Hashing It Out
Hashing It Out

Episode 47 · 2 years ago

Hashing It Out #47 - MakerDAO pt. 1 - Nik Kunkel

ABOUT THIS EPISODE

MakerDAO has built a token that is reasonably stable with the USD. We have the great pleasure of interviewing Nik Kunkel, Head of Backend Services at Maker. In this Part 1 of 2 on Maker, we talk about the cryptoeconomics behind creating a stable token pegged to the USD. It's essential to understand this before getting into the technicals found in Part 2 of 2 (next episode, #48). This is compelling work, and truly amazing how well it's standing against the throes of the fluctuating markets.

Links

https://makerdao.com/en/

https://www.reddit.com/r/MakerDAO/

https://twitter.com/makerdao

Now antering kindwelcome to hashing itout, paanfor retock to the Seck intovator's mindblock in infrastructureand decentralized networks. We diev into the weeds toget at Wyan how peoplebuild this technology, the probise they fase eon the way I'm listing o learnfrom the best in the business. You can join our Rek Wokin back Castin out episode, fortyseven and as always I an her house stuck acort petty with my cooast, mytrussy co host Calin Crochet, Sai Hallocon show I don't know who that is my real name's,crug R, rgright tick with usy. Really, I'm really Craig Right. I think that'smy actualwhythe w! Why can I get that actual name out of my mouth properly?Like every time I pronounce t, it's got like a little thing going on, so whatit's like, Ri, Wi, Idur Aexao, go. I go full elmerfudd. Whenever I try and pronounce his name. I don't understand that one UM forthis episode. This is something thatI've been wanting to do for a long time. Actually, Um or H, be U to maker DalTeck Cunkell had a back in servce as you make bow, and this is a system thateveryone h has become quite big. In fact, I think the last lethio mediumblog said they have gained the largest user group so far in terms of docentralized finance. But it's IT S. it's I guess their main product is astable coin called the Dow, but the way in which it works, isn't isn'tstraightforward. I'v always wanted to try and get at least qualities, texplanation or have a conversation with someone about how all of this works andtees to put together and why people back it the way they do so welcome tothe show Nick. Why don't you give us a quick introduction as Tolihe? You are Oyou got ino space and and and be concerned from ther surl Um. So I got TAT inte space aroundlike two thousand and sixteen. Actually, I was working at IB on Hyperleger,which is their kind of propritary kind of private, watching, Um and kind of you know, Etherim was wasmaking the rous and UH. I I got hooked on that and I wastelling my boss Hey. You know. I know e spend a lot of resources on thishyperleger thing, but you know etheorium ind, like public chains likethis. Is You know really where this is the game changer at kind of moment andh? You know it was always like shut up and eventually I was like. No, you knowyou shu up like I'm Gonto leave and, and I joi makeer and Um. You know it's a great story like you made the right choice, my oldboss at IBM ADM at Atkonsensus last week, O Walpert Yeah, wo Werbert YeahMhm, and it was funny because, when I left he told me I was making thebiggest mistake in my life and now he's a he'smuch nicer bodies like Sher Te,Hubble, pithere, Lat's, let's, okay, let's uh again so the way I think most people have a viewof maker and die. Is that Makeer? Somehow produces die, maybe enoug,that's the case. TSCLEARED OT, the crace we'Lt get under that, but whatmost people are doing in terms of the disincialized FINANC interacting withguy? Is there basically bonding Ath, just somewhere wo et o what that is? It'inn, it's overclatealized and thenby doing that they meant die which, as that they put eft somewhere and theyget seveny five percent of that Ind. So...

...if you put O hundred and fifty dollarsworth of F and E, smart contractor we' talk about you're, going to mint ahundred dollars a die as a loan, that's Cllateralizedby nertherm, and then theydo stuff with that. So let's get into how the hell that worksso firs ourself, is that accurate is that is that an accur representation?'CAUSE! That's what my reading was. That's wh t thi Coriro was that'sfairly ACR Y, okay, cool, okay, let's before we get into that part of it,which is kind of like I say what most people could to consider. I don't evenknow about that. It's very interesting part and that's what most people areusing, but it's backed by this massive other system, the macodol. What is that so I mean we're Gonto have to realizeyours, we're not trying to be. You know, M we're not trying to be likea bank run by some. You Know Maker Corporation, what we've done here aswe've created an autonomous of smart contracts like that operate completelyindependently from from everything we do right. So when when, when you say, for example,someone comes and walks up their eath and we give them die, it's not USgiving itto them it's. It's. Basically, we've enabled people to take out aloanfrom themselves and that's kind of what I what I think the big Dan Changer ishere right Um. So you know just to kind of reiterate what you said earlier. I Auser coms and you know they th have some eat and you know maybe they wantto tap into the equity in that eath, where they they don't want to sell, butbut they'd like to tap into some of that equity ry. Then we reallow them todo that or maybe they want to you w go along on Atheoriom. Maybe theywant some capital to h to play around within in the market, but to see a goodbuying opportunity. So we allow people to tamp into the equity and inthorasisOye hold on. So that's actually the verse of what I think crit said. So, ifyou so like my understand is that if you you lock up o e hundred and fiftypercent of the dieyoreceipt, so you receive a hundred die. If you put in ahundred fifty dollars, I dcaractrsthat's correct. So so that'show's that, like not like that's that aloner locking up the CLARAS ASE Thi's,like taking less than what the actual collateral is worth Sothat, you don't really lose ownershipof the comlateral right. It's not like. We take away your eat, tea is stillyorse. It's just! You need to pay back your loan in order to get it back. Soit's it's essencially a secure, a securead loan. If you will right on your own undern essen o before weget into the wise causethere's a lot of them. I kindo Wan I want to get towhere guy comes from and wiit stable and what maker it is because there's awhole nother thing which, which is the maker token and that's use forgovernance over ow diismented andte. Indeed, I guess the interest rateassociated with Di so sur what s? What is that whatismaker? How does that work and how does it? How does that function across theelike the the set of smart contracts, that people are Oiso? So what we'vereally done here is we've identified that there's Um there's a couple fdifferent types of users in the system so far weave only talked about one.We've talked about the type of person that is trying to effectively. You knowaccess the equity in there in their eath, but there there are other users to ite. So,for example, there are people who just want to use that, and this is actuallythe most common user. We have right. They they don't want to take out a lolof Hem. They don't wanna, like you know, lockup eave or anything like that. Allthey want to do is they want to use die Righ, so you can hedge, you can use dia hedge right. So if you think the market is about Totur Rie, you can sellyour eathin to die. Ond O can sellwer other tookens that can happen to dieand write and Mor protected because the...

