I have a strategy that uses daily values to establish direction, tick bars to enter and 5 min bars to exit.
The strategy is optimised for each instrument and the results are used to drive my live environment.
I've been getting some poor and unexpected results with CL, so I had a look and there are holes in the tick data, for example a gap from 27/9 13:24.52 to 3/10 00:00:01. Trades were taken live but not in the backtest .....................
further inspection shows there are other gaps - so I cannot trust the data or the results of my optimisation runs.
This potentially renders the entire backtest / optimisation circus supported by NT worse than useless.
What can you/I do about this ?
I've got to have quality data otherwise all bets are off.
Comment