How market statistics can lie
“There are three kinds of lies: lies, damned lies, and statistics.”
Mark Twain popularized that quote over a century ago. It couldn’t be more true now.
Pundits constantly bombard investors with wild economic and stock market claims backed with charts and data. Much of it is bunk. But it isn’t hard to see through it if you’re armed with a few simple tricks.
Those tricks come from one of my all-time favorite books: Darrell Huff’s 1954 classic, How to Lie With Statistics. It’s humorous, illustrated and non-academic—a great guide about how people torture data to force them to fit their argument. Here are its five most useful nuggets.
1. Beware stats from surveys
Surveys are only as good as their representativeness. But most have baked-in biases and aren’t truly representative. Political polls, wrong in so many recent elections, prove this. So do opinion surveys of all stripes. As Huff writes: “No conclusion that ‘sixty-seven percent of the American people are against’ something should be read without the lingering question, sixty-seven percent of which American people?”
2. Don’t take “averages” at face value
According to the Social Security Administration, average U.S. annual wages in 2017 were both $48,251.57 and $31,561.49. How? Technically there are different kinds of averages. The higher figure is the arithmetic mean. The lower is the median — the number in the middle, with half of wages higher and half lower. Both are “right,” but the mean average often gets skewed by outliers. You can’t assess an average until you know what kind it is and what skew it may hide. Keep this in mind when digesting climate stats. As Huff showed, Oklahoma City’s average temperature from 1890 – 1952 was a cool 60.2 degrees—but temperatures in that window ranged from -17 to 113.
3. Consider “axis” fraud
Usually, in economic graphs, the horizontal line or “axis” shows dates and the vertical axis shows quantity or size. It’s easy to make anything look way wilder than reality simply by stretching the vertical axis scale or skipping a whole chunk of it. In 2014 someone did this with the Dow to make that year’s market movement look just like the run-up to the 1929 crash. It went viral—scaring many. Properly graphed, the two periods looked nothing alike. Blatant fear-mongering, lying with charts.
What could the next recession feel like?: Stories from past downturns are a guide
AI at work:: Machines are training human workers to be more compassionate
4. Know your profits.
Does an article about corporate profits report earnings per share? Or does it talk about gross operating profit margins? Is it discussing return on assets? Or on equity? As Huff explained, “You can, for instance, express exactly the same fact by calling it a one percent return on sales, a fifteen percent return on investment, a ten-million dollar profit, an increase in profits of forty percent (compared with 1935-39 average), or a decrease of sixty percent from last year.” Writers and companies will often pick the one that best suits their angle.
5. No correlation without causation
As Huff explained, once upon a time, a study showed non-smokers got better college grades than smokers. But did not smoking help students get better grades? Or did bad grades drive more people to smoke? Is it a mere coincidence? Tyler Vigen’s hilarious website, Spurious Correlations, shows how easy it is to draw false conclusions between two totally unconnected things. Did you know annual per-capita cheese consumption correlates with the number of people who died from tangled bedsheets? Or that U.S. chicken consumption correlates with oil imports? People pull similar stunts with market data every day. That’s where we get seasonal myths.
For more, get the book.
It’s cheap and a breezy 144 pages. Your 401(k) will thank you.
Ken Fisher is founder and executive chairman of Fisher Investments, author of 11 books, four of which were New York Times bestsellers, and is No. 200 on the Forbes 400 list of richest Americans. Follow him on Twitter: @KennethLFisher
The views and opinions expressed in this column are the author’s and do not necessarily reflect those of USA TODAY.