Jayant Krishnamurthy I What’s Stopping DeFi from Competing with TradFi I Pyth I Pragma San Fran 2024

[Music] so without further Ado let's welcome our next speaker giant to talk about why it's very hard to compete with trafi with Defi and what we can do to make that a much much much better experience so please give jant a big round of applause thank you caric um so hey everyone my name is jant I am the CTO and co-founder of Duro labs and we are building p network if you're not familiar with p network I'll tell you a little bit more about it a little bit later in the talk and the title of this talk is something like uh what's it going to take for defi to compete with trafi or why can't defi compete with trafi and I think the answer to this question is uh actually very simple which is that well defi sucks so you know now you know the answer um thank you for coming to my talk uh no uh but seriously I do think that as builders in the space we need to recognize that there are a lot of problems with Defi and that the actual experience of using defi today is not going to meet the needs of the vast majority of users and that is fundamentally why defi is not replacing tradire and just to uh drill down on this a little bit I want to do a comparison of the experience for users using D versus trafi and look this experiment I mean this analysis is not meant to say that there aren't problems with trfi I mean try opening a bank account or moving countries or you know God forbid you have a business and you need to open a business bank account there are a ton of problems with the traditional Finance system however those are not problems that people encounter every day those are things that you hit once in a while maybe once in a year you get really frustrated but then you kind of forget about it and I think it's important to realize that for the vast majority of use cases the traditional Finance system is actually pretty good and so I want to think about that in terms of three dimensions which are um first of all speed the traditional Finance system is actually pretty fast if you want to place an order or like buy something on Robin Hood or you know a platform like Charles Schwab your trades will settle very quickly less than a second it's just like using a normal website so it's a pretty friction free experience the whole process is pretty efficient the second one is cost so on all of these platforms you have essentially zero cost trading right market makers with like pay for order flow enable zero cost trading the liquidity is really really good it's very very efficient to use the traditional Finance system the third thing I will say is trust and look I know defi is all about you know eliminating trust in third parties but in reality the banks the Brokers are pretty trustworthy they've been around for a long time I'm sure all of you have money in these places I doubt that any of you are worried about losing that money right now so from a trust perspective the system also works pretty well and that is really what we're competing with in defi right it's a system that's actually pretty good and just let's see how defi Stacks up against that traditional Finance system so for defi um first of all let's talk about speed depending on your blockchain you know things are going to settle in a couple seconds that's not counting all of the other you know messing around you did with your ledger to sign the transaction whatever that's easily going to add a minutes right so it's a much slower process to interact with defi than it is to interact with trafi uh the second one was cost so defi is much more expensive than trafi in a couple different ways obviously gas costs will be Salient to everybody here you've all had to pay for gas costs you know those are coming down as we improve the te the technology but there are other costs like liquidity and slippage which are just worse in defi than they are in trafi and this makes transacting much more expensive the third thing is Trust uh I'm pretty sure that well some of you probably have your money in defi protocols and uh you're probably worried about losing it right there's a lot of hack risks there's a lot of trust issues at lower levels of the stack with things like infrastructure um that could actually cause you to lose all your money so uh just based on this comparison I think it's safe to say that defi is not as good as trafi for the vast majority of people's use cases and to give you a preview of where I'm going with this at pit what we trying to do is we trying to build infrastructure that enables decentralized Finance applications to compete with trfi applications so we're trying to kind of bridge that Gap over here and I'll talk a bit about how we're doing that in a minute um and the reason that we're doing this is actually you know when we started building pith we sat down and we said we really asked this question which is what is preventing defi from competing with trafi what are the actual problems and the thing that we realized is that when you're in one of these situations like the blockchain which is sort of a new paradigm for build applications um you really need to end up in a situation where the new um well the way this works is basically look there's a new paradigm let's say it's the internet there will be someone who is solving a problem using the internet as technology and there will be an incumbent who's solving the problem in an old fashioned way and in order for people to really shift from the old Paradigm to the new paradigm the new paradigm needs to offer them a solution that is better in some respects and no worse than their existing solution in all other dimensions right this makes it much easier to sell someone on making the jump and this is also when it becomes obvious to people that this is the right way to do things it's much much much harder to get people to transition when they're going to get some benefits but they're also going to lose something in the process right just psychologically that is difficult and so uh we can actually look at some examples of this so let's look at an example from the internet I think this is another example of a classic kind of paradigm shift in how people build um applications and you know the thing that the internet really brought to the table is zero marginal cost distribution so this means Amazon could offer an effectively infinite number of products for sale right they could have an infinite number of pages in their catalog right and that's a benefit that's something that users would want but if you go back in time to say 1996 and try to do a comparison between Amazon and the Sears catalog the Sears catalog actually looks pretty good so you know with the Sears catalog let's say you're looking for something to find there's like 10 pictures per page you can flip through pretty quickly maybe a second per P um second per page it's pretty easy to go through that and find things and buy them whereas with Amazon on a 56kb modem it would take you about two minutes to load 10 pictures I think it'd be pretty hard in 1996 for people to think that this really slow website is the future