DAPPCON 2019: Why Finance Can be Better on Ethereum - Ethan Bennett (MakerDAO)

Transcript

I'll go ahead and get started I do want to mention first though that this is sort of an experimental talk it's about some stuff that I've been reading and really interested in lately and in preparation for it I am asked just like pages and pages and pages of notes and then I had to distill it down into this and so we'll see if I distilled it well all this to say very open to feedback and I hope you get something out of it we're gonna talk about a few reasons that I think defy is really promising and in particular we're gonna talk about reasons that I haven't heard people talk about so much before so the bottom line here is that defy or more broadly a new ground-up

financial system allows for new approaches to economic thinking that in the traditional system and in academic economics can be very boxed out pushed to the side because of entrenched ideologies within those institutions so the questions I'm gonna pose are do new approaches to economic thinking have a place in the legacy system and do they have a place in defy and then we're gonna talk a little bit about complexity economics and we're gonna talk about it for two reasons the first and the most important reason is because it's a somewhat arbitrary example of a new approach to economic thinking that we see and defy and which is pushed to the periphery in academic economics but then the second reason is I find it really

interesting and hopefully some of you do as well as some of you do too so with that I will dive right into this giant wall of text you're about to see apologies in advance but it's a really great quote that I think sets the stage nicely for these points I'm going to make this is a quote from Yanis varoufakis who is a the former finance minister of Greece now writes quite a bit of really excellent content this is neither here nor there but a lot of stuffy old white men like to call him a bad boy which I find just like really awkward and hilarious so I put that on here so the quote is economics and universities began to evolve from the 1950s onwards as the queen of the social

sciences and what gave it discursive power and monopoly power within the academic environment was the claim that it was the only social theory which was peddling universal truths to be proven by mathematical means and it succeeded however in order to close the model mathematically the only way to solve the equations is by making assumptions that distance the model from what really exists in capitalism so for instance you have to assume that there is no time and there is no space because if you allow time to interfere with your model or space to enter it you end up with in determinism in other words you end up with a system of equations that cannot be solved or that have an infinity of possible solutions and then you have no predictive power you can't say this is what's going to happen so you have a very interesting inverse Darwinian

process the more successful economists were at creating models that said precisely nothing about capitalism the greater their success in the Academy they create wonderful abstractions aesthetically pleasing models that I spent quite a few years studying in the same way that you go to a museum and you look at a piece of abstract art but you don't expect to find the truth of capitalism in its form so I think this is an important quote because first of all it makes the point that I was trying to make that academic economics can sometimes have a gatekeeping effect towards new ideas in economics but also it reflects the sort of abstractions that you find in economics or in academic economics which sort of keep them from representing the reality of financial markets so my argument here is going to be that complexity allows for

more accurate models but first a couple of quick definitions economics in the course of this talk we're going to refer to as the study of how outcomes or structures in the economy derive from the behavior that produces them and further the definition of complexity we'll be working with is that elements in a system react to the pattern that the elements together create in a recursive way so we'll talk a little bit more about well we'll talk a lot more actually about how this relates to economics but from the beginning that's the basic definition we're going to use so back to the idea of particular abstractions keeping models from accurately reflecting the reality of markets in standard economics we have a few major assumptions the first that

we'll be talking about is the assumption that equilibrium is the natural state of markets the second is that actors within those markets tend to act rationally and the third is that all of the actors within those markets have equal access to information and that information can be spread throughout that market and I a perfect or an uninterrupted kind of way with complexity we're gonna assume pretty much the inverse of all these things non equilibrium or dynamism is really more the natural state of the economy that actors within the economy can behave unpredictably and irrationally sometimes may be all the time and actors may know more or less than other actors so just a quick spoil oh yeah not quite as the spoiler yet to just dive into that a

slight bit more some words that can define standard economics in this context we tend to see order we see well-defined problems deductive rationality wins in most cases stasis and there's no concept of time I its pristine its pure its elegant it's artificial complexity economics we tend to see a lot of contingency people will act in one way or another depending on the situation at large there's indeterminacy our models can't necessarily predict exactly what's going to happen all of the agents within the system are trying to make sense of the system that they're involved with all the time so there's openness to change there's path dependence which means that small changes within the economy and also in the behavior of individual actors can make really large differences

