According to Dierk Droth, the VP of Engineering for NinjaTrader, the upcoming NT7 beta will include a Monte Carlo simulator.
Monte Carlo sims are a way to produce random data to simulate different types of market scenarios and events, so that your strategy can be tested in a way that will preclude it from being curve fitted to historical data.
Although I am not currently pursuing automated strategies, I do still believe backtesting is a way to improve your trading. I use backtesting to find better values for indicator settings, or to just prove in a general concept if my theory works or not.
Additional details on the Monte Carlo simulator, as well as other NinjaTrader 7 features, are expected to be released within two to three weeks (by the end of June 2009).
Additional info on Monte Carlo Simulation:
Investopedia
Wikipedia
Mike
Email subscription
Labels
- about me (14)
- adxvma (2)
- automated systems trading (20)
- backtesting (8)
- books (2)
- CCI (4)
- channelthis (3)
- collective2 (1)
- double ma (9)
- downloads (15)
- easylanguage (1)
- eco2 ergodic (10)
- eot indicators (4)
- indicators (33)
- macd (2)
- market profile (1)
- mathematical analysis (4)
- money management (11)
- multicharts (4)
- murrey math (3)
- mysql (2)
- neural networks (2)
- ninjatrader tips (7)
- performance (3)
- poll results (4)
- polynomial (4)
- price action (11)
- prizes (2)
- psychology (14)
- scoring (3)
- symmetry (4)
- trade results (9)
- trading tips (26)
- trend bars (5)
- ubuntu (1)
- velocity futures (1)
- videos (24)
- vma (4)
- webinars (6)
- x-trader (1)
4 comments:
The main use of Monte Carlo simulations I've seen is to randomize the order of the gain/loss trade sequence produced during a backtest, rather than "generating bars". Presumably, if your backtest isn't too contrived, the specific order of the trades isn't that relevant to how the results will turn out in the future over the long run. This is especially important with leveraged trading, where a few losses in a row can blow up your account. A Monte Carlo simulation can tell you what percentage of the time the randomized trades result in a blowup or a gigantic gain. Some Monte Carlo simulators assume the trade results roughly follow a regular distribution, but some are not so strict.
Anonymous,
Thanks for the info. Monte Carlo is a new concept for me, I look forward to learning more about it and using it.
Mike
Hi Mike,
are you still using BWT?
David
David, not really.
Mike
Post a Comment
Note: Only a member of this blog may post a comment.