insignificant volume for a seven day period. There has definitely been higher. There's definitely been lower. It's middling range, but there is very little price movement. So there were a couple of spikes in both directions, but that tends to happen when somebody is desperately trading on a very high time preference. And so they incur a lot of slippage, or they end up depleting the order book all of a sudden and they experience either particularly high or particularly low prices. That tends to happen with DAI because I think the liquidity on these order books is low, is relatively thin and so in short timeframes, it does experience a fair amount of impermanent market impact from either selling or buying behavior. |
| # / 01:09:02 | Vishesh Choudhry | So, my point is I think the most reasonable expectation, if none of those other stats have significantly changed in the past week is that there was some de-leveraging, someone gave up on their ETH long position and so sold some DAI and either got a really good rate for it or took a haircut because they were the ones who had the high time preference and they were just trying to sell quickly. There were some trades that went down into like 95 cents. |
| # / 01:09:34 | Vishesh Choudhry | Given that, I think it was just someone somewhere sold a significant amount of DAI and then someone else bought it up and decided to supply it on a secondary lending platform to get some interest out of it. I don't think it's like a big macro trend or anything is my point. And I think this fluctuation, this actually does touch on your question. This fluctuation will continue to happen where the utilization on these secondary learning platforms will go up, the rates will go up. Then somewhere, somehow some more DAI will be sourced. Some over supply will find its way onto secondary lending platforms. Rates will go down a little bit and then you'll oscillate through this I think to the extent that some oversupply exists. And so I think this is just an indicator that okay, a little bit of over supply does still exist, which means given the supply went down significantly, demand didn't take a hit. |
| # / 01:10:35 | David Utrobin | Cool. Thanks. All that make sense. Well, I had a question for Primoz, in the very beginning of the presentation you said that there is a little thing that smooths the data. What does smoothing data mean? |
| # / 01:10:56 | Primoz Kordez | So it just calculates the moving average, so simple moving average is just average is the last time like seven days and it makes an average of seven days so that the line is much more smoother. |
| # / 01:11:13 | David Utrobin | Got it. |
| # / 01:11:26 | Vishesh Choudhry | So maybe actually I'll ask a question of Primoz or we can just start a topic discussion. One of the things that I'd be looking at is and I've shown this a couple of weeks ago, was transaction volume of DAI. So transactional volume of DAI picked up rather significantly especially compared to trading volume of DAI, which has been low to middling range for it's longterm average. To me this is primarily as a result of a lot more usage in smart contracts and a lot more DAI being passed back and forth through the DeFi ecosystem. But one of the metrics that I've tracked over time is the amount of DAI that's being held by CDP owners. And so it's interesting to then talk about completing the cycle on that for selling DAI for ETH. |
| # / 01:12:22 | Vishesh Choudhry | And so, you had mentioned that one of the metrics that you were tracking was the addresses that own CDPs that then go and sell DAI for ETH so you can track the leveraging behavior. I've found that very, very low, like all time high, like 4% of DAI is actually held by CDP owner addresses. So maybe you can pretty much answer a little bit about how you think about tracking, how that DAI moves hands if it moves like multiple addresses, right. So people may have multiple contracts or multiple addresses that [inaudible 01:13:09] chain ownership of those CDPs and subsequently where that DAI changes hands. Right now what I'm thinking is that very little DAI is held by CDP owners, by and large when they are minting DAI, they are passing it off rather quickly. Do you have any thoughts on that? Is that what you're seeing as well? |
| # / 01:13:36 | Primoz Kordez | Yeah. The inventory that you're talking about, the quality inventory, so that's the amount of the DAI that the owner hold is pretty low. I mean, I showed now it's 200,000 before the time points, it was 5 million. And I guess that inventory gets higher when they're leveraging more. Because some CDPs are pretty sophisticated and they do transaction actually in, you could say in the same cash, they just issue it and they just send it to exchange and it's done. So we don't see it on the address or it's like immediately happening. But some, they trade it really slowly, so they always have some amount. They don't issue one million of DAI and then they have this DAI on their address and then they sell it. But what they do, issue I don't know, 50,000 of DAI because that's the only thing they could sell at the moment because the markets are still not that liquid and they don't want the slippage. |
| # / 01:14:40 | Primoz Kordez | And then they do this in small increments. Right. That's why when you do a snapshot of a CDP owner or a DAI holdings, it's actually pretty small at the moment. But if you like we did, check in 24 hours and you just sum it up and then you see how much they bring it back, the number is much much higher. Right. Because it's this time differences that's how these things are happening. Does this satisfy your answer to your question? |
| # / 01:15:14 | Vishesh Choudhry | Yeah. That answers my question. |
| # / 01:15:19 | Matthew Rabinowitz | Cool. We could also follow a logical act of a rational lender or borrower, right? If you're borrowing against your asset, simply holding the DAI in your own wallet burning interest doesn't really make sense either. You'll get rid of it. |
| # / 01:15:38 | Primoz Kordez | Yeah. What's interesting, at some points we saw 5 million of DAI in the inventory so, I guess at some points, some people don't mind holding it for some time. |
| # / 01:15:49 | Cyrus Younessi | Can you go back to, or can you talk about the chart that had the mean liquidation value? The difference between that, the one where you said the access was messed up on the left side. I didn't quite catch what that was, but it sounded interesting. |
