JustData.com.au » Journals » Stock Investor » Jason Mitchell » Constructing and Trading Flag Patterns

Over the last few weeks we covered what a flag pattern was, the rules regarding its construction and how I measure targets. In the original articles from my newsletter I showed a real time example of a flag trade so that readers could see our management methods. I have included the notes of this trade directly from the newsletter as an example of a successful pattern.

When they work they are easy to trade. This is why I have also included an example of a flag trade that did not perform and the subsequent management of this. It should be noted that these notes were prepared between the 12th and 26th February 2005.

Looking at the XAO review we have decided not to add any position trades to our case studies or web site portfolio again this week. While this may surprise the newer readers, long time readers have probably already guessed this would be the case and have realised that this is partly why there have been no new charts on the web site lately. (We are not suggesting we are bearish yet).

This week we have decided to look at a shorter term trade using a pattern as our setup and trigger as this is in line with our own trading strategy.

The way in which we found this was very simple. We simply chose a letter (completely at random) and quickly scanned through each stock in the database. We were specifically scanning for patterns and in the first pass we looked for ascending triangles. We did not see any of these and so went through again looking for flag patterns. We found a stock that was developing a possible flag pattern and have decided to show how we may analyse this type of opportunity.

The above chart shows the stock price as it was at the close on Friday going back to April last year. The stock recently broke away from its base in a sharp upward move on heavy volume and has pulled back slightly on lower volume. The tops of each bar in the pull back have been in a straight line allowing us to plot a short-term downward trendline. Using this we place a parallel line and note that it is touching the bottom of price action. This is a short-term channel of a few days to maybe a week and creates the flag. The strong rise prior to this creates the flag pole and combined we have a flag pattern forming.

Looking at the JICD (in the previous chart) we can see that it’s peaks have been getting higher, it crossed zero on the breakout and has now returned to the trendline and has turned up and may be about to rebound away again. This is the type of setup we like to see with the pattern. We would also accept a rebound from the zero level however the trendline is more bullish.

I have taken the height of the flag pole and projected the height of this above the upper line of the flag. This was projected based on a break on Monday and will be re-adjusted if the break does not come at this time. The value of the line on Monday will be $0.71. I have drawn a line across the value of the lower trendline on Monday and a move below this invalidates the pattern.

We will buy as soon as possible on a break above the upper trend line. Using a contingent order at $0.72 on Monday could get us in – we will buy at $0.72 or $0.73 and will adjust these values each day there is no breakout. I can not chase prices any higher based on my desired risk reward strategies. As I trade full time I place my orders manually.

The above spreadsheet shows the position sizing for the trade. We expect prices will move quickly once they break above the pattern and as such we have used a very tight stop to get the biggest position size and the best risk reward ratio. We understand that this may work against us and that using a tight stop can actually reduce the success rate of a system. We are happy to take this risk however due to the high probability of the pattern.

If we get in at our preferred level of $0.72 then the trade offers a risk reward ratio of 4.34:1, which is acceptable. If the trade does not work we will have $336.00 at risk based on the stop and this could be larger quite easily. The possible profit that is available $1458.65 based on the target.

Now, this is where low price shares kill traders - if we need to enter at $0.73 (only 1 cent difference) then this puts $468.38 at risk (an extra 39.4% more than the purchase at $0.72). Based on the target the reward will be $1,301.70, this puts the risk reward ratio back to 2.78:1. We prefer trades that give us a 3:1 ratio although due to the probability of the pattern we will still add the position to the case study at $0.73 if need be. The plan is:

The following notes are the update on this trade for the Following Week.

Last week we decided to look at a shorter term trade using a pattern as our setup and trigger as this was in line with our own strategy. We elected to use a pattern trade and came across a flag pattern that was well constructed and offered an adequate risk reward ratio.

The chart above shows the successful completion of this pattern as it exceeded its target on Friday. This article shows the management of the trade.

On the Monday we added the trade at our preferred price of $0.72 just after the open, we believe better trading could have gotten an entry at $0.71 as the market settled down, however for the sake of the article we take an entry at $0.72. The price did not shoot up on this day until just after lunch time so there was plenty of time to enter with this particular trade. As soon as the buyers stepped in however the price shot to $0.76 and rose from here. The last chance to enter was on Monday morning before the big buyers moved in.

Price continued to move up on the Wednesday to hit the target of $0.83. It was not possible to exit at this level on the day and the highest available exit was at $0.82. The huge number of buyers that had moved in however on market depth showed that there was likely to be short term (intraday) support at the $0.82 level. Using this knowledge (and knowledge of an increase in the dividend) we held off exiting until the Friday and took advantage of the price spike to exit just above the target at $0.84. We feel this was easily achievable on the day and prices in fact went to $0.85.

There was no need to confirm this exit strategy with any indicators as it was a set target. The trade returns $1,596.70 in one week which is equivalent to 16.66% and the trade ended up showing a risk reward ratio of 4.75:1. In a less bullish market we would have taken an exit at $0.82 on the previous day.

