Right on IronFist. My win rate on patterns was about.....you guessed it....25%. So, I started fading them and my win jumped. This was one of the ways I evolved as a trader. Everyone has to find what works and what doesn't, make adjustments, and find their groove.
Why I won't trade patterns
- Thread starter ProfitTakgFool
- Start date
I tried that too - i.e. large volume break up confirms the move. Funny, that didn't seem to work either. The lights came on for me when I began studying the statistics of the financial markets. Funny tho, all I ever learned in grad school was that statistical analysis of the stock market doesn't work - non-stationary and unit root problems proved it so. I understand the non-stationary problem but for the life of me I can't recall what the unit root problem is. I can't tell you how many proofs to this end I had to sit through. Interesting stuff tho.
Hypo,
Here's what worked for me. I don't know if you have ever studied statistics but stock market data looks like an F-distribution, or skewed left or right. Within that F-distribution there are normal distributions. The normal distributions can be traded, the trick is knowing how to find them. My strategy is this...find a normal distribution within the day and fade the number that goes in the tail, or the part that represents 2.5% of the data. Not by coincidence, that data just happens to be the move coming out of a pattern. So, hypothetically, my win rate should be about 95% (2 tails are 5%). The problem is, when we move out of a normal distribution (pattern) we can be converting back to the F-distribution, which represents the entire day. Is there a way to solve this problem? The answer is yes -- risk and money management.
With all three of these tools - stats, risk, and money mangement I have managed to get my win rate to 90%.
That is the reason I won't trade patterns. I just do what I know works.
Here's what worked for me. I don't know if you have ever studied statistics but stock market data looks like an F-distribution, or skewed left or right. Within that F-distribution there are normal distributions. The normal distributions can be traded, the trick is knowing how to find them. My strategy is this...find a normal distribution within the day and fade the number that goes in the tail, or the part that represents 2.5% of the data. Not by coincidence, that data just happens to be the move coming out of a pattern. So, hypothetically, my win rate should be about 95% (2 tails are 5%). The problem is, when we move out of a normal distribution (pattern) we can be converting back to the F-distribution, which represents the entire day. Is there a way to solve this problem? The answer is yes -- risk and money management.
With all three of these tools - stats, risk, and money mangement I have managed to get my win rate to 90%.
That is the reason I won't trade patterns. I just do what I know works.
Indeed I did study improbability and sadistics. But I prefer to think of your tails as having run to strong S/R or to volume exhaustion. Different names for the parts of the elephant.
That's basically what I'm looking for, exhaustion moves. From a statistical standpoint, they are very hard to isolate so you can't just take something like "sadistics", lol, and apply it to the financial markets like an exact science. It's takes a lot of trial and error before you find your way.
Good talking with you this afternoon. Happy New Year!
Good talking with you this afternoon. Happy New Year!
"So there is dilemma one -- which way will it break? The odds are 50/50."
and this is supported by what evidence?
really, the lack of understanding of math, game theory, etc. by supposed "traders" astounds me.
just because a situation has two outcomes does not mean the odds are 50/50
that's about as basic an element of statistics as there is.
fwiw, i don't trade "patterns" either, in that i don't make any decisions SOLELY based on a shape. do i take patterns into account, as a small part of the greater whole? yes/
and this is supported by what evidence?
really, the lack of understanding of math, game theory, etc. by supposed "traders" astounds me.
just because a situation has two outcomes does not mean the odds are 50/50
that's about as basic an element of statistics as there is.
fwiw, i don't trade "patterns" either, in that i don't make any decisions SOLELY based on a shape. do i take patterns into account, as a small part of the greater whole? yes/
I use a 40oz, er, sma for support in an uptrend and resistance in a downtrend ... jeez you guys are sooo complicated. 
But seriously, what we see as patterns only exist on the lower fractals of price action, on the higher fractals, price is only in an uptrend, or a downtrend ... period.
Good trading,
I
But seriously, what we see as patterns only exist on the lower fractals of price action, on the higher fractals, price is only in an uptrend, or a downtrend ... period.
Good trading,
I
Agree, a bull pennant in an uptrend, logically, has a greater chance of an upside move but you still have to consider the fact that a bull pennant could trade into an ascending triangle and trigger your stop in the process, or make a sharp break below support, for whatever reason, and head northward shortly thereafter. The stock market is nothing more than a random number generator and random number generators do, in fact, create patterns.
Somone a while back on this board (can't recall who it was) said the key to trading patterns is doing it on the right stocks - i.e. GOOG, AAPL, RIMM, and the like. Well, if this is the case, then you aren't trading patterns, you are trading fundamentals. Stocks with poor fundamentals don't behave well when they form patterns because they don't have the fundamentals to support their movement unless the move is a result of speculation, rumor, or pumps from the likes of Cramer.
Don't get me wrong, I think fundamental analysis is heavily flawed as well. Why anyone would buy a stock soley because it has formed a bull pennant or because it has advanced 200% on good fundamental news is beyond me. Could it go higher, could you make a profit? Sure, crazy shit happens in this business but in the long run you're destined for failure by buying high and expecting to sell higher, just because you see a pattern.
Ask the sharpest minds in the investing business and they will tell you they buy great companies with superior economics when everyone else is selling them -- here's the key, when everyone else is selling them. Tell them about your bull pennants and they'll probably get a good chuckle.
Somone a while back on this board (can't recall who it was) said the key to trading patterns is doing it on the right stocks - i.e. GOOG, AAPL, RIMM, and the like. Well, if this is the case, then you aren't trading patterns, you are trading fundamentals. Stocks with poor fundamentals don't behave well when they form patterns because they don't have the fundamentals to support their movement unless the move is a result of speculation, rumor, or pumps from the likes of Cramer.
Don't get me wrong, I think fundamental analysis is heavily flawed as well. Why anyone would buy a stock soley because it has formed a bull pennant or because it has advanced 200% on good fundamental news is beyond me. Could it go higher, could you make a profit? Sure, crazy shit happens in this business but in the long run you're destined for failure by buying high and expecting to sell higher, just because you see a pattern.
Ask the sharpest minds in the investing business and they will tell you they buy great companies with superior economics when everyone else is selling them -- here's the key, when everyone else is selling them. Tell them about your bull pennants and they'll probably get a good chuckle.
As price approaches apex of pennant one really will do himself a hell of a bonus by looking at something like MACD's histogram, some will monitor the DOM/volume etc. I prefer histogram for its easy visual aspect. Why use an indicator? Aren't they lagging and therefore shit? ImE, NO they are not shit. Looking at price along approaching the apex I don't see what is the likelihood of price breaking out or breaking down, histogram on the other hand indicates where it is most likely to break. No holy grail there, just a great 'navigational' tool.