It is one of the hottest trends in investing.
Hedge fund managers in particular have always sought an edge over competitors; now they’re vying for new data sets that their competitors don’t have, or haven’t thought of using.
Still, many investment managers are struggling to put that data to good use. Almost a third of hedge fund and asset managers recently surveyed by Greenwich Associates said they had “difficulty understanding/working with data sets that are not customised.”
Funds are struggling to find the right people to hire to analyse all the new data, says Matei Zatreanu, founder of System2, which advises hedge funds on integrating alternative data to their investment strategy.
He helps run Augvest, which organizes events around alternative data. Previously, Zatreanu worked on data-science efforts at King Street Capital, a $19 billion New York-based hedge fund firm.
Business Insider spoke with Zatreanu about the future of alternative data and what kind of skill sets hedge funds are looking for when hiring for the new roles.
This interview has been edited for clarity and length.
Rachel Levy: What is the future of quantamental?
Matei Zatreanu: In short, it’s the future of investing. If you’re talking with people in finance programs — the Whartons of the world — and you ask those students in investment clubs how they think about investing, to them it feels natural that, obviously, you’re going to use any data to make an investment.
The analyst is still in control, but now they’re augmented by this data and technology. There’s still a human decision-maker, but their decisions are now more accurate, more predictive of what they’re trying to do. It just seems to me — and many others — like this is a natural progression of investing.
This is something that has been done for a long time, and there’s definitely a history of people doing deep due diligence on companies they’re investing in. With this data and technology, they’re doing that, but on a larger scale. In general, once you figure it out, it’s going to be the future of investing.
Levy: What kind of people are hedge funds looking to hire if they’re interested in building out their alternative data teams?
Zatreanu: It’s an interesting question, because you’re basically trying to figure out what the research analysts of the future will look like — what the technologists and PMs [portfolio managers] will look like. Each is a role that has existed with most funds, and going forward we’re still going to need them in some capacity, but they’re going to look slightly different, if not fundamentally different.
What does a junior analyst look like? Right now it’s typically someone who went to a good Ivy League school, then an investment-banking program at one of the top banks, then a hedge fund poached them to analyse the fundamentals of companies.
What we argue is that that’s no longer enough. They need to understand maths, statistics, and computer programming.
They don’t need to be experts in these things, but they need to have the language to communicate with people who are experts in them. For instance, you’re not expected to work at Macy’s before you can cover it and invest in it, but it certainly helps. You need to go through the s— to understand the subtleties, like how to listen to the nuances in someone’s voice on an earnings call.
We, as research analysts, have been abstracted from the businesses. We’re sitting in our Midtown offices analysing businesses throughout the world that most analysts have never stepped foot in. And yes, it helps to listen to the calls and read through everything the company puts out, but on the other hand, wouldn’t you much rather get much more granular on the details of the operations of that company? And that’s what data can offer, being able to look at how promotional a retailer is, how the variance across the different locations around the world, the differences understanding customers at a high level — not as summarized on top line of a financial statement, but understanding the individual customers. Like, “Why has Rachael continued shopping at Macy’s? Why has she stopped shopping at some other store when another store came online?”
Levy: How would you be able to tell that Rachael made that decision?
Zatreanu: It depends. You can look at a credit-card data. That gives info on an individual person. The huge difference is, unlike marketers, I don’t care that your name is Rachael and that you live at a certain address. I just care that a person with ID 345 used to shop at a Whole Foods but then — all of a sudden — an Aldi opens up in your neighbourhood.
And now I see that person with ID 345 no longer shops there and I see an Aldi transaction. That means that Whole Foods potentially lost a customer.
You can see what the customer loyalty looks like and how fragile is that loyalty. As soon as the new sexy store opens up, do my customers shift over? That can speak volumes to the future of a company … but like I said, I don’t care that it’s you, whereas the marketer does so they can send you advertising. I just care about the trend at large.
Levy: So it sounds like you need to have more of a hybrid analyst who knows this data is out there, on top of the other skills they’d already have. You have also said that sometimes you have a hard time finding people who meet that criteria and other skill sets, like finding people who could go out and get the data. Could you speak to the challenges in hiring?
Zatreanu: The skill sets that you now need for an analyst or data scientist at a hedge fund — you need them to have the typical data-science background, which is an understanding of maths and stats, computer programming, and also domain expertise in finance.
What I would add to that for our purposes is also someone who has the soft communication skills. People who can go out there and negotiate, and then a broader curiosity about how the world works. Like you know, you’re picking up a rental car and you ask the person at the desk a bunch of questions, like, “How many cars do you have in this garage now? How is business? How do your systems work?” Those are the components of a successful data scientist, analyst, whatever you want to call them.
