Backtesting result on Grid strategy with DPG

Backtesting result on Grid strategy with DPG

Queenie Wu                                           29 April 2020

Today, as usual, I am testing CM & Undeadearthworm EA.  I am thinking different settings for EURUSD.  How to win the trade?  In H1 diagram, a wave like pattern always happens.  Should I use breakout strategy? Trendy strategy? Reversal Strategy?

Yesterday, I tried testing the trend following strategy using the indicator of “sk”, Ultra-Signal indicator and EXP iCustom EA.   The result is OK, with profit factor at around 2.  

Today, maybe, I can try using the Grid strategy.  The problem is how to minimize the risk?  If the currency moves too far away in one direction without pullback, the whole Grid will collapse.  In recent 2 years, EURUSD was moving gradually downward, maybe, we can try sell only Grid strategy.

I searched the old sets on Dark Personal Grid EA that I have.  I modified one of the old sets with the following settings:

Currency: EUUSD H1

Backtest pathform: IC Market  Period: 2018 – April 2020

Version 1:

Lot size 0.01   Trade direction: Sell only

Grid coefficient 0.5. (lot size: 0.01 0.02 0.03 0.04 0.05 ….. )

Minimum distance to start the grid at 100. 

Distance multiplier 0.5

Trading day: Mon, Tue, Wed

Trading hours: 1:30 – 8:00 (Broker hours)

Broker GMT: 2

Unexpectedly, from 2018 to 2020 Apr, I got a 45 degree upward sloping straight line.  I cannot believe my eye.   The profit factor is higher than 7

Version 2: (I only changed the Grid coefficient to 2) Much higher risk as much more deposit is needed.

Lot size 0.01   Trade direction: Sell only

Grid coefficient 2 (lot size: 0.01 0.03 0.05 0.07 0.09 ….. )

Minimum distance to start the grid at 100. 

Distance multiplier 0.5 (See diagram below)

Trading day: Mon, Tue, Wed

Trading hours: 1:30 – 8:00 (Broker hours)

Broker GMT: 2

Unexpectedly, from 2018 to 2020 Apr, I got a 45 degree upward sloping straight line.  I cannot believe my eye.  The profit factor is higher than 13.8!!  A $1000 account grew to more than $8000 in two years. 

Potential problems:

The above setting is suitable for the market situation after 2018 only.   EURUSD is in general on a downward trend.  For the year 2017 and 2016, the backtest results failed.  The main reason is due to the long term upside movement in one direction.  This, of course, opposed to the “Sell only” strategy.

As only “Sell only” trades were placed in the above testing, am I curve fitting the past data?  Should I believe the above test results and put the EA in real account??

Possible solution:

Maybe, only one direction trade should be allowed based on the market direction during the trading hours.  Once market direction changed, we have to manually change the EA to the oppose trading direction.  Then, the EA will become partially automatic only as I have to make the decision of the big trend direction.