Backtesting Again

I've obtained tick data for 99% modeling quality, recently, and start backtesting Wallstreet Forex Robot Again. WSFR is almost forgotten by now. Last summer, this EA has taken consecutive losses and drawdown period had last about three months. Although the bot has nicely recovered after that and ended with a profitable year, so many people, including me, has stopped using it and moved to other EAs. So I decided to keep this EA for at least three years with serious money.

Why backtesting WSFR EA again? Well, I aim at better compounding results by finding as many profitable pairs as possible. So far, the pairs below are the potential pairs.

GBPUSD
EURUSD
AUDUSD
AUDJPY
CADJPY
USDCHF
USDCAD
USDJPY

Many people believe GU and EU are the only worthwhile pairs to trade for WSFR, but I found some very interesting fact recovery mode.

Pair        Setting     Capital  Risk  RecoveryMode   Result
AUDJPY  EURUSD   $1,000   10%   False               $4,799
AUDJPY  EURUSD   $1,000    5%   True                $9,320
(Period: Feb.8,2009 to Feb.8,2012)

See? If you enabled Recovery Mode, you would have made it doubled while risking only 5%. I've set AutoMM Max:20%, so the max risk is 20%, but it's acceptable according to my backtest. Moreover, RecoveryMode made the equity curve smoother.

However, this doesn't apply all the pairs.

Pair         Setting  Capital  Risk  RecoveryMode   Result
EURUSD  Default   $1,000   10%   False               $37,458
EURUSD  Default   $1,000    5%   True                $11,811
(Period: Feb.8,2009 to Feb.8,2012)

For a pair with already high winning ratio, RecoveryMode is not so effective. Of course, if you compare with the same 10% risk, it would be a different story. So I'll have to keep the backtesting going and find out which pairs to trade, which setting should be used, and whether RecoveryMode should be enabled or not.