How hedge funds employ AI to facilitate trading

Machines are now making trades at the world’s top hedge funds. But they’re not doing away with humans just yet.

Key takeaway

AI is proving relevant beyond the tech world. More and more hedge funds – including some of the top firms around the globe – are incorporating AI and machine learning into key trades and investment decisions.

Probabilistic Logic

Ben Goertzel and his company Aidyia made headlines in 2016 when they launched a hedge fund run entirely by AI. It makes trades on its own. Their core technologies include probabilistic logic – a concept created by Goertzel, among others, for use by MindAgents within the OpenCog Core. It involves “using probabilities in place of crisp (true/false) truth values, and fractional uncertainty in place of crisp known/unknown values.”

Aidyia’s machine analyzes “everything from market prices and volumes to macroeconomic data and corporate accounting documents.” These machines essentially make their own market predictions and then “vote” on the best course of action.

Funds from Sentient Technologies, to Two Sigma and Renaissance Technologies have been trading using AI for years now. According to Goertzel, as more firms adopt these technologies, the race for further innovation will be on. “If everyone is using something, it’s predictions will be priced into the market,” he warns in Wired. “You have to be doing something weird.”

The shift to machine-aided investment decisions

Executives at Winton Capital, one of Europe’s largest hedge funds with $30.6 billion under management, stated that at Winton nearly every vote on a prospective investment decision is made by a computer. However, they noted, when it comes to the big, final decisions, humans are still involved. Winton has an algorithmic approach to investment, and of their 400 employees, 200 are data scientists.

Of the top 10 highest-earning hedge fund managers in the world, today the top three are all ‘quants’ – managers that rely heavily on computer systems in their investments. Only four managers still use mainly traditional methods, where the decision-making is still primarily driven by human analysis.

Bridgewater Associates, notably the world’s largest hedge fund with $160B under management, built a machine learning algorithm taken straight from its employees’ brains, according to a NYT article from December 2016. Intended to do more than improve accuracy, billionaire founder Ray Dalio claims he wants to “ensure the company can run according to his vision even when he’s not there.”

Their team, the Systematized Intelligence Lab, is led by David Ferrucci who worked on IBM’s Watson. They have famously taken AI internally as well, with meetings recorded and staff asked to grade each other throughout the day using a ratings system called “dots”. These ratings are incorporated into “Baseball Cards” that show employees’ strengths and weaknesses.

Man vs Machine?

We’re nearing technology today that can tackle one of man’s most significant advantages: the ability to learn on his own. Today, deep learning and AI technologies can improve themselves based on exposure to new data inputs, by finding patterns.

Many hedge fund managers, however, make the case for continued human involvement. Winton notes, “There are big tasks at hedge funds ripe for automation, such as performing large-scale, recurring calculations for assessing risk across portfolios…[however,] people will be running software at every stage of the process.”

Jordi Visser, investment chief at Weiss, believes humans still have the upper hand when it comes to recognizing patterns, thanks to intuition. “The industry’s survivors will be the ones who imbibe technology into their processes,” he said.

The Outside Insight here?

The Outside Insight here? Computer-aided human decision-making is the next norm on Wall Street – with humans continuing to steer the course – and those that don’t invest in machine-learning technology to assist in decision-making will soon fall behind the curve.

Recent Articles