Will much advertised Head & Shoulders pattern send market lower?

There is a potential bearish Head & Shoulders pattern developing in some of the major stock indexes. I’m sure you’ve heard about this pending pattern, that is if you’ve watched the financial news, read many blogs, or have been on twitter. You can clearly see the left should and the head already formed on the chart of the S&P 500 ETF above. There is a well defined neckline that is almost horizontal. The right shoulder is beginning to appear to be building a symmetrical shoulder to the one on the left.

The psychology of the pattern can make sense. A first impulse move is put in. Then a correction appears that finds support. Then another impulse move to new recent highs, but usually with less duration that what preceeded the pattern. Then another retest back down with the market holding near the previous support level. Then a third rally that fails to get back up to the level of the second rally, with the market usually stopping close to the level of the first rally. Finally the market again retests the support line that held twice before. Traders and investors get discouraged and begin selling, thus driving prices well below the neckline support. Bulls often rally the market back up to the area of the neckline in one final test, which usually fails. Then the market rolls over to begin a large decline.

The above paragraph is what is found in most beginning books on charting. Indeed the head and shoulds pattern, and the inverse head and shoulders pattern found at market bottoms, are eye catching when seen on past data on a price chart. The successful patterns tend to pop off the page. However, on closer inspection, many more failed pattern can be found. The eye tends to gloss over the failed patterns.

The larger question regarding this impending pattern, as well as major chart patterns in general, is if there can be an edge in trading such patterns. This particular developing pattern has been mentioned numerous times in the last week or two. I’ve heard it and seen it mentioned dozens of times already. There are more chart readers and technical analysts than ever, and they’ve all read the same books, and I would guess that every single on of them are watching this pattern. I’m not quite sure how there can possibly be an edge in what everyone sees. If a major pattern presents itself, it can become self fulfilling for a time. However, professional traders are also aware of the pattern and could be quick to fade the move. It will be truly amazing if this pattern completes itself as many expect. What everyone sees is usually wrong. There can be no edge in betting with the majority. The cycle indicator on the chart could easily roll over if the market heads back to the red neckline, but I’d be careful going with the heard at that neckline. I’ll be watching this carefully and will update as the pattern develops.

There, I got through a whole post without commenting on Al Frankin…..well, almost.

6 thoughts on “Will much advertised Head & Shoulders pattern send market lower?

  1. Hello. I’ve read most of your indicator articles (from 2007) and read in your post that you are using the double stochastics on the present chart, but what is the “white” line on the subchart? Ehler’s adaptive cci?

    Your older articles this year (mar-jul) also show a lot of double stochastic charts [and that white line]. Do you favor the dbl stochastics now over the adaptive cci formula? Just wondering if the huge bear market made you tweak your indicators and how you use them.


  2. Larry, The white line is just the 5 period double stochastic. The black line is the 10 period. I tried using an adaptive period, but prefer to just use the 10 and 5 period as I am used to how they interact. I do prefer the double stochastic to the CCI. I still refer to the CCI in some cases, such as how long the CCI stays beyond the 100 lines as a clue to when trend may end, but other than that I use it very little. I haven’t tweaked anything for the current market. I’ve had the same parameters for years.

  3. Thanks for the response. I have been trying to use the CCI smoothed with a simple moving average (6 period CCI, 5 period SMA) to find cycle tops/bottoms, but have not been that successful. Bottoms are generally more accurate that tops, as the CCI can stay overbought for a long time, unlike the market, which tends to reverse extremely oversold conditions quicker than overbought.

    A question about your double stochastic indicator. I started to fool around with it. I can’t seem to reproduce your daily chart of the SPY. Are you essentially taking the result of the stochastic indicator (found in almost all chart packages) and then feeding in that value into the same stochastic indicator with the same parameters (except, the values for high/low/close are just the output values from the first stochastic function).

    I’ve been trying to do this and have noticed that using an EMA or SMA [as part of the original stochastic function] shows very different results.

    Just wondering if you saw the same and have any comments on what you found more accurate…ema or sma…I am using a length of 9.


  4. Larry,
    I think the difference is the smoothing. I use 3 period emas on both, so a 9 period would give a different look. Also, if you are using your platform’s basic stochastic as a starting point they might be using a different smoothing. It probably doesn’t really matter as long as you are comfortable with what you have. I didn’t backtest or optimize anything, so different periods or smoothing might work fine.
    So here is how I set it up: I programmed this from the standared formula for the stochastic, which would be a fast stochastic, and applied a 10 period lookback and then smoothed the result with a 3 period exponential moving average, which basically results in a Slow K line. Some software packages use a 3 period simple, some use a 5 period exponential, so result will vary. I then plug the output of that into the same fast stochastic formula, again using a 10 period lookback, and also smooth with a 3 period exponential moving average. I then multiply that by 100 to make the numbers eaiser to read. The white line is the same thing, but I use 5 as the lookback in both the first stochastic, and then again in the stochastic of the stochastic.
    Let me know if this makes sense and you get similar results.

  5. I got it Doug! Thanks!

    After I posted my comment yesterday, I started to play around with lower value ema’s, and when I hit 3, it looked a lot like yours [close enough for gov. work]. I am using basically the same formula as your stochastic so I am happy that I could at least reproduce some of your work. Now I have to find what works best for me (as you have stated many times on the site). I’m going to try different ema’s and sma’s for my trading style, and for stocks…btw, it seems like you like to trade indicex futures…I’m not sure if you trade stocks, but did you find in your research that these “cycles” occur in most types of markets – commodities, stocks, bonds, etc….that is my next research project…looking at these other markets.

    Thank you very much for a great starting point with your articles on your site! I’m gonna cut bait and teach myself to fish some more 🙂

  6. Larry, I do seem to favor indexes, but also use this on stocks. I even used it on daytrading tick charts of index futures. So have been using the double stochastic in all markets and time frames that I follow. I made no attempt to tweek it for each market. I just use the same setup. But it is only a guide. Price action is the main thing to watch in my opinion. I also use a couple of other oscillators that I rarely put on the blog. But I don’t get trading signals from any of them. They just help me read the price action. There are many articles you can find by Walter Bressert on this indicator. I have a link in my resources tab. Also, those INO videos that I have a link to on the left side of the blog have many Bressert lectures recorded that you can listen to, along with PDF files of the handouts. They were done about 10 years ago at trading conferences, but information just as valid today. I think there are at least a half dozen by Bressert, and many excellent presentations by Linda Rashcke. She doesn’t use the double stochastic, but has some excellent lecyures on trading using price action and very simple indicators.

Comments are closed.