A Ninjatrader Problem

E

EliteTraderNYC

For those of you who are interested listen to this, read it and weep....

I uploaded NT to a rented VPS just recently and installed the 64 bit version. On my home PC ive got the 32-bit version.

I then exported all of the tick data for ES 2011 to a text file, then reuploaded it on the rented VPS.

I then uploaded my strategy in its exact form to the VPS from my local PC.

I backtested both versions, while the profits were similar for ES, the same exact code in backtest produced a totally a different list of trades occurring at somewhat different times, for example my local would produce a trade at 9:45 AM that closed at 11AM while the server version would produce a trade at 10:30 am and close it at 1PM.

I think the problem probably has to do with the way the Betterrenko bars are created when it loads the tick data.
 
Quote from EliteTraderNYC:

For those of you who are interested listen to this, read it and weep....

I uploaded NT to a rented VPS just recently and installed the 64 bit version. On my home PC ive got the 32-bit version.

I then exported all of the tick data for ES 2011 to a text file, then reuploaded it on the rented VPS.

I then uploaded my strategy in its exact form to the VPS from my local PC.

I backtested both versions, while the profits were similar for ES, the same exact code in backtest produced a totally a different list of trades occurring at somewhat different times, for example my local would produce a trade at 9:45 AM that closed at 11AM while the server version would produce a trade at 10:30 am and close it at 1PM.

I think the problem probably has to do with the way the Betterrenko bars are created when it loads the tick data.
More...

Open historical data manager in both environments, and check the tick data for any day within the date range of the export-import period. Are they the same? If not, did you accidentally make a time-zone shift during your export-import?
 
This problem appears to be being caused by some anomalies in the formation of Betterrenko bars. there are some variations in one set of data versus the other for some reason even though its supposed to be the same set of tick information. It may be because of some type of corruption.
 
Since we are discussing NT, I want to talk about another issue on automating the break even.

So for example, I create an order management strategy for 2 contracts. I set the 1st with x as my stop loss and y as my target. I then set the 2nd contract with x as my stop loss and w as my target (w = 2y). I then set the profit trigger for the 2nd contract to the variable y.

So what should happen is that once target y is hit on my 1st contract, the 2nd contract's stop loss should move to BE. If this is not correct, please let me know. I think this will work fine.

Now final question. Let's assume I believe the highest probability of y being reached is if y = 4. I also believe y = 6 is a very good probability. Now using these values for y and or one of your suggestions, what is the best value to set w at or is using w as 2y good enough to meet standard risk vs reward setting and good trade management?
 
I know this is an old thread but if you guys are still around...

why are there so many "membership temporarily revoked" on futures.io? (formerly Big Mike Trading)
 
You must log in or register to reply here.
Back
Top