Ep 256- OneChronos Co-Founders Kelly Littlepage & Stephen Johnson

[MUSIC] >> Hello, Mr. FinTech newscast, my name is John, and with me, as always, is, and just for this podcast, I'm going to call you other Steve and we'll find out what about that in just a second. >> It's so hard, it's so hard to be the other person in the room or the other, it doesn't feel good, John. >> Steve too? >> Is that better? >> Steve too, Steve too works, Steve too. >> Yeah, old Steve, I like that one. >> Old Steve is a lot better than old Steve. >> I'm technically a Steven, so we could, we could not, and that's our, that's whatever guests for this episode. That's Steven Johnson and we also have lucky enough to have Kelly Little Page, the co-founders of One Chronos. Welcome to the podcast. >> Thanks so much for having us. >> Yeah, yeah, thank you, before we jump into everything, just a little bit on your take on the markets these days. >> Just a hair of volatility and unknowns going forward. >> The Fed isn't raising rates at least as of this week. What's your take overall on the world of the markets these days? >> We're so far on the market microstructure side of it. We don't really have anything interesting to offer on the global macro side. We have seen significant increase in trading volume for the general market and for us. But as far as which way the world is going, if you have that crystal ball, we would love to see it. >> Yeah, yeah, yeah. Yeah, actually, that's Steve's answer always to me is that I'm so talented, I'm going to do well no matter what. >> Exactly, that's power through it. >> That's what I call power through it. >> I don't care, I'm going to be fine. >> Yeah, yeah, well, you're certainly building some interesting tools for the market. Can you tell us about alternative trading platforms and how you got started in this world? >> One chrono, as you said, it's an alternative trading platform in ATS. We match trades for predominantly institutional traders. And the thing that differentiates us from there's a lot of ATSs and there's a lot of exchanges. The thing that differentiates us in this market is that we do so via something called a combinatorial auction. And that's supposed to more typical limit order book dynamic that's used in other trading venues. And we got into this world coming from the institutional trading background and also coming from an academic auction theory background and releasing the opportunity that there was to give institutional investors tools that were more directly aligned with their high level portfolio intent and would allow them to reflect the views that they had directly into the market. That's how we got started building one chronos in 2016. >> You're working with a person who's both on a noble price and an Emmy, which is crazy, by the way. Can you give me an idea what exactly is a combinatorial auction? And how does that differ from the way that things have been done in the past? >> The word combinatorial in this sense means packages of goods and it can also mean compliments and substitutes. So we're talking capital markets and trading, popular compliment that people might have as a Paris trade. So by X and sell Y, and X and Y might be stocks that are involved in some merger arbitrate or some corporate action, or they might be completely unrelated as companies but correlated on some level, so relative bets. And the way the world is wired right now is there's no way to ensure ademissity for those types of transactions in a typical exchange or ADS trading setting. The commenterial part of one chronos is that you actually can input bids and they say can because it's optional, we also take trades just like any other trading venue does. But you can input bids that say something like I went to buy X if and only if I can sell Y at some ratio. >> And that's really where the commenterial auction is at its core. >> And is that, it's not only for two things, so I want to sell it on a buy or can you actually add other things? Can you add other variables to the strategy as well? >> Yeah, so Steve, with all that trading jargon? >> That's right. Yeah, so it is scalable to packages of pretty much arbitrary size and composition. So right now for US equities, we're only live with the packages of two, it's called pairs. And a lot of that is because it's just a common style of trading in US equities. There's a lot of scenarios where institutional traders are interested in rotating from one stock or another or have correlated views on two different stocks. For example, if there's a merger about to happen. So the concept though is really scalable to whole portfolios. You could insert a bid that looks like I want to move from portfolio A to portfolio B or portfolio D, EF, etc. As long as I only get one of them. And the idea behind that is that if you give people the ability to express their full intent instead of trying to go one piece at a time, they can potentially price a lot more aggressively or they can offer larger sized orders. And one analogy I like to use on that is if you're making a recipe, say you're making pesto or something, you go to the grocery store, you pick out the olive oil, the pine nuts, salt, etc. And it turns out they don't have basil. So the key ingredient is missing. You can either put all the items back and give up and you just wasted all this time going to the grocery store or you have to find another store and risk that maybe they don't even have basil. Either maybe this is some shortage that you didn't know about what the what combinatorial auctions. So this exact same concept exists in trading where you don't know exactly how much it's going to cost you to execute each of the different pieces of a basket until you're already done executing them. In current market structure. So what this gives people the ability to do is say like I want this all or nothing so they're not left hanging with, you know, unfilled pieces of the