Hello support,

I've tested a multiple instrument strategy on 1hour candels with the historical tester interpolation at the "4 ticks at OHLC option."
This gave me incredible results.
But when testing the same time perdiod with the option "Tick on open" or "Tick on close" the results are very bad.
I figured out that these big differences stem from the bad implementation of the OHLC interpolation.
Historical tester sends open, high, low and close of all instruments at the same time to the strategy, which is in my opinion a look into the future and so no reliable method of backtesting with multiple pairs.
For example, the high of EUR/USD might come 10 minutes before the high of GPB/USD, but historical tester sends both highs at the same time to the strategy.
Clearly this cannot work.

Any feedback on this is highly appreciated.

Cheers,
Juergen

Yes, you are right. The OHLC of different instruments are given at the same time. Interpolation method of Hourly candles are not using any other time frames, so it is not possible to detect the time of Open, High, Low or Close. Using interpolation is a bad idea in your case. Use All ticks.

Thanks for the confirmation.
I have two proposals:

1) Put a comment on the OHLC interpolation documentation that it might look into the future.
2) Make the OHLC option not the default one

Thanks,
Juergen

Hello Support,

I now have also strange results with the "Tick on open" option.
When testing 1hour with this option I still get very good and unlikely results.
But when testing the 1minute period, the results are very poor.
Is it right that "Tick on open" sends all open prices to the strategy when a bar opens?
I tried to simulate this within the 1minute period, but I get different results.
Can you provide a way in which one can simulate the 1hour period together with the "Tick on open" option of the historical tester on the live market?

Thanks,
Juergen