...is stable and so that kind of brings into goesinto your question. While you, why is thy stable and how does this and CRtoken I kind of play play at school and the role? The way that I like to thinkof the mpro token is, is really a way of segregating risk, so as a die user, you don't want to haveany risk. You want to have something stable, so you're trying to get rid ofyour risk. Well, the Makor token is the exact opposite. It's trying to absorball of the risk reward of the system, and so it's Righ it. It's very volatilein that sense, because it's Isicotet so essentially if the system is runwell and by rung well, you know it means that Um. You know there is lots and lots of Dyeiand circulation and all that die is generating lots of fees and all thosefees go towards basically buying up and burning TMKO Toket, sothat makes them care flationary if it's wrong. Well, if the system is not beingrun well, Um, you know there are Um, there's a bunch of underwater loans andwhat I mean by that is that if w we talked about earlier, how youknow you need to be minimum hundred and fifty percent collateralized on yourloan? Well, the reason we have that is because, if IDRO, if the price of beefdrops like we need to liquidate your lown, we essentially need to to margincall you I M, and so as long as we margin call you above a hundred percentcollateraization. So that's fo, one to one. You know value of eth to the valueof the dye that you borrow right. You know the system is fine, but as soon as you cross that threshold,where the value of the E is worth less than the value of the loan, now youhave it underwaterlone riht. No one will ever pay back. You know a hundred. You know dollars to get back.Ninety seven dollars worth it e Righ, it's not rational. So how does thesystem kind of stay solvent during that period and that's where the mkinortokenkind of comes? In so remember, I said Riskal War, while this is the this isthe risk, so the MPR token has to cover all of the bad debt so, for example, nthe previous example, I said you know et you're a hundred and fifty dollarsworth of Eath drope to ninety seven dollars and your loan was a hundreddollars. I means there's a three dollar difference that and care holders needto cover the way they cover. That is that thesystem automatically start commenting new NDKR and then selling them in Kartuntil the deat is covered. Okay- and that actually makes somesense m to me- and- and I kind of I kindo like that-approach a little and that it puts a speculative side of it- really does balance out thethe lending side and the investment side and like the Grosside, with thespeculate deciding like how. How well is the markets going to perform andthat kind of thing, and so you're actually kind of creating a derivativein the maker token, whereas the actual dietself has tried to sort of like linkthe two logical enities of theorium and the? U S dollar together. So that makessense to me and also the whole fact that, like Y, if you drop to ninetyseven dollars, the whole system is kind of compensate four that loss in threedollars, et Cetera, etcetera t was out of like a hundred tokens. You lostthree O riht, whatever Um but H. I think what kind of bothers meabout that scenario is that it doesn't at first blush sound, very likeresistant to things like flash crashes, and so I'm wondering how does that workin terms of like very radical changes?...

That happened? Suddenly so so, actuallyI I think it is covered, so one one thing I forgot to mention about the MPOtoken is I that it's governance stoken. So the reason that why uncarefulgersare willing to absorb the risk alword of the SYSTEMIS, because theiradministrating system they make all of the important decisions. So they saywhat type of collateral is good collateral for these loans. Rit Rightnow we're just using e, but in later later this year will be relucing,multi collateral die, which will enable all types of all types of tokens, andso it's governance, 's job to say. Well, which token? Should we do? What shouldbe the minimum like collateralization ratio? Should it be less than o ehundred and fifty percent? Should I be two hundred percent? Should it be? Youknow four hundred percent Um and you know what should be, what shouldthe fiest for that particular type of collateral be and- and these are allkind of things that are relevant to the quality of that Cillotal- that arerelevant to the risk of that clatter. So when you say oh well, aren't youguys not accounting for a clash crash? I say e. We do count for a flashcrash.So basically in your model, you have to say: Well, you know. I think there is ayou know, a point. You know five seven percent chance that there is a flashcrash in this year, and so, if we have you know this many wollns outstandingand this collateral type and we're taking on that type of risk, you knowhow much do we have to be minimumly coloneralized. You know how many feesdo we have to charge to make that worth it in the end, what Nkr holders aredoing is they're, providing their an insurance policy right, they're,basically providing an insurance? No, for you know insuring this Collat andthey're saying these are the premiums that you have to pay right, but then okay, but then you're,depending on the mkr users and I'm not really sure how that's set up at themoment. We could get that into that next figure, depending on the peoplewho are participants in the NKr token, to set the rules for the to correctsoits. It's basially a voting mechanism and ninvestment mechanism, as well asyour busiclly voting with your money, which I love by land of that in everyform. Whenever it comes out, I think voting wih for money is an amazingconcept that we need to explore more, but that's another topic Um, but they're voting with their moneythrough the Dow and Um you're, depending on them to makerational decisions or to know other variables involved with this. So there's a social problem there M, andso, when the whole system collapse. If that social problem sort of manifests,I don't, I don't think it's a social problem. I think it's it's a feebactofwhere, if mkr holders are not educated, ind making good decisions- and they youknow, don't make data driven decisions, theymake decisions based off of hight. You know we want to get Shcoin xascollateral, and you know we WANTA have the claneaization bese Super Low andthe INTERES PCO Berlou they're going to get screwed and in that scenaro theykindof deserve it. But the point is that the risk is segregated, so themaker holders making bad decisions they're the ones that get punishedthrough dilution right and someone else can basically come in and watch thisand go like this is being terribly manased. They can buy a bunch of outcare for cheap because it's been so terribly managed. All of the votersthat made these terrible decisions. Their voting power has been diluted, and now you give the opper. Youbasically give Um extrinsic value to anyone that can come into the systemand start making good decisions well, you'v. just what you just propose then,is a possibility of a hostile takeover of the maker, like F, if a scenery likethat were to manifest H, not fair,...

