The Darren Aronofsky film Pi features a mathematician with the uncanny ability to perform complex arithmetic in his head. Notwithstanding his talent, he uses a computer to make stock predictions. In one scene, after recognizing a predictive pattern in a 216-digit output, the character becomes so overwhelmed that he passes out.

This scenario is science fiction, but computers have long had greater processing power than humans. When used to build artificial intelligence (AI), as its datasets grow, so does a computer’s advantage.

Rules-Based Systems

Rules-based investing can be traced back to the industry’s formative literature. Books written by Benjamin Graham and other legendary investors explain investment rules that practitioners have tested and implemented over the years. This early reasoning provides the foundation for many of the rules-based processes used to make active investment decisions today.

Explicit rules remove emotion from investment decisions and limit human error in execution. Systematic approaches often use the same information that discretionary investors do — market data inputs with buy or sell decision outputs. The rules driving these decisions are designed by humans, but computers maintain an important advantage: they can apply the directives to more markets and instruments than any individual could possibly analyze.

Rules-based systems represent AI in its basic form as a series of “if-then” statements used to make decisions in the place of a human. These systems let investors take the logic in their heads and codify it to process information more efficiently and consistently.

Machines Acquiring Knowledge

A static rules-based system is intelligent, but limited because it cannot learn on its own. Machine learning occurs when a system processes data to assess its predictive power and then, using the information it has learned, improves its process for the next iteration.


One way systematic investment strategies acquire knowledge is by borrowing logic from genetics. In evolutionary computation, investment rules evolve through selection using parent and offspring decision trees. Successful rules survive, much like genes in reproduction, and when computers can combine each strategy’s best characteristics, it does not take a lifetime.

AI is also applied to investing via an extension of machine learning known as deep learning. Deep learning seeks to recognize patterns in financial markets through a process similar to facial and speech recognition. It uses neural networks to process information, effectively connecting virtual neurons by mimicking the design of the human brain.



While still in its early stages, this form of pattern recognition is positioned to disrupt traditional forms of technical analysis. Advances in processing power have made it possible to identify patterns through filters with more depth than a simple chart of prices.

Deep Learning and Big Data

The volume of financial market data is enormous and growing daily. Deciphering signal from noise in this environment represents an opportunity of equal magnitude. In this context, applying deep learning to the universe of big data offers the greatest investment potential for AI.

While the performance of most algorithms decreases with increasing amounts of data, deep learning is less affected by this influence. As a result, the intersection of deep learning and big data is where most research is currently focused. Ultimately a combination of these disciplines will be available in the future.

Investing has traditionally been a cyclical business, and the technology being developed today is a departure from this. Being able to analyze more information with greater intelligence than what is humanly possible is a structural change in an industry well equipped for innovation. Investors naturally focus on where they can have an edge, and combining their strengths with those of AI will increase the value they create with each investment.

If you liked this post, don’t forget to subscribe to the Enterprising Investor.

All posts are the opinion of the author. As such, they should not be construed as investment advice, nor do the opinions expressed necessarily reflect the views of CFA Institute or the author’s employer.

Image credit: ©Getty Images/JIRAROJ PRADITCHAROENKUL

Andrew Dassori

Andrew Dassori is the Chief Investment Officer of Wavelength Capital Management, an independent alternative investment firm based in New York. Prior to founding Wavelength, Dassori was a portfolio manager at Credit Suisse focused on global macro investment strategies. While there he was also a member of the firm’s Global Citizens Program through which he worked at Equity Bank in Nairobi where he was responsible for a team building default risk models for microfinance loans, and represented the bank in meetings with the IMF, World Bank and other economic policy institutions. Dassori received a BSc from the London School of Economics.

Source link