recipe, so to speak. So you are coming at this. It seems like you both have backgrounds in in Quant rating and in consulting. What sort of market gap did you try to address with this product here? Yeah, so when I started out my career, I was focused on the portfolio manager side of it, the question of how you generate alpha for this portfolio. That's kind of what I was brought in the due and then I got very interested in the question of rights. So once you have the alphas, how you reflect them into the market, which is the execution, triggering side of it. And I had just come off of my academic background was in a mix of computer science and auction theory and math. And this was the problem that people were talking about for a very different asset class, which was FCC wireless spectrum. So it was kind of clear that the same issues that come up in wireless spectrum, investors or in this case, people building networks had the exposure problem, what Steve just, Steve just described, not necessarily being able to complete packages of goods and then having to think very strategically about how they've been in auctions while worrying that they wouldn't be able to complete that package. There's a very direct analog to what happens in institutional investing where you're running an optimization for your portfolio based on expected transaction costs, but expected is definitely average and backward looking. So I said the real power for institutional traders would be if you can reflect your objectives directly into the market and not have to try to model out or think about historical execution costs. So that's that's really what got us started talking about one kernels at a high level. The unfortunate reality of this at the time was what we were discussing is combinatorial auctions weren't computationally possible at the speed and scale of capital markets, meaning that we had talked abstract way about how to do them, but the time that wasn't possible computationally to actually run these types of markets. Got it. And then did you come across Paul and I said there was more academic or more base around trading and how to use what what what he's built for the financial markets? 100% on the academic front. So I first became familiar with his work when I was getting interested in auction theory as an undergrad and he get rid of a very famous book, the standard textbook for this called putting auction theory to work. And the way that he presented the challenges that auction can solve and the type of beneficial outcomes they can create in markets was this extremely compelling. Got it. Okay. So that that's sort of the the way in which you you get you're going to this. Yeah. And I think we got to pick up our our conversations. I mean we were just talking about what we're going to have for lunch and you guys are chatting about common putting combinatorial actions into into the market solving holes. I think we got to step up right now after Steve. Right. Yeah. Let's go have some Italian after this and we'll check pine nuts. Yeah. Absolutely. It makes sense. Yeah. And and you you've actually so you've built us also the you you're trading upwards of you mentioned four and a half billion dollars daily. Is that is that the number that you're at right now? It was actually growing quite a bit since that part of part of it is where you alluded to at the markets before. But yeah, we've we've had some days that are upwards of seven billion dollars now. Wow. And who are the customers who are using your platform right now? Yeah. The direct the direct users are all you know the large banks and broker dealers that you know people know the know the names of indirect customers. It's a mix of you know asset managers, hedge funds, etc. But institutional traders are who we're currently through the institution list. So yeah. So this is a tough market. I mean usually I'm you know people see how hard it is just to get the first customer when they're they're like a regular FinTech or some consumer facing side. But this is seems like a whole other level in terms of difficulty convincing an institution to trust your system to trust your your solution even with your your impressive backgrounds. How did you start getting those first clients? Yeah, it's it's not easy and it's actually even harder than than just getting you know an institution like a large bank to trust the system. It's you need several of them to do that before you even start trading since it is truly a marketplace right. So it goes by different names the chicken and egg problem or the the cold boot problem for any new marketplace. And yeah, it's it's not easy. And for us I think you know before we were live the the message we always heard from potential customers was this sounds like the most innovative thing I've heard in capital markets but it's going to be really hard to get it off the ground. And and that was not come back when you had like five other people or something. Yeah, yeah. So it's the the mechanism when we describe it as you can tell it sounds pretty complicated right. So that's um it's hard to get over the hurdle for for a lot of people but what we've done is make it so the actual users it's not complicated at all for them all the complexity is behind the scenes. Now the complexity is interesting and it's why we after launch of generally outperformed other dark pools and other trading venues. But the we've hidden all that complexity to away from the subscriber the users themselves. So that was a big part of how we were able to get um over the the initial hurdle of those first. Yeah, how did you get that convinced those first first one or two people. Yeah, it's um it's really a mix of kind of the peel of the first principles of as former institutional traders we can speak the language. There's total agreement on what some of the problems are as far as you have high level portfolio objectives. You don't necessarily have a direct