Home Page
 
Leon Wilson
Leon Wilson is the author of the highly respected books "The Business of Share Trading"; "Breakthrough Trading: Revolutionary Thinking in Relative Analysis" and "The Next Step to Share Trading Success". Leon's primary focus is position trading. He is a firm believer that to succeed in any business it is necessary to always be looking for a better way of doing things. Part of his trading approach is to question conventional wisdom and push the boundaries of accepted technical analysis theory.
In this article Leon looks at trading patterns.

PATTERNS

     

With programmes such as MetaStock and BullCharts, it becomes easy to immerse ourselves in a world of complexity. There is nothing wrong with complex trading techniques providing the complexity is relevant to the approach, however complex techniques are not essential for successful trading.

Patterns are without doubt one of the most consistent methods available to the trader for defining probable price behaviour. It concerns me the number of people who still consider pattern trading and pattern analysis in general as some sort of sorcery and witchcraft. Patterns define probable crowd movement and the larger the crowd becomes, the greater the probability becomes of us anticipating its direction. For those who doubt the validity of pattern analysis we will have a quick look at pattern behaviour in a slightly different context. If we go to the football, what is the probability of the crowd going to the food stand at half time? We do not know the exact numbers or in what order each individual will be served, however what we can realistically anticipate is that a large number of patrons will be going for food. We can also expect the same patrons to return to their seat before the match resumes. We do not know when they will choose to sit down, however the probability lies with the majority being back at their seats before the match resumes. If we go to the football every week, then we will see this behaviour repeated each week and in the same manner. The crowd will consist of different individuals each week as the visiting team changes each match and while the crowd differs each week basic crowd behaviour remains unchanged. The larger environment determines general crowd behaviour. Take the pattern skeptics to a football match in order to watch the same behaviour repeat itself week after week and they will quickly state that this is predictable behaviour, which involves nothing more than the application of common sense. Apply the same observations to the stock market and all of a sudden it becomes witchcraft and sorcery. Basic crowd behaviour remains largely unchanged regardless of the environment. It may appear slightly different on the surface however once we look beyond the fringes and delve into the core of crowd behaviour it does have a level of consistency regardless of the surroundings.

  • Personally, I divide my patterns into two groups, trading patterns and directional patterns. The primary difference between the two is that trading patterns tend to be short term with regard to maturity, specific in development and allow for precise target projection.
  • Directional patterns fall outside of specific development criteria and often develop over the longer term. Targets can be projected from directional patterns however probable trend direction is the primary intention behind the concept rather than identifying precise points of price action. Targets tend to regions rather than specific values;

The most common misunderstood directional pattern that the majority of us are familiar with, would be the head and shoulders. It is surprising how many times you will see the head shoulders pattern applied over the short term with price projection shown. True head and shoulders are long term reversal patterns applied to primary trends. Target projection is usually applied from the neck line and the target is reliable when constructed correctly. When it comes to actually trading off a head and shoulders pattern the benefit is with primary trend entry signals. The successful upward penetration of the neck line with a bullish head and shoulder is an entry signal for a long position. The depth if the pattern is used to project the upside target. Naturally the same applies with the successful downward penetration of a bearish head and shoulders pattern. The downward penetration of the neck line signals a short trading position while the depth of the pattern allows for an expected downside target. Forget this non sense about exiting positions on head shoulders development. Most of those who advocate the concept of exiting on the breach of the neck line forget two basic facts with disciplined trading. The majority of traders would have closed the position, if not on the initial retracement from the first shoulder, then definitely on the pull back from the head. To suggest that we know in advance that a head and shoulders is developing therefore we should ignore the first two retracements and exit on the bearish penetration of the neck line is dangerous thinking. The pattern can only be confirmed once the second shoulder develops which is usually somewhere around 80% of pattern development and to suggest that the exit should be delayed on this assumption is ridiculous. If the trader is not going to exit on either of the first two retracements then they are unlikely to exit on a penetration of the neck line. Discipline traders will have closed the position well before the development of the second shoulder while the undisciplined trader will continue to remain with the position regardless of ongoing pattern development. Head and shoulders are better suited to entry techniques as the pattern is complete when the entry signal is generated. Exit signals based on a neck line exit requires the presumption that the pattern is currently developing and will complete as expected. This is not disciplined trading.

Trading Patterns
The majority of trading patterns are continuation by design with the triangular pattern being the preferred pattern of analysis. Consolidation is also a pattern in its own right that receives limited attention. In recognizing patterns we need to have some understanding of the intention behind pattern analysis. My interpretation behind the purpose of pattern identification is:-

Pattern analysis is a method of defining current crowd behaviour with the intention of defining expected crowd direction and probable destination.

