Putting the 'De' back in 'DeFi'

ETHCC 3 · March 2020 · 30:43

Cassidy Daly and Abbey Titcomb examine decentralization in decentralized finance.

Speakers

Topics

From ETHCC 3, hosted by Ethereum France.

Transcript

okay sorry for the delay I'm gonna get started Abby's gonna join in probably just a couple of minutes she's finishing up another top tryst next door but I'm gonna go ahead and jump into it and chill she'll join so today we're gonna be talking about putting the D back in D Phi and so what we really mean by that is that we don't really think D Phi decentralized finance is truly decentralized today and where we're gonna go today what is do you find what do we mean by it what actually makes a D Phi project decentralized and

having a closer look at some of those D Phi projects a range of decentralization actually between them and then closing with what can dowse actually do for D Phi to start opening it up so what is D Phi and there are lots of different definitions and most of them touch on some aspect of what decentralization means but they really range between Bank the unbanked super vague but also important but we think that probably the definition that really brings the D from d fi to the forefront is the the one at the bottom here that decentralized

finance is the movement that leverages decentralized networks to transform old financial product products into the key words here are trustless and transparent protocols that run without intermediaries and we think that that's probably the most important thing in a D Phi product compared to a C Phi product or a centralized finance product so first starting with D Phi means according to coinbase the operations are not managed by an institution the code is transparent on the blockchain gaps are designed to be global from day one

there are permissionless to create permissionless to participate a flexible user experience and interoperable but this this also includes things like that it's open source that it's auditable transparent global sense of censorship resistance these are all good things but definitions like this actually result in half-baked FinTech solutions that use crypto and this is actually Abby's quote here she walks up to the front so let her take it for me hey do we have two mics or just one okay cool

hi guys sorry I'm late I was just finishing up another talk right next door but you were saying it perfectly right is that definitions like the coinbase one although that captured her like the inherent like global characteristics of defy solutions like generally result in like a leniency with definitions of decentralized tech that results in a lot of like kind of fake news around like how we're treating defy and so yeah this is a spicy thing you delivered it really well so if we want to hold ourselves to this definition of decentralized tech which is a movement that leverages decentralized networks to transform like

Rhian vision these old financial products into trustless and transparent protocols that run without intermediaries we need to be very transparent about which defy projects are actually trustless in transparent and with running without intermediaries and which ones aren't so what actually makes defy decentralized coming to a real definition here that's more concrete that we can actually start applying to projects the first thing is that there's minimal and distributed intermediary custody so that's things like admin keys are not controlled by one owner and and there are tools that we can use to do this such as multi SIG's but if that multi stick is controlled by one come

vini that's not really actually useful right Dow's are another good tool that you can use but if the towel now is actually only controlled by a few entities that's also not really using the tool the way that it could be ideally there are decentralized price Oracle's as well and a lot of projects do you focus on this and that's something that's a positive in the space but this is something that can also be improved and made more decentralized and then also distributed sharing of benefits so for example that means that all of the fees that are collected by the protocol are not going into a Treasury that's controlled only by the company and that's also something that

we've seen so following up on that we also have the concept of having open and decentralized governance for technical upgrades so how we're defining parameters changing them upgrading protocols as they develop you know having an open and transparent process into how these decisions are being made is like an inherent quality of a decentralized system and obviously we'll talk about where we see a lot of projects lacking the next one is you know going along the bank the unbanked meme which is how actually globally accessible these projects are so what is the KYC is there kyc necessary and how

accessible these products are to actually use is like a very important requirement into how we can actually create you know trudi five projects and then the last one is you know the how the operations you know that operations the assets the brands like the communication channels or the logos or the money you know naming are not going owned by one company and so this is what we see a lot as you start to see these projects you know emerging is how are these projects controlled do they exist outside you know they can operate without intermediary custody but are they technically really operated by a company and

