One Weird Trick to Good Tokenomics with Nihar Shah

awesome thank you okay let's get started on the next talk one weird trick to good tokenomics so my name is nihar Shah I'm a researcher at Jump crypto and this is my slightly tongue-in-cheek title and presentation on tokenomics now every every good crypto talk has to has to start with some sort of Meme and here's my obligatory contribution to that um but this one in this one I actually argue it's a little less tongue-in-cheek than it may look that indeed tokenomics may seem very very complicated but over the course of the next 20 minutes I hope to convince you that in fact there is some basic Simplicity to it that can you that you can exploit when designing your protocols so the framework that I want to present to you is this one around a North star I'll begin by sort of setting the stage and acknowledging that questions around token design are indeed very complicated they are very bespoke they can be nuanced and often when you're trying to design a protocol that has some interesting technology and apply a layer of tokenomics on top of it the tokonomics and sometimes can be much more complicated than the underlying technology but the argument I'm going to pitch in today's talk is that all of these questions become a lot easier if you have what I call a North star some guiding principle some basic Foundation or structure that allows you to take any of these questions complicated bespoke crazy in their own rights and you can actually make some Headway on it and so what should be our North Star when we address the question of tokenomics to present that you'll have to engage me in a slightly tortured historical analogy by which prior to the Advent of a modern sailing equipment Sailors would look up into the sky and they would try and find the North Star and use that to orient their maps in their compass and then find home accordingly and the North Star for those of you who are astronomers will know it's the star called Polaris but so the lore goes there were these ancient Sailors who sometimes are rather these inexperienced Sailors who would sometimes get tricked by a much brighter star a star called Vega that's slightly off North and in doing so they would then be waylaid and never reach their destination and so that's the analogy I want to use when I think about tokonomics too that the North Star should be this this thing that I'm going to come back to time and again incentivize Civic Behavior sustainably too often when I talk to folks who are in this space their North Star is this Vega this incorrect one the one that burns a little brighter shines a little brighter but really way lays them from their end goal and it's that they're always thinking about tokenomics in the context of say Revenue plans transaction flows it's this very operational mechanical thing that's inexorably linked with perhaps the business development side and I don't think that's how we should be thinking about tokenomics by contrast there may be slightly more subtle North Star but the one that is I think important for the protocol's long-term success is this one incentivizing Civic behavior sustainably that is I'm going to give some examples about how a token should be a coordination mechanism you shouldn't think about tokens as a business strategy you should think about it as a way to power social consensus that it's a carrot and from time to time a stick used to get a decentralized group of people into actually making Headway in some common shared setting so let's begin with a little bit of Economics 101 in economic theory people always want to maximize their own utility but as we know sometimes these actions call what economists call external sorry they cause what economists call externalities externalities are cases where someone's actions impose costs or benefits on unrelated uh third parties and so these are external to the person generating the action often the canonical example is pollution if I set up a factory and I say convert raw materials into widgets and I Pluto out into the air that's great for me but the pollution will annoy and bother my neighbors that creates external harms on my neighbors that I don't necessarily take into account unless say a government forces me to do so and so that's one classic example of a negative externality externalities can also be good I think the canonical one here is open source software if I write say a useful piece of code and I put it on the internet and tens of thousands of people downloaded and they also get use out of it that's great for them again I don't generate or I don't benefit from any of those actions except perhaps in very indirect manners and so those benefits are again external to me and so externalities are a classic market failure and they're typically characterized by either diffuse benefits and very concentrated costs in the case of the open source software example or very concentrated benefits and very diffuse costs in the case of pollution externalities exist all over the world around us I can come up with many examples in the real world and governments have a role to play in the real world on trying to enforce and mitigate these externalities trying to unlock the positive ones and shut down the negative ones but in crypto there is no such uh there is no such Central Authority and so that's where tokens can come in tokens Can Be the mechanism that manages those externalities in crypto let's maybe make it a little more concrete um let's go and let's talk about gas fees which we all sort of take as given and we have some basic understanding of why they work and why they're useful but now let's actually apply it in this language and see how how um how much Headway we make so to begin let's imagine a really really simple example imagine you've created a layer one you've got no malicious actors everyone on it is well intentioned but you also decide you'll have no gas fees well what ends up happening we all sort of know how this one plays out protocols and participants face uh no penalty for being wasteful in their compute and so the network just becomes unusable the protocol protocols and participants impose too much of a computational burden then other people can't get their transactions or their computational