Sunday, January 10, 2016

January Is a Cloudy Crystal Ball

Given the vast number of variables that might be used to predict stock prices, it’s possible that even a high degree of accuracy of the January barometer is a purely random outcome.

“If you have a truly random variable, and there are, say, 60 million possibilities, it’s impossible not to find some pattern somewhere,” said Nassim Nicholas Taleb, a professor of risk engineering at New York University, author of “The Black Swan” and a former derivatives trader on Wall Street. “You might well find a correlation between changes in your grandmother’s blood pressure and stock prices. But that’s a spurious correlation.”

He added, “It’s surprising we don’t have more weird correlations, given the vast number of possible variables and the large number of markets we have.”

Professor Goldin said that if you flipped enough coins every year, you would most likely find one that predicted stock prices with 100 percent accuracy. “You could call that a magic coin,” she said. “It has an impressive correlation looking backward. The problem is, it wouldn’t give you any sense of how well it’s going to do going forward.”

That seems to have been an issue lately with the January barometer, whose accuracy in recent years has been slipping. Over the last 10 years, it gave false negative indications in 2014, 2010, 2009 and 2005, and a false positive in 2011.

[---]

Professor Taleb said that despite this week’s decline in share prices, he would not be buying United States stocks now because of high valuations, rising interest rates and the continuing plunge in commodity prices. And while that decision has nothing to do with a bearish forecast from the January barometer, he doesn’t entirely dismiss such market maxims.

“One thing you learn as a trader is, don’t bet against the folk wisdom,” he said. “The odds are this is a purely random outcome, but sometimes there’s a kernel of truth in these sayings.”


- More Here

No comments: