The Cheat Sheet

Is it a buy or a sell, Mr Robot?

Jinoy Jose P | Updated on February 13, 2020 Published on February 13, 2020

Cut to the chase; are the machines taking over stock trading?

Well, hold your calls; for now at least. You’ve got enough time to sit back and analyse. That said, a new study from Indiana University in the US has found that very soon, robots may surpass humans at picking stocks. The study offers the first comprehensive analysis of robotic investment recommendations.

Okay, is this a ‘wow’ moment, or not?

Depends on where you’ve put your money. To elaborate, the study — Man versus Machine: A Comparison of Robo-Analyst and Traditional Research Analyst Investment Recommendations — by Braiden Coleman, Kenneth J Merkley and Joseph Pacelli of Indiana University’s Kelley School of Business, says investment recommendations generated by “Robo-Analysts” (human analyst-assisted computer programs) conducting automated research analysis are more balanced and in sync with new developments.

That’s cool! And Robots can’t be lazy and biased, I guess.

Well, that’s another matter, as there is ample proof available now showing algorithms can be biased. But what’s interesting here is, as the Indiana University scholars rightly suggest, advances in financial technology (fondly called fintech) have transformed product offerings in the financial services industry and are enhancing the way markets function.

Interesting. But how important a role does AI have in stock trading?

Pure-play AI is still a novelty in the markets, but algorithms and robots were embraced over a decade ago, and have been creating extreme fluctuations at will thanks to the wicked minds they belong to. As JP Morgan’s recent report nicely explained, more than 80 per cent of the stock market is now on autopilot mode now. Computers have taken over trading in almost all geographies, and complex algorithms have been making deep impacts in the markets, introducing new forms of financial frauds that pose gargantuan challenges to regulators and policymakers as more and more investors and traders paint their wagon to cash in on the new “Wild West” with the markets.

Scams are in the offing, I’m sure.

There are many tumbling out of the closet already, because stock markets have, as their critics say, an uncanny ability to be prone to manipulation. The fact is, traders love computers because they fetch them money. If you look at the history of the US financial markets, secretive hedge fund managers such as Renaissance Technologies, PDT Partners or DE Shaw benefited from such programs. If you remember, just last year in the US, reports suggested the existenceof algos that trigger trades based on tweets and create volatile swings in stock prices. As ace finance writer Michael Lewis notes in Flash Boys: A Wall Street Revolt, his interesting work on high-frequency trading manipulations in America: “The US financial markets had always been either corrupt or about to be corrupted.” Critics of financial markets say this is true for markets everywhere.


Now, to put the Indiana University study — which concludes that automation can replace at least part of the research task and generate value for individual investors at potentially lower costs — in perspective, with the amount of proof on the accuracy and reliability of algorithmic interventions in the markets (especially at picking stocks, the DNA-function of stock trading) there are chances that more and more people will start depending on such tools. At least some constituencies of analysts, mostly old-school puritans, believe that this herd mentality can cause many dangers in the foreseeable future.


In his nail-biting thriller novel The Fear Index, Robert Harris has introduced a financial dystopia where a scientist creates a computer software to manipulate the markets and make billions. The novel is a decent reminder of what can happen if things go wrong, and of the urgent need for regulators and market players to formulate ways to understand — and limit, if need be — the role of AI in financial markets, before it’s too late. As a start, and in a first, Nasdaq recently started using AI on its stock market activities to detect “irregular and potentially malicious trading activity”.

A weekly column that helps you ask the right questions

Published on February 13, 2020
This article is closed for comments.
Please Email the Editor