Artificial Intelligence (AI) promises to usher in a new age of productivity and performance in the investment industry.

With the recently acquired ability to understand and produce writing, speech and images, and the long-demonstrated power in predictively processing high-dimensional data in complex domains, AI-empowered systems have the potential to revolutionise every corner of the industry.

Read: Nvidia becomes Tesla’s successor as market flips from EV to AI

The firms able to position themselves to catch the most significant productivity wave since the internet revolution of the early 2000s will realise an order of magnitude step change in the efficiency and effectiveness of their processes.

The technology isn’t a magic bullet, but its potential impact on the competitive landscape is immense.

The most successful firms won’t simply deploy every bit of AI they can find. Rather, they’ll need to carefully evaluate where and how to strategically apply these tools and technologies based on a considered understanding of the business and investment process implications.

Rezco has been investing heavily in our next-generation, AI-centred investment management tooling over the last six years through our partnership with Alis Exchange (recently featured by Google Cloud as a lighthouse case study for innovation in the industry).

Given the latest wave of interest in AI, we thought it would be a good time to share some of the thinking that underlies our AI strategy with clients and to invite commentary and debate on where others agree or differ in opinion.

Accelerators vs. Augmentors: Understanding different AI impacts

We can broadly divide AI-driven opportunities into two categories:

Process accelerators: These AI tools optimise processes by dramatically increasing the speed of the investment process. They don’t necessarily do things better than humans but work at speeds far beyond human capabilities. A powerful language model can synthesise key investor opinions, earnings call transcripts, and research reports within seconds, giving a human analyst a substantial head start.

For example, it takes about 20-30 hours of detailed work (reading sell side, earnings calls, reports, prior coverage, market and news commentary) to synthesise a good “key drivers” and “bull case/bear case consensus summary”. A decent LLM (large language model) can do that in 20-30 seconds.

Maybe it gets to a slightly worse end product, but it is 1/3000th of the time and effort. Investment managers are in the information processing business and the speed with which information needs to be processed to remain competitive in public markets is experiencing a step change.

Augmentors: Augmentor AI tools meaningfully improve a person’s capabilities. Perhaps this could be AI that identifies unique anomalies in trading data hinting at potential fraud or software that recommends portfolio rebalancing based on highly complex risk considerations.

Augmentors elevate human skills.

We’re already using AI to materially augment the sophistication of our global equity screening process. The old style of taking a week to build a model and produce a report, post earning updates is struggling in a world where a lot of the alpha can happen in the pre-market after an earnings announcement.

In global equity, the traditional long-only style is struggling to compete against passives on the one hand and massively resourced pod shops on the other. Good AI models are able to find the patterns in the data and help PMs react quickly. But this is just the beginning; the complexity will grow exponentially, so the gaps between the early and late adopters will increase.

Rethinking investment analysts

Analysts will need to adapt. AI’s information processing abilities could render traditional analysts’ jobs obsolete. Instead, analysts of the future will need to showcase stronger critical thinking, problem-solving, and synthesis skills.

The goal becomes less about information production and more about leveraging insights – a shift more closely aligned with a portfolio manager’s traditional role. The fear of AI-caused job losses is natural, but we need to focus on the competitive edge.

Firms shouldn’t simply fear AI but rather fear falling behind those who are harnessing it powerfully.

Embracing (but not blindly accepting) AI

A valid and common criticism of AI tools is, “it’s a black box” or to hear managers espouse the “explainability” value of simple Machine Learning 1.0 models.

Having said that, there are probably only ten people in the world who really understand the inner workings of frontier foundational models such as GPT-4, and this doesn’t stop it from being really useful to millions of users. Tools like ChatGPT are impressive, but it’s vital to use these powerful black boxes intelligently.

It’s tempting to fully trust output, but the key is learning by doing. The more we interact with AI models, the better we understand their capabilities and limitations. Verification of AI-produced work remains essential.

Invest in your foundation: Data and digitisation

Catching the AI wave requires having something to ride. Like the internet revolution before it, firms need to have the data and digitisation foundation in place to capitalise on the AI opportunity or risk being left helplessly paddling in a vain attempt to play catch up.

In the funds industry, laying the data and digitisation foundation for AI is a particularly acute challenge. Data moves through the fund’s ecosystem in a flood of disparate data formats, definitions, and templates. These PDFs, Excel files, emails and CSVs settle within the inscrutable crevices of SharePoints, server drives and email mailboxes like the accumulating detritus of a hoarder’s basement.

Might an honest investment professional’s job description read, “Help wanted. Experience tracking down and
manually transposing data.”

To truly make the most of AI, we can’t ignore the prerequisite work. Firms need clean, structured, and well-managed data as the building blocks for AI to be helpful.

Today, too much valuable data exists in messy formats like PDFs and emails, hidden and scattered within the digital workplace. Data management and digitisation investments are just as crucial as deploying the fanciest AI toolsets. Commentators on the “Pod Shop” world often talk about the high systems and technology barriers to entry to be able to compete.

Read/listen:
Bear market truths and how tech changes the face of investing
From learning investing ‘on the job’ to AI fund management
Is the AI story priced for perfection?

Our thesis is that having a coherent and well-executed digitisation and data strategy will similarly become table stakes for managers moving into an AI world.

The question isn’t “Have you spent millions on alt-data?” but whether you can effectively create a digitised manifestation of your investment process that enables the seamless collaboration of people with AI systems.

Conclusion

AI’s ability to generate text, speech, and images will change the way investment funds work. Firms that thoughtfully prepare will unlock powerful productivity gains.

Just as the internet revolutionised industries, AI is poised to do the same for those who plan strategically. This isn’t just about cutting costs. It’s about increasing capabilities, and the most proactive firms will find an unmatched competitive advantage in the years to come.

source:How AI will transform the investment industry – Moneyweb