Jason Mitchell Articles
Constructing and Trading Flag Patterns - Part 4
 | 

In our final article on Flag Patterns we look to methods I use to locate trade opportunities and briefly cover the difference between the two types of entry strategies - conservative and aggressive. Finally I provide a number of recent Flag Trade opportunities and how these patterns were presented by the charts.

Finding Flag Patterns

There's some good news and some bad news about trading flag patterns and that is they are not the easiest things to find in the world, but they can be found. There are basically three methods that can be used.

  1. Eyeball search
  2. Price Volume Breakout Searches
  3. Specific Flag Detection Searches

Eyeball Search

The first method we have mentioned is the eyeball search. We know that this does not suit some of our readers however the truth of the matter is that it really is one of the best ways. Further to this the eyeball search gets you looking at a large number of charts repeatedly. We feel this is one of the best methods of practice when learning about technical analysis.

{Filter}
C > 0.10
AND
C < 1.50
AND
(Mov(V,50,e) * Mov(C,50,e)) >100000

You can run a search that restricts the stocks that you will look at and then eyeball the remaining candidates. For example you may use a search that finds all the stocks between $0.10 and $1.50 with the average volume being higher than $100,000. You may adjust this as you see fit.

Alternatively you can do what I often do (and copied from Daryl Guppy) and choose an alphabetical folder at random and scan this. For example you may choose the letter M and then look through this folder. This was how we found MCR. It is possible to use any folder including a list of your preferred stocks, the ASX300 stocks, a list of stocks from fundamental reports or any criteria that you feel comfortable with. It really does not matter how you choose the folder as can be seen by the random letter approach.

The main thing is to make sure that you do not spend too long at any one chart. 3 - 4 seconds should be more than long enough. If you don't see the pattern it probably is not there - move on and don't over analyse. If you are in doubt leave it. There will be more opportunities around the corner.

When I scan through the stocks then I am looking for one of two things. The first is a possible flag pole. I save any charts I see that are possibly developing a flag pole in to a separate folder or write them on a list called Developing Flags. This allows me to monitor their progress and see if a flag pattern develops.

The second thing I am looking for is a flag that has started to develop. This is what we saw with MCR last week and we only just saw it in time. The trick is again to look for the flagpole and then assess the subsequent development using the rules covered in part 1 of this article. The flagpole due to its change in range is the easiest thing to spot.

Price Volume Breakout Searches

The price volume breakout search can come in many forms and is basically searching for a flag pole developing. Once these have been found it is necessary to put these in to another folder and monitor there progress to see if they develop a flag pattern. In Daryl Guppy's book Chart Trading (I think) it has a search for a price volume breakout. This means price has increased by more than 5% and when the volume is greater than 50% higher than the average volume.

The formula as Daryl has it is already in Metastock as: Equis - Price Volume Breakout

I however have altered this formula to get rid of a lot of the stocks that it brings up. The main change I have made is I do not want to know about stocks worth less than $0.10. I generally do not like trading these types of stocks due to the excessive price leverage. Everyone thinks price leverage is good but it goes both ways and can become difficult to set reasonable stops. As such I wipe them out.

The second change I have made is to get rid of the many illiquid stocks where the chart is full of days with gaps in them. If the stock is showing excessive gapping I find these become much harder to trade and further to this the majority do not give me any reasonable trading opportunities. Inserting a simple code helps us to get rid of any stocks that have had more than 15 gaps in a period of 50 days. The alteration to Daryl's formula is:

ColA Close
ColB Ref(C,-1)
ColC ROC(Close,1,Percent)
ColD Volume
ColE Mov(Volume,50, Exponential)
ColF ((Volume-Mov(Volume,50,Exponential))/Mov(Volume,50,Exponential))*100
Filter V1:=If((GapUp() OR GapDown()),1,0);
Gaps:=Sum(V1,50);
colC >= 5 AND colD >= (colE*1.5)
AND
C > 0.10
AND
HHV(Gaps,50) < 15

We have found this search is adequate for looking for flag poles however it should be noted that it will miss those that do not rise more than 5% in a single day and it will also need to be run daily to ensure that you do not miss any developing opportunities. From this point the flag still needs to develop.

