Entries Exits Dual TimeFrames problem

[saltminer]

Hi guys,
I have a strategy with two bar series, both daily, with stocks loaded into them. The strategy requires measuring relative movement between the two stocks, so I have all my processing in BarsInProgress 1. If I use BIP 0, of course, the data I get from BIP 1 is not up to date yet.
So far so good.
When I take a position, I take it in both stocks at once, like this

(... in BarsInProgress == 1)

EnterShort(0,Inst0Quantity,EntryNameString);
EnterLong(1,Inst1Quantity,EntryNameString);

I expect the strategy to enter at the next day's Opening price.
However, only the instrument in BarsInProgress 1 enters the next day. Instrument 0 takes an additional day to enter. Exits are similarly delayed on BIP 0.
This is clearly not what I am after, so perhaps you can advise what might be going wrong.
Here is data from the Output Window using TraceOrders...

23-Feb-93 (BIP==1) Entry: Inst 0 Short 1761 shares, Inst 1 Long 842 shares.

23-Feb-93 12:00:00 AM Entered internal PlaceOrder() method at 23-Feb-93 12:00:00 AM: Action=SellShort OrderType=Market Quantity=1,761 LimitPrice=0 StopPrice=0 SignalName='T1' FromEntrySignal=''
23-Feb-93 12:00:00 AM Entered internal PlaceOrder() method at 23-Feb-93 12:00:00 AM: Action=Buy OrderType=Market Quantity=842 LimitPrice=0 StopPrice=0 SignalName='T1' FromEntrySignal=''

Execution='NT-00000' Instrument='RIOTCT' Account='Sim101' Name='T1' Exchange=Default Price=11.88 Quantity=842 Market position=Long Commission=0 Order='NT-00001' Time='24-Feb-93 12:00:00 AM'
Execution='NT-00001' Instrument='BHPTCT' Account='Sim101' Name='T1' Exchange=Default Price=5.67 Quantity=1,761 Market position=Short Commission=0 Order='NT-00000' Time='25-Feb-93 12:00:00 AM'

As you can see, the TraceOrders of the Entries are the same date, 23-Feb. THE EXECUTIONS ARE DIFFERENT DAYS. This is not a stock data issue as it happens every time on every trade.

Please help.
Cheers,
saltminer

[NinjaTrader_JoshP]

saltminer,

This is what happens in backtesting. When you submit from BIP1 to BIP0, the message is received on the next bar which becomes the signal bar and then is filled on the following bar which is the trading bar.

[saltminer]

So the way Ninja is designed I have the choice of either...

a) Calculating in BIP==0 which I can't because I don't yet have up-to-date data for BIP==1

or

b) Calculating in BIP==1 which I can't because orders are delayed a bar

Correct?


What about if I write my own filltype? Will BIP be available inside that processing? I could set BIP==0 to open on current day, not next day (I've already written logic like that successfully for something else one time, although it didn't try to access BIP.)

stymied saltminer

[NinjaTrader_Bertrand]

Saltminer, your best option for backtesting would be to add a finer stream for simulation of intrabar fills as explained in this sample - http://www.ninjatrader-support2.com/...ead.php?t=6652

[saltminer]

Thanks, yes I get that concept, but I am backtesting over 15 years of daily data. I don't have any data finer in granularity than daily.

I suppose I could add a third bar series that is the same as BIP[0], take the trade on that. At least then it will fire on time, even if I can't see it displayed on the chart.

But back to another question... is BarsInProgress available inside the Fill processing code? I could modify a FillType if it is....
Thanks.
saltminer

[NinjaTrader_JoshP]

saltminer,

No, you work within one context at a time. I suggest you try and get additional granularity since that would be the preferred way to proceed.