Saudi start-up launches AI stockpicking app

• Almutanabi ‘predicts price directions’
• Approved as adviser, not broker
• Seeks $4m series A funding

A Saudi startup has launched what it says is the country’s first artificial intelligence-powered equity trading application, in the hope of persuading investors to trust their stock picks to a computer.

Almutanabi, which became available to paying customers on 30 August, offers trading suggestions for Saudi and global equities for a monthly fee. Its founders say the app currently has fewer than 100 users – all retail investors – but they hope to expand their customer base to include large institutional investors.

The technology is homegrown and built from scratch, according to Almutanabi’s chief of AI, Omar Alolayan.

The app compares a range of information related to stocks, including trading patterns, company cash flows and incremental share price changes, to make a prediction on where a company’s share price will go next, he said: “These are all of the things that nobody as a human can comprehend.”

Alolayan believes that the AI’s ability to process large amounts of data far quicker than humans can translate into better trades over time. But AI experts warn that the bar for beating the market is a high one, even for the best computers.

Other AI-powered stock pickers have been launched outside of Saudi Arabia. Investing platforms such as eToro, Public and TradingView all offer their own AI bots, offering support and advice to retail investors.

Alolayan believes that the concept is likely to grow and become a standard feature of equity trading, citing its ability to assess a decision with impartiality.

“Machines don’t have feelings,” said Alolayan. “Whenever it makes a decision, it makes it based on maths.”

The company is approved by the Capital Markets Authority to act as an adviser, but not a broker, Alolayan said, meaning that the application is unable to conduct trades on behalf of the user, who must approve suggested trades. Almutanabbi is in talks with Saudi trading apps to embed the technology with a brokerage and reduce the steps needed to approve the app’s stock picks.

The company is self-funded by Alolayan and co-founder Mohammad Alsallum and is hoping to raise around SAR10 million to SAR15 million ($2.66 million to $4 million) in a Series A funding round.

• War damage forces costly rethink of AI infrastructure
• Saudi stocks weathering oil export concerns
• GCC investors should look past Brazil’s election uncertainty

The use of AI in trading is likely to grow, but is not without its limitations, according to Vasant Dhar, professor of data science at New York University and author of Thinking with Machines: The Brave New World of AI.

“Predicting markets is the hardest prediction problem out there, period,” said Dhar.

AI offers some advantages over humans, he says, particularly in the speed at which it is able to operate. Computers already take the lead on so-called high frequency trading, where trades are often made in fractions of a second based on algorithmic calculations of market changes. And for more traditional day trading, Dhar said, AI is fast becoming a more standard tool among investors.

Ultimately, though, the unpredictability of markets, where stock movements are idiomatically equated to a “random walk” for their tendency to rise and fall, is a problem that could stump computers just as it has baffled humans.

“It doesn’t matter whether you’re using math or AI or God’s own arithmetic,” said Dhar. “Nothing will help you predict it, because the phenomenon is inherently random.”