...saying that well take over it's, it'snot a it'. It's a plotocracy, riht ie it it's a lutocracy and it's you votewith your money, a that. If I made Yov decisions to go reboarded, if you makebad decesiens, you don't Gare award it, and you know from a dying user's pointof view like the ultimate goal is just tha die estat right, so whatever themoney managers can of have to do on the e MPR side, that rescue is isolated tothe. How is THY staple may explain that, butI haven't quite Ike Rockyet. How is how? How are you keeping that peg sure so sothere's a couple of mechanisms we actually use and a couple that we havein the Pipelik, so the primary mechanism Um, is actually quiteintuitive if you think about it. So if e the system itself always kind of USie as one dollar right, but that may not be in line with what Marcus aretrading a right markets are absolute and theyll. You know have perfect pricediscovery and they'll find O. U know what what they think is worth. So,let's come in t a scenario where die is trading above a dollar in the markets.Well, now, there's n arbitrage opportunities here because effectivelyanyone can mit die right. Anyone can lock up some Yoath and Minc some dieand effectively flood the market with this newly men to die right and so theact of flooding the market and selling it you kno at a premium at this aboveone dollar crisis, very appealing, because after you've, you know dumped abunch of supply on the market, then you reach parody with the Pegagat right andnow you know you can basically pay back your loan and you can you've arbitragedthat, like you know what is it? Three cents spread o two: Some spread: FiveCet spreat whatever it is. The opposite case is when dieis tradingunderneath the dot right and when dies treading underneath the dollar. We haveto understand that every single dye that's in circulation, has beenborrowed by someone. So there are constantly people with open loans whoare deciding you know all the time. ShouldI pay back, mylond E, Not Pa, and if you can pay back your low atninety seven cents on the dollar or ninety eight cents on the dollar.That's pretty appealing right! So you effectively, you know it's effectivelya negative interest. Loan right like say, I borrowed a hundred die yesterday and today die drops Te,IAE cents and I you know, return back. I buy back. You know the dye for ninetyeight cents and like I've, now made money from taking tat a low. So sothose are kind of h. The primary mechanisms m. The secondary mechanisminvolves governments UMSO, the secondary mechanism, is u you need to have a balance like anequalirium between the demand for taking on leverage and the demand forholding die Rit like a very simple concept: R, it it'sjust supply and demand like how do you balance supply and demand and thesupply comes from. You know how many people are trying to take out. You knowloans with eath cllateral, so they're, effectively going evil, Yo long oneathlie deman from Leveragon and how many people actually just want ta whole dieright. That's the that's the demand! So what we've seen in the last few months-or maybe it was not months but like the last month or so- was that I I wastrading underneath the peg Um. You know it was, and it was pretty consistentfor H for a while. We were trading between like ninety six or ninety sevenand ninety eight cents, and the reason for that was because there was too muchdemand for creating die right for taking on leverage vursus demand forholding Niet, and this was because the...

...market was, you know, extremely bullish.People thought that you know the bottom was in and the market was going torebound and everyone wanted to be in. You know on the speculation side, theydidn't want to be on the stable hedging side, and so the lever that we have toplay with here is that we can control the interest rate on the loss, and sowe went all the way you know when we watched we had an interest rate of halfa percent. Currently we have an interest rate of nineteen and a halfpercent so effectively. Every week we have a public governance type of meeting whereall of the and carholders C meet and discuss. YouKnow How is the peg doing? Do we need to raise rates? Do we need a lower raceand ye w? We haven't just gone up like we've gone down a couple of times aswell. When H DI was training above the pig and effectively Y, you can look at it.Like you know alone, you know going along oneath. You know with a ten percent annalone like that's pretty appealing,but once you start hitting nineteen and a half percent, that's not as appealingto many people, and so that's how you suppress this demand for creating newdye and that's by suppressing that demand. That's how youv reach thatequilibrium stay around a dollar ther question: Um is the interest rate dynamic? Is itvariable across the lifetime of a loan or I it Esthatic armime? It's it's adynamic is's, a dynamic rate M. So so that's intr point you bring up right ina HAP. It has to be dynamic because otherwise, right, if we change theinterest rate only on new loans coming in right, that doesn't give us the theoath that we need right to get existing users to close out their positionsright ut. It has to be that way. I just feesas though like no one's willing to give up their loans exactly well. If you think about it,the stability fee is nineteen and a half percent annually right now. Whatdoes that mean? I if someone takes out alone, that means they think eath willgo up more than nineteen and a half percent in a year from now- and you know I I can't blame them like I, Iwould imagine a good pochaablare overlatertonsense that, like there likethey, basically okay. I imagine this is a scenariobased on my understanding of how things work so far is that people haveclateralize, throigh eate for die and then done it again to buy Mortin andthen colatalize that from more die and may be a number of times so they're,basically like a few iterations in associated with their originalcolateral. What type of affect do you think thatcould Hav orlike w? What wh t is there danger in that there has to be dangerin that, so Si yes and no right, so the you canachieve up to Ou know by doing thecycling that you just talked aboutby locking up eth. You know drawing dye by using that die to buy more eat,locking up that he'sdrawing even more dive rite. You can achieve up to threex leverage, but the thing is that no matter what you're always hundred andfifty percent collateralize and Writn, if you have, if you're riding on thatMax three Xleborate, you know if it even goes down a penny, I you getmargin cald now the risk that you're talking aboutis um the kind of a y. u what do you callthis in trading? You call it kind of cascading kind of liquidations rightwhere you now, if one major whale say with a huge CDP, gets liquidated rightand now you have, you know a million dollars if eas get deliquinated at once,Rin that pushes the price of be foundyven more, which causes morepositions to go, riquidated and so on and so forth. That is effectivelyalready a factor accounted for in the risk model that you use. A governanceuses one defiting, the risparameters...

