AI has become a mainstay in the financial services sector over the last decade, playing an important part in everything from cybersecurity to chatbots.

Yet, despite the huge amount of data generated by financial markets, machine learning and AI algorithms are still not a big part of the investment decision-making process.

In their study, Barbara Jacquelyn Sahakian, Fabio Cuzzolin and Wojtek Buczynski analysed 27 peer-reviewed studies by academic researchers published between 2000 and 2018 that described different kinds of stock market forecasting experiments using machine-learning algorithms.

The academics found problems with cherry picking in the studies: most of the experiments ran multiple versions of their investment model in parallel, with the authors almost always presenting the highest-performing model as the primary product of their experiment.

"This approach would not work in real-world investment management, where any given strategy can be executed only once, and its result is unambiguous profit or loss - there is no undoing of results."

This problem, along with a failure to account for the black box nature of AI algorithms and a lack of consideration of regulations, may explain why the market has not yet fully embraced AI.

In fact, the handful of AI-powered funds whose performance data were disclosed on publicly available market data sources generally underperformed in the market.

"As such, we concluded that there is currently a very strong case in favour of human analysts and managers. Despite all their imperfections, empirical evidence strongly suggests humans are currently ahead of AI."