Relations to Montreal AI Eco-system
Montreal AI landscape is large with many institutions specializing in research & application in all industrial sectors, except finance/asset management.
What makes finance sector so neglected?
First, these organizations have no specific/focussed mandate for the financial sector; second, the unique nature of financial data. In all other industries the data have high signal-to-noise ratio. For example, it is by now trivial to train the algorithm to identify a cat or a tree with high precision. A cat and a tree do not change their appearances from one day to another.
Financial data have very low signal-to-noise ratio, as for example to make a buy or sell investment decision is not trivial. Moreover, as investors’ behavior changes with trading strategies becoming more and more sophisticated — the nature of the data dynamics changes as well. Therefore, the “cat” in financial data that one could identify yesterday will most likely have a different appearance tomorrow. Therefore, the algorithms used for example for driverless cars cannot be directly applied to the financial data.
Thus, the explanation is simple – the lack of research on applications of AI in financial sector is due to the lack of talent. It takes a different skill set, the combined knowledge of research in finance and computer science. This is a new discipline, AI in Finance, which is not quite established yet within educational institutions.