The National Association of Individual Investors surveys its members via the Internet each week to see if they are bullish, bearish or neutral on the stock market, and then publishes the results on Thursday. Over the past two decades, the survey has become a compelling contrarian indicator: If the results are very pessimistic, for example, it usually signals that the stock market is about to rise.
But few people know that the sample size of this survey is very small and the statistical method has many loopholes, making it almost statistically worthless.
ADVERTISEMENT
According to the association, only two to three hundred people take part in the survey every week, which is less than 0.2% of the association’s 150,000 members, and most of those who take part in the survey are retirees.
David Madigan, professor and chair of the statistics department at Columbia University, said that from a strictly statistical perspective, the NAII survey is essentially worthless. What is particularly troubling is that the survey relies on people voluntarily reporting to their bosses.
“You have to worry about whether the people who volunteered to take the survey are biased,” Professor Madigan said. “Why did these people decide to take the survey? You can’t assume that the results of the survey represent the feelings of all members, but maybe the views of these 200 participants can predict the direction of the stock market.”
Indeed, despite its significant statistical limitations, the survey has been surprisingly effective at marking market turning points, especially stock market bottoms.
ADVERTISEMENT
