The Impact of Generative AI on the Hedge Fund Industry
By Bilal Rashid (LSE BSc Economics; Intern at Brookfield, Cinven and Goldman Sachs)
Since the launch of OpenAI’s ChatGPT LLM in November 2022, awareness of generative AI tools and the benefits of their use has increased dramatically. The innovations we are currently observing represent a transformational force across industries, with the potential to drive substantial changes, whilst also bringing the potential of negative repercussions if the associated risks are left unmanaged.
With regards to the hedge fund industry, a survey conducted by AIMA found that 86% of hedge fund managers surveyed currently allow Gen AI tools to be used by employees to support their work. In particular, it is expected that AI can help save time on administrative tasks and drive cost savings, with its potential influence reaching across the whole span of corporate functions. The extent to which AI is currently integrated within these hedge funds varies, however.
On the one hand, there are many funds that allow the use of Gen AI, though are yet to formalise any form of AI policy document. Whilst it is encouraging that nearly half of the funds without policy documents are in the process of drafting one, there are other funds that have been more frontfooted in their advancements. In particular, 10% of firms surveyed make use of an internal Gen AI application, providing the benefits of greater customisation and data security features relative to generic open-access versions. These internal models do, however, require greater levels of investment, perhaps explaining why half of firms with company-specific models are large multimanagers with over $20bn in AUM. An example of this is Man Group, which developed ‘ManGPT’, an internal model that assists with programming queries, text rewording and compiling meeting notes (amongst other functions), all within a secure and controlled environment.
The rapid progression of Gen AI tools shows no signs of slowing and brings great opportunities, as demonstrated by the aforementioned example of ManGPT. Moreover, there is a growing belief that AI could support portfolio optimisation through extensive data analysis, alongside delivering greater efficiencies in alpha generation (through research insights and better use of alternative data) and trade execution (through analysis that better evaluates trading signals and optimises orders). Despite these exciting benefits, there will likely be challenges around data security, errors made by models and costs of training. Nonetheless, as in vestment continues and these tools mature, we will likely see a bifurcation in performance between those with and without AI models, with the latter likely being smaller hedge funds that (relatively speaking) lack the required financial resources to innovate and maintain a competitive edge. This bifurcation in performance is especially likely given the potential for AI to drive improvements across a broad range of firm functions, whether that be sales, marketing, legal, compliance or other areas.
Source:
https://shorturl.at/xMiNE
https://shorturl.at/oISSz
https://shorturl.at/5M15x
