Fernando Martinelli: Balancer – The Automated Market Maker Protocol for Liquidity #350

Transcript

this is epicenter episode 350 with guest fernando martinelli [Music] hi i'm sebasti with you and you're listening to epicenter the podcast where we interview crypto founders builders and thought leaders on this show we dive deep to learn how things work at a technical level and we fly high to understand visionary concepts and long-term trends if you like epicenter the best way to support us is to leave a review on apple podcast if you're on a mac or ios device the easiest way to do that is to go to epicenter dot rocks

slash apple today our guest is fernando martinelli he's the co-founder and ceo of balancer labs balancer is a generalized automated market maker protocol or amm it allows anyone to place liquidity into a pool and to put it to use so on one side we have liquidity providers who generally seek to balance their holdings and they can do that by placing it in an amm protocol like balancer and they'll get rewarded with trading fees and on the other side we have traders who are just looking for the best rate possible one way to look at balancer is as a generalization of unit swap however balancer pools aren't restricted to the same 2020 split between two tokens a balancer pool can support

up to eight tokens with any distribution for example one could create a pool with 70 percent wrapped eth 20 die and 10 link and balancer uses smart order routing which ensures trades gets sent to the pools which provide the best rate possible so essentially balancer pools can be seen as a sort of self-balancing index fund which pays you for contributing liquidity to the platform so the inner workings of balancer are quite complex so thankfully we have mayor and sunny on this one to break things down and help us get a better understanding of how the protocol works and also the token economics which uh allow it to function and the governance of the protocol enabled by battle tokens this week the interview debrief is free

for all subscribers and for everyone to hear and sunny and mer go on for an extra 15 minutes and they talk about token economics and the business model of balancer if you like the interview debrief you'll want to check out epicenter premium to get these every week as a premium subscriber you'll get access to a private rss feed where you can hear the debrief after every episode you'll also get enhanced features like episode transcripts and interview chapters that allow you to easily skip to specific sections of the interview you'll also get access to roundtables discussions and bonus content that we put out from time to time and you can sign up for representative premium at premium.epicenter.tv and with that here's our interview with fernando martinelli so today we have on with us fernando

martinelli who is the ceo of balancer labs where they are building the balancer protocol it's great to have you on fernando great to be here thanks for inviting me guys yeah so me and me here have been like you know following the balancer work for a long time you know i think my hero was very interested in sort of like multi-asset pools i've also been really interested for a while in just like you know more advanced parameterization of uniswap or amms because like i have this like giant blog post i wrote like a couple months ago where i'm like oh i want to parametrize everything in amms and then someone's like oh you should go check out what balancer is doing and it's like oh wow this is really cool yeah happy to finally like you know be

able to chat with you about this stuff so can you tell us a little bit about your background and how you got into crypto in the first place what were you working on before balancer if anything so i'm a mechatronics engineer and i did a masters in robotics and image processing i also did an mba at the cerban university in paris worked for bain company as a strategy consultant in germany i actually created a few startups already in the past so i started very young when i was 14 i created this cyber cafe that was very very successful because it was the time where people were addicted to counter-strike and i was myself too i figured i should create the my own cyber cafe

i went to europe and after i worked for bain company i created this company that allows people to prepare for interviews called prep lounge i got back to brazil and created an energy drink company like quite different and how i got involved with crypto in the first place was late 2012 i got introduced to it by a friend and at first i just like dismissed it completely like this is a ponzi scheme you're only going to buy this thing if you believe that others will buy it and the value will grow just because like people are buying into it and then second time i looked at it early 2013 it really struck me like how like revolutionary this is and i kept kind of uh paying close attention and when ethereum came out it really like kind of uh caught my

attention because i think that bitcoin is very is great for like a base um money layer but ethereum has this flexibility that allows you to create the financial system instead just of a coin right i was very excited with stable coins that was something that i really believe we need to get mass adoption since then i still do and ethereum allowed that and that got me like to be participating in the early days of the maker dial community this is where i got to meet roon and nikolai and all the guys in early 2016. i collaborated a little bit with maker dao doing some research on the control mechanisms for the target rate feedback mechanism that

didn't get applied and now there's like reflexer that's actually reviving all those control ideas with amine and stefan so yeah that's kind of an overview of how i got to here and so a lot of the like early balancer team is from like the maker community right to like nikolai and like um mike mcdonald as well and was there anything in particular like there was a problem that you guys faced in like designing maker and you're like okay i need to go build this balancer is gonna be the solution to some problem we were facing there it wasn't a problem we were trying to solve for maker it was so balancer started as a research project at block science which is an

engineering research and development analytics firm and that was kind of me alone collaborating with them and uh yeah the reason why balancers started in the first place was like this idea of how can i keep a portfolio uh stable in terms of uh percentage of value distributed like you want to be positioning trading and you don't want to have like to actively go and buy and sell tokens just because one of them went up by a lot so you want to have this kind of peace of mind of being passively investing in a basket and and having it being automatically rebalanced for you that was the the beginning of everything and yeah we we spent off of block science and created bouncer as its own company

and i showed this to nikolai he was extremely excited because he had this problem i had the problem myself too like how can i put my tokens provide liquidity with them if uh back then only there was only uniswap and you can just like put 5015 nikolai as being one of the maker dow founders he has a lot more mkr at least i suppose than than eth so it would be great for him to have an 80 20 pool right with mkr and eth and he really loved the idea he then kind of joined as our chief architect and tech advisor wrote a lot of the code and then mike mcdonald joined as a cto and co-founder as well and yeah this is kind of how we got to here

so if the original goal was to figure out how to like do passive portfolio management something more similar to like set protocol i would say then what point did it transition to then becoming like also liquidity providing it was a consequence like when you want that to happen then you need someone to do the trading for you right so the idea is that it's a two-sided market where you have like the right prices in place for people to be incentivized rational players to be incentivized to be doing the trades you want so if your pool needs to sell a for b you want to let people be marginally incentivized to buy b4a or yeah do the opposite trade it then creates this massive amount of

liquidity where like if everyone's using this common protocol which is balancer they're all like creating pools and adding tokens to them that creates this massive liquidity that uh makes it very easy and gives great conditions for people to train i'm really curious about the generalization of uniswap right so you have unisop which is this constant product market maker in which there are two assets x and y and it's a pool which is 50 50 and then in some ways balancer extends that formula to it introduces more terms and it introduces the exponents in these terms and it's a generalization of uniswap who had the brainwave of generalizing unisport like that

and how did that idea come about it was me so i came up with that formula back in early 2018 and since then it was a long like process of modeling and simulating things and lots of things that formula that aha moment came while sleeping i had like tried dozens and dozens of things and uh yeah written a hundred maybe pages of uh mathematical scripts because one thing is to get that simple and kind of uh small or very simple formula the other thing is to prove that that formula like if you take partial derivatives of two tokens and you prove that your pool is going to always have the same value in each of the tokens so it seems

very simple but it was a lot of time that i spent uh yeah working on that i i also did uh during my undergraduate studies i did uh advanced mathematics at the university so that helped quite a bit yeah it was pretty much myself maybe before we dive deeper into the math or could you maybe give a broad overview of what is balancer and what does the protocol do balancer is an amm protocol for programmable liquidity and amm there stands for automated market making it's a complicated term but it's nothing but the protocol trades uh automatically that means that balancer allows for the creation of flexible

liquidity pools that are continuously self-rebalanced so the protocol makes sure that those pools are rebalanced and that's what we just spoke about the nice thing about those pools and if you look at unit swap you see that it also does that if you go to contracts uh that unit swap has you're gonna see that the if value and the token value so if you're talking about unisoft v1 you can you can see the evaluation token value you're going to see that those values are almost the same unit swap always keeps that those pools at a 50 50 percentage distribution of value now balancer allows you to do that with many tokens not only two and not only like the same weights so uh you can have twenty percent token a twenty percent

token b and sixty percent token c you can have up to eight tokens and you can also choose the trading fee that you want your pool to have and that's fixed also for uniswap at 0.3 so you can think of a balancer pool as an index fund where instead of the lps paying a fee for the administrators to manage the fund and sell the tokens that go up which is what happens in conventional finance it actually pays the lps for providing liquidity to those funds so it kind of inverts that idea of i'm paying for an index fund manager to take care of my funds and rebalance them for me it's actually i'm getting paid for letting people use

my liquidity in that index fund to trade so yeah maybe kind of finalizing the answer you asked about like who are users what's in in it for them there's two types of users it's a two-sided market on the one hand you have the liquidity providers who are looking to both balance their holdings so they don't want to be overexposed to one of the tokens that they hold so they want to be selling those tokens as those tokens go up and the other uh use case is you're making trading fees right as people sell buy sell buy you're charging a small fee out of each of those trades so you're accumulating more and more funds so it's a way in which you can generate some interest or some some trading fee revenue the other side is you have uh traders

and they want to trade a for b they don't care where that's coming from they just want to have the best rates or or conditions possible and balancer can use tap on the liquidity of all the pools to offer someone a trade so one way to think of balancer is that balancer takes two different parts of the financial supply chain or the two different parts in the traditional financial supply chain and it kind of merges them into one so on the one side you have the asset management supply chain which is me shipping some dollars into fidelity and then managing the index fund for them uh for me and this fidelity has you know like trillions of dollars of assets

on its balance sheet and then on the other side you have the market makers which are operating on the different exchanges the new york stock exchange all of these exchanges and these market makers are people that have the assets and are always willing to quote prices on on the assets so they have us dollar and british pounds and they are willing to always quote an exchange rate between them the nature of a market make a traditional market maker is that they need access to assets so that they can make the market and the nature of an asset manager is that they have lots of assets which have been sourced from all of the people that own the in index token and normally you know like these two are quite siloed these two

functions of finance are done by different entities one of these entities has asset and the other needs assets but the twain never meet in traditional finance but balancer is almost the system where it's a smart contract where both these parts of the supply chain exist inside one smart contract so you have users that want an index supplying assets and those assets being used to make a market and the market maker is the smart contract itself so it's like it's compressing down the entire financial supply chain into set of smart contracts and allowing for a new kind of efficiency to emerge out because of that compression

i couldn't have done a better job at uh kind of doing that analogy than you did that that was uh extremely good and i totally agree with that the idea is that balancer using ethereum which is an amazing protocol an amazing revolution for society it just like spreads the fees that are now going to fidelity to nasdaq nyse to all those regulated entities that have to spend a lot of money being regulated and paying lots of salaries this is why you have to pay an index fund manager uh for them to to manage your funds for you the beauty of ethereum and yeah decentralized finance is that you don't need anyone for that because you have the smart contracts

controlling that and it's totally trustless and permissionless so yeah we're just combining those two sides of the market and yeah spreading the fees across them so no one's is getting in the middle anymore and and this is not only cheaper for everyone it's more efficient and faster and has a lot of advantages so one of the questions i have though here is i understand why it does well as a trading venue but when you say it acts as a fund like portfolio management you know when you do portfolio management you usually want to rebalance occasionally right so in your portfolio you have

like 80 bonds 20 stocks and then suddenly the stocks 2x or 3x and now you have way more stocks and then at some point you like rebalance your you know you sell some stocks to equal out but the problem is you know i feel like the portfolio management the profit works because it happens at like sort of longer intervals where you know you let the stocks rise for a week and then you rebalance but i feel like the problem with doing it on balancer is the rebalancing is almost it's like it's too over eager where it rebalances too quickly it's a continuous rebalancing and so

you're rebalancing before you even get to realize the profits of the stocks rising so for that reason is it really act as a good portfolio management tool in that case that's a great question sonny so actually there is a way to simulate what exactly what you're saying like wait for there to be some profit for one of the tokens to go up by a given amount before you start selling that and i wrote an article on that which is how to use balancer pools with high fees to simulate swing trading so when you have a high fee let's say 10 your pool is only going to sell or it's only going to be profitable so let's say ethereum is 200 your pool with a 10 fee

is selling at 220 and buying at around 180 right so what happens is the price is stuck until someone trades and if the external market price goes from 200 all the way to 221 that's when someone's gonna uh buy from you so you're gonna be selling at 220 and then if the price goes down and whenever it's between though that in that bandwidth like between 180 and 220 you're not selling nor buying because it makes no sense for external rational actors to pay uh more or less than the current market price so if it falls down to 180 then actually you start buying eth so this acts exactly as you're saying so

that the fact that you can customize the fees allows you to choose like what strategy you want so if you want to be rebalancing very often that would be a zero fee or if you want to rebalance like very sporadically when prices have huge variations that's a higher fee so you can control that with the fee which is pretty cool right yeah that's really interesting you never thought about that because we're used to the 0.3 percent on uniswap which is fixed right so that's why people i think never thought about that but that that's actually pretty powerful and pretty cool in my opinion is it possible to make these much more resilient where like you know sometimes the fee you know you want it to be higher for some reasons when maybe there's higher volatility or maybe

you want to be lower when there's lower volatility or how dynamic is are these fees or is it like you know you set a fee one time or it's like okay we're doing 10 for this pool and it's stuck there great question i think in order for me to be able to answer that i'd have to kind of explain a few things uh real quick we have two types of balancer pools today in balancer v1 which are the private pools and the shared pools so the first ones the private pools are owned by an address and only that address can add liquidity to those pools and because there's only a liquidity from that address that addressed the owner of the pool can do whatever they want so they can do it exactly what you're saying they can start with a high

fee add more funds and then change weights add new tokens remove tokens and then decrease the fees they can do whatever they want because it's their pool now we also have the the equivalent to uniswap markets which are the sharepools which are immutable so you cannot change anything and that's necessary because other people can put money into that pool as well so if you allow the creator of that pool to change and add a new token they would be able to just add their token that they own 100 of the supply and drain the funds of everyone else so that's why shared pools have to be immutable the nice thing about balancer is that you can have a private pool that's controlled and that's actually composability on ethereum it can be

controlled by or owned by a smart contract not by an address so that smart contract is what we call or that whole kind of structure is what we call smart pools so smart pools are actually private pools at the base layer like they're seen as private pools by the protocol but the owner of those private pools are smart contracts that can be gateways for external liquidity and one of the like the use cases that i think is the coolest like one of the coolest ones is exactly what you said is like imagine a surge pricing pool or search pricing fee pool when there's a lot of demand for liquidity think black thursday everyone like is desperate to close their vaults because they're going to get liquidated

