If you’re one of the few people to interview with Point72 Asset Management, you’ll almost certainly be asked a question that gives an insight into where Steve Cohen’s family office thinks hedge funds are heading.
“A model predicts a company's revenues based on four third-party data sources. They suggest that the company’s revenues will be X, but consensus is Y. How do you know how to make a bet? Coming up with an answer requires a whole load of statistical analysis related to that,” said Matthew Granade, chief market intelligence officer at Point72 Asset Management, speaking at the London School of Economics Alternative Investment Conference today.
If you’re starting out in finance today, you need to understand how the industry is evolving over the next five years, he said. This doesn’t just mean that computers are taking over, but that the people involved in the process (or at least those that survive) need to understand that programming, statistical analysis and the traditional “creative” human processes need to be combined.
Granade is a big proponent of the so-called 'quantimental' approach – namely, using huge data sets from external sources to support the decisions of more traditional fundamental stock-pickers.
“At Point72 we have portfolio managers who use the best of both,” said Grenade who started his career at Bridgewater Associates, where he eventually became co-head of research before leaving in 2012.
Investors are not relying simply on sell-side research or corporate access to companies anymore, he says. Instead, vast quantities of data come from external sources and it needs to be interpreted correctly. The likes of credit card data, satellite images of crops to give an idea of potential yields, or geolocation information from mobile phones to get an insight into purchases, for example, are not necessarily best absorbed by human traders, he said.
“Hedge funds are learning from quants,” he says. “If you look at something like portfolio construction, for certain inputs the maths is pretty much a pre-determined thing. Discretionary hedge funds still have humans sitting there doing that work. A computer can do it better and faster than humans.”
Not surprisingly, Granade says skills are changing and anyone starting out right now needs not only to understand what is needed currently, but what’s going to be needed in the near future.
“It used to be that you started out in investment banking and then went into private equity. You got very good at financial modelling, amazing at Excel, and then maybe went into a hedge fund. This misses how the industry is changing,” he said.
Granade acknowledges that there’s going to be increasing demand for data scientists and quant analysts. But there will also be a “minimum bid” for the skills required of any analyst or portfolio manager, he said.
“Firstly, you’ll be great at the analytical process, be able to structure problems well and be able to apply the traditional creative investment process to data,” he said. “But you’ll also need statistical know-how and to understand quantitative techniques, as well as a technical ability to manipulate data using some sort of programming language – either Python or R.”
Finance is still an “apprenticeship” business, he said, so key to a successful career is joining a company that understands where the industry is heading.
“One of the things computers are not going to replace humans is during human contact – understanding what a person wants and why,” he added. “As an example, we send investors into to corporate access meetings to think about body language of the executives and those kind of things.”
Photo: Getty Images