Fogo is redefining high frequency trading on Solana

0:01I'm Ashen Addison from the Cryptocoin Show and today on blockchain interviews we have Doug Kolkit, initial contributor to Fogo. Doug, welcome to the show and thanks for taking the time. Hi Ashton, thanks for uh thanks for having me. I'm excited to be here. Yeah, excited to talk about Fogo SVMs, Salana ecosystem, uh what this layer 1 is doing, why it's

0:24unique, uh why it's needed, uh institutional, DeFi, Tradfi, and so much more. I would love to kick off our conversation by first of all hearing a little bit about yourself uh in blockchain and and how you got interested in contributing to Fogo. Then we can talk about all those details. Yeah. Yeah, definitely. So quick background on myself. I spent uh most of

0:46my career as a uh high frequency trading quant in tradi. Uh started off at at Citadel uh when they were kind of spinning up their early HFT group. Was on a number of uh smaller teams trading different trads equities and futures around the world. Got into crypto uh in DeFi summer. I was I was a me searcher uh for a bit. So I was on the trading side. um did that for a bit and then

1:10started uh my first project ambient which was uh originally still is a uh AMM an Ethereum an EVM based AMM um we kind of built that out we saw some of the limitations uh from that model some of the limitations from general purpose blockchains uh as well as kind of some of the limitations of just spot versus like perk trading and and kind of an AMM you're limited to to certain types of

1:34markets so uh I met uh met my Fogo co-founder Rob uh almost a year ago now and he kind of had this vision for a L1 that was built uh built for trading um and so built for kind kind of the what you would need to get minimum viable level to compete with with Tradfi or centralized exchanges on a blockchain uh but as decentralized as possible and and from the the ambient side we had kind of

2:02been looking at different chains and like kind of what would we need and and Fogo kind checked off a lot of it. Now the biggest thing was SVM instead of EVM, but even that like was kind of kind of shifting uh over time. It was kind of clear that like a lot of activity was moving to Solana. So that's uh that's I kind of joined joined the project. Uh co-founded it with Rob and uh Mike

2:24Cahill or the the Pit the Doro team were also behind Pith and uh yeah, we've been we've been building ever since. That that's a great intro and I I think not not as many projects have really thought this through but we need to have you know trading is one of the most important things currently in blockchain and moving digital assets around and there's so many dexes that are like what

2:50layer 1 do we build on? You know there's nothing perfect. It's like, well, that's because we need to build a layer one that's dedicated for trading for a great decentralized trading experience, especially if institutions going to come come along. So, I feel like this is something that's needed. Yeah, absolutely. And I, you know, I'd say block blockchains like originally

3:09going back all the way to like Bitcoin, like blockchains were not built for a lot of the conditions that you face in in trading, right? So, you know, the traditional use of like Bitcoin is right, it's transfers. You have a certain amount of assets in your wallet. I send to you, you send to me. Right? These types of transactions aren't super timesensitive or super order sensitive.

3:28Uh you know, if you know, you delay mine a few seconds, it's not the end of the world. If you put someone else's before mine, again, these are just, you know, we're sending money to each other. And so, right, the traditional blockchain model is great for that. What's really difficult though in a trading context are the we're trying to trade value like

3:44against each other from different assets and these values are fluctuating all the time, right? Right? And like these are why high frequency trading firms exist, make billions of dollars a year because they're very good at this. Um, but blockchains, you know, by and large the blockchains we have today just don't move fast enough, don't have kind of some of the guarantees around that

4:04where they're just a ton of games and therefore like a ton of value that gets extracted and and to your point, right? like institutions even ordinary traders don't want to come in but institutions definitely don't want to come in and you know deal with things like sandwich attacks or uh you know just like bundling that that type of stuff that that we've seen. So definitely I'd love

4:22to know more on what kind of pieces you put together with the co-founders on okay what do we need to have a layer 1 that is dedicated and efficient for high frequency trading. Obviously like the finality of a transaction is important but there must be so many other pieces. Yeah. So I mean like you said the big big one is definitely uh definitely latency. So uh you know one of the

4:50limitations with general purpose blockchains are their the validators are or you know the networks global globalized. Um you know what we realized early on is that even if the best software in the world there's just limitation to how fast uh the speed of light goes. So one thing we realized is okay like collocation would be important uh not only because of that but also

