ZKPodcast: Digging into DAI with Rune Christensen from Maker

Transcript

[Music] welcome to zero knowledge a podcast where we explore the latest in blockchain technology and the decentralized web the show is hosted by me Anna and me Fredrik [Music] today we'll sit down with Runa Kristensen the founder of make er down to talk about stable coins what maker Dow is and how it works [Music] before we start we want to say thank you

to this week's sponsor no fund no I found is building an e rc 20 compatible open-source technical environment which enables the tokenization of real-world assets assets like equity in a company or real estate no fund does so by implementing basic protocols that bind an off chain asset with its own chain representation the company is currently looking for a VP of engineering to help them develop their cutting-edge technical environment check out the job offer at no - comm slash careers or review now funds github at github comm slash my friend so thank you again no fund for sponsoring this episode and here is our interview with rune of maker [Music]

today we're going to talk about stable coins and we have Runa from maker Dowell with us which is one of the most popular stable coins right now and yeah I think it no one could have missed the die and sort of all the buzz that's been going around with that so I'm really excited that we can have you on and talk about what this thing is yeah I'm really excited to be here today so looking forward to a great discussion let's start off as we often do with a quick intro to you and yeah who you are maybe what you were working on before maker and how you got into this yes so I got into I got into crypto through Bitcoin

back in 2011 so I was really like a very early you know Bitcoin believer in the early days and then what actually happened was that I basically lost a lot of money on Bitcoin and that's what drove me into looking at stable coins so from there at some time around 2014 I got into years which was arguably the first of one of the very first blockchain 2.0 projects and among other things they actually invented the stable coin through the basic concept of using collateral to bec like using volatile crypto assets - back then a stable asset on the blockchain and then for a variety of reasons

sheriffs never really gained critical mass like never really gotten major momentum I think primarily because they were trying to do a lot of things at once so they actually had so many early inventions that still only are being rolled out today on a serum like she has had all that in the very early days but unfortunately there was just so much that none of it really had like a single major breakthrough was was bit chairs just it was based on more Bitcoin like I guess the theorem didn't really exist them no so bit chairs was its own completely new architecture and in fact it still sort of still exists today and this is a quite big project but also kind of like that the technical legacy of bitshares has been

is today sort of represented by eus no so there's it's this whole branch of of kind of like blockchain philosophy based around you know delegated proof of stake which is more focused on trading off decentralization for performance and business was was the first machine that implemented this idea and and had all these very advanced concepts including the stable coin but yeah like I said it didn't really get the major traction that everyone was hoping for at the time and then me and I and some other people from the bitchiest community then saw aetherium s the great opportunity to you know build in an ecosystem where ever you know we could just focus on just one

thing and try to execute it as good as we possibly could which is just try to be the best possible stable coin and then the etherium ecosystem from there can build all the other stuff around the system right so that's really how the mega project started four years ago and since then we've just continued to try and actually build it up who are those original members who made there who's the original maker dough group well so I mean obviously it's primarily me in terms of actually founding the project and sort of getting excited but then you know I was also joined by nicolai Michigan who was actually a core developer on pitchers

after a while as well as James rady who is like a key community member in so we were we were sort of the initial three guys that really got the project off the ground and then immediately we were joined by a lot of other people both from the richer's community and also just like from the general assyrian community because our focus from the very start was just trying to build something you know so sort of keeping her heads down focusing on coding and designing and building real products what time exactly like what time frame did make her don't really start like is it right at the beginning of the etherium project when

it the Genesis block like what kind of at what point would you say is the beginning well I mean so I started I started really the the fundamental work in like over Christmas 2014 and then the project was announced you know in I believe it was the 19th of March and reddit in 2015 but the other co-founders and some of the team only really assembled in the summer of 2015 and up until that point they were a lot of like you know essentially pivots and like redesigns of the core design but then around summer of 2015 that's one that was when the project really

crystallized into essentially the the format yesterday right and sort of the overall design the overall philosophy around how the project should like what it should look like and how it should be governed and rolled out over time so as we're talking about stable coins I think it's obviously good to start with a definition of that and I think there's there's many ways to achieve a stable coin and we've seen some tethers that are like centralized you you put a dollar in someone's bank and they give you one token on ethereum or what have you so can you maybe walk over a little bit of like what are the different types of state stable coins and how are these housed stability and

achieved in these kinds of stable coins yes so they're generally accepted to exist three types of stable points however I'm like I mostly really focus on two of them but I'm also just like briefly get over the third type but that so the two main types are the centralized stable coins and a centralized stable coin is something like tether and then there's a decentralized stable coin and that's something like dye and really the way a centralized stable coin works is it's incredibly simple it's it's quite straightforward and some would say pragmatic right in the sense that what you have isn't centralized trusted custodian who holds fiat currency in a

bank account and then directly controls and issues a token that can then be redeemed for for money in this Bank account so that's really great because it's very simple to understand how it can remain stable and it's very easy to do arbitrage in this model of course the downside is that you effectively lose most of the advantages of the blockchain and you rely on a single counterparty and potential single point of failure as well as introduces a potential risk of too big to fail in terms of a single centralized stable coin absorbing a huge amount of assets under management and as a result we see I mean we sort of see a natural trend to where they said like

there's just so many of the centralized stable coins right so because they're centralized they sort of naturally decentralized through diversification so we really have I mean at this stage probably 50 or more of them that exist and are being being set up and most notably we even got JP Morgan creating a centralized stable coin so it's getting quite it's getting quite mainstream to certain extent some even call it the new ICO in terms of how its this a very standard thing to do nowadays for people who want to do something in the cryptocurrency space and and one can also that it's an extremely profitable thing to do because if you have a ton of dollars in your bank account while

