Simple Forex CobraSystem Thoughts Please? | Page 5 | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

SIMPLE EA for a simple method needed!! 2 replies

Trading is this Simple? Finally Explained simple and clearly! 46 replies

Wasabiboys MT4Stats Good I think, tell me your thoughts please 3 replies

Please post a simple editable EA for moving avarages 2 replies

Programmers: Simple SendMail Line. Please Help 15 replies

Attachments: Simple Forex CobraSystem Thoughts Please?
Exit Attachments

Simple Forex CobraSystem Thoughts Please?

  • #81
  • Mar 17, 2012 6:48am Mar 17, 2012 6:48am
this indie overlooks adx over several timeframes... it might be a help but you can't backtest it ..so it's not easy to set some rules for this... if all are green your are in the middle of a full blown trend ..at the beginning of the trend it is okay if some are blue...but red values on the lower timeframes are warning signs for a trend reversal usually
Attached File(s)
File Type: ex4 Matrix FxTraderPro Trend or Range.ex4   16 KB | 517 downloads
  • #82
  • Mar 17, 2012 7:34am Mar 17, 2012 7:34am
I can see potential, GBP-JPY actually quite well behaved on Friday, as it is considered a volatile pair.

Will watch in the coming weeks, not a coder, but from what I've seen discussed in the "Experts Tool Box" thread while making the EA for "Trading Made Simple" method, this would be a piece of cake to make an EA and backtest.

Attached Image (click to enlarge)
Click to Enlarge Name: 0001.jpg Size: 142 KB
It takes a lot for me to IGNORE you, but you can try.
  • #83
  • Mar 17, 2012 7:43am Mar 17, 2012 7:43am
Quoting yongezy
When do we move SL to BE?
Can someone point to me in the manual?
Of course I will read again to see if I can locate it.

Reason ask is to prevent from losing a trade like the image.
Any idea?
Page 35, basically, when 10 pips ahead, jump trail 10 pips until at Break Even, then revert to the EXIT strategy.

That trade you show would be fine, not even close to being stopped out

Attached Image (click to enlarge)
Click to Enlarge Name: 0002.jpg Size: 163 KB
It takes a lot for me to IGNORE you, but you can try.
  • #84
  • Mar 17, 2012 8:06am Mar 17, 2012 8:06am
I see trades everywhere, too bad I have seminar all next week



but I will catch up with your trades each night, good luck

Attached Image (click to enlarge)
Click to Enlarge Name: 0003.jpg Size: 136 KB
It takes a lot for me to IGNORE you, but you can try.
  • #85
  • Mar 17, 2012 5:00pm Mar 17, 2012 5:00pm
Here I found a series of 6 entry signals in a row within 10 days, where I would have probably lost money.

Does anyone have any idea how to avoid those trades?
Attached Image (click to enlarge)
Click to Enlarge Name: eurusd_2701-06022012.jpg Size: 216 KB
  • #86
  • Mar 17, 2012 8:51pm Mar 17, 2012 8:51pm
Quoting Rob_Trader
Here I found a series of 6 entry signals in a row within 10 days, where I would have probably lost money.

Does anyone have any idea how to avoid those trades?
Have you checked which trading sessions those candles are in? Seems to me that most are Asian session, not the best time of day to trade.

And at least for one of them, moving the stop loss to break even after 40 pips would prevent any loss at all.

Hope that helps, E.
It takes a lot for me to IGNORE you, but you can try.
  • #87
  • Mar 17, 2012 10:24pm Mar 17, 2012 10:24pm
Quoting Erebus
Page 35, basically, when 10 pips ahead, jump trail 10 pips until at Break Even, then revert to the EXIT strategy.

That trade you show would be fine, not even close to being stopped out

Erebus

Many thanks. I got it too. Yet to understand how-to use "trailing stop" and nor the concept. I set SL and BE manually.
  • #88
  • Mar 18, 2012 5:12am Mar 18, 2012 5:12am
Quoting Rob_Trader
Here I found a series of 6 entry signals in a row within 10 days, where I would have probably lost money.

