Stablecoins, built for DeFi: introducing the stablecoin pentemma - TokenBrice | Liquity

[Applause] oops thanks for the kindly for winter hi everybody and welcome today we're going to be talking about stable coins but not just stable coins I want to tell you about risk analysis for stable coins and I think there is a lot of uh a lot that can be improved here we're very at the Stone Age of how we model risk for stablecoin to the point that I even disagree with the most common presented model for those risks which is a well-known trilemma you might have seen we'll talk about 3dma between scalability ability to maintain the fragrance workflow that's stupid and and the decentralization of the stablecoin so that's how we're gonna go first I want to take 10 minutes maybe around that to walk you through my framework of analyzing stable coins risk and then we look at example because that's the most interesting way to actually like make sure you got the framework and I believe yeah I believe you will see also why this framework is relevant with those examples because this framework enables you to essentially establish that some models were nonsensical from the get start including UST so you could from the white paper of UST assessing it with this framework conclude that the only way out for lust is to explode but first let's go a bit on the topic so why are stablecoin such a big deal I'm sorry pretty simple you got the numbers of screen D5 tvl is like what 50 billion stable continually is two three times that end of the story so stable coin is like 3d5 at least of course there is stablecoin in D5 so you cannot exactly compile numbers as such but essentially it's a big deal I think we all agree on that so no that's the usual Freedom as it is presented so you know back decentralization scalability pick two because you cannot achieve three that's usually how people goes and so I try to map some of the Stables on that trilemma to help you understand it a bit so usually what you see is there is very few stable coins trying to maximize for decentralization most go for the Peg scalability Side of the triangle and this is where you have all the centralized the taxes TUSD usdc USD TB USD gusd but also the pseudo decentralized as I call them so the dye the frax the meme everything harnessing a decentralized infrastructure but using a centralized collateral essentially uh so this is and I'm sorry I'm not an artist but this is my actual revised to them uh we're gonna go uh in into so you see there is a bit of subdivision here because I need five categories to properly as a stable coin and the main split is essentially that ability to keep Peg and decentralization are both split into two subcategories essentially so you're left with technical exposure what are the risks your stablecoin is taking and that include any protocol it's interacting with then you have the collateral nature what are the properties of the collateral you're using and how does it impact the system you also have the collateral counterparty which is essentially does your collateral require a third party to behave as it should be behaving to function properly and the two bottom part are more like the scalability dimension I guess and that's the liquidity model so how is the liquidity maintained on your stable and how sustainable are those strategies and on the other side the governance model governance model that can pose different risks to the stable coins as we will see so I'm just splitting that a little bit more here so you get the full context uh but yeah in terms of collateral native there are a few items to consider we'll go through them all in in the specific slide don't worry uh so this is more like of the plan and for the sake of time I will walk you through it already with like live example because it's easier to understand so collateral nature is a question of what threat to my system is that collateral posing and for me you analyze this with four subcategories the first is a question of being endogenous versus exogenous and it's it's big words but essentially it means is your collateral correlated to your system or not and of course you want it not correlated to your system ideally so you know this is why for instance ethereum is an exogenous collateral for for any protocol because the SDM network is your protocol if you made doesn't have a stable coin so necessarily if you put a stable coins outside the system but now if you take something like lunar UST Luna is an endogenous collateral to the lunar USD system it's part of the system so it's a reflexive system which means Max risk reflexivity is a bit of a yeah the Bain World in Risk Management then you have other characteristics that are important so volatility of the collateral the liquidity of the collateral and the additional risk this little D is where billions are lost essentially because people kind of forget about it so you know it's like you go usdc good collateral right it's necessarily uh exogenous uh it has very low volatility slash known it's a stable coin it has immense amount of liquidity it's one of the most liquid token on the iso and blockchain but it can be censored and can be frozen so it pose a massive amount of risk and you know this could be this need to be assessed as uh when you look at usdc escalator then I give you Luna but it's pretty much you know as we say in French shooting on the ambulance at this point because you do see that Luna is pretty much maxing every risk factor possible for for for for this situation so it's purely endogenous it has high volatility it has low wish liquidity and it has maximal amount of risk so just by that you can conclude that anyone considering Luna for a stablecoin is insane um another Kunta party risk is uh I wanted to take a side for this because it's often misunderstood so here I want to explicit that risk because it depends on the situations of how your assets are stored essentially so here we look at in the first two it's easy and Bitcoin in both case all right still is in Bitcoin it's just the first case you are self storing those Bitcoin with your own wallet essentially you're securing your own access to your wallet versus on the second case you are putting them on let's say uh coinbase and of course as soon as you deposit them to coinbase you have the coinbase