The examples above are interesting and add genuine value to us, but they aren’t necessarily transformative to our investment process. Our decision-making process is still broadly the same, but we’ve now created more time to research those decisions. What would be more exciting for us, would be reaching a point where we can use AI to help us make better decisions for the portfolios to improve investment outcomes. The work we’ve done so far suggests that this is still a work in progress across the industry.
One example of a mainstream use so far relates to “unstructured data” as a tool to help analyse sentiment. Unstructured data refers to information that is not easily usable in an easily packaged format. An example might be the relative usage of positive words or phrases compared to negative words in CEO corporate earnings presentations, which across the broad market can give a sense of executives’ sentiment or confidence for the near-term future. The AI aspect of this is how the computer “learns” what words or phrases are positive or negative through a process of trial and error.
From our perspective, AI is but one example (albeit a particularly interesting one) of a tool that can help us improve our success rate in delivering our ultimate goal – making good decisions to improve investor outcomes. Part of our job is to make sure we stay at the forefront of both academic and practitioner research to give ourselves every opportunity to succeed in this goal.