these this kind of like invisible hand of centralized you know forces that come in the form of companies or projects or founders can take a toll on the decentralize aspects of d5 projects it also does mean that it's open source that there's an auditable history and since there's a persistence but we think that these points actually listed here are more important than those things in terms of really getting down into the details of these dphi projects and if they're truly decentralized or not so let's actually start to dig in to a few projects so to start we thought we'd look at dydx right so dydx is a non

considable trading platform that enables decentralized margining trading and if we look at dydx that actually satisfies like all of these requirements so there's no unchain mechanism for any party to upgrade the logic does the base protocol of dydx and the team only recreant retains the power to put the protocol into like a closed only mode in the event of like a major bug or vulnerability and so if I skip ahead now going along like how does the protocol then get upgraded if we go to how its open how its governed like how technical upgrades or governs and they actually upgrade implements so they actually implement upgrades through the deployments of second layer contracts so

these are on top of the base protocol so it's not the base product pole that's being updated but it's actually contracts that are written in deployed on top of it and these contracts can be won they can be written and deployed by anybody and the user of the protocol is actually in control of what contracts and what upgrades really that they want to use so that creates like a very open and transparent process for how a protocol is being managing governed but also creates a more accessible way for you know how I interact with this protocol and so going back up to the top is that these second layer contracts you know again the decentralized price Oracle's

it's the second layer contracts enhance the margin calling mechanism so it adds this extra protection for traders so it's not really decent rising the price of Oracle's but it's adding this like additional layer of security that has and kind of adds to the decentralized aspect of it but then they also enable margin tokens and so these margin tokens are freely tradable fungible and transferable and they represent ownership interests in like dy/dx margin positions and so these definitely satisfied the distributed sharing of benefits but again it's like built on top of the protocol and not like really baked into the base protocol and then there's no kyc needed and the operations assets and brands are not owned Y by one

company there is a dy/dx team but again this protocol like exists on its own you know like if the team has the ability to have this closed only mode but outside of that liked it very much so is resilient and autonomous in its own right so then we go to compound so compound got a lot of flack by some like medium articles and we know that they've grown so this is like first one but then we'll have another analysis on compound with their recent updates and but so compound is an algorithmic autonomous interest rate protocol and so the it has a lot of so if you look at the first requirement mam it doesn't really have that much distributed intermediary

control so compound has like four different administrative positions that if compromise could lead to the drainage of all lending pools and this comes directly from amin salim Inez analysis of the compound contracts that kind of led to us realizing that like a lot of functionality of the contractor architecture was like really controlled by these for admin keys including the price Oracle's and all that different stuff so there's not that much there's a lot of like intermediary control in the custody process for the compound protocol but they do have like kind of decentralized price Oracle's so they launch this thing called the open Oracle system which enables developers to

contribute in leveraged pricing data from a distributed price feed by kind of creating like a standard that allows reporters Society key value pairs that interested users can post to the blockchain and so this was in an effort specifically decentralized as one part of the process which is really cool and the third is the distributed sharing of benefits so one could say that using the protocol you're receiving some of the benefit because you're you know receiving the interest but inherently in the protocol you know there's this centrally controlled treasury of collected fees that compound is producing and pooling and that's written in their white paper which kind of

actually leads to like a very centralized little aspect of this in relation to this requirement and on top of that the fee that they're collecting is only determined by them yeah exactly so then on to the open and decentralized governance for technical upgrades of course we'll touch on how what compound is recently launched with their governance token later but before this you know the process was again designed to be only upgraded by these admin keys that are as a result of the first requirement above so the important contracts are proxies which like simply point to a separate contract address which holds their implementation logic and the admin keys can kind of change

this at will but the protocol does not require kyc to use but compound is like a vc-backed company which is like not a bad thing but like say at it as it is right and so the operations assets and brands are owned by one company so I I think we're we were a little harsh on compound and I'm gonna be a little harsh on synthetics as well so synthetics is a crypto backed synthetic asset platform that provides unchained exposure to real world currencies commodities stocks and indices and and we would say it is pretty c5 actually there is a proxy contract architecture which has

