By – Mr. Jayant Manglik, President of Retail Distribution at Religare Broking Limited
Remember Paul the Octopus, whose accurate predictions in the football World Cup 2010 brought him global fame? Before each game, two food boxes with different team flags would be offered to him and the winning team would be selected depending on which box he ate from first.
With Paul’s record of 12 correct out of 14 publicly done predictions, soon other animals too were in the fray, including porcupines, hippopotamus and parakeets!
However, given the use of artificial intelligence with with high computing power, it is still much more logical than an octopus. But that begs the obvious question – can it be used to predict winning stocks? Is Goldman already using it for this purpose? Can it give an unfair edge to some, similar to high frequency trading?
Even if it is not exactly accurate (it cannot be or it will defeat the purpose of markets and lead to a completely new order), maybe it can be better than alternative stock selection models being used currently.
One could argue that the Warren Buffetts of the investing world do precisely this, except that they’re human and they do it mentally. In the case of machines, trading calls will need to analyse human behaviour too, but the long-term outcome of stock prices often depends on individual brilliance or personal integrity or macro or global factors, most of which may yet be outside the machine’s ability to comprehend and analyse.
The (un)certainty of a Kim Jong Un threat is always hard to evaluate, as is its impact on markets. In any case, the options are defined and built into the machine by humans, though artificial intelligence can add its two bit here.
In other words, the statistical part will work fine, but markets of course are simply too complex. And there aren’t just two teams here with 22 players, it’s about thousands of stocks with millions of people buying and selling simultaneously all over the world. Being able to predict markets with any level of certainty fundamentally involves predicting the broad future and probable impact of each event on us. That is a big ask for any machine, by any standard.
But what is inevitable is what is already happening. The use of technology in finance is omnipresent now. Fintech is everywhere. Raw computing power and technological innovation have disrupted many processes and outcomes. Increased use of AI and big data is inescapable. Cryptocurrencies and blockchain technology, regetch (regulatory technology) and insurtech (insurance tech, no prizes for guessing that one), ICOs and robo-advisors are already changing the face of the financial services industry.
Yet, markets have been and will always be near impossible to call consistently. For man or machine. That is the nature of markets – uncertainty is the underlying principle and fear and greed will always be factors which no amount of algos can possibly replicate. Technology will facilitate better analysis, but the only people who will always make money from new technology are its developers.
“Source The Economic Times”