way to express those to the market where things get more interesting is how do you strap that on to existing capital market structure in a way that's non disruptive to existing workflows. And the thing that we really iterated on was focus on this big picture value proposition and the path there really the 10 year forward to what capital markets could look like. We'll also reiterate this to you's point that you can plug one chronos in treat us exactly like any other trading venue make no changes to your systems whatsoever and still get a really good result out of it just trading the simple limit order type workflows that people are used to today. So going at it from both making the reward high and the cost low is really what was key to getting us off the ground here. There were a link to take a chance on you. We had we are very grateful for for early supporters in getting to launch for sure. And you know very complex system you need a lot of hands in there. Were you able to self fund or did you raise money at the beginning? We raised money. Yeah. Yeah. You know unfortunately it's there's a lot more similarities between launching an exchange or an ATS and like going to space than there are to most startups. So so it's to your point before getting people to to trust the system. You know this is critical infrastructure for very high stakes transactions. So you have to be very carefully. It's it's not a cheaper, easy process to get launched. And then there's the whole regulatory aspect of it which takes time and capital as well. So we did why our first funding was through Y Combinator the startup accelerator in the Bay Area. We did that in 2016 long time ago and um we're feeling like you've aged prematurely. Definitely aged faster than I think we would have otherwise. It's all been worth it though. Yeah. And as our partner from my teammates pointing out to it another YC company from our year did in fact put a satellite in space before we launched. Oh hey yeah yeah that's great to the conversation when you guys meet up. Oh yeah you're from our year. What have you done? Oh I don't know we launched some rockets into space and said like what are you doing Kelly? Um trading at this nice. Writing writing legal documents and writing. Yeah. We're literally rocket scientists and yeah. Yeah yeah. So did you have the problem where a lot of people a lot of potential investors didn't really understand or end up getting what you were doing. We've we've been fortunate that the investors that are in the space and know it well. Really understanding people that know know this world. People that know the world understand what the problem is that we're going after and that's what got them the conviction. It's uh it's like anything you know you've got to segment your audience basically and um what we're building is not historically anyway then for folks outside of capital markets the interesting crossover we're seeing now just given that there's data and there's also a lot of talk about what the economy is going to look like for AI and AI agents it's kind of like oh cool that's what one premise has been building for the last nine years at this point. So there is just a lot more general understanding of what we're doing now in 2025 when there was in 2016. I've always wondered what is the regulatory world around private markets. I thought it was uh there's not too much involved but uh you mentioned getting over that hurdle as well. Yeah there's there's a lot involved for US equities at least so we're drawing regulated by Fender and the SEC so Fender on the broker to the other side and then SEC is the one that oversees what's called uh ATS the framework that alternative trading systems. I think the SEC is the ones they can send me to jail. Yeah yeah they have teeth that's for sure. They're yeah it's uh it's uh the regulation that we operate under is um basically gives us the the ability to trade like we're a stock exchange uh so we can match buyers and sellers of any publicly listed stocks but um but we don't have the ability to list new stocks or or sort of govern what uh listed companies do. Nor do we want to for that matter or our business model and like what we're focused on is very much around helping helping markets function well via this new matching mechanism. It's not that we have any interest in becoming a listing exchange. You mentioned you've been at this for nine years or is the system where you want it to be now or are you still building up more capabilities? More capacity? Yeah well we'll never stop building more uh more capabilities and then more capacity. The interesting thing on the capacity front is that um there there's a lot of things that are good about starting out with US equities as our our first asset class. For example the regulatory framework is very well understood. There are lots of venues people are used to plugging in new venues but um the the amount of data and the system requirements are extremely high and higher than almost any other marketplace in the world so we kind of have the capacity that will all the capacity will ever need or within called an order of magnitude that we have now that will ever need so uh but that being said from a feature standpoint and from a interop standpoint every marketplace is different so foreign currency and and European equities which are very related to US equities and we're currently expanding into all have their own nuances and their own trading behaviors and so we'll have to um we'll have to extend the the interface that people use to trade to meet these markets where they are and then it's even more the case when we start getting into real-world economy marketplaces outside of capital markets. Is there anything that you would consider like a limiting factor some obstacle that you would uh love to to get out of the way that would that would help your company grow and video chips. Getting on the the best podcasts. Brett