Essentially ANY search that looks for a sharp rise in price will help pick up flag poles. Many people make use of the Rate of Change function, which allows us to look for stocks that have moved a certain amount.

One of the other searches that we created and have used occasionally is what we term the flag finder search. It does bring up a lot more candidates than the above search, as it does not require price to have moved 5% in a single day. It looks for a spike in the price over the last five days. As such this search will sometimes bring up candidates that are just beginning to form a flag.

{insert in filter}
ATRP:=(ATR(2)/ C)*100;
V1:=ATRP/Mov(ATRP,50,E);
V2:=If((GapUp() OR GapDown()),1,0);
Gaps:=Sum(V2,50);
Pole:=(V1 >=2.1) AND (PDI(3) > MDI(3)) AND V > (1.5*Mov(V,50,E));
(BarsSince(Pole) < 5) AND (PDI(14) > MDI(14))
AND ((HHV(Gaps,50) < 15)) AND (C > 0.10)
AND ((Mov(C,50,E) * Mov(V,50,E)) > 80000)

The best way to work out which search is best for you is to run both over a folder and scroll through the candidates. Choose the search that gives you the candidates you feel comfortable looking at. Of course many of the results between the two will be the same and I have found most people prefer to use the price volume breakout search and monitor the developing activity for a flag pattern. Most methods revolve around looking for a flag pole. Eg. Price volume breakout, or the flag pole search mentioned above.

If you do use the search for the flag pole and run a separate folder watching the stocks as they develop then we would suggest that you delete any stocks that do not conform to the rules on a flag pattern immediately. There is generally no need to watch the stock once it has been invalidated in any way and this will help keep this list very small. (Often within a day or two a flag pattern can be ruled out).

Specific Flag Detection Searches

We are not going in to this in too much depth except to say that these are available. I haven't come across any that I could confidently say picks flag patterns with any great success, although there are some interesting products.

Daryl Guppy's newsletter ran an article on CPFinder, which is pattern recognition software that integrates with Metastock format data to look for a number of chart patterns. If you want to see the review that Daryl gave of this we would refer people to his "Tutorials in Applied Technical Analysis" Newsletter, October 30th to November 20th editions.

I personally do not use the software but I acknowledge that it does find patterns and can be very useful. We would recommend checking it out to see if it suits your trading needs. Make sure however that you have stricter rules regarding your pattern construction than the program - it will pick up a lot of "flags" that are not flags. This is not a major issue as the program helps to save time - it does not make our decisions.

Entering In to a Flag Trade

Generally it is accepted that there are two methods of entry in to a flag trade. One is considered a more aggressive approach and the other a more conservative approach. The aggressive approach aims to enter in to the trade as the flag is forming. This often allows for a better reward if the trade goes ahead however there are a few issues that make this type of trade more difficult.

The second or conservative method is to wait until price breaks above the downward channel formed by the flag. This is considerably easier in some respects however it can mean missing out on opportunities as the price can move exceptionally quickly once it breaks out. On the site and in our newsletter we tend to show the conservative method more as it is easier for people to use however we do often take a more aggressive approach and buy during formation occasionally.

The Aggressive Approach
As mentioned this approach involves taking an entry in to a flag trade as the flag has formed and in anticipation of a breakout. This method takes some skill and confidence and our experience has shown it is a more risky method and has a slightly lower probability of success. When you get it right the rewards can be great - get it wrong and things can go pear shaped very quickly.

In order to demonstrate the aggressive approach we will use the chart from our recent trade example MCR but approach it in a different way. Our actual example and personal trade took the most conservative approach and this was partly to do with the time at which we found it. However we had anticipated the breakout when we looked at it on Friday.

