TradFi vs DeFi: How Vertex Is Shaping the Perp DEX Landscape

Transcript

welcome back to the rollup High Roller Community as always nothing on the rollup is to be ConEd as Financial advice we are just an Educators in the defi space toss us a quick like give us a share a repost send it to a friend who may want to watch this and finally give us a sub or a follow it means the world to help support credibly neutral educational content in def5 we're here to educate Empower and enrich you and let's jump in what's going on guys welcome back to the rollup where here with Darius vertex just sandwi in between salon and Adas for fees darus how does that make you feel sir feels good maybe we got some growing to do which one have we got to beat

salana or avalan I think uh I think you're ahead of of uh salana maybe 24 hour we saw some stat on on defi llama oh yeah no no stop till Bitcoin up only Bitcoin Bitcoin yeah uh Rob how you doing you just came back from the Pudu party p is fun uh I'm doing good I uh was certainly on a bit of a just a bit of a go there for uh about 48 Hours traveling and then and then having some fun like ready to get back to work because uh we've got a lot of a lot of work to do to overtake Bitcoin as a network with uh with vertex

that's sick yeah seriously what's the uh what's your what's your state of the pur Market take here and 20 December 11th 2024 what is the state of per because um there's a lot of per markets going out people have this idea that perss are money makers they're just money printers yeah yeah I mean it's a difficult one right so everyone thinks it's easy money and just spin up a p STX uh get a little action in there and off you go easy money but um see a lot of platforms come up and die in the last even six months right um I don't want to speak ill of the dad so

I won't mention any particular names but for a while there say a lot of people had the wisdom of um going clone GMX it's such an easy thing to like clone their V1 smart contract put some special spin on it put it on another chain or put it even on Arbor and then users will come and they'll use it and that wasn't really the case and actually there was like a lot of complexity that had gone into how GMX had built their business for example so they built a really successful uh referral program and they did a good job of appealing to very particular users who wanted High leverage or wanted to be

onchain only and then um I think the real lesson was like they built some kind of moat around what theyd built right so that that business wasn't instantly replicable and therefore people could come and take business away from that which is really a lesson you can see from any centralized exchange right they try and build a Mot around themselves it retains users I think it's the same thing with per so I think a lot of users plow in and just say oh it's easy to be a per de anyone could be a per stack but if you look at the list of per staes the number of per stack is actually making money is really slow you know because if

those guys are making you know probably the rough rule of thumb is if you're doing less than aund million of volume a day which sounds a lot if you look at fees and everything else if you doing much less than that probably ngmi right it's just not going to happen for you and unless you got a plan for how to get over that and keep over that it's not going to be good for you you might last for a while but it's not going to be a fun existent and so the big thing for us with vertex was like What mo can we build for ourselves you know I think it's probably safe to say we have like the lowest latency highest

rer architecture we have probably competing with dydx but like most market makers and therefore liquidity on bid offer um we have the cheapest fees we're building out we have a referral program but I think we can improve it you know like all these Matrix if you can be top three on everything you're probably going to be successful otherwise you know thanks for coming I mean is there even any room like this quite seems like the winners are esta how do you come into this face and win them it's like and and from a perspective of a user man

it's like I'm using vertex and I trust vertex of millions of dollars in what world am I going to a GMX for and what world am I switching my my place of choice you know yeah I mean yes but also you know crypto's a bit fast forward right so like what might feel like a long time or lifetime in crypto probably only six months you know and the fact is that all these like onchain Solutions are still only I can't even remember what stat I put up the other day but I think the whole perk deck space like last Friday was like less than 5% of what binance did that

day right so there's still root for like huge growth everyone can get bigger but I think in the mean time you need to plan for like how do you get into that top three to five and preferably you know top one to three because otherwise it's extremely competitive there's not enough market share right now to feed all those mads in five years yeah probably but between now and five years there's a lot of living to be done and how do you do that I see kind of like two parallel directions here where there's one there's one school of thought that might say that uh in order for uh those like to feed all those Mals it means that like all of they all of them can win but

