We should not be surprised that, on average, artificial intelligence bests stock analysts in forecasting earnings. Or that rule-based strategies produce superior financial advice on average than a private banker. Even before the recent advances in generative AI, the merits of investing in a systematic way existed. While such techniques will not find that elusive percentile of stocks or the market inflection that make the excess returns, they do have proven value.
AI developments show that we can go beyond rule-based recommendations, though. Macroeconomics, accounting and statistics are the three pillars of investment. Large language models get top scores in advanced exams of these subjects. We also know that LLM’s can summarise vastly more context and crowd wisdom than a human can, which should be very helpful for macro strategy. Hence, if AI can help in financial decisions, why is it so difficult for an analyst or portfolio manager to get on the bus of change?
We can find some clues in the work of data scientist César Hidalgo on how humans judge machines. When we use a program, we zero in on the performance of the tool. Thus, any prediction error by that program will make our financial professional lose confidence in it. And in most of the cases, it does not matter if the algorithm is on average better than the human. Our financial adviser will let her intuition and experience take over.