someone else is you know having blockchain tokens you can collect interest on those dollars in your bank accounts yeah that's absolutely true although I think I mean there's also a lot of like it is it is more costly than you would think though so it is kind of a financial engineering exercise to really ensure a profitable set up in particular if you don't have really really massive volume but really the primary focus seems to be towards cryptocurrency trading and speculation and also very focused towards centralized exchanges and sort of doing the you know some of the most traditional activities that people have been doing with cryptocurrency ever since Bitcoin so you mentioned you

mentioned three different types but you've really only listed two what's the third the third type of stable coin is what's called well it has many different names right some call it algorithmic simple call it senior shares some call it like self-referential and essentially what this type of stable coin design is meant to do because it's really more of a design there there aren't really many of them that have ever existed in the wild is to basically use a very very advanced algorithm to then create stability out of nothing essentially and the problem is of course then that there's there's a potential risk that

the whole thing can collapse if suddenly there's a loss of faith in the algorithm and like the most the most famous ones of these projects was the project called basis which raised more I believe more than 100 million dollars in their pre IC o---- from various high-profile investors but then ultimately actually shut down citing regulatory risk and and I would say just general uncertainty around the model right and and basically this fundamental question of how like what happens and how do you prevent this thing from just crashing a fun fact is that they were actually the best performing ICO of 2018 because they then returned the funds to

the investors and investors only took a small cut of the money they initially put in so that made them the best crypto investment ouch lost the least yeah maybe we can go back to sort of the fundamentals of why would one want a stable coin yes so I mean well so like I was saying right this there is these centralized teams I mean just tether which obviously it's really used for just like large-scale crypto speculation on all sorts of markets that don't really have access to to feel banking which typically is because they're just not considered legit enough essentially to be able to get banking

relationships and that's where tether is a really useful substitute and of course there's also just this additional flexibility and and liquidity really that you can access when you can have much faster transactions between like Fiat Fiat dominated denominated transactions between two different centralized exchanges for instance I always saw stable coins or like I mean when tether when I first heard about tether the way I understood it being used the most was just if you were dealing with really volatile currency prices you could always kind of like quickly go back into something that's still crypto so you don't have to go through the banks and like do some sort of fiat exchange you can kind of go in and out of this you know staple US USD

in this case backed token and that meant that like transactions between them could be faster it like yeah obviously yes there's like this cross you sort of mention these like centralized exchanges working together but I really saw it just as like traders being able to go in and out of USD but not actually going to fiat that's at least the way I thought it like it was basically to play these this volatility really quickly yeah absolutely although I mean it's it's not really like if you just operate in one exchange it's not really faster to sell your crypto assets or stable to just sell it into let's say USD and coinbase for instance the

question is whether you're able to get an account on coinbase in the first place right because that's where a lot of the USD T for instance is used in in areas where there isn't necessarily that much coverage of like exchanges with established banking connections so it's it's it's a lot extent it's about like yeah bringing out the power of the of the USD to areas that don't have easy access to it right now let's move on a little bit to talk specifics about maker and what it is that your guys are doing so I think it's useful to start with some definitions we have a bunch of terms like die maker there's the maker token there's CDP's / CEOs and might get into it what the

difference there is how would you present these different tokens and sort of what exists and they make our ecosystem yes so definitely the most important is the dice table coin right and so the dice table coin is different from something like the centralized table coins we just discussed in that it is decentralized and that means that on one hand it doesn't have this direct Redemption ability so it doesn't have like a custodian that'll always promise to redeem it one-to-one for for money in a bank account or something like that but then on the flip side what it does have is the entire sort of fundamental power and advantage of blockchain technology to the full extent of you

know of how that technology works right in terms of being fully decentralized fully transparent no single part of failure no central counterparty and as a result also no too big to fail risk and sort of potential moral hazard and principal-agent problem all of these basically problems that to some extent resulted in Bitcoin being created in the first place right if you consider it a reaction to the financial crisis so to to to some extent what we're doing is we're really trying to just build on top of Bitcoin was initially trying to achieve and that's that's where we've ended up with die as kind of this this asset that

it's just fully decentralized peer-to-peer digital cash so in your model in the three sort of that you had listed it was centralized decentralized and algorithmic is maker decentralized and algorithmic or only decentralized because it's it is backed by something right it's not a pure algorithmic thing it's backed by ethers as I understand yeah and in fact both algorithmic stable currents and the kind of stable coin that the dye is are decentralized so I think it better when you talk of them in terms of a three types of stable going actually a better way to distinguish between them is to call what dye is a collateralized stable point and then and algorithmic stable content is also

sometimes called and uncollateralized stable corn because that's kind of the fundamental difference right so and and exactly you're right right so dye is right now backed by serum tokens as collateral and this is really just like the very fundamental logic of how it can remain stable right it is basically that for every die in circulation there is an excess of aetherium token value locked in transparent and decentralized smart contracts on the etherion blockchain and it's through that just like that basic logic of having an an over collateralized pool of assets that creates the just like I mean the value basis really and also just the trust

that ensures that I actually traits around one dollar in the market I know that we want to go more into like these particular examples but let's continue with our definitions so we have the die the stable coin that I think a lot of people are really familiar with there's also MKR the maker token what is the maker token so while die is this decentralized stable coin that actually is very simple to use right it's really just a stable coin it's a it's a cryptocurrency with price ability anyone can really access this Exodus and you know send it on the phone use it to buy stuff with MPR's kind of like this complete opposite of that

right so NPR is a token that's only meant for highly specialized individuals that want to actually participate in the governance of this entire system and fundamentally be a part of the process that regulates the stability of Dai and ensures that it actually can remain stable over time it's maker a tradable token still though yes so so MKR is a lot more like a traditional cryptocurrency right so it's a it's a it's an asset that's tradable exchanges and it has that variable price and the reason why people buy it is because it enables you but to again to participate in the governance of Thai and be a part of the process of regulates disability of diet and then as a result of that also obtain some value from this

