The trouble with a half-decent investing story is it can survive the truth, and that lingering impression can cost real money.
In Learning from False Stories, a working paper by behavioural economists at the University of Cologne and the University of Bonn, participants were shown an AI-generated video of former Federal Reserve chief Jerome Powell warning of a major banking crisis.
They were told the video was a deepfake, that Powell never said any of it, and that it was irrelevant to their investment decision. And yet they subsequently invested less money in JPMorgan. Despite knowing the video was fake, they acted as if some part of the story might be true.
A fluke?
READ MORE
The researchers devised another experiment to find out. Imagine you own Company A. You are told an analyst has raised its price target, so you become more positive about A. Then comes the correction: sorry, the analyst was actually talking about Company B. You should return to your original view of A. When the information consists of dry numbers, that is more or less what happens.
However, repeat the experiment with a story about the company’s apparently exceptional management: its executives are visionary and disciplined, think long-term, and have created a strong corporate culture.
Then comes the same correction: sorry, that was about Company B. This time, the correction does not fully work. People continue viewing A more favourably. Basically, a number can be corrected, but a story makes us imagine something. Once we have pictured a competent management team, a banking crisis, or perhaps a trillion-dollar future for a company colonising Mars, we cannot quite unimagine it.
The factual claim can be removed but the impression remains, leaving a “qualitative residue”. Markets are awash with stories, and AI-generated deepfakes intensify the problem. It seems that even when investors know something is false, part of us may continue to act as if it is true.
