The Evan Reich Episode

Welcome to the Alternative Data Podcast. Welcome to the Alternative Data Podcast, powered by CFM. I'm Mark Fleming Williams. In this episode, I speak to Evan Reich, a veteran of the alternative data space who has worked at Millennium, SAC Capital, Quandle, and most recently Verishion in various data roles. In our conversation, Evan and I discuss a wide range of topics from how to hire a good data sourcer, to which data types are overrated, to whether exclusivity is a thing in the data market.

In the podcast we mentioned Evan's excellent conversation on Matei Zatrianu's System 2 podcast, which you should also check out if you like this one.

So in this episode, I am thrilled to be joined by Evan Reich. Welcome, Evan. Thanks, Mark. Um, Evan, you've just uh corrected me on the spelling of your name, so that's a good start. I'm doing that correctly. Um you are a very familiar face in this world. You've been around But um It's a very polite way of saying I'm old, but thank you. And you are at all the events and um So you are. But um let I think probably

Um, with others, if you've listened to other podcasts, what I've done is kind of go step by step through their background. But I think with your background having been so um like varied and in different places. And actually, you know, you've worked for some big names. You work for SAC Capital. You worked for Millennium, you worked for Blue Mountain, you've most recently been working for Verition. In between, you worked for Quandle and it got bought by Nasdaq. You have been in the data.

You were uh you were S and P originally and originally you're at Credit Suisse. So my word, you have done uh banks, you've done providers, you've done buy sides, you have Explored this world from every direction.

That is an even more polite that is an even more polite way to say that I'm old, so thank you for doing it twice. But you are entirely correct. It's it's one thing being old and it's another thing actually making use of your time in the uh in the aging. So uh so no, a lot of a lot of people have not

uh been so interesting in in all the in all the things that they've touched. But um so what I want to do actually instead is I think if we if I if we talk in broad terms, um, then it means that you can tell me what you think about things without saying which employer you learn. Um and which hopefully allows you to be to be free on we get Evan Reich's um uh pure opinions. Um without upsetting anyone, sorry.

It's all good, no worries. Um without upsetting anyone. Um and and then we can finish up by talking about what your new your new uh endeavour and project. Sweet. Okay, cool. I've written down a few kind of random questions um and maybe some more will will um occur to me. But I mean, okay, let's just do a little bit of of who you are. Yeah. Can we summarise your background? So most recently you've been head of data strategy and source.

Um, you've have you and you've you've been a you've been market data before, so you've touched alternative data and market data. What other what other labels can we put on your

I think when I think about it, it's really, you know, initially it was I was a hedge fund investor at one time. I worked in private equity at one time. Um, but then sort of having moved into the research space. I think in some order it was doing doing research in data before we had named all these things. We didn't have categories. Then it was doing the thing we now call data engineering, but before we had put a name on it, we didn't know it had a name. We just thought we were loading stuff.

Um and since then really, you know, being sort of, you know, head of date of strategy and sourcing for all manner of data. It was never really market data, it was always just all the things, particularly since I got to all of Um, but that's that's probably the right way to think about it. And if you had any

special skills which really define you? What would you say if you're like the thing which is kind of your your um hellboy left arm, uh, you know, what is the what is the thing that you are that that you are perhaps you know, best in the world at or whatever, best best uh part of you. Well one excellent reference. Two, um I mean other than my overwhelming snarkiness in general of that whole thing, which of course everyone is well aware of, um, I think, you know, initially

People didn't get into data like this. You sort of ended up there through a randabout, you know, odd, you know, island of misfit toys kind of a thing. Um, so initially it was really just having an unusual skill set and no particular specific box to put them in. But I think as time goes on, you wear a lot of hats. And I think you would know this, you know, in in

I in your other job. Um, what we do more than anything is sit in between a lot of people. So there's a lot of, you know, speaking in different languages that all sound like but really the ability to code switch and really understand people at their level and make those connections because quants, fundamental people, research people, salespeople, legal and compliance people.

They all speak a language that sounds the same on paper but is in fact not. And I think that's really the at the end of the day that's the challenge is understanding all of it, putting it together, synthesizing it and then spitting it back out to a different audience. So Brilliant. Okay, so um You have been b in data sourcing roles, among other things, for a long time. How has the data sourcing role evolved in your during your

Yeah. I mean, obviously the biggest thing is really just the degree to to to the number of folks you need to know. And I mean, y y y you know, sort of alluded to being old. You know, the advantages you do one of these kind of things where it involves a lot of connections. And you do it for a long enough period of time. Um, and the people you meet initially who are, you know, the junior folks that are your peers.

you wait, you know, ten, fifteen years and they end up are the people that are running things. And so and they spread out consistently. And so I think as long as you invest in and maintain those relationships It's made it easier. It's much harder to walk into this market now than it was, you know, ten years ago when there were I don't know five vendors, ten

Now there's hundreds upon hundreds and even keeping them straight is challenging, but the tools have gotten better. So that really helps obviously, you know, LinkedIn and and various AI ways to stay in touch with people and those sorts. things. But I think the biggest changes really have been just the explosion of of number of places where things reside.

And the and the quality of the information you can get. I think you know, years ago if you went to folks and and initially had said, Oh, we have, you know, data about transactions and people were nervous And now people are like transaction data, that's the most basic of all things. So it's really been that evolution and just the the people's comfort with data um and how much of it we're all willing to put out there in our daily lives.

speaking and then the explosion of uh vendors as a result of that.

