Registered User
Joined: 4/18/2005
Posts: 4,090
|
I recently requested a book fromthe library and discovered that it is not available. But I began to wonder about the various probabilities of candlestick patterns.
I had such an enjoyable time going through charts and plotting good/bad signals that now i'm considering going further with reguard to figuereing out which candlestick patterns truely have the highest probabilities. I only know a few and base my decision of of just those... or instances that I see the intent of price but I don't know the name of the candle arangement.
so maybee we (those who use candlesticks) could discuss our favorite patterns and look into their probabilities.
I was taught engulfing white, springboard and draggons tend to be highly proffitable patterns and took it for granted. I have a feeling that spring boards are more proffitable but that is becasue they have worked well for me in the past... but there are times when they can guarentee a bad entry.
The book in question was "High proffit candlestick patterns" but not having acess to the book I have no idea as to what is actually covered. Anyone who has read this book is encouraged to add feedback.
What if interested persons each took a few of their favorite candle patterns and tested them over a few hundred charts and then we posted our results here. the rub is being consistant and not grooming that data.... which can be difficult.
We should decide on a study format and all use the same one. Perhaps it should include date ranges (to help guage effectiveness in general market conditions). Three trends.. long, intermediate, short term. Good signals/bad signals and some cut off point for teh signal... EX: if price dosent move up/dn after "x" bars it's considered false. Whipsaws might also be a good catagory, or that would go under false signals?
I'm also interested in learning new patterns that have higher probabilities for proffitable trades. perhaps they could be incorporated into my(our) various strategies or help me(us) make inprovements. It would be nice knowing that statistically this signal out performs that signal... so that when we are narrowing down decisions for entries its easier to toss our the weaker signals.
|
|
Registered User
Joined: 4/18/2005
Posts: 4,090
|
The other "RUB" is recognizing those patterns at the time.. It's easy to look back and say "taht was indeed a springboard. Allot of times candle signals arent picture perfect. but we need to start somewhere.
|
|
 Gold Customer
Joined: 11/11/2006
Posts: 359
|
Scott: You may be interested in "Candlestick Charting Explained" by Greg Morris. He did candlestick pattern analysis on 1,505,359 days of data with 148,984 candle patterns and ranked them by win%, frequency of occurance, average % gain. fHe does this for each of one day thru 7 days after signal.
In addition, he used 14.6 million days of data over a period of 13 years to rank the performance of the popular indicators, using their buy and sell signals. DM,MFI,%D.%K,MACD,ROC,etc. He gives the number, average performance, average buy and hold performance and average percent difference.
Also includes a chapter on "Candlesticks for Traders" by Ryan Litchfield.
An interesting book. I haven't found this much research on candlestick performance, Or indicator,anywhere else. You read a lot of "I believe's" and "this is a strong..." but not much hard data.
Mammon
|
|
Registered User
Joined: 6/22/2005
Posts: 81
|
Scott: the book you are looking for is available at (URL removed by Moderator) a web site run by the author.
iwanaphish
|
|
Registered User
Joined: 12/2/2004
Posts: 1,775
|
QUOTE (scottnlena) I'm also interested in learning new patterns that have higher probabilities for proffitable trades.
Scott, according to IBD founder Bill O'Neil the cup with handle pattern is by far the most successful chart pattern year after year for 50 consecutive years. He has analyzed in detail the top dozen or so stocks that made the biggest one year percentage gains for each of those 50 years and noted the type of pattern each rocketed out of. FWIW he says triangle type patterns are the least reliable. The other small handful of bullish patterns per O'Neil (but not as strong as cup with handle) are the W/double-bottom, flat base, cup without handle, and ascending. He has rather specific and detailed characteristics of what makes up a legitimate pattern, but all in plain english easily learned...no far out math needed:)
|
|
Registered User
Joined: 4/18/2005
Posts: 4,090
|
FPTRY
Thanks. That is interesting as well with the longer term pattern arangements. Is there a time period to the ideal cup and handle?
Originaly I was thinking short term trading but I want to start working on my ability to hold longer term stocks. I had discovered and read that Triangles have a 50% ish predictibility rate.
Has O'Niel written on the toppic or is his work only available through the web site?
|
|
Registered User
Joined: 10/7/2004
Posts: 2,181
|
Wm O'Neill's "The Successful Investor" should be on the shelf of anyone aspiring to proficiency in technical trading/investing. And it should be read annually.
(Just having it on your shelf will probably provide very little assistance; I think it serves quite poorly as a talisman.)
|
|
Registered User
Joined: 1/28/2005
Posts: 6,049
|
Scott
I would suggest finishing the TSV MS analysis first.
Before you get sidetracked by candles and handles.
thanks
diceman
|
|
Registered User
Joined: 12/21/2004
Posts: 902
|
QUOTE (scottnlena) FPTRY
Thanks. That is interesting as well with the longer term pattern arangements. Is there a time period to the ideal cup and handle?
Originaly I was thinking short term trading but I want to start working on my ability to hold longer term stocks. I had discovered and read that Triangles have a 50% ish predictibility rate.
Scott, I think you need to check out Encyclopedia of Chart Patterns by Thomas N. Bulkowski (2000). He set up a database using 500 stocks from 1991 to 1996 and analysed all those kind of stats about each chart pattern - generally gives about 10-15 page analysis of each pattern.
For example, he has chapters on 4 different types of triangles - Ascending, Descending, Symmetrical Tops and Symmetrical Bottoms.
For Ascending:
Short-term bullish consolidation
32% Failure rate
Average rise - 44%
Volume Trend - downward
Premature breakouts - 25%
Breakout distance to apex - 63%
Throwbacks 58%
Percentage meeting price target - 89%
So Also - Head-and-Shoulders-Tops; Triple Tops
(and that's just the first page of the chapter - he goes on for 15 more)
Hot to identify
(in-depth discussion of 8 identification characteristics)
Types of failures
General Statistics:
Number of formations in 1991-96 database - 725
Reversals 196
Consoldiations 529
Failure Rate 230 or 32%
Failure Rate if waited for upside breakout 13 or 2%
Avg Rise of Successful Formations 44%
Average decline of Failed Formations 21%
Most Likely Rise 20%
Breakout distance to apex 63%
Start to breakout - 2 months (64 days)
Lots and lots of charts, diagrams, tables, analysis out the wazoo
Once he finally rnns out of analysis, he then devotes a couple of pages to Trading Tactics for each pattern for short, intermediate, and long-term traders
Finally he closes out each chapter with a sample trade on that pattern.
And it goes on for 650+ pages from Broadening Bottoms to Weekly Reversals, Upside.
I know you started off this thread looking for probability stats of candlestick formations, but if you're also interested in chart formations, it sounds like this is something that you might be interested in.
I also would recommend a book that I'm currently reading called Trading in the Zone by Mark Douglas, which is about the psychology of trading ("Master the market with confidence, discipline and a winning attitude"). It's quite interesting in that goes into a lot of basic human psychology (fear of failure, etc.) and ways to think about trading that get around some of the ways that we are hardwired (he compares a successful approach to trading as comprable to a successful approach to gambling - playing the odds but always recognizing that at any time anything can happen.
The quote that hit me hard and reminded me a lot of myself (and maybe some others I know):
... why do you think unsuccessful traders are obsessed with market analysis? They crave the sense of certainty that analysis appears to give them. Although few would admit it, the truth is that the typical trader wants to be rigth on every single trade. He is desparately trying to create certainty where it just doesn't exist. The irony is that if he completely accepted the fact that certainty doesn't exist, he would create the certainty he craves: He would be absolutely certain that certainty doesn't exist.
Sounds a lot like embracing the force, young Skywalker
|
|
Registered User
Joined: 12/21/2004
Posts: 902
|
Whoops - of course in the above post I meant to post Scotts's quote rather than Fptry's (makes a lot more sense with Scott's quote LOL)
QUOTE (scottnlena) FPTRY
Thanks. That is interesting as well with the longer term pattern arangements. Is there a time period to the ideal cup and handle?
Originaly I was thinking short term trading but I want to start working on my ability to hold longer term stocks. I had discovered and read that Triangles have a 50% ish predictibility rate.
Has O'Niel written on the toppic or is his work only available through the web site?
|
|
Registered User
Joined: 12/21/2004
Posts: 902
|
well obviously I did quote the right quote. I need more coffee...
|
|
Registered User
Joined: 12/2/2004
Posts: 1,775
|
QUOTE (scottnlena) FPTRY, Is there a time period to the ideal cup and handle? Has O'Niel written on the toppic or is his work only available through the web site?
Great thread you've started Scott. Per O'Neil the shortest period for a legit cup with handle pattern is 7 weeks, longest 15 months. The best of the cup/handle patterns avg. about 3 months. Minimum time for a handle is 5 days. First day on the pattern is first down day from a prior uptrend of at least 30% that starts left side of cup. Last day of pattern is the day price breaks .10 above high of handle with volume at least 50% above avg. daily volume. Lots of other criteria go into it, really fun to learn and study it imo. If you check out the graphic chart in post from several days ago on VOLC you can see what I thought at the time was a good cup/handle pattern pre breakout. But it failed soon after, win some lose some, but the example was a good idea of what an O'Neil cup/handle chart looks like I think.
His best seller "How to Make Money in Stocks" contains detailed analysis of cup/handle and other patterns. It's on the Worden recommended book list. His website "investors" has even more detailed dissection of the patterns.
Tobydad mentions his other book "The Successful Investor," and I understand it too is a very well received book. It seems a lot of traders unfamiliar to the IBD/O'Neil school of investing think it's long term buy and hold, which is very inaccurate. O'Neil believes in short term (weeks) to long term to all inbetween. His philosophy is to sell when the chart says to sell, whether one day or one year after initial buy. He places equal emphasis on technicals and fundamentals.
