The broad realm of asset management has been slow to fully adopt the big data wave and there is still a lack of data-driven decisions being made by active managers.
However, the exponential increase of intangible assets over the last twenty years is only increasing the need for firms to adopt an organisation-wide big data strategy to fully understand assets and understand where improvements can be made to current business practice.
It is safe to say there is a lack of appropriate knowledge about data science, data mining, and the potential artificial intelligence has to revolutionise asset management. This is coupled with a lack of data scientists compared to the number of jobs available – indeed, data scientists becoming the premier league footballers of the IT world!
Another constraint is the lack of any substantial overlap between a data scientist’s qualifications/experience and an active manager’s qualifications/experience. Why would an investor with twenty years of experience start trusting the opinions of a data scientist fresh out of university? With the current dearth of industry-proven use-cases currently available, it’s pretty easy to dismiss data science as hype.
This isn’t likely to change in the short-term and AM firms need to think smart about big data and AI instead of fundamentally changing processes without a full understanding of the consequences.
This means educating asset managers in the act of data science and letting them witness the insights that can be generated first-hand.
A current blocker to consider is the nature of current software offerings and their suitability for direct use by asset managers. Dedicated financial management tools focused on big data are in their infancy (compared to generic industry agnostic methods) and there is little proof in the field that the application of dedicated ‘FinTech AI’ tools have managed to produce benefits anywhere near the scale that they are expected to achieve.
On the flip-side, industry agnostic methods require considerable experience in statistics to decipher the output from machine learning algorithms, and nobody can expect business users to jump head-first into SPSS or R to try and take advantage of their multiple data streams.
Big data or AI tools focused on asset management need to be developed in partnership with an active AM firm. Otherwise, the use cases will be built on guesswork and there will be no trust in suggestions for business change made by AI.
For AM to truly take advantage of the big data buzz there needs to be a change in attitude to how it is adopted. We strongly believe smaller firms should begin experimenting with AI/Big data by building use cases and using an easy-to-use AI tool to test predictive models against real-life BAU. By working in partnership with a smaller software development firm rather than adopting an expensive offering from a major player, AMs can begin to understand the fundamentals of how AI works and what data science can do to let business users unlock insights that transform business processes and get an edge in the world’s most competitive industry.
To get a jump start on AI organisations need to focus on the following areas:
• Joining separate data sources into a single version of the truth
• Applying deep learning algorithms to historical data to ‘train’ artificial intelligence tools
• Democratising access to insights to properly assess the impacts of data-driven decision making organisation-wide
• Partnering with software development firms to build use cases based on real life business problems
• Allowing business users to get experience in data science without having to adopt a risky data scientist-based strategy