Please Note: Blog posts are not selected, edited or screened by Seeking Alpha editors.

Me, Myself And AI: The Inevitable Machine Learning Revolution Of Investing

Much has been made of AlphaGo, a Google developed artificial intelligence (AI) algorithm, which beat the South Korean grandmaster Lee Sedol at the notoriously complex 3,000 year-old game Go, in March last year. And for good reason. It represents a step change in computing intelligence – one that was reaffirmed at the end of May when the Google algorithm defeated Chinese world number one Ke Jie by three games to zero.

The comparisons with Deep Blue’s famous chess victory over chess grandmaster Gary Kasparov nearly 20 years ago are stark, and the implications are profound. Deep Blue was impressive but essentially a very sophisticated programme which acted according to certain rules. AlphaGo, on the other hand, had to adopt and master deep learning techniques in order to win. You could even say that it had to ‘think’ for itself.

Finance is no stranger to the powers of computing. Analysis of complex financial data has been aided by technological advancement and larger data sets. And the industry has benefitted from greater brainpower. Cuts to US space programme funding and a loosening of regulations during the 1980s led to large swathes of mathematicians and physicists entering the world of finance. Their presence brought an increasing focus on quantitative models.

Active quant investors seek out the largest available data sets, either of the highest quality or the most relevance, and then analyse this larger opportunity set on a systematic basis. The output of this approach comprises signals or trading strategies that could be exploited before being uncovered by others without access to the requisite technology. More recently, quant managers have started using similar techniques to create ‘smart beta’ products.

Smart beta can be described as a ‘third approach’ to investing that combines the best features of both time-honoured active management (i.e. the potential to outperform a capitalisation weighted index) and passive management (i.e. simplicity, transparency, scalability and low costs). Blending these risk premia factors together into a multi-factor approach can help the investor to obtain the best possible risk-adjusted outcome. This is where AI also fits into quant investing.

AI has the potential to open up sophisticated active quant approaches to a wider set of potential investors. By accessing and processing huge amounts of data, and using machines with speech and vision capabilities, AI investment approaches will be able to identify previously unrecognised patterns in markets. What gets developed by active quant investors – new data sources, unstructured data, greater analytical prowess, more insightful pattern recognition and prediction, or dynamic factor timing – will have a trickle-down effect and will be incorporated into smart beta approaches.

However, much like the AI achievements in the board game arena, efforts on the quant side of finance have been restricted to narrow, controllable frames of reference (e.g. a specific factor). Dynamic systems, such as economics and finance, have a far greater number and complexity of inputs and variables than those found in board games. Environments that involve human behaviour are inherently unpredictable, meaning their typically static models have proven to be limited - that is, the models ‘work’ until suddenly they don’t. They can struggle to spot structural shifts or rare ‘black swan’ events.

A far broader understanding of the world needs to be incorporated into investment models. The leap from machine learning to deeper learning techniques appears to be the next frontier of quantitative investing. The potential is there for machines to learn by recognising failures and adapting accordingly.  

But while leadership and management, deep critical thinking and more creative tasks can all benefit from technological advances, they also require the human touch. Bank of England Chief Economist Andy Haldane puts this in terms of a dog trying to catch a Frisbee. The subtlety and complexity of a dog’s movement when successfully catching a flying Frisbee is currently well out of reach of any robot. The process cannot be easily reduced to an algorithm. Similarly, interpreting and interrogating certain data sets is beyond complex.

Machines or algorithms can begin to overcome such issues by using deep-learning techniques - but we return to the ‘breadth of understanding’ issue. While algorithms that learn dynamically could analyse new data sources that were previously overlooked or unavailable (such as satellite imagery offering clues to retail footfall), the question remains whether machines can be taught to intuitively understand the whole picture.

This is where the real challenge lies for AI and investing. Data is not information, and information is not insight. The transition between these distinct states is fraught with dark alleys and potentially expensive dead ends. Sifting through the noise and identifying the correct signals, whether by hand or by machine, remains the true quest. Humans and machines can help one another, but alone they are doomed to failure.