Man and Machine: Human Ingenuity and AI in Quantitative Finance

Artificial Intelligence (AI) is the hot topic of the moment. The simulation or attempted replication of human intelligence using machines is either already a reality, or about to become one in many industries, and speculation is rife that AI will take over a growing number of jobs.

Does that include jobs that rely on human judgement and experience, such as medicine, law, and even quantitative finance (QF)? Will robots one day become traders and investors?

"The consensus for many of us in the field is a clear 'No'," says Singapore Management University (SMU) Associate Professor Tee Chyng Wen from the Lee Kong Chian School of Business (LKCSB).

The Current Status Quo

Why? Well, while AI may have advanced to the point where it can successfully beat human players at chess and perform facial recognition on humans and animals, it has not quite reached the peak of its abilities. Therefore, Professor Tee believes AI is unlikely to render the role of humans redundant in QF just yet.

Current AI is trained to recognise stationary two-dimensional images, and to operate within systems that have clear fixed rules laid out for optimal machine learning. More has to be done in terms of pattern recognition and reasoning before AI can even be considered for independent operation in the financial markets, where the rules are highly dynamic, unpredictable and subjected to frequent changes.

So, as of now, AI is merely performing the role of a trustworthy assistant when it comes to making trading and investment decisions. Apart from that, machine learning is most commonly applied in Transaction Cost Analysis (TCA) and optimal execution strategies. TORA, a global provider of advanced investment management technologies capable of seeing through trading cycles from end to end, is one such conglomerate that began this practice back in 2017.

Cross-disciplinary expertise is key

In order to stay on top of the game as technology in these fields evolve at lightning speeds, Professor Tee advises individuals seeking careers in finance to become "cross-disciplinary experts" who are versatile and well-versed in drawing expertise from varied fields, in order to tackle the industry's new frontiers.

The SMU Master of Science in Quantitative Finance (SMU MQF) programme aims to groom individuals of this calibre. Through a full-time 12-month or part-time 24-month curriculum, postgraduate students learn from rigorous courses and case studies designed to equip them with both theory and practice.

"Apart from core financial knowledge, they also learn mathematical modelling and computer programming," says Professor Tee. For instance, the module "Machine Learning and Financial Applications" explores how neural networks can be used to support the quantitative finance industry. Designed with multi-disciplinary purposes in mind, the course prepares its students to be comfortable with integrating data science and finance together to navigate the world of QF.

This is a handy way of closing the gaps for data scientists who lack finance intuition, and traditionally trained finance students who may lack the technical expertise of data science, explains Professor Tee.

Cooperative competition to shape the future

A lot more is in the works when it comes to manipulating AI to further optimise the field of QF, starting with an amiable culture of "co-opetition" — a friendly, cooperative competition among various stakeholders.

"Practitioners and researchers are cooperating by sharing their latest findings in academic journals. At the same time, there is a competition to see who will be the first to figure out [how to better use AI in trading or forming portfolios]," says Professor Tee. Based on how quickly machine learning has entered our lives and made its presence felt, these developments are bound to impact the world of finance in the near future.