Traders, I need some comments. Here's a website which shows a lot of candlestick patterns. It also classify them into continuation and reversal patterns. Do you think these candlestick patterns are reliable (it doesn't need to be 100% correct, but I expect a higher percentage of accuracy, say 75%)? Any comment?
Hi WmWaster, I used 1015am - 1130am est as one example only. Thus, its not the only time I trade nor is Wednesday the only day I trade. Therefore, key market events occurs on different trading days and at different times during the trading day. To answer your question, if I get a pattern signal at 12pm est...I don't ignore my pattern signals. The concern that I keep emphasing here is the trade management that will be different depending upon what's causing the price reaction and depending upon when the key market event occurs. (ex. Fed Chairman speeches/comments have a stronger reaction in the afternoon in comparison to the morning) (ex. I tend to stay in trades longer for a continuation pattern signal in comparison to a reversal pattern signal eventhough it can be the same pattern signal). Take a look at the Bullish White Hammer pattern. Most candlestick books call this a reversal pattern. Reality is that its very dynamic and can occur as either a continuation pattern signal or a reversal pattern signal (see Trading Hammers (revisited) thread). Simply, I'm concerned with the adjustments in the trade management after entry because I've matured in my trading where I now know that just because its a Bullish White Hammer pattern doesn't imply its occurring in the same type of price action. However, when I first started trading... I traded a Bullish White Hammer pattern in the morning exactly the same way I would trade a Bullish White Hammer pattern in the afternoon and exactly the same way if such occurred in the overnight trading session. The flawed perception was that the pattern signal defined the price action. Thus, I thought the price action was always the same. In reality, its the price action that defines the pattern signal and the position must be managed based upon the cause of the price reaction and when such occurred. Mark (a.k.a. NihabaAshi) Japanese Candlestick term
What's your definition of reliability??? Lets assume your trading QQQQ and it moves .50 in your favor after your entry but doesn't reach your 1.00 profit target and then pullsback and takes you out for a loss. Would your stats show that as a failed trade or a profitable trade for your statistical results??? My point is that another trader may only have a .45 profit target and you two could have completely different reliability results while using the exact same method. Simply, just looking at a percentage number all by itself may not be the true measure of reliability when that reliability is based only upon where you had exited the trade. Mark (a.k.a. NihabaAshi) Japanese Candlestick term
In either case I think it helps to follow a stocks historical data and see on average how much it rises / falls within your desirable timeframe. No point in entering a stock if on average it only rises by .25 or .45, and your looking for a .50 gain. - nathan
Thanks for your compliment. Although I can be profitable now without learning new things, I'm always eager to learn new things and improve my profit-making & loss-control ability. Actually I'm just wondering why there's no writer (as far as I know) who realise candlesticks & their patterns are dynamics (this is what I understand from my experiences), and publish a book or some resources to deal with it. It seems they would like to leave all these work to every candlestick (price action) learner, or they just don't realise their straightforward interpretation of candlestick won't work.
I understand that. My question is also an assumption. I just try to figure if you would trade a signal if you are confused with the cause of this reaction (ie you don't know what causes it rise). Really that's the problems of candlestick books. They stubbornly sort every candlestick pattern into one specific group. In reality, they can be both depending on the context. You method of trading is quite interesting. I start to understand why you need to know what events occur. They not only help you time the market, but also tell you how to interpret the trading signal when it occurs. Eg: Maybe you will be more confident to a breakout signal when it is at 11:30am where a key event occurs around this time. I don't realise patterns can be time-specific too. Although I'll try to figure out when most reaction occurs. Eg: For my Hang Seng Index, there's usually more volatile within 30 minutes after the market opens. The trend in afternoon is clearer than that in morning. But all these just tell me to concentrate more on specifc time periods. But if a trading pattern occurs exactly in about the same size, position, & market trend(eg does this pattern happen in a upward or downward trend?), I would treat them exactly the same without concerning when it occurs. I would like to hear more how you relate causes of reaction & time of occurrence to interpret the same trading pattern differently.
My definition of reliablity is how well the signal works to reach the profit target of the trader. There's no point to set one's own profit target if we shouldn't expect such a high target for this kind of signals. We are here to learn its methods (including the profit target). We will accommodate ourselves to its methods as long as the method is worth learning (ie a good combination of "risk/reward ratio & accuracy"). If the method does work but not work so nicely (eg the trader exits too early and limit his profits), we will then try if we can improve its method. After all, what we need is the statistical results based on the trader plans. If the results look good, we'll learn and incorporate them into our present trading plans. Does it answer your question?
Does anyone read a book called: Candlestick Charting Explained (Morris, Gregory) ?? What do you think about this book? Does the author focus on candlestick application, not just a mere explanation of candlesticks?
I commented on this deficiency in my original posting. To amplify more specifically, there is a profound basic difficulty. As you go from period to period of chart display (fractals) you will find that the appearance or lack of appearance of the types of candlesticks changes considerably. I stated this as a distribution of frequencies of appearance. Because this change occurs, the result is that the sequences of appearances of the bars also is differing on different fractals. Therefore, it is not possible to "write" a text or guide for speaking about how sequences of bars provide information or guidance on trading (making money) using bars in specified orders or combinations. The utter failure of the 1 minute bars was noted by inference and implication. For comparison, consider indicators that are "tuned". Understand "tuning" as setting proper defaults. Since the advent of indicators, I have "tuned" them according to the character of the environment of markets. Think of pre and post market operation relative to the advent of the PC. The 30 part sequence of the MACD, properly set, can be used as a signal repertior for making money. I advocate many of these for various maoney making approaches (think 80 successful edges. Candle sticks do differing sequences for differing fractals and thus they do not generate "signals" consistently. I could go into many other facets of this theme. If you read my IF1 , IF 2 model, you will see some of what I mean as a standard. Names of bars cannot be substituted for this strategy. Recently the software for the relationship of leading and lagging indexes was finally established mathematically in a statisitically significant manner. This kind of effort cannot be duplicated using csticks as named or coded since the coding changes from fractal to fractal. In June we are assembling a national team to work on coding and archiving my life's work. We will present and camtasia the interrelationships of indicators, their staticitcally significant sequences and the method of tuning indicator defaults according to the character of the trading environment. This will be done on six levels from "assessors (pre-novice) to expert. This includes information architecture, coding, writing, editiing, etc. for the construction of three kinds of web sites. Just like differing manufacturers make lubricants, the sellers have cross over lists so that replacements for one brand can be done with another. Part of our information architecture will include this. You will be able to see the connections among all the approaches of those mentioned in this thread. We are making it possible to handle the stumbling blocks of all types of approaches by showing the the relationships to other strategies that avoid whatever. It is like in the problem of having a unified theory a person just switches from relatavisitc to nonrelatavistic for intellectual convenience. A a theoretical phyics practioner from profession to profession, I have been able for a long time to do the equivalent with most of the available trading strategies. up to now, I never took the time to integrate all of the aspects of sucessful trading, I just simply used a critical path to be able to extract the optimum from any type market at any time. the oil brand crossover model will be used to correlate the majot money making thems over all time to allow people to see why this or that does not suffice when a better solution is available. We will also include the software for doing stuff mechanically. The above is stated to you as a way of providing some substance for my views. I do not use candlesticks and I feel that in the up coming documentation we are doing, the fundamental limitations of the candlestick approach will be clearly stated.