of e-commerce okay so this is again another paradigm shift but the interesting thing about these Paradigm shifts is as the infastructure that powers these applications improves um adoption kind of grows along with that so here's a graph that I have that shows the adoption of Amazon as a function of time and I've marked on here you know some of the infrastructural improvements that have kind of you know where they came along in that journey and I think you know you can ask this question for yourself but you can ask you know when did it become obvious to you that Amazon and E e-commerce were really the new paradigm for doing shopping and for me I think that was somewhere in the mid 2000s and I think it coincides pretty closely to when broadband internet became widely available because with broadband internet you could actually browse the Amazon catalog it didn't feel like a slow and clunky experience and then you could get the benefits of the added selection right so that's kind of the place in time when the experience was no worse from like a browsing perspective for Amazon and so hopefully this framework for thinking about things makes sense to you and I think we can do kind of a similar analysis for blockchain so blockchains do bring a fundamental Improvement to you know um what what do I call it I forgot what I called it but anyway they bring a fundamental Improvement to the technology which is uh they enable programmable trust it is a setting where the people using the system trust the incentives of the system to ensure that it operates correctly versus trusting some you know middleman effectively like you would in web 2 and you know this adds security and things like that which are properties that people value right so there are real values to this Innovation and I think this is actually the reason why a lot of us are here today is because we're excited about this vision of the future the thing about it though is that we have not yet reached the point in time where the infrastructure that powers web 3 is good enough to compete with the traditional Finance system so to do that let's do a comparison of say like d versus Charles Schwab right with Charles Schwab you have something like 300 millisecond latencies on your trades you have zero commissions this is much better than what you're getting with your defi apps today um however I think if you look at again an adoption curve as a you know function of the infrastructure you can see how as infrastructure is improving you know it's improving the experience for all the applications and it's also we're also seeing sort of concurrent adoption in wallets and things like that so you know we're on a similar sort of Paradigm Shift the growth curves are all going but we still need to work on the infrastructure in order to really enable that next generation of applications right I think we're getting better but we're definitely not there yet to the point where we're really going to start competing with trfi so what we are doing at pit network is that we are trying to build defi infrastructure that enables applications to be supercharged and really compete with their traditional Finance equivalents and we think about that on these three dimension of trust speed and cost and you know when thinking about that we've thought about a number of different products and solutions that we can release that kind of help applications in different areas and so I want to talk a little bit about some of those and then I'll talk a little bit about uh places where you can see how these infrastructural improvements are actually leading to new kinds of applications so the first thing that pit released which is sort of the flagship product are price feeds these provide uh prices for something like 500 assets on 75 blockchains different environments those kinds of things and when we were uh approaching this problem in 2021 uh the reason we were attracted to this was because it was clear to us as people with a financial background that market data would be a really critical component of any sort of defi solution data is critical to all of Finance if you've ever worked at a trading firm you will know this um and the other thing that we realized is that uh you know look this is an oracle right it's trying to bring state from the offchain world onchain but we realize that this problem this type of Oracle problem really needs a specializ solution in order to be um competitive and the reason for that is that market data is highly latency sensitive it is most valuable when it is first created and then the value decays as a function of time and it decays pretty quickly so if you have ever Googled a stock price all of that data that you're seeing that's publicly available is delayed and you're effectively getting the the free version of the data that doesn't have any value if you actually want the valuable data you have to go to someone like NASDAQ and pay them for a real-time subscription and these subscription costs tens or hundreds of millions of dollars so NASDAQ and these other uh exchanges make something like s billion dollar a year selling Market data to hedge funds and other people who are latency sensitive so it's a very big business and we wanted to basically bring that type of Market data onto the blockchain the first obstacle that we found is that the blockchain is not a good spot for businesses like NASDAQ because their whole business is based on selling a subscription where you can't resell the data right it's all about hiding the data whereas if you want to put it on the blockchain It all becomes you know World readable right so it breaks their business model so what we did instead is we found um other parties to participate in the protocol like hedge funds and market makers who have access to really high quality data but hadn't viewed that data as a source of Revenue right it's sort of a byproduct of their operations and by bringing these uh participants into the protocol um we found a really low latency and high accuracy source of Market data that we could broadcast to the world and we combined this with a with a pay peruse model where people can pull the prices onto the blockchain and pay every time they do so and this combination has really enabled the system to scale to you know the 500 feeds and the 70 blockchains and all those things today um so I'll talk a little bit more about these price feeds I think this is the product that we've had around the longest and has actually enabled the most kinds of applications so I'll talk about that at the end but I want to talk about the other products first so um oh okay skip that um the second product that we launched is onchain random numbers so this is kind of an analogous situation to the the market data except it's sort of for for gaming every time you're building a game you have a question about do you want to trust the publisher of the game or do really want to have a secure protocol for generating the random numbers so that everybody using it can feel confident in the randomness right that your pack of cards was was generated fairly and those kinds of things and um when we were