Algos Could Take Over Trading in Just Five Years, Experts Say

by Malynnda Littky
  • Strategies based on algorithms and predictive tools take over as machine learning grows more sophisticated.
Algos Could Take Over Trading in Just Five Years, Experts Say
iFX EXPO Asia Team
Join our Telegram channel

Last month’s iFX EXPO Asia 2019 panel on prediction tools explored the intersection of Algo Trading and machine learning. We gathered a team of data science evangelists to describe how investors can use AI-based trading tools to exploit nonlinear market patterns.

https://youtu.be/HP2ii6t-b5Q

Algo trading goes mainstream

In the last two years, the attitude towards AI trading models has shifted drastically, and most institutional investment firms are now either using the services of algo trading providers or developing their own predictive algorithms. In part, this has come about because traditional funds management strategies have failed to perform according to market expectations.

According to Ksenia Semenova, Chief Business Strategy Officer of Cindicator, the major driver towards the adoption of algorithmic trading was the introduction of tokenized assets. She noted that “Cryptocurrencies started with algo trading, and this is a trend where all trading will be switched to algo trading…”

Coders and data scientists dominate

The experts also took some time to discuss trends which have gone under the radar so far. More traders are using their coding skills to develop strategies specifically for use by other retail investors, while institutional clients are turning to data scientists to create algorithms.

AI masters imperfect information

Panelists also discussed the potential of machine learning. Tiantian Kullander, Co-Founder of Amber AI, pointed to the success of Google’s Deep Mind, which can now beat humans in real-world games involving imperfect information. Kullander stated that “If you think about the markets, these are also games of imperfect information, because nobody ever wakes up and buys a stock at random.”

Yaron Golgher, CEO of I Know First, added that AI can improve trading in two distinct ways. First, algorithms can be used to confirm manual trading decisions. Second, predictive tools can be used to systematically select strategies as the basis for AI-powered investment products, such as ETFs and mutual funds.

Humans still necessary

Currently, 80 percent of trading in the U.S. is done by machine, and the panel expects increased implementation of AI in the future. The panelists made a distinction, however, between the use of automation and machine learning. In today’s cryptocurrency market, programmers are still responsible for creating and testing the algorithms. It may take several years to collect enough data for humans to give up their active role in the trading process.

This article is part of our iFX EXPO Asia 2019 recap. You can view below previous series sessions, or visit the official iFX EXPO YouTube channel to watch all of the iFX EXPO Asia 2019 sessions and get additional insights from industry insiders.

Top Experts Debate the Future of Institutional Crypto Trading

ESMA Moves Push Industry Leaders to Focus on Asia

Last month’s iFX EXPO Asia 2019 panel on prediction tools explored the intersection of Algo Trading and machine learning. We gathered a team of data science evangelists to describe how investors can use AI-based trading tools to exploit nonlinear market patterns.

https://youtu.be/HP2ii6t-b5Q

Algo trading goes mainstream

In the last two years, the attitude towards AI trading models has shifted drastically, and most institutional investment firms are now either using the services of algo trading providers or developing their own predictive algorithms. In part, this has come about because traditional funds management strategies have failed to perform according to market expectations.

According to Ksenia Semenova, Chief Business Strategy Officer of Cindicator, the major driver towards the adoption of algorithmic trading was the introduction of tokenized assets. She noted that “Cryptocurrencies started with algo trading, and this is a trend where all trading will be switched to algo trading…”

Coders and data scientists dominate

The experts also took some time to discuss trends which have gone under the radar so far. More traders are using their coding skills to develop strategies specifically for use by other retail investors, while institutional clients are turning to data scientists to create algorithms.

AI masters imperfect information

Panelists also discussed the potential of machine learning. Tiantian Kullander, Co-Founder of Amber AI, pointed to the success of Google’s Deep Mind, which can now beat humans in real-world games involving imperfect information. Kullander stated that “If you think about the markets, these are also games of imperfect information, because nobody ever wakes up and buys a stock at random.”

Yaron Golgher, CEO of I Know First, added that AI can improve trading in two distinct ways. First, algorithms can be used to confirm manual trading decisions. Second, predictive tools can be used to systematically select strategies as the basis for AI-powered investment products, such as ETFs and mutual funds.

Humans still necessary

Currently, 80 percent of trading in the U.S. is done by machine, and the panel expects increased implementation of AI in the future. The panelists made a distinction, however, between the use of automation and machine learning. In today’s cryptocurrency market, programmers are still responsible for creating and testing the algorithms. It may take several years to collect enough data for humans to give up their active role in the trading process.

This article is part of our iFX EXPO Asia 2019 recap. You can view below previous series sessions, or visit the official iFX EXPO YouTube channel to watch all of the iFX EXPO Asia 2019 sessions and get additional insights from industry insiders.

Top Experts Debate the Future of Institutional Crypto Trading

ESMA Moves Push Industry Leaders to Focus on Asia

!"#$%&'()*+,-./0123456789:;<=>?@ABCDEFGHIJKLMNOPQRSTUVWXYZ[\]^_`abcdefghijklmnopqrstuvwxyz{|} !"#$%&'()*+,-./0123456789:;<=>?@ABCDEFGHIJKLMNOPQRSTUVWXYZ[\]^_`abcdefghijklmnopqrstuvwxyz{|}