3 Ducks Based Back Testing

Trading Systems Free Forex Trading Systems
1

As some of you are aware, I’ve been doing some backtesting based around Captain Currency’s 3 Ducks Strategy. I agreed to move it away from the Strategy Specific thread because I was firstly moving away from a strict 3 Ducks, and secondly distracting from the main content of the thread.

Anyway at the time I was testing using my own code written in Python, which was working, but I was finding that everytime I wanted to examine the results, I’d have to mess around in a charting package comparing my results, to the charts etc. So I’ve rewritten my stratergy into a NinjaTrader Strategy, so that I can click on results and examine them quicker etc. As a very quick test I started with only 2 parameters that are subjective, stoploss (trailing), and profitstop. Everything else strictly 3 ducks. No trend detection, no extra rules, if the price is above buy, if the price is below sell. (Oh and I tied stop loss to the trade size, so that in theory a stop loss would take 2% of the starting bank)

I test on the last 18months, and optimised on the default 8 currency pairs that are available in NinjaTrader.

Thought people might be interested in the results as a starting point.

AUDUSD - 229 Trades - 199pips Profit.
EURCHF - 58 Trades - 79pips Profit.
EURGBP - 75 Trades - 543pips Profit
EURJPY - 164 Trades - 716pips Profit
EURUSD - 212 Trades - 974pips Profit
GBPUSD - 155 Trades - 389pips Profit
USDCAD - 108 Trades - 746pips Profit
USDJPY - 121 Trades - 399pips Profit

Total For All currencies - 1122 Trades - 4045 pips profit

Now my disclaimer - these are back test results, they run different stops on each currency, and I’m new to NinjaTrader so I’ve possible made a ****up. In the meantime this is probably a good start, to fine tune things, and maybe add some of the rules we were discussing before.

2

The main physiological problem I have with the 3 ducks, is that it loses more often than it wins, actually in it’s most basic state (with my chosen stops) only wins about 30% of the time, what makes it work is the size of the wins and loses. Somewhere in my mind I equate the % of wins, to the length of a losing streak, and hence potential draw down, personally I’d like to see a win rate above 50%. Playing with just the EURUSD again for simplicity. If I look for a 1 pip rise over the last 3 5min candles, I increase the win rate to 57%.

My secondly problem (again physiological) is that there is a 98pip stop loss on the EURUSD (and similar sizes on other currencies), this to me is too much of a risk unexpected news events on a couple of currencies couple be really nasty, and we all know that half the problem with strategies is sticking to them when it’s hurting. My theory is reduce the pain a little and it’s easier to hold on. So I reduced the maximum Stop Loss to 50pips which brings the profitability to 55% which is still within my target.

Anybody got any thoughts?

3

Drum, thank you very much for posting your results.

I just wonder, did you subtract the spread? (for 1122 trades it is more than 30,000$).

Please share your stops and TP’s

Good luck, fxwin

4

Where were your entry points based on here, more the 5m char or the 1h chart? Were entries closest to the right or moved more left? Was price above the 5m sma when going short than entry below the 5m sma and via versa for long? Just curious here I like that you put the effort into it! Talk to you soon friend