Breadth Volatility a Good Sign
May 12, 2004
One of the hallmarks of the past couple of weeks has been the sharp reversals in breadth on the NYSE. By ďbreadthĒ, Iím simply referring to the number of stocks that close higher than the day before minus those that close lower. We have seen some very large numbers, mostly negative, but very volatile.
As one way to view this volatility in breadth, I looked at the standard deviation of the breadth figures over a 10-day period and compared that to the total number of issues traded on the exchange. Over the past 10 days, the standard deviation of the advance/decline figure has been 1631. This means that 68% of all the readings should have been within 1631 issues of the average.
The absolute number isnít all that important Ė what may be important is the fact that when we divide 1631 into 3479 total issues traded on the exchange, we come up with a breadth volatility figure of 47%. This compares to 13% near the peak in January, when the number of advancing and declining issues remained very stable over a 10-day period.
So is this good or bad? Well, each moment in the market is unique, but if history is any guide at all, then this type of violence in breadth is very positive in the intermediate-term. We only have a few other comparable instances on which to base our opinion, and there was often some short-term pain, but looking out 3 months or more, the market performed extremely well.
In fact, out of the 34 days that showed a breadth volatility reading equal to or exceeding the current one, the S&P was higher after 90 days 34 times (a 100% success ratio). Iím always leery of anything that suggests something is 100%, and by no means do I intend to suggest that we will for sure be higher 90 days from today, but it is a compelling figure.
The series of charts below show each of the distinct occurrences, with the first day of the extreme breadth volatility reading highlighted by the dotted vertical lines, along with the volatility figure and the return in the S&P 90 days later.
After the 34 days that showed such extreme volatility in the breadth readings, the S&P was an average of 10% higher after 90 days, and again every one was positive. One year later, each of them was still positive, and the average return climbed to a very respectable 17.5%.
Certainly the arguments can be made that the breadth figures are skewed by non-operating companies, or that decimalization has complicated historical comparisons, and I believe each has some merit. However, I donít think either one significantly changes the basic idea that we have experienced some violent changes in breadth over the past couple of weeks, and that has historically been a long-term positive sign.
© 2001-2007 Sundial Capital Research, Inc. All rights reserved. Disclaimer. sentimenTrader.com is a trademark of Sundial Capital Research, Inc. Sundial Capital Research, Inc. PO Box 341 St Michael, MN 55376