Stochastic crossover

[snotrag12]

I am trying to use a stochastic crossover in my strategy. I wasn't quite sure how to do it, so using the wizard I clicked on stoch. then zeroed out period d & smooth, left period K at 14, hit cross above then on the right box hit stoch. zeroed out period k & smooth, left period d at 7. I used this to trigger my entry. It seems to work although not every entry seems to coincide with a stoch. crossover on the chart. Is this the correct way to do this?

Thanks in advance

[NinjaTrader_Matthew]

Hello,

If I'm following you correctly, you want to leave all the input parameters the same so that you're comparing the same exact series on both ends.

If you're checking for a cross, you want to select the "plot" you wish to check for the cross over.

I've attached a screen shot that shows the Stochastic K CrossAbove the Stochastic D.

[snotrag12]

Thanks, Matthew, I set up my strategy like the screenshot, with the action being to enter a trade, but it is doing it after the crossover. For example, the crossover on a 60 min. chart is at 9:00am but the trade isn't executed until 10:00am. Is that correct?

[snotrag12]

I also noticed on the sample ma crossover that the entries are occurring after the crossover.

[NinjaTrader_PatrickH]

Hello snotrag12,

Thank you for your response.

If this is occurring on real-time data set the CalculateOnBarClose option for the Strategy (when enabling or editing the strategy on the chart or Strategies tab) to False.

Please let me know if I may be of further assistance.

[snotrag12]

Actually this is occurring on strategy backtests, and I did have calculate on bar close checked

[NinjaTrader_Bertrand]

In backtesting this would be expected, as only in realtime or Market Replay testing the intrabar formation would be available to work with CalculateOnBarClose = false.

In backtesting the first possible trade location per default is the open of the next candle after which your signal was triggered.

[snotrag12]

In reference to the previous post where Bertrand informed me that in backtesting, the entry occurs at the next candle following the trigger, as opposed to realtime testing, I was wondering if anybody has compared the two, and if it makes a significant difference in outcomes?

[koganam]

In reference to the previous post where Bertrand informed me that in backtesting, the entry occurs at the next candle following the trigger, as opposed to realtime testing, I was wondering if anybody has compared the two, and if it makes a significant difference in outcomes?

Entries are delayed by one bar: it usually makes a significant difference.