5:09because collocation is is predictable right so if I'm an institution I'm trading you know I want predictability about how long it's going to take for my trade to hit the market um all that stuff so we we built Fogo to be colllocated um but but what we do is we call it multilocal consensus where instead of trying to move all the messages between different validators we

5:31bring them all together at one place which is great does a lot of performance, right? Obviously comes with centralization risk. So the way way we deal with that is uh it's called multilocal consensus and makes it very easy within the network to move from collocation center to collocation center or even fall back to just fully global consensus. Obviously it's slower but you

5:50know that's that's a good fail safe. So you know things like a given data center goes down somewhere. Obviously the whole blockchain doesn't go down. Um so we think it's actually very resistant. it can move anywhere in the world in a kind of a matter of seconds when when if things ever do fail which you know hopefully in a happy case they don't or if there are any issues. Uh so so that's

6:09one thing. The other thing is around uh there are just kind of abusive behaviors that are uh just kind of the nature of block production. So right like in traditional financial market most things are are are first in first out right and that gives you a lot of nice properties. So you send a trade uh your trade gets executed first because it reached it

6:28first. And what's really important is someone else can't see your trade, hold it back or you know put their own trade in front of your trade, wait, delay their trade, make their own trade go faster, right? There are all kinds of shenanigans you can play, right? When you when you can do that, right? And and we see this, right? Like when you have blocks and uh obviously the longer the

6:45blocks are kind of the more shenanigans there are to play, but you know even even in small blocks there are shenanigans people can play. uh rearrange things that doesn't exist that doesn't exist in uh or exist to a lot it's a lot harder in in treadf uh I'll say that um so when you have this concept of of a block in a blockchain and you have the leader who kind of has

7:05total freedom to you know put the transactions in a block however he wants or including leave out transactions uh right that's that just like makes the kind of the level of trading that you know institutions and and tready requires just it's not possible. So at Fogo, we we've done a lot to put rules into uh you know both at the social layer and the technical layer what

7:29validators can and can't do um in particular around certain transactions being prioritized against other transactions. Um and as well as right we have this concept of a of a curated validator set meaning if there's some abusive behavior by a validator um you know majority of the stake can can kind of kick them out. So you can't always do things completely at a technical layer,

7:50but like people are very very clever when there's money on the line, but we kind of have this social consensus fall back to to kind of police abusive behavior. Yeah, it's very interesting. And I'd heard these stories in Tradfi with high frequency trading about these firms that, you know, they would move their servers, you know, from the other side of town to down the street and like

8:13that would help with the few milliseconds faster. Um, and this notion of, you know, localization, although it's distributed validators, is that sort of copying that Tradfi idea of like get the servers close to make it faster? Exactly. Exactly. And like what what uh like you're talking so what what actually you you typically see the evolution in in Tradfi when they go

8:35electronic is a lot of times the exchanges don't initially have this collocation and so like the exchanges data centers here and then like you said people go around town and try to find the fastest connection. Then the exchanger realizes well wait like this is you know not great for our customers um you know why don't we have our own data center and and then the next

8:52evolution of that was people would figure out which rack in the data center was actually closer um and then start doing that and so then the exchanges did another thing where they said okay everyone has exactly the same length cable uh so you know even if you're next door you're going to we're going to give you 100 foot cable just to make things uh equal so yeah I mean you see the same

9:12thing with uh you know distributed uh validators Except the problem is the infrastructure costs are way more if there are you know Salana right the validators are across probably 100 different data centers and and just imagine kind of the the level of cost that would need to kind of be near all of those. Uh so yeah, we just we just said it makes sense. So we're going to

9:31go jump to what we know the end state is and uh yeah, when you have when you have a data center like that, uh you can go to a trading firm and trading firm can say, "Hey, this is something I'm used to dealing with. This is very predictable. Uh I know exactly how long it's going to take me to get my orders in into the network and you know, this is I'm a lot

9:48more comfortable with this and therefore willing to you know, show a lot more liquidity uh you know, put trade a lot more and kind of all that other stuff." Mhm. You mentioned in your history working on an AMM on Ethereum virtual machine. A lot of people that are just getting into decentralized trading are familiar with the DEX is like unis swap and they know that Ethereum as a

