AI Will Not Replace the Analyst
Artificial intelligence and machine learning application may provide financial analysts with invaluable tools that process greater amounts of data, but they will not supplant the role of the research analyst completely, according to one industry watcher.
Much of the concept of a fully automated research process comes from the world of high-frequency trading and its quantitative trading strategies, noted Paul Rowady, research director at Alphacution Research Conservatory.
“To say that quantitative research is a fully automated function within the trading world is to confess your blindness to what actually is required of trading strategies with slower turnovers that require more diverse data sets,” he said.
Rowady also attributed the idea of a person-free research department as a knock-on effect of the continued consolidation within the high-frequency trading space like Virtu Financial’s planned acquisition of KCG Holdings.
As more firms drop out of the low-latency arms race, they are left with limited options. “Either they have to get out of the game because they cannot compete on speed or they have to go the other direction and find automated trading strategies that have slower turnover frequency are less dependent on speed,” he noted.
The changes might include trying different asset classes, regional markets, and longer holding periods.
“But with each of these moves, they are assuming that there isn’t any incremental experience necessary to make those moves successfully,” said Rowady. “Each of these shifts require different data and, therefore, different skills.”
Most high-frequency trading strategies that rely on sub-second updates and exit their positions at the end of the trading day do not need data outside of market data. “You don’t really care what the ticker symbol is or what industry it is in,” he said. “You care about its liquidity and volatility and that’s it. Ticker symbols just represent a different thing that you can trade.”
But as firms adopt strategies that hold positions longer than a day, these new strategies introduce gap risk in the open as well as margin, portfolio construction, and other concerns, which advance the complexity of the research function that cannot be fully automated for quite some time, according to Rowady.
The bank can access data science, artificial intelligence and machine learning for new products.
AiPEX with Watson simulates a team of analysts and traders to identify potential investments.
Machine learning models systematically scan newly arriving, anonymized data to identify anomalies.
The Cobalt programme launched in 2018 to help fintechs collaborate with the asset manager.
Users with different skill levels will be able to undertake machine learning and advanced analytics.