significant control of the network it has a centralized price Oracle so both of these things you know would would argue that that synthetics is actually controlled by one user one positive site here is that there is a distributed sharing of benefits of the rewards and but at the same time the there is no open and decentralized governance on the other hand it's globally accessible so anyone can use synthetics but the operations assets brands are all owned by synthetics and although this is a critique of synthetics the team has expressed a desire to actually open up the project add an element of governance as well probably hopefully opening up

their contract architecture so this is just to say that these projects they really exist on a spectrum and there's a spectrum between being really decentralized and being centralized and they're not stuck in one place right they can actually start to move towards the decentralized side and open up the project and so we really hope that projects are gonna head into this direction this is just a small sample that Abby and I kind of feel like we're gonna get a lot of photos like we just put our thumb in the air and we're like well let's see what we think but I think that the important point in the

analysis and the creation of this spectrum is that there is like this specific lack in the governance and control of protocols and platforms like at the end of the day these protocols can satisfy a lot of their requirements with the decisions that how these protocols are parameterised and how changes are made and how they're developed can kind of be abstracted away in different implementations of governance of implementations of governance and just in general like how the protocol people decided to develop the protocol so jumping into what dows can do and defy so how do we define those and defy and why do we think it's

important so we think down membership need to be open and not limited to a specific group that down members can actually propose and vote on changes and their decisions are not blocked or modified by a central authority and that they receive direct or indirect incentives from participating in this and that actually aligns the incentives all of the participants in the system and so that Dow's kind of now our introduces this tool that can give us new means of introducing transparent and open governance into the technical upgrade process of protocols or in the decision or the you know parameterization and so that's a interesting thing that a lot of

different projects are exploring so um let's take a look at a couple projects that are incorporating dowser trying to incorporate dowels and so the first is make your doubts we all know maker maker is pretty famous but maker is a decentralized or aggress dedicated to stabilizing die well sigh and die and maker employs the maker token for governing critical parameters of the maker protocol and so it really follows this formula of you know for-profit dows introducing token incentivize governance and it satisfies the dowel requirements at a glance but if you dive deeper it kind of uncover a little bit of stuff right so um down membership is open and

not limited to a specific group because anybody can buy maker tokens and anybody can contribute you know and the contribute to the governance process but there's such an unequal distribution across Wales like a 16 that limit actually the access of this down to the common person and so the governance process kind of becomes closed off because of this very substantial unequal distribution of tokens across the memberships I mean so technically the proposals and the decisions can't be blocked or modified by a central authority but a lot of things make this like really

convoluted right so again if there's an unequal token distribution with --eline parties maybe aren't aligned with other parties these whales control most of the votes and actually limit code you know aren't one central authority but are a collection of central authorities that are acting under the you know of the intentions of a central authority and so although inherently the use of a Tau is it can't be blocked or modified it can and it has been before and also the governance is kind of abstracted away so in this governance UI that maker uses it only shows votes done by the maker Foundation

and the maker foundation still holds this central authority over the maker token whether it's actually in the token distribution or actually in the as I said like invisible hand of the foundation and how it has its influence that it can exert in the maker ecosystem but you know it does satisfy the last requirement and because incoming interest rates back this like by Bern which reduces supply of the maker token so I mean if a be started Adel with a thousand tokens representing ownership but she kept 999 of them like what is even the point of me having the other token there is probably no chance of me ever influencing anything and I think this is something that a lot of projects

haven't actually spent enough time looking at is the distribution of the governance power and just because anyone can hold maker tokens doesn't mean that they're and just anyone has influence or the ability to truly participate in maker dough so look at compound they're down membership is technically going to be open and not limited so this is their plan going forward but right now it's just distributed top-down from the the compound company and at the same time though it does include delegation so those that hold these compound governance tokens can give voting power to non token holders so there's an

element of openness that they're already trying to implement in the system but we do hope that eventually this becomes a bit open going forward anybody with 1% of these compound governs tokens can propose a governance action so that is something that makes it open to be able to propose on decisions and I think this is something that is important at the 1% they're using for spam prevention basically and so it is an element that their I we would say this in the green they're doing well but at the same time it's it's not public yet incentives come and the ability to influence the development process and so there's a

