Investment Management industry is grappling with three key challenges; dramatic change in client expectations since the appearance of smartphones, litany of regulations since the 2008 financial meltdown, and disruptions from new technologies like AI that can upend long established business models. Moreover, the data deluge is making it very hard to distinguish noise from information and fund managers are finding it very challenging to find the elusive alpha. Investment management firms can no longer run on reputation inertia but need to step up to the challenges posed by the new era.
Artificial Intelligence (AI), as Dr. Andrew Ng says, is the new electricity and as such it can be used to shine light on the dark pools of enterprise data and transform business models by identifying new revenue streams.
Of course, there are lot of other fascinating things in AI like language translation, logic programming, transfer learning, reinforcement learning, etc but the block and tackle AI in the enterprise context is basically a collection of technologies and mathematical techniques that allows us to take the data within and outside the organisation and using that data to predict what’s going to happen or even cluster it to find patterns that we didn’t know existed.
However, AI in enterprise context is distinctly different from AI in the consumer space. Actually, there is an asymmetry in AI adoption in these two domains. Firstly, enterprises do not have as much labelled data as in the consumer space so deep learning algorithms that need huge data are not as useful. Secondly, enterprises have much more stringent security and privacy requirements compared to the consumer space. Mathematical techniques like differential privacy etc. are still largely in the research domain. Thirdly, Deep Learning, on which the last decade of amazing progress of AI rides, is not very explainable or humanly interpretable. Enterprise problems are very different, for example, how to read graphs in a pdf file or contextualise a text to the business domain. These are still hard problems for AI and require human intervention. And finally, there are adoption issues in the enterprise. We are talking about business processes that have been built over decades and even centuries so when we talk about AI in enterprise we must deal with these processes and we cannot just ignore knowledge that resides in the business processes, which includes IT systems and human workforce.
Given the aforementioned challenges in deploying AI at an enterprise level, we are yet to see its transformational impact on the wealth industry particularly for the HNW and UHNW clientele. However, AI would surely be the differentiator in the next 2 to 3 years. It will have an outsized impact on Investment Management businesses in areas like Front Office Advisory, Risk Management (including Fraud Detection) and Back Office paper-based operations. Firms must consider with urgency how AI could be deployed to open new revenue streams, creating new business opportunities, creating new product lines (ETF 2.0) with smarter Beta, identify the elusive alpha or evaluate PE investments. IT services companies should seek to combine their understanding and knowledge of Enterprise IT in Investment Management, experience in the IT service sector, and strong partnership ecosystem to add tremendous business value to their clients.