| # / 01:16:11 | Primoz Kordez | It's not really that interesting. Yeah, because it's the mean. That's where it went, let me try to find it. No, I think it's next one. No. It should be here. So you mean this one? |
| # / 01:16:30 | Cyrus Younessi | Yeah. |
| # / 01:16:34 | Primoz Kordez | Yeah, this tells you ... how to put it. So let's say DAI supply, whatever it is is 75 million and you would say collateralization ratio is of course, 50% higher. So you would multiply with 1.5 until it gets to a number, what, 100 million and something? And then you would say, "Okay, was the collateral, the whole collateral in the portfolio, how much is it worth?" I think currently it's ... no, I need to check, but I think it's, yeah, it should be 300 million. And then you say, "Okay, if price drops by 50% like here when you measure the distance from whether 200, Ether price to 88, all the depth would be liquidated." But you're looking just at the main, right. So that's how- |
| # / 01:17:31 | Cyrus Younessi | Okay. I got it. |
| # / 01:17:33 | Primoz Kordez | Yeah. It's not that relevant because this wouldn't happen because you have CDPs which are more collateralized, some less. But the distance between two make sense. I mean, to have an average portfolio picture, the risk assessment. |
| # / 01:17:47 | Cyrus Younessi | Isn't that just the collateralization ratio? |
| # / 01:17:52 | Primoz Kordez | Yeah. Basically it is. Yeah, it's more or less the same thing I guess. |
| # / 01:17:59 | Cyrus Younessi | Okay. Yeah, that's what I thought. I was just making sure. I don't know if it was more than that or not. Okay. That's cool. |
| # / 01:18:16 | Cyrus Younessi | There is a lot to go over here. I think we're going to spend some time going through these charts. You also mentioned that you guys were working on some other stuff. Like you guys already have future roadmap ideas for more metrics. Any idea for what those might be like? |
| # / 01:18:39 | Primoz Kordez | No, but I mean, as soon as we started this I saw how many stuff could be developed or shown. So it just happens with time. Probably the biggest challenge will be identifying DAI holders as much as we can. It's a bit of detective work and you need to readjust addresses and so on. But it's good to have it. |
| # / 01:19:15 | Maksim Ravno | We could have that signaling, because basically we have on the platform and identification and signals anomalies or important events. So if it's really a request for this kind of a feature. |
| # / 01:19:29 | Rich Brown | Yeah, I was just about to mention that it'd be handy to have important signals or a hyper-relevant tab of a snapshot of the health of the system or some particular views I think might be useful. Because, there are a lot of things to click through on that page. |
| # / 01:19:52 | Maksim Ravno | So you would mean there is a kind of health check and then some flags where you select, okay, let me know if there's something on this screen, which I believe important it just suddenly changes in part the threshold or specific percentage or some sense of direction. So I think- |
| # / 01:20:08 | Rich Brown | Yeah, something like a specific dashboard would be interesting. The alerting mechanism though, that's something I consider. Sorry, what was that Maksim? |
| # / 01:20:21 | Primoz Kordez | No, I just wanted to say that what's perhaps missing is there's so many charts and of course not all of them are really relevant. Some of them really it makes sense to look at them occurring now and then, but some of them could be used on daily basis. Obviously we would need a front end dashboard just to focus on those I guess. And then signals and so on. |
| # / 01:20:44 | Cyrus Younessi | Yeah, I think we've already internally brainstormed some ideas for what a front-end like a public facing, almost like a retail dashboard would look like. So if there's an average user coming onto an analytics site I think they'd want to see some of the more important CDP statistics, stability fees across the various collateral types, DSR when it comes. And then for example just speaking at a glance, I think some metric of distribution of DAI is useful. So there's 90 million and 10 million of it is in Compound or whatnot. Inventories for market makers, if that's the thing, maybe anonymized- |
| # / 01:21:36 | Rich Brown | Maybe you're on to something that we could look at it by demographics. So like, I'm a voter, what do I care? |
| # / 01:21:43 | Cyrus Younessi | Exactly. |
| # / 01:21:43 | Rich Brown | Or I'm voter, interested in the stability fee, here's my single pane of glass or I'm a risk analyst for ... all right. I'm pretty interested in the DSR. Here's my single pane of glass. That would be really cool. |
| # / 01:21:54 | Cyrus Younessi | Right. Stuff like that. Yeah, maybe we can sync up offline and discuss some ideas around that. |
| # / 01:22:07 | Maksim Ravno | Yes. [inaudible 01:22:13] |
| # / 01:22:13 | Cyrus Younessi | Especially when there's going to be multiple collateral types. It's going to be interesting, actually if you think about it and imagine this whole website times a hundred collateral types. |
| # / 01:22:26 | Maksim Ravno | Yes. I can imagine a lot of fun. |
| # / 01:22:40 | Rich Brown | All right. We're coming up on this 30 minute mark of the Q&A aspect of the call. I think this might be a handy place to end the meeting. |
| # / 01:22:52 | Rich Brown | Thanks Primoz for the presentation. There's a lot to digest and I'm hoping that we'll see some more interesting graphs in the future calls. Thank you once again Vishesh and David and Matthew for your presentations. Please continue the discussion in the forums, how to look at the signal request for what the asset party list might look like and have a poke around because there's a lot of interesting chats going on in there and a lot of decisions that need to be made and touching on Matthew's point, we have a series of polls that we have identified in the last couple of months. And in the next couple of weeks, we're going to be able to act on those. So let's put some effort into resurrecting those threads and maybe refining them for implementation. |
| # / 01:23:47 | Rich Brown | Thanks everybody. That's it for this week. See you next week. |