counterparty risk so this is kind of what I wanted to explicit the same asset you can have a different counterparty risk depending of its custody model then the usdc example is just here to explicit to that whatever the case study if an asset is having counterpart series because of how it's structured you cannot mitigate it so to put it simpler there is nothing you can do with usdc that would allow you to extract yourself of the risk pose by Circle it's emitter circle is the counter battery of usdc end of the story and yeah I guess I've kind of forgot a dimension but just in case East and Bitcoin are arguability only Assets in the world without counterparty which make them pretty interesting collateral so governance wise and then uh yeah we're going to the example governance you kind of have to understand how can the protocol evolve over time this can pose different threat and risk to the stable coin so there is this topic of collateral degradation and I think maker Dao is a prime example of but you have also the risk related to governance itself remember being the stable coin that's for the governance attack and you have another thing that could be the governance bloat so governance too slow too heavy too not efficient essentially and that also hinders the growth of the project so talking about collateral degradation I like this example because it's very straightforward it's essentially three years of making Dao history where you start with the Chad make it outside where only East me backside so maximum amount of resilience and then maker MCD comes out nuclear collateral types were included and also the infamous price stability module leading to the accumulation of usdc and other centralized stablecoin in reserve the collectible of maker Dao degraded because of the decision of governance we are now in the pigeon stance where it goes further because real-world assets with trusted parties are embodied as collateral and also trusted on-chain assets such as for instance LPS managed by another protocol that are relying on a multistig so essentially maker is accepting all kind of risk and needs collateral now and it's a scaling strategy but here we're talking about risk so of course it's you know increasing risk for us um this one is interesting the technical exposure technical exposure often people understand the direct technical exposure so you know you interact with liquid TL USD you're interacting with liquidity's contract you're assuming the risk related to those contracts same thing with others but what is often missed is you're also assuming the risk related to any protocol that protocol is using and here is a prime example we have a stable coin called AG Euro that was using a usdc escalator and those usdc were deployed into Euler Finance to produce yields pretty cool so far except Euler Finance got wrecked and the money was lost at some point ag0 had no collectible lucky for them they got a refund everybody got a refund that's very rare in D5 when that happens but essentially for me it's still you know of course the project got saved because of the refund but I really hope the infrastructure will evolve in light of what happened uh and then last category is the liquidity model often overlooked to but there was a key factor for instance in terra's UST downfall where liquidity was about 25 x undersized uh and and definitely contributed to the downfall so here you have few things to look at the liquidity size the cost of maintaining that liquidity and overall the uh how efficient that liquidity is and the goal is obviously to assess is this sustainable can the project keep doing that 10 20 30 years from now do does it even need integration from the project and so on so you have the framework I hope I wasn't too long now let's go into a bit more comparisons so it will make more sense so here we're looking at LUSD stablecoin from the liquidity protocol immutable code using only is collect form we're looking at usdc centralized stable current produced by Circle and we're looking at the infamous UST from Terra Luna that blew up about what a year ago uh and it's just going through the five categories I gave you but articulating them and we I guess we don't need to dive through the full thing but what I want to highlight with this side is listen to this and yeah I guess I'm going to ask you a question after that is just listen to the speech and tell me if you would invest in such a stablecon project so the collateral is highly correlated to the system extremely volatile with slow liquidity the control party is essentially the reserve of the Terra Foundation that are hard to assess the governance model could be described as a theater with maximal amount of centralization the tax infrastructure are upgradable contract with massive dependencies to all the contracts such as incor and finally this liquidity structure is maintained by bribes and the foundation money and and all those activities they were doing my point is really to show that I don't think any of you in this room would put a dollar in that concept if it was presented to you as such and this is just like from the terabyte paper you could conclude that so my point is also I think it's really important that we communicate more clearly on the risk and this is what I'm trying to do today so let's compare some Stables to get a better sense of how this articulate so here the longer the bar the more risk is posed by the state category so you know it's going to be pretty straightforward collateral nature with liquidity it is usdc and other assets so you do understand that obviously the risk posed by the collateral nature is maximized on the Frac size minimized on the liquidy side same thing for the counterparty there is no control party for self crystal is which is the case for liquidity there are counterparties for usdc whatever the flavor governance model Equity has known so no risk flux has one posing different amount of risks they are improving on that front uh potentially the the the bar could be lowered to the half potentially soonish but I am judging what is live I do not give a care about your promises of decentralization ah then technical exposure so here essentially flux is interacting with tons of protocol and so this is why you see a various a large difference I would say in terms of direct