Joe Doaks Backtests Unlikely Strategies

Discussion in 'Strategy Building' started by Joe Doaks, Jan 16, 2008.

  1. i'm curious, how did you come to this conclusion..(no stops/pt's)
    and, if hypothetically, you had a system that did produce a positive result w/o stops/pt's, would you then insert these signals?
    thanks,
    :)
     
    #11     Jan 17, 2008
  2. LMFAO, you should put this stuff in a book.
     
    #12     Jan 17, 2008
  3. Eddie, the speed ratio is the ratio of the settings of the two EneMAs, par exemple, 100:25, which specific rat-io will do you no good at all intraday. But I am not joking. EMAs work because The Street's algorithmicists are too dumb to handle anything more complicated. For years I have tried to popularize the adaptive Kalman filter, to no avail, because they can't understand it. It's engineering, as opposed to pure math, haha! What do we sparkies know?
     
    #13     Jan 17, 2008
  4. BubbaBig, that is a great question. The porblem with uptimization is that it leads to breathtaking stops, like many points. Every system I ever tested gave lousy results with rational stops (I use 3 ticks for NQ, and am thrilled to escape with four). So I have a theory. If I have to optimize it, it doesn't really work. The corollary is the following. The more you complicate a good idea, the more random and unreliable the test result is. Mind you, nothing I use now is even remotely testable. I learned that scam from Jack. And no, I wouldn't trade anything so simple that I could test it. That takes all the excitement out of trading.
     
    #14     Jan 17, 2008
  5. MangleBread, I would not tarnish the memory of the great and late lamented Bruce Babcock Jr. by writing a book on backtesting. His could never be improved. But I do appreciate that you get it that backtesting is funny.
     
    #15     Jan 17, 2008
  6. LOL, it wouldn't be a book on backtesting.

    It would be a poetic treatise on the insanity of the markets ... I'm sure it would sell quite well.

    :D
     
    #16     Jan 17, 2008
  7. Are you thinking Dante? Milton? Kierkegaard? Or Hunter Thompson?
     
    #17     Jan 17, 2008
  8. Definitely Hunter ... are you kidding me, Long-Term Capital Managment, Victor Niederhoffer, Sub-Prime Mortgages, the list goes on and on.
     
    #18     Jan 17, 2008
  9. I haven't proved or disproved to you that you should buy an up close. Should you? You don't fucking KNOW, do you? But let's suppose that I did, and that you liked "Buy Monday". Thus begins madness. You think, if I should buy Mondays, and an upclose, how much more would I make if I only bought upcloses on Monday? See the attached. A lesser intellect than mine would optimize this, hahaha!
     
    #19     Jan 17, 2008
  10. Breadmangler, one could hardly outdo Mackay or Kindleberger, but then again, they never met Jack Hershey.
     
    #20     Jan 17, 2008