...for a theory effectively. That is theliquidity risk of a theorm. It is how much you knowhow much the way you can thinkabout. Liquiado risk is how much eath can you dump on the market withoutcompletely? You know bottoming out th the price of ebelow below our buffer.So we have this concept of a debth sealing in our system, and so it is themaximum amount of dye that can be minted or a particular Acet. Currently,for the end, the debt ceiling is a hundred million dilar, so ry. Now theDI supplies that around eighty million eighty one million a few months back, we were at ninetyseven million and people were asking us. Are yougoing to raise the debt seiling Ar Goin to raise the best silingy, becausepeople were concerned that they wouldn't be able to take out that theywouldn't be able to generate any more that and at the time you know, We'erelike we were looking at the liquidity kind of a eaft anlike? No, you know wekind of feel comfortable with it ad a hundred million Rinot, but you alsohave to realize that the liquidity risk is a constantly changing thing. I Ethdoesn't have the same liquidity every single day. Right knothere are days thesupersuperhigh volume their days with no volume their days where you knowthere are very, very fine and spread some days where the market is more thin.So it's it's up to MP holders to constantly be reevaluating these thingsand making adjustments. You kN W cosstantly okay, so I want to get a little bitinto the user experience in a moment before I do that. I think it'simportant of a cover like how. So when I hear like I'm locked upmy collateral and I'm received, I is it y trainable exchangeable. I meanlike because to me that sounds like it would mean that you can't really domuch with it. It just says that you have this much in value locked up, but I'm not really sure how, like thesystem changes when you're able to trade it so i'Mnot, even sure, if it'spossible, can you so so? As a user, when you, when you lock up Athin, youcreate die, that's you're die. You can do anything to ith e nurse twenty tokenand we use it g US or lot rigt, and I I've seen that s t the problem is thatI'm trying to figure out is like okay. If I spend that to like I s, then thatmeans that I no longer have enough lets to say it could drop my collateral sothat I no longer have enough. Is that right or do I am? I misunderstandinglike if I te the die that you created iscompletely funtable with every other die and the alone has nothing is notinterlinked, with your die or completely disconnected so okayif any.If, if your loan position gets liquidated because the price ofbeithdropped it doesn't affect the dye that wastaken out with through that Lon I see so. Basically, I would still have thatDi that dies. Mine E butthat duyis fine, if, if, if it drops below then baseballthe'll just take off the amount of the drop and be like okay, you you receive,like I say, half your assets of what you collatteral we're, calling yourmargin, here's the rest of your money back, but we're keeping the restexactly. That'Si t wheres that margin called F go, so it goes to an so effectively. Whatwe do right now is. We have oricles that are constantly reporting the priceof Eath IAND and that's how we kind of know that we need to that. We need amargin. Call someone M A and an interesting point here. It's not us doing the margin, calling it's actually kind of random, like people, and youknow, funds and organizations and Boghts, and the reason why...

...people are doing these margian calls isbecause they get to have first sdips on buying the collateral effectively. Sowhat we do when a position gets margin Claus, we sell it at three percentbelow the price that the orcles are reporting. So that means that you knowinterestrading at a hundred dollars. We sell the eth at ninety seven dols andanyone can buy, and so there is a huge ecosystem of Boswhich are doing Arbhitruph and they, you know, they'll, look at unislopthey'll. Look at timer they'll, look at to die the look at radal relay to lookat dedecks to look at you know all the ZEROEC stuff, um they're, basically trying to arbitrage,write the e that we're seling with et somewhere else. I they have to buy itwith maker. They have to buy it. Witd die okay wel. This is so right, youralone to won it and die. So you we need to the system eeds to recover a hundreddie, and so we just sell your eath until someone pays us a hundred die foryour eath, then Eeran that Di Thati. No, now we're I work with athree percent difference. H, Wowho makes up for that differene,because you're Burninat die it's gon in circulation. Is that something thatmaker, like the maker, commudity absorts? Well? Technically, it's the the personwho locked up the t e Ethin Thelon rigt. That person is absorbing that hat threepercent loss. It's effectively. Just a diansentive toever let your your loan get liquided, riht half course mom. So one ofthe things that I want tomention here is that this mechanism of liquidating Um of liquating these loans is about to bedeprecated in our next version of of die an I will ti catall that andinstead it's going to be replaced with an auction mechanism, so there will bem Um, onchain auctions that you know have a that are basically selling thiseath and anyone can participate in effectively what happens if people are,you know we need to recover a hundred die right. That's that's what we needto recover and we have you know, let's say a hundred and fifty dollars worthof beef, so we put all of the eath in the auction and we wait for people tobid and we say hey, we need a hundred docks, so maybe you bid seventy five Di. I bid ninety five die and then callimcomes INI's like il I'll bet, a hundred bucks and now the auction changes into areverse auction, where okay, every we've reached this hundred dying leveland everyone we don't need to make more than a hundred die. Now the auction issaying how much less than the total eath are you willing to take. So youknow Callin's like well. I want all these I'm like. Well, you know, I onlyreally want you know. Oi'll take. You know a hundred andtwenty dollars worth at Yo teeath for a hundred iy, and then you know you comein. It's like you know, W A I'll pay like a hundred and four dollars. Foryou know a hundred dollars wit tab and no one outbids, you great, you won theauction. So this way m auctions are kind of a pure price discovery modeland I think not having liquidations linked to oracles is going to be a lot,a d more flexible when there is a liquidity,crunches right, so you you talked about it early. You know what happens if youknow a bunch of s. You know a bunch of loans, get liquidated right, inthsecascating liquidations. Well, if you sell that eath at a fixed price right,you may not get enough buyers, but if you have auctions, then for everysingle auction you theoretically have perfect price discover yeah. Well, oxosis perfect. Price tof discovery, something that's kind of like I, I meanit's, it's approximate enough tat...

