Combining big data technology with artificial intelligence (AI), machines should one day be able to learn to invest on their own, according to experts.
Some recently-launched hedge funds already depend on computer modeling for trading operations.
There is data suggesting that computers can do a better job than humans, Che Pinjue, a partner at venture capital firm Sequoia, noted in the Hong Kong Economic Journal.
Taking in a huge amount of information from news and social platforms, the trading robots can feed the data into tens of thousands of prediction models and figure out a suitable trading strategy.
“Put simply, artificial intelligence system can create an army of digital traders, and test their trading performances against historical data,” says Che.
“The best ones will then be selected and their trading strategies further refined through rounds and rounds of such back testing. Finally, a super trader can be created, like the AlphaGo that beat world class Go player Lee Se-dol.”
In March this year, Google DeepMind’s AlphaGo beat Korean professional Go player Lee Sedol in a contest in Seoul, marking a huge victory for the field of artificial intelligence.
While AI is expected to be adopted more in the world of professional investing, robot traders will however work best only in highly transparent markets.
Their investment horizon is also expected to be relatively short, ranging from seconds to a few months at the most.
– Contact us at english@hkej.com
RC