UPDATED BY
Sara B.T. Thiel | Mar 21, 2024

Investors have long tried to predict economic markets, often unsuccessfully. Lacking the gift of precognition, they've had to make educated guesses about what might happen based on research and intuition. 

Now they're increasingly relying on a powerful tech tool: machine learning.

A subset of artificial intelligence, machine learning employs algorithms to spot patterns in data and use that to make informed predictions about a subject's future behavior. In effect, computers can learn to perform actions without being explicitly programmed to do so.

Machine Learning for Trading

Machine learning is being implemented in trading and investments to better predict markets and execute trades at optimal times. “Robo-advisors” use algorithms to automatically buy and sell stocks and use pattern detection to monitor and predict the overall future health of global financial markets.

There's no shortage of data these days, and a lot of it can provide keen economic insights. The challenge is deciphering what's relevant and what's not. That's where machine learning comes in handy. 

In financial trading, it's used to parse massive piles of market data, find correlated patterns and apply mathematical analysis to predict where markets are heading. In some cases, trades are made almost instantly without human intervention.

As former Citibank CEO and fintech pioneer Walter Wriston put it, “Information about money has become almost as important as money itself.”

Deployed correctly, machine learning can provide that information more quickly and more accurately than traditional methods. It's no surprise, then, that money managers are so enthusiastic about its potential to improve their results.

In fact, a July 2018 survey of hedge fund professionals found that 56 percent use machine learning for a variety of tasks ranging from trade execution and risk management to idea generation and portfolio construction.

"We think changes in the investment landscape will be profound," Marko Kolanovic, the global head of macro quantitative and derivatives strategy at J.P. Morgan, has said. “Investment managers who are willing to adopt big data and machine learning will have an edge.”

Check out these companies that use machine learning to improve trading and bolster their bottom line.

 

Citadel

Location: Miami, Florida

How it’s using machine learning: Citadel is a global financial institution known for its technology infrastructure and quantitative research capabilities. In an interview at MIT, Citadel CEO and co-founder, Ken Griffin, discussed how the company uses machine learning: “In the market-making business,” Griffin said, “you see a real application for machine learning because you have so much data to parametrize the models with.” 

 

J.P. Morgan Chase

Location: New York City

How it's using machine learning: Because J.P. Morgan Chase works with thousands of companies and millions of customers around the world, it has access to troves of data about spending history and macroeconomic trends. In order to make more sense of that data, the bank uses big data analysis and machine learning to predict where markets are headed and keep track of variables that might affect market trends. 

 

Morgan Stanley

Location: New York City

How it's using machine learning: Morgan Stanley uses a number of machine learning-powered robo-advisors to help investors manage their wealth via algorithms that lead to better-informed investments. While these robo-advisors typically cost less than their human counterparts, the two often work in tandem.

 

Goldman Sachs

Location: New York City

How it's using machine learning: In 2000, Goldman Sachs employed 600 traders at its New York headquarters. By 2017, the company had traded all but two of those traders for 200 computer engineers. The reason: Goldman automated its trading process, swapping humans for computers that run complex algorithms and perform other types of analysis to predict which trades will be most profitable.

 

WaveBasis

Location: New York City

How it's using machine learning: According to a financial theory called the Elliott Wave principle, a market can be predicted by wave patterns that reflect economic trends and group psychology. The WaveBasis platform helps investors measure, track and analyze those waves — in part through machine learning algorithms — so they know when to start trading.

 

Bridgewater Associates

Location: Westport, Conn.

How it's using machine learning: Bridgewater Associates is a hedge fund that manages about $160 billion in assets and uses machine learning algorithms to automate investing. The company also uses machine learning as an automated “coach” that can guide workers and notify supervisors if employees are feeling overworked. 

 

Two Sigma

Location: New York City

How it's using machine learning: Two Sigma's tech-centric trading is guided in part by machine learning. The firm invests in public equity, fixed income and alternative investment markets across the globe.

 

Numerai

Location: San Francisco

How it's using machine learning: The open-source Numerai's trades are guided by machine learning algorithms from numerous data scientists who are compensated with the company's proprietary cryptocurrency. Since the value of that cryptocurrency is determined by Numerai's overall effectiveness, the scientists are incentivized to create the best trading algorithms possible. 

 

Tino IQ

Location: Campbell, Calif

How it's using machine learning: Tino IQ’s algorithms scan stocks across the market looking for specific patterns that indicate movement in one direction or another. Once the patterns are recognized and validated, the stocks appear on the Tino IQ app along with "buy" or "sell" recommendations.

 

Kavout

Location: Seattle

How it's using machine learning: Kavout is an investment platform that uses machine learning and big data to provide insights about stock trading. Regularly updated “K Scores” ranging from 1 to 9 help stock investors determine whether to buy (higher) or sell (lower). The company ranks between 3,600-3,800 stock tickers each day.

 

EquBot

Location: San Francisco

How it's using machine learning: EquBot uses both proprietary algorithms and IBM Watson to find promising investment opportunities. It creates AI-powered ETFs (exchange-traded funds, which contain a mix of stocks, bonds and commodities) that are designed to perform like professionally managed mutual funds but without the fees.

 

The Voleon Group

Location: Berkeley, Calif.

How it's using machine learning: Voleon uses machine learning algorithms and statistical models to parse large amounts of market data and make financial predictions.

 

Sigmoidal

Location: New York City

How it's using machine learning: Sigmoidal's machine learning consultants look for ways AI can benefits their clients, including those in the financial trading sector. The company designed a machine learning neural network that analyzes financial portfolios and predicts expected returns for each asset. 

 

Wealthfront

Location: Redwood City, Calif.

How it's using machine learning: Wealthfront is an automated financial advisor service and app that uses machine learning software and other forms of AI for financial planning, investment management and banking. Because there are no human advisors, fees are relatively low

 

Taaffeite Capital Management

Location: Bala Cynwyd, Penn.

How it's using machine learning: In a process also known as "quant investing," Taaffeite Capital Management uses machine learning algorithms to spot patterns in market data that ideally result in profitable returns. "The real opportunity is to develop general learning systems that are not focused on specific, known patterns," Taaffeite co-founder Howard Sio has said. "So instead of just finding trend-based patterns in market data, it is looking for all profitable patterns in data.  That is, you are no longer limited to profiting from simple patterns that you can easily describe and understand."

Great Companies Need Great People. That's Where We Come In.

Recruit With Us