The Exchanges

Q And, and the firm you came from before Third Point, World Quant is, is much more on the sort of pure quant side, right? Is that correct?

A Sure. So World Quant was fully systematic, which was more of a thesis of, if we consume more data than anybody else in the world, we can find more signals, create more alpha. I think on, you know, where we sit now, or where I sit now with Third Point, our team is thinking, there's all this data out there, and there's so much noise, we have to decide where do we spend our time, and what can help us better understand the names we have, and Rather than mining data for new ideas, we're almost looking for data to help us understand the ideas we may already have, or to, you know, create very sophisticated screens, you know, looking across all these companies to kind of dwindle down a smaller list of names that we can work on with a fundamental manager.

AI assessment note: “Sure. So World Quant was fully systematic, which was more of a thesis”

Q obstacle is that, like, you have literally no idea how the hedge fund uses the data, so, which makes it very hard to then go to the next hedge fund and try to do the same sell, because obviously the hedge fund industry is notoriously secretive, um, is that, um, Is that what you see as well? I mean, you, you basically just raw data dumps and say thank you.

A I mean, I think that's changing. I think in the quantitative hedge fund world, you'll have the top 10 players that you're not going to get any feedback from, but at the same time, they're going to be a client and not really complain unless the data just doesn't show up on time. So it's almost an easy, you've sold it, make sure it's being delivered and you can move on. I think on the fundamental side, especially Where you have dedicated data teams that are looking at things or analyzing it. There's a lot of back and forth discussion. You're going to get a lot of feedback, and I would say the better data teams at hedge funds are really helping these companies build products. Because a lot of times getting the raw data is fine, but if we're going to have to do all this work and you have a team and you really want to make the sale, we'd be happy to help you, you know, learn how to build products that'll make our life easier.

AI assessment note: “I mean, I think that's changing. I think in the quantitative hedge fund world”

Q Interesting. And the, the, um, um, I guess, how do you, how, how do you know, um, whether machine learning works? How, how long does it take, uh, for any of those, uh, sort of quant techniques to prove their efficacy?

A I mean, I think it's a lot like how we test a lot of the data sets we have, but you really want to see it run out of sample, arguably for a year. Obviously you can slice the data so you can do different training periods, but I think, you know, you truly want to understand if it's not point in time data where you don't have that time stamp to understand when you could have actually executed that trade, or to make sure that they didn't overwrite the historical data, you really want to see, you know, a few quarters of how it actually performs before you tell, you know, management that you really want to put money behind just an algorithm.

AI assessment note: “you really want to see it run out of sample, arguably for a year.”

Q And do you think as a, you know, at the financial services industry level, this percolates down from hedge funds into mutual funds and other types of investing, this general, like, more quantitative, more data-driven, machine learning-driven approach?

A Yeah, I mean, I think that data in general is just becoming more prevalent and available, um, and I think that long only is just how money is managed, You have more information. So, you know, in the early years, people used expert networks to really understand a healthcare company or a technology company. Now you can buy data to kind of give you that insight, and data's becoming more and more available from so many different vendors that it's not just one monopoly that has it. Um, so I think as the hedge funds are maybe the first movers, it kind of, ah, kind of moves all the way down, and the exciting thing is to see who are the new, you know, data vendors that are gonna come out.

AI assessment note: “as the hedge funds are maybe the first movers, it kind of, ah, kind of moves all the way down”

Q know, initially made me smile, and still makes me smile to, to, to, to, to, I mean, a supportive smile. Um, but, um, uh, at the same time, I actually have heard cases where that is true. Um, so what, what's, uh, is that, is that just like a one in, um, you know, thousand or million type occurrences? Uh, how should startups think about selling data to hedge funds?

A Yeah, I mean, I think the multi-million dollar data sales was maybe something that you saw three, four, five years ago. Um, I think now that, you know, people are understanding what data is actually worth, uh, they're maybe not trying to get exclusive access to just one data set because they know it's just a bigger part of a mosaic. Um, you know, things are changing. I think now you have data hunters or people that are really just focused on buying data, so it's becoming more transparent on what vendors are out there. You know, we're not needing consultants to help us find the next credit card transaction data set. Um, I think selling the hedge funds, you know, most hedge funds in this day and age are very open to seeing what's out there, understanding what, you know, the offer is, but I think, you know, you have to have something unique with a lot of history. I think there's a lot of support because hedge funds, you know, are making investment decisions off of this, so if it's not coming in every day or if there's, you know, the data changes and there's all these issues, The more that we have to work to make it, you know, kind of help us, the less we're going to really pay.

AI assessment note: “I think selling the hedge funds, you know, most hedge funds in this day”

Q Is that like basically a big arms race, right? Presumably you use credit card data one day, but like if everybody uses credit card data, then your edge just disappears. Is that the right way to think about it?

A Yeah, I think it brings more volatility into, you know, events, earnings, especially for firms that are trading around these events, where everybody's looking at the same credit card data. And a lot of it is there's third parties that are creating analytics that are maybe More affordable, but they're all using the same underlying data. Um, so it almost becomes that you have to have that information just to understand what everybody else is doing. Um, if the data comes out two weeks before earnings, the stocks are moving a lot, but everybody knows it's because of this data. You may not have a opinion about that company, but you understand why things are actually changing. So I think it brings different opportunities. I think also, you know, using this data is very different from a quant fund who's placing lots of trades, both long and short at the same time. Whereas an activist or long-term investor might look at this as just, ah, another input into their longer investment approach.