the centralized exchanges must lose do you think that's a necessity to survive for for a per STX on a chain to survive it means kind of the downfall or at least downsizing of centralized exchanges or kind of as you alluded to in five years there's enough growth overall for both the centralized and decentralized for for both of those party still in yeah I mean it's interesting because binance is about half the market right so they're just using binance as a rough yard stick they do like 50 to 100 billion a day the whole Market does maybe 100 to20 billion dollars of

trading a day you know FX alone trades a trillion dollars a day as an asset maybe it's two trillion I'm not to dat with the latest numbers but it's in that region 1 to2 trillion crypto's still tiny they we could grow the whole industry could grow 5 to 10x it would not be unreasonable given the number of assets the way things are growing stuff that's going on you know like it would still be a 50x away from FX right and therefore like at that point what happens well you know all these per taxes when if

they just a sustained market share would become a bigger number of grow know this Dynamic with exchanges the longtail of exchanges you see on centralized exchanges the top three to five make money top five to 10 do okay if you're in like 20 pluss it's not an amazing business to be in right all exchanges thrive on liquidity and activity you want want to be the near the top of that scale or else your business probably isn't viable long term and liquidity tends to pull on the biggest venue so um right now means buy that previously that's been other exchanges that number one spot is super

super important so I think that the the this this idea was brought to light basically like the end of 2020 is and I like def was like oh well we need per SES this is going to be the next big thing kind of blew up um and and now there's you know like you say tons um is there any other verticals that you're more bullish on than per texes going for what is more bullish than offering people high leverage on chain where there volumes dramatically higher offchain is a great question uh

I guess I have a sort of old Boomer style brain right so I just look at like what's being successful offchain that could probably successful onchain right using that shortcut um well you mentioned per dexes P's obviously a super successful product with a lot of product Market fit on centralized venues that's translated over into relative success on chain um same thing's been true with like borrow lending platforms there a lot of centralized borrowers lenders that has translated to success of things like ARB um I think I think for me this is a bit

of a like there there's two answers like a boring answer the really boring answer is like well payments you know look at things like Cal that probably becomes more integrated into like payment system systems um that's kind of hard to get excited about but in truth that's probably a big chunk of like where crypto's going I think the other part and this is something we're trying to do with vertex I don't think we're all the way there is more integrated platforms some people try to do that through composability we've tried to do it through integration within our own system but like the ability to do more than one thing on

chain so being able to trade spot per have that link to a money market I think dydx V4 is trying to do something similar to what we have on vertex in that s um you know you saw something similar on the old mango markets on Solano which kind of had a bit of a rebirth recently like there a that move towards more integration I not just like one vertical One D like more integrated system kind of analogous to things you've seen in like social media where it used to be like Facebook was just a thing you could post stuff to your friends with and that was it where now it's a full picture sharing video

sharing messaging system people tend to move towards just having like one platform for things I think and so I think that's probably the next iteration is like integrating all these things together either third parties that integrate are their Union I don't know name what person that's G or systems like vertex that trying to integrate spot pers money market all in one say interesting interesting say yeah I've seen that with uh with um lending markets like you mentioned OB they have like a swap tokens feature now right where you can just swap directly in the app honestly it's just a pain to click around multiple apps to do bridging all

that kind of stuff um I mean if you're doing everything on chain you know we have oneclick trading so you just press and then you just press press a button and it works so easy but so how miserable is the amount of button pressing you do in crypto where you know like we all know you're supposed to check every transaction the fact is I mean I've been in this space for a few years now I still don't check every transaction as thoroughly as I should and then afterwards you panic and go oh God I'm going to lose all my money it's like all my money is going to get taken away but it's it's really bad ux so I think those ux improvements would be huge as well over time I mean man going from oneclick