positive like the positive regulation of the system right so you use have the potential to earn some money if things are going well but you also stand the risk to lose some money if things are not going well exactly so this is really the fundamental logic of the incentives that drive some care that you have right there's like an alignment of interests between MPR holders and die holders in the name garrulous want to make sure that I remains stable because then they make money and if they make decisions that then end up potentially destabilizing Thai they take the the hit so to speak directly through actual dilution of them care token so so that way there's just like it's just that the ability to vote

in the system actually is used to possibly interact with it alright so in my if things are going well the definition of well is the dot and die is stable yeah if did I remain stable they make money if it becomes unstable they lose money I think what would be really and I think I want a better understand this and maybe we can through the next definition but like how I'm curious how these two things actually interact so the next one the next definition in the maker definitions is the CDP the collateralized debt position I have a before we start on that I have a question is that a is that a term that actually comes from finance or is that a term that's used more in crypto

yes so that's the term we just made up okay why the reason why we call it a CDP right is because the CDP the collateralized debt position is the smart contract that holds the collateral right so right now that means it's it's my contracts where that holds aetherium tokens and then it essentially locks that collateral behind debt when ever dies generated and that is fundamentally how the dye supply is managed so the CD peas are the mechanism that ensures that for all the dyeing circulation there is also an excess of collateral in the system that backs that circulating dye and ultimately that is done through through the debt and so CDP's as sort of

an asset as and as an instrument you can use as a user its first of all meant for advanced users so it's not really something that regular users of dye and regular people have to to to interact with because it's it's a little bit I mean is it's actually quite an advanced financial tool but what it essentially allows you to do is to get leverage or get the quiddity out of your theorem without having to sell it right because you can deposit your theorem into the system and you can then borrow dye against it you do you borrow that dye against your theorem collateral without having to sell your theory on collateral right so for instance what most people do then is they buy even more theorem and what they've achieved then is a

decentralized margin long in the system and you can really compare this to actually taking out a mortgage with a bank right it's not like you pledge the deed of your house to the bank you then take a loan from the bank and after what you have to keep paying back the money you borrowed plus an interest rate and only when you pay down the entire value of the mortgage do you then again fully control your house right and fully own it and meanwhile if let's say there's a huge crash in the housing market or something that you've you fail to make your payment's the bank actually has a claim to come in and potentially seize your house and that dynamic exists with maker as well right so that's the fundamental

core dynamic that insurance that I can actually remain stable in a situation such as the theory of market crashing over time right because the system holds the theorem in these collateralized debt decisions as long as as there they are in these CDP's there is a strict requirement to maintain a certain collateralization ratio meaning that if the system detects that the value of collateral in a particular CDP so basically a particular person took a loan now hasn't maintained enough collateral to essentially back that loan then the system will take his collateral and sell it off on the market and through that be able to contract to supply what you just said there so like

how is it tracked do you actually have to keep the dye somewhere like how would you track if they still can pay it back well so that is just inherently all sort of endogenous view tracked in the CDP you right so a CDP isn't is as my contract that can for instance whole let's say $200 worth of aetherium and then have that locked behind $100 worth of debt right and then the sir right now this this minimum requirement which we call a liquidation ratio is one hundred and fifty percent so that means that if there's suddenly a huge crash in the price of a theorem and it falls by you know let's say 30 percent or something right suddenly now the price of e like you know the value of your theorem

collateral be the only one hundred and forty dollars backing one hundred debt in the CDP and that's where you've fallen below the one hundred and fifty percent liquidation ratio so that's when the system detects that now there's not enough collateral to fulfill this requirement it then immediately liquidates the position so on as this as this sort of market was crashing was just like CDP's consistently liquidating because it would have just constantly gone down well so it was actually incredibly interesting to see how how all of that played out might be really think of the big bear market of 2018 this dies trial-by-fire because it was launched almost

exactly at the top of the market and then it just rode through it all the way down and and it was first of all it was surprisingly resilient in terms of the pig holding the entire way and also it actually kept growing during this time which is also quite interesting right so people kept actually hanging out more and more CDP's even as a market kept falling and also surprisingly people who are incredibly good at maintaining the CDP's so we actually saw a lot less liquidation so that we expected in terms of people actually hitting their minimum requirement rather people are just incredibly disciplined in adding more collateral as they saw their position getting close to to being very risky right but there were these events of

huge liquidation so which typically happened after if there's been like a consistent so the steady downturn and then it was followed by a massive sudden crash that would very often cause just like a huge string of liquidations right so to better understand the way that the maker CDP model works like I'm definitely there's there's an article or two that we've been using and prep for this that we're gonna add in the show notes because I think also seen like like kind of graphics of like what happens and how the theorem is locked and then converted it to dye and like at what point does that need to be paid back I guess would be really it's really nice to see graphically so I'll add that

in the show notes so speaking of collateral analyzing a house as an example I think that's an interesting one because we're sort of in this eath space you know if you imagine a house you can't buy groceries with the house like I can't sell you a portion of my house to buy bread so obviously I go to the bank and trade my house for some dollar so I can use those dollars but that's not really how it works in reality like you do this upfront but in the maker down model it is similar that like if you're a large eath holder you might not be able to do what you want to do with those eath or you might want to

do margin trading or other things so you go and you trade those in you're willing to pay that fee that that it costs to take this loan in order to get something that is more liquid or that is stable or that you can do other things with but if you're not real if you're not already a large eath holder like how how does it make sense like what's the use case for me to say trade my dollars into ether into die - then have blockchain dollars because I don't you know I there is a fee so I'm trading my dollars for slightly less dollars and I would obviously never like do a clutter ice that's in a bank this way you sort of why would a regular

person obtain died yeah I mean so and that's a that's a great question right because well what we're seeing with the centralized stable course is that they focus a lot on kind of like the traditional speculation use cases of cryptocurrency die is really has proliferated much more towards kind of exploring or this new frontier of just how you can actually use blockchain technology to build products and services that go beyond speculation right so I mean one example is that die essentially powers the defy ecosystem and movement which is this whole quite recent new space that essentially has developed and a theory right where what