Rather than focusing on triangular patterns we will initially look at consolidation analysis. Consolidation is very specific in its actions and the outcome is often consistent in its behaviour therefore such developments are also entitled to be included as pattern techniques.

Crowd Consolidation
Computershare experienced a break above resistance at $2.00 on the 14/08/2003 (chart 1) following a period of non-trending activity. There is no obvious development of any trading pattern at the point of the initial break out. With no trading pattern clearly visible, we are unable to determine any probable upside target at the time of the break. Price action briefly rallied then paused throughout September creating a period of consolidation. It is common for consolidation periods to occur mid trend or rally, so with this in mind we can gain some insight to further upside should prices recommence to move in an upward direction.

Chart 1: Computershare

It could be argued that consolidation periods are not true trading patterns, as we tend to focus on the bulk of price action while ignoring the occasional extreme high or low. Trading patterns such as flags and triangles tend to focus on the extremities of pattern development and are very precise in their construction. With consolidation analysis we are aiming to evaluate group behaviour in general and this is where we come back to our football match. We can not anticipate the movements of an individual spectator with any certainty, however we can anticipate the movement of the majority with a fair degree of probability. It is these same reasons as to why I focus on the majority when analyzing consolidation periods. The masses have temporarily paused and are congregating in the same region. As I am interested in the group as a whole I focus on the core of consolidation. The occasional individual that strays away from the group and hangs around the fringes of the pack is of no major concern. Finally a clear break develops and we have the first signs that our group of traders are on the move again. Such breaks are usually accompanied by an initial increase in volume.

With chart one, the top of our consolidation area is in the vicinity of $2.88. This has given the initial rally a value of $0.88. Should price action break from our consolidation period then the upside target will be in the region of $3.76. I originally selected CPU as it was shown in the daily update at the time of the consolidation break. With hindsight I wish now that I had not selected CPU as price action performs on cue and hits the target price of $3.76. While I am not keen on perfect examples real life trading does carry a level of consistency with CPU behaving in the expected manner.

Having identified a period of consolidation how could we trade it assuming that the trader is not already in the position.

  • If the stock is still trending, then we can enter on the day following the first close above the top of our consolidation box used for analysis;
  • The upward move should be on above average volume and the closing price should be in the top third of price action on the day of the break. The closing price finishes above any price action belonging to our consolidation period;
  • Upside targets should be treated as regions rather than specific values, even though we have the ability to draw precise values for expected upside targets. This is due to the more generalized approach taken to consolidation analysis;

The upside break from consolidation signifies potential trend continuation therefore the likelihood of a rally or trend style trading opportunity exists. From this perspective we can apply our normal trailing stops such as ATR or relative percentage. Price action will often break upward then briefly return to the trade near the top of the consolidation period before continuing with up trending behaviour. Treat the break out from consolidation no differently to a break of resistance where we commence with our trailing stop from the most recent low. If the trailing stop commences below the lower extremities of our consolidation range then the more conservative trader may wish to place an initial stop 1 tick below the lowest low of the consolidation range. Personally I am not so keen on such an approach as it places a precise value on general behaviour. A trailing stop such as ATR that is applied from the low immediately prior to the upside break out is determined by specific price behaviour that is directly associated with the trend that is now potentially underway.

If price action is non trending then treat any upside break out in a similar manner however we should consider the application of the minimum close technique or similar to ensure that price action moves a sufficient distance from the consolidation range prior to entry. Many traders have found that once price action moves away from non trending activity by more than 3.0% then the break often becomes sustainable as the influence of previous price action is lost. Percentage values are acceptable however the minimum close technique draws its values directly from price action making upside entry levels relevant.

While price action is not always tightly compressed it does consolidate within a fairly narrow trading range as experienced by ComputerShare and generally speaking can be easily defined.

  • Consolidation periods are analyzed as continuation patterns. The direction of the pattern break out is usually consistent with the prevailing trend;

I have found that these developments tend to be fairly consistent over the shorter term, however if the intention is to analyze from a longer term perspective then weekly data would be a preferable alternative. As for down trends, consolidation periods do occur on occasions, however I have often noted that price action will regularly exceed projected downside targets.

