Technical Analysis Doesn't Work

Discussion in 'Technical Analysis' started by rcanfiel, Jul 16, 2007.

Thread Status:
Not open for further replies.
  1. \


    typo, sorry, should read " don't use time---PF and.....".

    PF--point and figure

    here's an explanation:

    http://stockcharts.com/charts/pointFigure/


    surf
     
    #221     Jul 20, 2007
  2. #222     Jul 20, 2007
  3. #223     Jul 20, 2007
  4. To have reliable and realistic testing results, one needs clear and undisputable rules. In math it is clear, because each rule can be expressed in a value. That's to me the advantage of TA as i use it. If i run three times my system through a number of trades, the results are always identical.
    This is never the case if you "look" at charts and use "various" systems. Each time you might see something different, so what's the value of the testing results? Three runs will give three different results.
     
    #224     Jul 20, 2007
  5. ok--no issue with your premise.

    if you cracked the code, more power to you. no argument here.

    however, my associates who work on such things have reached different conclusions.

    regards,
    surf
     
    #225     Jul 20, 2007
  6. Statistics? And would you be using historical price data when employing your statistical might? The same historical price data that, according to you, has absolutely no value in assessing possible future price direction?

    And of course you never said TA was not subjective. You continue to claim it is entirely subjective and, to prove your point, that is the manner in which you evidently use it by your own admission.
     
    #226     Jul 20, 2007
  7. have you considered that statistics may be applied to TS and automating tape reading, rather than past data on a chart??

    surf:D
     
    #227     Jul 20, 2007
  8. And how is T&S not historical price (and volume) data? The very historical data that, according to you, gives no insight to possible future price direction? Who said that TA is exclusively limited to the eyeballing and intuitive feel you report to be getting from your charts? Is that not just a convenient strawman construct so that you can readily knock it down? (And, oddly enough, continue to employ it in exactly that manner.) I think we need to remain consistent if we are to make any headway here.
     
    #228     Jul 20, 2007
  9. The fact that your associates came to other conclusions doesn't mean automatically that they are right. Perhaps they took the wrong road. This has nothing to do with intelligence ( because in that case i would surely loose), but with originality combined with logic and strong analytical thinking. One must see the relations that others don't see.
    In the medical world researchers look for the interactivity between different cells to find out which cell has what influence on the other, and is causing diseases. Once they understand the relationship they can start develloping a drug. Those who don't see the relationship will state that the disease is incurable.

    Two things are very important in observing systems:
    You need mathematical proof about things that should work. 1+1 is always 2. If i have to go long at 2 there is never any discussion about the objectivity of the rule. And the calculation will always give 2 on the same quote, no matter how often i rerun the calculations.
    If you trade on formations , or on trendlines, things get very difficult. When do you have a valid formation or a good trendline? And if you're in a different emotional mood (very optimistic or deeply depressed), you will see things dfifferent. So how can you do reliable testing on these kind of things? Lots of improvisation and results that vary very much due to the factor "luck".

    I tested a system with rigid mathematical rules on over 1000 trades. The statistics from this test are very reliable because the rules are fixed. So if i follow these rules i will have results that will be probably very close to the testing results.

    I use math but not the classic TA. To perform well you need things that others don't have. So throw away all the books and know indicators because they have all been tested already millions of times without success.
     
    #229     Jul 20, 2007
  10. I agree that TA consists of Price/Volume at its core, that is a given. The problem becomes your inability to see the simplest of scientific fact that a Price bar built with specific non-varying Volume is a more stable environment to view Price Action in than Prices bars build on time, ticks or range.

    You call this mystical or pseudo science but ever Mathematical/Logic's academic I've spoken to, which are many and respected, calls it, "Elimination of the variable aspect of an equation". This is alien to you because you also think of Logic as a product of Statistics which it isn't. Statistics is rampant with variables which is the exact opposite of Logic.
     
    #230     Jul 20, 2007
Thread Status:
Not open for further replies.