Trading to having to whip out The Ledger and confirm all the transactions and pay gas is like oh my gosh it's going backwards it it feels it feels disgusting to do transactions um in where you have to click The Ledger every single time but there's also tradeoffs with that which is something that um I'd like to jump into kind of in in a little bit further on this ux and kind of hers not and reason why I was asking you mentioned quite a bit of the institutional interest in vertex now Craig G Ro Perpetual uh Futures contracts kind of a zero to one moment for crypto where binance and bitmax offered these Futures that were Perpetual right they were you could have them open forever now this

this isn't something that is offered in the stock market or in in in you know different types of uh trading venues I mean I think Forex has leverag which you keep open for um you know for as long as you like but there's often like excessive fings and and uh decays and in the summer you've got you know High beta plays like you know gold you got J num if you want to go super aggressive but again it's a lot of beta or a lot of DK if you want to hold those things for longer than a day or two days but the zero to one per Futures you got these these high lever positions you kind of hold for a long time with reasonable funding rates is this something that you have to explain to institutions as to why it's why it is this way in crypto or

is there any education that you have to do when it comes to onboarding like larger players from from trfi that they're not used to when it comes to Vertex or shps in general yeah so I mean just to be clear most of the institutions we deal with are crypto native or pretty deep into crypto systems so they're usually very familiar with per I think um perss are like a serious innovation in terms of they really solve a problem for Market that's fundamentally based on physical asset I you need to know that you can get a i Sushi token uh when you trade spot and that's instantly transferred to your

wallet so that like physical difference makes leverage very difficult um it also makes fungibility very difficult because if you're on different exchanges or different blockchains like those tokens might not be available it might not be bridged the bridge might not be healthy or be safe there's like a lot of risk there right perps are great because they just solve that say right all you need is an oracle somewhere to trade it bang you trade it and you sudden you have a return profile and an exposure to that thing um I think in traditional Finance it's not as needed like they have other efficient mechanisms that exist but I think like

crypto really runs on perss right if you didn't have perss I don't know how you would maintain rational pricing and fundamentally fluid market so yeah it's something I have talked to my Trad five friends about they get it once you explain why it exists and what it does they instantly usually say oh yeah that's really smart that's actually a useful thing to our work and see why it's helpful um so I was very funnny enough I you mentioned bit Max the first real trading I did in crypto was Trad bit Max Cash and Carry back in like 2017 you could

make I feel like I'm making up I swear you could make like 100 to 150% doing Cash and Carry On shitcoins on bitmax just running you know Bitcoin to bitcoin to alts and then have different tokens laid up against it it was a really good time to be a tough market so uh yeah yeah me playing around in those things for a long time what why do you think the derivative space in trafi grew in a Divergent Direction Why do you think then then crypto we evolved in this direction where perps became kind of core to this alternative economy why do you think perpetuals did not find

that same like efficient mechanism in trafi um because the history of the market is fundamentally different so like much as I wouldn't I don't want this to be the case it is the case crypto is a speculative asset class right people come to crypto because they want to speculate on the price of it um and perss are perfect for speculation they don't require like a rolling of the future you don't need to take delivery of anything it's easy the history of like tradire Futures Market it's all in hedging you if you were a farmer that

wanted to hedge your Corn Harvest you sold the corn Futures and then a guy that wanted to buy corn for Milling at the delivery of the Futures date would buy the future and at the end of that contract you wanted to deliver corn and get your money and the guy wanted to give you money and get the core at a pre-agreed price right that was what the Futures were designed for and so TR grown out of that history and so the liquidity and all the depth is in that and all of the system is built into managing those curves having these Futures contracts and they have official efficient mechanisms for say you

mentioned FX I was an fx Trader for many years um FX spot markets role they call it a tomx so the day before delivery you buy or sell your spot and then buy the day afterwards and you just keep roll ring it basically you can do that wholesale at very cheap prices like it's super efficient and easy to do but it's all built for institutions and fundamental people that actually want to take delivery of these things in crypto is a lot of retail who are non-experts so they you know they don't want to take delivery of a thing they just want if number goes up I make money if number goes down I lose money I understand what I get and so per are perfect for that

