Quote from TriPack:
This isn't in reply to anyone in particular, just thought I'd add my perspective since this thread is on candlestick patterns, in particular *intraday* patterns.
Candlestick patterns are generated based on the particular timeframe you are viewing. If you are looking at 5 minute bars, then you will get bars that show the activity during that time. Let's say you are looking at the bar that starts at 9:30 AM and ends at 9:34:59. You may have a particular candlestick pattern on that bar. But suppose your start time is altered by a minute in either direction. Is the doji in the first instance still a doji in the second or third?
With candlesticks, we get the problem of the timeframe creating the pattern in some instances. Why is this a problem? Is the pattern real or significant? If it is a significant pattern then in my view it should appear in multiple shifts of the timeframe you wish to view. For instance it would appear in 5 minute candlesticks starting at 9:28, 9:30 and 9:32. In other words the pattern would not be affected by the particular starting time of the bar. The same problem exists for candlesticks drawn on tick charts. One tick either way could mean the difference between a certain pattern appearing or not appearing.
I found when I looked at it this way and started analyzing candlestick patterns on multiple shifted timeframes (5 minute candles, starting 2 minutes apart), or even on multiple timeframes (like 5 min, 7 min, 12 min) that what looked like a pattern in most cases didn't show up as a pattern in other cases.
And when you start looking at multiple timeframes or start shifting the start time on the timeframe by a minute or two, you are really just doing the same thing as a moving average. So in other words I don't see any great benefit to using the patterns over what a moving average gives you. And I'm not that particularly fond of moving averages because of the time lag involved.
Anyway, there are folks who use candles well and profitably and I salute them, and I'm sure they will take issue with what I have said. I understand that daily candle patterns are more reliable than intraday patterns and so none of my remarks above were aimed at daily candlestick analysis.
I do look at my charts as candlesticks, not for the patterns, but for the easier view of the OHLC relationship.
Thanks TriPack,
Interesting view.
However, what you've describe I don't know anybody that has tried to use it that way...unless I've misunderstood what you were saying.
Your the first trader I've ever met that's tried analyzing candlesticks this way.
From what I understood...your saying if your time interval is
altered by a minute in either direction...
sounds like your saying your using a different chart interval than your 5min chart example.
Thus, if I
initially started using a 10min chart interval starting at 0930am...that means I will see a completed interval at 0940, 0950, 1000, 1010am et cetera...
However...if I decide to analyze the candlestick formation that's forming at 0939am (1min prior to 0940am for the 10min chart interval)...
(altered or shifted)
that means I'm using a 9min chart interval and not a 10min chart interval.
Therefore...chart intervals I should be analyzing are 0930, 0939, 0948, 0957 et cetera
Isn't this like saying your going to get a different interpretation when using MACD (12,26,9) in comparison to using MACD (5,15, 7)?
Of course the info will be different.
Also, if your using a software program that starts at a different time...for example...starts at 0934am instead of 0930am and you want to use a 5min chart...
means your first 5min chart interval will complete itself at 0939am est...
Its time to get a new charting program because that's not normal if your software starts at 0934am (displaying prices) instead of at 0930am est (your missing 4 mins worth of data)...
(a software that doesn't collect history...starts collecting data based on the moment you log on)
especially if you want to properly analyze candlesticks.
Simply...I've never heard of anybody viewing...for example...a 5min chart interval in 2min sections...
that means your attempting to analyze a 5min chart and split the time frames into 2min, 2min and 1min...
Candlesticks were not designed via the method you've attempted to use them in the past...
I've never read anything like your description before.
Thus, if your going to use a 5min chart and you want to analyze lesser intervals within that 5min chart...
analyze each interval equally...not unequally...
Therefore, if I was going to break down a 5min chart into equal intervals...
I'll view a 1min chart...not a 2min chart because the last interval (1min) will not carry the same weight as the prior two intervals (2min).
Please correct me if I misunderstood what you were explaining.
Also...I've done the stats work based on my personal use of daily candlesticks and intraday candlesticks...
>>> There are candlesticks on the daily charts that are
less reliable than intraday patterns...
>>> There's also candlesticks on the intraday charts that are
less reliable than daily chart candlesticks...
Simply...I guess it comes down to what candlestick patterns you tend to trade.
For example...based on my trade methodology...
Bearish Engulfing is more reliable on intraday charts than daily charts while trading the Eminis.
Bullish Engulfing is more reliable on daily charts than intraday charts while trading the Eminis...
Lets breakdown my use of Bullish Engulfing a little further...
On the daily charts where its very reliable for me...I don't need a confirmation via a divergence signal...
However...because its not as reliable enough for me on intraday charts...I need a confirmation via a divergence signal for that Bullish Engulfing...
Note: There are two types of Bullish Engulfing I trade...regardless if its on a daily chart or intraday chart...one is high probability for profits and the other is low probability for profits...there are 3 other types of Bullish Engulfing I wouldn't touch with a ten foot pole.
all based on my trade methodology.
Your trade methodology will obviously be different from my own...especially if we are using a different confirmation system...
Thus, our results more likely than not...will differ greatly.
Something I tell traders often...
we all use different trade methodologies...
For example...to a particular trader that trades pullbacks...he only looks for Ascending Triangle formations...
To another trader that trades pullbacks...he only looks for Flag formations...
Imagine if one of them say that pullbacks don't work while not disclosing their trade methodology for trading price pullbacks.
Simply...something works or doesn't work solely based on our trade methodology or the difference in such.
Here's something that happened between my spouse and I about a month ago while on a picnic...
She was trying to show me that the cloud above looked like a telephone...
I couldn't see it...what I saw was a shade or blind you put on a window.
Once again...I don't think candlesticks were designed to be used via
shifted (maybe not a good word to explain this) interval analysis that produces unequal chart intervals.
I know a guy that uses 11min chart interval as a price action only trader (no indicators)...
he told me a few weeks ago that the volume in the last interval is odd in comparison to the volume in the last interval while looking at a 15min chart interval in which the Eminis closes at 4:15pm est
P.S. I think you should try analyzing candlesticks via equal intervals...your analysis should improve greatly (maybe not to where you want it to be at) in comparison to unequal (altered or shifted) chart intervals.
NihabaAshi