people do is they build these advanced second layer financial services on essentially on top of the dice table coin so there's projects like compound unit swap and and a whole range of others that support die SS their main stable horn which is of course because these are all decentralized finance products so they need to use a decentralized stable coin if they want to represent stable value otherwise they essentially lose the benefit of their decentralization so that's kind of like one type of interesting you know like avenues for for use cases and demand for this system right that that dies is how

you enter this like new ecosystem of of decentralized applications and then another area we're focusing a lot on is Financial Inclusion and adoption in basically where we you know like underserved communities right so like places in the world where financial infrastructure isn't very well developed or they're just like our different types of problems from the monetary system and maybe one of the best examples is Argentina which is a country that just suffers from like quite severe inflation and where people really already use a lot of black market dollars essentially right so dollar bills that are traded against the local currency as an actual store value so

here we're seeing like a lot of interest in adopting dye as this just like digital version of the dollars that they can just hold on their phone that's like easy to access there's no restriction on getting access to it and it means you don't have to hide a bunch of bills under your pillow for instance if you if you're like you say the speculation use-case doesn't really make sense it doesn't make sense to trade your dollars for slightly less dollars on a blockchain if you're just gonna hold those but if you're it does make sense to trade your dollars for slightly less dollars if the utility of those increases which they do if you want to use to centralize finance applications on aetherium or if you want to have

faster like you know transfers or if you want to do cross-border transfers in a way that you can't really do with regular dollars so essentially you're paying this fee to increase the utility of your dollars is that a fair way to say yeah but just to be clear so you don't have regular use of tie you don't pay the fee or kind like go through the process of doing the collateralization right you'd really just buy dye from the market but someone actually right so it has to be in there somewhere yeah but that's I mean but and that fee is essentially the you know I mean that's the that's essentially the cost of credit right or the cost of capital so assuming that this system is properly

balanced which is like definitely a challenge in the early days right sweet and this is actually one of the things to experience it right now that there are pretty significant swings in the interest rates of the system to basically deal with this question of how to balance out dye demand with supply but in particular once we release this feature known as the dye savings rate which is coming with the full version of the system that's being released sometime this year we will really get a much more liquid and much more stable market that will ultimately give users effectively the ability to just trade want to want into dye and very very large volume and then even there will be this this you know if

the reason to hold dye which will be that there is actually going to be a savings rate on it right so you will actually to earn savings over time from holding diet which kind of goes back to the earlier point about how centralized stable coins monetize by taking the interest rate for themselves where dye is basically so efficient because it is fully decentralized that it actually has a surplus that is able to be given directly to the holders of the stable coin so this the this actually one of the points I wanted to to go back to was this connection between dye and the maker token and I feel like it's somewhere in here as well like in the CDP like is there a fee or anything like how to make how do the maker token holders do they earn anything from the

creation of dye that's the sort of fundamental connection right that like I'm careless earn based on the size of the dye market effectively and and if this specific cash flow comes from the interest rate that is paid on borrowing money from the system right so what's called the stability fee on collateralized debt physicians and so right now in the current version of the system which is still just a beta it's called single collateral dive so it only supports a theory Mescalero currently the way you pay the stability fee is that you just pay it directly with NPR and you have to burn this NPR when you pay the fee so for instance if you have a fee of $50 because the interest rate

was was you know whatever right like you the interest rate was 5% then you had $100 and we had $1,000 and you know you waited a year and I have to pay $50 worth of interest rate right then what that means is you have to burn $50 of MKR when you want to pay down the debt like paint on the principal debt that you initially took out and then also mentor which you were your collateral so the maker token holders don't actually receive the fee they just there's less maker tokens out there and so their tokens are more valuable yeah exactly right so it's it's just a different mechanism for a cash flow the reason why we chose that mechanism is because it's just more technically efficient but from an

economic perspective it has the same effect right in terms of ultimately transferring the value to the mpi token so a CD like an individual who wants to create a CDP also needs to be and make mkr holder that's right or at least they need to acquire a cure the moment they want to close this evening and is that like in this like the CDP itself does that live within a wallet you could say it lives with it I know it's smart contracts but like is it related to one wallet I'm just curious if like does that wallet need to have MKR ether and dye in it for this to work yes that means the CPS are owned by aetherium accounts right so the way you specifically pay this ability is that

you have MKR sitting in your aetherium account and the moment you want to pay down your debt for the CDP the you know the CDP has to be able to access your NPR and and through the whole you know like the way that token transactions work in aetherium and then use that to also pay down the necessary stability fee is that stability fee then paid at the end yes so that stability fee is paid at the very end when you when you close down the CDP right and so one of the things one of the main criticisms of this model was that it was very inconvenient that people had to actually go and buy a pair on the market right so kind of like an extra transaction out of nowhere essentially right for no reason really so reasonably what has been added to too many of the front ends because

there's actually many different dashboards right that give access to CDP's way and that really is this your case of how maker is is this more of a platform right so there are many different dashboards that that give people access to managing the CDP's and some of them have implemented a feature where they will take care of actually buying them PR and paying down the list ability fee behind the scenes essentially so the user only has to pay the you know the the deaths die and don't have to worry about going to an exchange and acquiring other cryptocurrency this model actually changes even more once we launch the full version of the system which is called multi collateral die right where first of all we'll add support for multiple collateral types well but we'll also add the die savings

rate so the this this savings rate that's applied to die holders and in in multilateral diet and 'mentally handles all of the fees and all of the the rates and died only so you don't have to pay like you cannot pay your stability freedom care in sort of internally in the system's logic you have to pay it and die and the system then actually takes that die and does open market transactions to then burn the API token the way like when you pay your your debt you actually pay two types of debt right you pay this die savings rate which goes to the die holders and then you pay the stability fee which is essentially the risk premium that's specific to your

