This is a guest post by Wayne Whaley, CTA:
I first learned about the importance of extreme one day up vs. down volume in the mid 1980’s, while studying Martin Zwieg’s book “Winning On Wall Street“. In his book, Dr. Zweig suggested that any day when up volume leads down volume by a 9 to 1 margin is a sign of positive momentum for equities. He later points out that having two such occurrences in a short period of time is very bullish.
On Friday March 5th, using my unsanitized NYSE data, we experienced a 13 to 1 “Up to Down Volume Day”. We had also experienced a 10 to 1 day on February 16th. I scanned my database back to 1970 for all occasions where two such days had occurred within 21 (approximately one month) trading days of each other. All days where there were more than two such days were considered repeats. Also in my version, to squeeze out the really juicy returns, there could not be a nine to one “down” day in the last 21 trading days. I found nine instances.
“Back-to-back” 9:1 Up to Down Volume Day Signals:
*Volume Index=Up Volume/(Up Volume + Down Volume)
Also, note that there were no 9:1 down days over this period.
I was a bit surprised by what popped up on the screen. Every data point returned at least double digit returns over the next 252 trading days, with the exception of the January 1987 instance. But as you may recall, the market was up sharply 9 months later before succumbing to double digit bond yields and crashing in October. Unfortunately, adding the caveat that there could be no nine to one down days in the last 21 days, eliminated many good signals such as the spring of 2009, but this was required to keep the 252 day record perfect.
In his book, Zweig credits Lowry Research with doing the initial research on the importance of the one day up vs. down volume. If you read Lowry’s most recent update, much of their work focuses on the importance of getting a 9:1 up day that reverses a 9:1 down day.
Interestingly enough, we just got a little bit of the best of both Zweig’s and Lowry’s approach. The two 9 to 1 up days that we experienced this month were preceded by a 9 to 1 down day 21 trading days ago on February 4th. If you screen for occasions of 9:1 up days that were preceded by 1 single 9:1 up day and 1 single 9:1 down day in the last 21 trading days you get the following six data points:
Two 9:1 Up Days plus “One” 9:1 Down Day in the last 21 Days:
One should keep in mind that I did not consider instances with more than two 9:1 up days as they were considered repeats and I found that the existence of 4 or more tends to suggest the market is overextended. Now, here are the results if we combine the two studies:
To be honest, if this study were authored by someone else, I would be a bit skeptical and wonder about the bullish curve fitting going on with the various constraints. So I’m going to grant you your skepticism, but regardless, given that Zweig and Lowry have similar analysis in their back pockets, I’m not inclined to want to go short the market on an intermediate basis.
Enjoyed this? Don't miss the next one, grab the feed or