At the same time, internaly for candles we calculate the candles using the HA system ( while still using in live the normal candles). This way we can assure that even if we use HA candles, we avoid repainting, and its legit.
We first calculate the HA candles based on the EMA 50 values, and after that , we use that candle properties to apply to EMA 100.
Once we have that, for entries we have the next conditions :
sell = o2 > c2 and o2[1] < c2[1] and time_cond buy = o2 < c2 and o2[1] > c2[1] and time_cond
For sell : Our open from HA 100 is bigger than Close from ha 100, and the previous open is smaller than previous close For long : Our open from ha 100 is smaller than close from ha 100 and the previous open is bigger than previous close.
Then we have 2 options :
If we wnat to go only long , which is my prefered version ,or the original one where we go both long and short.
I found that the best results are in general around bigger timeframes, 1h+ , 3h works the best so far on my tests.
For exit we have 2 versions : 1 lets say we had a long signal, as soon as we have a short signal we close the trade. Viceversa for short. 2. Is based on price % movement. In this case I use 7.5% price movement of asset.
We have no TP in use for this system.
For the purpose of this test I use 10.000 $ account. For test I use 100% of it, without any leverage. I use the SL based on price movement , which is a very risky tool, since it can fluctuate even at 20-30% of our capital. For comission I used 0.1% for each deal, and a slippage of 5 points.