right there's no delivery problems it's all fungible and that one contract in that exchange just easy so um yeah I think that's the difference yeah makes sense when you're uh when your livelihood is on the line and your business is different type of uh speculation or at least uh risk management than when you're uh trading shikin yeah not sure I would have put it as colorfully as that but yeah that's probably the uh right I mean my my uh my family here asked me what the best use case is for crypto and without beating around the bush it's to it's providing a a a young uh young

casino with tons of upside for people there's a lot there's a lot of use cases I hate to be so pessimistic but here we are I mean there's definitely what take I mean if I was going to be a bit more optimistic I'd say like you I worked for years in large investment Banks where you know a lot of the systems were 20 30 years old lacked interoperability very inefficient um blockchain genuinely solves a lot of those problems you know like being able to trade atomically see the settlement happen on a public database have that be mutable recognized it will take a long time for

banks to implement such things but I'm convinced actually like a lot of Finance will move that direction it might not be on open permissionless networks as a lot of crypto Maxis might hope but there will definitely be change and I think there'll definitely be convergence where bits of tradire will drift towards crypto and bits of crypto will drift towards tradire right where people become less maximalist and see that there are maybe better ways of doing things so it's kind of fun to be in a young industry right where you can help form that be part of the change as things move forward yeah and there are certain Strate like you mentioned Cash and Carry

there are strategies that emerge maybe similar sort of maybe Arbitrage strategies that emerge as there's private blockchains with significant liquidity kind of parallel to the public blockchains with significant liquidity that we see now that could uh kind of pave a path for new strategies to to develop like again you mentioned Cash and Carry what are some of the strategies you see available in the market now and I mean there's as many strategies as there are crypto Traders right now right because I yeah if You' have said to

me if you'd have said to me three years ago I think the most efficient way for someone to get rich in crypto is go to as many random daps as possible click a few buttons deposit some money in and out and they'll probably get rich I would have just told you to shut the up and the truth is that would have been a super high alpha way of running your money for the last three years right with all the ups and downs and everything else if you'd have said when salana was on his ass buy a couple of salana and steak and gitos says off dude like what the hell that would have been a very plus EV thing to do you'd have made a lot of money

so you got you know I'm I'm probably too Boomer to see that and to like go after it right but it doesn't mean there's no value in it you know two years ago people were making loads of money flipping nfts um but you know like people consistently make Industrial Level money doing Cash and Carry very dependable what you're going to get see the money you deploy your cash you just make consistent returns it's like hard work right um You can trade options there's like a bunch of option strategies you can do there's trading you know we're integrating with pair protocol who are pair trading app you can do pairs

trading it's like a lot of money to be made in trading correlated Assets in crypto because there so many of them um there's just so many things you could do right but like on that uh go on on that integration D that will be able to trade uh trade coins against Bitcoin against e via pairs tooling on vertex yeah so they'll have an integrated tooling where you can come in um you can leverage our cross margin engine you press a button on P it executes the trade into Verte I think they have two Market going to be integrated with one is GMX I think they're up with already they're up with

us now they currently like running through U that integration but you know like that integration is big for people because they like that they can look at it on pair they can do stuff there that we don't have chart one asset versus another trying to evaluate how that correlation is operating etc etc the the nice thing with those sorts of Integrations is you could never look at vertex vertex just becomes a liquidity layer underneath other apps right so for us that's when we talk about composability that's what we really care about is how do we get more volume go rever you know you want to integrate

with the people that will give you the highest return to that metric um and as such you want to be working with aggregators THS doing lots of volume other protocols because then you become power you become multiplied out right pairs client base becomes out of client Bas you know at different margins so they might charge a small markup to then trade on vertex But ultimately we make our bit in the middle and then they might be implicitly trading you against another protocol is integrated so suddenly you've got like these Protocols are trading with each other really but they're doing it all through vertex's stat and that's kind of where we want it take things is to be

