I’ve had the good fortune of some extra time over the past few months, which has allowed me to explore the ever-expanding world of large language models (LLMs). Like many, I’m captivated by their potential – so much so that I currently subscribe to two LLM powered tools and I’m confident I am getting at least a tenfold return in enhanced productivity.

Yes, there’s plenty of hype – and potential bubbles – but it’s thrilling to consider how these models might transform entire industries. As an analyst, and having worked at an established South African asset management company, I can’t help reflecting on how LLMs will affect buy-side analysts and portfolio managers.

Machine-Enabled Insight

At its core, an analyst’s job is about retrieving information, summarising that information, and most importantly, applying reason. LLMs excel at retrieving and summarising vast amounts of data very quickly, slashing the time spent sifting through financial statements, broker reports and regulatory filings.

The application of LLM into AI tools like ChatGPT, and frameworks that streamline building with LLMs, like LangChain and n8n , make it easier than ever to simultaneously incorporate internal data and wider, real-time web data. Services like AlphaSense and OpenBB are already advanced in this regard.

So, that leaves the reasoning part. The good news here is that AI is also evolving at breakneck speed, thanks mostly to massive investment by foundation model providers like OpenAI , Meta and Anthropic. Each new-generation model, like ChatGPT o1 for example, has made significant strides in incorporating sophisticated reasoning capabilities.

On top of this, multi-agent frameworks like Agency-Swarm and CrewAI demonstrate how multiple AI ‘agents’ can collaborate to solve complex problems – similar to how an analyst will weigh various angles of an investment thesis. While the AI is probably not yet advanced enough to replace human analysts, the current trajectory and the sheer quantum of financial investment in these models suggest a highly disruptive future.

Boosting the Analyst’s Edge

For many, this kind of disruption is terrifying. The knee-jerk response is to worry that AI will completely replace human analysts, but I see it differently. Being able to offload repetitive or data-heavy tasks allows analysts and portfolio managers to focus on deeper strategic thinking and creative interpretation. It also broadens the ‘field of vision’, enabling analysts to spot more opportunities and connect dots that even the best LLMs might miss.

We work in a competitive environment where every edge matters. Any additional mental bandwidth can generate real alpha. For firms that adopt AI early, there’s enormous potential for improved profitability and better outcomes for clients.

Unknown and Under-Gunned

South African asset managers face a new reality. Regulations have come into play much sooner than many expected, allowing clients to invest more offshore. Unsurprisingly, this has triggered a landslide – clients are shifting assets in existing portfolios from local to global investments.

Local investment managers suddenly must compete with the world’s most sophisticated investment firms. Perhaps more dauntingly, they also have to convince powerful asset-consultants, the agents in between the manager and the client, not to dispense the usual career-preserving advice of re-allocating assets away from the local manager to well-known global managers like Schroders or BlackRock.

South Africa boasts very few genuinely global investment managers and, with a few exceptions, evidence suggests that South Africa-only investment managers are under-gunned in a world where clients are rapidly broadening their horizons and seeking global opportunities. But instead of being a negative, this challenge actually presents a tremendous opportunity.

South African Asset Managers Could Benefit Even More

Why? New technology often disrupts markets, and since LLM and AI are still in their infancy worldwide, South Africa’s investment managers should grab this opportunity to leverage the technology to leapfrog competitors.

There’s no reason more South African investment managers shouldn’t be globally competitive. South Africa’s local investment industry enjoys global respect, with many managers having built-up enviable investment track records. There’s a long legacy of South African companies showcasing their abilities and skills to become world leaders in many sectors.

By harnessing new AI tools, local managers can significantly boost their global relevance. In other words, by adopting LLM-powered research, modelling and monitoring – and adopting it now – managers can leverage proven investment philosophies and processes to punch above their weight on the international stage.

Doing this depends on local leadership teams investing proactively in the right technology, training and partnerships. Waiting on the sidelines won’t work. It’s a recipe for disruption. Those who lag will be left behind.

The Way Forward

We’re still in the early days of seeing how LLMs will reshape the industry, but it’s an exciting time for analysts and portfolio managers who are prepared to adopt these new tools. After all, which analyst or investor doesn’t want to combine deep domain expertise and a solid investment philosophy, with technology that lets them see further and move faster?

While no one can predict the full impact of AI on investment management, most agree that early adopters will have far greater influence over how the technology ultimately shapes the industry. AI is more than just efficiency gains. It also gives us the chance to enhance the core human qualities that clients truly value: judgment, creativity and empathy.

I’m optimistic about the future for South African managers. We have great investment teams, a strong research culture and profitable investment firms that are (hopefully) willing to adapt. Paired with the power of LLMs, these strengths could be a springboard to global relevance for South African asset managers.