By Vaibhav Sagar, Senior Technology Consultant, Open System Tech
A tech consultant reviews CLSA’s Eugene Kanevsky’s recent article about a new generation of trading algorithms.
In a well-argued article, Eugene Kanevsky shows how complex algorithms and machine learning-based artificial intelligence (AI) processes are helping stock trading firms make rapid decisions based on hard objective data, eliminating human emotions and biases.
Kanevsky, who is global head of electronic trading at CLSA, describes how his firm’s “ADAPTIVE” technology builds rules and identifies stock categories, then adapts in real-time as the significance of data fluctuates and trends shift. The system categorises stocks and predicts future price movements based on historic trends, short-time changes, market capitalisation and liquidity in order to automatically implement trading decisions.
If a stock is in one category today, it could be moved to another a few months later based on changing patterns. The process is constantly evolving and learning from itself.
ADAPTIVE creates a data science framework, and decision-making is conducted by a proprietary neural network that can read patterns and anticipate future price, volume, volatility moves based on pattern repetition.
Clearly, many brokerages are in a race to build the latest and best trading technology, providing tremendous career opportunities for quant and algorithm development experts. However, Kanevsky stresses that although AI machine-driven trading has an ever-increasing role due to its speed and continuous learning capabilities it still requires human intervention and monitoring.
Therefore, it should be considered as an aid to a sales trader rather than a replacement, an example of humans and machine working together to create greater efficiency and reliability.
We’d love to hear your feedback on this article. Please click here