bit more like a yeah liquidity later as I said rather than just a yeah re recently we saw the the fee switch on Unis swap how do you think about fees when it comes to vertex and other other participants trading on vertex and kind of the fee landscape in the in the ecosystem overall overall onchain trading's been too expensive for a long time um even with the Advent of arbitrum optimism uh various ZK rollups like gas fees came down but protocol fees really didn't

right like if you looked at say I I'll reference GMX just because they were on um arbitrum when we started you know V1 had fees in the region of like 10 basis points plus there were like a bunch of embedded swap fees that UNL in the protocol you look at Unis swap it's typically it's five to 30 basis points depending on the pool to execute a cost to execute a trade if you look at I think synthetics are down to like three basis points now three to five basis points I can't quite remember where their fees are but like these are quite High compared to centralized exchanges it tends to be a lot cheaper

um so that's something we've thought about is you know for the long-term Health the protocol how do you charge fees that are as low as possible in a effort to grab market share because ultimately the increase in market share will more than offset what you've given up in fees you know vertex would never be a top three deriv de if we charge 10 bases player trade this would never happen but because we focused on people that do really high volumes I.E trading firms you know pruma type people that they turn their book over a lot those guys are never going to trade if it's 10

basis point of trade they just won't do it so you got to find other ways other Market segment's other clients to bring in and use your platform yeah yeah I want to get into the weeds a bit as to kind of how that works so you know your trading engine is kind of a hybrid model um you know I'd love to hear your opinion as far as kind of how your model differs and compares to something like an app chain model as long that's like it's on it's almost like it's on app chain but it's it's it's not so when a user deposits into vertex from arbitrum you know where do those funds go what's the trading engine and kind of how do you how do you how

are you guys able to enable this uh oneclick trading this very low latency engine um using your model kind of that the hybrid onchain op chain yeah so we have our own offchain sequencer and that services our contract solely right so in many ways vertex is like a really conventional de me not totally conventional it's got perks and spot and man markets but those all sit in smart contracts on arbitrum where things get a bit more funky on the technology side is we have this sequencer that sits above it runs an order book so you can send orders in there and if they match we send them to chain but it also allows us to like Fast

Forward chain state so so long as our sequence is running we can keep receiving orders and transactions from our users even if arbitrum went down short term or became very congested we can keep processing and we can do that really fast we've like Rewritten the whole evm in Rust so it operates super low latency it means then every transaction that happens we just fire it off to chain so from a custody and um a sort of transparency point of view in terms of trading it's exactly the same as doing an onchain a fully onchain protocol in terms of speed and

convenience is like 10x is more like a centralized exchange because you come in you given signing rights locally so that you've got like a key running but you can't withdraw so when you use one clip trading you just get this super fast process you press the button it trades but all your assets are still safe you know that everything that trades is printed on chain so we like to think it's a nice hybrid kind of compromise solution that helps people get the best of both worlds so then you must be pretty pretty bullish on the idea of app chains for perex is as well kind of iated execution environment so uh we like the idea of app chains I think where

we well we have a problem and my co-founder wrote an amazing thread on this anyone that's interested wants to nerd out should go and find alwin Pang on Twitter and go and how looks through his profiles he don't have that many threads but this one was very interesting and alwin's argument was well if you want to run a low latency setup high frequency trading generally operates in the region of like single digigit millisecond latency you need to be really really fast TR and the problem with a distributed chain is that any reasonably distributed

blockchain will need to reach consensus consensus will need to occur across a number of globally distributed nodes and those nodes might be you know one side of the world to another that can add latency in the region of like one to two seconds just because of the speed of light you can't travel fast enough to get to single digit millisecond latency and what that means is if you want to have super low latency blockchains then what you end up doing is collocating the servers in one place right we've seen a number of blockchains do that but at that point a number of

