[B]Philosophy behind the system[/B]
First off, let me say I’m just starting my Forex journey. I have not yet placed a single trade for real money. I’ve gone through the Babypips School of Pipsology and have read pages upon pages of this forum. I’m still trying to figure out what kind of trader I am. I decided to enter this system in the contest because I want to see some more money go to charity.
I’m drawn to longer time frame price action trading because it’s like an art form. With enough experience and screen time, one can develop a trading method with a high winning percentage, and who among us doesn’t like to win 80% of the time.
On the other hand, a mechanical system with even a small positive expected value can make money like the casinos make money. They take a small edge and exploit it over as many trials as they possibly can. The law of large numbers kicks in and their 1% edge turns into billions of dollars over time.
MA Fracker has been an exercise for me to see if I can create a purely mechanical system with a positive expected value based on historical data. I believe my back test results show that there might be some merit in pursuing this system further. If I stop seeing squirrels and shiny things, I may continue backtesting and begin forward testing.
[B]Indicators and parameters[/B]
Currency pair – USDJPY
Timeframe – H1
RSI(14), levels at 30, 50, and 70
Fractals
EMA(5)
SMA(13)
Why 5 and 13? Well, they’re Fibonacci numbers so they must have mystical, magical powers, right?
[B]Entry[/B]
LONG – Enter at market on candle close when EMA(5) crosses from below to above SMA(13) by at least one full pip* AND RSI(14)>50. DO NOT ENTER is RSI(14)>70. *Example - 80.009 IS NOT above 80.00 but 80.010 IS above 80.00
SHORT – Enter at market on candle close when EMA(5) crosses from above to below SMA(13) by at least one full pip* AND RSI(14)<50. DO NOT ENTER if RSI(14)<30. *Example - 79.991 IS NOT below 80.00 but 79.990 IS below 80.00.
A note about entries…a cross is not valid if EMA(5) has not been on the opposite side of SMA(13) by at least one full pip. Example, say EMA(5) is 80.05 and SMA(13) is 80.00. EMA(5) then goes to 79.996 on the next candle and SMA(13) is still 80.00. On the next candle, EMA(5) goes back up to 80.05 and SMA(13) stays at 80.00. This IS NOT a valid long signal because EMA(5) never went one full pip below EMA(13), therefore I never consider it below and if it’s not below it can’t cross above. I know it sounds complicated, sorry. Read it a few times and maybe it will make sense. This rule seems to prevent SOME whipsaw.
[B]Stop Loss
[/B]
On a LONG position, place your stop 1 pip below the most recent down fractal.
On a SHORT position, place your stop 1 pip above the most recent up fractal.
I chose fractals because it seems like a fairly logical place on the chart where, if broken, proves that the trade is bad or wrong.
[B]Take Profit[/B]
I tested two different take profit methods, use either one or a combination of the two, back test results are at the end.
METHOD 1 – Take profit at the same distance as your initial stop loss.
METHOD 2 – Move stop loss to break even when price moves the same distance as your initial stop loss and implement a trailing stop that is the same size as your initial stop loss.
[B]Early Exit[/B]
Close all positions if EMA(5) crosses to the opposite side of SMA(13) by at least one full pip.
Close LONG positions if RSI(14) goes above 70 and then crosses back below 70.
Close SHORT positions if RSI(14) goes below 30 and then crosses back above 30.
[B]Backtesting Results[/B]
I manually back tested this system on a Pepperstone MT4 platform using the F12 feature to step through the charts one candle at a time. I’m a newbie, so when I started this I didn’t think to include spreads in my calculations. I only took trades between 06:00 and 22:00 US Pacific Time (GMT -8) because that’s when I’m not sleeping. I also did not take any trades on Friday morning Pacific time, basically no trades on Friday during NY session. Results will most certainly be different on a back testing data set that takes trades 24 hours a day.
These results are based on 173 trades taken from October 2009 to October 2010. I had intentions to back test from 2009 to present but those damn squirrels with shiny objects… I have not forward tested this at all.
METHOD 1
Wins – 83 (47.98%) Losses – 86(49.71%) Break even – 4(2.31%)
Total pips won – 2568 Total pips lost – 1916 Net pips – 652
Avg. win – 30.9398 pips Avg. Loss – 22.2791 pips Avg. Risk(SL) – 39.3 pips
Expected value per trade – 3.7688 pips
Longest winning streak – 5 trades Longest losing streak – 6 trades
METHOD 2
Wins – 66 (38.15%) Losses – 86(49.71%) Break even – 21(12.14%)
Total pips won – 2727 Total pips lost – 1916 Net pips – 811
Avg. win – 41.382 pips Avg. Loss – 22.2791 pips Avg. Risk(SL) – 39.3 pips
Expected value per trade – 4.6879 pips
Longest winning streak – 5 trades Longest losing streak – 6 trades
Nick