Taking an aggressive entry requires very good timing and nerves of steel. The reason for this is that we need to be sure that the flag has formed. We then need to ensure we do not enter too early and then need to try and time the breakout. It would not have been so difficult in the example that we use however as we look at the plan you will see the downfall's to this approach.

Looking at the chart we would start drawing the upper trend line on the day marked 2. This is the 2nd day of the flag formation. At the same point we draw a parallel line touching the lows. With this chart it was easy, however in other charts we would place this tentatively. We would not wait for confirmation of this line with this pattern (this brings up an interesting point which we will cover later).

On day 3 we would be reasonably confident that the pattern had formed - it could still fail - these patterns are no guarantee that they'll hit the target, but we are comfortable that the flag had formed. It would be a very aggressive entry on this day however as prices closed in the middle of the day and did not even get near the top trend line. Some traders would be happy to enter however so let's look at the issues with an entry at this point. We take an entry at $0.70 on the close.

While it is not hard to take an entry on the close we need to define our position size for risk management. This sounds easy at first, the pattern is invalidated by a close below the lower trendline so this is our exit point - this is fine except the trendline is going down - this means our stop is LOWERING. This is a dangerous way to trade and for almost all cases we would NEVER recommend that you lower a stop. So where do we place the stop. If we place it below the value of tomorrow's trendline and price does not break out tomorrow then we need to lower the stop to the value below the line on the next day and so on. Our stop could be lowered 3 or 4 times before it takes out or we are stopped out. Let's look at what happens in this case.

For the example we use the entry at $0.70 and a stop below the next day's trendline. The trendline itself is at $0.67 so if we say $0.66 then our position size available is only 11000 shares.(based on maximum risk and trade size of $500 and $10,000 respectively). This costs $7,729.95 and puts at risk $499.90. Based on the target the risk reward is 2.74:1. Readers know that we prefer 3:1.

On the following day price does not break out and the stop needs to be lowered. It needs to be lowered because the lower trendlines value is $0.66 and so a close at this level does not render the pattern invalid. To lower the stop to $0.65 then puts $627.90 at risk suddenly going over our maximum allowable limit. The fact more is now at risk means the risk reward ratio has dropped to 2.84:1. If this kept going on for a few days then we are just putting more and more money at risk.

This illustrates the importance of two topics when dealing with flags. The first is we need to try and anticipate the breakout and not just enter and wait - this takes a lot of skill and many hours of watching charts on an intraday and end-of-day basis. The second is that the steeper the flag the quicker this stop will lower and the more difficult it is to take an aggressive entry.

By the time we get to day 4 the prices gap up on the open and hit the value of the downward trendline. Despite the small range of this day the gap up on the open is a bullish signal and a sign for some aggressive traders to enter using market depth for timing. Entry on this day was available at $0.71 and $0.72. Again however we still face the dilemma of the lowering stop and this time the stop is further away. This again illustrates the point we make in that timing is the successful trade off between risk reward and probability. The chance of continuation would seem slightly more probable on this day however the risk reward is reduced.

Based on the exit below tomorrow's trendline value we can only take a $6,279 position, which still has $500 at risk. This also puts the risk reward ratio back to 2:1.

I have heard the solution that a flag pattern should only take up to 10 days to form and 3 or 4 have gone so why not base the stop on the value of the trendline in another 6 days, the floor is going to be the severely reduced risk reward. Even based on our most aggressive entry point at $0.70 if we had used the trendline value of 6 days away the position would have been reduced to $5,134.50 and the best risk reward down to 1.79:1. If it did take six days the risk reward ratio would be less, as the target would also be reduced. This is an unacceptable alternative in most cases. We believe this should be assessed based on the risk reward calculations.

Aggressive entries are best made in slower sloping triangles when we are as sure as can be of a breakout. At this point it is often better to use a tight stop. This does reduce the success rate of the pattern as a tight stop can mean being whipsawed out of the trade, however our experience has been that during a bullish market the small reduction in the success rate is well worth the increased risk reward ratio.

