Less trades when testing every tick?

[Jowin6789]

Hey all,

I've been strategy testing using on close of candle rather than every tick as it's quicker and at the moment I am just getting the code to do what I want it to do before then properly testing with every tick.

I have tried every tick a few times and find it takes out less trades than when on close of bar and the amount of money made or lost is vastly different. 

Also, i have found a few times that I can change settings around like this, then move everything back to how I had it before and I then get massively different results to what I had before on the same settings.

Is there a guide to setting up the strategy tester? 

And is the reduced amount of trades possibly due to trade executions failing? 

Thanks for any help, this is losing my confidence in the results I get! 😒

[Anthony Garot]

Jowin6789:
Hey all,

I've been strategy testing using on close of candle rather than every tick as it's quicker and at the moment I am just getting the code to do what I want it to do before then properly testing with every tick.

I have tried every tick a few times and find it takes out less trades than when on close of bar and the amount of money made or lost is vastly different. 

Also, i have found a few times that I can change settings around like this, then move everything back to how I had it before and I then get massively different results to what I had before on the same settings.

Is there a guide to setting up the strategy tester? 

And is the reduced amount of trades possibly due to trade executions failing? 

Thanks for any help, this is losing my confidence in the results I get! 😒

The following answer describes the MT5 tester. If you are using MT4, there may be some bits different.

1. Most of my strategies consider a closed bar, so I use "Open prices only" mode for a "rough approach" to testing. Using "Open prices only" is seductive because it's so much faster!

2. When I feel I need more accurate results, I switch to "1 minute OHLC."

Depending upon the strategy tested—yes—the results can be very different between these two modes. Sometimes hugely exaggerated results, like lottery-sized results!

On the other hand, I have some strategies in which the results are quite similar between the two modes, so I feel more comfortable optimizing using "Open prices only," at least while working on the code.

Ultimately, I think the answer is: the finer granularity you use in the tester, the more accurate your back test.