...counts in reality. It works really wellbecause it does create a marketplace. It does depend on the nature of theparticipants in the auction Um. It doesn't necessarily represent value onthe whole, but it is a really good way of just getting things done. So Ireally like that again. actioning is just another type of voting system. Inmy mind, 'cause money is just boing Um, so yeah pretty much. So what I? What I I'm curious about is howyou chose that aupction model Um, you know, and whether or not th, theactual auction bits themselves are visible and how you built that Um. Soin other words, is this an enchaine auction Um? If so, how are thetransactionls completed? Can People see the bids? Are these closed or open bidsand then what determines when the, when the reverse action occurs? Is it parodyor is it time frame or both ogreat questions by the way, so the everything is completely anchane, not controlled by US and completelytransparent. So when we launch MIL teclateral die rihe there will be akind of auction daskpor that anyone can go and participate in auctions. They'llsee every auction wey'll see every bid, they'll see who is making these bidsyou know, part of being you know, on a theory means every single action isfundamentally transparent, which I think is great m. So on your kind ofquestions on the specific SOM actions Um, the auctions have a kind of maximum time that they can run, but they also have a time after bidexpiration. So, for example, an auction t could run for twenty four hours Max maximum duration, but from a bid, duration, point of view t couldthe auction could end at any time in that twenty four hour period after noone has bed for more than one hour and sosomewone you'll see very quickly?Is that bid will come in? You know immediately because there there's an incentive toto kind of be first liht, because threthere's also this kind of skipfactor right. If you you have to be at least this much over the the previousbid. Rit like you can't just you know, increase it by like one fenny or oneSutosi right. You know, there's a some constant factor that you have tobe larger than than the previous bit, and the reason that you also don't havethis kind of you know. Micro, you know, micromicro like outbitting out biddingo ianything is because there are gas cots that people are realizing right.So it's actually from a you know: Paphilos plenifvew, it's actually verycost ineffective to have to do a lot of bit. So the the simulations that we'vedone show that people are, you know going to come in around. Like you know,eighty eighty five percent, you know immediately with with the first bid totry and and walk in some some kind ofprompositionng there does. Does that make sense. That makessense. So what made you choose the the auction? What other actionsmethodologies did you look at, and why did you intimately go with the once wereach limit? We go into reverse auction system mechanism. I think that makes alot of sense, but there's there's a lot of different ways to do in option. Imean a lot, and so I was just wondering how you arrived at this partie I mean II agree. We actually use UM differ, there's three different types ofauctions in our system. Actually, the liquidation of the collateral is only one of them Um. Forexample. We also have the this maker...

Byinbern mechanism, so I think Imentioned earlier that you know all of these H. outstanding loans are generatefeats right and those pees are used to effectively buy up MPR and then destroythem in CR and wit. Now, there's Lessan can supplys. So each and cower is morevaluable right 'cause it has a larger clam on defease Dowan. It has a largerkind of voting power than it previously did so. The way that we do that Byinburg iseffectively that we just got paid a bunch of Dye. You know if ea now wehave a bunch of die, and so we sell that die in an auction and people haveto bid in UNCR, and so we say: Hey, you know we have. You know we have thirtyDi that we've collected in fees and you know we want to get some MPR. You knowhow much MPR is anyone willing to give us for this three Di. You know peoplewil Bideve ben eventually, you know someone someone wins to that auction Um.The third type of auction is what you know. This Thi MPR dilutionthat we talked about right. So what happen is when you have underwaterloans and the system needs the kind of a mince new MTHEY R and recover somedying or to make up that difference in the in the Lon Um, and so the way thatoption works is that the system said this ation says youknow we need to raise. You Know Twenty Div Um, how little n Kr do we need togive you for you to give us twenty dollars, soeveryone is binding. Twenty di twenty dies. T one die, but you know callingyou may say: Well, I want five maker and then corecan come in's like youknow what I'll do it for just one maker and then I'll come in and be like I'll.Do it for like half of a maker? And you know no one underbids me now. I youknow, pay my twenty die for half of Makr. So there's a lot of different UMauction formats in our system, feeling, like very, very particular type of RolsM, due to kind of their they're characteristics. Yeah, and I don't wantto pick on Um Pik on, like the auction park parts of this too much and ther'sa lot. We sle you have to cover wel yeah, one of the things I'm thinking islike God, if we had better privacy mechanisms or the transactionsthemselves could be hidden until a reveal like we would have basicallyavailabldin for things like vicary options, which would save people toneon transaction fees and so and also would arive at a pretty reasonableprice. So I was just kindof curious with the thought process behind youknow the reverse auction beep. First oction in the revers, the naturalaction and the reverse auction was but m. It sounds like a lot of thatrevolves around just the nature of the desetialized Systemat, using as a court.Truth Mechanism M not having closed bids readily available in any like realway. Um or maybe maybe it is- I just don't know about it, but I just don'tsee a current way to do that. So I just kindof curious Um. Yeah I go ahead said makers is, is a ES, the alsending loans of the feesthat hapepen as used to by upmaker and burnit. So then what I was, what IDconsider a saint right, Otio like kind ofun credit, typical terms formodelring things, you a sinks and sources that would be a sick, so I'mtrying to is there? Is there an alternative source, an that type ofthing 'cause? You don't have like a stable equilibrium for this. For thesefor this value flow, then, is there a potential future where maker is liketcirculation to Makeheris so low that it's it's not usable Um? I wouldn't say that so I think th,the one that you have to think about it is that these stability fees are constantlyconqle constantly coming in, and so these auctions for buying mkr m youknow in exchange for selling these DIY...

