stevehopwoodforex.com
https://www.stevehopwoodforex.com/phpBB3/
Print view

Preemptive Zig Zig
https://www.stevehopwoodforex.com/phpBB3/viewtopic.php?t=2792		 Page 4 of 4
Author:  snailbeard [ Tue Oct 22, 2013 12:15 pm ]
Post subject:  Entry Filters
One of the most useful entry filters is blocking trades which would cross a D1 Bollinger band. Checking the proximity to a major BB is my top priority. The improvement on the profit factor by this filter alone takes this trading method from a modestly profitable method to a significant level of success:

We have reached our target of profit factor >= 1.5
However, this has had little impact on the drawdown, which is due to consecutive losses.
The next priority is to find the best way to reduce drawdown (without reducing profitability)

Here is the new report from testing over one year again:

Bars in test 405775
Ticks modelled 808278
Modelling quality n/a
Mismatched charts errors 0
Initial deposit 10000.00
Total net profit 1974.76
Gross profit 4851.76
Gross loss -2877.00
Profit factor 1.69
Expected payoff 17.17
Absolute drawdown 209.44
Maximal drawdown 661.00 (6.32%)
Relative drawdown 6.32% (661.00)
Total trades 115
Short positions (won %) 56 (62.50%)
Long positions (won %) 59 (79.66%)
Profit trades (% of total) 82 (71.30%)
Loss trades (% of total) 33 (28.70%)
Largest
profit trade 127.28
loss trade -115.14
Average
profit trade 59.17
loss trade -87.18
Maximum
consecutive wins (profit in money) 13 (794.40)
consecutive losses (loss in money) 7 (-593.76)
Maximal
consecutive profit (count of wins) 794.40 (13)
consecutive loss (count of losses) -593.76 (7)
Average
consecutive wins 4
consecutive losses 2
Author:  nonlinear [ Tue Oct 22, 2013 5:58 pm ]
Post subject:  Re: Preemptive Zig Zig
Stay at it, SB. Enjoy reading your progression on this.
Author:  snailbeard [ Wed Oct 23, 2013 10:22 am ]
Post subject:  Bugs bugs bugs
Thanks for the encouragement, there have been several occasions when I'll really felt this was beyond me and I was not the only one here at the mollusc farm thinking that!

Bugs in the simulation:

I'm not sure whether this bug improves results or makes them worse.
However, the Virtual Pending Orders are supposed to expire each time that we look for a trade setup.
In this instance it results in a losing trade.
Hopefully it won't be difficult to figure out what is going on...

Thanks to CrappyTester for not being completley useless:

2012.06.04 22:45 ZZ VPO SELL ORDER SET UP!
2012.06.04 22:45 VPO: Pair: AUDUSD, bActive: 1, Type: 1, Lots: 0.100, Price: 0.97203, SL: 0.97503, TP: 0.96603, MTPT: 0, TS: 2012.06.04 22:45, TE: 2012.06.04 23:00
2012.06.04 22:45 AUDUSD ZZReadyStatus : -1
2012.06.04 22:45 updateZigZagReadyFlag(): iZZReadyStatus != 0
2012.06.04 22:45 useVirtualOrders && (iZZReadyStatus != 0) && bM15BarChanged
2012.06.04 22:45 VPO: Pair: AUDUSD, bActive: 1, Type: 1, Lots: 0.300, Price: 0.97203, SL: 0.97503, TP: 0.96603, MTPT: 0, TS: 2012.06.04 22:45, TE: 2012.06.04 23:00

2012.06.05 00:00 Testing: It's a new day - make sure trading is enabled.
2012.06.05 11:14 bVpoHasBeenTriggered
2012.06.05 11:14 open #13 sell 0.30 AUDUSD at 0.9716 ok
Author:  snailbeard [ Wed Oct 23, 2013 11:15 am ]
Post subject:  Reduced drawdown
And the answer appears to be
a) some loss of winning trades
b) reduction in the drawdown

Profit factor 1.74
Maximal drawdown 509.52 (4.49%)
Relative drawdown 4.52% (472.00)

There is some scope for optimisation for the duration of VPOs.

Can we squeese another 0.5% or more off the drawdown?
Author:  snailbeard [ Wed Oct 23, 2013 6:28 pm ]
Post subject:  Red letter day
It's a Red Letter Day!
- everything just seems to fall into place today...

I can stop tweaking for a while now and start running this on several demo accounts:
There has been a drop in the number of trades but the profit factor is up to 1.8 and more importantly:
drawdown is just over 3% so it should be able trade more pairs simultaneously :D

So this should work as a manual trading method or automated,
but perhaps it is time to move further development of this method to the auto-trading section.

Although, I find weekly and daily pivots useful when trading from charts, they are not used at all in the simulation. There is plenty of scope to improve this method further. The overlapping trades are currently disabled and various settings should be run through the optimisation process each week to curve fit to current market conditions. Also many of the rejected trades can be turned into scalp trades.
I currently split the trades into three parts - total guess - so there is a need to experiment with lots of variations on this.

Don't forget that you can project the Fibonacci lines from higher timeframes into H4 and M15 - I was planning to include them for locating strong support and resistance levels, but it is left for the future now.

Also a variation of the method would look for Bollinger band breakouts.
Another variation would be the location of channels by finding lines that intersect multiple zigzag pivots.

Happy Trading!

 Bars in test 405775
Ticks modelled 808278
Modelling quality n/a
Mismatched charts errors 0
Initial deposit 10000.00
Total net profit 1741.78
Gross profit 3925.36
Gross loss -2183.58
Profit factor 1.80
Expected payoff 19.14
Absolute drawdown 57.00
Maximal drawdown 339.00 (3.26%)
Relative drawdown 3.26% (339.00)
Total trades 91
Short positions (won %) 39 (64.10%)
Long positions (won %) 52 (78.85%)
Profit trades (% of total) 66 (72.53%)
Loss trades (% of total) 25 (27.47%)
Largest
profit trade 127.28
loss trade -115.14
Average
profit trade 59.48
loss trade -87.34
Maximum
consecutive wins (profit in money) 9 (494.40)
consecutive losses (loss in money) 3 (-297.00)
Maximal
consecutive profit (count of wins) 494.40 (9)
consecutive loss (count of losses) -297.00 (3)
Average
consecutive wins 3
consecutive losses 1
Author:  intcust [ Wed Oct 23, 2013 7:24 pm ]
Post subject:  Re: Preemptive Zig Zig
great read just came across it keep up good work
Author:  nonlinear [ Thu Oct 24, 2013 3:40 pm ]
Post subject:  Re: Preemptive Zig Zig
SB, I am going through the same process as you on another system, but you are months ahead of me. I am currently at a PF 1.79, with a 35%DD, but that is without implementing a SL to filter out some really bad trades. Try the same test on another correlated pair to see if it holds up and also over different periods of tick data.
All times are UTC Page 4 of 4

Powered by phpBB® Forum Software © phpBB Limited