on the patterns that emerge from that system so it's organic its algorithmic and it's full of messy vitality this all comes from Brian Arthur who was a researcher at the Santa Fe Institute and one of the pioneers of applying complexity to the economy and there's just a ton of really interesting writing out there if you're interested in it so that's a good place to start so can we view markets as complex systems I would say yes for two major reasons at least in the context of crypto the first is that there's a fundamental uncertainty in everything we do the agents within the system don't really know what they're facing so they're always trying to understand the system that they're in they're trying to explore the system that they're in and

they're constantly adapting to the changes that are happening within that system and also we can think of finance is really a collection of different technologies right we can think of that even before crypto but especially in crypto now that's particularly relevant and technology is always changing at all at all levels of the system so that keeps finance and in particular it keeps defy in a constant state of disruption and that's really that's really important because if we remember that recursive definition with which we defined complexity that leads to this situation where the particular actors within the system are constantly adapting for the system itself the system is constantly changing because of that constant adaptation and so the

actors have to continue to adapt so I'm gonna go into some examples of complexity to try to explain why it's so why I think it's really interesting in the context of crypto but first I want to give you a look at where we're going because I don't want to just dive down a rabbit hole and expect you all to trust me that it might be relevant at some point so the reason that it's relevant I'm gonna pull it back to maker governance and in particular some of the stability fee issues that we had throughout the early parts of this year and point out that in crypto and particularly in maker we don't assume that crypto prices are converging towards an equilibrium that would be ridiculous right volatility is the natural state that we assume we also don't assume that crypto traders act

rationally necessarily this is kind of a loaded statement because of course there's a ton of crypto that's based on game theory which does assume some degree of rationality but I'm not gonna talk about that that's kind of out of scope and third we don't really assume that crypto traders have the same level of knowledge right especially people who are opening CDP's it could be people who are experts about crypto it could be people who are experts about finance or it could be people who just learned about crypto yesterday and we're trying to appeal to all of these different ranges of expertise and my argument is that makes a healthier system so to move into some examples I'm gonna move pretty quickly through these but they're pretty interesting and they allow or they're good things to read into if you're interested in this topic so the first

set of examples that we're gonna look at I particularly focus on emergence which is I'll define it as an emergent pattern a pattern can appear when a number of simple agents operate in an environment forming a more complex behavior as a collective the whole is greater than the sum of its parts and small changes on the agent level can translate to huge changes in the macro level patterns so I'm gonna show you some examples of this with cool gifts the classic example is traffic oops well I guess the videos aren't gonna play in the pdf version that makes sense now that I'm thinking about it so this is a cool video of traffic I'll just kind of explain what why traffic is an emergent pattern so we have you know traffic forms because

individuals are driving the cars in particular ways with particular strategies for driving and particular like maybe they're in a rush maybe they're not in a rush and so we talk about traffic as if it's like a tangible thing but it's really not it's a pattern of behavior from the individuals and just by understanding the individuals behavior we can't exactly predict what traffic is going to look like we can predict maybe certain things about like the volume of traffic or how long it might take you to get somewhere overall but we can't predict that a traffic jam is going to happen in this time in this place another example this is so cool I wish that you could see it it's been going all over on reddit lately so I look it up but I answer a really good example to like based on the simple

rules that the individuals with an ant colonies follow you end up getting these extraordinary patterns forming among ants that you know in certain cases they like farm fungi in other cases like in this video there's some really cool footage of them like forming together in a clump and this guy like picks it up and drops it and it's really creepy and it would have given you all nightmares but like in a good way all right so just just look it up it's pretty cool so immergence we'll come back to why that's important especially in terms of markets but the main point is that if the economy is a complex system the markets that form from it emerge in this way right we can't really predict how markets are going to behave just by looking at the individuals behavior within that market we can

predict certain things about it but we can't really predict when the booms are gonna happen when the busts are gonna happen and to what degree so the second category of examples I want to point out really quickly is how this how this translates translates to like analyzing the data within markets and within other you know complex systems so generally this falls under the umbrella of what's called agent-based modeling which is an approach to computational modeling that focuses on diversity and dynamism of the agents within that system rather than solely focusing on the macro patterns of the system so what that means as we'll see in this example is if we were to model what the stock market might look like given certain conditions certain price conditions that sort of thing we can get

a general idea of the shape of market movement but we can't see with any degree of granularity uh where the spikes are really going to happen so this is a really cool example that actually was done by that same guy quoted earlier Brian Arthur I and the Santa Fe Institute originally in like 1988 but it's continued to be updated over the years and just for a quick tour of what's going on here the top line is a normal model based on neoclassical economics and the bottom line is agent based models so the way that they do that is just from a very high level they'll program in certain expectations of individuals behaviors and they'll