collateral type which then goes to them care holders by burning them care yeah that's right okay I want to get more into multi Clara collateral that's a hard word to pronounce I want to get more into that later but I just wanted to touch on as well like what is the motivation for someone to actually close out their CDP because like I have some died but I don't have an open CDP so someone out there is missing their die to be able to close it out and likewise if I'm if I want to buy something in a store with died then I can open like i trade my eath through a CD PE for die

and then I send off my die to buy my thing and now you know I've extracted the value that I wanted why would I ever care about closing out this EDP well so I mean just like in the very simple sort of case it's it's all about getting access to your collateral right so it let's say you you're your heaven aetherium back CDP and the reason why you do that is because you want you need to spend some money on something but you wanted to keep your exposure to ease then later if the price of ease has gone to the point where you think okay now it's not really it's not gonna go much higher that's the moment where you would want to then go in and buy back your dye pay down your debt get get your ass out and then maybe sell the

easter dice so you can just have the the stable exposure or something like that right there's actually also like a deeper aspect to this question right which is in general how does the system ensure that there's a balance between how many people want to hold dye and how many people want to hold CDP's and when there are new people coming in to buy more diet how do you match that up with when people want to open more CDP's and sell more time and all of that is determined by the dye savings rate so essentially the interest rates of the system right so it's really a balancing asks act to try to get the aggregate demand for people who want to use CDP's and want to borrow money from the system to match up with the aggregate demand for people who want to hold dye and who

want to earn the savings rate over time and the the basic logic is that if there's not enough people who want to hold dye and there's too many people who want to borrow diet that means is there the savings rate is too low right the interest rates and the system is too low and you have to increase it and basically the same goes for the other way around right you have to decrease interest rate if there's too much demand to hold die and not enough demand to borrow dye essentially I guess you would you you just buy die off in exchange if if all you want to do is use die for something but if you actually want to use this EDP in its proper you know full value of having liquid dollars while keeping exposure to ether because you're

intending to do something with that in the future either through margin trading or just like through exposure and and then trading it back later then you'd go through this EDP but otherwise you really wouldn't yeah definitely and it's really two different it's two very effectively opposite types of use cases right yes die holders are sort of regular users like regular people who don't really need to know much about finance or advanced blotching stuff who just need to use money on the blockchain and maybe in many cases they actually use that without even knowing it because it's being used sort of under the hood in some sort of FinTech solution so that's basically regular uses and the majority of uses just use die and don't worry about the

other stuff and then CDP uses of these quite advanced users who actually rather than access money and access stability they want to access effects to be the opposite right they want they specifically want volatility and leverage because they not only want to hold the the speculative or volatile collateral asset but they also want to borrow dye against it and actually go and use that dye to buy something like some other volatility assets for instance yeah would there be any point of like taking your eighth putting the CDP getting some dye tradie for 'if putting in the CDP yeah he's like is there any reason to go through that chain multiple times or no yes if people

do that all the time so and that is really just I mean of course there's a there's a pretty natural limit to how you can do this and in general it's just done in like when you really want to do it you do to be to do it in two steps no sorry three steps in total right so it could be something like you start with let's say $100 worth of aetherium right you can then use that to generate let's say $50 worth of dye right let's it is 66 right it's always a isn't it fifty percent or does that change that that over collateralization well so I mean you can you don't you enough like so 66 like that you have the ability yes to like to generate 66 dye with a

hundred dollars worth of collateral but you would never have a user actually generate that sort of directly because that would put them at the absolute limit right I see okay so you usually would take out much less than that yeah I mean in the practice what we see is people typically collateralized by three times so that's kind of that's the average collateralization it was Justin right so that would be someone using a hundred dollars worth of collateral to generate 30 died pretty much right but also what what the more sort of aggressive magic traders tend to do is to collateralize to X so they end up having a you know Yeah right like I mean they would for instance they would like maybe they would like to end up in where they have $200 worth of collateral and then 100 debt in the CP right and

you can do that in two steps starting from just having a hundred dollars worth of etherium right if you basically generate 50 die at a time two times in a row you can actually go through this interaction without ever exceeding or like without ever really approaching the risky risky state where you you could potentially get liquidated it's kind of interesting to see margin ratio so I think crack and coin based they support like 2 to 5x ratios or something maybe they increase it recently I don't know but typically like in the traditional FX trading like the currency trading world they're kind of famous for having fifty to a hundred x margins and it's just

these insane margins but it's kind of because currencies are typically very stable and like they don't actually have that much volatility so you need like 100x margins actually get some some exposure to it but it's interesting I mean well I guess we'll see if theorem ever becomes more stable like if the margin ratios of people go up we had sort of touched on this earlier that the MKR token holders can participate in governance of the system but what I'm curious about here is like what exactly can they affect because it sounds like the individuals like like do you say holding or running a CDP is they holding a CP holding a CVP so the individuals

that are holding a CVP they can choose like how much they take out or what have you what are the maker the MKR token holders deciding so the M character holders actually decide on exactly this issue that Frederic was just talking about right so what is like the sort of the correlation and the connection between the fundamental risk of the collateral asset and then the type that the kind of terms you get to take out a loan with that asset on right so for instance if you want to use a theorem is collateral obviously that is a quite risky asset right so that's why it has this significant or collateralization requirement for instance whereas on the other hand if you want

use another stable coin as collateral suddenly there's a lot more room for leverage right you would actually get to these Forex levels of potentially 100x if you're using another currency as collateral right simply because there's so little fluctuation that you can you know you still have that wiggle room essentially in terms of how how quickly the price can move and how how safe like how safely the system would be a literally create a position and ensure that it remains over collateralized as a whole so that that's really like it's really this fundamental question that lamp care holders they they deal with and so that also means it's very much related to the future where there will

actually be multiple collateral types so in the current version where there's only theory was collateral there hasn't really been much like too much kind of like experimentation with with trying to change too many of the variables of the system other than just continuously adjusting the stability fee so the interest rate of the system to try to balance the dye pick but in the but basically the entire governance framework is all based around building off traditional risk models of assets such as bonds and stocks and commodities and so on that are essentially how banks they do margin lending on these assets in the real world building of those models we're able to construct an open-source framework that allows us to actually scientifically determine how we

should set all of these various risk parameters as we call them right the various terms of the different like of the different collateral types and different ways of using these collateral types in the system that ultimately give the system the best possible conditions for long-term stability of died right and that really comes back to the whole fundamental incentive of mpr right which is that M care holders want to make sure that die remains stable which really means that it mean remains over collateralized and it remains without too much concentrated risk so it doesn't it isn't at the at the mercy of a potential Black Swan event just crashing all the collateral at once and suddenly leaving a huge so empty hole in the in the inner solvency of the system because