the nodes are probably controlled by the team they have a number of controls over how that thing set up and then you start to ask questions of like well what what genuine value are you getting out of this being a blockchain rather than just being a single node and the truth is you end up with all the negatives of being a blockchain with none of the distributed decentralized positives you were looking for in the first place and so there's like this expertise and need problem where for us we view that that was kind of a false narrative so for us it's like no if you want to be fast performant and

the faster you are you know you might say why do you need to be fast doesn't need to be that fast the reason it needs to be fast is because if you want Market makes to show good liquidity the faster it is the braver they can be about the orders they put at the top of the book The braver they are about the liquidity they put up the better price they will put and then the better execution it is for you as a user so even though it doesn't feel important if you want to have the best ux which means cheapest liquidity deepest liquidity you need to be super low latency and that's what our setup gives versat chain which you know might be pretty performant but it still might operate in the I think I heard the

guys from hyper liquid I was on a thing with jaff from hyper liquid his setup operates at like a 200 to 900 millisecond that's awesome for an app chain it's still pretty slow compared to our setup and it means that like Market makes have to just be a bit more defensive which is is understandable you know and it's how you feel about that tradeoff for us that tradeoff is not worth it for them it is that's a post unit yeah they have lower liquidity overall not good of an experience nothing on them it's just the the design choice and I know Rob's got something um just though the difference between the the single sequencer set that you guys

are are hosting versus a app chain with centralized nodes all in one Geo geographical location it doesn't seem that different but it's the it's the facade of trying to run a decentralized app chain on top of a layer two that kills the performance and speed so one can argue then that this need for assuming that that you you can generally agree with that then the need for their for Speed May outweigh the security of said uh execution environment that's a tough one to argue when it comes to such large amounts of capital what's kind of the what do you think about that

it depends how much security benefit you think you get from that distribution like just to bear in mind our setup is such that we can't touch your funds there's no ability for us to do anything in that regard it just means that we run the single node that operates the order book right so it's really about order book security rather than asset security and if you are compared to trust on the AUD book level basically we're asking you to trust us with the order book in a way that that sort of system isn't we think that that is a decent payoff because you have the

security of knowing that if we were playing around with those with that order book setup probably all our market makers would leave yeah I was going to say how do we know that that verx isn't Co on a bitx and for a uh Haze and Gathering all the orders and liquidating me yes it's really it's like a social enforcement right where like market makers are making and if we were reordering their transactions they'd see it immediately in kind of performance times and have things operated they'd stop providing liquidity it's also that like we run open liquidations rather than a closed liquidation system that say a bit Max would run so like anyone can pick up liquidations and execute them into our

environment which means that we can't Jazz around with that either um because we also run our liquidations versus external Oracle we don't run them versus S order book so it's not just like can't just pick a number out of our ass execute it and make that thing happen it has to execute versus the external Oracle there's a number of safeties there where what it then ultimately comes down to is do we execute the orderbook as we said we would you have a bunch of professionals that look at this in like real time all the time that would leave if we didn't that's kind of your safety mechanism and then I and sorry for all the just C questions bra just got this last one how do we as users and speculators on token or on vertex token

or users of The Da how do we know there's no preferential treatment for any specific market makers um interesting as far as information passing like hey you know or who gets the information first or you know yeah yeah so again it's like a social proof type thing I mean you you can never know even with some of these like let me put it another way if you trade on chain you're familiar the concept of me right the risk of us doing something preferential is on a time frame that is lower than our like latency round trip

cuz the the amount of reordering we can do can only happen in that like 5 milliseconds 5 millisecond just on the order of 10 milliseconds if you're trading on like pure on chain environments you are open for like a block or two right depends where your order gets picked up and executed so that might be over a course of like a number of seconds if not minutes so the value of risk is much much lower and as I said it's kind of being monitored by others so if if we had a preferential deal we'd have to give a preferential deal to everyone at the