scope it's pretty similar but the indirect scope flux is exposed to like 50 each protocol liquidity is exposed to itself and that's it and where it's hardest to assess is the liquidity model because keep in mind remapping the risk posed by the category here so if we are mapping the efficiency flux will have a very long bar and Equity against a smaller one but here we're mapping the risk and so flux essentially flux liquidity model is extremely efficient but also extremely risky so if you map the risk posed by that you obtain something quite sizable I would say liquid on the side of lower means but also much lower risk posed by this model and you keep in mind that this is now but Trax has a governance so maybe six months from now it's different as I was saying USD came out recently violated governance procedures to release pretty interesting right so let's let's look at what's in it's in there you have uh so far axis as collateral and they're voting to add State ease fraxis was added based on a curve team decision despite them insisting on the diverting it pretty interesting but anyway we're not here to talk about this but you understand that frax is uh I mean maybe you don't but it's an East LSD and so far there is no detail being shared on the validator model meaning that it's not possible to make a risk assessment of fraxis as a token or this assessment will conclude the risk is infinite or I cannot measure it so of course this collateral is supposing quite an amount of risk the governance model here I put it Max risk because again not only it's not clear but it's also unrespected so it's really hard to just measure what's happening in terms of technical exposure we want something a bit bigger than liquidity there are more components to the system obviously but you know I got to be honest for what CRV USD achieved I think the technical exposure is quite minimized and the liquidity model here I picture it as less risky than LUSD we have to be honest simply because remember that curve is a tax has a lot of fire power and capacity to attract liquidity natively and so it's kind of a bit projecting here because remember that Chevy USD is what A month's old but yeah I would say probably over the short term it's safe to say that the liquidity model will pose less risk so I know you might be like Bryce okay it's fun but nothing is is better than LUSD right and I mean the way I see it is I describe it as you know when you play those role-playing games and you have those diagrams for stats you see so like you have strings and endurance agility and dexterity and like if you increase strength it will lower a bit your endurance it's kind of like how I see it you know those five dimensions and the way we say it now is yes in terms of absolute values if you average it out probably LUSD tops it but LUSD tops it with a bias you can see on the chart which is minimizing the risk on the first three dimension and I'm gonna do this one right after I wanted to do this one first so this one is interesting because it's the first day born since six months of this framework that achieves a score equal to LUSD so here you have something of relevance potentially and that's essentially a liquidy fork but they are going to Peg it to the Chinese one CNY Fiat so that's the main difference the rest is similar the other difference is it's a fox they don't have investor and so they allocate more of the token the healthier token the featuring token to liquidity management and so this is how you can explain the difference between the two because For the Rest it's clearly an easel Fork so technical exposure is a bit higher for alternative that's the name of that protocol simply because they use more oracles than liquidity they need to fetch the cni price against USD on top of the USD price against ease but then the liquidity model poses less risk because they have more means to sustain it so yeah I thought it was an interesting one uh yeah I just added this little not two this is based on the white paper specifications it is not live yet so let's hope this ship as such this is another gravity liquidy fog gravita that came out recently gravital main changes is there is governance and they accept liquids taking derivative tokens as collateral and I guess with that you can understand the differences in terms of response that collateral is much more risky as seen and they have governance very young governance very trusted so obviously Max risk on governance too but it's quite normal for a young project and obviously the technical exposure is also bigger and is likely to rise over time because they have additional infrastructure plane but then here again younger project more means more token they have potentially more means to sustain the liquidity which means the liquidity model is less of a threat um so yeah I hope this is helping you see things now I want to get you got this you're really like 99 above like 99 of the pack in terms of stablecoin including stable coin Builders and I am serious about that no we go a bit further in the actual subtlety uh because it's really hard to compare stable coins at face value essentially so here I wanted to show you something where it's this concept of algorithmic market operations uh sometimes it's called fed it has different names but it's essentially a system for stable coins who can mint arbitrarily the stable coin we'll use that conditionally to supply liquidity across D5 so that means you will have stable coins that are minted with zero collateral but only under certain conditions this has relevance in terms of scaling but it also leads to confusion in how you analyze the stablecoin model because if you look at LUSD versus flux at face value you will find 1 billion fracts and about 260 or even more maybe 270 80 million USD now so you'll be like okay faxes full-time Lux market cap LUSD is an end as fax funder would say a year ago but now you look at collateral and you see what happened at collateral you have about 730 million watts of East for liquidy you have not even 200 million dollar worth of stable for flux so liquidity has roughly four times the collateral of flux already despite at face value having a fifth of its supply of force of its Supply something special here okay let's see what the difference is of course the ammo flux