But in the early days, I mean uh in the early days it was about uncovering data which, you know, it was like, Oh my god, I didn't know exist. And so was there and I'm just, you know, I'm wondering, was there more of a kind of imagination involved in the early days of what might work? And now it's more, you know, it's It's a less imaginative job and perhaps a more kind of

managing efficiencies job. Is there is there that kind of arc going on? I think the I I think the components in the early days were more of you had to be quite creative to come up with things, but you also had fewer varietals to work with, if you will. So it was like, you know, you had like three apples and it was, you know, well, which of these three apples should you pick? Now you go to the farmer's market. I live next to Union Square.

There's like 17 apples to pick from. And it's like, well, which of all these flavors do you want to deal with? So now you have the ability to combine far more things in far more ways. Um But you know, there's a lot less you can do, I think. I mean, like we're we're we're still waiting to see. This is all still evolving.

But, you know, initially, you know, when I started doing this, not everyone had cell phones all the time. Like we weren't collecting data about ourselves as much about everything. So there were just less tools to work with. So you did have to be very creative in how you applied it. I think the other

really big change has been in access and that initially you had to, you know, connect to files, you had to bring them down, you had to parse them, people had to put in all that work. So the lift to access a piece of information too theoretically connected to other

was far greater and now I mean I think there's a lot of people in this space now that have never seen data not in Snowflake. They don't actually know what form it actually exists in. And I think pretty soon the the analogy I've been using with a lot of people is you know, in AI land is a lot like the witches in Macbeth.

It's a big pot where you throw in ingredients that people may not understand. I have new tale of salamander, data from here, or this thing from here, and then you say a spell over. Um and right now most people are writing their own spell books, but at some point in the future you're gonna get you're gonna see people I think who have a spell book they got. And a pot of ingredients they don't understand. And or can't

And they're gonna say that spell and hopefully money flies out of the pot. But if one day it doesn't, um all they're gonna really know how to do is say the spell louder and faster, kinda like, you know, an army of darkness when he's like clattu, barata since we're getting geeky with the Hellboy stuff.

But it really is that kind of thing. And so I think it's really important for the people that are creating creating those pots and putting the ingredients in for the end users to use in in in all stripes to really be careful about what's in there because uh if you combine the wrong ingredients

It can explode or an army of Deadites can come and get you and that would If in the early days there were three apples and now there are seventeen, um presumably in the early days then you could test all three and discover by rigorous research which one was the best.

When it's seventeen, it's harder to test them all. And so is there more reliance on the data sourcing team than there was back then? Because actually you're having to differentiate for the PMs or or the quant researchers or whoever, um, you have to do the first sieving process and so your job is more and more, I don't know, more involved. I think nowadays you see that You definitely have to get talk more and have that conversation. I think it's interesting and having been on both sides of it.

to see you can really see the difference between people who understand the why of it all. And like you do deal with some people in some places where they're like, is your data point in time? And is your data this? And you can kind of get the sense from speaking with them that they don't re

They're asking these questions'cause they know they're the questions you're supposed to ask, but they don't really understand why. And so it's it's sort of interesting to see that dynamic. But again, back in the day there was less data, but it took far, far longer to load it. I mean, literally we would procure data, you know, back in the day in data engineering, we would get like an array of hard drives.

Would show up. We actually built a wall out of hard drives for fun because we had received so many hard drives full of data. And it's like, man, that took time and energy and work back in the day. Uh uh w what is now known as I guess El L Seg tick history, but used to be called Reuters Tick History was all housed in a server in Australia for bizarre historical

And so if you tried to pull data over the wire from that, it took a very, very long time to come in. So they would literally send you like terabyte drives of this stuff. And that just took a ton of time, even to just get it off of there and and do anything. So in that sense, it's it's changed in both sides, but I think definitely you see a lot of people in the space now who need to sort through a lot. And so you need

a lot of folks to do that at least as of a few days ago until, you know, AI takes over some piece of that. And if you don't understand why you're asking the questions of, you know, is the data point in time, how frequent are the up? If if you're just parroting words and you don't know the meaning of them and and the depth behind that, um well that's AI could ask those questions. So it's probably important to understand the why and the how and how that all comes together and really

closer to the rationale for a lot of people. So I think that What do you when you're hiring for a for a data sourcing team, what do you look for in a potential hire? I mean, you always have to think about who the constituencies are and what people are gonna be doing. So the simple reality is if you're working at a hedge fund, you know, the PMs and the investors, those are the people you are there for.

First and foremost, you have to work well with them. So you want somebody who's really able to have those conversations and do a good job and can synthesize. PM time and analyst time is an incredibly valuable commodity. It's the most valuable thing a hedge fund has really for most funds is their people.

in time. And so being respectful of that and using it well, I think is the most important thing. And then, you know, following that, you need people who can speak well to the vendors because you're representing your firm to them. Handle that well.

Judgment. And then you want somebody who ideally, I mean, you've been in these rooms, can walk into a room and be comfortable going over and talking to everybody because that's how you find the interesting stuff. There's a lot of people in our space and in the larger.

who have interesting things to say, interesting things to offer. They may not be good or comfortable out there and if you're the person that can extend that olive branch to them and talk to them and make them comfortable, they may well um be more comfortable speaking with you and and it's it's easier to do business with your friends. I think some combination of those things is probably the right side of the Oh, and people who just wanna work all the time.

You know, there's always more to do and there's always people working. I think the nature of the hedge fund business is that because it's public markets, um 247, 365, maybe not New Year's, but every other day, there's a market. Be doing something. And so, as this people supporting those people, um, you want to really understand the importance of being available and and and serving.

You just said a lot of people person things, um nothing technical, nothing about imagination, creativity, nothing about knowledge. um pre pre existing knowledge, uh are any of those things, would you add them or really people person and then you can teach them the rest? I think the I I mean ideally you'd love somebody who has knowledge. You'd love somebody who brings technical capabilities.