Interestingly there was an announcement made on the Monday and on the Thursday in this stock. The first was an announcement regarding a high grade drill intersection and the other was an increase to the dividends that were being paid after the release of the Half Yearly report. Without any knowledge of these announcements we have taken a good chunk of profit from their effect on the market.

As can be seen the management of this trade was relatively easy as it did what it was supposed to. This is an example of a conservative entry as we waited for the break above the pattern. The time that we can get in to trouble on these patterns with a conservative entry is when they fail to move ahead as expected or fall below our stops and we do not act.

Any one can manage a flag pattern that hit’s the target but all successful traders will tell you it’s about how you manage losses or things that go against the plan that makes a trader successful. As such you will rarely see much on failed trades as people generally don’t want to discuss these. The rest of this article is devoted to showing management of a pattern that has not worked.

Managing a Flag Pattern that hasn't worked
The following chart shows a flag pattern and its subsequent failure to reach its target. Going back to last week’s article we can note a few points of weakness in this flag as it was forming which hopefully would have kept our readers out. The first is the fact that the flag has started just below the top of the pole. The higher the flag the stronger and more likely the pattern is to reach and exceed its target.

This flag started about 85% of the way up the pole, but still this was not up the top. The second issue is the angle of the flag which we estimated to be about 25 degrees. As we noted last week we do not generally measure the angle of the flag however we do take in to account if it is overly steep. Leon Wilson suggests that any flag with an angle of less than 45 degrees is less likely to hit its target. Combine this with the fact the flag started at 85% of the pole height and we should be aware of the reduced likelihood of this pattern reaching its target. This turned out to be accurate in this case.

Before we discuss possible management methods we should remind readers that the flag pattern is a short term trading technique. We are not in this for the long haul – we are expecting momentum to drive prices quickly to their target as they did in the recent trade example of MCR. A flag trade usually takes between 2 – 6 days to hit it’s target after it has broken out although they can take up to 10 and can reach it or exceed it in as little as one day. So how could we have managed it had we entered? For the sake of the exercise let’s say we took a more conservative entry on the breakout. It doesn’t really matter where we take the entry as in this exercise we aim to show the management if it fails to reach its target.

The first thing we should understand about this pattern as I said in the first article is – we are not trading the price action but the excitement of a crowd. We expect the excitement of the crowd to push prices up in a reasonably quick timeframe. We call this type of move a rally and it is produced by an increasing momentum in the share price. We do have tools that are designed to help us measure the strength of this momentum and there are a number of momentum indicators out there to choose from. We have a preference for the JICD, the Williams %R or the RSI. There is no need to use all of these however and one should really do the job.

The chart below shows the breakout point and our entry (circled).

It also shows our chosen initial stop at $1.03 indicated by the thick red line. This is chosen as price has not closed below this level and should the excitement of the crowd continue we would expect momentum to take over and for prices to stay above this. If they do not we can accept we were wrong in our analysis and get out. Being wrong is going to happen – this doesn’t matter too much. It is being able to judge how we will know we are wrong and then acting on this that matters. Being wrong happens to every trader.

As we have mentioned waiting for the breakout sometimes gives us a better chance of getting a good position size. We show the position sizing spreadsheet (available from our members section) that we use for most of our trading and an entry on the open as prices break out (this is a short term trade – timing is everything). This entry allows us to take 9466 shares at a cost of $9969.25. Based on the stop loss level at $1.02 (just under the stop at $1.03) we have 343.90 at risk. Using the target this gives us a slightly better than 3:1 risk reward ratio. If we used the figure of $1.01 however the risk reward ratio drops to 2.45:1 (this is not shown).

So far we have the following parts to our plan:

Unfortunately some people leave it at this point and are happy to enter the trade. We however have only looked at 2 out of the 3 possible outcomes. The first outcome is price falls back and we exit using our initial stop. This is easy and MUST be followed. It’s going to happen and that does not on its own mean you are a bad trader. A number of unsuccessful trades however would make you re-think your strategy on flags.

The second outcome is that price reaches the target; this has been covered and is easy. The third outcome is the price goes sideways or moves ahead but fails to reach the target. For every trade we need to develop a plan for the three possibilities of price action, which are:

Earlier in this article (and in the first) we said that we feel the pattern shows excitement and that we are trading this excitement. If the excitement slows down our theory is that we would also expect to see momentum slow down. Combining this with our expectation that price will rise fairly quickly, the first thing we could do is to say we will re-assess the position if it has not reached its target within 3 days (short term trades call for close monitoring). At this point if we see a bearish divergence between our chosen momentum indicator and price action we could take this as a sign that momentum (and therefore the excitement) has slowed. We are trading the excitement so if this happens it is probably time to get out.

Seeing as we do not trade knowing the data on the right hand side of the chart, the easiest way to show this is by a day-by-day blow of how we may have managed it using our methods and ideas. This will show us the chart as it would have appeared at the point of a decision and not how it appeared in hindsight.