...is Um is happening all the time I'shappening continuously. So the system is essentially the permanent buyer ofNCARE, and so you know, as you know, well right now, ther're one millionmakers, so you know there's a lot of people a lot of maker in circulationthat people can sell. But, as you know, more maker gets destroyed, m yeah,there will be less maker available, Tho purchase and that's naturally how youget a price increase in in N Kr. It's not bad that there's less maker incirculation right, you know, Mkr goes has you know is token with eighteen decias we caneven imagine a world where there's only one M kr les, but you can split it up.You know the one, the ten to the minus eighteen and t ijui, think about that.A lot of the Times cause a lot of people talk about burning mechanismsabout actually thinking about what the pole supply could evengially dindledown to say, for instance, this works for fifty years, and you have thisconstant, slow burn of things. If you don't have that ability to break thingsapart, appropriatel 'cause, like the price of a single dye, is arbitraryif it could be broken up at Te. Eighteen de, it's just a unit for for,like T at's easy for humans to understand, but unless you have that ability tobreak it apart, then you may have petential problem way down the line itif incredibly successful. I just try and always consider those types ofthings. 'CAUSE collegy theory works really well forthese sites of Value Floas, I'm trying to figure out how token ECNOMICT works.I'm not sure I makers a huge project. You have atremendous amount of equality, people who look inthese things, but there areprojects that are I as just KINDOF, curious Onlik. What what the answer that was and likerefreces, like you, know, runaway runaway, either inflations ordinflations of OFA Gian token? What that could mean for the system as awhole? It's it's tough to suss out, and Ithink that's one of those things that I was tryng to get at earlier is it. Itseems, like the the crypt economic model behind this is extraordinarilycomplex. So how do you vet this model rig like? How do you like look? Who canlook at this and actually go? You know this works M, any sort of degree ofconfidence whowho's behind that is that kindo going to help you with yourconfidence question there cory or not. I didn't know that Ting about alof.This is that you get a lot of really great price discovery and and metricsfrom the use of the system. As it's been used for a lot of times, you cansee I stuff in action right like the workall in itself, is a really goodway of figuring out what the actual like decent price atheum should be.Coraying H, especially when you have these these exaptions, and so in a lotof ways. The maker SYSTEMIS IT STUBIC natural price, discavery witthebestonhow people feel in the entire market. Wutl. What's interesting to meis that you say of a theory. I mean that was kind of going to lead into oneof my other questions. It's like, if you open up other assets, is this tothe setralist exchange nick. What do you think about that? Is this actuallya desentialized exchange in principle, bat you're, starting with a very simpleconsept of a stable coin, thats revolving around the? U S, dollar asthe basis for bilding that out. I I don't think we're we have a duecentralized exchange. I think we have kind of a decentralized credit facilityright if we thinkthat. What I'm really excited about is the fact that you know right now: kind of crypto project,tokens or or only the first step yer. You know you have a lot. The kind ofgreat promise of blockchain is that, instead of having kind of localizedmarkets, we have globalized markets. Rightso insteat of you know an Amazonshare, just trading on the New York Stock Exchange. You know during openhours, you can have a tokenized representation of f an Amazon stock.You know as a as near C twenty token and they ha kind of what I'm saying.Isn't it slet's Yu say if the die is the foundation for w the principle ofvalue determinator for what this is?...

Okay, an yeah is stable relative to U Sdollars, which may may not be relevant in a hundred years. Who knows, butlet's Yu say for right out stable in terms of Q dollars. Okay, anybody couldtake that die and then go and bid on Um collecting a like cllateral or some sort of valuefrom another another system. Another, like let's say, said: OFETHER IT'LL, beI don't care light point okay and then I am actually using the same token,which I used, which I put my cholateral in and got or I purchased off Rabroader market, but I actually got this die in hand. It's given to me, or it's some manner I can actually use,have to purchase another asset. How is that nod? In exchange? There's alsomargin trading. There's there's lending their speculation like it has a lot ofthe features of a decentralized exchange Um. Only it's logical insteadof like some sort of coordinate, setralize fforwhich, I guess, is thepoint of e centralize exchange to begin with. So can you differen? Maybe I'mmissing the pointere ut to me? It sounds a lot like you're buildingtoards an actual desentialized exchange. I I think we're definitely enabling Ummarketplaces M, which I I think is amazing, but we'renot closing the marketplace right agree were giving people access to credit andwhat they do with that die right is up to them, whether they want to go onsome decks and then buy more tokens, whether they want to go to cinbase andconvert that in the dollars and go purchase a car or a home or Um. Youknow just pay some bills Um, you know that's that's kind of where,where our role kind of ends is we give you the die now you go and use it yeah,it's basially about the infrastructure behind it, the sentralizchangs, whatI'm kind of hearing that it something co Ti Taroen! He explains itself. It'sReh wired for a quality oe't exchanged yeah, some type of UH itlikew, mosttreading things do like when this first started do ther had to peg to the thecoins. All training paiers were pegged to BNC, and then people started doinglike stable coins like tether, which eallkind ofyou a Olot. If you want tobut e, you need something to peg every trading pair too, and what's Nice aboutdie. Is that te in the future of ditill be multiclateral, which means it's notrelecated to the risk of it. A single entity, sinceit crit, it'spotentionally, much more stable than all her things, not relying upon asingle thing to make it staple ani, sopatly, a good point, so n in anotherkind of stability mechanism that that we didn'ttouch on early, but so as soon as you have more types of collateral than justeat. Now the job of uncarfolders is almost to kind of do a portfoliomanagement strategy. where, if you take all let's say you have ten differenttypes of collateral, it's. How do we make this entire portfolio ofcollateral and optimize it for the least risk? Somaybe we only want ether to take up. You know, fourteen percent of that youknow of the total corfolion, so we capped the this debt ceiling. I wastalking about EALLE RECAP: The maxilal Mount of die. You can mince m againstEthir, you know a you know, say some number and you no, then you say well,there's this other token digits. You know, and gold is not you knowcorrelated you know. Priceogold is not corrolated with the prices of Cripto,and so oh you know. We probably want a little bit bigger exposure of you know.Digict so maybe you'd be say. Twenty percent of the total collateralcorfolio should be digits because because it allows us to derisk theportfolio and what you find is that, even if you add risky types ofcollateral as like say, for example, you have something super supervolatile,like I don't know like a...