they don't care if they're paying 0.351 or 2 percent right they just need liquidity so this is when that smart contract that controls and owns that pool says okay it's high volatility time i'll just like crank up my fees and i'm gonna make this bowl more profitable and then what that means is that it's like uber when it's raining like no one wants to drive and everyone wants to take the cab so what happens is prices go up so you attract more drivers to the streets and that's exactly what happens by increasing the fee of that pool in times of high volatility you attract more liquidity providers and that in turn makes that pool better for traders because the slippage is going to be

lower because the more liquidy you have the lower the slippage so you actually are matching both sides supply and demand which makes that smart pool a lot more efficient for both sides so i think that in the future we're going to see a lot more smart pools and we're working a lot on that and other teams as well as opposed to just shared pools that are immutable we're going to see a lot more of those smart pools the difficulty with a shared like with a public pool is that a public pool is probably good for trading but as a portfolio management solution a public pool is not so good because like once you set a particular set of assets and certain percentages for that assets that thing is kind of frozen forever

and can't adapt to changing circumstances that's not an ideal portfolio for anybody it is frozen but you can withdraw your liquidity at any moment so there's no lock-up period of liquidity so if you find that basket or like that the fees of that pool are not ideal then you can just move migrate your liquidity to the pool you like more and we've been seeing that a lot happening on balancer lately so pools that were big at the beginning now people realized okay the fee here was not ideal so let's move to a pool with a higher fee so even though the pool is static your liquidity is not you can migrate your liquidity anytime and we're also working with other teams and inside

balancer as well to create like those intelligent migrations or migrators if we detect that there's a better pool for exactly or a similar distribution or basket that you already have that tool with one click you can just migrate your your liquidity to a better pool that's more profitable or uh what not but it does require active management then yes if you're talking about the shared post but not if you're talking about a smartphone which is uh the search pricing so those smartpools can also change the weights can also change the yeah add new tokens and there's a very nice example of uh that i like to give about a few actually

but ryoti is this company that tokenizes real estate on ethereum they are building a smart pool that does exactly that like they are constantly adding new properties to realty so instead of like having everyone migrating to a new shared pool with an extra property what they do is they create that smart pool where they are the controllers of the assets they can add new properties they can remove old properties and if they have more than eight properties so balancer pool is limited to eight tokens today in version one what they can do is they can replace a property by a pool of properties so now instead of having just eight properties you can have eight different cities

each city is actually a pool that contains properties in that city so you have florida detroit and et cetera the nice thing about that is that your smart pool token so actually also smart pools are tokenized so you own a part of that smart pool that token does not change even though the like all the properties are being updated and changing the weights you can use that smart pool like the the mother token you can use it as a maker dial collateral because that token is canonical it's not going to change so that's really powerful of course you have to trust the guys at realty if you're already trusting them to hold those properties on your behalf like tokenizing those you're actually already trusting them anyways so smart pool if they want to do like

they would have to like implement their own tokenization to distribute shares because you know the balancer protocol just sees it as a private address so they have to sort of do their own sort of shares however they want and like deal with like how to you know when people add liquidity issuing shares and all of that smart pools will have that part so the controller of that private pool issues shares erc20 shares for people who provide liquidity to that controller the controller then passes that liquidity down to the pool it controls but then the controller itself issues erc20 shares for people who are providing liquidity we're working in almost launching and auditing right now

uh smart pool factory it's like a template that anyone can create their own smart pool and they can actually decide what rights that smartpool that they're creating will have i can choose to create a smart pool that can only change the swap fee i can choose a smartphone that can add tokens but not change the swap fee so you have all those like levers you can change or yeah your switches you can turn on and off and then you have that smartphone without having to deploy any code at all and all of those are erc20 compatible so the issue shares that can be used as in your c20 anywhere on ethereum i can sort of see two types of smart pools that or two broad categories

and there's like really a spectrum but there's one which is more sort of governance focused smart pools which are like you know you can imagine them basically being dows and similar to maker almost right you use governance to make decisions and then you on the other side you have you mentioned earlier reflector labs where it's no more automated like using smart pools and smart contracts to make better smarter algorithmic things that adjust the parameters and there's obviously a spectrum in the middle so which ones have more traction right now and which ones do you personally see as being gonna be more popular going down the road right now we're not seeing a lot of smartphones out there it's like really the early days

it's very nascent and incipient but the guys at pai dao did an amazing job they've already come up with their own smart pool they they didn't wait for us which was awesome they have this uh dow controlled smart pool where the dow can decide okay let's change the weight of this token because now they have the btc plus plus the usd plus plus spice so now usd plus plus it has usdt or usdc usdc has some like trust issues so let's agree that we should decrease that weight and then they vote and then the dow decides that the controller of that private pool like a smartphone has to change uh reduce the weight of that stablecoin that is

having problems so for now that's pretty much the smart tools we have around in terms of what's more promising that's a great question i saw one thing that i think will be huge for bowser is going to be treasury management so if you have a project or company or whatever that has funds on chain you want to have like some sort of management where you don't want to be overexposed so for example an insurance protocol they have different types of assets that they will need to use in case there are some claims someone gets hacked they need to use that insurance fund that insurance fund itself has to be risk managed so that if one of the assets goes up by a lot you don't want to be holding

99 of all the the insurance funds in that asset so balancer is perfect for that it will be selling that token as it goes up and if it goes down and then it rebuys it so it's a great way to use a smart pool another thing that another way i really like is the concept of liquidity bootstrapping pools and that is an article we can maybe link in the show notes by mike mcdonald our cto he talks about how project that's uh starting and has a lot of tokens that they want to sell like an ico they can start with a smart pool that has 95 of that project token and five percent eath or die and then over the course of six months to a year they will start flipping

those weights so at the end of a year you're actually going to end up with 95 die and five percent your project token so slowly you allow people to poke that smart contract and update the weights according to that schedule of a year and what you're doing is you're selling in a year you're selling like 90 of your tokens and along the way you're letting people trade you're making your token very liquid which is one of the biggest problems for for projects right they they pay like um a lot of money for private companies to be market making on centralized exchanges they pay for a list being listed on binance they pay for those market makers to make sure that there's some liquidity people can buy and sell and now they can do all of that on balancer without

paying anything and actually that may be a segue to the next topic but they actually are getting bow tokens in return so they're getting some tokens that will allow them to have a voice or to to vote in the direction the future direction of the protocol so they only have like advantages in my opinion about this like uh i eeo or ido initial dex offering i guess you could call it so let's like take the example of the uma token launch so we just had uma on the podcast last week and so for them what happened was they were trying to sell off relatively small percentage of their tokens of their total supply which is only about

two percent on unisoft they have to do like a 50 50 balance it turns out they weren't able to put that much east in either and it just the problem was as people bought in the price was just fluctuating way too wildly like it just there wasn't enough liquidity to make it usable how do you fix this in balancer is there somehow i can use the weighting system of balancer to have done this better so your project wants to sell your project tokens and let's say in case of uma they didn't have a lot of eth so what they could have done on balancer is they could have put let's say 20 percent of the weight of that pool in eth and 80 percent in uma tokens so the overall pool would have a lot more value

in it which would have made it more liquid so the slippage for the same size of a trade would have gone down so it would be much better for traders now the thing that i like most about lbps so liquidity bootstrap imposed that idea is that what you could do is you start with a very high price to avoid this kind of uh bank run like everyone's trying to be the first ones to buy because they know that the price is starting as like the seed price and everyone knows that um is very successful is a great project so everyone wants to be the first to buy and just to sell to dump on top of uh the late kind of laggards so what you can do with balancer is you start with a very high price and then what you do is you start

