Without allowing for any spread/commissions, I got 616 wins, 690 losses and 7 breakevens for a win rate of ~47% and a net loss of 970 pips.
Allowing 1 pip per roundtrip trade for spread/commissions, I got 545 wins, 768 losses and 0 breakevens for a win rate of ~41.5% and a net loss of 2283 pips.
Hence I'd be interested to know how you obtained your win rate of 87%. Otherwise I don't think it's worth writing an EA.
Of course all of this assumes that I understood your strategy correctly, and that there are no bugs in my hastily constructed XLS.
Attached File(s)
EURUSD5.zip
1.6 MB
|
243 downloads
____________________
[EDIT] I found a glitch in the data. Data was missing between April and October, but this doesn't invalidate the win rate.
Anyway, I re-ran the test, using data from a different br0ker server. This time I got 5375 trades between Jan 25, 2016 and Dec 6, 2016.
Without allowing for any spread/commissions, I got 2560 wins, 2783 losses and 32 breakevens for a win rate of ~47.6% and a net loss of 594 pips.
Allowing 1 pip per roundtrip trade for spread/commissions, I got 2227 wins, 3148 losses and 0 breakevens for a win rate of ~41.4% and a net loss of 5969 pips.
The XLS/ZIP file was too big to upload.
Interesting that the win rates were almost identical to the previous test. With an edgeless strategy, one would expect a win rate of ~50% before costs, which is what we get. I'm always reading that the best strategies are simple, but I disagree. I ran this test to satisfy myself that a strategy as simple as this would be unlikely to deliver any kind of systematic profit, given a reasonable size data sample.
____________________
[EDIT2] I ran another test, this time on USDJPY. I got 5396 trades between Jan 25, 2016 and Dec 6, 2016.
Without allowing for any spread/commissions, I got 2658 wins, 2706 losses and 32 breakevens for a win rate of ~49.3% and a net loss of 29 pips.
Allowing 1 pip per roundtrip trade for spread/commissions, I got 2445 wins, 2951 losses and 0 breakevens for a win rate of ~45.3% and a net loss of 5425 pips.
The XLS/ZIP file was too big to upload.
Again, a win rate of ~50%, and a significant net loss overall, after costs are applied.