Block box trading

Discussion in 'Prop Firms' started by Jimmy3, Jun 8, 2008.

  1. wenzi

    wenzi

    Well, people are backtesting on such shorter time frames. You can get level 2 tick data nowadays. I have seen backtested systems with an average 6 sec trade time on the ES.

    If you really believe in your system, I say just trade it with your own money. Or even start out trading a single car of the YM which trades similar to the ES, has a lower margin and smaller tick sizes. You will find out quickly how well the backtested system works in real life.

    And with a good system you can talk to investors or just trade it yourself from the beach in Aruba.
     
    #11     Jun 8, 2008
  2. Why aren't you running the system in real time? A months worth of daily real time profitable results would be more convincing than any backtest.

    Do I understand correctly - daily profit of 6 points ($300) versus max drawdown of 60 points ($3K)? I know I don't have the stomach for that risk/reward ratio.
     
    #12     Jun 8, 2008
  3. bespoke

    bespoke

    Sounds like just another incorrectly done over optimized backtested system on not enough data without any out of sample data (99% of the holygrails out there). I say that because you didn't provide any details. Maybe if you could provide us with some of your statistics along with how you tested your data we could give you a better answer. You might even stir up some interest from investors on this board. But most likely not....
     
    #13     Jun 8, 2008
  4. BackTesting?? Try walking forward testing at a company with and excellant paper trading system like IB. BackTesting in most cases is not as effective as Walking Forward Optimization and testing.
     
    #14     Jun 17, 2008
  5. To "prove" anything, you'll need both an extensive back-test, plus a live forward test and be able to explain why they deviate at times.

    From personal experience, I can tell you that it's hard getting an aggressive intraday strategy to match backtest, no matter assumptions for slip, commission etc.
     
    #15     Jun 17, 2008
  6. What works very well for me is Walk forward Genetic Optimization of a diversified group of stock, futures and currencies (at IB)
     
    #16     Jun 18, 2008