needs through um and this all sort of you know breaks down so if we didn't want to go into a token yet we wanted to just try and solve this problem what would be some examples of non-token solutions well the two I postulate here are well maybe we want to block gas inefficient users if there's a user who time and again doesn't seem to respect the compute constraints of the network we try and cut them out but of course what will happen in crypto they'll probably just create a new address and get back on the network okay so that doesn't work so well uh maybe there's a second one we try and block gas inefficient transactions if a transaction looks like it takes up a lot of gas let's try and shut that down but in fact this is a really hard technical problem it Bears some similarity to the halting problem we don't really know what gas inefficient transactions exist until we actually simulate them and at that point we're spending the compute anyways and this problem gets much harder when you think about it in the context of the real world let's now add back in malicious actors who for no other reason than their own trolliness want to try and shut down your network now what happens now what are the as we know these actors will submit a bunch of spammy transactions it'll overwhelm the network what are the non-token solutions here well now they start to get even more complicated maybe you try and maintain a black list of addresses well we already covered why a blacklist probably doesn't work so well in crypto maybe you start to get some anti-ddos technology this can work a lot of web 2 companies use this but now this starts to introduce uh non-trivial censorship risks maybe you say Okay gas fees are the answer but let's at least just charge them in dollars this starts to get close to what I think the most elegant solution is but this starts to also introduce an extra an external dependency on a different asset now you have to connect to Dollars you have to charge in dollars you have to do all of that and you all have to rely on say the banking system or some other intermediary to facilitate it and so we know what in fact the best solution is it is gas fees in the native token and so what I want you to do is take a minute and try and understand it in the language of externalities of a mechanism that gas fees aren't just some Revenue plan for making miners money they are a mechanism for ensuring that the network can function in an efficient and judicious manner now those who impose a negative externality the people submitting transactions pay for it and those who bear the negative externality the miners or validators who have to do all the compute they receive these participants are now incentivized to have good behavior to be gas efficient to be judicious spam Bots are now cut out of the network if they really want to spam and overwhelming Network well they have to pay for the privilege they're really no additional censorship vulnerabilities no uh dependencies on external assets this just becomes a much more elegant solution and as a pure bonus that isn't really related to these main points variable gas fees also allow us to coordinate on a relative importance of a bunch of waiting transactions let's go through maybe one more example let's think about security let's imagine a world where anyone can be a validator that casts one vote for consensus but what happens here what's the outcome what's the equilibrium outcome what's the externality as we all know security is going to be this diffuse uh diffuse benefit concentrated cost problem it's going to be under provided that even if there are altruistic validators even if I take on the hard work of setting up my own machine and trying to parse good transactions from Bad transactions it's not really going to make the network very functional because although I'm providing this diffused benefit I'm paying all the computational costs and my diffuse benefit does very little because spammy or malicious validators can completely overwhelm me again I'm not going to belabor and think about some non-token solutions but if you were really to think about those you probably get somewhere with identity perhaps validators have to show their driver's licenses or passports and you'll have to get some sort of dependency on that in order to get around this problem or of course we can use tokens and so that's the solution that again seems to be much more elegant here that validator stake tokens that broadly go up or get compounded if they perform good technical behavior and broadly go down or get slashed if they perform bad uh and malicious behavior so now there are incentives to provide security Now the network does become usable I will add this one Coda that gas fees I think are a really great example of a mechanism operationalized by tokens security is a pretty good example of a mechanism operationalized by tokens but it's not perfect in particular this does not necessarily incentivize minor diversity as we all know a few networks struggle with the problem that too many or sorry too few validators or too few miners can control too much of the compute capacity and this solution doesn't really get around that um and then the other one is that sometimes this links security to Market sentiment if the price of a token has unrelated fluctuations to the core use of them the network that can then either enhance the security which is could but also weaken it at the same time and so this is a pretty good solution but there is some room for improvement okay so enough of talking of the mechanisms we already know why don't we talk about some new ones well before we do that what I'm going to try and argue is that there's a general recipe or a cookbook that I at least I try to use when I approach these problems when I want to incentivize Civic Behavior sustainably what I'll try and do is these four steps the first step is to try and actually figure out what are the important externalities what are the positive ones that you want to unlock what are the negative ones you want to shut down this is almost always the hardest thing uh second try and find who are the people who are the actors who actually generate these who are the agents that we want to either incentivize or tax accordingly third we need