It seems you have JavaScript disabled.

Ummm.. Yeah... I'm going to have to ask you to turn Javascript back on... Yeah... Thanks.

Searching For Bear Tracks In The Tape at Trader’s Narrative

Searching For Bear Tracks In The Tape

This is a guest post by Wayne Whaley, CTA:

Over the last 12 months, I have posted several articles concerning the bullish nature of tape action during 2009, specifically the repeated thrust signals that have occurred over the last 12 months coming after the selling capitulation in October 2008.

Recently, I received an inquiry from a reader about what it would take for the tape to give me a bearish signal. Here is the start of a response: In January of 2010, I wrote a 17 page paper on reading the tape. The following is a summary of a tidbit of that paper related to negative tape action. My analysis is of an intermediate nature.

There are a couple of different tape observations that would cause me concern. The first is a long period of time with no signs of either a selling capitulation or a thrust signal. The second would be signs that the market is confused with unusually large numbers of issues going in different directions. Today, I will address the first of those two concerns.

bear growl

For a little background, my paper identified conditions related to thrust or reverse thrust (selling capitulations) that were extremely bullish. We have touched on many of those recently. I outlined in my paper, 51 of these signals from 1970 through 2009. Assuming that each signal is good for 12 months and eliminating repeats, the 51 signals condense to 12 time periods. Below are the results for the S&P 500 during the signals, assuming that you exit long positions 252 days after the last observed signal.

breadth volume price data table Mar 2010

Any time within the 252 day thrust signal period that a new signal was observed, the long exposure was extended an additional 252 days (approximately 12 months). For example, in the third period listed, an initial “price” thrust was observed on October 10th, 1974 and then a “breadth” thrust was observed on October 11th, 1974, January 6th, 1975 and January 6th, 1976 extending the exit date to January 7th, 1977 (252 days after the last signal on January 6th, 1976).

We reviewed the bullishness of the breadth and volume signals recently. Today, I want to address the question of how long the market needs to go without a signal before we should grow concerned.

Absence of Thrust Signals
As we have identified, the sweet spot for playing the long side of the market is the year after one of our previously defined thrust signals. If the thrust is powerful enough, the momentum generated can actually often carry the market beyond the first 12 months after the signal.

For example, in the above table, if you looked at the first year after the 12 month signal coverage had dropped, you would find that the S&P still returned a respectable average return of 10.9% that following year and was up 10 of those 12 second years. Not the kind of odds, you want to bet the house on, but above the 8.0% annual average and certainly not a period that you want to rush into short positions. But the longer the market goes without a thrust signal the more vulnerable the market becomes. Cases in point, in descending chronological order:

2007 - The market peaked on October 19th, five years and three months after the last thrust signal on July 30th, 2002. A seventeen month decline followed (-56.8%).

2000 - The market peaked on March 24th, nineteen months after the last signal on August 31th, 1998. A nineteen month decline followed (-49.14%).

1987 - The possible exception. The market peaked on August 25th, 1987, eight months after the last thrust signal on January 8th. A very short, but painful, two month decline followed (-33.24%). The January thrust signal was very accurate for the 3, 6 & 9 month time frames but eventually succumbed to the three rounds of Fed tightening from April to October. Thrust signal advocates will argue that it was indeed the strong momentum of the market that allowed it to fight the forces of the Fed for as long as it did and will also point out that even with the crash, the signal was still able to post a 1.47% one year gain.

1981-82 - The market peaked on November 28th, 1980, three years after the last thrust signal given on November 11th, 1977. A nineteen month decline followed (-37.2%).

1976-77 - The market peaked on September 21st, 1976, two years after the last thrust signal on October 9th, 1974. A seven month decline followed (-16.6%).

1973-74 - The market peaked on January 11th, 1973, thirteen months after the last thrust signal on December 1st, 1971. A twenty one month decline followed (-48.15%).

Absence of a Thrust Signal for an extended period is not a guarantee of an impending bear market, but is often an indication that the market is in the late innings. Two years without a thrust signal would be similar to a pitcher approaching triple digits in pitch count. As long as the market is sailing along, let her go, but one hard hit ball and it’s time to hit the exit.

For the record, the last thrust signal we experienced was on September 10, 2009. That is not guarantee that the upward trend will last until late 2011. As mentioned, there are other bearish conditions that my model searches for, as well.

If there is interest, I’ll try to address those at another time, else I’ll wait until they are observed to share. However, I wanted readers to know that I am not a perma-bull and I am actually keeping a close eye out for bear tracks. There is just simply no sign of any to date.

Enjoyed this? Don't miss the next one, grab the feed  or 

                               subscribe through email:  

2 Responses to “Searching For Bear Tracks In The Tape”  

  1. 1 W. Peary

    What database and software does Mr. Whaley use to do his studies? I would like to do some historical testing of various indicators and market performance. Thank you.

  2. 2 Wayne W

    Mr. Peary,

    “What database ?”

    In 1983, I ordered my initial database from a database provider. I want to say it was CSI, but I couldn’t really tell you for sure. I have been updating my data daily since. I currently get real time data from Prophet. I find that this daily data, along with most of the market summary data that is published in the financial sections of newspapers (USA Today) and internet sites (Yahoo and CNBC) differ. The following day, I defer to the WSJ and at the end of the week I double check all my numbers against Barrons. Those two are normally in agreement and seem to be the final resting place for official numbers. After checking against Barrons over the weekend, I back up all my data to a backup computer and then secondly to a disk.

    “What Software?”

    I have been asked this question a couple of times before. I am a retired engineer, and use Fortran (the engineering language of the 70s and 80s) to write my own software. As I have mentioned. it is hard to teach an old dog new tricks, but writing my own software allows me the flexibility to study most any topic I wish from whatever angle I desire. I couldn’t even tell you what engineers use these days. Maybe one of our readers could update me, as I am out of touch on such matters, but slightly curious.

    yata, yata,

    Since you mentioned studying indicators, I get a lot of good study ideas from articles I read online. After reading the ‘Ides of March’ article on here this week, I did some number crunching to determine if there was statistical evidence supporting the March Pivot Month theory that you may read on here in the next couple of days. I found the Encyclopedia of Technical Indicators” written by Colby and Myers to be a good source of ideas to study. Also, the Stock Market Almanac (Hirsh) and Stock Market Logic (Fosback).

    I have fancied trying to publish a similar book some day. The subject that I think I could tackle that would be bring something different to the table is the idea of allowing the computer to determine how to interpret indicators rather than the analyst telling the computer how to interpret indicators, and pattern recognition analysis. Hopefully, I can stay healthy long enough to get to those. I still consider myself to be on the steep end of the learning curve, or said another way, the more I learn the more I realize how much more there is to know.

    But more info than you ask for, maybe continue another day.

Leave a Reply