Could artificial intelligence someday control your stock portfolio? A North Carolina State University research project is looking at ways to use machine learning to maximize returns.
“Machine learning is one way of doing that because it can combine the data and the knowledge much better than a human being,” said Mehmet Caner, professor of economics and co-author of a recent paper on the project. His team developed and demonstrated an AI program that can help financiers meet specific risk and return goals when dealing with massive portfolios.
Caner’s research specifically deals with large portfolios – think hundreds of assets. And it centers on improving an established metric called the Sharpe Ratio, which helps investors understand returns versus risk.
Specifically, the ratio is the average return earned in excess of the risk-free rate per unit of volatility. It gets extremely complicated when a portfolio contains hundreds of stocks.
Many machine learning applications in the space are like a “black box,” they allow researchers to analyze what happened. But Caner’s project focuses on the future, on how to advise portfolio managers on what to buy, sell or keep in their portfolios.
Caner said the work shows it’s possible to use AI for just such advice – though he’s never actually used his technology to boost his own portfolio.
There are obvious commercial applications for his project. Caner cofounded a startup, Magnus AI, in Turkey, based on the technology.
While there’s nothing stateside in the works right now, he said licensing the technology some day is not out of the question.