In many recent conversations with investors, the topic invariably shifts to AI. A retired fund manager I spoke with recently summed it up succinctly: “A car won’t drive itself; it needs direction or at least guidance. But if you’re headed to Manchester, it’s far easier by car than on foot.”
This encapsulates the crux of AI’s role in high-value intellectual activities like investment. It’s not just about whether AI can outperform humans, but rather how AI can enhance human capabilities—speeding up processes, improving accuracy, and reducing costs.
A lawyer at the same gathering remarked on how, when he started his career, a significant part of a trainee’s role was document discovery—sifting through extensive paperwork to extract information for interpretation.
Similarly, in the investment sector, there was a time when companies prided themselves on having “the information advantage.” This often involved a team of numerically adept but relatively inexperienced individuals combing through annual reports to populate databases with financial metrics, enabling investors to focus on stock selection.
AI holds tremendous promise in revolutionizing the resource-intensive field of investment analysis. Can machines truly outperform humans in this domain? The answer is nuanced.
The so-called “information advantage” was essentially about having some basic knowledge, albeit at a significant expense of time and money. AI now promises to accomplish this work in a fraction of the time and cost. This shift parallels historical transitions, where thousands who once worked with horses in 19th century cities found themselves needing new livelihoods as technology advanced.
AI functions akin to a labor-saving device—a versatile vacuum or dishwasher. It could also enhance investment decisions by mitigating the behavioral biases that often cloud human judgment. Unlike people, AI operates impartially, driven purely by data.
Recent studies, such as those from Chicago University’s business school, have explored whether AI can surpass analysts in converting financial data into predictions about corporate earnings. The findings suggest that machines perform marginally better in this regard. Moreover, portfolios constructed using AI-generated earnings predictions demonstrate statistically significant performance improvements—a trend that may imperil jobs in various professional sectors, including more senior roles, sooner than expected.
However, AI’s apparent ability to simplify complex financial management tasks can be misleading. It presents answers swiftly and seemingly effortlessly, akin to the advent of screening software three decades ago. Then, accessibility was hindered by cost, whereas AI democratizes access but also obscures processes within its black box.
Personally, I once trusted the tangible process behind my Online REFS stock screening software, where I saw colleagues inputting data. Conversely, querying ChatGPT leaves me uncertain of its information sources.
AI capitalizes on our preference for shortcuts. We crave definitive lists of stocks matching specific criteria, often without scrutinizing how these recommendations are derived. Yet, as with early computing, the quality of input directly influences the output.
Reflecting on my career, I’m optimistic about AI’s potential for investors, yet relieved it emerged toward its end rather than its start. Many tasks I’ve performed over 35 years hold little relevance today—a parallel to adjusting bridles and mucking out stables. Now, understanding internal combustion or making room for new skills is imperative.
Despite AI’s transformative impact, I remain confident in human capabilities. We excel as social beings, valuing personal connections and nuanced understanding—qualities I prioritize in those managing my finances. I seek advisors who inquire about my family, not just offer automated solutions.
The AI revolution mirrors the disruptive force of the internet boom preceding it. It promises prosperity for a few highly skilled individuals while potentially marginalizing many others. A winner-takes-all scenario risks fracturing social cohesion, underscoring the need for thoughtful adaptation.
Indeed, AI is here to stay, demanding adaptation rather than resistance.
Tom Stevenson is an investment director at Fidelity International. The views are his own
Source: telegraph.co.uk