decreasing the weight of your project token and increasing the weight of each if no one trades the effect that that has is you're decreasing the price of your token so let's say you start with 95.5 a very high price and then no one buys and it starts decreasing the weight at some point the price will become interesting for someone and then what you do is like you completely get rid of that hype or that those kind of opportunist people that just want to buy and resell because people know that if no one buys the price will fall so they will wait for the right moment for them to step in and if anyone wants to buy before that they can buy but they know that the price is high and it will keep falling if people don't just keep buying keep

buying the price will fall that's so cool it's basically like a combination of a dutch auction with an automated market maker into like one exactly that's so cool yeah so it's a dutch option because yeah anyone can wait and until the the the price goes down but it's an amm because if along the way people start buying the price will grow right yeah so that's pretty pretty cool going back to the topic of smart pools and the more algorithmic style ones to me it seems one of the things that would be the most useful in in a better smarter amm would be essentially how i see it is when you design

a fee model there's three parameters you can take into account which is volume slippage and volatility and currently uniswap only takes into account the volume it's 0.3 times volume the one that you'd really want to take into account though is volatility if you want you want to be able to say that like hey okay as the thing is getting more volatile automatically increase the fee can i do that on balance or is that easy like how you know unit swap v2 they started like sort of keeping this like uh historical records of prices and stuff and you know what how hard would it be to construct a

volatility index on balancer that's consumable by a smart contract that's a good question so i don't believe that we want to focus on that because that's a quite a complex problem uh in and of itself especially if you're one if you want to do that on chain then it becomes even harder you need an oracle you need some like like some history of prices in the past there are projects that are doing uh doing great work on that for example open they're building on top of balancer as well they have some some ideas around volatility and how you can get that information on chain today what uh what you can do with bouncer is what i said like you have a smart pool

that it has a controller that can just change the trading fee and it gets like some external source of data it could be off chain someone like really like a monkey watching the the monitor and sending the transaction which is increase fee decrease fee or it can be like something more complex uh totally unchained using opens information so yeah now now you do like you do need some external source of information to to use volatility but the cool thing is that you can right balancer pools are flexible you can change the swap fee according to liquid to volatility what what data is available from balancer that i can like query in from my own smart contract could i

you know for example thor chain they are doing you know along with volume they also take into account the spread of a you know how what this what the slippage of a trade is and they charge fees differently based off of what the basically how much liquidity is in that pool right now is it easy to sort of like ping the balancer contract for that information so i can have my my smartpool be take that into account yeah there there's ways how you can control that so what what you could do is you have uh a trade function in the controller so you say like now you cannot trade directly with my pool because my pool is kind of turned off

the controller can set public swap true or false and then what they can do is like it's false so you cannot trade directly at the base layer like protocol layer of balancer if you want but if you want to trade you can talk to me and i will in this atomic transaction i'll let let it be tradable and turn it on yeah i'll turn it on i'll set the fee according to the amount you want to trade i'll let you trade and then i'll turn it off again and you get you want to trade with the fee that i want so that that's really cool yeah so there's also a team that's working on that yeah so essentially like in in in balancer are like the assets you put in there they are customizable up to eight

the weights are customizable how much portion each asset gets the fees are customizable and we also discussed that when you make the fees customizable you also get behavior that the trading frequency can also drop if you increase the fees trading frequency it can be dropped so via fees you get customizability of trading fees what's also customizable is is the control of a balancer pool so it can be an individual it can be a firm it can be a dao etc what is customizable is the switching on and off of a pool and what else well at what point is this too confusing for users you know one of

the things that attracted people to uniswap was it's like you know here's i have some ethan dye i don't know what to do like i'm just gonna put it in this pool forget it but now it's like you know what if there's like different smart pools who all are like you know for the similar pairs and they're like now now there's so many decisions i have to make again and i have to like look okay who which smart pool is better at what point do we start to lose out on what made amm so attractive in the first place awesome question i i do think that we're we have we're very like opinionated there we want to be the base layer the lego like the money lego the primitive that people use

to build cool applications on top of so if you want something simple sunny because your users are used to uni-swap and you just want to throw like two tokens there you can just use balancer and create your own ui for balancer that does exactly that even better like balancer has this native function that allows you to provide liquidity with only one token so you if you have a pull of eight tokens you can just invest in that pool using one of your tokens one of the tokens in that pool so you don't need to have all the tokens and like kind of add them all together you can but you you don't need to uh and if you want to do like something more complex like uh we we saw lots of projects interesting projects building on top of balancer at hack money one of them is like my

defy pi you can build so many things you can go to uma and you can like start this synthetic and then you can put those synthetics into a balancer pool you can change the the the the fees and other weights so you it's the spectrum of complexity that's on the app layer and we restrain like ourselves or we restrict ourselves to being there a very good protocol that is flexible that allows people like to to get creative on top of so in some senses you know you know there's this aragon project and the aragon project came and said we are building a framework for dows other people are going to come and build

the actual dows and our framework is so extensible and generalizable that we can conceivably match in a lot of the needs of of users and balancers something like that except for these these pools you're building a framework that's extremely extensible and then other people will come and build the applications that go to the to the end users but but flowing out of that kind of design is the natural question of how does so how does the base layer extract value in this in this economy of many people designing customized pools and many other people contributing

assets and and so on so that's going to become an interesting question when we talk about the business models around balancing yeah so the the answer to that is it really it really is up to the governance to decide when there will be if there will be a protocol level fee that base layer is all is where all the liquidity sits and it's also like where traders tap onto uh for trade so it doesn't matter if you have like a very smart pool that changes your the swap fee when there's like more volatile times or if that pool is managing a treasury of a dow it really doesn't matter for a trader who wants to trade a 4b

it doesn't matter where it comes from the balancer ecosystem as long as it's in balancer it will be used to get the best trade possible and at that point the the protocol can decide that the governors that hold bowel tokens the governance tokens of the balance of protocol they can decide okay from now on we believe that we are adding enough value to uh to the ecosystem already people are really using balancer a lot so we we deserve like to to get a small fraction of that it's up to the governors to the token holders and that's a tricky decision because like everything's open source and it's um anyone can clone and start a copycat so if you charge too high of a fee people can just fork it and use a copy without any fees

if you charge uh yeah if you charge no fees then at the same time people might not be so incentivized to be uh to be paying attention to everything to be making form to decisions so it's like in democracy like you have to pay the governor's president and and and yeah at some point that could be the case where there is some fees they're going to uh to balato can hold us to to kind of incentivize them to make the educated right decisions uh for for the good of the prodigal yeah so i think maybe what's interesting here is to actually go into what the ballot token is unless sunny has a different question well what i wanted to say was like so you guys pitched this as

a governance token and this whole concept of a governance token you know xerox started like picking their token like years ago as oh the point of this token is a governance token would it be fair to say that the like true value of governance tokens is in its potential ability to add fees eventually that extract like value or do you think that there is some larger external value to governance tokens outside of that as well that's a very good question philosophical one i think if you're like a rational actor sunny the only point is like uh discounted cash flow of

future cash flows right so how much uh i can make with uh with this token that i'm holding in the future and there you have to factor in what's the percentage of chance that there will ever be a fee and then all those calculations for a rational actor will define how much the token is worth today and uh if you're like if you're like a a philanthropist or you you like defy and you just want to be like part of the decision-making process you might just want to pay for for about token because you want to like give your say and and and have a voice into how things unfold in the future so i think they're like yeah if you if you're thinking rationally then it's the probability of there being