AI assessment note: “it almost becomes that you have to have that information just to understand”

Q Are there some examples of data that was just, that didn't work? Like, you know, you had big hopes, um, or you or like, you know, friends of yours in the industry had big hopes for, and that turned out to be like super not interesting.

A I mean, I think we all hear about the satellite data, and it's interesting maybe for specific use cases, but I don't think that there's a lot of firms really counting cars in parking lots, and that's what's really making their, ah, decision making. And I think a lot of these data sets, we really need to think about how do we combine them. And that's where we spend a lot of time is, you know, you have all these different types of data. How do you connect them to each other? If we have information about, you know, where people are, what they're spending, what the demographic is, how does that map back to the type of company we're looking at? That's probably much more exciting than, you know, there's 50 cars in the parking lot today, but tomorrow there's a cloud and they couldn't count.

AI assessment note: “I think we all hear about the satellite data, and it's interesting maybe for specific use cases”

Q And, um, and so what, what, what does one do with this data? Like, how do you know if it's gonna be helpful or predictive, or do you just, like, acquire a bunch and run experiments? How does that work?

A So for us, I think our approach is more of what type of information can make us better understand a company. So if we're looking at a specific company, I think we try and think, if we were a data science team within that firm, what information would we have internally to understand Better how our sales are looking for the quarter, right? Because there's a lot of information that comes out quarterly for these companies. We're trying to better understand for different line items, maybe different regions, different products they sell, how are they performing? So it could be credit card data, and does the panel really represent that company, and is it clean, and how has it looked over time, and so, you know, from a basic correlation to what is it predicting going forward, and really understanding maybe the industry they're in, and then maybe we'll tie that with macroeconomic data. What's the imports and exports look like?

AI assessment note: “from a basic correlation to what is it predicting going forward”

Q And last question on this concept of selling to hedge funds. Is there a Uh, compliance aspect to this, the legal aspect to this around, uh, insider trading, like data provenance?

A Sure, so I think definitely over the last few years it's become more and more of a topic and more and more of like what our day job is, is there's due diligence questionnaires, there's log legal processes, um, if it's a software or some sort of, um, platform that we're going to be using that we're putting our information in, it's just going to be more and more, um, compliance and security checks. I think that a lot of firms are getting good at having a set, you know, amount of paperwork. You fill this out. We have our calls and we can move forward, but it's not a, you know, put your credit card in and buy the data set.

AI assessment note: “there's due diligence questionnaires, there's log legal processes”

Q Great. So maybe switching side, um, you know, from, away from data and more into the algorithms themselves. How much, how much, uh, you know, machine learning and AI is actually used, uh, with, with all the, all of this data?

A You know, I think on the quantitative side, it's probably more prevalent just because you can, you know, systematically trade and not really understand what the algorithms are, you know, doing. I think on the fundamental side, we have the explainability factor. So if we come up with a great model, and our team shows that it's made X amount of returns over the years, they want to understand, like, what is it actually doing? And can you actually explain what factors it's actually picking? So I think it becomes a little bit more difficult, and you have to think of ways of, How do you maybe deliver information or a list of names that they should go look at based off of the machines, and then there's a layer of due diligence.

AI assessment note: “I think on the quantitative side, it's probably more prevalent”

Q suddenly everybody just became an AI and machine learning startup, right? And from the little I know about the hedge fund world, it seems like everybody now is, like, at least raising money on the basis of AI and machine learning. So are you a believer that ultimately that creates, like, a different type of success and hedge fund, or is that more of a marketing and fundraising type positioning?

A I mean, I think it's, in the hedge fund world, you need people to be pushing the innovation because, you know, it's been an industry that historically, outside of the quantitative space, hasn't really embraced new technologies and, you know, leveraging alternative data. Um, I think that it's bringing smarter people into the industry, which is a good thing, and, um, how we utilize it and what it does for our investors. It doesn't have to be just using AI and machine learning and, you know, cutting edge NLP For systematic investing, it can be to cut through all of the noise and then deliver something that our fundamental investors could then go talk to management about and like be more intelligent when they're having conversations.

AI assessment note: “you need people to be pushing the innovation”

Q data driven approach. So it's interesting that it applies to the hedge fund industry as well. And, um, so how does that work? So, so you, um, maybe help us understand what, what does a machine do and what do you do? Something come out of the machine, and then you, uh, somebody on your team brings it to the PM for an investment decision, or is that more automated?

A Yeah, I mean, I think this is the big question of how do you integrate data science within a fundamental investment firm? Um, the culture and how do you actually use these tools that give us more intelligence now we're making investment decisions. So, a lot of it is how do we automate what a traditional investment analyst is currently doing, make their lives more efficient, But then also, how do we incorporate all these unique data sets to give them a better understanding of the opportunity? Maybe argue for or against the thesis they already have. Um, if you're taking a deep dive into a consumer name, you probably want to understand the transaction data or the location of same, of people walking into those stores. But how do you take all of that as just another addition to what they're looking at? I think the unique thing for us is we have a very small, concentrated portfolio, so we get to really dive into specific companies. Um, really, you know, going back and forth with the analysts on is this a good opportunity, is it now the good opportunity, or can we track this and use, you know, technology to keep track of things and have those ideas kind of populate back up to us.

AI assessment note: “this is the big question of how do you integrate data science”