...just making up tha mater fro Sheperim. I um it's not bad to add that as acollateral type, there is no bad collateral. There is only notsufficiently accounting for the risk right, so you can say: okay, we're onlygoing to give you a quarter of a percet total exposure in the total cllateralporfolio, and you know what we're going to charge you fifty percent interest,and you need to be three hundred percent colaterats right. So it's headsfun, yeah. So as long as you adequately account for the risk, you can actuallylike. Nothing is bad every like there's, nothing good, nothing, bad! It's justall numbers you know need good models of the governor's part then can assessthat prisk. If you don't have goes by his WAG TI. NOCARE had USS like we're,not sure like how these makopeople are going to react to like this kind ofstuff. Well, my question there is like once you do. multiplateral die peoplestart doing doing these risk models. There's a bad choice on a singlecollateral affect the other. The other ways of the undertokens like this onesay you' tell you: You'd, have a terrible rescascessment for one of thesituations and that just falls flat that that really really screw over therest of the makers. NUCO system a only effects and CR token holdersyeah, but like one of them, so you really need like get 'em all right. I Mtrying to think of like a Zer way to like as an a way doing. multiclateralGy is e way of diversifying orisk across a multiple different asset class,ar multiple assyes right that have different risk profiles, but it'sSAFFOR instance. If, if it Soso happens, you screw up the risk assessment of anultravollacal thing, you don't want it to like, have a leveraged effect on the rest of theecosystem. You would like it only to be relative to whatever tings thit. Don'thave a leverage effect. So, like saying, like you know this tronicaline that wegave a quarter of percent exposure in our corfolio right completely fallsflat and Groes to zero. That's OK because it's only a quarter percent ofthe total portfolio, so there's a very minimum amount of like n kr that eds toget diluted o what what you're kindof going into is what if they get it likecompletely wrong right, like they thought this thing you know was supersafe and it got you know, should occupy ten percent of the collateral portfolioand you know really should only occupy that orover percent. I mean in thatcase yeah, like they, they screwed up when they're going to get and prisgoing to get diluted. When that, inevitably, you know all those loansfall flat, so I think the question that I I would want to ask is a personparticipating is this is how do I prevent myself from being subject tothe tragedy of the Commons you explaini little Mar Fel Pal Budget,but idiots or a cabalof idioncs or a bunch of people who are manipulators insome way, just maybe Socialy, Engineeri, bad narrative or something come in,create something that actually hurts. My my standing, ok m how as anindependent user as somebody who's, just like just a guy in his room likeAbento, have enough. You Know Dye tond make this interesting to me. How can Iprevent myself from being damaged by them, making poor choices on behalf soas as a dyeholder, you're, really you're, actually really safe, becauseibetmaker atter, a artoken which has a market cap of like six hundred, a fiftymillion which you know can get diluted a lot to cover so or I think of outthis way. There's eighty million die out there, backed by five hundredmillion dollars worth of east, which inturr is back by another six hundredand fifty million dollars worth of Nkr. I guess a dieholder, like you know I,in terms of you K, ow, when companies go bankrupt, rigt there's an order of ayou know, there's a hierarchy in which creditors get paid out. First dihlornumber one in the system always always always hoks sorst in the trunch, an theSD innela. So as a dyeholder, even if...

...the system goes tits up, you get yourone dollar back. So that's! That's! That's Paramo, sowhat Um? I really meant maker, I said, die Iapologized if you're a maker holder- and you seelike it's, not a passive job. This is not a token that you just kinda likebuy and hold at air portfolio and like see how it does in five years. This isa very active type of thing because there are constantly decisions beingmade and if you see bad decisions being made- and you know you recognize thatthey're bad, but you you know you're Yo know you mentioned tragedy of theCommons right. You have all the the minions voting for something and andyou're like no, this is terrible. Then you should sell and it's profitable foryou to sell right because you can buy back in later after the effects of the denegative repercussions. Of thatdecisions have played out. That's a great modelin theory. I Oudimagine people T it's. The rational way to do things is that what you're,seeing in terms of the use of m of maker holders, what percentage ofvoting are you getting from the total amount of maker? So e? It's beenramping up for a while, I I think anchane governance has been write, hashad traditionally very, very low races of participation. Um, I think themaximum we've Gottenusfar has been out of the total one billion maker. We gota hundred eighty five thousand m PR, a vote Um, and so so that's abouteighteen, a an a half percent m. What I really like about that, though, is thatm. It does not include the mkr that we have in our dead fund, which is about.I think it's a little less than three hundred thousand n Kr right. So reallyit's. If you look out at the Torbal Ane of IX, seven O R, seven e percent ooday,and if a hundred EIGT K is voting, that means we're getting aout thirty percent,almost thirty percent of Alionpavo, also the vcs who have and funds whohave acquired Mkr, have not started voting yet and locked up at I havnbaced, dollar and Racon. They have not voted and they have significant. Youknow I mean you can look up the token contract o write. Ah, it's alltransparent. You can see exactly how much they have so they haven't startedvoting it either. So in terms of the the Little Guy Um, you know voted we'reseeing what I think is among the highest kind of rates of voterparticipation in the entire space. I guess that number is like the effect IVcirculation isn't quite shown by the number of Macerbans if Youso what it'slike. This is an engineering pocast after all, and the CRYPT ECONOMIS S isextremely interesting. In fact, it's super interesting because engineeringand crypt economics AREB, are are so closely interclined, especially at thispoint in history. That may not be true forever, but for right now, it's likeyou cannot be an engineer without being aware of the crypto economics side ofthe thing that you're building, I find that very fascinating, but we've hadalmost an entire hour of podgast and our hour along piecass talk aboutalmost purely lecrypts economic side. Let's o a little ti do too, but Ireally need to know some stuff about the engineering you'r back in serviceslead tellone's about the makeer back in why co? We Trust Tus tell us about theoracles. What are you using for those like what Um? What is it built in? Howdo you? How do you keep things honest if you're maker Dow is a smart contractand it resides on therium h? How do you integrate other coins into the CIscenario when they don't have that same? You know like a facility available, so wh wh.What is this is that first of I wantto, maybe start we're almost that an hourhere and that's a lot to discuss. Would it be worthwhile Tho just have a wholenother procast on engineering, I'm happy to come back, coon Plat, youthink about that, like basically like we just force peope this ones, a d.They get the concept of maker before...

