Article Tools
Try this: Imagine two graphs, each charting the performance of a different stock. The first resembles a mountain range, with deep crevices and soaring peaks. The second graph is more of a jagged line, a series of short ups and downs. Now, if you had to guess which stock was the riskier investment, which would it be — the Rockies or the ripple?
Most people select the mountainous graphic as the riskier stock pick, but it turns out this decision is rooted more in skewed perception than statistical reality, according to a study recently published in the Journal of Consumer Research. Subjects were shown a series of stock charts, some with short up-and-down movements, or run-lengths, and some with longer runs. Participants overwhelmingly identified the stocks with longer movements as riskier bets, even when they were statistically identical in other respects to ones with shorter runs. Dubbed “the run-length effect,” this phenomenon suggests investors may be making baseless decisions as they peruse the dozens of websites, newspapers, e-mail newsletters and other sources that increasingly rely on graphs to present stock information. Stock graphs are so widespread, the study’s authors even wonder if we need guidelines for them similar to the ones governing the labelling of food or prescription drugs.
“Everybody is sitting at their computers looking at these charts on Yahoo and Google,” said Sanjiv Das, a finance professor at Santa Clara University and one of the study’s authors. “But you really get less information from a chart than from statistics, even though a chart looks better, because your brain doesn’t access all the information.”
Confronted with a year of stock data compressed into a single chart, the study suggests that our brains focus on a few points, namely the minimum and maximum prices. But when investors concentrate on the highs and lows at the expense of everything else, it distorts their view.
Das and fellow researcher Priya Raghubir of New York University found this bias toward short run-lengths to be remarkably pervasive. In one experiment, they recruited volunteers through an elementary school’s parent-teacher association. More than 95% of the participants were college graduates, three-quarters had a household income of more than $100,000, and over a fifth worked in the financial sector. The researchers showed them two charts with identical means, variances and other statistical characteristics. Ninety per cent of this educated, affluent group identified the stock with the longer run-length as riskier.
For some, the research illustrates the limitations of stock graphs as analytical tools. “It might be fun to look at them, or to try and guess where a trend is going, but if you’re going to start using them seriously as an investor, that would be a dangerous route to go down,” said Eric Kirzner of the University of Toronto’s Rotman School of Management.
Kirzner argues stock graphs reveal a company’s past performance but offer little insight into its balance sheet or earnings potential. He speculates that graphs have real value only to investors with skill at technical analysis. But Das’s research challenges even that assumption. Increased levels of education and trading experience make investors even more susceptible to the run-length effect, he said. “We believe the people who are better tuned into the data effectively overdo the bias.”
Das’s research becomes increasingly relevant as more investors use the Internet to both glean stock information and direct their own portfolios. E*Trade, an American online brokerage, was adding 1,000 new accounts each day at the start of 2009. In Canada, Scotia iTrade, reported in March that their new account openings had increased by 62% over the previous year. “What used to be the domain of the professional investor is now very much available to the online investor,” said Duncan Hannay, managing director of Direct Investing for Scotiabank.
Hannay claims stock graphs can be useful tools — in conjunction with other information sources. One online investor, who asked his name not be used, uses graphs to track commodities or indexes, but said they are useless for judging particular stocks. “They are too easily skewed,” he said. “You can get whipsawed very quickly.”
Das argues it may be time for financial regulators to require statistical data be displayed alongside the pretty graphs. Our brains can’t assess the risks, he argues, so the government will have to help.
Rate this article