a protocol level fee in the future so yeah yeah so as my hero is saying let's maybe talk about the bow token itself and so you know you guys have been very public in saying the goal is you want balance your labs as a company to disappear why is that the goal and how have you designed the battle token to achieve that goal so that that kind of uh is something that is in the in the dna of the team that built balancer we're like uh the purists that like it more like the early days of the maker dial uh developers we don't want to be like a centralized company that's controlling the protocol and getting

fees yeah we we want this to be like uh uh really a community common good and we want this to be self-sustainable and not to depend on on us we're like doing doing the the the dev work right now but we're seeing already a lot of people building tools on top of a balancer you have uh pools.vision predictions exchange so many nice tools that people are building and this is exactly the direction we want to go to in the near future uh say like four or five years video is that it will be just like ethereum or just like uh like linux it's it's a an open source software or platform that people are just incentivized to be building on top of and and improving

that that's kind of the idea we don't want to be centralizing any decisions or how this should evolve how have you designed the token what is the bad token and what does it do because as i understand it like at some level balancer is an is an idea it's it's this design of how to build pools then the another asset that you have is the open source solidity code which is so general and extensible that you can build various kinds of pools with it and then you have a token so how does the token exercise any influence over over the code or or the design and like what is the connection between these

these elements great question so the the the answer there is that we knew that it's hard to start with an on-chain governance system very complex with lots of parameters that governance have to decide it's it's just like not realistic to start that way so what we did instead was let's start with something that has no admin keys has no kill switches has no upgrade ability that's balanced if you want you you deploy a pool that pool will live for as long as ethereum lives that that was like our like objective let's be very conservative and there's no way you can game the system or control it there's no governance token for v1 now in the next versions not sure if v2

already but v3 before probably will start getting the bow token more intertwined in the code and getting more of unchained governance like spells like you have in uh with maker dalek you just propose this uh this new kind of feature on chain and then you cast a spell and people vote on it and if it's approved then it automatically gets implemented at the code level at the like on chain level now what we're doing is we're already using balancer tokens for people to signal to vote on chain but by signing messages only like carbon voting they're not sending transactions so by by doing that we can iterate very fast how how the

like the governance and the decisions take place for example the bowel liquidity mining uh program we're distributing tokens every week to people who provide liquidity on balancer we want to provide liquidity or give bio tokens for the liquidity providers that are providing the most useful liquidity what is the most useful liquidity that's a very subjective thing and instead of like doing like comp did and saying well we're just going to distribute like one com token per block half half to landers and borrowers and that's it now we decided to say well it it could be more complex than that and it is uh let's just like do everything off chain distribute like decide how we're going to distribute that off chain according to some rules that the

community itself will iterate on and will improve on and they they do that by using their bow tokens and voting for uh for changes for new factors and that has worked very well so far it's it's really interesting to see how the community has proposed new things and they got approved then yeah it's it's really great to see to me like the fundamental thing fundamentally the issue appears to be that you have pools and yes individual pools have utility to these traders and to the portfolio holder liquidity providers agreed but you have essentially like to design an open source code and

and that can be implemented wherever it can be implemented in cosmos sdk substrate agorax different people can can do the same thing there doesn't appear to be anything which is like a network of pools somewhere in which there are 100 pools each having some assets together doing something where each pool is getting more utility out of that network level coordination if you had a network of pools of some kind then the balancer token could have some claim over value because like it's orchestrating that network but that seems to be absent from your system doesn't it no it's not uh it is there it will be

more so in you in in future versions but an example of how this is already the case um here is when you have a trade you you can use our sor you can trade with individual pools so you can just say wow this is a nice pool it has a lot of eth and maker mkr i'll just trade with trade with that pool but this is not the ideal uh the ideal move for a trader the ideal thing that they should do is to use our smart order router the smart order router looks at all the pools that are on on balancer and it sees all the pools that have eth and mkr and then splits your order in a way that is optimal for you to get the most return possible so whenever you're trading on

balance or you're not trading with a against a specific poll you're trading against the protocol and the protocol uses that network as you said 100 pulls and chooses okay a thousand each i'll send 200 200 200 and 400 and that that way you get the most mkr back and it's no no different than what actually one inch is doing already one inches is doing that network effect across the different protocols right so it looks for liquidity on unit swap and balancer on oasis and it gets like the distribution that suits uh yeah that trade or that returns the most um from from this trade yeah i was just gonna ask that why what's the benefit of doing that is isn't that gonna be someone almost

expensive to do on chain and wouldn't like doing a off-chain aggregation of so what would be the benefit of going through this versus letting one inch use their node to query like every balancer pool and figure out optimal routing we don't do that on chain so apologies if that that sounds like that no though so our sor is a javascript package that you just execute client-side it asks for pull information from our subgraph and then runs like some some sor algorithm and then prepares already okay this is the trade that is going to be optimal

and then on chain you just execute the trade you just send like i want to trade with that pool that much that will that much and then you you execute it we do have we're working right now on our on-chain slr because many protocols need that to happen on chain they just want to say on chain balancer i want to sell 100 ether die um please help me you cannot access javascript packages right so we're finalizing our on-chain sor as well but what i told you about was off chain and yeah so sure one inch can do that can poke all the pools individually but it can get very messy because we now have about a thousand pools so what we do is like we make it transparent

for them and uh give them already like a figure of what the the balancer ecosystem can offer in terms of bass liquidity and then they use that as if it was just like a single unit swap pool they can treat that much better they cannot handle a thousand different exchanges which would be if they had like a thousand different pools each of them like uh treated like a unesco pool so we're kind of facilitating the interaction the interaction of aggregators or integrators with balancer with our own sor one thing i've always wondered is that these balancer pools so each individual pool is at some level paying arbitrages to discover

the exchange rate between assets so when you think of something like ease and mkr it could conceivably be the case that 200 different pools are paying separately to discover that same price could that not be the value prop of the network of pools and balancer that somehow this intelligence is it allows the sharing of this intelligence in some way and it makes a network of pools more intelligent on chain and that's the value prop that even if you want to build your own pool go and join specifically the balancer network because the pricing intelligence is the best there

totally uh absolutely so what we'll see um here is that we're not going to have a hundred pools that are very similar because this is total it's rationally wrong right you're not going to put your create pay 50 or 100 dollars today to to to build to deploy a pool or you're not going to put your money like a thousand dollars in a 80 20 maker eath pool if in a pool that has another thousand dollars if there is the same exact poll with a million dollars right because that one million dollar pool is gonna be profitable for arbors to trade and generate fees uh with uh variations of price a lot smaller than the variations of price in a 1 000 pool right

so you're gonna see naturally the liquidity coalescing or revolving around some sweet spots and the nice thing of bouncer is that we're not aiming to say what the sweet spots are we're not gonna say say 0.3 percent is the right fee no it's up to you if you want to uh if you want to put 0.5 percent if that is a profitable choice what will happen is that people off chain will just observe all the pools we'll see that your pool was a good choice and actually my liquidity will migrate to your pool does that make sense that to me that doesn't make sense because to me one of the value props was that hey i can build my own portfolio now some person can come and say

