5. Time of day, many traders miss the fact that some patterns fail at certain times of day. Several chart patterns have a high tenancy to fail and certain times of day, let me offer a few examples. 1. Breakouts, for one, have a high failure rate if they occur over lunchtime ( 12est-1pm est ). Breakouts that occur over this time period tend to fail or have limited followthrough due to the lack of volume that accompanies this period of the day. 2. Reversal patterns have a high failure rate during the last 30-40 minutes of the day as most traders are unlikely to fight the trend for very long into the close. Shorting resistance or buying support is far riskier a proposition going into this time period. 3. There are 4-5 " reversal periods " that occur during a normal trading day. 9:50,10:15,11:15,14:15,15:15 ( these are approximations and meant as a general rule ). Chart patterns that occur during these timeframes that market possible reversals of trend have a high failure rate. While none of these " time period " rules are written in stone it is good to be aware of them. Always refer to more than just 1 time period when entering a trade based on a chart, be aware of the time of day, understand possible news effects, and be conscious of underlying price support ( fibonacci for example ). Allen
T/A is like a roadmap drawn 100 years ago by a drunk. The major features of the map will probably stay the same, but if you read it as the end all you will get lost, confused etc.
Let me offer somewhat different take on the question 'do TA work". Please keep in mind that this is just my approach and I am not adept in TA, so it might be of no value for some. To me, TA have no PREDICTING value, in a sense that this or that setup or pattern do not allow to tell in advance which way price goes. Their value is different: they put structure in what seemed to be chaos. Using Brandon's map analogy I would illustrate it like following: You observe car on the road trying to figure out which way it headed. There is a crossroad ahead and until car reached it you don't know which way it choses. When it makes right turn you can assume with high enough probabaility that it goes to the right. Can it stop and back up? Sure it can but still, betting on right direction after right turn you have odds working for you. At the same time you know what signs indicate that your initial assumption is wrong: car reaches crossroad going backward and turns left. Exactly this way setups work - they show you road signs. Let's cite simple example: Your setup is breakout over 20 with support at 19.75. There are two road signs: break of 20 indicates going upward and break of 19.75 indicates downward. Having those signs identified you got your set of IF-THEN scenarios. You also got your safety net: if price moved through 20 and dropped to 19.75 and broke it - oh well, driver was drunk That's what I meant when said that TA didn't offer me predicting value. They haven't told me which side was going to be broken. But they did offer me the structure and favorable odds. This is one more way in which "Tade what you see"... works. So, for me "Do they work" is nor even question... sure they do - as long as you read them to your advantage and do not expect them to just tell you what to do. As usual - it's always the trader... Best regards, Vadym
For those interested in the success rate and failure rate of about 50 different chart patterns, take a look at "Encyclopedia of Chart Patterns" by Thomas N. Bulkowski ISBN 0471295256 While the book was written for swing traders, I use the patterns described for day trading. It's my bible.
jperl I have one problem with this book. The data he crunched was only over a five year period (91-96) and I wonder if this is enough for a true picture...
The tremendous popularity of T/A has reduced its effectiveness IMO. Whats the old adage about the market does whatever hurts the most people at any given time. If everybody is following the same patterns I dont see how they can work. I think the market can only support so many people. Those following well worn pathes will come to a dead end. This year should weed out the weak hands. I hope I'm not one of them.
Buklkowsi typically has about 100 or more data points for each pattern. The error in his analysis would then be the inverse square root of 100 or about 10%. I can live with this.
IMO the prettier the set-up the less likely it is to work... Personaly I would rather look to play a failure of a seemingly "picture perfect" pattern... I think the problem is that when you have a picture perfect set-up that every body can see, they were either 1) piling in early, anticipating the actual trigger, which makes the pattern almost artificial in a sense or 2) they are all gonna jump on it at the same time and when it comes time to throwback to the trigger / breaking area nobody will be left to support it, as there are few that missed the train... In either case failures can be very lucrative with classic patterns... BUT this is in no way an attempt to denounce the validity of patterns such as flags, wedges, tri, cup w/handles... They are patterns that have stood the test of time and will be around imo as long as the markets are... PEACE and good trading, Commisso
Here is a good analogy on the best setups. When I was starting out and found some one to help me, I was scanning one night and was really excited because I had found "perfect" setup. When we talked that night I was really excited and showed him my pretty setup. He said that yeah that is really nice. Then he said, if this setup was a pretty girl you met at the bar, would you want to take her home and screw you? I figured that yea, its a nice setup, so if its a girl..sure. He then said well, when it screws you don't be mad, its what you want. I thought he was nuts, took the swing setup, and it stopped out. Thats not to say you should not take a nice setup, you should..but if everything is setting up just perfect, you might be carefull.
I have seen some perfect setups that worked great. But I don't like it if they are found by a lot of other people and everybody knows about it. If Wordens, newsletters and web sites all say look at this tgreat pattern about to set up it most of the time doesn't work. This is why I think it's good to monitor some newsletters and even CNBC to see what the majority are thinking and also trade some stocks that aren't the very favorites.