also give them memories to some degree of how other agents within that system are acting and how the system has responded to certain exogenous events so you'll see actually that they follow a pretty similar pattern and I actually I listened to a talk from Brian Arthur when he was delivering this data and he said if his first impression was like well damn like this whole thing isn't that different from standard economics after all but then you start to look at it a little closer and you see that actually in some critical places there are some differences like right here in particular you can see that it follows the same general shape so you might be able using standard economic models to say there will be a dip in the market sometime around here but what you don't see is the extent of the dip which does

become clear with this agent-based model and of course that's important in markets because that's where all the money's made that's where all the interesting stuff is happening if everything stayed gradual and at equilibrium then no no trading behavior would happen and I wouldn't work you know all sorts of things that we don't really want to happen so volatility isn't such a bad thing in these cases the second example is I'm gonna rush through there's a lot that could be said about it but this is a an example of the prisoner's dilemma game in game theory which is run for 70,000 or so iterations and it's designed as an agent-based model you see here the zero ones and

eventually lots of ones and zeros these represent the memories of the agents in terms of what strategies their opponents are using how those are changing and how the system changes in response to it and so what you might expect to see after a certain amount of time is that a certain strategy would win out you might think that especially in the prisoner's dilemma there aren't that many strategies you can use but in fact what you see is really a vibrant ecology that alternates between stasis and dynamism and various strategies will react to other strategies gaining prominence and mutate based on that they'll learn how to take advantage of other players strategies in such a way that you you don't really see a rational outcome here and in fact this doesn't stop because it reached some sort of equilibrium it just

stopped because that was the end of the experiments arbitrarily so I will put a bookmark here because I think I don't know why it did that I think there's a lot to be said about how this how complexity and game theory meets especially in terms of like consensus algorithms and stuff like that but I don't have a lot of time and that would take up a lot more time than even I had in the beginning so if you're interested in talking about that though it's something I'm really interested in lately so we'll move out of these examples and try to bring it together so me complexity is an example of a reasonable approach to economic thinking the kids push the periphery by entrenched ideologies and it's actually

very strange that it does because complexity as a viewpoint is something that's very natural it's something that borrows a lot of intuitive logic from from like nature for right ants was a good example traffic is something that comes somewhat naturally and so it makes sense to a lot of people intuitively and the only reason really that it isn't a part of standard I can academic economics is because we lacked the computational power to make effective agent-based models until the last few decades really so just as proof of that here's a quote from Adam Smith where he says it's agents banks consumers producers government's react

to the aggregate outcome those agents together create this is essentially talking about complexity and not just Adam Smith but you find similar quotes from people like John Stuart Mill Karl Marx like people who the Academy really considers like the essential academic thinkers we're talking about this so I think it's a perfect example of like a pretty non subversive economic idea that really should be a part of how people view markets and model those markets but because of how rigid academic economics can be it still gets pushed off to the side and it's still sort of a niche interest so now we'll talk about yeah why that's important to defy so these

limitations like the entrenched ideologies that prevents new ideas from becoming part of the mainstream they just don't exist in defy and I do want to mention that that's not necessarily specific to defy it's not specific to aetherium what it is specific to is a new ground-up way of thinking about economics and that's something that I think is really important that's happening here so I will tell you why that's important and I'll also tell you why it's or how it's actually happening like in real life right now so as I mentioned before I make our governance does not assume that crypto prices are going to be reaching any sort of equilibrium at any point we don't assume

that people who have CDP's are acting rationally and when they see the rate go up we're going to close at a certain point we don't assume that our CDP holders are even aware that these rate changes are happening necessarily we don't assume that the mediums or the media that we publish these changes on are reaching everybody and this leads to tighter iterations on governance and it also leads to more adaptive approaches to setting these rates and I think that's really important because at the very least the last seniors giving us reason to think critically about how we might improve the system the way that monetary policy is set and the way that we depend on you know a few individuals to set that policy and

omnipotent sort of sorts of ways and I'm not trying to like be a like obnoxiously skeptical of central banks or anything but I do think that it is relevant to maybe take a new approach try something out this is a fairly contained experiment and I think it's very it's uninterrupted a feedback loop between markets and people who are governing those markets and critically you see that's really not the case when you can win I for guidance like notes are dropped that maybe rates are gonna raise eventually in the future it causes all

