Well, for comparison purposes, I decided to run backtests for the monthly method (aggressive preset) on FXDD. Almost without exception, the results were inferior to the Alpari-UK results. Here is the table:
FXDD Backtest Monthly Results
(Aggressive Preset)
|
Month
|
2004
|
2005
|
2006
|
2007
|
2008
|
2009
|
|
Jan
|
-69.3% | -67.6% | -20.7% | +81.7% | +157.5% | +192.2% |
|
Feb
|
+9.3% | -75.6% | -70.2% | -1.2% | +80.4% | +146.6% |
|
Mar
|
+60.6% | -69.0% | -70.8% | +46.2% | +126.7% | -73.7% |
|
Apr
|
-70.0% | +22.2% | -69.6% | -23.0% | +147.3% | +159.9% |
|
May
|
+34.3% | -68.3% | +105.0% | -43.9% | +125.2% | -73.6% |
|
Jun
|
+14.7% | +152.1% | -68.8% | -66.7% | +106.5% | |
|
Jul
|
-77.6% | -70.0% | +100.2% | +47.1% | +132.4% | |
|
Aug
|
+105.2% | -7.9% | +104.7% | -65.8% | +118.9% | |
|
Sep
|
+106.9% | -18.7% | -41.3% | -76.2% | +172.3% | |
|
Oct
|
+50.7% | -76.1% | +90.3% | -64.8% | +255.5% | |
|
Nov
|
+12.2% | +116.2% | -68.3% | +19.8% | +219.6% | |
|
Dec
|
-2.4% | -76.7% | +37.2% | +0.0% | +248.9% |
Here is the comparison on a yearly basis (using the monthly method):
2004: Alpari-UK (+99.8%); FXDD (+34.6%)
2005: Alpari-UK (+88.7%); FXDD (-56.1%)
2006: Alpari-UK (-36.2%); FXDD (-12.7%)
2007: Alpari-UK (-8.6%); FXDD (-37.7%)
2008: Alpari-UK (+31,903%); FXDD (+4,997%)
2009: Alpari-UK (+114.6%); FXDD (+88.5%)
I think there are a couple of conclusions we can make here:
1) As was suggested early on in this experiment, backtesting is never the best way to test a method, and these results prove it. Using the same pricing data, we got strikingly different results with the two brokers. How do you explain that? I’m not sure we can. I am guessing that we could test this with 10 different brokers, and the results would be all over the board.
2) Looking at either of the backtest sets, I think it is safe to say that Euro-Blaster has done unbelievably better in recent months than in the past. It has been hugely profitable over the past 1.5 years. Is this because the EUR/USD pair has become more volatile? Is it a permanent change and can we expect this type of performance to continue? Of course no one knows for sure. One thing is certain, forward testing is your best bet, and especially so with this particular EA.
0 Comments until now.
Comment!