Compression Patterns
Taking this type of behaviour to the next level, price compression develops in a similar manner to that of consolidation however as the period of consolidation develops price action becomes increasingly compressed as volatility continues to abate, often to a point of being almost non-existent. Eventually price action will explode into life often in a violent manner. Generally speaking, the longer the period of compression the more buoyant price action will be that follows. With regard to compression patterns, they are most effective in stocks that are not prone to this type of behaviour. Property trusts will often develop in a slow and consistent manner giving the appearance of price compression. This is not genuine compression behaviour but rather a normal characteristic of price action due to the sedate nature of property trusts.

Breville Group Limited experienced two compression patterns in close proximity which is good as I do not have to go looking for a 2nd example. While both patterns may look similar, one is a clever ring in while the 2nd is the genuine article.

Compression Pattern 1
Both patterns look similar however the first compression pattern is a wolf in sheep's clothing. Compression patterns will often develop following a notable high in price action. Price activity then drifts sideways without producing a higher value until the patterns concludes with a healthy upside move. I have drawn the ideal region for compression analysis with pattern one in light gray so that it does not become too dominant on our chart. Price action consistently trades above our initial high that led to the period of compression without experiencing an upward move. The Initial break out occurs on above average volume and the closing price finishes near the high of the day. While the break may meet the necessary rules of an entry signal it is not a clean signal due to the attempted break two days earlier. Added to this, price action was compressed however it did not contract as the pattern matured. Price action remained basically unchanged until the break finally eventuated. Volume remained consistently above average while the pattern developed which is not consistent with true compression behaviour. When the pattern break finally eventuated it was not on excessive range. While our first compression period may fit the general criteria for compression behaviour it is a poor representation of the genuine article.

Chart 2: Breville Group Limited

Compression Pattern 2
The second period of compression behaves in an ideal manner. As the pattern develops, price action continues to compress becoming increasingly tighter. Volume has basically evaporated while the period of compression develops and price action drifts into the central region of compression range. The break was on good range and on above average volume. The signal is unambiguous and very clean, while the closing price occurs in the top third of price action. The break out with our second period of compression is decisive leaving no question in the traders mind that the pattern has most likely concluded. While the period of compression developed, price action made no effort to move beyond the initial high.

Pattern Identification
  • Price action continues to compress as the pattern matures usually drifting toward the centre of the compression range;
  • Volume will largely evaporate from the market;
  • The upper extremities of the compression range usually coincides with a definitive high price, however with this said this is not always the case;

We can trade a compression pattern in the same manner as we would a consolidation pattern. Quite often a healthy rally will develop following the period of inactivity therefore we apply the usually methods for position management such as ATR trailing stops and trend line placement as price action develops.

Generally speaking, I have usually found price projection from compression periods can be fairly inconsistent. While the pattern is often a leading sign of buoyant price action, the upside range can often be unreliable. As with a consolidation pattern, we can analyze periods of compression as patterns of continuation however we should always wait until the pattern has concluded before considering entry. The key with identifying compression periods is the continual tightening of price action while remaining non directional. This setup is more important than its positioning to a recent high. Compression patterns do not occur with the same consistency in down trending price action. Compression patterns should be treated as continuation patterns and in bullish markets can be reliable on 80% of occasions.

Triangular Patterns & Projection
Silex Systems developed into a traders dream in December 2003. Decisive price action with such clinical precision does not occur very often and when they do we can often miss cashing in on the opportunity for a number a reasons. We may have no available trading capital or just simply failed to identify price behaviour at the time. It is not possible to monitor all of the stocks listed on the ASX therefore we should not kick ourselves if an opportunity develops outside of our normal spectrum, annoying as it may be. The first setup with SLX was the appearance of an ascending triangular pattern. The construction parameters are very specific for trading patterns especially when trading on daily data over the short term.

    Construction Rules
  • The perpendicular side of our triangular pattern "A" (chart 3) must consist of buoyant trading activity and develop without pauses. The starting point for our vertical side is the first day where the upward move commenced which is normally indicated by an increased trading range;
  • Using the absolute high of our initial rally we place a line of resistance in order to define the top of our pattern and then project to the right of screen.
  • A further close above the line of resistance renders the pattern invalid if less than 50% developed and complete if more than 50% developed.