in my portfolio eth is 10 and maker is 20 and some maker maximalist will come and say hey in my portfolio maker is going to be 80 and each 10 percent and both are valuable and both get their pools and the balancer protocol should be written in a way that both pools are viable so i can't i don't see i don't see the two views as being consistent where you want people flexibility over percentages so those are different poles whereas on the liquidity side you're saying things between each and maker should get consolidated into one pole isn't it the case that portfolio management actually requires there to be

lots of pools with each ethan maker and discovering prices independently yeah what i tried to just to say is that for similar pools and exposures then you're not going to see so much fragmentation but definitely if if there is a an lp that wants to be 90 exposed to eth and 10 to mkr it's not compatible to uh ninety percent uh mkr timbers and eth right so they there will have to be two different pools for those two different uh needs but what i'm saying is that there's not gonna be a thousand pools so 90 10 89 11 88 12 because there is like some some

threshold where it makes sense to kind of uh give up a little bit on your exact exposure that you wanted to being part of a very close exposure but uh to in a pool that is huge already it's used very often and it's generating lots of trading fees and it's very profitable does that make sense so going back to uh you mentioned a little bit ago about distribution of baal and so one of the interesting things that you guys are doing is you know people who are providing liquidity on different balance or pools you're sort of air dropping bowel tokens to them and this is sort of a you know it's a solves two problems of one how do you distribute a token to users i mean okay you give it to

liquidity providers and also you incentivize liquidity but then you know with this whole crate defy craze with like liquidity mining and yield farming and all this stuff that's been going on you're actually like you know the first one we've had on who's like you know been part of this whole you know game i guess um what's been like sort of the takeaways from this process and i also remember reading that there was like you know some sort of like weird things where people could game the liquidity mining and so there have to be like what are the problems you ran into what solutions are you guys working on so yeah i i absolutely think that this

is a first first and foremost a way to distribute ownership of your product code to people who really care about it and those are users those are the liquidity providers who are making it possible for traders to trade so that that was the the decision like how again getting back to the subjectivity of usefulness for liquidity how do you say liquidity is useful well usually how much people are trading with uh with that pool but that is very easily gameable you can just wash trade and uh that would be like a catastrophe very hard to to come around so we decided to say well if uh if you add liquidity to balancer then you get bow tokens

and we knew that this this like was prone or very likely gonna be gained because our idea was to let the community step in and come up with cool suggestions to improve this process so what we initially did is it's just like a percentage of total usd value of the liquidity you provide proportional to the full liquidity balancer and and and your token has to be listed on coin gecko as long as there is like a price tag to your token because if there is none you cannot talk about us usd liquidity uh that you're providing right so you have to have um two tokens in a pool because you have to be able to trade uh two tokens that have a price tag on on

coin gecko and then very quickly we uh yeah the community and we knew that already but we wanted kind of the community to step in and and uh can't come to the same conclusion community realized that if you have a pool that's 98 too that pool is uh even though there's like a lot of value locked in dollars that pool doesn't facilitate a lot of trades because if you trade a little bit that will already like change the price a lot so you have a high slippage for uneven pools what the community came up with was okay let's introduce a ratio factor which says that pools that have very uneven weights they're gonna get less uh bow for the same usd

value locked and that's just like one of the examples and other examples are um hard packed tokens so if you create a pool that has die and a die both are listed on on coin gecko and you can put like a lot of money there you know you have no risk of people like uh of of impermanent laws and then we can get back to that again but there's no real risk of volatility there because you know those tokens are packed so if you have a pool like that with a 0.5 fee there's not going to be one single trade on that pool because people can just use the ave contract to uh to wrap and unwrap die to a die right yeah so we created this hard pack

which says like you're gonna be slashed by a lot if you just provide that not so useful liquidity and yeah and then so on and so forth we we have um a very complex kind of system to distribute liquidity today which has evolved according to the demands of the community wait so how does how does the slashing work it just like a governance based slashing like oh sorry uh maybe slashing is the wrong word you get penalized for providing liquidity that's bad hard like hard packed so you actually don't get as much bow for the same usd as you would if you had eth and maker you know mkr and heath so the same ten dollars here uh for die a die will get a lot less bowel tokens

than ten dollars for mkr and eat if that makes sense and this is all sort of possible because the bowel isn't being distributed by like a smart contract or something right it's just being distributed okay that was exactly that point like uh that that was really what differentiated us from compound like compound took a long time actually to to bake that into uh the smart contracts already and we believe that it's we're so much in the early days there's so much to learn and to improve that um we decided to to to keep that off chain because anyways like if you look at compounds governance it's very concentrated the team has a lot of tokens so they can still decide pretty much the outcome of the votes right so there is already

uh like us a social contract there that they will let the community decide but still you have to trust the founders that they're not going to vote and it's the same for balancer right uh we the the vote the team advisors and investors uh in the these early days uh well we haven't distributed so so many bow tokens yet so we could even if it was unchained we could control the governance process if we broke this uh social contract that we have to not vote and let the community kind of uh build and vote with uh with their tokens so if we already if the community is already trusting us to keep that social contract why not just do like things off chain which is a lot more

efficient and easy to work on and to iterate and to improve right so that was our thought process so that that makes a makes a lot of sense right like especially with it feels like with a protocol like balancer there is a lot of there's so many tokens and there's quite a lot of emergent complexity around liquidity mining here right and maybe you want to be able to tweak the distribution process as as things emerge like maybe it's more valuable to to pay out the tokens to some particular token that's in short supply for example or there could be a role for human subjectivity here one question that does come to mind is that you know balancer labs has

has raised financing and so is the fine is all the financing raised by selling tokens or are these equity financing rounds of a company separate from the now yeah so this is actually we've been quite innovative and uh i think we were one of the first ssfg so uh agreement for future standard agreement for future governance what that means is that investors uh not only acquired equity but also the pro rata share of governance tokens that that equity maps to so what we did is balance our labs the the the company that raised uh venture capital with uh accomplice

placeholder and other other investors it has uh the right to get 25 million tokens out of a possible cap in the future of a hundred million tokens so if you have one percent if you've invested one percent in the equity of balancer labs you have one percent of the 25 million tokens so it's um so the investors both got equity and uh tokens and this is probably not going to be the case for our future fundraising rounds because yeah we we kind of uh want to make want to make it more simple like simpler by having this future fundraising fund which will be used for

uh yeah for selling tokens in new series yeah that's really interesting so what you're trying to do in this sefg format is you're trying to make your investors indifferent to where the value of accrual occurs so if the value accrues to the token they have a claim on it and then if it accrues to the company they have a claim on that so you you are essentially trying to make them indifferent so that it gives you flexibility yeah though we have been very uh upfront and and candid about the equity not being worth anything very likely in the future because since

the beginning we had this in mind that this pro we want to make this product like a common good that is going to be owned by the the bowel token holders and uh the the company ideally will be dissolved and the equity is not going to mean anything at all so that that that's hopefully uh the outcome that we're going for that's quite a radical decisions do you find it hard to convince uh people to join your team and because like your equity is going to be worthless so why how do you incentivize your employees then that's a great so this is not something that's going to happen overnight next year

so it's probably a long like a longer process like five years maybe to 10 years and what i what i and and that also includes myself not only our employees right um we're here so what i what i think is that even though the company itself is going to be dissolved that doesn't mean that uh the team that's now working for balance or labs won't be able to work further on balancer right because then we're gonna have like we're gonna find some other common good ways of financing like bitcoin grants or yeah whatever means we we we can find to finance the development on the platform uh we'll probably be involved with that like