so thank you for your help with that. One box check here um we're we're never going to be done from a computational standpoint so Steve kind of disreferred to the system scaling part of US equities has a enormous amount of traffic from border entry. The other side of this is these auctions that we run are fundamentally unsolvable optimization problems at scale. Our IP and our direction of travel is around being able to find very good solutions to them quick enough to be relevant to the speed and scale of capital markets and that's something that we will always be investing resources at getting better and that's frontier research type stuff so it's a great fun problem but if you ask me hey could you snap your fingers and one thing is um done it's that it's computer scientists would say P equals NP and this problem goes away and we're able to operate these types of markets without those computational constraints. So you mentioned that the role that I place on your platform and I'm curious some other thing I have you looked at other mathematicals are doing this like combinatorial optimization or even H quantum computing is well that seems like a fairly easy use case for this. Yeah so the quantum stuff is interesting and somewhat related to the P equals NP joke that is made for the computer scientists but there's actually some initial hardness results that indicate quantum is not going to be useful for this problem which I wish wasn't the case so there's not only the problem of when quantum hits a viable technology but if it would be applicable to this particular optimization problem. Got it so it's basically it's it's going to be just all AI all sort of traditional compute for now. For the foreseeable future anyway the there are some places where quantum could conceivably help with very narrow definition of quantum could conceivably help with the optimization but the the problem that is being optimized is just fundamentally a very hard one from a computational standpoint. Can you give me an example of a trade that your platform now can execute very very efficiently and effectively that maybe would be much more difficult or really impossible if you use a non-AI way of doing this. So basically the the question is how has the technical execution of all these traits evolved in the past called a 10 years or so. Yeah I would just so I'd predicate this with we use machine learning and classical AI to solve the optimization problem. It's not that there's like kind of AI adjudicating these trades or matching anything but as far as like something that's possible now under this platform that wasn't it sounds embarrassingly simple which is a peristrate just the ability to buy X and sell Y as an atomic unit that's some ratio or better and that's just not the way that capital markets are wired right now but on one chronos you can put it in order that says I want to buy this if and only if I can sell that at some price or better and it's it's a it's a way of creating with way for the market because people are able to in cases price those individual wigs excuse me price the package more aggressive way than they would the individual wigs because they solved for the exposure problem the risk that they would face otherwise if not completing the Pesto recipe. What has been has there been any really tough point along the way where you thought you might quit or just abandon this or you really had to pivot the business away from what you thought it would be. Yeah I guess feeding for myself first is kind of no and no it's it's funny how identical what we were pitching and doing was nine years ago to what we're doing right now it's been completely consistent on that that we are commercializing commutorial auction starting with capital markets and more generally building something for the real economy in the future that message has changed zero the way that we present it kind of in the capital markets overlay of hey how do you practically use this day one that's something we refined a lot and as far as quitting it's never something that I thought about it's it didn't go into it with the framework of let's see how much progress we make it is this I'm going to do this until it's done and we're nowhere close to done but we've got a lot more proof points than we did nine years ago. Yeah I guess well you're coming from this world but in both of you and I guess you that just shows how much you really understand understood and understand the the challenge and the market opportunity whereas what we talk to other companies that are kind of maybe on the consumer side things can be different than what you expected or consumer behavior can be different but this is a little bit different in that way yeah yeah exactly and on the market side of it you know from from day one we've been having conversations with every market participant every type of participant we can and even in 2016 the the consensus has always been that you know what the the market structure that we've been proposing makes total sense and is better than the market structure that exists today that's it's it has not been controversial at all over the years it's just a matter of can we get people over the complexity and can we get people to adopt these more interesting futures or features and and so now that we're seeing that happen it's it's very easy to in retrospect say yes but I think the just the fact that there was so much agreement amongst our prospective customer base over the years that this was genuinely a good idea gave us a lot of wind in our sales and made it pretty easy to keep going and the only thing I would add to that since you brought up the consumer behavior point two is like see when I definitely wind into this because we felt like it was something that had to be done it was just like very obvious to us that this company need to exist and it really was not that we wanted to like start a startup or keep pivoting