Technical purists will consider a pattern invalid if the pattern experiences an additional high that breaches the pattern top. In today's more volatile markets I consider a close above resistance as a pattern breach;

  • We then connect the starting point low with further rising low points on the ascending side of the pattern. A close below the ascending side of our pattern requires the line to be repositioned.
  • The time associated with pattern development is not as important as the nature of development for the pattern. Some patterns will develop and mature in a matter of just a few weeks as experienced with Silex while others will a number of weeks.
  • Once our trend line of pattern support is confirmed, downside penetration of the trend line by the closing price renders the pattern invalid;

My rules for pattern identification may appear somewhat restrictive and there are plenty of examples where patterns that almost qualify go on to reach the expected target price. In bullish markets an almost pattern will often hit the target, however once the market becomes weak or bearish, then the reliability of ring-in's drops dramatically. Why the stringent rules? Have you ever listened to kids drawing faces in the clouds? They can spot their dog and teacher along with a myriad of other things. We look to the heavens and see a heap of heavy clouds rolling in and the only thought on our mind is the cancellation of tomorrow's fishing trip. When it comes to the images portrayed by the clouds, we look, but can not see. Why, because there is no money attached to image recognition. Now when it comes to the market and the lure of greed steps in, all of a sudden our imagination returns and we begin to see patterns everywhere. Leave the imagination to the kids; they have a gift for it, not us. We have rules in place to protect us from our biggest adversary, ourselves. Poor pattern identification in weak markets is not a good combination for survival.

Ideally the upside break out needs to occur at approximately two thirds to ¾ of pattern completion. The earlier the pattern breaks especially before 50% of pattern completion, the increasingly unreliable the break out signal becomes. Like wise for patterns that fill completely, such occurrences are also somewhat unreliable on the majority of occasions with regard to achieving consistent upside targets. In reality, pattern completion in nothing more than trend continuation therefore projecting triangular pattern is somewhat irrelevant. It is nothing more than a steady trend than has penetrated a short term level of resistance. Patterns that experience upside break outs within the range of two thirds and ¾ of pattern maturity are usually the most consistent signals generated from pattern development. It has been my general observation that correctly constructed patterns will reach upside targets in excess of 80% on occasions with up trending markets and as high as 70% in bearish conditions.

Chart 3: Silex

Looking at chart 3, the vertical side of our triangular pattern has a value of $0.39. This value is then projected upward from our line of resistance to give an upside target price of $1.32. Looks easy doesn't it, yet most of us never seem to catch these opportunities. Such short term patterns are seldom caught with end of day trading. Generally speaking, the shorter the pattern is to maturity the quicker it will reach it's upside target once the pattern breaks. By the time most of us are aware that the break has occurred, entry is not a practical choice as price action is already nearing the upside target. Many of us have the skills to identify triangular patterns but have difficulty with trading them, especially the short term patterns such as the one experienced with SLX. Before you can trade such a pattern setup you are required to have the following two criteria. Without these elements and tools we should not waste our time chasing short term trading patterns as appealing as they may be.

  • To have clinical and decisive execution skills;
  • Access to contingent order facilities or a live data stream;

The most difficult aspect of short term pattern trading is the trader. Many of us consider that the pattern is difficult to trade, which is incorrect. It's not price action that is the stumbling block in this situation; it's usually the trader. The inability to buy and sell on cue and without question is easy retrospectively but devilishly difficult to achieve in real time. That stuff called emotion seems to get in the way. Another problem is often the inability to differentiate between pattern trading and position trading. Some traders will catch the pattern break a little late, then miss closing the position near the target price, so a pattern trade becomes a position trade, that is justified with some feeble excuse about using the pattern as an entry trigger. The lack of discipline or inappropriate tools fails to receive a mention. Get yourself two hats and label one pattern trader and one position trader. Wear the hat relative to the trade in progress.

Never mix the two techniques, you are either trading the pattern or you are trading the position. You can not do both simultaneously as the methods of risk management, position exposure and trade execution will differ accordingly. If you are trading the pattern then focus solely on the target prices and forget all about the possibility of further upside beyond the target price. With SLX we have decided to enter at $0.94 and exit at $1.32 so this is what must be done. Ignore any temptation that may lead us to do otherwise as pattern trading is about clinical and disciplined execution.

The most effective method available for pattern trading without live data is through contingent orders. Contingent buy and sell orders allow for trade execution with relative precision and will remove much of the emotional attachment providing we have the discipline to set the orders in the first place.

First Published: 17 May 2005 - Copyright © Leon Wilson

This document is copyright. This publication, which is generally available to the public, falls under the ASIC Media Advice provisions. These analysis notes are based on our experience of applying technical analysis to the market and are designed to be used as a tutorial showing how technical analysis can be applied to a chart example based on recent trading data. The author and publisher expressly disclaim all and any liability to any person, whether the purchase of this publication or not, in respect of anything and of the consequences of any thing done or omitted to be done by any such person in reliance, whether whole or partial, upon the whole or any part of the contents of this publication.