As a finance major, whenever news comes out that artificial intelligences are set to become the future of financial trading I get a little nervous. Many people think that sure, we could eventually create a program that is able to make real time decisions based around the market. But a lot of other people say that while they would be able to analyze concrete data, there is still too much left to interpretation for that to truly happen. So which side is right, and should I switch my major at this point?
In the world of trading, some tech startups have already begun to experiment with using AI’s as their only way of trading. Back in January tech startup Aidyia turned on their full tie automated hedge fund program. On it’s first day of operation they saw a 2% return on their investment. ( source ) While this isn’t a significant return considering they could make more just investing into a safe S&P index, it does prove that we have the capabilities to make a positive return without human intervention. According to Preqin nearly 10% of all operating hedge funds in the world make a majority of their trades using computer generated models. By continuing to allow the computers to create more and more models, firms are hoping that the programs will begin to recognize patterns in the way that humans do and trade on what might happen in the future.
As AI continue to get smarter, their return on the investments will potentially be able to outgrow that of a conventional trader. This article from August outlines the newest player in the AI trading game, a program called “Emma”. What they say is different about this program is that it’s programmed to not only look at the securities market, but also what is happening on a global economic scale. In the six months before they had unveiled the project they had it trading a set of stocks that it determined would be smart investments. Since then, their portfolio size has grown by 30% and they are donating the extra proceeds that it makes to charity because for now they are just interested in seeing how far they can push this AI growth.
This isn’t just something that is happening in the U.S. market either. Later this year in Japan a new AI developed by Mizuho will work along side human traders in an effort to help them make more informed decisions. Their goal with this system is to use the computer to show historical trends and interpret that into the future, and then use the actual trader as more of an evaluation of whether it’s actually worth their investment.
For now I don’t believe that traders and finance students should be worried about AI’s completely replacing them, as they are probably more useful for simply aiding in the job. But it is an interesting future change in the way the industry operates and will become an important thing for anyone interested in the markets to know how to work with.