But again, I mean, if somebody was really technical and couldn't talk to people, you'd say, well, that's that's a back-end data engineer. That's somebody where you don't want them dealing with that. They're not somebody you'd be looking for on necessarily. And again, over time, I think as we see the explosion of of AI, you know, I have some friends that are working on a variety of uh AI enabled, you know, data loading tools, things like that.

that are really fabulous at what they're doing. Um and the more of that they can do, then maybe you may need fewer people in those roles. But that human interaction part, that is not something that is easy AIable, replaceable, what have you. I think when it comes to the the creativity, that is really important. Um, but I think that's something that comes over time where if you're bringing somebody into the role initially

They they don't know enough to be creative in the right directions in most cases. Um, but I think I mean, y you know, the analogy I've used again over many years is like, you know, you get a bunch of pieces of what to think about And you, metaphorically speaking, um, take a bunch of peyotes, sit on a mountaintop, go on a vision quest, think about how to bring these pieces together, and then you say, hey, what can we do here? And in some cases it works, in some cases it doesn't.

doesn't uh work on the vendor side or the utility side or what have you. But those are great when they happen, but that's not the that's not the day to day of the gig. The day to day of the gig is connecting A to B and making sure that everything is just moving. Do you see hiring someone from a competitor as being a opportunity to a

get a whole load of, you know, data set ideas that the competitor has and also the person who can hik hit the ground running, or do you see it primarily as a bit of a flight risk'cause if they've moved once and they can move again? Um I I personally don't think of people in those terms. I mean, I've always tried to take everyone in this business as like individuals and think about them as people. Um, so I've been very fortunate that for the people that have worked for me,

You know, I've I've I've known some of them before, you know, they do they they they worked with me. I think some of them, you know, we kind of had had moved around each other. directly, but knew a lot of the same people in common. So I think look, the biggest compliment somebody can give you is to go, you know, work for you, learn a huge range of things.

The time is right, and they've grown and hopefully evolved and developed their skill set to go out and do that somewhere else. There isn't always a place to move up. So somebody leaving Because because they've become so good that they need to spread their wings and fly is uh Honestly the highest praise you can have is as as a manager of people, as a friend of people.

Um, so I don't think that's a problem. I think if you're bringing people in specifically to say like, uh, let's find out the thing that the competitors know, um, you know, again, that that that's good for an hour, but after that they still work there, they still have all the other set of skills are far more interesting and important. Certainly there are people that run funds that have very different takes on that because they are in an extremely competitive

of universe. Uh as the data sourcing team, I don't think that's primarily where one's head has to live because it is by its nature a collaborative enterprise. And so I think there you want to be as collaborative as you can. you can given the industry you're in. So do you see data sourcing as a big strategic risk in terms of an awful lot of IP is held in those heads, or do you see or are they in an ideal world, are they fairly quarantined?

Don't hold that much. Or do you see it actually that um to we're in such a commoditized world these days that to be honest, knowing what I think what you find is that there's real differentiation between different situations. So if you're working in a place, you know you know, a few fundamental people or what have you, the likelihood that somebody on that side has, you know, super advanced knowledge of the strategy. Again, the strategy is taking place largely

Heads. So there may not be, you know, hugely differentiated things. It's about those people. you're doing the more quantitative the fund becomes um and the more there's like specific timing elements and things of that nature that there can be real ip in there and that's why you see such sensitivities and long lockups and things like that

in that space and you want to be careful about that. I think also, you know, as a data sourcing team, there are times when there is, you know, real specialty products that are made. You know, in my time at at Quandle You know, the the single most successful product they have. Um, and I'm aware that there is a limited audience to which we sold that at that time, that that continues to be sold. Um, and there is some proprietary nature to that.

Know how people treat that and what they do with it. I think that's a personal integrity thing, more than anything, is being very selective in that process. And and being respectful to your prior employers and your future employers and your peers and your colleagues and and not, you know, just being a good person at the end of the day for lack of a better way to put it. But not everybody sees it that way.

Um so from a data sourcing perspective, a a big fund will get data from all sorts of places. Um it will get it from the big uh what from banks, from prime brokers. From big ugly data providers that of which everybody knows. Some of them are literally gorgeous, beautiful They are they are beautiful. They are beautiful. I just try to hide my truth. Um but no they are. Uh if you're listening, you are. Um but uh and then also from um uh all sorts of uh kind of alternative data providers from from

are smaller. Um and the alternative data providers can be found in catalogue companies often. A lot of them can be found in catalogue companies. How would you consi how would you break down what you see the natural um percentages of those. What percentage from PBs? What percentage from I'm PB. Exchanges, shall we say? Um what what percentage from big well they all keep merging periodically, so again there's like half of these things where you'd say like percentage do you think should come

From catalogues and actually, should a self-respecting data sourcing team what percentage should be coming from off-catalogue companies? Yeah, I mean So while catalog guys are friends of mine, so I feel bad saying it, but I mean on some level I think like if you're really gonna have a really high performing data team, they shouldn't need a catalog. I mean, like, like that is not the people who should be using the catalog. It's fine for the PMs and analysts to use that. Some people enjoy that.

Um they should be doing that and kind of dipping their toes in and whatever. But as a data team, I mean, frankly, if you're having people that are referring back to the catalog all the time, they're not out in the world enough. They don't know enough of what's out.

more than that. Otherwise just employ the capital to do more than that. I think you know in terms of the Which some people do. They just give access to the catalogue to all of their PMs and and off you go. I think that's the best use of it, honestly, because that's where the people can, you know, for for the people that are more visual learners, um, who don't want to sit down and talk to folks who don't have a strong sense of what they want, it's it's it's worth appreciating.

