REAL unbiased statistical studies ?????

Discussion in 'Trading' started by marketsurfer, Jul 11, 2002.

  1. has there ever been statistical studies on the reliability on technical analysis for predicting future price moves ?? how about studies on candlestick analysis ?? if not this may be an area i will pursue.


  2. furthermore, i am speaking of real studies. not those created by "true" TA believers like the folks at MTA.

  3. BKuerbs


    Search the internet and you will find a lot of studies.

    One page: read article #26

    As to more conventional indicators, I know only of one book: "Technical market analysis" by Bauer/Dahlquist. Though their conclusions are very wrong: They studied the performance of indicators in the american market from 1985 to 1996 to claim that nothing beats buy and hold. Very true in a bull market!

    There are other books that claim to study the performance of indicators/chart patterns, but in my opinion they lack the statistical rigour required for such a study.


    Bernd Kuerbs
  4. thank you, bk !!

  5. the subjective nature of chart patterns causes a big problem when judging their accuracy. There are no universal official objective definitions for chart patterns. One man's cup and handle is not another man's cup and handle.
  6. sorry, i need to disagree with your post--dot/slash. there is consensus on what constitutes certain chart patterns. a double bottom is a double bottom, the breaking of a 20 period moving average is the breaking of a 20 period moving average, etal. there is even more consensus regarding candlestick patterns--a doji is a doji, a harami, a harami---etal. . .you are correct that certain patterns are subjective, the majority are objective.