Supposed Breadth Thrust Signal Actually Just Missed
There has been a lot of chatter this week among market breadth aficionados about a supposed Zweig Breadth Thrust signal. This is a rather famous signal developed by the late Marty Zweig, who passed away earlier this year.
To generate a signal, the 10-day EMA (AKA 18% Trend) of Advances divided Advances plus Declines needs to go from below 0.40 to above 0.615 within 10 trading days. The basic idea is that a sudden flip from really negative breadth to really positive numbers can be a sign of initiation of a strong new rally. The late Mr. Zweig found that it needed to happen in a short enough amount of time to be genuine.
According to Zweig's original criteria, the market just missed triggering a signal this week. There was a dip in that EMA to below 0.40 on June 24, but it would have needed to get up above 0.615 by July 9 to qualify for the 10 trading day criterion. The EMA did get up to 0.620 on July 11, but that was 2 days late. So no cigar.
Within the last 30+ years, there have only been two genuine Zweig Breadth Thrust signals using composite NYSE data, which was what Mr. Zweig used to develop the signal. Both of those ended up being really great signals, but 2 signals across 3 decades is not much to hang one's analytical hat on.
Some analysts have tried to cheat their way into a signal recently, by focusing on breadth data for smaller subsets of the NYSE list such as common only issues, or for the SP1500 stocks. But those subsets are not what the signal was developed for, and so transplanting a signal and its parameters from one set of stocks to another without testing that move is to make a prejudicial assumption.
Back in July 2011, I wrote about how Gerald Appel borrowed a portion of the math that Zweig used to create his Breadth Thrust rule, and used that in a trading system that Appel named "Time Trend III". Appel discarded the requirement that the 10-day EMA (AKA 18% Trend) of A/(A+D) dip below 0.40. Appel focused only on that EMA rising up above 0.615 as a "continuation signal" which took away the system's rules for going short. His presumption was that gobs of strong breadth were a sign which said one should not get shaken out of an uptrend that is underway.
The problem is that since the late 1980s, when Appel first formulated these rules, the market has perhaps changed in its personality such that these signals are about as likely to signal the end of a blowoff top as they are to show continuation.
Since it has been more than 10 trading days from the below 0.40 reading to the above 0.615 reading, the latest instance does not qualify as an official Zweig Breadth Thrust signal. But it would qualify as an Appel continuation signal, as shown in the lower chart. The problem is that a lot of these supposed continuation signals have turned out to be just blowoff tops. In fact, most of the recent signals have carried that meaning rather than conveying a message of upward continuation.
We are operating now in an era when the Federal Reserve seems intent on having its way with the market, and with vanquishing any and all illiquidity dragons. The question for analysts and investors is whether $85 billion a month worth of stimulus is enough, and whether the renewed excitement over the punchbowl not being pulled away is real or just another blowoff. Recent history suggests that the blowoff scenario is more likely.
|Jul 01, 2011
A Breadth Thrust Signal
Jun 06, 2013
May 16, 2013
Source: The McClellan Market Report
About Tom McClellan
Tom McClellan Archive
|12/12/2014||What Hindenburg Omens and an Oil Crash Mean||story|
|11/14/2014||More Good News for Employment||story|
|11/07/2014||More from Investors Intelligence Data||story|
|10/31/2014||Investors Intelligence: Initiation of Strong New Uptrend||story|
|10/17/2014||Summation Index Rate of Change||story|
|10/10/2014||COT Data Show a Bottom for Copper||story|
|09/12/2014||Gold - Watch the €1000 Level||story|
|08/28/2014||A-D Line Foretells Job Openings||story|
|08/22/2014||Small Cap Relative Strength||story|
|08/07/2014||2nd Presidential Years Are Different||story|