Education Resource

Statistics Don’t Lie, But Statisticians Sometimes Do

Master Instructor Blog

There were two articles in the September 14th issue of the Wall Street Journal that illustrate the danger of relying solely on statistics. First off, Jason Zwieg’s “Intelligent Investor” noted that over the past 20 years, the Dow Jones Industrial Average has delivered a 9.8% annual return while the S&P 500 returned only 8.6% annually. That would appear to set up a clever offsetting trade: Buy the diamonds then go short an equal amount of SPX and you’re a sure winner, correct?

Free Workshop

Well, not necessarily. Since 1930 the S&P has actually outperformed the Dow, 9.6% per year counting dividends compared to 9.4%. According to Kenneth French, a finance professor at Dartmouth’s Tuck School of Business, “there is no way anyone should confidently say the expected return on the Dow [is] higher than the expected return on the S&P… any differences you observe are almost certainly noise.”

The two indices are weighted differently: the higher the price of a stock in the DJIA, the greater its effect on price moves; on the S&P stocks are weighted by total market value. In addition, the Dow is just 30 stocks, and those stocks are periodically swapped out (we just lost Alcoa, the last under-$10 blue chip), making it even harder to make accurate comparisons to the broader market and further undermining the statistical reliability of the result.

The second article is, if nothing else, entertaining. “Ask Ariely” columnist Dan Ariely describes a game you can play at parties based on the statistical certainty that if you flip a coin, over time it will land 50% heads and 50% tails. In the game you tell everybody to flip a coin and remember the outcome. Then you announce the winner is “tails” and say that everybody who got tails gets a drink and the losers get nothing. If you’re popular enough to have a large party, you will get some statistically significant information from the results. If over 50% of guests announce they are winners, then some percentage of your friends are liars.


The moral of these two stories is that you can do pretty much anything you want with statistics by choosing the ones that support your case and presenting them in the right context.  Among traders, the loser is the uneducated investor who takes them literally in making market decisions. That investor would do a lot better to follow the simple, rules-based strategy taught at Online Trading Academy which allows us to anticipate market turns based on price and price alone—regardless of what the statistics say.

This newsletter is written for educational purposes only. By no means do any of its contents recommend, advocate or urge the buying, selling or holding of any financial instrument whatsoever. Trading and Investing involves high levels of risk. The author expresses personal opinions and will not assume any responsibility whatsoever for the actions of the reader. The author may or may not have positions in Financial Instruments discussed in this newsletter. Future results can be dramatically different from the opinions expressed herein. Past performance does not guarantee future results. Reprints allowed for private reading only, for all else, please obtain permission.