The hedge fund world has been evolving dramatically over the last few years.

Just like in other industries, software, data and AI/ML have been playing an increasingly important, and disruptive, role.  Many hedge funds have been scrambling to embrace this evolution – not just to gain an edge, but also to avoid becoming extinct.

Certainly, quantitative hedge funds have been making heavy use of software and data for a while now.  The “quant” funds rely upon algorithmic or systematic strategies for their trades – meaning that they generally employ  automated trading rules rather than discretionary (human) ones, and they will trade tens or hundreds of assets simultaneously.

But another big part of the industry, the “fundamental” hedge funds, had been operating very differently.  Those funds will perform a bottoms up analysis on individual securities to  value them in the marketplace and assess whether they are  “undervalued” and “overvalued” assets.  They’ll often have a much more concentrated portfolio.

In part because the entire hedge fund industry has been performing generally poorly recently (years of performance trailing the stock market), there’s been mounting pressure on hedge funds to evolve rapidly, particularly fundamental ones.

A couple of years ago, Third Point made a big splash when they hired Matt Ober, who was 32 at the time, to become their Chief Data Scientist.  Dan Loeb, the billionaire founder of Third Point, was a prime example of a fund manager who had reached tremendous success through a fundamental approach.  His efforts to hire Matt away from his previous employer and make him Third Point’s head quant was widely viewed as a sign of the times.

I was fortunate to get to sit down with Matt at our most recent edition of Data Driven NYC.   The video is worth watching in its entirety, of course, but I jotted down a few notes below:

 

 

Leveraging data at fundamental funds vs quant funds: 

• At WorldQuant (quant fund), the thesis was more “if we consume more data than anybody else in the world, we can find more signal, create more alpha”
• At Third Point (fundamental fund), the thinking is more “there’s so much data out there, and so much noise… we’re more focused on using data not to mine new ideas, but to better understand the ideas we already have, or to create very sophisticated screens to dwindle down to a small list of names that we can then work on with a fundamental portfolio manager”

On alternative data:

• “In the early years, people used expert networks to really understand a healthcare or a tech company… now you can buy data to give you that insight… data is becoming more available from so many different vendors”
• “Alternative data” means any data that not financial or fundamental data… at first data may feel like it’s alternative but we’re moving to a world where we’re just using more data
• Our approach is what type of information can help us better understand a specific company…  like how sales are going to look like next quarter… so we ask ourselves, if we were that company’s data science team, what data would we want to see? And maybe tie it to macro economic data
• Satellite data is interesting, but I don’t think there’s a lot of firms out there counting cars in parking lots
• We spend a lot of time thinking about combining data sets, that’s a lot more exciting to us

On startups selling data to hedge funds:

• The multi-million dollar data sale is something you saw maybe 3 or 5 years ago… now people understand what data is actually worth, and they’re not trying to get exclusive access to just one data set, they know it’s part of a bigger mosaic”
• It’s becoming more transparent who has what data, we don’t need consultants to help us find the next data set
• Most hedge funds are open to seeing what data is out there, but you have to have something unique, with a lot of history… at least two years of daily data.
• Some funds won’t tell you much feedback, but the better data teams out there will provide feedback and help you build data products

On AI & Machine Learning:

• “Everybody is experimenting with machine learning and AI… but [for fundamentals], the simpler it is, the better”
• “AI/ML is more prevalent on the quant side… on the fundamental side, we have the explainability factor… [our investors] want to understand what the model actually does and can explain the factors… it’s more about delivering a list of names that they can look at, based of the machines, and then there’s a level of due diligence”

Impact on broader financial services industry:

• “Hedges funds are the first movers, but this will come all the way down” (mutual funds, etc.)
• AI taking jobs is always a concern in all industries, but for the more complex strategies (e.g. structured products), it’s about understand the pieces and AI will just be a toolkit that will enhance the industry overall

NOTE:  If this topic is of interest, see my previous post: “The New Gold Rush? Wall Street Wants your Data”