M care holders then have to actually cover that directly right they actually have to to recapitalize the system and essentially bail it out through direct and automatic MPR dilution that the system just it detects there's a shortfall and immediately it starts diluting him here so that means it creates new and care yeah that's the value of their em care goes down yeah you can think of it as as if the bailout for the financial crisis has been financed by you know the big banks and the regulators and kind of like the people who had been responsible for actually doing it right like the people who have made the wrong decisions if they then had some sort of like financial stake in the system as a whole that was then diluted and essentially

taken from them if they if they feel properly regulating the system so what you've just described sounds very much like I mean you you basically just said this you have one token right now you only are like the the the main maker governance process right now it would only be focused on the stability fee you have maker holders that can vote or I mean I guess how does that actually work as it just sort of like here's the stability fee that's proposed by this system we think this is the right way and then they vote or is it all kind of automatic so how does that actually happen now basically the approach of the governance process is to focus on first getting what we call yeah well I mean

getting scientific consensus rights are getting so going through this scientific governance process where the goal is to initially get risk experts and kind of people who actually are able to to articulate and reason the decision systems correctly to then get a consensus around what is what does the data show us right now that the best course of action is and this is still like a very like it's it's still incredibly difficult to do that in a you know a very precise manner so is quite a lot of kind of like just trying to see what is the best what is the best guess and how the system should be controlled from here from a perspective of stabilizing ties so this is when you

know this is a fundamental question of if the dye price is too low we have to increase interest rate if it's too high we have to interest rates right so so that is a you know so first there's this what we called the governance meetings and generally just like the community process of people discussing the situation and and and sort of aligning on the data on that on the facts and then from that ultimately there there is like like I mean there there's there's the consensus formed around what do we believe like what does the community believe to be the best course of action and it based on the consensus F vote is then proposed and that vote then as long as the consensus process and so the scientific governance process was

followed correctly in the first place that vote then always passes right and the vote is really about it's almost like the elders are acting as an Oracle to the process of whether the governance actually hadn't correctly or not but this is actually proposed by people like it is it's actually it sounds like like some is proposed and then like one would need to actually stake or do something with their maker tokens like it's nothing automatic it's not like if you don't answer it just like automatically goes somewhere it's like you actually have to act yeah that's correct so and there's even it's even like a double what we think of as a to vote model so initially there's what we call a poll where it's basically a yes-or-no vote

where you get to like signal with your own care of whether you believe that this is the correct decision and whether you believe the governance process probably followed sometimes is actually a multiple choice vote this is a and we've actually in fact we've done that for the very first time this week and and so that that's a way for kind of the in care holders to give their final input on whatever the question at hand is after has already gone through this like a scientific debate process and once a decision has been made through the poll then you get this the the the next face where MPI hold is actually a like effectively stake their MP are

directly on you could sort of think of it as a smart contracts that implements the new state right so there's a smart contract created that's put on the blockchain that actually contains data that if it is voted in as what we call the exit proposal right so it's kind of if it obtains the most votes in the system through people essentially staking their mkr on that particular proposal then it gates administrative access to the call maker system and it's able actually directly changed the state of that system and then the data of the proposal then executes and just like chain-like directly changes the state of the system right and that's and then the stability fee for instance is then immediately

updated from that moment on so this process as a whole then means that that you have this like sort of this multiple this multiple phases that sat very broad right and sort of narrowed down until the moment where it's just okay we are going to actually execute there's logic into the system and change it based on what we decided I know that like this past week you this was this question of the pet the pig was lost was it the first time the pig was lost well so I mean so die has always been us well that is a soft pick right and this is fundamentally because it's not backed one-to-one by a custodian yes it's not a centralized stable right so there's always been like a spread right there's always been a difference between you

know like how much die you can buy and what price you would you buy it and so on especially in the early days because it's just low liquidity in its very early right but what happened was that for a longer period of time the stability fee so the interest rates in the system were just not set very well essentially a rather they were set well initially and then they were not changed quickly enough to accommodate the change in in the eath market specifically it's because they were you know like we sort of bottomed out and suddenly there is this much more positive sentiment of people thinking that the price will go up so that resulted in even more CDP's getting opened ultimately that just resulted in like I mean the the rate was simply too low for the amount of demand it was for opening CDP's and then there

was a process to continuously increase the stability fee but ever increase was effectively too small and had very little impact and that's why now like increasingly what the governance processes then been doing is it sort of been more and more aggressive and how these increases have been made in order to try to get the figure out well the balance in point is that'll actually start getting a real reaction out of the market in terms of returning diet back to this more like firm position where the spread really you know where you can you can sell it like slightly below one dollar and you can buy it slightly above one dollar speaking of the dollar price something that I'm very curious about is Oracle's and how the dollar price actually makes

its way into the system like who's inputting it how and like how do these Oracle's actually work yes so many people think of Oracle's sort of call Oracle's really the weak spot of every decentralized application right Oracle's are definitely this you know very very important part and also a bit more you know like the part this will connects a decentralized system to the real world right and then the definition of our Oracle is like some sort of external system that then delivers trusted data on to the blockchain right and and of course because maker needs to be able to detect the price of a that needs to be able to perform liquidations in the