sorts of infighting among people in financial markets president fighting in political environments not to and it really at every step of the way interrupts the feedback between markets and the monetary policy that's governing those markets and if we review the economy as a complex system the healthiest way it can possibly develop is organically which means and I interrupted feedback loop that's allowed to it's so organized in whatever way it might and I'll explain why that's important here so right now we see defy sort of traditional financial as and some institutions being sondik reported over on the blockchain but if

this is a complex system and if we do see self-organizing structures emerge from it I would say what we can expect is to see more and more truly novel things coming out of it but novel institution is not our instruments and other ways of doing things they aren't just copying the way that financial markets work today and I think that's really important because the more distinct economies and the more you need to be tons the more of an effective hedge it becomes on the inherent financial risk in the traditional system and to that end to me if we create a distinct option which both retains the standard crypto like value adds like

Financial Inclusion and also becomes a workable financial market with accessible credit that can set its rates appropriately with tighter iterations and adapts to new technologies effectively we don't really need this to be clipped traditional finance right in fact it might be kind of an apocalyptic situation like that all came crashing yeah so what I really want to stress here is that forty five to be successful it doesn't need to take over the financial world it just needs to keep developing as it is and we need to keep letting whatever comes out of it come out of it which doesn't mean we should be cavalier about governance or anything but it doesn't mean that you know maybe

we're setting you bar a little high when we were thinking about like the entire financial world in the short term and the reason that's important is I think there's a lot of really incredible things that we're doing already right like maybe there's not a lot of people in Argentina that can use dye but there are some and if some people get some value out of it in the stressful time it's already really good thing right so maybe over the next few years more and more people can use it and that's really good I think so that's just a weird screenshot new approaches to finance might be the only way really to incorporate new ideas and economic thinking into real monetary policy and into general practice as I

mentioned before that's not specific to t5 it's specific to a new approach in general but I think the most promising new approaches to finance are happening here in d5 but I also think that objectively an appreciation of complexity has to be a better and more adaptive approaches to monetary policy I guess you thought that in the case of restoring the dye peg and as novel economic structures financial institutions and financial instruments emerge Defiler continue to be better what is it before I'm very interested in

talking about these things like I want to so feel free to start a conversation with you about any of that stuff reach out if you'd like to I that's my twitter handle I just stupidly joined Twitter like two weeks ago I'm really behind but so that's my Twitter and you can email me too if you're late to the party like I was yeah I don't know do I have any time for questions or I think we can give Ethan a round of applause to start with [Applause] I counted 23 things we've got time for one question if there's someone who did

very good hand up Chris Powers just interested in defy stop really interesting talk I really think it's cool to be bringing these ideas into this space and I think there's a lot of just natural experimentation that gets to happen in defy and play these out I guess like starting with the beginning when you're talking about complexity theory and basically the failure of economic models it kind of seemed like then you're basically trying to model irrational behavior which kind of you know there's some doesn't only make sense and I guess when I think of like how investors react in the world or users react in the world it's not so

much like oh we are acting rationally to some weird some weird event or oh there's some like upper bound or lower bound oftentimes those decisions are based on this thing called judgment right and I don't know like I guess is a fool's errand to try to model judgment in a lot of ways because judgment is like our gut feel right it's these things that we know like aren't rational but they do affect us and so I just kind of tiers to how you kind of square that and whether yeah yeah that's a great question and actually it's a pretty common criticism of both the complexity and especially emergence to the extent

that I had like a funny sound effect I thought about playing every time I said the word emergence cause it's like magical people kind of treat it like it solves everything but I think it's mostly important as a counterbalance against full trust in models that assume that rationality so we might not be able to say this like judgments a great word for we might not be able to say that a person's judgment is going to be here or there on a particular day but what we can start to do potentially is starts to figure out like what the bounds like what what reasonable variance we can set on those models too so temper our expectations that we get from those models does that answer your

question right and I think even from a less quantitative perspective the reason why maker governance is a good example I think is because we have calls every single week to talk about raising the fee or keeping the feasts table and basically none of these things are taken for granted there so you end up you hear a lot of conversations about like okay if everybody is rational then this rate should deter people from opening CDP's maybe cause people to close them but we don't really know so let's raise it and let's come back next week and let's see what happened and I think that that's a much healthier approach