Day 2 (Tuesday)
The Day After the breakout prices did move ahead however we notice that the close is at about the midway point of the day. This would give us an upward shadow. The JICD is rising showing momentum continues to rise.

There is no need for any concern on this day other than to note the fact price closed at about the halfway mark.

Day 3 (Wednesday)
The next day is very short in its range and the open and close are at the same level showing a lack of dominance by either group. Combined with the fact that that the close was the same as the previous days close we can see there had been a lack of short term momentum. There is no need to be overly concerned on this day although we again note that the momentum is not very strong.

Day 4 (Thursday)
By day 4 we are beginning to become a little more worried as price has failed to get very far in the four days and the close has been at the same level for three days. There is no momentum here at the moment and we have placed a very tentative short-term trendline that points downwards on price action. On the JICD we have placed a horizontal line at the last peak. It is too early to plot a divergence as the 2nd peak has not been formed but the line shows that the indicator has failed to make a new high at this stage.

It is possible that some traders would be considering an exit at this point as price fails to move ahead. This is a valid strategy however I would be inclined to give it a little bit more time seeing as it has not retraced. An exit on the close of this day would have given the trader $318.74.

Day 5 (Friday)
Now the closing price has fallen and the short term downward trendline has had its third touch. Further to this a new low for the week had been created. The JICD indicator had shown a divergence as the peak had been formed (JICD now headed down).

This is the point that I would have considered taking an exit (if I hadn’t already). There is enough evidence to show that the excitement had died down and five days is a fairly long time in a flag pattern. An exit on the close at $1.08 would mean a small profit of $224.08 after brokerage.

Remember we are trading the excitement and after 3 days of nothing and the close actually getting lower we were happy that the excitement had seemed to disappear.

Day 6
Had I stayed in the stock until this point I would have decided now that price and momentum were well and truly gone. The close had dropped back again and the range of the day was very short again showing that there was no dominant force at this point. Obviously the divergence in the indicator is still visible and is even more pronounced now as the JICD continued to fall.

An exit on this close would have meant the trade did not lose money and we would have taken $129.42.

Day 7
We have now had 7 days without one day of upward momentum in the last 4 days. This 5th day that lacked any major movement would have been a definite sign that this pattern has failed. It has been given every chance by this stage and has literally gone nowhere. Price still failed to close above the downward trendline.

Patient trading may have been rewarded here with a small profit of $224.08 on an exit on the close. Often in these patterns patience is not always the greatest thing. I would have exited on day 5 or 6 (and possibly earlier if better opportunities had presented themselves.

I have also included the following chart, which illustrates other momentum indicators showing a divergence by the time we suggested we would take an exit. We usually use the JICD however appreciate that readers may prefer to use other indicators. We have shown the Chande Momentum Indicator, the Williams %R and the RSI. These are all shown with their 14 period versions. We would not recommend using all of these and skill with one should be more than enough.

Before we leave it for this week, I thought I’d show another failed flag pattern. One reason is to show some issues with the flag and the other is to make sure readers know these patterns do not lead to definite results. They do put the balance of probability on our side if well constructed, although this is not.

This chart shows a stock where there has been a flagpole form. There is a somewhat definable flag created as price retraces. The major issue is the speed of, and the gaps created, during the retracement. The flag starts below the top of the pole and the angle is too severe for my liking. This is not a high probability pattern.

The subsequent price action did experience a break out however price pulled back from this level and it is unlikely that an exit would have been achievable at a profit.

This article has been taken from a series on 'Trading Flag Patterns' by private trader Jason Mitchell and his newsletter the StarTrader Report. In further articles Jason covers some of the more important points on trading flag patterns including how to find flag opportunities and how to enter in to a flag trade (looking at both an aggressive and conservative entry methods). He also shows a selection of actual examples of flag trades and patterns to help people recognise the differing ways in which these may appear on a chart in real time. During this series of articles Jason showed two real time flag trade’s (i.e. chart shown before the entry) being MCR and CEY. These returned 16% and 14.57% respectively – both within 7 days.

Raw Price Signal:	Flag Pattern – Looking For Break
Confirmation:	JICD trending up and possible rebound off trendline (5 periods) Maximum position size available and risk reward of 4.3:1 on preferred entry Sector Trending Up and outperforming the general index.
Initial Stop:	$0.70
Position Size:	13,805 @ $0.72 for a cost of $9970.05 or 13,616 @ $0.73 for a cost of $9969.63
Risk:	$336.00 (plus) on entry at $0.72 $468.39 (plus) on entry at $0.73
Management Tools:	JICD (5 Period) to measure momentum if prices fail to reach target. Candlestick Patterns.
Plan:	Buy on the breakout from the flag pattern using a tight stop for risk reward. Sell as prices hit the target. If pattern fails to hit target within 5 days re-assess the stock and position. Sell on weakness of the JICD (eg. Divergence)

Constructing and TradingFlag Patterns - Part 3

Constructing and Trading
Flag Patterns - Part 3