vitalik is not paid by by the the ethereum foundation as at least as far as i know but he's still involved he's still doing work uh and there's many people are working for for ethereum with other kind of incentives in mind not not necessarily getting a salary from a centralized entity so this is how i envisioned the future of balancer so given all this like liquidity mining that's or incentivization that's been happening one thing we see on balancer is basically it's beating uniswap massively on liquidity but not yet on trading volume so what why is this and what are sort of what's your plans on we've got the

liquidity now how are you planning on attracting the volume trading volume great question so this is because we've been around for a few months and unit swap has been around for a lot longer and unisop is integrated with pretty much everything in the phi like whatever platform or front-end or like um distribution app like zirion or or zapify they've all like been integrated with unisoft for a long time so uniswap is seen by everyone it's also like a brand people think of trading they go to you know swap decentralized trading go they go to unesco so um it's a matter of time until we we have like integrations by xerox api they're like finalizing their integration

or kyber network they're also finalizing integration one inch has started already uh integrated with balancer from day one so periswap is also working on on integrating balances we need time to kind of uh get integrated right so that that's the first thing and the second thing is that actually unit swaps liquidity is the most efficient possible because it's always 50 50. so there's a lot of liquidity on balancer there's like a 200 million dollar pool one pool has 200 million dollars now which is yfi and the curve y curve token uh that was a 98 too so it naturally has effective liquidity

the effective liquidity that deadpool has is a lot lower than 200 million dollars right so it's hard to compare volume between uniswap and balancer because uh balancer allows for like flexibility which means that no not all liquidity will be efficient or the most efficient for traders because that means they are like the pulsar 50 50 or third third third or 25 25 ton fight and that's not the case what would you say is the most unique or interesting pool that you've seen or maybe it's not been deployed yet but like you know that people are working on there's so many um so i think lbps

are gonna be a huge thing as i said uh earlier sunny like the the idea of uh the ideal like initial decentralized offer or decentralized exchange offer this flexibility of having a smart pool that like has this schedule of weights flipping and letting people not only buy your tokens but also create liquidity all wrapped in one to me this is very powerful one pool that i'm really excited to see uh coming out and probably i don't know if you talked about that last week you had uma uh here in the show the there's this idea of having a perpetual synthetic pool so what um does is they have this um

this awesome system that allows you to create any synthetic that has an expiry date and often what you see is people seeking exposure to gold without an expiry date so they want to be just long gold but that doesn't work so simply like so so easily for a product like uma to do directly because uh maybe they explained that but they need this expiry date in order for the underlying price and the synthetic price not to deviate a lot so what you can do is you can have different months like synthetics with expired dates of different months all in the same uh in the same pool and then what you do is is a smart pool that as soon as one of the expiry dates is

getting close to maturation you start decreasing that weight to start phasing out that that gold let's say september because it's close and you start facing in gold november so and that is naturally done by arbitragers that do those trades for a marginal profit so your pool is actually leaking a little bit of value but you know that your pool will always have like three or four different months that are being recycled automatically so not only you create this one place for people who want to have exposure to gold november to trade with gold exposure september people can trade between those different uh different months you're also creating an exposure and a

perpetual exposure to gold which is the erc20 token uh of the pool so that pool is issuing ese 20 tokens that represent exposure to gold full stop so that that's amazing have the option to to have go december and you trade for gold september but you can also buy the pool token the bush here which is just exposure to gold and uh yeah to me that's that's a very interesting example of uh of pool that will be soon launched and we're working closely with the uma guys on that yeah that that you know i was shining with my friends at open and they that's you know another problem that they face with unisoft and that's also why they're looking at balloons to kind of do something similar as well

cool so what do you what's like the future like you know what's scattered throughout like the docks it mentions balancer v2 but there's not a lot of details what's what what is what can we expect to see so battle survey 2 will be a lot more efficient in terms of gas costs we're going to go to this single vault architecture so instead of creating like a new smart contract for every pool that's deployed you're actually just creating like an internal accounting for the pool you created inside of this big contract that is balancer v2 that will allow us to do cool things like huge flash loans so all of a sudden you have like all the mkr from all the pools

sitting on the same contract so we can have like flash loans that are huge and that also will allow you to do arbitrage between pools in a much more efficient way because uh all you're doing is like you're changing internal accounting balances of the the pools inside the same contract you're not sending erc20 tokens like across different pools you're not paying for all those your c20 transfers all you pay is at the end this is what you're gonna get token a so then balancer vault transfers you that amount and gets from you that amount of token b that you sold so that will create a lot more uh a lot more possibilities for for arbitrage and for traders

uh yeah and we'll we'll have like nice features as well like more flexibility in terms of uh for example you can have circuit breakers so you can set your pool to stop selling if one of the tokens goes below price x and this is something that uh gnosis has been asked asking for because if you have prediction markets one of the tokens can just like outright go to zero and that is a problem because the pool gets drained so having circuit breakers like allow allows you to say okay so whenever one of the outcomes is known the pools will just freeze because the prices will go beyond the threshold that i'm i'm willing to to let people trade against

lots of nice nice features and cool things that we're getting like feedback from the community like you should do that and you should do this we'll do a lot of uh yeah improvements also in terms of flexibility it's gonna be very cool awesome thank you uh yeah thanks for taking the time to chat with us and teach us more about balancer yeah thanks a lot senior here so yeah if you want to know more about balancer you can go to our discord channel you can go to uh you can follow us on twitter at balancerlabs i'm fc martinelli on twitter as well yeah it's been a pleasure guys great questions and yeah looking forward to chatting more offline thanks so what did you think of uh

the project i mean i've always felt balancers a great project yeah i i i wrote like this like long blog post a couple of months ago called like the uniswap or darwin dowrifying uniswap in making the uniswap network and i wrote this like long thing and then as soon as i published it people are like wait you should go check out balancer because that's exactly what they're doing i'm like whoa because like i looked at balancer early on when you first showed it to me but back i always just thought it was oh it was just like you know multi-dimensional unit swap that you know but like i didn't realize that they were so focused on like

all this generalizability generalization which is what i find super fascinating because i just think that whole yeah like i always thought that look okay this x times y equals k that there's no way that that's the optimal like formula that's just like you know someone just pulled that out of their hat and it's like but i'm excited to see that you know they're so focused on parameterization and especially i imagine they're going to get even more you know we didn't get to talk about this on the thing but you know i'm willing to bet that you know they're going to even start doing other types of curves not just like the same constant product constant product one but like you know we have there's other products hyperbolic yeah there's other ones that curve finance which are more flat for stable

coins or you can i don't know you can do like some wacky stuff probably that um yeah so i'm excited about it from that sense right i i just feel that that it's a it's a very elastic primitive and it it will end up entering into lots of different applications and much of much of the use cases of something like balancer aren't just understood or imagined whatever i love about this project is just its elasticity and how it can mutate and fill different niches so what do you think one of the questions i have

do you think amms in general are a long term sustainable concept though like just this like it feels like what balancer with all their parametrization they're making amm or they're giving amms the opportunity to be smarter than x times y equals k but at the end of the day i still feel your automated market maker or you know algorithmic market maker whatever you want to call it like you know algorithmic market makers like even real market makers are algorithmic the difference is that a smart contract market maker doesn't have access to external data and i feel that because of

