August 28, 2020
Market rushes
How well informed are traders during a panic?
Yastremska
Financial markets usually work well, but they do occasionally go awry. From mysterious swings to spectacular crashes, seemingly bizarre movements have led many close observers to conclude that traders are prone to going “crazy” or “completely mad” from time to time.
However, market panics and frenzies aren’t as irrational as they appear at first glance, according to a paper in the August issue of the American Economic Journal: Microeconomics.
Author Chad Kendall found in an experiment that people often trade on their first instincts to beat other traders to the punch. According to his model, this strategy is actually optimal in some cases. But in others, he showed how laboratory participants succumbed to irrational biases.
The takeaway is that people panic when they should, and they panic when they shouldn't.
Chad Kendall
Previous work has highlighted how herd-like behaviors drive markets in unexpected ways—what John Maynard Keynes called the animal spirits. But Kendall highlighted the role that information and timing play in trading decisions.
The key in his study was that it takes time to figure out what a stock, bond, or other financial asset is worth. And that delay causes a dilemma for traders.
“They can trade on the low-quality information they have right away or they can wait and try to find better information,” Kendall told the AEA in an interview. “But the problem is, if you wait to try to get better information, prices are going to move against you on average.”
This tradeoff pushes people to act as quickly as possible—what Kendall calls a market rush.
Today’s swings in the stock market offer an example.
Experienced traders bought and sold quickly to news about the spread of COVID-19. They understood, without knowing the exact impact, that hotels and restaurants would likely suffer while medical supplies boomed.
At a moment when news is just breaking, it’s not prudent to spend time on in-depth analyses and miss what intuitively appears to be an opportunity.
“Nobody’s really doing any research at that point because everybody’s trying to get ahead of everybody else,” Kendall said. “So what that means, at least in the short term, is that prices can get pushed far away from fundamentals.”
Kendall set up an experiment in which he could control all the information to make the incentives more clear-cut than is possible in the real world.
He recruited eight subjects to trade stocks with a computer over thirty trials. Participants could buy or sell an asset once as the computer released information about its value over several periods. The computer updated the price of the asset as participants either bought or sold.
At one extreme, the initial information provided was quite good. In this case, the theory worked perfectly. Participants rushed to buy or sell so as not to let the price of the asset move against them. As a result, they made as much money as they could on average.
At the other extreme, when the initial information was very poor, participants didn’t wait for more information as the theory would suggest. Instead, they waited a little bit but then traded too quickly.
Rush or wait?
The figure below shows how participants behaved under two different scenarios—rush (R) and wait (W). In W, the initial information about the asset is poor. In R, the information is of higher quality. The x-axis indicates one of the eight trading periods, and the y-axis marks the cumulative share of trades prior to each period. Solid lines are for the last five trials (Late), and dashed lines are for the first five (Initial).
“The takeaway is that people panic when they should, and they panic when they shouldn't,” Kendall said.
He found that participants were prone to chasing market trends. This ultimately led to bad information getting incorporated into the price of the asset.
The results have implications for how quickly trades should be allowed to take place. Currently, trading on stock markets occurs almost instantaneously. But Kendall’s work hints that markets may work better if trades are only allowed in fixed intervals, and people have to take a breath before rushing to trade.
♦
“Market Panics, Frenzies, and Informational Efficiency: Theory and Experiment” appears in the August issue of the American Economic Journal: Microeconomics. Subscribe to the monthly Research Highlight email digest here.