3 to 1 Ratio

Discussion in 'Strategy Building' started by DraXon, Nov 22, 2002.

  1. no, the random test is just to see if your strategy really has an edge or not. You do the random test, then compare that to your actual system test results. If your tests beat the random tests then you have an edge.

    This is acrary's edge test. It beats the heck out of just backtesting and going with something that makes money, because if you don't compare it to random, you never know if it is really your system that has the edge, or if the profitable backtest was just a result of mkt conditions which may never repeat.

    The trick is finding the proper random correlation.
    #31     Nov 23, 2002
  2. acrary


    I thought you guys were done. Here's the results this test:
    #32     Nov 23, 2002
  3. acrary


    Here's the results for your 30min. breakout:
    #33     Nov 23, 2002
  4. acrary


    And here's the results of the 60 min. breakout. Both were done using the 1% of SP 500 target.
    #34     Nov 23, 2002
  5. acrary, something doesn't look right with your first test. If no commissions and slippage are factored in, then Avg W/L loss should be exactly 3.00, instead you got something more like 2.67 ??
    #35     Nov 23, 2002
  6. Thanks for your time. I appreciate that.

    #36     Nov 23, 2002
  7. acrary


    The difference is caused by the neither the profit target or stop loss being hit. In that case I put in a exit MOC to make sure we start fresh each day. The shortened trading periods before holidays caused the biggest problem.

    A smaller part of the problem is that the market may have gapped above/below the stop or limit order (in which case we end up with fills that are above/below the price points).
    #37     Nov 23, 2002
  8. acrary


    Just for grins, I thought I'd post the results of one of my daytrading edges over the same period. This is a breakout edge:
    #38     Nov 23, 2002
  9. 0008


    Is 3 points for the ES too tight? I think it is OK. Even for the NQ it may not be too tight. Its well-know volatility already gone.
    #39     Nov 23, 2002
  10. The tests presented in this thread do not confirm this, that is, as far as profit is concerned. However, if you want to minimize drawdowns, 3 pt stop-loss is better.
    #40     Nov 23, 2002