I have volunteered to put up the cash to have the EA produced by Mr Pip. We are working behind the scenes to see what works best. I will spare no expense to put together a top notch EA based on simple yet effective approaches.

I am reading and developing in silence.......

Don

__________________

"A man wears one watch and can tell time. A man wears two watchs and doesn't know what time it is"

old chinese proverb.....

www.ForexRobotTrader.com

Note: 100% no loss is obtained when no type of stop loss is employed

If I see a signal late, sometimes I will jump in the trade. like this morning for instance.I went long on the gbp/jpy with a limit at the next pivot point, and this is what happened.

Attached Images

2007-09-11_150823.jpg (10.5 KB, 770 views)

I have a new signal on the gbp/jpy, but i'm still waitin for the right candle-stick pattern to jump in the trade.

the new trade will be a short.

EDIT: I have found a bug in EA that makes it open multiple orders, so these results are invalid. I am working to correct the code.

Well, I have taken the original EA that zmax wrote, stripped out all unnecessary code and ran it on H1 GBP/JPY. Next will be to try getting out at the first zig zag. I ran it 1/1/07 - 9/7/07, and it lost money. However, I ran it 1/1/06 - 9/7/07 and it returned big.

There's a huge drawdown, so I added code to optimize the stop loss - testing that now. The number of consecutive losses would test even the hardiest of traders I think. But this system is profitable - extremely so even with its simple rules. 500% in 21 months? Yes, a huge drawdown, but did I mention - 500%?

Bars in test 15298
Ticks modelled 5665635
Modelling quality 90.00%
Initial deposit 10000.00
Total net profit 53287.95
Gross profit 98153.71
Gross loss -44865.75
Profit factor 2.19
Expected payoff 59.01
Absolute drawdown 4774.05
Maximal drawdown 94583.25 (71.86%)
Relative drawdown 71.86% (94583.25)
Total trades 903
Short positions (won %) 627 (1.12%)
Long positions (won %) 276 (21.01%)
Profit trades (% of total) 65 (7.20%)
Loss trades (% of total) 838 (92.80%)
Largest
profit trade 4381.43
loss trade -480.79
Average
profit trade 1510.06
loss trade -53.54
Maximum
consecutive wins (profit in money) 34 (87310.59)
consecutive losses (loss in money) 587 (-29735.98)
Maximal
consecutive profit (count of wins) 87310.59 (34)
consecutive loss (count of losses) -29735.98 (587)
Average
consecutive wins 6
consecutive losses 76

Thanks,
burn0050

__________________
He who laughs last thinks slowest

burn0050,

Man you have to use a filter if you want the EA to be successful all the time. Also remove the "start/end trade" at candle open and let it make the trades when the arrow+nonlag appears + filter confirmation.
It has to enter the trade as fast as possible, because u might lose lots of pips if you let it waiting.


Thanks, Zmax

Hey Rob, would it be possible for you to also produce a NonLagZZ indicator that also shows its repainted lines in a different colour? Because the system comprises of not just the arrow, but the line as well. So if we knew when both repainted, it would make our backtesting much more powerful.

Thanks.

I have been working on that and so far no luck. I modified the indicator to place arrows as I did with the zigzag-pointer indicator and the arrows only appear when the zigzag reverses.

I will post it when I get it to work.

Robert

I have been working on an EA for this method. The biggest problem is when trading by the origianl rules of placing the trade as soon as the arrow appears the backtest takes a VERY long time.

When trading on closed candles, the EA can use the following code to speed things up.

if (dtBarTime == Time[0]) return(0);
dtBarTime = Time[0];

This code allows the EA to only check the indicator once per candle.

With the original rules it is necessary to check every tick.
The arrow can occur in the first few minutes of a 1 hr candle.

I also believe the stoploss rule needs to be clarified. Is set to 30 pips above the high that created the arrow, some hour candles will have 2 or 3 losses before the close of the candle.

So far the EA is profitable using open or closed candles.

I am also looking at CCI to confirm the exit. One buy trade visually closed for 700+ pips at the first arrow but CCI was still rising signaling to stay in the trade. There were 4 more arrows saying to exit the trade but CCI kept the trade open. The final close was for over 1400 pips. If the trade had been closed, the arrows would have caused loss trades because they signaled a short. Each arrow was on the nonlag ZigZag line.

Robert


Robert

Thank, Marko for giving posting Move like a Shadow... Indicator. I just tested it out and I saw it repaints and that unfortunate. So here I am out again trying to find a filter.

Zmax,

I removed all of the candle code - I'm not sure what you mean by 'remove the "start/end trade" at candle open'.

I wanted to see what it would do without any filters first, because there could be ways to optimize it using stop loss and money management first. Even so, it has an awful lot of consecutive losses, and the only way I see around that is to use a filter. I also wonder what it would be like if the system didn't trade during NFPs.

I haven't done any forward testing with this yet; I wanted to see what I could get to with back testing. Live testing is a whole different ball game. I am getting some errors in the journal - I don't think it's opening and closing all the orders correctly.

Thanks,
burn0050

__________________
He who laughs last thinks slowest