until we figured out something that worked commercially I don't know exactly what to be doing other than this probably something in academia but it yeah it's it's a problem that we felt had to be solved as opposed to we're going to keep pivoting until something works well you you would have found some other problem and made a company around that that's true and we've found a couple hundred of those problems along the way so it turns out we building a company you end up on a lot of little side quests so yeah yeah sure you you can't start anything without running into things that you shouldn't do yeah I've always liked the story of a Toyota for that matter as well as far as you know what's one of the biggest pain points that some people have it's getting the phone system working and you end up building exactly that so yeah right right so you mentioned expanding into other asset classes and those are all pretty like you said pretty different worlds do you have to hire up for people that that have experience in that or can does the logic still make sense either way yeah absolutely so the technology is no different across the the different asset classes it's the same platform same code that runs everything which is which is key for our future but it's very important the higher experts in each of these different asset classes because they each have their own nuances so we've already we've already done that we have a great team in place in the UK and in the Netherlands we have offices there and for FX we got a great team in place in the US here so the the hiring for those is done and it's just the matter of getting to launch at this point yeah congratulations that you guys are not afraid of taking on some some difficult tasks to just taking on an all new asset class and expanding the trading portfolio so is there any company that you would love to partner with or that you kind of admire you'd love to work with so certainly a lot of companies we admire for different reasons the way that we think are of ourselves in this market overall is just we want to be the the marketplace of marketplaces we want to be that API that allows people to spin up a marketplace for anything from widgets to stocks and the extent that things are regulated regulated assets we want to get people a straightforward way of operating an efficient business that doesn't just keep reinventing the wheel over and over again well look at finances as stands there's all of this infrastructure we built up for people doing fundamentally the same things and no economies of scale that comes from socializing the cost of that infrastructure so I would say we'd be excited to partner with anyone that thinks some way about hey how do you how do you kind of how do you build that future and how do you treat one kernel more as a marketplace of marketplaces is there any law that you think that if you could with your magic wand would you go away and would make your business easier to run but also safer don't say insider trading we'll have to drop the whole thing it was not remotely going there and it's it's actually it's really hard to point to one anything in isolation because a lot of laws as you'll discover around capital markets there's a programmer would say there's a series of unit tests and you might see something where you're like why is that in there and then of course there's some reason why it's in there because at some point someone or something accidentally or deliberately took advantage of it not being in there so when I think about what overhalls of market structure could look like a lot of what I do think about is this kind of what does this more combinatorial and future marketplace for AI look like and I think it is going to be important that we get those regulations right and I think it's going to be important that we put laws into place and revisit laws that make sense in a human world and maybe less so in an agentic sense and vice versa I think it's going to be really important to get all this stuff right that'd be my main common on see if might have a different take I was going to go a similar direction where it's it's almost like what rules would you add instead of delete so I mean crypto currency is a perfect example of it's been around for 15-ish years and is still more or less unregulated with the exception of some state intervention and that lack of regulation has created a lot of business risk for large institutions to participate and that's why we've seen these crazy cycles in the in the industry so it just that actually the lack of regulation in some ways creates barriers in in these little pockets of the marketplace and Dekelli's point I think part of that delay is like it's very far to get these regulations right so I think yeah I think there's there's a set of rules around things like crypto and things like AI agents and and especially things like what data and what kind of you know what kind of data is it okay to put in a llm to just send out to chat gpt these are things where companies are coming up with their own policies and and at the end of the day there's just a lot of trust currently that exists or has to exist between you as the user and the llm platform so that those are areas that I think about a lot when it comes to regulation well one thing I can predict for sure is that you guys won't be bored in the future or never I'm never running out of things to do around here yeah yeah you got a few capital markets and what uh what the future stays going to look like well we wish you all the best taking on some interesting market problems and providing some very interesting solutions good luck with that and thank you so much for joining us on the podcast thanks for having us on thank you both share the great questions yeah that's Kelly little page and Stephen Johnson co-founders of one cronos please hit subscribe to keep up with the latest in fintech news and thank you for listening