In the previous article we have been reviewing an interesting strategy for AUD/CAD, and at the end, I asked you to do homework – to suggest how to improve the strategy and how to get more profit out of it.
At the beginning, I want to say that I did not want to hear the “add an indicator X with period Y” reply because it cannot be verified whether the answer is correct or not. There are many variants, and I was more interest in the way of your thinking than in exact solution.
To repeat – what is the strategy doing?
Timeframe: M15 (works with every timeframe up to D1)
LONG input (for going short it is the opposite):
CCI (20) <0
SL: 129 pips
PT: 55 pips
What were your suggestions?
Most replies were concerning the CCI change. Shifting period to for example -50, an addition of the second CCI, etc. There were many variants. From my point of view, those are good ways, but the CCI itself is not enough. The results will not change essentially, and we will not solve the situation when we enter the market on absurd levels.
What was the most original answer?
David came on Facebook with an idea of adding candle formation as a filter. This is the direction I like :).
How did I progress?
To eliminate the moments when the CCI oscillates around zero, I need to know the level for entering the market. So I need to follow the price to decide on when to enter according to its behaviour. Because AUD/CAD is a market that goes rather sideways, it’s a good way to monitor the deviation from the medium value. So I added a simple filter, which must not only meet the CCI condition, but the price for going LONG must be under BollingerBand and for going SHORT above it. I’ve done a little bit of money management because the strategy has changed and the optimal space has shifted.
How does the result look like?
How did the parameters change?
The profit / DD rose to 12 and the number of trades fell to less than half. If we optimise SL and TP values with a 30% change, then all options are profitable, but it is not entirely optimal.
The strategy is very interesting in its simplicity, but I would say that the robustness is not ideal for M15 timeframe. So now we have two options – to test the strategy in some sessions and limit our signals only to specific time-zones or to move to H1 to find out how it works and how to improve it for a higher TF.