point where you you give a preferential deal to everyone it's not a preferential deal anymore so like it kind of argues itself out do you know what I mean almost yeah fact that we're controlling that node is a bit more security than being in an open network where it's just a bit of a dark alleyway and you might get mug yeah it basically seems like the trade-off from security to speed is is not necessarily A trade-off in the sense of you lose one or the other but by opting for extreme speed the ability for M or for information passing to happen within the time of orders being placed uh is so small that it's nearly impossible to pass that information in any sort of way even with the fastest Bots or information passing you know no

but also the thing is if we did start doing that it would get picked up by the market makers right you basically got the like smartest guys in the market are monitoring this to check that you're not around with people so it's pretty rigorously enforced it's just a social enforcement it's not enforced at the chain level I try to make sure my stops with safe you know what I'm saying yeah can't say that you can't say that they're safe on all all per dexes from what I've heard about your Trading starts the last 48 hours um yeah I I think I mean you guys already picked up and and kind of in the pocket of like when the information

is corroborated like when it when it is pulled together in the order book and that it's ready for like sequencing onto the chain for settlement that's kind of like the part because you you spoke and I find this fascinating about like the the exec like the way that you guys are able to arrive at best execution uh because it's so latency Market makes are able to provide their best pricing because they can take more risk with the lower latency then the sequencer picks up the orders maybe can you like talk us through just like how powerful it is to have onchain settlement and you already

kind of went over this at a high level about how you're rather resisted to me like what what other components and value do you do you collect by posting these orders that get matched in the order book directly to the chain so just to be clear on the mechanism so the I think to be clear probably market makers appetite for risk is always the same pretty much what the low latency does does is it means that that that risk concentrates in a smaller amount of time so they can show bigger orders

right if you a bigger order out for one second I'm just using this as a rela is like the same as a much smaller order out for 10 seconds right in terms of how much they can reasonably expect they might lose so the more that you draw down their latency level and their cost of placing and replacing orders the higher you know exponentially they can get higher and higher to how much liquidity they put at top of book so you know if it was instantaneous then they could really show like almost infinite liquidity right because they know like they just pick what they going to trade but you never going to get instantaneous so the smaller you can get in terms of time the better it is now in terms of

like the relationship the sequencer to the chain and what's powerful about it is because we operate that ourselves all orders that come in operate sequentially right so order one comes in that takes priority over order two then order three then order four and we do what's called like time price priority how orders operate on like a conventional centralized Exchange in traire but operating at true time price priority without any preferences to anyone means that again we remove very ability for market makers because they know for certain they will get executed

at a certain level if their order hits in front of the other guy they know that you send it the only risk they have is on the latency rout because if their order got in before someone else did it it's whether someone set an order within that like five to 10 milliseconds that'll determine whether or not they were wrong if you like right where if the latency is 1 second or 10 seconds or 30 seconds your risk becomes a lot more and if in the midst of that someone could Meb I you were in front but then they say nope not anymore they put you behind and now this guy executes first then you execute then someone else behind you that can really mess you up right so just by being consistent fast

in the order predictable you add a lot of value to those guys that providing liquidity and then the guys that providing liquidity can be bra we going to show better liquidity for Traders and that's a virtuous circle got it so there's in the five millisecond of latency there's several orders that come through the system in these batches if you will and the the these kind of like these batches of orders are what then get sequenced by the vertex sequencer they're not even batched right they the continuous processed the only thing that's gets batched is when an order matches so you

might say like I bid at 100 someone offers at 101 someone bids at 10.5 no trades have happened right we just retain the orders at the sequencer level at the minute that you say I'm 100 a half bid someone else says I'll sell at 100 a half trade happens when that trade happens it gets sent to the chain and your balance change on the other person's balance changes that's how it works the rest of it we maintain a state on the sequence and the reason is market makers and Traders might send millions of orders a day in terms of gas if you printed that all on chain would just add up and make