looking at this problem we thought that the existing Solutions didn't really meet the speed and cost characteristics that people would want they're a little bit complicated they require complicated cryptography and this is just hard to do on the blockchain so what we ended up doing is we ended up launching a new protocol which is powered by um it's sort of a version of a class very classic cryptographic protocol which is a two-party commit reveal so if you're familiar with this protocol it's a system where two people can work together to generate a random number and neither party has to trust the other party um in order to prove that the result was fair so as long as I do my part of the protocol correctly I know that the result is random and it doesn't matter what the other guy does um and so this a very simple protocol actually uh it's sort of embarrassingly simple if I told you how it actually worked um but it's actually very effective on the blockchain because it's very cheap it just uses hashing um and it can be done very fast so this has been a very um this sort of a newer product for us but it's actually already powering a bunch of different applications uh finally the the third thing that we've just launched is a product in the me space so I'm sure most of you are familiar with me or maximal extractable value for those of you who are not um me refers to the value that validators or miners can extract from the network by reordering transactions so on any blockchain there will be transactions that are valuable whether it's a lending liquidation or arbitraging an amm pool there will be transactions that are inherently permissionless and whoever does them first gets some money and because the miners or the validators have the ability to choose who gets that money what ends up happening is they end up you know earning all of this value right so if you're M with flash Bots or these kinds of auctions those are the mechanisms by which the miners end up earning all of this meev and I think me is is a really serious problem to think about in defi because if you think about me from the perspective of a defi application meev is a cost that money is coming from the users of the protocol it's coming from the person who lent the funds it's coming from the LPS and the amm pool and what that means is it means that defi applications will not be able to have the same efficient pricing as their trafi analogs right they're effectively paying an added cost to me when the Trad versions are not so in order to really be competitive with trafi we need to think about ways to get rid of me or minimize me and so we were thinking about this problem and what we realized is that there are a number of different kinds of me but one major class of me basically comes down to the fact that applications don't have a way to get efficient market prices for um transactions that they want to perform so a lending protocol would be a canonic example of this where you have some collateral that you want to liquidate in order to repay your loan so you want to sell some ethereum get some usdc the way that lending protocols do this today is they offer a bonus so they say I will sell you this ethereum at a 5% discount and whoever gets it first gets that money but the problem with this is that in reality you could pay much less than 5% and still have someone do this operation right there are Searchers out there who would do it for 25 basis points probably less than that and so all of that remaining money all of that 4.75% ends up becoming meev that's just lost from the protocol and so when we thought about this problem what we realized is that if these protocols could just get efficient market pricing for the for the transactions they wanted to do this me would simply go away and so what we've created is um this product called Express relay which we're calling a priority auction and what this is is It's A composable Primitive that you can integrate into any other protocol where you can basically add one line of code and say I want this trans action to be auctioned off and then the protocol will auction it off to a network of Searchers um we have some of the top Searchers in you know in the world like winter mute participating in Express relay and they will bid on that transaction and all of the money uh from that process will go back to the protocol and you can transmit it back to your users perhaps the person who loaned the collateral or whatever you want to do with it and so this actually eliminates me it eliminates um that whole leakage of value from the defi ecosystem um okay so I want to talk a little bit about uh some of the some of the ways that we can see how you know these kinds of infrastructure are really powering new types of applications and so one of the trends that I think probably most of you have seen is like there's been a lot of pers protocols created recently like over the past two years there've been like probably a hundred I don't know um and that trend is actually under the hood it's actually partly because of better Oracle infrastructure so pit launched um our initial version of the product was on salana we launched it in 2021 um but we launched a version for evm and other blockchains at the beginning of last year and this is basically right before the rise of perss and if you look at the protocols that have been built today 60% of the derivative protocols in existence use pit oracles and the reason for this is that before you had this high frequency low latency Market data it was basically impossible to build a derivative protocol if you have an oracle that updates every 10 minutes there is no way that you can use that to Mark somebody's trade it just it's not going to work right so the availability of this better pricing infrastructure is kind of the the secret sauce under the hood of the rise of a lot of these derivatives protocols um so this is really cool but the thing uh the thing I want to say is even though we're improving no I where this slides here hold on even though we are um improving the state of infrastructure you know we've now created made it possible sorry this is a graph that shows the total transaction volume on crypto derivatives and it shows uh I don't know if you can see that line but there's a line there and below that it's sort of Defi and above that it's kind of trafi or you know centralized exchanges and you know we've with better infrastructure we've made it possible for defi to even you know have a nonzero value here and it's growing quickly but you can see that there's still a ways to go to compete with trafi right the volumes that are being done on defi are almost nothing compared to the centralized exchanges today so there's still work to do to improve this infrastructure to the point where defi can really compete with trfi um and that's basically what we're doing at pit you know we're working to continue to improve the infrastructure to ensure that defi doesn't suck um so that's what we're doing that's my talk um thank you so much for your time if you're interested in chatting I'll be around today and tomorrow just come and say hi [Applause] [Music]