76 Reply by angat (edited by angat 2012-12-14 23:20:17)

  • Registered: 2012-10-08
  • Posts: 84

Re: Live Trades different than in FSB

Popov wrote:

FSB probably thinks that the entry signal was before the exit. That's why it cancels the entry after position close. Check Journal by bars carefully for that bar and see if there is a canceled entry signal at the corresponding level. In such case, try using scanner with intrabar data. It will improve the backtest accuracy. Is this bar marked "Ambiguous"?

Yes there was a cancelled order. That was exactly it. So why does FSB differ from FST in this instance?

At least I know that this was just a rare occasion where there was just an overlapping entry signal with an exit signal in one bar.

I'm trying to figure out which one was right. I am trading one lot. Was FSB right to cancel that order. Or was FST right to instigate a new trade after the TP was met from the first trade?

I mean it was a valid signal, but FST shouldnt have traded it, if there was an active trade in progress. But if the entry signal came after the first trade was closed, then FSB should have displayed the trade as normal?

I have zero ambiguous trades for this strategy-data set.

EDIT: Are you saying for my 15minute data to be super accurate I really need to have the corresponding 1minute and 5minutes data for that timeperiod?

  • Popov
  • Popov
  • Lead Developer
  • Offline
  • From: Bulgaria
  • Registered: 2006-11-30
  • Posts: 7,988

Re: Live Trades different than in FSB

Is the bar marked "ambiguous"?
Did you try scanner?

Popov's Website

78 Reply by angat (edited by angat 2012-12-14 23:27:28)

  • Registered: 2012-10-08
  • Posts: 84

Re: Live Trades different than in FSB

Popov wrote:

Is the bar marked "ambiguous"?
Did you try scanner?

I tried the scanner, no ambiguous bars. I will add though, that my 1minute and 5 minute data are not relevant to my 15minute data. I could update them and rescan to reveal an ambiguous bar.

So to answer my previous question, FST seems to be working as expected and FSB not (as I havent given it the correct 1 & 5 minute data). My logic is right?

  • Popov
  • Popov
  • Lead Developer
  • Offline
  • From: Bulgaria
  • Registered: 2006-11-30
  • Posts: 7,988

Re: Live Trades different than in FSB

So to answer my previous question, FST seems to be working as expected and FSB not (as I havent given it the correct 1 & 5 minute data). My logic is right?

I think so. Updating intermediate data may improve the backtest.
What if you try different interpolation method?

Popov's Website

80 Reply by angat (edited by angat 2012-12-14 23:53:52)

  • Registered: 2012-10-08
  • Posts: 84

Re: Live Trades different than in FSB

Popov wrote:

So to answer my previous question, FST seems to be working as expected and FSB not (as I havent given it the correct 1 & 5 minute data). My logic is right?

I think so. Updating intermediate data may improve the backtest.
What if you try different interpolation method?

Every interpolation method gave same results, however I'm only using 13 weeks of live data so far and adding to it every day from mt4. Going to update the 1 minute and 5 minute data tomorrow.

Thanks for your help.

  • Registered: 2013-07-03
  • Posts: 106

Re: Live Trades different than in FSB

i have still that problem sad

  • footon
  • footon
  • ◄≡≡≡►
  • Offline
  • Registered: 2009-08-17
  • Posts: 2,969

Re: Live Trades different than in FSB

aysamatik wrote:

i have still that problem sad

What problem?