little bit of ways that compound can can grow going forward but on compound is that I mean they just introduced this I forgot how long ago but this governance token really resembles more of a community governance approach than a for-profit incentive a token incentivize governance approach which is super nice right because it's maybe taking the bad parts of maker governance and in pairing the good parts you know with the governance token in which you know you could actually have an influence on proposals that are happening so although it's centralized now you know the plan is to progressively decentralize that and that's super nice and this is also coming a long way from not having any governance at all yeah so they're

definitely going in the right direction here and then a really nice finishing case study here on DX Dow they really check all of our boxes so DX Dow the initial purpose is to govern Dutch X protocols parameters so this is smart contract logic token whitelisting etc the down membership is open so reputation was originally allocated to all of the users of Dutch X Pro to call but basically any anyone who wants to lock down ether or other tokens that are part of Dutch X protocol can get reputation as well and anyone with an aetherium wallet address can create proposals for DX tau which is really cool that it's just open it really

opened for anyone and then participation in the Dutch X is providing these indirect incentives so the ones that would want to participate in DX Dao or the one using debt checks and there there are aligned incentives there so we really think that they check all of the boxes and doing a really great job in terms of using this tool in a way that would actually be impactful to D centralizing the project but at the same time it's not getting a ton of use and so these much bigger projects like maker down compound are not using these kind of tools and this project is also not getting as much traction probably as they would like so pros and cons there as well so also

there's you know the incorporation of dowels but there's also just the incorporation of like additional governance layers that have Dow s qualities right so Mellon implemented the Mellon Council which is basically an aragon dow for on chain voting regarding a bunch of different things with the Mellon project so that includes inviting new members adjusting the is it am goo or am jus I don't know I've never said it out loud price updating the subdomains and updating protocol parameters and making decisions about inflation spending so this is super nice because although it's a specific group of people and it's again not like really built in to the protocol it's just like layer on top

that is still introducing more transparency and openness into the decision-making process regarding how Mellon is being managed and deployed and developed and then the second is like you know swab so like you know soft kind of has this like invisible Dow built in into how it works right so it allows any next participant to list any token for trading by allowing them to create a liquidity pool and then these participants actually like own shares in the pool that entitle them to the pool revenue and so this satisfies that like you know financial and said that the incentive Dow requirement and and kind of creates more of this distributed like incentive alignment like more into the

actual protocol and into how it works and so that's like another really cool layer that we'd like to call out so the problems that we think that does can really address starting with reducing the risk of financial systems and they really do this by increasing the transparency of systemic risks within the system they can actually start to align the incentives much better of all the stakeholders that are involved open up access for participation in the governance processes and increase transparency into that decision-making process itself so these are really key areas where dphi projects can start to used out to open up different layers of

their systems so this is another maybe just slide that we threw Isaac cannot take but again there I think that does need to be like maybe not maybe not even a narrative shift but like a realization that like a dows don't have to be for profit token inside advice governance right like we can be looking into progressively introducing more open and transparent governance into D Phi projects and protocols that reduces systemic risks of centralization right like yes we all progress it like progressive to decentralization is a thing because it's safer to introduce centralization at the forefront but if we aren't like very aware of how that centralization can propagate and like

continue and like leave residue then again we're just gonna end up with these projects that like don't really satisfy you know these and don't really like live up to the defy definition and so just to go back to that is that open accessible Dow's can enable open and accessible community governance that have aligned incentives models financial or non-financial and we're seeing that implemented in projects now and exploration into that for other projects is something that should be encouraged and supported so to end and and draw this together we want to focus on progressive decentralization so not just mending governance tokens controlled by VC Wales not just recreating centralized

financial tools that already exist today and just putting it on a blockchain and not introducing systemic risks into this emergent new ecosystem so building towards a true d phi nice these guys I'm not sure if we have any time for questions do we how do we can we take some questions or okay all right