the whole Market I was wondering that it's GNA cost you a freaking load yeah it would cost infinity right you wouldn't be able to run the business so that's why we keep that piece offchain that's what we view as like the low value part right placing and cancelling orders does not create economic difference to you or and doesn't matter you don't need high security for that you don't need blockchain level security you need blockchain level security when I say your money's moving Andy to Robbie now you say okay yeah I I want that to happen on chain so that signed transaction you've done we send that to chain and then the changes happen in the smart contracts and you to see the

assets move on chain which is really what people should care about yeah they it I I find it fascinating to think about the the state of the blockchain as this array of balances in people's wallets when things change when balances change like you said like someone's money goes up the other person's money go down because this trade happened that affects the state on chain but then there's also the state of orders that exist on the vertex chain if you will with like in very rough terminology there's a state of the order book and this rests on vertex's order book and the vertex sequencer maintains the state of these orders it's just it it's like I'm coming

to realize kind of like the the uh state of of orders that happen offchain and then this is very different from the state of balances that happen onchain 100% yeah yeah 100% And it's like you're saying this it's all of the important actions that actually drive economic almost like transfers of value that that require hosting on chain because those actually impact the state of the chain whereas when I set a stoploss it's not like I'm I'm making an impact on somebody else's balance or on my balance correct nothing's happened at that point right you just set an intent basically and so that part that part doing that using using vertex is order

book and mechanism is kind of what separates the uh you guys from the ones that that which what is why the experience if we're going to Shield ver why the experience is fantastic now now obviously you have to understand the you know the trade-offs of a single sequencer and there's kind of a centralized Point affiliate which I'm sure you guys have thought through that many of times and we can kind of dive into that at a at a later time as we're kind of wrapping up into into that but um you know I it seems like that that kind of methodology and and thought really you know yeah you can onboard huge market makers yeah that's it you know yeah and you know like we think a lot about how to maintain uh stability if you look at our downtime since we launched has been

almost zero um and that's because we basically designed the whole system not to have failures at that sort we have a lot of failover within our architecture so yeah it's a good question but you know people that have used salana and I am1 uh will know that blockchain is not necessarily a recipe for no down time so but the uh it would be somewhat of a of a catastrophic event not just vertex but all uh all rollups all kind of sequencer based operations with when you have a Mis A misalignment of State uh yeah it's it's bit of it's bit bit of an issue

yeah so um because we maintain our own State interestingly we're like less reliant on arbitrum in that sense than other protocol actually we had our own kind of fail over if arbitrum were to fail we can keep sustaining state within vertex until AR were back up and running were able to like realign state so yeah they gone down much recently man 2021 arit kept freaking going down that team works a lot on you know robustness and fail over and I think they do a great job you know work with them a lot so yeah I think they had no doubt and that fail safe is uh you guys

keep keep the the offchain state of the orders as well as records of the onchain state of the balances in case arbitrum were to go down corre so yeah the the orders runs without arbit from anyway that never touches chain so that bit we sustain but then the yeah onchain State we maintain as well and so if chains unable to pick up State we keep state in our sequencer until arbit want to come back up and then we fire all this down we actually do that because we pay gas we actually do that if the chain ever becomes congested which does happen periodically right and gas fees might go astronomic we might pause for half an hour on withdrawals or settling transactions on

chain until chain activity diminishes and then we set send all our transactions off so um yeah it's a mechanism we thought about a lot and yeah we keep a line but it's very powerful actually it's kind of more complicated than it sounds but it makes you out yeah it's interesting well appreciate the answers to the tough questions we wouldn't be doing our uh our duty in this world if if we weren't asking the tough ones and I'm sure there's either tougher ones that with a technical lens that we could get into but as far as today goes then I appreciate the uh the the time and the uh the interesting Dynamics from your perspective from uh the background and also on the technical side so um loving

the platform and um keep Crush at man pleasure I'm glad to be here thanks for having me thanks for coming on Cheers