system it relies very heavily on Oracle's right and and and actually what we've done in the current beta right where we just need to get this you know sure that the system works and get it running was to essentially build an Oracle that we just knew would be really robust so we used a model like a very simple model where it's essentially like like right now it's 14 trusted but actually anonymous actress that very few people actually know the identities of these people in order to prevent like protect any sort of attacks being being performed against them and then they were they were included into the system at the launch of a system

right but actually they are under the control of him care holders so it has never actually happened but them here holders have the ability to update this set of Oracle's over time and then what happens is these 14 Oracle's they all run actually like all they all run different types of software so this they get focused on having diversification of the infrastructure right so they don't all run the same Oracle system rather is like different types of software that ensures there's like different methods for actually obtaining the price in the market and then pushing it onto the blockchain and into the system and then on the blockchain itself is what's called media Nizer contract so basically a contract that then takes these

fourteen different input sources and then just calculates the median of the input they're giving and then ultimately pushes that median into the the core die system itself so that way we ensure that there is there is some level of decentralization to some level of resilience against compromise or failure of some of these Oracle's but you never have you ever thought of using like like in the interview we did with Jenna from melon port they're like working with khyber like there's other sort of price discovery Oracle's out there do you ever either like test yourself against these or have you thought of using those as well yes so I mean one of the main focus points for the for the upgrade of the system to the full version with multi

terminal die is to significantly upgrade to Oracle's right because it's really all about getting Oracle's to state where they are completely resilient against attacks right and actually what we found is that the main like the main important feature that needs to be be added under system it needs to really be a part of really every Oracle and every decentralized application is not even like making the Oracle itself more robust because it is never like no matter how good you make an Oracle you're never going to be able to completely guarantee that it won't fail and it won't be compromised and potentially be used to launch an attack against the system where the data is manipulated right so instead what you need to do is you need to have a

secondary really like a fallback mechanism or like an it like really a like a defensive emergency mechanism that ensures that even in the worst case scenario where the Oracle starts inputting let's say a bunch of zeros or like a very very high number or just some random data you need to be able to guarantee that you will always be able to protect the instead and and recover from it and so that is really the first big upgrade we're adding to the system and the way we achieve that is through a combination of two mechanisms the first one is the Oracle security module and the Oracle security module is we actually have a whole bunch of these security modules but the Oracle security module is like what they do is that they just

impose a delay on the data so the Oracles input the data through the regular mechanism of computing the median and then that median then actually sits in the Oracle security module for an hour and so it just sits there and waits and then after an hour it is pushed further into the core of the system and during this hour-long security delay the community then has the opportunity to actually protect the system from an attack by doing what's called an emergency shutdown which is actually a mechanism to completely shut down an entire system and like settle all positions and gracefully unwind the system essentially and like immediately through with like full guarantee completely unwind any some risk in a

system and it's like any sort of operational risk in the system and then of course from there you have to be able to also quickly recover and and get it all back up again and basically migrate to a new deployment so there is this whole process around how the system can protect itself from various types of attacks do it like a temporary shut like shutting everything down pausing everything making sure nothing blows up and nothing goes wrong and then redeploy itself have every one microwave to the new system and then receive operations and actually this can be done quite smoothly once the infrastructure is in place this actually leads me to the next question which is about security and how you even designed a lot of this stuff so like are you using statistical modeling

in any way to figure out what could like like how are you basically designing a lot of this risk profile what tools are you using well so I would say there's there's there's mainly three types of risks to consider right so first of all there is really like the fundamental like economic and financial risk of the system right so this is where we use traditional risk analysis and and like frameworks like Value at Risk and this kind of traditional financial tools that enable us to to as a community get consensus around how like how much interest should we charge on collateralized loan with you know US Treasury bonds or something like that

right so that's kind of one kind of risk and it's really handle in the in a traditional way that you handle this time at risk in the real world but of course upgraded with the power of open source and and community governance and then a second class of risk is crypto economic risk so this is really the time of risks you're exposed to by being a decentralized system on the blockchain that is controlled by you know the community by voting tokens and also by by like this these trusted data sources such as the Oracles right so so this is a class of a text that that include things like like someone buying up all of them PR and doing some malicious vote with it or someone taking control the Oracles and trying to inject malicious data into the system and the current

economic attack vectors are all covered by the emergency shutdown I just talked about right so like basically what you do is you have delays on any table of sort of you know proactive change the system and then you almost have a much easier way to be to do a reactive defensive shutdown of the system and then a redeployment that can mitigate whatever it's like whatever cost attack in the first place right so if you have some bad Oracle's you do redeployment where they're not included right if you have someone buying up a significant chunk with NPR and trying to vote something terrible in you do a redeployment where his and Kara stake is not included right yeah so the final type of risk is just fundamental technical risk right so this is the Dow style problems right basically the risk

that there is some sort of vulnerability in the code itself and that would enable someone to just like go and steal all the collateral something like that right and to a large extent this is the most serious type of risk at least in the early days right especially in the context of you know like I mean I actually saw the Dow happen live right right before my eyes and how like the assets were being drained out of the the smart contract right so that is really more than anything else the main thing that we've been the main problem we've been trying to solve with maker and I mean the only real solution is just really good code right it's you like not make terrible coat with exploits we had Martin Lynn fallin just who is doing formal verification with

you guys so I know that you're using at least that as a technique I'm curious to hear more yeah about what how do you see smart contract development and the security practices around that because it is an extremely difficult area I mean I'm sure we could talk like an hour on this but just like your general thoughts and like thought process and how you design a building so our contracts to ensure that quality yes no really I mean well so I mean so so fundamentally it is all about writing good code right and following it best practices and not making something that's bet but yeah there's like approaches to do that right and I think the most important is just simplicity right so like to have that try to really have the simplest possible