And by the same token, um, if the folks that are doing the real work are are busy looking through that, it's like, well, what are you doing with the rest of your time? It doesn't take that. Um, I think, you know, with regard to, you know, the breakdown of everything else, I I think again it's very strategy and environment specific. So, you know, having worked, you know, back in the day at Blue Mountain.

You know, first and foremost, the giant bond fund. Um, there's, you know, a slower cadence to those markets. And so as a result, you know, the data flows in certain ways, use it in certain ways, it has primacy. Um, I always think it's helpful, you know, sort of to be a good data person in the data community to always be thinking about like where are you? What is the firm? What is the role you're in? What is the what is the theoretical competitive advantage you can bring to that?

I was at NASDAQ, you know, trying to find, you know, partnerships for Quandle. It was always, look, what is the competitive advantage that somebody will have in partnering with a Nasdaq Quandle and what are the disadvantages? Because it's a huge organization, if somebody came along and it was a partnership that was like, we can make$10, yeah, and that's not really going to move the needle.

Um, so there's going to be, you know, a need for the size to be larger, but there is a certain amount of prestige and air cover and other things that come from parts. stock exchange. And so if you understand those benefits that each situation brings and and and debt that it brings to every scenario, then you can kind of, you know, e in in in each interaction look and say, why is this the right one to be having? But it's super situational. I mean, if you're at Citadel

You're have a one specific positioning, if you're with some startup fund, it's a very, very different story. So you really just have to understand that and and understand.

your own your own universe if you will and then say, Okay, what am I bringing table and and be upfront about that with people that not every partner's gonna be the ideal partner. Not every vendor is gonna be the ideal vendor. If if I'm trading, you know, infrequent bond data and somebody says, I have sub nanosecond, whatever, it's like, yeah, probably not a good fit. And that's not a problem. It's just you don't you're not doing the same

I don't know if that answered the question. I don't remember what the question was. Well, I wanted to try and pin you on percentages. Like if you had a hundred data sets and you're in a big fund and you work for a bit a few big funds, then how many of those are coming from bank?

exchanges. Yeah, really. How many of those are coming from big data providers? Yeah, so many of those are coming. So so as a perfect example there, I mean like Blue Mountain was a great big fund, but it it it it was not the same style fund overall as, you know, millennia. Back in the day. Now. And so when you look at those, I don't think you can put percentages on them. I I think it's way too specific to the strategy that people are running.

And even for bit like like yes, there are certain big funds that look, you know, reasonably similar from the outside. I mean, to me, if I look at like dot on No I don't even want to put names on it. Don't put names on it. Don't put names on it. Like the more the more anonymous it is, the freer you should be. If i if if I look at Jane in Hudson River, sorry for the people that are there if you don't feel that way. To me as an outsider, they look very similar.

But I mean i i in most other cases I think everything is so very specific at the ground level of, you know, exactly which markets are you're trading. With cadence are you trading? It's all very specific to those use cases. And I think honestly it's impossible to put a like like it oughta be this, it ought to be this.

In a perfect world you want to be using everything from everyone all the time. Um but obviously there's a million reasons why that's challenging for money and time and every other thing. Okay. Um okay. Are you seeing any assets now getting covered by data which where previously there were none? And I'm probably talking alternative data here. Obviously, you know, Mark. But actually, you know, new forms of alternative data coming to help.

To trade an asset. It is really interesting. I mean, I remember very distinctly back when I was at Blue Mountain. I don't know, I was have to look at my link. We'll call it twenty fourteen. It probably was later. Um and I remember getting data from the folks at that time, it was market, now part of SP. Um, and we got like for something we got all the loan data. And I remember they emailed it over. What are you talking about? And like all the loan data and I was like, there's a trillion loans

Like and and this is all the history. How does this file fit in my email? Why is it only ten thousand rows? And they were like, This is all of the loans that have ever traded in the secondary market like more than a couple of times. I was like, Are you serious? Like like that's all of them? So you've seen a lot of expansion in data since that time aro around alternative data and non alternative data, around just as the markets have expanded.

Trading those markets has expanded. You've seen a lot of foreign growth in a variety of things. Um, but also you see um I had a coherent thought. I totally ran out of my Um, you also see, you know, a lot of non n non public equity traded names that are more important. the moment. So I think you have seen that expansion across all areas um broadly and it's it's it's been interesting. Um again I think I forgot

Any asset any assets which previously I mean Right, yeah, okay. Um I mean equ equities is probably where alternative data began and maybe there's been an an increase of macro usage recently. But have they I don't know, yeah.

There's definitely been growth in the macro side. There's definitely been growth, you know, in sort of the future side. I think, you know, people have been trying to electronify the bond markets in in a variety of ways and electronify all non electronic markets for a long time. So I

pushing out in those areas. But I don't think there's been like massive C changes in that. There is definitely a lot more data in l in sort of the futures and options space than there used to be. Um but in p in part that's also because those markets have expanded more than they It's uh I think it's largely more of a global expansion, exchange expansion front in that area. per se. Um have you got any interesting uh tips in how uh data is found not in catalog?

Um, I mean to me the most interesting folks and and and and I I I guess I'm happy to say it'cause I'm not trying to source data, you know, at the moment, but also, you know, like man, everyone's gonna know Um is it's always been more interesting to me when you see a large entity. um coming to a place to have a conversation about. So somebody from like a really large company making up, you know, like like Walmart.

like that. Um, and and at one time there was a friend of a lot of ours who I think was there for a period of time with And it was a kind of thing like when that happens, those are the really interesting folks to connect with because if you had to if you tried to call Walmart and just say like, hey, I want to talk to the data.