Supply is 82 percent produced without collateral liquidity doesn't have enamel it's not possible to Mint LUSD without supplying a proportionized amount of each collateral so if you compute all of that into what we call the outstanding supply of the stable coin which is the supply that is not produced by an ammo or if you like it better the collateralized supply we don't do 180 million for facts liquid is not moving an inch because there is no ammo you make of that what you may but I really wanted to highlight that you see if you're going to cut market cap or coin Gecko and you see USD at 260 million flux and 1 billion and you conclude Frac scaled more than LUSD you are missing the story because you need this context on ammo to understand that essentially 80 of the fracs have no backing uh and and you know it's a very different risk model essentially and Outlook so how do we go from here what's next what do we do what do we make of that I think you have some ideas already but as you may have seen uh it's very Lively recently in terms of fox we're seeing a lot of what we call CDP collateralized position protocol launching um and they like especially to support the liquid sticking derivative of is so the sde sprocket is fraxis and so on so we've seen quite a few one launching most of them are either liquidy or reflexive Forks and here on screen I put some ideas for you want to be focused if you want to improve on the model essentially what we had so alternative we talked about it that's a blue one I'm gonna kind of skip gray using it too with the stable so let's focus on the two of the one cutting the Box on the bottom left and on the bottom right it's tie a reflexive Fork that launched just a few days ago so here um yeah just some more context on the left is this is gravita you can see the various tokens accepting as collateral on the right this is cut in the Box it's an interesting piece of infra because this is an immutable protocol just like liquidy enabling you to leverage stickies so in my risk analysis it's literally the most resilient protocol to leverage your stake is right now and it's barely discussed in all space because people are more eager to share about a bigger protocol that tends to be more lucrative I guess even though cut in the box is extremely profitable for for those who know what they're doing um so yeah this is more getting into a bit of the meme series sorry thinking about what's next uh and I guess the guys from reflex are doing a bit better than our set liquidity on this front where they have this money got linked to talk about their reflexive folks uh and it's a really interesting idea so you know essentially to put it simply if you want to Output a stable coin right now you have two battle tested infrastructure you could honest that's one of liquidy and the one of reflexive the each come with different buyers I would say liquidity is a minimized infrastructure maximizing for resilience I would describe reflexor as a more flexible infrastructure allowing you to decide what are you packed too you know that's that's uh um the controller essentially so you can make you can use your right infrastructure to Output stable that are literally tracking any asset you might need gold silver whatever so to wrap this up I wanted to bring a man that is not one of my darlings but still helping us think about what's happening right now so this guy is Friedman you might know him recognized Economist liberal and so on but it has some interesting thoughts so the first thing is this question about results you know is think about how central banks are operated who decide and what kind of uh Overlook there is on their policies uh and then think of how or central banks in crypto the rise and liquidity the die and so on are operating to me it's obvious that you know uh a central bank is operating with minimal transparency close to no accounting no consequences for the people who are taking the decisions and no ownership we are already doing much better I think because we have transparency how we decide if there are changes people are held accountable we look at the result and learn from here so it's it's already interesting to think of kind of feeling one of my next out too uh this one is more foot falling because you know Milton Finland was doesn't like government so much and so it was saying if you let them manage the send they will find a way to run out of it why not uh but it's kind of also you know to have in mind in terms of how of Fiat currencies is managed recently so I'm not going to give you the stream and I don't want to turn this into something political but keep in mind for instance that I don't have the exact figure in head but it's something like a third of the total supply of the dollar was produced over the last two years let that sink in a stable coin with 15 yearly inflation they did it uh and yeah the funny thing is this Freeman years ago in like the 70s 80s he said something he said like get rid of the fed and put a computer in its place and uh you know it was quite Visionary because if you think about to be honest that's more like the reflexive motor but that's literally what reflexor is doing the central bank is now a series of algorithms taking decisions on the interest rate uh and stabilizing the system based on predictable rules and I don't know about you but to me this is a massive Improvement in how we manage our currencies and eventually I would love to see them this logic coming down to the Fiat even if it's probably really optimistic so I guess yeah I wanted to give you context on analysis risk analysis for stable coins but also kind of share this there will be all kinds of stables they will use all kinds of collateral track all kinds of assets using all kinds of infrastructure to maintain the Peg and price stability and liquidity with governance or not on various chain cross chain or only one chain all kinds of flavors that's a good thing that means the diversity that means loads of concurrent models and that means essentially you know the market crashes will do the purge over the next few years or maybe we'll have 100 stable then we have a massive East crash deleting 10 Stables fine at least we know the 90 remaining are able to stomach this 40 East crash that cause the other to crumble also and you see the process