[guruji]

A morning star doji is a 3 bar pattern with bar1 (long and dark), bar 2 (doji) gaps down and bar3 (long white) gaps up.

I did a Back test over SP500 stocks over 15 years and could not find this pattern.

Have you seen it

[wrbtrader]

Originally Posted by guruji »

A morning star doji is a 3 bar pattern with bar1 (long and dark), bar 2 (doji) gaps down and bar3 (long white) gaps up.

I did a back test over SP500 stocks over 15 years and could not find this pattern.

Have you seen it

I just went to your blog and it saids you have found 10 examples of this pattern while you said in this message post you didn't find any. Which statement is correct ???

Regardless, there are several variations of this pattern and your requirement of the high of the doji be less than the lows of both interval 1 and interval 3 is an extremely strick variation.

I also want to remind you of an earlier conversation we had about Japanese Candlestick patterns. Each pattern has there own different variations. Thus, it's absolutely critical that if you're going to backtest, test any particular candlestick pattern...you test all the known variations you can find or else your testing results will be flawed and may be different in comparison to what traders are actually using in real money trading.

For example, another variation of the Morning Star Doji pattern is that the high of the doji is less than the close of interval 1 (dark line) and less than the open of interval 2 (white line). Simply, this particular version doesn't use the Lows as a criteria as a measurement against the high price of the doji. However, in this particular version the lows of the dark line and white line are not less than the close=open of the doji.

Something else you may want to consider. You're testing your particular strict variation on the S&P 500...you testing it on the Index only or have you included results of the SPY and Emini ES. Also, I'm going to assume you've been testing this pattern on multiple different time frames (monthly, weekly, daily and intraday charts) to have a decent sample amount versus testing on one time frame alone.

[guruji]

A few comments:

- I am being very generous when I state on my blog that there are less than 10 instances on this strict definition of the pattern. I have NOT FOUND any instances of this pattern
- The back testing was NOT done on the SP500 index. It was done all the 500 stocks that made up the SP500 index in 2010.
- Yes, I agree this is a very strict interpretation on the morning star doji. I plan to back test a more relaxed interpretation soon.
- I find is interesting that particular pattern that is well written about in literature actually does not exist

Originally Posted by wrbtrader »

I just went to your blog and it saids you have found 10 examples of this pattern while you said in this message post you didn't find any. Which statement is correct ???

Regardless, there are several variations of this pattern and your requirement of the high of the doji be less than the lows of both interval 1 and interval 3 is an extremely strick variation.

I also want to remind you of an earlier conversation we had about Japanese Candlestick patterns. Each pattern has there own different variations. Thus, it's absolutely critical that if you're going to backtest, test any particular candlestick pattern...you test all the known variations you can find or else your testing results will be flawed and may be different in comparison to what traders are actually using in real money trading.

For example, another variation of the Morning Star Doji pattern is that the high of the doji is less than the close of interval 1 (dark line) and less than the open of interval 2 (white line). Simply, this particular version doesn't use the Lows as a criteria as a measurement against the high price of the doji. However, in this particular version the lows of the dark line and white line are not less than the close=open of the doji.

Something else you may want to consider. You're testing your particular strict variation on the S&P 500...you testing it on the Index only or have you included results of the SPY and Emini ES. Also, I'm going to assume you've been testing this pattern on multiple different time frames (monthly, weekly, daily and intraday charts) to have a decent sample amount versus testing on one time frame alone.

[wrbtrader]

Hi guruji,

I don't follow stocks so I really can't comment upon any of them nor do I know what time frames you're testing on. However, I do trade futures and see the pattern on intraday charts of futures and forex trading instruments.

Also, most of the documentations I've seen about the pattern involved futures or forex. In fact, I can't even remember the last time I've seen chart examples of this particular pattern via stocks.

In addition, most of the literature I've seen doesn't use the strict variation that you've mention in this thread. I've seen at least 7 different variations of this particular pattern discussed in different literature (books, videos, websites and forum discussions).

Think about it, your strict variation has the high price of the doji LESS than the lows of the interval 1 (dark line) and interval 3 (white line). Basically you're talking about the entire price action range of the doji interval gapping below the entire price action range of the black line and white line. That's a very rare price action and a particular variation I don't see discussed too often. The few times I've seen such occurred within the context of key market news causing the entire range of the doji to gap down.

Simply, the Morning Star Doji is a commonly discussed pattern but some of the stricter variations are not commonly discussed in any literature I've read.

guruji, out of curiosity, have you compared your results of any particular type of candlestick pattern to some of the candlestick websites that post daily scan results of stocks (e.g. stockcharts.com) to determine different variations being used that contrast your own variation ???

Originally Posted by guruji »

A few comments:

- I am being very generous when I state on my blog that there are less than 10 instances on this strict definition of the pattern. I have NOT FOUND any instances of this pattern
- The back testing was NOT done on the SP500 index. It was done all the 500 stocks that made up the SP500 index in 2010.
- Yes, I agree this is a very strict interpretation on the morning star doji. I plan to back test a more relaxed interpretation soon.
- I find is interesting that particular pattern that is well written about in literature actually does not exist

[Corey]

def from Greg Morris's book

Look for a tall black candle in a downward price trend. The next day, a doji appears and its body gaps below the prior candle’s body. The final day is a tall white candle whose body gaps above the doji’s.

I think your interpretation of the pattern picture is too literally. Here is how it looks in real life.

[wrbtrader]

Originally Posted by Corey »

def from Greg Morris's book

Look for a tall black candle in a downward price trend. The next day, a doji appears and its body gaps below the prior candle’s body. The final day is a tall white candle whose body gaps above the doji’s.

I think your interpretation of the pattern picture is too literally. Here is how it looks in real life.

Thanks Corey

guruji is still testing a rare variation of the Morning Doji Star as of January 27th 2011 as shown at his blog link below instead of testing commonly appearing variations of the Morning Doji Star.

http://blog.kilotrader.com/2011/01/m...star.html#more

However, he did say the following here at Traderslaboratory (December 2010) that implies he does know he's testing a rare variation.

Yes, I agree this is a very strict interpretation on the morning star doji. I plan to back test a more relaxed interpretation soon.

Yet, he continues testing the rare variation. Also, his interpretation of a Doji is incorrect as shown at the above link via many of the images having Hammer Lines instead of Doji Lines. Thus, it's possible he has included Hammer Lines as a possible variation of the Morning Star Doji. Yet, if that's the case, I strongly disagree because Hammer Lines are not a variation classification of the Morning Doji Star. The other possibility is that he accidentally posted Hammer charts with the Morning Doji Star charts in his blog. Regardless, he did state early upon his arrival here at Traderslaboratory he's new to Japanese Candlestick patterns and hopefully he'll develop a better understanding of the definitions along with revealing more information about the "trade management" after entry.

In my opinion, the trade management after entry has an enormous impact on backtesting results, simulator results and real money trading results of Japanese Candlestick patterns. I've seen results vary from 30% to 80% of a specific candlestick variation just by changing trade management rules after entry.