Oh, thanks Hohandy for mentioning "Trading in the Zone." I had forgotten about that book but have heard it's a good one...I need to check it out.
|
|
Registered User
Joined: 9/21/2005
Posts: 566
|
This is a good thread. Perhaps its easier to backtest with candlestick patterns more than other measures because candlestick formations easily can be re-created as formulas. There might be degrees of price divergence but for most candlestick patters its either black or white.
I wonder if there are there more academic studies out there that have analyzed all the various popular technical indicators for historical effectiveness?
Some good traders (who have historical track records of accomplishment) have told me they only focus on some combination of:
closing price
support resistance
moving averages
trend lines
volume
|
|
Registered User
Joined: 7/24/2006
Posts: 131
|
For platinum users, cyberdawg, who heads the 'Disciplined Swing traders' club has done pretty extensive work testing candlesticks. It's called "New Candle Finder PCF with backtested candles". You'll find it under the "Reports" tab. I will e-mail him and see if he would mind me posting it on here.
Kevin
|
|
Gold Customer
Joined: 11/11/2006
Posts: 359
|
Scott: You may wish to try the free 30 day trial of North Systems CandlePower 6 Pro. It is a great companion with TeleCharts and has about everything you can know about candles and indicators. Should answer all your questions about these subjects.
Mammon
|
|
Registered User
Joined: 3/21/2006
Posts: 4,308
|
Scott, I strongly agree with Diceman here. You have proven that TSV works. Martha has given us all the candlestick knowledge that we need. Trade the TSV system like a business, protect capital at all times. The ones that get rich do so buy being focused...
Good luck..
|
|
Registered User
Joined: 4/18/2005
Posts: 4,090
|
Yes... I'm with you. My thinking is this. Most people ... at least new to charting tend to think in terms of candle pattern period. IMO candle patterns in and of them selves are a small part of the picture .. nearly worthless in and of them selves. so I have developed the background or context in which I want to see my candle signals. I have found my entry alert. But I still enter off of candles. I'm curious to find out which candle patterns are the most effective.
I still generate ample potential entries on any given night. IN the narrowing and selectino process If I know that engulfing whites outperform dragons as a rule by 20% then I may be inclined to take the engulfing over the other pattern.
|
|
Registered User
Joined: 11/24/2006
Posts: 9
|
QUOTE (tommy21) Scott, I strongly agree with Diceman here. You have proven that TSV works. Martha has given us all the candlestick knowledge that we need. Trade the TSV system like a business, protect capital at all times. The ones that get rich do so buy being focused...
Good luck..
Who the heck is this Martha everyone keeps talking about? Please point me to a thread or her web site.
Goodness, I'm not trying to be funny, but the quotes about Martha are kind of scary in a Jim Jones / David Koresh sort of way.
|
|
Registered User
Joined: 2/8/2006
Posts: 73
|
I agree with you about the Martha thing....
|
|
Registered User
Joined: 3/21/2006
Posts: 4,308
|
Martha Stokes is a senior technical analysis instructor for Techni-Trader (Technical analysis trading courses). She also writes articles for Stocks & Commodities magazine.
Most of us that have taken her courses and read her articles consider her somewhat of a Guru.
Her teachings focus on using candle stick formations supported heavily on good sound technical analysis technics... She also advocates the use of the Worden Telachart software. (that is how I arived at this forum about a year and a half ago).
|
|
Registered User
Joined: 4/18/2005
Posts: 4,090
|
I must say that for the beginner the training is good and fairly priced.
|
|
Registered User
Joined: 4/18/2005
Posts: 4,090
|
There is a nother thread here when most candle patterns are converted to % true indicators for testing. So i'll start running my own tests soon.. ut at a glance.. I gotta say that candles in and of them selves aren't looking that great. Which is as I supsected. They are next to worthless unless they come in the right context... then they move up a notch... but We'll soon see.
|
|
Registered User
Joined: 1/29/2005
Posts: 104
|
I like Greg Morris' book as a primer on candlesticks. I have since read Stephen Bigalow's "High Profit Candlestick Patterns: Turning Investor Sentiment into High Profits" is excellent reading. He gives a simplified, easy to read description of the patterns he favors. His web site also provides a free weekly chat room that reinforces your training. You can find his site by googling his name. I highly recommend this site.
I have attended many stock training programs and find candlesticks to be the most consistent at finding good stocks, futures, commodities, etc to invest in - short term and middle term.
|
|
Registered User
Joined: 4/18/2005
Posts: 4,090
|
Perhaps I will stand to be corrected. I do use them becasue they make sense to me and I like that you can clearly "see the story of price" Bt I personaly think as you narrow the time period the more noise gets into the picture.
|
|
|
Guest-1 |