that it will always be worse than like a real market maker who can quote you prices but like has access to real world data and it seems that what's hap like what's gonna happen is how i see automated market makers is that this is what they're doing they in a traditional market maker you have someone with capital as well as knowledge or information or expertise and they're providing quotes and it's the same entity amm seems to be splitting these into two different categories one is the capital providers and then

there's the information providers and the capital providers are like these dumb capital fighters these provide capital and then all the information benefits are going to the arbitragers um and it seems that like the arbitragers are screwing over the capital providers and unless you can get your fees to be high enough like and this is the whole impermanence loss situation right where oftentimes the if the if it's not unless you have enough volume your capital providers really get screwed over um and the a smarter market maker could do better so what do you think about this i i totally agree so that that is a

downside of this gen this generation of automated market makers that in the sense that the automated market makers are they are blind and they are dumb blind because they are unable to see what's happening on other markets and they are dumb because they don't have the information but they don't even have the mechanisms to include that information into their own decision making so would it be fair to say balancer is making them less dumb but it's not making them less blind yeah it's not making them less blind right so it's still blind so i i do agree that uh this lack of

blindness is is a disadvantage for automated market makers but it's also the case that we are so early in the game no yeah i mean unisop and balancer combined this entire space is two years old and there's an eternity of 50 years 100 years to develop this thing and a lot can happen and i think eventually their blindness will go away uh i'm not sure how they will acquire information from the real world because it's a hard problem so but i do feel that it's just that the design space of these crypto protocols is just so

massive that the challenge will be surmounted and we will have automated market makers that are way more efficient than today's race variety i think that's the optimistic case are they gonna remain for more niche assets or do you think that like they're gonna overtake like certainly where i see them going is that they are like great for like weird niche assets like options or like um like wi-fi tokens or whatever but it's like prediction market are they gonna take over the btc usdt pair like where that pair is just like

the biggest pair on like all centralized exchanges is it gonna take over that i think that's unlikely that's unlikely i think i think i think you've identified the market it's the long tail of assets that automated market makers are really suited to today and that's the initial niche they end up populating now now the question is beyond that initial niche can they invade other niches and then when they do btc usd what not only btc usd but one fine day in the future you want to do all euro usd volume or automated market makers is the future

like that i think it depends a lot on how efficient these things can these things can be made made to be and i think it boils down to the question of what kinds of data can be given to these automated market makers and how much intelligence you can put into them and both are hard questions because getting data means developing a protocol by which you can trust the quality of that data and then intelligence probably means a lot of computation and on both those dimensions blockchains are very limited today now i think if those dimensions were to

be solved market makers would grow a lot but but if if they remain remain at their current state i think they are more suited to this long tail of assets what what do you think i i i think that they're yeah i don't know it i think that until they start and until they figure out how to do that solve that blindness problem it will as soon as there's a larger market for um that the that pair then the amm it starts running into problems when the largest pair for that market

is the amm then it works fine because the because then the amm is the price it's setting the pricing but when the real price whatever that means is being set somewhere else then that's where the capital providers start to really get screwed over because the arbitrages make all the profits there right yeah and it's one of those reasons i i haven't contributed any liquidity to a unesco pool because because it like as a like if i think of my ether or my bitcoin like really liquid crypto assets i hold in at some level i'm holding these crypto assets because

you expect them to go up in price but when assets grow up in price and you've put them onto a liquidity provider you're making losses at exactly that same point so so at some level that that that that is what has prevented me from actually putting assets into into the market makers now yeah so i think i think for very niche assets like the real team real estate tokens or these y tokens or the the situations is different but but let's see how it evolves i'm

actually quite bullish on this space even though the solutions don't exist today my my instinct is that the design space is so massive that it will end up being solved but it will take 20 years you you were um asking quite a bit about like the network effects of the balancer protocol and you seem kind of skeptical do you still are you how do you feel after that discussion so my impression is that balancer has worked out a really good pool design and at some level you know i mean if this was to be an economics paper or something if if i imagine balance as an economics

paper this might end up becoming i don't know one of the most cited papers of 2008 2019 in the far future i will design an ideas level of that pool design it's amazing that's that's my impression but i'm not sure about the code quality right how how the design translated into code but let's assume it's it's average it's better than average or whatever but i where i feel when i feel balancer is kind of lacking is that there is a great pool design there's this code and then there's this token but the question of how this token actually captures value

is is not well addressed now of course you mentioned sunny that the token will someday say that hey there's a certain fee that you you can add add to these pools and and that is how it will capture value but then the question just shifts to being okay if the token holders decide to add a fee what prevents these liquidity providers from migrating to the next protocol and when you ask that question what will prevent them from migrating to another protocol it's then that you kind of realize that okay for this to be really viable there has to be some network effect associated with multiple pools now now the answers that that were given were

okay smart order routing but to me smart order routing seems like a completely off-chain play like chorus can come and build a smart order order router for balance as a centralized entity and capture the value around around smart order routing which is what one inch is doing which is what one is doing so it doesn't feel a good way by which the protocol itself captures a defensive mode and captures value and i think like to me with current balancer that is the gap that i would i would think about when i would make a

decision of whether to purchase bad or not because if if that gap can be solved then i think this is going to be one of the great investments of the crypto space it could be on the investment of the crypto space but the uncertainty around them not being able to solve the gap is what what would make me skeptical of buying the bad token what what is your sense yeah i i agree i think the smart order routing isn't the solution um i one if yeah like you said there has to be some

way of where the networks all the the pool of network of pools are somehow sharing information with each other such that they can help balance each other out in that way um that's sort of almost that's sort of the key i think that has to be figured out because earlier in the conversation we were talking about how these pools are blind now okay so an individual pole is blind but even if you could make it less blind by having the that pool get info from other pools

then it's not the perfect solution but you're making it less blind but if you can make things less blind then there could be some kind of network effect that's my impression he mentioned flash loans right and in the v2 is there a way where we can somehow restrict the flash loans such that you're allowed to use flash loans as long as you're you're arbitraging like within balancer pools so you can use a flash loan from this pool to go arbitrage another balancer pool but you're not allowed to but somehow place some restriction where you can't use the about the flash loan to go do you know you can't go use it to do something on

uniswap or compound or whatever right you know free arbitrage amongst balancer pools only that that seems like uh you know as there's more liquidity then the arbitraging basically gets better and better on balancer alone that that would be really interesting that would be really interesting so i think some mechanism like that is needed and i also think it's it's probably coming but if thinking from an investor's perspective i would want a very clear answer to that before buying the ball token well you can't buy the bowel token anyways right you have to go well i'm sure that you can buy on a balance or pool

but yeah perhaps i can do some liquidity mining to get bad tokens i'm not sure right yeah i mean it's it's certainly i i feel it it's a really interesting uh project and i i respect the idea a lot my nitpick is you know how the idea translates into business model but we have time to solve that right like this is a long game crypto is a long game so well looking forward to seeing how it turns out yeah i hope you enjoyed the interview debrief and you can get access to it every week by becoming a premium subscriber as a premium subscriber you'll get access to a private rss feed where you

can hear the debrief every week and you also get enhanced features like full episode transcripts in your podcast player and interview chapters that allow you to easily skip to specific sections of the interview you'll also get exclusive access to roundtable conversations and bonus content we put out from time to time to learn more and become a premium subscriber go to premium.epicenter.tv thanks [Music]