We could have just shown you a historical example and said "An aggressive entry here with this position size leads to this profit". The reason we show the position sizing and initial stop methods however is because these are the issues that we actually face with a trade as opposed to just looking at one. When you do these calculations you often realise the trade offered a terrible risk reward ratio and had we tried it a number of times we may have lost money.

The Conservative Entry Approach
We do not intend to go over this in detail as the trade example in section 3 based on a real time example shows exactly how we would have traded it given the point we found the pattern. What we will do however is recap the fact that the conservative approach is entering as price break's out above the pattern. This often relies on a contingent order or full time access to the market to get in. It is essential however to have a limit as to how far prices may be chased and this limit should be based on the risk reward ratio available.

As we are waiting for the breakout it is more acceptable to run a tighter stop as the pattern has a relatively high probability of not coming too far back in to the flag (or even falling back in at all) - but it can happen. We would refer readers to last weeks update and the previous week's trade example for an example of a conservative entry.

We do not have a preference for either, as long as we are happy with the risk reward ratio. We do believe that waiting for the breakout is the easiest and least stressful option.

Summary
Over the last few weeks we covered what a flag pattern was, the rules regarding its construction, how we measure targets and last but not least the management of these trades once opened. This week we will wrap up our recent articles on trading flag patterns with a quick summary of some of the more important points that we covered. These have been set out below and then we have included a few examples of flag patterns and how they appear "in the wild".

    Pole Construction Rules:
  • Created by 2 - 6 days of continuous, strong, upward price action (can take up to 10 days although this is less common).
  • Should have no major retracement or "kinks".
  • Flag pole may have upward gaps
  • Down days tend to be small in range and have little effect on the upward momentum
    Flag Construction Rules:
  • The flag is made of two parallel lines
  • The top line is placed first using two points. The bottom line is placed tentatively to this using only one point.
  • Flag should be at the top of the flag pole
  • Generally take less than 6 days to form. Can take longer but this is the exception. If they do, they generally tend to be weaker patterns.
    Height Measurement Rules:
  • The flag pole begins on the first strong up day.
  • If there is a gap on the first day then the gap is included in the height of the flag pole.
  • The height of the flag is projected from the upper line of the flag at the point of breakout to measure the target.

    Invalidation Rules:
  • Close below lower trendline of flag
  • Price falls below the low point of the flag after the breakout.
  • Some people use the 50% mark of a candle or the flag pole.
    Entry / Initial / Management Stop Rules:
  • Aggressive entry requires anticipating the breakout.
  • Conservative entry generally easier (on break)
  • Place initial stop one tick below the lowest point in the flag or chose an appropriate chart based stop.
  • Sell as target is hit. Note however: flag patterns in a bullish market often exceed their target.
  • Sell on lack of momentum and upward movement after breakout.
    Methods of Finding Flag Patters:
  • Most methods revolve around looking for a flag pole. Eg. Price Volume breakout
  • Eyeball search of folder(s) looking for the flag pole and studying subsequent price action
  • Specific Pattern Detection programs such as "CPFinder"

The picture shown above for the setting of targets is just a lovely little flag pattern isn't it? Unfortunately not all patterns are this easy to see. We have included a few more flag patterns from actual charts so that you may see how they differ from the above text book example. The charts on the left show the pattern as it developed and the charts on the right show the subsequent price action. These were all recent opportunities (at the time of writing) and two are taken from actual trades.

  |  

First Published: 20 May 2005 - Copyright © Jason Mitchell

This document is copyright. This document, in part or whole, may not be reproduced or transmitted in any form or by any means, electronic, mechanical, photocopying, recording, scanning or otherwise without prior written permission. This article needs to be viewed as educational reference only. It is not intended, nor is it to be regarded, as investment/securities advice or any other advice.