10:09blockchain, you know, it's it's not instant confirmation and there are these issues. And then more recently, we've seen more trading on the Salana ecosystems. uh specifically with memecoins, although they're a great sandbox maybe for institutions to see, you know, hey, we can trade fast, we can trade cheaper, can you really nail down the the difference in just creating

10:32another deck on Salana and utilizing the speed and the and the cost of of Salana versus actually creating Fogo as a layer 1 with the SVM? What are the main differentiators and advantages between just having a regular DEX and a whole layer one? Yeah. Yeah, that's a great question. So like one one of the you know the core issues with you know the EVM or Ethereum especially the layer one

10:58is like you said the block times are just very very long and we were early on very advocate very strong advocates of of lowering block times and I think finally some of the Ethereum people are coming along to this or at least they're no longer talking about lengthening block times which was a little bit crazy for for a minute there. Um but yeah I mean even in an AMM context what you

11:18find and we've done a lot of research on this is that longer block times uh basically going back to that kind of block builder has like this total power over over the block uh the longer the block time is the stronger their power is and and the more they extract from the system actually. So even from an AMM context, uh shorter block times are just a lot better lot better market quality

11:40and and you know I'm sure sure a lot of your listeners are are probably familiar with kind of the results showing that you know LPS in AMMs or unis swap often times don't really make money because the yields they get are not compensated enough for like the volatility that they're taking from the assets and then that's been a big problem for a while.

11:59Um and I think one thing with Salana why they took off is because you know the block times were shorter and I think kind of that those problems were not as serious. You look at Salana AMMs um you know uh the returns are uh you know if if not positive a lot less negative and and oftentimes positive depending on the coin and how much activity there is and

12:19so FOGO right like we take it a step further. So Salana block times are you know 400 milliseconds. Uh right now in test net we're running 40 milliseconds and and I hope by by mainet uh if if not shortly after mainet we'll be you know even even shorter than that uh because of some of the collocation and scalability things that we're focusing on. Um and right and I think right that

12:41just improves market quality uh even even a step further. So, we're uh we're excited about that. And I think with the a with the exchange itself, the DEX itself, there are other integrations we can do being this close uh to to the L1 itself. Um and and particularly that's around how certain transactions are prioritized relative to other transactions. So, one problem with

13:05general purpose blockchains are they're not transaction aware. Um, and so, you know, in a club context, this becomes particularly important because market makers are out there, they're putting out quotes. Um, prices change very quick. Binance, you know, moves very quick, they exchange. Uh, market maker needs to adjust their quotes. Uh, as the price moves at Binance, if the

13:25blockchain isn't aware, what can happen is, right, people can pay a higher priority fee, get their their their trades in to hit kind of that stale liquidity, and and the market maker is taking a loss on that, right? And the people who are benefiting are are very small trading firms who are very good at trading fast. And the people who are losing are obviously the market makers

13:43themselves, but ultimately the end users who don't see as much liquidity or more expensive liquidity because you know someone has to pay for ultimately pay for uh kind of that that extraction from the system. Definitely. And you mentioned there uh in the test net having almost 10 times faster millisecond finality than Salana sounds really great. And at the beginning you

14:05mentioned last year with your co-founders. Can you talk a little bit about the timeline in where it's at right now? How long does this kind of platform take to develop and sort of what's the road map to the main launch? Yeah. Yeah, that's a great question. So, uh, you know, the biggest thing is we're not we're trying to use components here, uh, that exist where we can. So, we're

14:28le leaning very heavily on the the fire dancer client. So, we kind of made a decision early on to do a a single client, uh, single client system and and to go with the fastest Salana client, which is technically right now the the Franken Dancer client. Um but when fire dancer is fully ready will be the full fire dancer client and kind of we've seen scalability uh in terms of

14:50throughput and and latency uh improvements from that relative to the standard Solana client. Uh there just have been modifications though we've had to make even though relying on existing software uh you know it's not as simple as forking things because we need higher reliability, we need higher throughput, we need a lot lower latency. No one's ever really tested the Salana software

15:12uh at kind of these levels. So, we're, you know, we're finding a lot of things out uh testing things, just kind of improving improving the network. We've had our uh I'll call our infrastructure builder test net up since since March 1st. Uh our public test net with uh some apps deployed will be out in in July and then we're, you know, targeting uh pretty firmly targeting mainet uh by the

