Flag Pattern on Gold Chart — Good News for Gold Bulls
You can see a flag chart pattern on the oil chart (4-hour time frame). The chart is named so obviously because it resembles a flag. It consists of a ”pole”, marked by the vertical yellow line on the chart, and a ”flag”, or the two parallel lines on the chart. The ”pole” marks a strong move of prices, while the ”flag” shows a small correction of prices in a small range (marked by the ”flag” or the parallel line on the chart) as prices take a breather after an initial strong move. This chart signals that the correction is a temporary pause before the prices will move in the direction of the initial strong move, in this case the next wave will be bullish. It’d be welcomed by gold bulls, who may be quite disheartened after the weak January performance of the precious metal. The price should rally after it breaks through the upper level of the ”pole”. Click the image to enlarge it to a
If you have any questions or comments regarding this chart pattern for the gold, please, feel free to reply below.
I’m always puzzled when people use chart theory to forecast. Perhaps I just don’t understand it, but it reminds me of pareidolia, the psychological phenomenon where people see patterns in clouds and other randomly generated sounds and images.
What percentage of the time would you say this pattern holds? Every time? Do you have any evidence to support this claim? If you tried to fit a trend line to the graph above, it would have a negative slope. And this is bullish? I’m sure if you expanded the time frame that could change, but still.
[Reply]
enivid Reply:
February 6th, 2011 at 6:47 pm
According to the statistical research performed by Thomas Bulkowski in his Encyclopedia of Chart Patterns, Flag pattern has about 13% failure rate.
[Reply]
Jake Reply:
February 6th, 2011 at 7:20 pm
Over what time frame is the flag pole defined? For how long must a bullish run occur after it in order for it to qualify as a success? For what time frame must the flag pole appear in order for it to qualify as a flag pole? In the graph above, we are talking about a 32 hour period (note, I am not referring to 32 trading hours). Is that long enough? How large of a change do we need? We could change the Y axis and the flag would look much differently. The whole concept needs to be better defined before something as exact as a percentage “failure rate” can be calculated. The same is true for most other chart theories. They require too much hand waving.
The beauty of statistics isn’t reporting means or correlations, it is determining what is a statistically significant relationship and what is not. Bulkowski neglects to perform even the most basic tests of statistical significance, as do most chartologists, potentially because it would show that such theories do not stand up to rigorous statistical analysis.
[Reply]
enivid Reply:
February 6th, 2011 at 9:15 pm
If I understand you correctly by “flag pole” you mean the whole flag pattern, not just its pole, right? Or otherwise it doesn’t make much sense…
Bulkowski tests the patterns on daily timeframe, but in highly liquid full-time markets it’s used on any timeframe.
Bulkowski lists 19% average rise after the bullish flag pattern. With the duration of about 62 days.
From a few days to 3 months according to B. But that depends on the chart timeframe used.
Why not?
What do you mean?
Yeah. Still the pattern would be recognizable. It would even look better if the chart is compressed on Y axis.
Bulkowski tested the pattern with automated software – clearly defining the needed parameters. Do you think he did it all manually?! That’s quite crazy. Who in their sane mind would even try to test the patterns with “hand waving”? Unless we are talking about 5-10 occurrences. And in case of the flag pattern, it’s not like that.
But that’s what he does. Defines a pattern, runs a test and registers the outcomes.
Then their studies could be easily disproved with a statistical study that shows no real value behind the patterns. Currently, I see the study that proves the patterns (by the way, not all of them, in B’s book, he lists some patterns as insignificant), and I don’t see a counter-study for that.
A side note about the chart analysis. It works not because I (or someone else) see there some magical patterns, but, first, because many patterns signify a particular traders’ behavior (like a correction/consolidation after a strong trend in case of the Flag chart pattern) and, second, because many patterns became so popular that they work simply. as the self-fulfilling prophecies.
@ Jake
*facepalm*
[Reply]