S8gqk2ry 1405009242

Brain scans could be used to predict financial bubbles

Experimenting with bubbles. Flickr/Indigo Skies Photography, CC BY-NC-ND

Brain scans could be used to predict financial bubbles

Experimenting with bubbles. Flickr/Indigo Skies Photography, CC BY-NC-ND

Some shares have new owners every second. Today much of the buying and selling is done by computers, but some still rely on human intuition – the gut feeling of the experienced trader.

“Nobody can predict the market, but traders are expected to,” Richard Taffler, professor of finance at the University of Warwick said. “This creates anxiety.” So emotions must play an important role in driving financial markets. Understanding what happens in the brains of traders as prices move up and down could possibly tell us something about the markets future developments.

In a study published in the Proceedings of the National Academy of Sciences, Alec Smith at the California Institute of Technology and his colleagues conducted group-behaviour experiments. For each of the 16 rounds, they had between 11 and 23 students, who played a game that simulated a market situation. For every round of the game, Smith monitored the brain activity of three of the participants with functional magnetic resonance imaging (fMRI), which highlights parts of the brain based upon increase or decrease in activity of that region.

The market game starts with an asset at an arbitrary value. As the market develops over time the asset price increases. The experiment is setup such that a market bubble always forms, and then it bursts causing the asset to return to the initial value in a very short time.

Smith and his colleagues found that activity in a brain region called the “nucleus accumbens” (NA) correlates the market price – so if the price goes up, NA gets more active. “This brain region was previously associated with emotions, fear and pleasure,” Smith said. “It makes sense we find this region active.”

In the next step the researchers compared the participants who, at the end of the market game, turned out to be high or low earners to relate the brain activity to a specific trading outcome.

The difference is that low earners buy around peak price, while high earners sell their shares around peak price. The researchers suspected the cause for this behaviour may lie in the “anterior insular cortex”, located behind the forehead. The insular cortex is active during bodily discomfort – when you feel pain, anxiety or disgust. The anterior region is also activated by financial risk.

Smith found that around the time when prices are about to hit the peak (that is, a bubble is starting to form), which is also the point where decisions of high and low earners start to diverge, insular activity increases in high earners, but shows no change in low earners.

The upshot is, if you could measure the brain activity underlying certain emotions of an active and successful trader in a critical market situation, you could predict how prices may change. “We think that irrational exuberance is likely to play part in price bubbles”, Smith said, explaining why they focused on the brain region involved in a lot of emotional processing.

Richard Taffler, professor of finance from the University of Warwick, isn’t sure that such predictions are possible. “It is not clear how we get from undergraduate students, in a confined laboratory environment, without a real potential of loss to a real world market situation,” he said.

Alec insists the behavioural game is not too much of an abstraction from reality and we can use it learn about the principles of price bubbles. Instead, Taffler proposes that we should observe and learn from behaviour and price development in real-world market bubbles to capture the complexity of the market.

“It is not clear that you can use the traditional scientific approach to analyse social behaviour”, Taffler said.

On the other hand, the experimental setup does not only give insight into market research, they could help understand other cases in which human groups badly judge the value of actions or events. The next step is probably not to connect every Wall Street trader to an fMRI. This study is only a proof of principle. And anyway, fMRI scanners are too expensive, even for bankers’ deep pockets.