You are never gonna find that person ever. That place is way too large. Whereas for a startup data company, if you call them, they have you know five employees. If you call and say, hi, I want to talk to the data, you'll be talking to that person in 10 minutes. So I think Finding the navigable ways through large complex organizations or more opaque organizations. Um that's always been the I I I I think the more

framework and the place to go and that's the trick because finding those people in those places is really hard. And I'll say that even in the large organizations that that that that are good at this, we'll call them, you know, stock Um, as you tal and and futures exchanges, as you meet people that are really good at those places over time.

Um really connecting with them and staying in touch with them is important because again, if you have a thousand employees at a place that are doing all range of things, knowing the four people that can really get things done who can really cut

can really help with you know understanding and communicating and translating those organizations is profoundly valuable. So I I I I I think that would be the big tip is like the people inside the big organizations are the ones to really find because the little organization A lot of people think the most interesting stuff is the start up y thing that started eight minutes ago and it's like it can be but very often doesn't have a doesn't have any history. Well yeah, that too. Um yes, so yes.

Someone like Jeremy moves to someone like Walmart. then that is great. I didn't say Jeremy, you said Jeremy Jeremy Who. Uh but uh but anyway, moves to Walmart and um so that's immediately like a ready made and the fact that someone is there with

that background means that, you know, Walmart are exploring the space. Yes. When there are so when something uh fortunate like that doesn't happen, then you've got big companies with lots of data who haven't necessarily thought sell data to h hedge funds.

Um and so then you're embarking on a really long process because you need to not only pi find the right person to pitch the idea to, but then they need to get by Unless you're talking to the CEO.

They need to get buy in further up. Oh, even the CEO has to get buy in. I mean, we've heard plenty I I've heard plenty of times in a variety of contexts where, you know, people were doing things and like the board came in and said, This isn't a thing or the new CFO came in and was like So it it really is in those moments when you get a chance and and these relationships.

Years. Um but where you that was coming my question. Like how long how long how long would you expect it to take? You've reached out, you've found someone who is relatively senior, you've uh how long would you expect this unlocking a fresh data source, how long would you expect

I expect it to I expect it to take forever and that it's never gonna happen, which is why from my point of view it's about putting planting as many seeds in the ground as you can and investing as much time as you can to try to help

seeds grow and then if some of them come back as plants in three months, six months, one year, two years, ten years, um, that's just a happy thing and it's great when you can do it. And and there's absolutely no time or any of that'cause there are just so many factors that can go into these things for certain places, concerns.

And changes of personnel and market forces and regulation and you name it. So in those cases I think you just do as much as you can as often as you can and when w when and if they come back, you're just thrilled and say, Ah, look at the wonderful thing I made. Um it's really just that. I don't th I i if you try to set your watch pod you're gonna be very Is it worth Is it worth all that effort?

Two hundred percent. It's it's it it it's the best thing. It it it's it's such a it feels good. I I don't I'm I'm not I'm not I'm not I'm I'm not surprised that it feels good. But the question is, you know, we aren't we don't live in a world of exclusivity. on on the whole, normally, unless you're gonna tell me that uh you I mean some do. But uh increasing increasingly so in some cases. Um yes. But I mean but the point is the but the point is what you're getting yourself as a head start.

Um and actually if you put in a year's work uh tending to this seed and watering it and revisiting it and then you've got a head start at the end, is it is that enough payoff for all of that effort?

I think I I think it is, but I think it also goes back to the understanding where you're coming from and that if you're i i if you're from one of those places where, you know, the they lean into the exclusivity thing, it's like perhaps you get there first and you do say like

Let's do this just us. And you have those conversations with, you know, just those couple of places. If you're not that place, then it's like, look, you might never get a look at these kind of things. It might be the you would never be at the table of, you know, the two, three, five, eight people. So by being very open and helping create it and do that, it's the only way you get a seat at that at all.

Um, but also by helping be there as you put it together, if you if you can do that, you really understand how the levers and the gears work and you have a lot more control. You can tailor it to your specifications to a degree. And I think all of those All of those facets have utility, but you kind of have to start from a, this is the fun, this is the joyful part. And then again, the expectation is these things never turn into anything. Like, like it's just it's not going to happen.

And so when they do, you have to be happy first and then say, boy, I hope this works out because when you start, you don't really know what you're gonna get. And it's possible that by assembling this thing that is unique in the world and amazing data from an amazing place, somehow it is exactly equal to Yesterday's closing price. And we'll be like, wow, that would be really a shame if it's if it if it's statistically equivalent, but the it's it's very rare you know that ahead.

Hm. If you've so when you hear so you think these two five eight limited exclusivity opportunities, you think that there is when you hear that, do your ears prick up? Do you think that that is uh that's where the that's where the action is these days? I think that's where other people think it is, in the sense if you hear about people spending, you know, tens of millions of dollars for some form of exclusivity

Um clearly somebody thinks it's worth that much and that is interesting. Um and then it becomes, well, okay, if somebody feels, you know, that way about that, is it because of their positioning? uniquely useful to them? Is it that they really do feel that if a bunch of people had it, it would be problematic? And then, you know, how do you attack that problem? What's interesting about it? Are there other ways to develop? How do you how do you think of it? Um

It's an interesting problem. It it's it's it's it's it's a fun logic puzzle to try to solve and I think at the end of the day for, you know, a lot of people in the space, um, that can be the fun side of things is figuring out how to reverse engineer. work through those processes. So It's fine, is I guess I'm trying to say. How does uh uh how having been on both buy side and provider side, um how how has that influenced your approach to price of a data set? Um I don't think being on both sides

feeling about price. I think being around this space for a long time has influenced my feeling about price, which is look, if you talk to The majority of end user PM folks, they're just like cheapest price, do anything you need to do that. And I certainly know that there's people who run. Um, I think that as you build relationships with people over time and you've been here for a while.

