Discussion in 'Technical Analysis' started by vladiator, Dec 2, 2002.
Just came across this article today.
Check it out.
This guy doesnt know his support from resistance. Trading is not about testing bootstrap statistics.
It is impossible to test whether technical analysis as a whole works or whether is doesnt. Anyone that attempts to.. should be discredited.
The only thing that one can try to test.. is whether a certain technical strategy actualy worked.. not the entire notion of technical analysis.
Technical Analysis works just as well today as it did ever since liquid markets were invented.
--Let the traders trade..
The thing is Ta is still valid but one has to evolve and stop using the classic methods based on trading patterns and oscillators.
One has to be quicker. Before you could have a nice signal and all the time to enter your position, today, this is not the case anymore. when you see thje signal it's already to late. traders have already bought or sold the stock so you start chasing the stock and this is very very dangerous.
As more and more people know about TA, it is key to know it and be on a move before them...
The second aspect is that market makers always create whipsaws in order to generate volume and trigger stops and entry executions. This phenomenon exister before but today, it is more obvious.
No it's not nonsense. You didn't get the point. Testing TA the way you suggest is not implementable. The data is not there. Secondly, there is a myriad of strategies that would have to be tested etc.
Some of them mightt do better now, other would not, yet others would remain unchanged. The only practical way to see if there was a structural shift in how well TA works is to see if the pros using it are successful or not. Arguably, if it works better in some period than in another, those who specialize in it would do better or worse depending on the period. There are special cases/outliers of course, but the overall approach is valid.
I dont think you get what I am saying.
There is no way to test if TA as a whole works or not. Period, end of story. Anyone that tries is fooling himself.
I dont get it... what is a pro supposed to say... yes it works this year.. but last year it didnt work so well. Maybe he would be talking about his own technical trading system which has absolutely no relation to whether or not TA as a whole actually works.
My point is.. the only thing that can be tested is an actual technical trading system... via something like tradestation or wealth lab. However.. this individual system has nothing to do with the overall notion of whether TA works.
This is a very interesting article and I advise people to read it first before making a conclusion about it.
The article has 2 main points:
1. volatility in commodity markets has decreased
2. autocorrelation of daily price movements has decreased, in other words, trends are less defined.
For their purposes, it seems the authors define TA as a trend following system...thus this definition doesn't cover alot of the other aspects of TA.
I believe the authors could have gone one step further and asserted that commodity markets have become more efficient as more participants enter and thus reducing the profits of common, trend following TA-systems.
If they found out that TA worked they would stop being academics and go get some $. On the other hand, if they are employed and put $ in a 401k they add to the liquidity so let those sleeping dogs lie anywhere they want to!
Like Hunter [age 2.5] says ''squigely lines and numbers'' It's not just squigely lines but squigely lines and numbers.
Patterns still work just as they always have because human behavior has not changed. When the markets are no longer driven by hope and fear then patterns will not work.
I'm speaking of triangles and such. But oscillators do work too because they are quantified behavior, as patterns are charts of behavior.
Nothing has stopped working. And no new understanding is required of any pattern or oscillator. I don't get that part of the argument at all. Trendlines aren't different. Ma's work the same.
If anything, all this stuff works better because there is now some self-fulfilling action to them thanks to the internet and cheap commish.
Well, at least I make money with them. And good money too.
Perhaps there is less fluctuation in the prices of commodities resulting in decreased opportunity for trend trading profits. That only makes sense as technology smooths out the fluctuations in crop production and herd growth. No mystery there. Oil, metals, grains, and meats are all going to smooth out over time. Cocoa, a one country commodity, and coffee similarly, they will continue to fluctuate more, but less and less as Juan Valdez is replaced by Folgers.
Geez, I bet these guys got grant money or tenure to write that paper. And they say there's no such thing as a free lunch.
Separate names with a comma.