CandleStick Body and Close Analysis

Forums ProRealTime English forum ProBuilder support CandleStick Body and Close Analysis

Viewing 5 posts - 1 through 5 (of 5 total)
  • #50206

    I thought I posted this but it disappeared! So apologies if it now appears twice.

    In my search for the elusive ‘edge’ I wanted to analyse what effect the size of candlestick body compared to its range has on the probability of the next candle closing in the same direction. I also wanted to see what effect the close position within the range had. So I wrote a little code to do this. I then added a filter of ATR rising which seemed to give a little edge if set to a very short period. This ATR filter can be turned on or off.

    I added an equity curve based on closing each position at the close of the next candle just to see what the result was. Interestingly most markets seemed to stop being profitable from this simple entry and exit criteria somewhere around the 90’s.

    Although this is technically an analysis tool rather than an indicator I wrote it as an indicator rather than a strategy so I’ve posted it here. Any comments will be appreciated – especially if you spot something wrong as I wrote it after a couple of glasses of wine!

    From a few quick tests on daily markets it seems that an edge of 2 to 4% can be gained over the 50/50 chance of random entry by considering the close position and body size – especially if the ATR filter is added.

     

    3 users thanked author for this post.
    #50208

    Forgot to add the ITF file so here it is.

    #50813

    Thanks Vonasi, very interesting concept. Did you try something with optimisation of the few variables with ProBacktest? You should consider it seriously.

    #84345

    Interesting concept @Vonasi, thanks for the code. Is there anyway to change the length of hold period (for the equity curve calculation), it’s hard coded?
    I’m wondering what would happen if the hold period was lengthened from one bar, sometimes a strong candle intimates bigger impending moves?

    #84346

    I’d forgotten this one!

    It was really just coded as an analysis tool and not a strategy. It could be re-coded for longer hold periods by just changing the candle analysed to be close[n] instead of close[1] and by adjusting the profit /loss calculations to cover the longer period.

    I’m a little too busy with boat maintenance at the moment to do it myself but feel free to give it a go. The code is quite logical and unusually well commented compared to my recent work!

Viewing 5 posts - 1 through 5 (of 5 total)

Create your free account now and post your request to benefit from the help of the community
Register or Login