Price is important, but there's other things that can also matter in terms of responsiveness, in terms of if things break, in terms of getting first looking at the first. And so you have to be sensitive to that trade-off at various times to say, like, if if I get you to give me the absolute lowest price humanly imaginable. then you are very unlikely to give me any goodwill or grace in any other areas. And so the question becomes,

You know, what what what are you willing to do and how does that work? And and do you want more flexibility and relationship capital or do you want more actual dollars? And I think you need to take a long view of that because it is something that, you know.

data person at a fund, you're representing the fund. And an analyst will be like, I just want this for five bucks. And it's like, yeah, that's your view, because you have to deal with this in these next 10 minutes. But over the next few years, I am thinking more towards you know, what is it?

renewal and how does this go and what is happening and what about the new products you may be rolling out and you want to save that powder for that relationship and not just give it all up to the one guy who's like save me eight dollars tomorrow.

Um, there's always a trade-off on both sides. And I think the more that you can be more upfront and honest about people's preferences and priorities, then hopefully you can find a middle ground that works for everyone. So you can have a a positive relationship because it is a relationship. I think um I think I think you are using the yeah.

Both sides there,'cause it sounds to me like you're remembering being a provider and thinking, I mean, did you did you have in mind wh with your customers, Oh, that guy's paying double what that guy's paying, so I'm gonna have I'm gonna give him more service essentially? Yeah. Um I never think in those terms, but by the same token, I wasn't the person servicing those accounts. So it's it's look again, it's it's you can speak the same language in terms

English and and communicate differently. There's no question that sales and support people are incented in certain ways and have certain drivers and very and PMs are ha very often have a very different makeup and sort of how they are compensated and how they think about side. And I think that's where you have to just again be aware of those choices. and say, you know, how can you bridge those gaps well and deliberately and and and serve the larger purpose while

um serving the actual larger purpose. I guess the way to put it is is is to serve the larger purpose if you're at a fund to say like look the goal is to have the fund make money so everybody So that's what you're there for. Um, and by the same token, that only works if you have, you know, the raw materials in which to run your factory. And that's, you know, the data and the research and everything you're taking in. So you want those processes to work'cause if they shut off, um

you're in serious trouble. So it's it's it's how do you do that and how do you think that through. So are there any um data types that are popular or Still that still ha still hanging around.

There's th th there's always the new hotness, right? Like I remember, you know, going a while back, there was like, you know, oh there were twelve trillion location data providers and geolocation providers and satellite providers and you know there was A bit of a story there for, you know, them, but not as much, which is why we don't see ten million providers like that.

Was that a COVID? Was that a COVID story? I mean, pro part of that probably. Um, but by the same token, you know, there were bits that just weren't working before that.

Um, I think now look, there's so very many, you know, folks with like AI and providers on everything now. And I'm sure You know, there's utility in in in some of them, but the idea that everything is an AI provider of anything, no. And I think, you know, when you think about some of the other I I think anytime you have people taking publicly available

And synthesizing it into something. And I think, you know, about like things like patent data, where over the years there's been a lot of folks with patent data. And there is some value in patent data, but

If people are saying, like, we can get this data for free from a free website and oh, we're going to process it a bunch, and like, ah, there's 80 of us with this value. It's like, guys, it's public data. Like, let's be clear. You're not taking some specialty thing and turning it into something.

Um, you have to wonder whether, you know, if you look around and say like 50 other people have already done already, it's like, am I really bringing anything to this? Like it was really so psychotically valuable.

there would be a better story here. And so I think any time you're dealing with truly public information, you have to really think through and say, like, is this is there really something here that someone hasn't tapped into yet and and d does this really What would you what would you uh look to use?

I wouldn't personally look to use it for much, but that's because the most of the funds I've been at have not had have had very very short term kind of investment horizons. I think the longer the horizon, there's a real story there. I think, you know, in talking with some folks you know, with like sovereigns and things like that. They have very, very long horizons. Fascinating conversation I had some years ago.

some folks at one of the sovereign funds and they were talking about how like they're tracking things like population trends because if there's gonna be a lot more babies in ten years, they're like, We should buy the Pampers Company because people will need more diapers for all of their children. Um, which is not, you know, the way hedge funds think about things. You don't think about like population trends. That just isn't the time.

So it's um you know, again, I think all of this is very situational to where you are and how to think of it and what you're doing. So there are definitely places where patents are incredibly important, but if you're trading, you know, intraday equities, probably Are there any geographies?

Um are there any geographies that you see. I don't know. There's incre th th there's increasingly more in in in the Middle East in Dubai now, so you know, it's tough to say. But any sorry you were saying. No, but they're still trading the US, aren't they? Yes, of course. But um where are there any geographies for data that you see as a kind

Eldorado, not an El that's a bit extreme, but uh a world of opportunity right now. I mean obviously China for a really long time, just because like data there is complex and dealing with it is very complex. So that's been the for a very long period of time. I y I I think Europe overall continues to be

Because I mean like the normal thing PMs would say is they'd be like, I want the data for the US and China and Europe. And it's like two of those things are a country, and the third one is a collection of many, many countries that don't speak the same language and don't have You know, privacy laws and every other thing. And so, like getting that getting that entity together and you know the folks that have put in the work to actually collect.

Across that massive of that massive countries that doesn't include the UK because you know Brexit. Um, and so like it's really it's really a challenging exercise and uh it's it's why it's an interesting exercise. But it's also a mental thing of thinking about how the world has changed in data. So, you know, coming from the US.