...disguise like how they're able to dothis type of things, an guarnt I'm down for a part, two yeah sure, let's do it.I think I think I don't. I don't want to rush this right. I think this is animportant concept. It's maker is a its a very complicated system, and it's inmy pinion done very well. I would like to maybe try this again and discusskind of th engineering, dsitoe yeah, that that sounds great to me. If Y U'R,if you're use ox Siko hearing me by now, like O, go I ar no, no, no, I D, Not I've already had plenty of peopletrying to say like she could make her on, because I have no fucking Cud owthet work and it's just it's being used a lotties being passed around. I manage a lot of good qoint bounies it status and that's what everyone wants in terms oflike like, like a usuaal currency, that they can, I guess hedge or focomfortable with the value they get bas done when theyr work start doing, work.Things like that. So there's a lot of uascases here to wrap up, maybe we'llstart with, is I like? Is there anything that you wish? We would haveasked you that we didn't get around to that we're not going to talk about thenext hour. Oh Man, Um kind of the where you knowwe've talked about where things are wow. I want to talk about where things aregoing. PERFEC, you all the kind of exotic type of things that you can bilgon top of Dye, like the entire financial system, like this entireecosystem of like terivatives and slops and everything you know, contracts fordifferences. Everything is based on having kind of a stable coin as ' primitive,and so as soon as you built that it's really like opening pandulors box andyou can build all kinds of neat stuff on top of it yeah this year. Maybe thelast of the last part of last year is definitely like the rise of stablecoins or better stable coins or an attempt to be a better stable coin, andI think, a 's speaking to the importance of Um its requirement to build financialsystems. So I they're going to see a battlethere in terms of wo, like maybe market share, of stable coins, but based on the way a theory works. This seems likethe best way to do it, an a centralize market driven fashion. I I certainly think so, like hat gourse,I'm I'm protocintialization in a lot of context, so maybe I'm biased there but o. What are you most excited about? Whatdo you want to see built? First, first Um, I mean there's a lot of low hangingfruit here, so I think you guys touched on earlier like Oh, these are variablerate loans Um. You know most people when they take out alone. They don'twant TA variable way. Long Right, I wasn't going to say to take plelike,that's kind of part of the problem with the financially harslyat impression Ito,nay. But but the thing is: that's not really a problem, so what you can do ism. basically, some third partienity can just use our infrastructure as a backend and say calculate the risk. You know what isthe risk that the the rate is going to change. You know in the next year, andyou know what is kind of the bounds of what it would change too, and so youknow a d if the rate right now is nineteen and a half percent when theygohm like, I think you know the rate's still going to go up later to hear, buthow high will it go? Maybe it'll go up TI. Maybe there's like a forty percentchance. It goes all the way up to twenty eight or something like that,and so they'll calculate that risk and they'll be like okay. So you know if Isell it at a fixed rate to the user at twenty four and a half percent, I can.I can make money yeah. So you see a lot of things like that that seem of anobvious pobleor, the risk of the variability of the rate and will offerit to users. You know repackage begain, as a fixed, straight low, could see a lot of things bilt. On topof this, I just like it for Escroratin any typeof vestro service WH. You enter into a...

...contract that that money guess put up,but you would like to know that the value of whatever service you'rerendering is going to be the same when Yhou're done rendering it and that'swhen that money gets released. So tepor I die on on Auger, I think, is a greatFiti atwont to speculate on. You know half of an eath when you can say don'tWantto, you have like a five thousand dollar position on this B, I mean:What's stopping pap from doing a lot of these things right now is the fact thatthey can't reliably model around the price of t e coin you're able to model around the dye.There's really. No, it's basically a bridge between paypound and the CRYPTAworld. Like that's the way I look at it and maybe and like that's what we need.We talk about adoption and, like that's the missing key. That's why people whoworking on thing Ath t other is because they know that's that's a missing linkbetween is is at stable point. Now, I'm going to be honest with you, it's verydifficult to determine, especially from one our interview reading a White Paperin a purple paper whether or not this actually works in practice, but I think,like t cause, there's so many details, ofol, so many attack factors, socialattack, bectors and, like I don't know enough about the actual trust side ofit yet, which is what I'd love to get into in the next episode on with you,but at first Blis it seem treasoable and sothat's exciting 'cause. A lot of things have' been I've talked. I was in Bostonat at a meetup for Cripto and M and they basically were like Um talking. I can't remember what the nameof the company was, but they were talking about like locking up literaldigital cash in ther bank, and that was what they were going to use tocollatralize a stable poin for you know from Incin, and I'm like that, justdoesn't fly, it doesn't work, so this is taking completely yo, centralized acompletely open market, a completely cryptoeconomic approach to it- and I,like that, that's exciting to me, but I need to see it play out before I havetoo much faith because there's so many variables involved with something likethis. So but that's just me, a lot of people are not like as cautious as I amso that's just where I'm kinda going on this and I'm really excited about thisproject. I think it's going to be amazing, especially when it works out,which it eventually will it well just fix it. If anything's not working rightright, so yeah M, that's sorry did did a Enon it down or up noute. I can't F.I can't Forkat tellbut yeah, I'm looking forward to to talk to you again,especially about some of the more m architectural size things also, a Tefor Hav me around this ecause t o. This is pretty also O, definitely lookdforward to the Nexto two and Gise. If you enjoy this foch your like button,they tune for part two which will schedule out and get O it. I itpossible Um, Yeh, sume,.

In-Stream Audio Search

NEW

Search across all episodes within this podcast

Episodes (108)