optimising Does it work

Discussion in 'Trading' started by jem, Aug 31, 2002.

  1. It has been said before but I will say it again -- throw out all your indicators and start from scratch if you use many of them to trade. The best traders use only one or two indicators at most, and they are generally the most simplistic ones (MA, Stochastics, etc.)

    If you are monkeying around with 7 indicators, you're heading in the wrong direction. All you need to know to trade intraday successfully is the MA, support / resistance areas, previous days open and close, weekly high / low range, etc.

    Just because there are over 500 technical tools doesn't mean that they are more useful as you combine and stack one on top of each other.

    You can still build a great house with just wood, nails and a hammer -- KISS (Keep it simple, stupid).

    aphie
     
    #11     Aug 31, 2002
  2. I agree.
     
    #12     Aug 31, 2002
  3. BillyG67

    BillyG67

    A while back i had purchased OT2000 the theory behind this program was optimizing and finding the right system\s in the back test period that performed the best. Backtest results from about 70-100%
    The problem with over optimizing at least with OT2000 is One day you will get a brand new Buy signal, the next day the stock tanks and you get a brand new sell signal,the day after you get a buy signal generated 20 days prior and so on,

    on the forward test I was getting results of about 30-40% on systems that backtested at 100%

    Optimization is good to a certain extent My feeling is find a happy medium and stick with it. don't over do it!

    Just a thought from a newbie!
     
    #13     Sep 1, 2002
  4. Biomech

    Biomech

    jem, I think you are asking this question on the wrong board. The dominant voice here at ET is the voice of the discretionary trader. Check out Chuck Le Beau's forums at www.traderclub.com, which deal mainly with system development. While they are rather disorganized, you will find alot of very in depth discussions on optimization and curve fitting, among other things. Lots of mathematically inclined folks post there, so check it out to get a bit of a different perspective than you will get here.
     
    #14     Sep 1, 2002