You look at like you know credit card transaction data and you're like, Yeah. And then you look at, you know, a large European country, Spain, and you're like and you look at the transaction data and you're like, This is so much smaller. It's like the country's smaller, the way they pay for things is different, all these things is different. And you have to really to reframe your mindset of

How does this thing that I'm used to in one scenario look in other places? And from a data perspective, that's a really interesting but also challenging problem. What's the trick to nailing China? Um

When I figure it out, I'll let you know. Um it's it's it's it's just it's an every boots on the boot boots on the ground? Um Th there's an element of that, but I think also it's really just I think you have to be really pure of thought in terms of what it is you really want to accomplish and then say how to get there.'Cause if you go in from like a give me all the information about everything, it's

Again, you look at like you know the credit card transaction data in China that's been around forever and it's like it it doesn't say that much'cause that's not how people pay for So you have to have a very, very, very defined sense of what you're looking for. And then in many cases sit on a mountain and have a vision quest and say like, okay, how can you know this given you know what's out there and and how is it triangulatable?

And I think that's where the creativity really really has to come through because in most cases you can't directly observe things. The data you're getting they will not say what it appears to say or or be reliable in that regard. So it's like what are the touch points and and that's where you really have to know w where things are coming from, how they're being developed and how to how to really work procedurally back all the way to say like is this doing the thing that I

But Evan, are you sitting in your in your favourite sushi bar in in New York having this deep five? No, they closed it. I'm so sad. Sorry, what? Uh or are you like do you need to go? Like do you need to become a Chinese uh specialist in order to understand exactly how credit c how they how they do their payments or are you reading a whole load of research about Chinese payments or are you hiring people there who can tell you about it? How are you how are you developing?

I do think speaking Chinese and and and to be clear, I do not. Um I do I I do think speaking Chinese is is a big thing in in in that regard. I do think there are geographies where language there is a different communicative way and a different openness you get in that regard. So China in particular, I think. you know, having Chinese language people to have those conversations is is meaningful.

having people that are I don't think there has to be boots on the ground in China, but like boots in the area is, you know, not insignificant. Um again'cause there's that sense of like, ah, our team, other team. Um and and then really just being very uh very aggressive in uh aggressive but open. Aggressive in, you know, sort of how many people you talk to and how many rocks you kick over, but open in the sense of saying like please

Anyone with anything, let's talk because it's you need to uncover these things and find odd ways into things. And I think that takes a lot of time and energy in relationships. Um who in our space is most at threat from AI? Um I mean w when you say in our space you mean like in terms of like people and their jobs or you mean like organizations or people, people, people. And they're just job descriptions essentially within hedge funds, data, all these things.

Data engineering, I mean, honestly, is is is a very obvious one because it's not a place you're winning in most cases. It's not not like anyone's like, I'm loading the files so much better than everybody else that like you're winning there. It is commodity, has been for a while.

Um so like if you could push a button and make all of that just work, uh I think basically everybody would. So that is that is significant and you have to really think about that positioning and what that And the value you're bringing in that exercise in the long run, that's the opposite.

I saw um Noah Smith's blog this week and he was talking about the fact that uh actually I think th I think he was implying that people have uh seen coding as being kind of art artisan like um and and kind of a deeply skilled trade like they did in northern Italy in the in the Middle Ages. Um but in truth a lot of coding is actually Waterford crystal encoding, same thing.

It's it's it's uh it's repetitive drudgery quite a lot of the time. And actually repetitive drudgery is exactly what can be automated and that's what we're going to see happen. I think that's where you really get into the like you know, you want people to be doing the clever thing, the insightful thing, the thing that really adds value on top of the thing you're taking in. If if all you're doing is taking it, loading it and making it available.

Um, that's a challenge. I think, you know, w uh a a personal bugaboo for me over the years has been, you know, they they have these roles in places they refer to as chief data officers. And very often like you look at the description of those roles and it's like database availability and security and whatever, that's not a chief of anything. It's like you it's an availability job. It's a glorified database. And so and now all the CDOs who listen to this are gonna be pissed off. But

Um but more than that it's like how do you take that and now what do you do with it? What is the thought press like do I you can bring in data as an engineer and actually add value and elevate it and say, ah, here's the clever way we link A to B and what does that mean? And like that is no joke. That is where like making those decisions

in tying data sets together and deciding the right way to do that and what to include and what to not include and what should be an option and what are the choices, that is a very, very valuable, powerful thing that needs to happen. Um If you're just, you know, loading it up and saying, here, that's not. And so you really have to say, like, what are the parts of this that involve thought and deliberation and choices?

and really focus in on those thoughts and deliberations and those choices because that's where you can lend the most value. That's where experience has value. That's where expertise lives. But also that's the thing that a computer doesn't traditionally make. Amazing choices and if it's the nature of it is a black box is like the one thing you never really want is a black box of choices you don't realize got chosen.

But don't you need thirty years of Daniel Sann style wax on, wax off drudgery in order to then be in a position to make the informed choices with your judgment that you've earned through blood sweat sweat and toil? Um I don't know,'cause as somebody that's had that, that's just how I got there. Like you'd have to ask somebody younger and more dynamic. I think, you know.

The look, ask the old people w w before they're gone and and find out how it comes together and that's some skill to learn. But also look, it's all present. I think like to the degree that people can go and look at that and understand it. I mean I I I do think it's interesting that um

You know, there are folks that sort of delve into the history of things and you don't ha in some ways everything in the financial services I think is always a history lesson. I think, you know, a as a random plug, there's a guy uh who has a substack called the terminalist.

