Ninjatrader Strategy Advice - Gold mine or ignorance?

[Maslo]

Hello, I'm a newbie to this forum. I have been playing with strategies for the e-mini S&P500 in Ninjatrader 7 for nearly a year. I have never actively traded, but I have been using real-time data to develop and test strategies, so as to minimize my risk when I do start.

My current fully automated strategy appears too good to be true, which makes me skeptical. However, I very well could have stumbled onto something. I just don't know. I have no experience in trading, but have done extensive research on methods and psychology, and I am good with code. The strategy is based on personal observations and this research. I run this strategy both in real-time and backtest, and the results agree with each other.

I would greatly appreciate someone with experience in Ninjatrader automated strategies to chime in. On paper in looks great to me, but I don't know how it matches up to other strategies. I posted backtest data to 2006, as well as monte carlo charts for profit and drawdown for this 10 year period. Thanks for your comments!

 

[NetTecture]

There really is not enough information to say anything here.

Geneally - NinjaTrader has a lot of problems with certain strategies. Problems normally arouse around the fact it does not properly backtest against amrket data, but against bars. As such, it can not properly simulare:

* Trading in Renko charts (as high and low of a bar may not correspond to price action)
* Trading with high volatility, or where the profit per trade for passive orders is low, as it can not do any simulation of bid/ask size and spread.

This puts limits on the simulator. It is workable, but if your strategy falls into specific behavior, it seriously misbehaves.

This LOOKS good. Let it run for a month against simulator, compare that with a backtest for the period. See whether they match up.

 

[Maslo]

There really is not enough information to say anything here.

Geneally - NinjaTrader has a lot of problems with certain strategies. Problems normally arouse around the fact it does not properly backtest against amrket data, but against bars. As such, it can not properly simulare:

* Trading in Renko charts (as high and low of a bar may not correspond to price action)
* Trading with high volatility, or where the profit per trade for passive orders is low, as it can not do any simulation of bid/ask size and spread.

This puts limits on the simulator. It is workable, but if your strategy falls into specific behavior, it seriously misbehaves.

This LOOKS good. Let it run for a month against simulator, compare that with a backtest for the period. See whether they match up.

Thanks for the information. The strategy does not fall into those categories, but your advise to verify backtests with real time executions is great. I will definitely continue testing before jumping in.