Roy Keenan from Yarra Capital Management shares insights with James Dunn from The Inside Network on the role of active management. The Inside Adviser
After the Hayne Royal Commission many advisers decided to steer away from large dealer groups in favour of becoming self-licensed. In the last six months, however, that trend has taken an abrupt turn.
New generative AI is a boon for fundamental fund managers: Kollo
As an ex (you may say, recovering) quantitative manager, I think there is a very significant shift in the AI narrative for asset management taking place around the world. ChatGPT, an advanced generative AI language model, heralds an early-stage but decisive pivot in our industry, handing the narrative back to those more inclined towards qualitative interpretation and language and away from those that have pushed for an analytic approach to investing.
While quantitative managers have hoovered up new innovations in data, processing power and technologies, this new wave of AI, focused on language and reasoning has the potential to even up the ledger.
Let me explain myself. The role of quant professionals like myself has been clear-cut till now: deploying sophisticated algorithms to predict markets, gauge sentiments and refine portfolios. Our approach, while reductionist in nature, distils asset management into its fundamental components, focusing on factor optimization to yield an effective risk-return matrix. This analytical and mathematical pathway has catalysed a sharp rise in passive strategies, alongside the advent of risk-optimized vehicles such as smart-beta.
Throughout this quantitative evolution, it’s been noticeable that fundamental active managers have mostly remained on the periphery, either adopting the tools sparingly or treating them as safeguards in their investment processes.
ChatGPT, however, presents an interesting inflection point in this journey. It’s a context-aware, interactive system that relies on language, and can respond with varying degrees of formality. It doesn’t necessitate a STEM degree or proficiency in coding; rather, it promotes a less analytical and more interpretive approach. There is no perfect prompt, there is no data cleaning, and in fact, its strength comes from its free shifting form, and ‘whatever you want it to be’ attitude. It’s almost entirely unrecognisable from anything else that has passed for ‘quantitative’ to this point.
For those active managers who have been mere bystanders to the quant revolution, as well as those who have been active participants, this is an intriguing proposition. ChatGPT is adept at narrating stock-specific stories, outlining economic trends, and bridging top-down macro-economic narratives with bottom-up perspectives. It can generate investment articles, fund memoranda, and even support in-house training on complex topics.
What stands out to me is that this entire operation doesn’t demand the presence of machine learning scientists, cloud infrastructure, or investment in costly data acquisition and cleaning. This presents an alternate vision for AI’s trajectory – one that is more human-centric and less reliant on technical mastery. It also empowers qualitative asset management teams to harness their intuitive and narrative-driven strengths, making their offerings more accessible to a broader clientele.
As a quantitative professional, I find this rebalancing of asset management focus due to a new breed of worth investigating. It signifies a balance between the quantitative and qualitative worlds, possibly heralding a more holistic future for our industry. The next generation of ‘AI native’ asset manager may not be the room full of servers of the likes of Renaissance Technologies, but perhaps a much more engaged, interactive and human.