15:36end of by the end of Q3. That's great. Looking forward to it. And you mentioned there the the fire dancer client is that like uh a dev environment for for Salana. I'm I'm not familiar with it. Yeah. Yeah. Yeah. So right basically um the right so or any blockchain network. Um you have you have nodes and like a blockchain is just a protocol. So it's a set of rules. Um and

16:00you typically you know you have nodes that participate in the protocol of validators or or just regular nodes and they run uh you know certain software. So for example in in Ethereum you have about five different clients. So uh you know there's there's one Ethereum but there's you know uh five different clients that you can choose to run. Um in Salana right now there there's

16:23traditionally been one client the standard Salana client which is now Agave. Um, and then Fire Dancer has been a project that's uh been funded by Salana Foundation to actually rewrite the core Salana uh the core Salana uh client from from the ground up actually to to just be a lot faster. It's being developed by uh Jump Crypto and and some of your listeners may be familiar. Jump

16:46is just a huge trading firm based in Chicago very used to writing low latency code. So kind of the performance from fire dancer should be significantly better than kind of even the fast performance from agave. Now the now the problem is is when you have multiple clients in the same network you can really only go as fast as your slowest client right cuz uh if one side can't

17:05keep up uh right like the network becomes unstable. So we thought early on that it would make sense to just do fire dancer 100% from the start um without having to worry about supporting slower clients and that allows us you know the network to move faster. And these are just kind of interesting things you can do when you're you're starting a network

17:21from scratch versus you have to support existing infrastructure. Mhm. Interesting. And with the Fogo being on know SVM and a Salana ecosystem, how important is the underlying tech for Salana and and Soul as a digital asset? uh because do you have to use soul as a transaction gas fee or transaction fees and like right now you know Bitcoin and

17:49Ethereum with the ETFs there's been a lot of institutional interest we haven't really got to that point of sole ETF's spot and and institutions gobbling it up as well do you think that I think that's inevitable but do you think that that's sort of a precursor to more institutional interest trading on Salana based chains Yeah. Yeah, that's a great question. So,

18:14to be clear, right, Fogo is Fogo uses the Salana technology stack. It's not it's not an L2. Um, it's its own L1. So, the the Fogo token is uh the the gas token and and the core token of of the Fogo network. Same way Solo is the core token of the Salana network. So, so it is a gas token. Um there'll be some interesting things we do with with the DEX and how we use the Fogo token uh you

18:36know as collateral as well. Um so yeah from from that perspective it it is its own own token. Now we do we do should have wormhole support on on day one. Um and obviously there's a lot of liquidity connected to Solana. So we do anticipate a lot of Solana liquidity um on Fogo as well and are excited excited about that. Um in terms of yeah institutional stuff

19:00I think the biggest barrier we've seen or at least personally I've seen with SVM stuff is banks move banks move slow um you know or financial institutions in general but especially banks move slow and to get stuff through compliance departments often takes six layers of bureaucracy um and just you know years and years I think banks so far have mostly built out their system for the

19:22EVM uh mostly because right like four or five years ago EVM was the default uh thing. Uh I think they will and they are building out for the SVM because at the end of the day the reality is like institutions are great but institutions want to be where the end users are. Um trading firms just don't want to trade and banks don't want to trade with each

19:41other all day long. They want to trade with regular retail users. Um so they are I I think they are building out SVM systems um integrating with the SVM kind of all the compliance tools they take. that will just be a cycle. But I I think we'll start seeing institutions come online in kind of SVM based worlds uh you know pretty soon. I think FOGO is pretty well positioned because again we

20:02can we can police a lot of abusive practices that you know financial institutions are just not comfortable with uh you know in in open fully open networks like Sloan. Definitely. And with sort of these different blockchain silos in DeFi, you know, the TVL on on Ethereum has always been the the just because of the network effect and and the length of history, it's it's been

20:24the highest and and Salana has been going up a lot. But being able to transfer value fast, cheap between different chains, how do you anticipate some TVL from the EVM ecosystem moving into Salana and moving into Fogo? Yeah. Yeah. Yeah. So that's a that's a great uh you know there's another great question. I I think the way I think about it is Ethereum is a great network

20:49for keeping assets or holding assets. Um Salana is a great network for moving assets. So Ethereum has a ton of TVL because there's just a lot of capital that's used to parking it and leaving it there kind of re re the applications you see that are successful like restaking uh maybe some lending stuff those work because those don't have to move around

