Good afternoon
I have programmed a system (see attachment) in which the entry conditions are calculated in 1 minute timeframe, and the exit conditions are calculated in 1 second timeframe.
So, for the entry conditions code I wrote TIMEFRAME(1 minute, UpdateOnClose)
And for the exit conditions I wrote TIMEFRAME(default).
Now I am trying to run a back test for the last 3 years.
My question is:
I wonder if this will this be too much for the processing capacity of the platform, as I want to test my code for three years with part of it in minutes and part in seconds.
Should I expect troubles or should I expect the backtest to run smoothly? Can you please give me any comMent?
Many thanks in advance and best regards,
Carlos
If you are able to display a chart of 3 years of data in 1-second timeframe for that specific instrument, then the backtest should run smoothly from start to end.
… which won’t happen because e.g. FUTU Holdings is (including Extended Trading Hours) 57600 seconds open per day = 288000 seconds per week. With 1M bars (each bar 1 second) – which is the maximum (for the Premium account) – this gives less than 4 weeks.
With only Regular Opening Times it is more (5.5 hours = 19800 seconds per day).
It may run sufficiently smooth for you if you don’t know any better, but in practice it is painfully slow compared to e.g. 1 minute (and again 1M bars). In my experience 5-10 times slower than timeframes based on minutes or higher.
If my math or otherwise is wrong, I am happy to be overruled. 🙂