can 1.5 pf strategy work

Discussion in 'Strategy Building' started by kenten, Jul 24, 2013.

  1. d08

    d08

    HFT rarely reaches more than 1.2-1.3 PF, so...
    Be equally worried when the PF is unrealistically high.
     
    #11     Jul 28, 2013
  2. kenten

    kenten

    Hi all. Thanks for all the advice.
    I need to make some explanations
    This is a non us equity. mainly devoloping country . I have no vendor to get the tick or 1 min historical data.
    Problem is the market conditions and players changing with time. So the latest conditions should prevail for some time.
    But there is no or little hft action and few systematic traders. not much volume but enough for me . These are great opportunities in my opinion.
    I am going on live trade and it is behaving as expected.
    I will pull the plug or adapt the system when /if live outcome changes.
     
    #12     Jul 31, 2013
  3. Anythign can be skill until it turns to luck or the other way around.
     
    #13     Aug 3, 2013
  4. A very practical way to "test" if your profit factor of 1.5 can be attributed to skill, would be to run a confidence interval on the trades. Excel has it all built in, so you don't even need to know the statistical calculations.

    I fully understand that confidence interval calculations work based on assumptions that are somewhat suspect for the distributions in trading, but if I see that I get a confidence for profits being positive of 90% or better, I would consider that much more significant than most other objective measurements.

    If anything, the fat-tail distributions seen for profit distributions, tend to make confidence calculations appear less favorable than they actually are (my opinion based on my trades).
     
    #14     Aug 3, 2013
  5. kenten

    kenten

    Whenever i do test an idea , i look at the profit curve. The more smooth it is the more confortable i am with the results
     
    #15     Aug 5, 2013
  6. Agree, timeframe is too short. Try years of minute data. Also, try up markets, down markets, and side to side markets.
     
    #16     Aug 15, 2013
  7. Dont forget to include slippage and commissions that can be a big negative.
     
    #17     Aug 15, 2013
  8. kenten

    kenten

    these are the live results from last month ?
    it is on %25 margin
    i have to accept most days were down trend days which may not happen in the future that frequently
    losses happened on choppy days
    i am working on some filters to reduce the loss on choppy days
    from what i see examining the chart a trailing stop may work, it will reduce the overall profit though.
    unfortunately i have to work with the same limited data

    i am currently searching the forum but any insight about effects of trailing stops and profit targets on sustainable success of a system will be helpful. sorry that sounded a topic sentence from an academic paper:)

    results
    7/26/2013 3%
    7/29/2013 -9%
    7/30/2013 1%
    7/31/2013 10%
    8/1/2013 -1%
    8/2/2013 -2%
    8/5/2013 2%
    8/6/2013 -1%
    8/13/2013 9%
    8/15/2013 1%
    8/16/2013 10%
    8/20/2013 4%
    8/22/2013 10%
    8/24/2013 -9%
     
    #18     Aug 24, 2013
  9. Sergio77

    Sergio77

    Not academic - very practical question. See this for example. I do not agree with everything but it is a good start.
     
    #19     Aug 24, 2013