so yeah all kind of stable of all types that's a good thing and we need to explore more because I think we just barely at the beginning of how we try to build a decentralized stablecoin and obviously I think I I didn't interview intelligence by starting with that but just in case you remember the beginning 130 million in stablecoin about 40 million D5 another metric we should add is out of the 120 million stable coins 119 million point eight are sensorable trusted there are ways you can prevent people from using them or freeze them or so on so essentially the main problem we have is we have loads of sable coins but they're extremely weak they are you know centralized for most of them and if not directly centralized open centralized by proxy so the core question for me is how do we solve this because let's not fool ourselves the current situation just doesn't work long terms guys like literally this is not D5 this is not resilient this is not censorship resistant and so if that's how we view things I mean that's not what I signed up for for five years ago when I started doing D5 so here I am today trying to push forward small resilience for stable coins hope you guys uh learn from that yeah feel free to ask questions if you have any I am token price liquidity manager at liquidy and involving a few other projects French Community Maverick and all beloved luchadores foreign thank you Bryce for an amazing talk it was a real pleasure um we're gonna check real quick we are we are out of time I do want to see with the audience if anyone has any question before we oh yes yes please go ahead let's think all right um well first I want to start by saying I'm a huge fan of LUSD and the liquidity model I I do think it's the model to enforce you know from the at least the centralized point of view and to give you context with regards to my question I'm part of synthetic Spartan Council and we have leveraged LUSD I mean if you definitely know about synthetics but maybe for the rest of the audience synthetics is also a stablecoin provider but apart from the Native way of minting susd which you haven't really talked a lot about in this presentation but we also leverage LUSD to wrap on top of it and mint SSD and it was our favorite stablecoin to leverage apart from you know the native one however the issue we had is that there just isn't enough liquidity for LUSD out there to satisfy our mean so anyway getting back to my question how do you you know from your point of view the LUSD getting to like 10x 100x of market cap because I I really want it to be there yeah I was a new really cool one and uh just to go back on this LUSD with synthetics it's pretty awesome because essentially you guys brought LUSD to optimism back when we went there to stabilize the susd and that helped susd for a while and essentially help us to have an established footprint on optimism when things picked up so it was really like win-win situation I really liked so anyway going back on the liquidity there are several ways to look at it um so the first thing is you know kind of putting it like that right now you have about zero two percent of all the ease in existence in the liquidity protocol I think we could easily go up to five percent without trouble so you know it's already quite a sizeable increase about 25x and that would lead to a supply of LASD with everything as same of 5 billion even with Chinese price on 1800 so that's already you know there is ways to scale it I guess the main question is indeed the liquidity and there were a lot of progress made on that front too thanks to new tooling so the main thing that happened before I use this liquidity uh and I will wrap it on this is stable swap stable swap I said it on stage yes what is table Swap stable swap is a formula used to concentrate liquidity on stable coins assuming a one one peg which is really good when you're doing usdc usdt but it's really one of the worst concentration you can harness when you're doing LUSD usdc for instance and your USD has a premium and so this a dampened the efficiency of value is this liquidity for a while because table Swap was everywhere and really hard to do anything else luckily for us unified V3 gave you a range so you can say I'm willing to provide liquidity on LUSD from one zero zero zero something usdc to 101 for instance and other tools such as Maverick that goes even further in how you can craft and shape that liquidity did wonder to help essentially contain this premium you have on ausd so to put figures we went from two percent average premium meaning you would need 1.02 usdc to buy a USD to know 0.6 average premium over the last month even 0.5 depending on how you measure so essentially you could buy one LUSD with one zero zero zero five usdc which starts getting acceptable uh it's hard to pick your worldwide there would be no premium on LUSD because it's stemming from its characteristic you know I often say that LUSD premium is also stable coins weaknesses priced on the market you know so like you have two things representing the dollar one can be censored the other cannot to make kind of make sense that the one that cannot be is smaller than the one that can there's you know you're not pricing censorship so I think there will always be the premium but containing it is really helping grow the liquidity further and yeah we are parallel to the premium containment the liquidity situation and then USD improved massively over the last year now you can really like set the sizeable amounts uh in the thousands of millions range without a sizeable slippage so it's a work in progress and it's also you know a lot of it is coming from community so for instance you know the gravita launch help build USD which in turn help further contain the premium so like we're reaching the stage where people who like LUSD liquidy can build all kind of projects and they will try to harness LUSD and that can result in improved liquidity and big containment anyway I'll be around if you want to ask me a question after the talk always happy to chat stables for hours long and if you have a stable coin in white paper you can shoot it to talk and price I read five a week at least and I'm happy to give feedback on all of them this is really amazing so yeah feel free so let's give a huge Applause for Bryce [Applause]