The problem is, we’re competing for those sills sets with the likes of a Google or Facebook or some sexy startup, and there are a lot of factors going against hedge funds — one of which is that their very mysterious. They have very little publicity. They don’t like to talk to the press.
Whereas the other startups are all about advertising. And telling people who they are, and what they believe in. And that speaks to the whole millennial ethos about wanting to work for something more than the paycheck. It used to be extremely easy to hire at funds. But that’s gotten harder.
The funds realise they are competing with the Googles of the world and they have to do these kinds of things to make themselves look less like the villainous hedge fund and more like the cool tech startup.
It’s kind of interesting. A lot of it is psychology and optics, because at the end of the day, it’s the same problem. A lot of funds were very sensitive about their use of credit-card data hitting the news because, you know — “These evil hedge funds are looking at my credit-card statements knowing what I am buying.” But at the same time, you have Mark Zuckerberg knowing everything about you, including who you are talking to, who you like, and not just that, but affecting the results of political elections, by curating your news you have access to. It’s the same thing.
You have rich guy Mark Zuckerberg using people’s private data without them knowing it, but they don’t get vilified. For whatever reason, if you’re a tech entrepreneur, you created your money by creating a real tangible product as opposed to just hedge funds, which are seen as gamblers. But it’s an interesting PR battle they’re facing, especially after the financial crisis, where they’re trying to attract the top minds of the field.
Before the crisis, they had no trouble hiring; now it’s much more of a marketing push to figure out how to incentivise people to join them.
Levy: Some people say, “Quantamental is a big trend, people are getting into it,” but they also say it’s relatively easy to find people who can do this. Some critics will say that, in the end, they all end up doing the same thing, doing bottoms-up analysis of data without really understanding the bigger picture. What do you think? Is quantamental going to become crowded?
Zatreanu: Fundamentally, I disagree with that. It’s easy to hire a physics Ph.D. and just say, “Now I’m doing data science,” just checking the box. The problem is, those people don’t have the domain expertise. They have never worked in finance. They don’t understand what matters. It’s basically seen like the head of the chess club arguing with the head of the football team — the nerds versus the PMs, analysts, and traders. The nerds will get stuck on some problem that is intellectually stimulating and really interesting but has no business value.
You can find people who have technical skills. The problem is they don’t have the understanding of what matters to these companies or how these investments get made.
Levy: You’ve mentioned the social aspect as well. What’s the challenge?
Zatreanu: Having good communication skills is hard to find. Where that is important is sourcing data sets. Once you’ve identified an investment thesis and the relevant KPIs [key performance indicators] we’re trying to measure, we start the discovery process, which is basically a due-diligence initiative, where we will look at the data ecosystem, you have the target company we’re interested in, and then we try to figure out who their partners are, who their customer are, what charities do they donate to, what invoicing systems do they use, what technology do they use for their payment system.
Understanding all these components, where there might be data and where that data might legally be purchased. Then after that, you basically start calling them up and telling them this is what we’re trying to do and negotiating that and convincing them that it would be a good idea for them and figuring out a good price for it.
It’s not a skill set most people have. It’s extremely important when you’re trying to source a data set that no one has thought of before and you’re talking with someone who has never heard of a hedge fund before, and you have to explain to them why you want to them to send you their database.
Levy: Do a lot of these conversations start in person? Do you have to fly out to Kansas or wherever?
Zatreanu: A lot of it starts over email or LinkedIn, then an email, and after a phone call, we’ll go in person. I’ve flown to Sweden, for example, to speak with a guy who had a data company we were trying to get data from — also to Hong Kong and Singapore. It’s one of those things where I value conversations on the phone or in person much more than over email.
Levy: So you’re trying to find someone who can code, is an analyst, and can socialise.
Zatreanu: Not everyone on the team has to have these skill sets. You can obviously compromise on one over another. But if you’re talking about the person who wants to set up these teams, they should have these skill sets. They might not need to be an expert in programming or statistical theories, but they still need to have intelligent conversations with the people who are the experts, to call out the bulls—. If your developer tells you this is going to take forever, you can push back and say, “What if we do it this way?”
If you’re a portfolio manager and you’re talking to an analyst, you don’t understand the companies as well as your analysts do because you’re not as into the weeds of it, but you can still speak intelligently about it. It’s a similar kind of thing from a management perspective. That person needs to understand a little bit of everything. And then on the team you definitely need to have people with these different skill sets.