to set up mechanisms to you know operationalize some sort of transfer to and from these agents maybe we reward them when they do well maybe we tax them when they do poorly maybe we move tokens around accordingly and then finally calibrate the mechanism's generosity and sustainably of a sustainability appropriately what we often find in good token design is that this fourth step undoes them that there are many examples of say interesting token designs that try and get at some core externality maybe liquidity uh mining or maybe uh airdrop stuff like that but then protocol designers always get the sustainability wrong they make the pro the rewards far too generous they never shut off the Taps and that compromises the networks or the protocols long-term ability to actually um you know do other sorts of interesting things with their community so here's the example I want to work through let's say we wanted to think about governance proposals we always want a healthy decentralized robust uh governance Community when we build a protocol what what would be the externality well who are the actors what's the mechanism that we might set up um maybe I'll pause for about 15-20 seconds take a minute to think about it and then we'll go through some of the answers okay let's uh let's give it a shot here is at least what I would uh postulate the core externality what's the real problem proposals they take a lot of effort they take a lot of effort for someone to sit down and write that person has to understand the protocol very very well um and they generate a lot of benefits but these are very large and diffuse benefits if I only I'm a small community or stakeholder in a protocol I'm not going to do this I'm not going to spend say an entire weekend trying to understand the protocol and bring over a well-crafted governance proposal if my stake in the protocol is so small that it doesn't really matter and so in fact we tend to see that governance proposals are almost always written by very large contributors that's fine but that seems to miss out on a lot of intellectual capital in the community that could help out so who are the actors well anyone can write a proposal but of course as those of you who have read governance proposals know there's a lot of heterogeneity in their quality there are some that are really really sharp and Crisp And there are others that seem maybe a little bit lazy and Fork from a different protocol and so given this as some groundwork maybe what's the mechanism that we want to implement well the one I sort of suggest in this stylized example is that we should generally reward proposal writing if people write proposals they take the hard work of actually spending their weekends doing it we should reward them with some stake in the protocol we should give them some token allocation now of course what will that do without any further checks might incentivize a lot of spammy proposals people writing really lazy shoddy bad quality ones and so the second part of that might be we only distribute this reward if a proposal Garners some critical mass if enough of the protocol believes that it actually was a valuable contribution to the intellectual Enterprise but then there's even some additional layers of nuances on top of that if we for instance only do that we only reward protocols that say pass what's going to happen well people are going to be incentivized to do all the slam dunks all the really easy low-hanging fruit proposals that are certain to pass and certain to earn them rewards but we don't want that we also want protocols to take some amount of risk we want them to stretch out their intellectual boundaries and Frontiers which means we need people to propose ideas that are sometimes really crazy and may not pass but at least put the community through a healthy debate and so perhaps the third layer of nuance is we give rewards to protocols that hit Quorum but not necessarily ones that just pass and then finally we want to really find those folks who seem to have a knack for this who really understand the protocol through and through and can write proposals again and again in support of it so maybe we even offer some sort of compounding of rewards for those uh for those authors that for the author that seems to have some real knowledge about the protocol we want to keep incentivizing them to come back time and again and so maybe we offer some increasing reward mechanism and then finally well what's the sustainability of this well I think here maybe it's a little easier at least for proposals perhaps the protocol says we'll use some portion of our treasury to fund this but I think the important thing here is to understand what is the long-term plan is there some long-term transition plan will at some point the protocol turn off the Rewards or are they comfortable with implicitly diluting all of their other community and stakeholders down the road and that's a decision that a protocol uh Community has to make up front and that's uh pretty essential for the sustainability so that's uh that's all I have for today um I'll sort of hang around for a couple of minutes if folks want to want to reach out and chat with me but I want to come back to this core Point your North Star when you address any of these questions from airdrops to liquidity mining to security to governance um and whatnot is this you should think about Civic behavior these externalities and try and incentivize them and incentivize them in sustainable manners maybe if that's a little complicated or a bit of a mouthful let's try a simpler question just ask yourself this question we want more of X behavior and less of why Behavior how do we get tokens away from the whys and towards the X's in short tokenomics should not be a business strategy it should be first and foremost a mechanism first and foremost a way of powering social consensus and this allows you to reduce nuanced questions into the same core strategy so I have here finally a little link um a co-author and I a jump have written an article about this we go through a bunch of other examples I didn't uh I didn't sort of spoil much of it in this talk so if you want to read up more please take a look at that and feel free to reach out if you have any questions thanks [Music]