While ChatGPT and its underlying generative AI technology has only been in the public consciousness for a few months, the world has since entered a rapid state of flux trying to understand how best to harness this revolutionary technology while dealing with the number of significant security and ethical question that it raises.

It has the power to change the face of every industry on the planet. But we believe that in the short term, the role of generative AI will be limited to supporting existing research and data gathering and not the primary driver of decisions. Our current knowledge of generative AI does not allow us to discern false positive with actual data, resulting in additional risk for portfolio managers. But that does not deter us from taking a longer term view of how generative AI can shape the investment management industry. Join us in our journey to reimagine how portfolio managers could invest in the year 2030.

A Flying Start

The bleary-eyed early morning start could be the first place the impact could be felt. A portfolio manager may be greeted with a real time dashboard where generative AI continues to learn through interaction with unstructured datasets. Alongside market data, generative AI could combine esoteric datasets such as sentiment analysis or keywords searches that generative AI has deemed to be most relevant. It could unearth hidden trends or ‘black swan’ events that were previously unseen, providing a unique view of the investment horizon faster than ever before.

The portfolio manager may be able to make real-time adjustments based on these insights, removing the need for morning investment meetings. Instead, they could be equipped and empowered to make the best decisions every time.

Uniquely Optimised

Though there are already ways to optimise a portfolio, the methods tend to be over-reliant on historical data and focus too heavily on comparing returns maximisation with the level of investment risk. Generative AI could refine this process. Using the learned knowledge of the manager’s style and investment philosophy, generative AI could create unique optimisation strategies, helping to create bespoke stock allocation suggestions and overlay the client’s investment and ethical policies to assist with further review.

AI has the potential to leave a footprint in the investment analytics space too. Traditional ex-ante risk models or performance attribution may be consigned to the analytical scrapheap. Instead, the models could follow continuous and iterative paths of improvement as generative AI seeks to tweak and optimise models based on investment style and market events, making the models more relevant to the portfolio manager and the client.

The truly transformative part of this process may be for generative AI to code, test, validate and deploy these analytics on dashboards in a fraction of the time it takes now. Together with the enhanced and novel forms of analytics, portfolio managers may be able to seek deeper insights than ever before.

‘Intelligent’ market making

Once the portfolio manager has decided on the trades to make, generative AI could influence the trading process. Market makers could have AI integrated into their systems to automatically determine the market conditions, helping to identify liquidity across trading venues globally to fulfil the trades. This could be across the traditional and digital markets, resulting in efficiencies exceeding what we have today.

Market makers could accurately forecast liquidity demands and hold more appropriate inventory levels by understanding the underlying liquidity patterns. This could help market makers to optimise the prices offered to make them more competitive, which in turn could reduce spreads and costs.

From automatically detecting the best time to execute trades based on market movements, AI could read market conditions and liquidity conditions, triangulating them and transforming market making into a highly intelligent function.

Completing the cycle

The uses of generative AI could be extended throughout all functions to support the investment decision making process, allowing investors to shift from being intuition-based to data-driven.

Portfolio managers could start using generative AI to support them on how to vote during shareholder meetings. Rather than relying on the contemporary news cycle to drive judgement, generative AI could determine which decision will boost the long-term share price. Predictive analytics could be incorporated to see if dividend pay-outs should be re-invested with the same company or re-invested in another way. Elsewhere, a portfolio manager may start to receive query responses from an AI ‘service rep’ at their third-party vendors who could provide answers more efficiently.

The power of generative AI would not be limited to the direct investment process, but swathes of operational efficiencies could be made too.

Final remarks – Is your organization prepared to tackle the potential risk and challenges before reaping the benefits from generative AI?

While 2030 seems to be a long time away, the industry would have to start thinking about the potential risk and challenges in 2023. Despite the numerous potential use cases identified, we concede that there are a lot more risk and technical considerations before any meaningful implementation could occur.

The actual integration of generative AI has also yet to fully materialise today. An organisation would have to consider where they are comfortable introducing AI and the areas that are ‘out of bounds.’ Having a clear vision detailing what needs to be achieved and what success looks like could help drive the development while of course being cognizant of the potential bias and arising ethical issues based in the data sets.

Northern Trust believes that generative AI could permeate across all industries. We are now actively experimenting with the technology in a sandbox environment to test several of the use cases illustrated above.

With the right safeguards in place and transparency on AI produced outcomes, the way that portfolio managers work could change substantially. The benefits that can be achieved would come from cost savings and efficiencies gained to delivering higher alpha.