business logic the simplest possible requirements from a design perspective and avoid putting anything into the core of the system that you can build in a second layer a third layer application right like from that just from that those design principles alone you already like that I mean law like I guess those are the very basics that would ever allow you to actually create something that's safe for the long run because you can't like fundamentally before you can do anything else you have to just have a simple starting point a simple to sign right and then in terms of how you actually implement this design formal verification and it like basically the ability to to sort of properly reason around what your your code actually does is then the the

critical aspect here right I mean but but it's also really important to not sort of over like overemphasize it's kind of like overly rely on the concept of formal verification by itself right formal verification is really more of a tool that helps you achieve what you really want to achieve which is good code it's not like you can sort of do like a magic ritual and then do some formal verification and now your code is safe right like you can still easily have a system that can get hacked even though you think some of it is formally verified if it's complicated and and you know that you know it just wasn't correctly made right but of course the the the practice of of trying to the best expect you the best possible extent

of doing formal verification is a process by self greatly helps with the goal of achieving great code and achieving just like you know a system that ultimately is simple enough and and built robustly enough that there are no major vulnerabilities in it I'm curious to hear what's in the future for maker Dowell I mean we've talked a little bit about multi collateral die I'm curious in particular like what the types of collateral you see there are like we're now starting to see like wrapped Bitcoin showing up on the theorem so you know it possibly gets other tokens in there as

well but I'm also curious to hear your thoughts on like eath - and that whole development and interoperability and like how that can bring a lot of you know actual diversification - to this multi collateral thing yeah so so so in terms of makers own roadmaps they're really three major milestones that that we're working towards right so first of all there is launching multi collateral die in sort of its its bare-bones format which in to some extent really is just launching dire as it was in its full its full version really so and and moving past this current beta stage into a you know a future-proof

full version of the system and that will initially mean supporting various kinds of like out of the gate various ill support various types of ear she's 20-20-20 basically the kind of stuff that's currently available on aetherium but which to some extent will be quite limited you know compared to theorem which is still obviously going to be the most significant type of collateral in the system just based off like the the market for these various tokens right but then what's coming next after multi Glenelg die has been launched and we sort of start focusing on on the next steps of the project is to actually focus on getting access to real world assets as collateral in the system right so and this is achieved through security

tokens and other types of tokens that actually able to implement a legal framework and ultimately represent a legal claim on a real asset and one of the the major efforts we're doing here is for instance with that we are have completely pivoted the racist XD centralized exchange that is the central exchange we're actually the first decentralized exchange and a theorem that we built initially back in 2016 the focus will be to then yeah like get access to regulated security tokens but in a way where on one hand they are fully compliant with the necess you know the relevant loss that makes them actually have you know force of law in terms of their legal claim to an underlying asset but at the same time they also from a technical perspective

exist in a format that allows maker and the maker covenants to interact with them and actually accept them as collateral into the CDP's in the system and that's going to really I mean that's when to some extent like this story of maker really begins right because once you get access to real-world assets there's just no longer a limit to scalability right you there's just this trillions of dollars of assets out there that actually can be very efficiently deployed as collateral in the maker system you know if adoption of theyself is also properly driven forward right and adopted by you know like both like various business solutions and kind of like company and b2b infrastructure but also things like the unbanked in the

development world who need access to to financial services and then of course the whole decentralized finance and and all of these like new this this blue ocean of new technology that's being built on the blockchain and powered by dialogic sent what about that e 2.0 question though what's most crucial for maker is to achieve some financial scalability right for the perspective of dai can grow any amount of time circulation and still remain stable because through access to real wall assets we can create proper diversification of the collateral base but then we need to get to the the next big question which is how to achieve technical scalability like how to actually support many transactions per second right and how to be how to

actually reach out and interoperate with all the various kinds of financial systems accounting systems payment solutions that exist the real world and this is where we focus heavily on blockchain interoperability solutions and in general like Christine you know solutions communicate like trust chain messaging and communication and various forms of two-way pics and like a really interesting example of something that's already running now and it's very popular is X tie sidechain to theory right where we have basically a very centralized but also very performant blockchain that then can can process dive transactions extremely cheaply for

like you know particular transactions where what you want is you know you don't really want to do like large-scale highly secure transactions you just want to do a lot of small payments very efficiently for instance one thing is deploying some specific side chains another Avenue is connecting up with and sort of interoperating with and being available on other sort of full block chains right so something like us right khyber is actually working on a bridge here there's also solutions like cosmos and polka dot that are coming out that focus a lot on blockchain interoperability and ultimately our vision is that eventually die will be available on every single block right let's out really our general approach to adoption and to to making our system

available right is that we don't want people to be restricted by their own choices in what kind of technology they want to work with right so it doesn't matter what kind of blockchain you want to use what kind of wallet what kind of whatever right what kind of signal solution you want to use we'll make sure that dyes available no matter what you want to do right and in eventually like it'll it'll be available because it's taking you possible to do so so there's not really a reason not to do it right it particularly if the demand is there those are not small ambitions oh yeah absolutely and this is something that'll take many many years to realize right and ultimately it will be a great community effort so so right now it's mostly driven by the foundation but increasingly the ecosystem is also just

getting its own momentum right and that's when the power of a decentralized community really starts to show right when these things starts just forming on their own and they don't actually have to be built by the foundation so much more that we could talk about this is very hard to wrap up this interview because I have more questions and I want to ask you we we know a lot of people in the community or we know we we also have had come into contact with other maker people I actually feel like this may be the beginning of a series a little bit because I think that there definitely is more to explore and I just want to say thank you again for coming on yeah thank you very much of course and giving us this awesome

overview and kind of introduction to this obviously much larger and expansive potentially expansive topic yeah it's it's quite something cool yeah thanks a lot yeah thanks so much for being here and to our listeners thanks for listening thanks for listening [Music]