Um and I very much enjoy the writing because it brings in elements of the history of things that even I didn't know. Um and it is interesting to understand where those things came from because all those choices that were made a long time ago are still present. in um i in in the data we see now today in all of those things. I remember a while back they were talking about adding

you know, a couple of uh decimal points um to one of the feeds and I was talking to a younger person, they were like, Like, why is this a big deal? I was like,'Cause they're transmitting a huge amount of data and they didn't get it and it's like if you remember back when you would download things over like a modem line and it took forever

It's like you add enough data points to something, any pipe gets more packed and they were like, Oh, I've only downloaded fast things. I didn't know that could happen. It was like it's still there, it's just bigger. It's just harder to run into those thr uh in into into those throughput limits, but they do happen, they can happen

And so as you think through those choices and understand how we got here, I think there's lessons to be learned in all of it. Um a lot of folks don't want to care about that. They just want to be here in the present and the now. But I think if you really want to do a really great job of what you're doing and you really want You know. And you really want to make a career of it and love it, then you should, you know, invest in the history of it and the time of it.

Th th there's a lot to be gained from there. So I think you can get thirty years of of experience by being an aggressive consumer and and patron of these kinds of things. But but you do have to seek it out. It doesn't just fall in your lap. It doesn't happen in your day to day. What is your big new um project? You've m left Veritian and now you are Chief Product Officer and Head of AI at B W Global. So take it.

Indeed. Um, so really, I mean, when I think to all of the really interesting conversations I've had in the last few years have really been, you know, around elements of AI. be done there. So, you know, BWG is is is the is the collection of two fabulous research shops that you know I've known for a long time. Uh BWG on the one Um and Creating a lot of interesting research and really and and really putting in the work and really putting in the expertise.

to bring these things together. And there's a lot of bits of that that are you know not what are just a pen portrait of P W G and B W G and O T R just for those who don't know. Writing actual research. So so so writing, you know, in in sort of the way a reporter might, like collecting data from the world or collecting, you know, viewpoints from the world and then writing, you know, sort of informed outputs of that, getting together on the B W G side.

panels of experts, so sort of like an expert network, but really different from them in the sense of getting experts together in a panel. to speak collectively. So you're not just asking one person questions, but hearing, you know, four or five or however many people speak collectively in their space about topics and getting to hear that conversation and put questions into that conversation.

Which is very, very different than pretty much what everyone else is putting out there. And also data. So Expert Networks is just where there's interesting. But you're talking about com like uh recorded, arranged conversations between experts, and you are a a fly on the wall.

At blessing in and feeding feeding questions into the group. Yes, which is a very different framework and one that I'm not really aware that anyone else has, and they've been doing it for a very long time. And so I think the process of creating those and then moving as we are into an AI driven world where people want to be consuming

content more through prompting and things of that nature. How to bridge those gaps to me is a very interesting question and a fabulous opportunity to really try to transform data and research and how we think about it. think that's a really interesting problem to solve. BWG as an entity is sort of, I think, very well positioned by virtue of doing these range of things and one or two things that are very different than everybody else.

um to really, you know, be creative and thoughtful and sit on a mountaintop and do some things that other folks are not doing and hopefully find a path. forward my experience and and you know, I'm I'm curious about yours. certainly heard from others is that you know in the AI world we live in, most people, I think the usual thing as a data team now is you sign up a bunch of data and you get the thing that says you can't load my data into an LLM.

This is inconvenient. Um, and depending on who it is, you're like, well, I guess we're just not doing it that way. But that's not making the consumers of this happy. The people that are creating the data, they're writing that in because they're too worried about where it's going to go because they're not.

Um, and so trying to figure out how to solve, how to get content that you know you you you make the time and energy and effort to create and being not just like okay with it being put into an LLM, but knowing it needs to be there, being happy to have

put in there, you know, running it through an MCP. Um figuring out the best way to make that process work for all constituents in that process and really have it make sense, that to me is a really interesting problem that is really necessary and worth solving and and that's the one that we're trying to solve. PWG and hopefully uh create a fabulous model for everyone in the world for how to uh put their data. Stay tuned.

I will do. Um, is there anyone you would like to hear from who might be listening to this podcast and should reach out to you? You're a very friendly guy, so you probably want to hear from everyone just have a chat. But yeah, I was gonna say that. I mean like

Honestly, it's it's it's if if you think you have an interesting angle on anything to chat about, you know, really anything to do with data, anything, anything, it it's always interesting. And I think like the biggest thing and I I I said this

So I know I was on Matea's thing a little while ago and we talked a little smack about you on that one. So to be clear, I can't start talking about Mate because we already did that one so we'll have to find a third person but I am truly thrilled to finally do this'cause we've been talking about doing this you and I for a long, long, long, long time and to actually do It's truly a joy and a delight.

Um very much and very very much like Michelle Obama, when they go low, I go high. I think uh that it's a great episode, yours and Matei, and um I point everyone towards it. I would strongly recommend everyone listen to it. It's got your video as well. So you get it That's good I I I I was gonna say that's'cause Matei's a good looking guy and between you a and between you and I collectively like he

He's bringing a lot more gravitas and and physical appearance than you and I can lend to literally anything. Nobody knows nobody knows what I look like. I might be a I might be just a brain in a jar. I know exact I know exactly what you look like. And and and and you are doing far better than I in that regard, if I may say. Because I don't think we I think we've covered different different ground.

Do them both. Um but awesome. But but regardless to go back to the original thing. No, really it it's it's it's always happy to talk to anyone about any of these problems and challenges because that is That's why we do this. That's where you find the interesting stuff. That's where the creation and the kismet and the magic happens, and it's in so many spaces and so many opportunities. And I think as long as you don't have a limited

of this is what you need in the next ten minutes for like, you know, one year, two years, fifty years. What are the things you can make happen? That's that's where the interesting stuff comes up and that's how also how you meet nice people. Like, you know, Mark. So Fantastic. Evan, it's been an absolute pleasure. It was well worth the wait. It's been a long time coming. Um thank you very much for joining and uh I look forward to seeing you in in New York very soon, no doubt. Very much so.