EA Results - Your Thoughts

Hi All,

I have been playing around with some EA backtesting lately. What do you guys think of these results? I have been looking for something reliable, one trade at a time with a decent r/r. Here are the results:

Symbol EURUSD (Euro vs US Dollar)
Period 1 Hour (H1) 2012.01.06 02:00 - 2013.04.30 23:00 (2012.01.01 - 2013.05.01)
Model Every tick (the most precise method based on all available least timeframes)
Bars in test 8312
Ticks modelled 23610262
Modelling quality 88.92%
Mismatched charts errors 0
Initial deposit 10000.00
Total net profit 18639.40
Gross profit 39739.70
Gross loss -21100.30
Profit factor 1.88
Expected payoff 162.08
Absolute drawdown 570.00
Maximal drawdown 3888.10 (19.61%)
Relative drawdown 19.61% (3888.10)
Total trades 115
Short positions (won %) 58 (58.62%)
Long positions (won %) 57 (71.93%)
Profit trades (% of total) 75 (65.22%)
Loss trades (% of total) 40 (34.78%)
Largest profit trade 759.90 loss trade -831.20
Average profit trade 529.86 loss trade -527.51

Maximum consecutive wins (profit in money) 16 (9451.50)
consecutive losses (loss in money) 4 (-2410.60)

Maximal consecutive profit (count of wins) 9451.50 (16)
consecutive loss (count of losses) -2410.60 (4)

Average consecutive wins 3
consecutive losses 2


No comments on this?

Did you use optimization while testing?
I test my strategies with python, so I don’t know exactly how this optimization changes behaviour of your EA, but, what I observed while testing with python was that if you change your parameters (settings of EA) to get maximum results from historical data, it is most likely not gonna work for future results.

Try using constant arguments and backtest across multiple currencies and see if results are consistent over a long period of time. Then forward test with demo, only then you can make a conclusion.

Thanks yhlasx, do you mind explaining how you backtest with python?

I have tried this EA on other pairs, but it’s only profitable on eur/usd

As far as the optimisation goes, I tried not to play with too many variables, and now I have it running on a demo account, so will see how it goes.