21:11that that quickly. Um so you know people are very comfortable with the Ethereum network. It's very decentralized. It's very secure by this point. But you mostly do things where you just, you know, click a button, park your capital and, you know, come back and, you know, whatever, six months later. Uh, Ethereum is not a great trading network. Um, some

21:28of the L2s I think have made progress there, but still not obviously Salana is by and far ahead. Salana is a great trading network. Uh, I think it's TBL is comparatively lower because people don't move to Salana to just keep their assets parked. They move to Solana to to trade. And at the end of the day, blockchains generate revenue when people take

21:49action. So if people are just sitting on assets all day long, that's great. It's nice little number for DVL. It's uh it doesn't really drive much value to the to the token itself. Um and so yeah, I think same thing with Fogo. Like we are a trading focused chain. Uh you know, reststaking a lot of these other applications. We're we're open. We're general purpose and and we might you

22:08know, have have some of those applications, but we're really focused on on trading. And so kind of our northstar isn't necessarily TVL. It's more things like volume, uh, you know, revenue to dexis, kind of kind of these things that are indicative of of trading and and activity. Um, so yeah. Yeah, it's a great point. And with high frequency trading, I feel like managing

22:29the volatility and of course having stable coins as a digital asset is important for any DeFi uh, experience. What do you think about the stable coin ecosystem right now and on Salana and and the legislation that's coming out for stable coins? I feel like that would push a lot of more institutions to be more comfortable holding stable coins and moving in and out of them on FOGO.

22:55Yeah, absolutely. Absolutely. And uh you know we are uh you know we're finalizing kind of our our stable coin strategy right now talking to a couple of different providers but but basically users can expect on day one of FOGO there will be a a canonical stable coin. Now of course people can bridge over whatever they want but but the DEX itself will support you know a single

23:14stable coin kind of as as the collateral uh the ecosystem will kind of coales around what uh what stable coin we do end up uh using. Um and so right I think like making a very clear uh stable coin inside the ecosystem like you said is super important for for institutions um they mostly right these are dollar denominated uh firms right they most they're trying to earn profits in in

23:39dollars they don't want to hold their whole balance sheet in uh you know whatever especially a longer tale newer L1 token um so yeah I I think in terms of that just more and more people are becoming comfortable with stable coins and that's been one of the big barriers to get larger firms on on chain. And it's still kind of pretty shocking how many how many firms that are even in

24:00crypto that are even doing you know very well in crypto big market makers at Binance um you know buy bit whatever are don't touch stuff on chain because they they kind of right like it's not worth them risking their larger business for kind of some of these regulatory issues. Um and so once we get clarity especially around stable coins I think it's right

24:18inevitable that they're start going to start looking uh onchain and start moving in that direction. Yeah I'm looking forward to it. And with the road map that you mentioned towards the mainet launch hopefully at the end of Q3 is there any way for people to get involved or you know test out before the launch and like to you know get there for make sure that they're there for day

24:40one? Yeah definitely. So I first I I'd say is you know follow follow our socials. Uh Fogo chain is our our Twitter handle. Uh we have pretty active pretty active discord that should be linked there. Uh so definitely I think we have a pretty good community getting built uh public test net uh where we'll have applications for users to play around. Um will be you know in in July

25:05should be early early July. So if if you follow our socials you should be kind of ahead of the curve on that. um actually get to try out Fogo and and see for yourself kind of some of the performance and for builders uh we have a a really active active ecosystem now and I think like kind of my one pitch for builders is definitely come to Fogo uh you know

25:24I'm an application builder myself a lot of the people on the team are uh you know we have fantastic support uh for applications and and we really you know are there to make make our application builders successful so uh definitely reach out uh either through the discord or or kind of to me directly um or anyone on the team directly and yeah excited definitely here to support

25:48support builders. Yeah, that sounds great. Doug, I will leave the link to the socials and and the dev stuff in the show notes below as well. I'm looking forward to checking out uh the test net and when we get closer to the mainet launch I would love to have another discussion on this and I really do appreciate your insights into Salana ecosystem high frequency trading uh when

26:09the institutions are coming and all the other great topics that we talked about and thank you so much for taking the time today. Thanks Ashton I really really appreciate it. Thanks for having me.