Digging Into The Head & Shoulders Pattern

Patterns based on bar charts came much later. The classic work on chart patterns is usually considered to be Technical Analysis Of Stock Trends, written by Robert Edwards and John Magee, first published in 1948. Edwards and Magee detailed the H&S pattern and the rationale for why it should work, just one of the many patterns they addressed. But they were not the first to do so. It was Robert Edwards's uncle, Richard W. Schabacker, who seems to have been the first to write about bar chart patterns. He introduced the idea in his 1930 book, Stock Market Theory And Practice. Extensive details were provided by Schabacker in 1932 when he published Technical Analysis And Stock Market Profits: A Course In Forecasting.

Real H&S patterns rarely resemble the precise lines seen in books, of course. The daily chart of the British pound in Figure 1 shows a tradable H&S pattern that formed in late 2009. This example includes a throwback, which is often seen. After breaking down through the neckline, prices tend to trade back up to that level. Throwbacks illustrate the idea that support, once broken, becomes resistance, and the short attempt by prices to rally higher often fails.

In addition to providing trade timing signals, price objectives can also be obtained from the H&S pattern. The measuring rule is to expect a decline equal in size to the range from the neckline to the high reached at the top of the head. In Figure 1, the decline failed to reach the objective in the initial down move. Prices rallied to the level of the neckline before falling lower, eventually reaching the objective.

Time objectives can also be derived from the chart. In this example, the pattern formed over a period of 37 trading days. Prices often make symmetrical moves, and the price objective in the British pound was reached 44 days after the initial break of the neckline. Some traders use the expected time frame of the move to help them decide when to close their trade. In this chart, the intervening price rally, the one that failed at the neckline, occurred before the expected end of the move. Traders who maintained their short positions even as prices rose were soon rewarded.

FIGURE 1: H&S IMPERFECT. An example of an imperfectly formed head & shoulders pattern. Chart courtesy of Trade Navigator.

The desire to label continuations as H&S patterns illustrates one of the problems with applying this technique to the real world of trading. In real life, it can be difficult to spot chart formations as they are actually setting up. It is much easier to spot them in hindsight.

THEN THERE'S VOLUME
Even when looking back, we can have questions about whether a perceived pattern is valid. Volume is considered by many to be one of the factors required to confirm a pattern. Decreasing volume while the pattern forms and an increase in volume on a break of the neckline are thought of as part of the pattern itself by many technicians.

The problem with incorporating volume into analysis is that there is no volume reported for many indexes. In addition, with structural changes in the markets like high-frequency trading (HFT), exchange traded funds (ETFs), and derivatives, volume has taken on a new meaning.

Buying and selling can be done for a variety of reasons and is no longer a pure directional bet. HFT has largely replaced the traditional market making activities, and trading takes place to capture small spreads or even to capture payment for order flow. ETFs buy and sell many stocks at a time, rather than just the individual securities that investors would buy if this option weren't available, in effect creating artificial buying and selling pressure in some stocks. Derivatives allow traders a way to obtain exposure to specific investments without adding to the exchange volume, which artificially decreases the reported demand for a stock.

THEN THERE'S HEURISTICS
With or without volume, many traders find an H&S in almost every chart they look at, rather than objectively drawing conclusions from the price action. This highlights another problem with the pattern, which is that traders often see what they want to see in a chart. This behavior, which in behavioral finance theory is called heuristics, is a problem-solving technique where the most appropriate solution is selected using preestablished rules.

One possible reason that traders spot H&S patterns so frequently is a heuristic called representativeness, which is defined as a reliance on stereotypes. We look for things that we think worked in the past.

While the H&S seems to be well defined, different technicians can look at the same chart and disagree on the pattern they see. The H&S pattern, and others, may be rooted in another heuristic, which is the aversion to ambiguity. People tend to create orderly solutions in their minds to complex situations. It has a calming effect, and it is almost certainly a partial explanation for why chart patterns have been studied so extensively.

Another study, "Humans Distinguish Price Charts From Random: Foundations Of Technical Analysis" by Jasmina Hasanhodzic with Andrew W. Lo and Emanuele Viola, demonstrated that people can in fact learn to distinguish real charts from random data.

Software and the increased availability of data have significantly changed trading over the years. Researchers have been able to program the H&S pattern, which makes it possible to test its accuracy.

The first such study was published in 1995 by the Federal Reserve Bank of New York. Staff researcher Carol L. Osler, now with Brandeis University, and P.H. Kevin Chang, now at Credit Suisse First Boston, reported their results in "Head And Shoulders: Not Just A Flaky Pattern."

They found that the H&S pattern could be used to trade a basket of six currencies, and the resulting "profits would have been both statistically and economically significant." (That phrase, which is directly from the paper, highlights one of the differences between technicians and academics: They don't speak the same language. Technicians would say something like "You can make money trading the H&S pattern.")

Lo, who was interviewed by Technical Analysis of STOCKS & COMMODITIES, and has been a contributor to it as well, studied the H&S pattern from an academic perspective as well as a technical one. In "Foundations of Technical Analysis: Computational Algorithms, Statistical Inference, And Empirical Implementation," working with Harry Mamaysky and Jiang Wang, he found that the H&S pattern, along with several other indicators, does "provide incremental information and may have some practical value."

Technical analyst Thomas Bulkowski is well known for having programmed and tested chart patterns. His tests show that the H&S pattern is one of the most reliable. The test results can be seen at http://thepatternsite.com/hst.html.

Success in trading requires discipline. The H&S pattern can be traded using only charts, but many traders would be better served by defining the pattern in a software package to avoid trading what they only think they see.

SUGGESTED READING

• Bulkowski, Thomas [2000]. Encyclopedia Of Chart Patterns, John Wiley & Sons.
• Edwards, Robert D., and John Magee [2007]. Technical Analysis Of Stock Trends, 9th ed., W.H.C. Bassetti, ed. AMACOM.
• Hartle, Thom [1997]. "On The Fundamentals Of Technical Analysis: Andrew Lo," interview, Technical Analysis of STOCKS & COMMODITIES, Volume 15: December.
• Lo, Andrew [2000]. "Foundations Of Technical Analysis: Computational Algorithms, Statistical Inference, And Empirical Implementation," NBER Working Paper Series.
• Malkiel, Burton [2003]. A Random Walk Down Wall Street, W.W. Norton & Co., 8th ed.
• Nison, Steve [1991]. Japanese Candlestick Charting Techniques, New York Institute of Finance/Simon & Schuster.
• Osler, Carol L., and P.H. Kevin Chang [1995]. "Head And Shoulders: Not Just A Flaky Pattern," published in 1995 by the Federal Reserve Bank of New York. http://www.newyorkfed.org/research/staff_reports/sr4.pdf
• Schabacker, Richard W. [1932]. Technical Analysis And Stock Market Profits. Available as a download from Traders-Library.com.