Does anyone have any idea how to avoid those trades?
The Manual for this System is very well written with good rules.
However, it was a bit of a let down on p 16 to read this:
Attached Image


and then find that Price would need to cross both zones before entry.
Was expecting a real innovative approach, such as using the deviation from mean to trigger an entry.
As is becoming obvious from recent posts, the late entry results in missing out on some good moves.
Have looked at 1H charts in search of a better signal and it appears that an earlier indication of trend change is when Price crosses a faster moving average such as 5 & 12 LWMA with Weighted close.
As can be seen on Attached chart, Cobra signals in Yellow/Green plotted with the 5&12LWMA cross signals.
Just need to filter out the whipsaws inbetween.

Any ideas welcome.
Ep
Attached Image (click to enlarge)
Click to Enlarge Name: cobra1.gif Size: 70 KB
  • #89
  • Mar 18, 2012 6:43am Mar 18, 2012 6:43am
Quoting EaglePip
Just need to filter out the whipsaws inbetween.
Any ideas welcome. Ep
Unfortunately you cannot have it both ways, very early in a trend, and no signal if it's not going to be the start of a trend. That would be akin to magic. All we need is a system that gets a big chunk of the trending move but doesn't lose too much in a range. There will be losses no matter what you do.

I did a manual backtest using the original rules found in the manual - 6 months, 2 pairs. Of course, 6 months is very little but the results are good. The key is to refrain from adding new rules or indicators when you have several consecutive losses (it did happen and it will happen again). If you come up with rules that keep you out of those whipsaws you also lose many of the nice early entries to follow a trend. If on the other hand you want to catch the big moves even sooner you get killed by whipsaws.

I'm not against changing rules or adding new indicators, but any such change should come only after some serious backtesting to confirm that overall, over hundreds of trades, it makes the system more profitable. What happens instead is that as soon as people notice a few losses they try to filter them out, unaware of the fact that their filter leads to more problems than it solves.
  • #90
  • Mar 18, 2012 7:09am Mar 18, 2012 7:09am
LazyPawn,
somebody once said that a strategy which looks promising must be backtested at least 1 year (maybe better 2,3,4 or more years).

I have seen so many "good looking" strategies which produced profits
for the last 2 weeks ( and usually that is how far some of us look back in the charts ) , but that is gambling.

Forward testing a strategy for several months on a demo takes way too long, so the only solution is an EA .

If the EA cannot show profits in the past, the strategy most probably will not produce profits in the future.

Well, I coded the Cobra-system into an EA and here is the backtest
on EU H1 for 2011 on Alpari-NZ broker.
http://i.imgur.com/xW0vT.jpg

I don´t like what I see there.
The EA is here:
http://www.forexfactory.com/showthre...26#post5486626

I´m afraid we need better filters and rules .
The EA has the option to use tradinghours filter. The backtest was without
it, I just finished a backtest with only trading London and NY session and
the results are a bit better but still horrible.
http://i.imgur.com/dGLI2.jpg
  • #91
  • Mar 18, 2012 7:22am Mar 18, 2012 7:22am
Quoting EaglePip
Have looked at 1H charts in search of a better signal and it appears that an earlier indication of trend change is when Price crosses a faster moving average such as 5 & 12 LWMA with Weighted close.
As can be seen on Attached chart, Cobra signals in Yellow/Green plotted with the 5&12LWMA cross signals.
Just need to filter out the whipsaws inbetween.
Your MA-cross look very good.
For the whipsaws we could create a locking - unlocking method.
What is that ?

Lock = no SL but at the SL where SL would sit, we place a pending order
in opposite direction of the active order with same lot-size .

If market triggers that level, we have 2 active orders with same lot-size.
One is long, the other is short.

From there market can go where it wants to go.
If it goes up, somewhere it will reverse and form a new Short-entry.

At that level we could close the Long order and go Short --->
we would have 2 active Shorts then.

Now we only need a certain move downwards to break even and there
we can close both Short trades.

In other words: Whenever price would hit SL, we will use the lock-unlock method to achieve a Breakeven .

Just an idea.
  • #92
  • Mar 18, 2012 8:30am Mar 18, 2012 8:30am
Quoting fxtr51
LazyPawn,
somebody once said that a strategy which looks promising must be backtested at least 1 year (maybe better 2,3,4 or more years)....
Absolutely. Most say at least 200 trades. I backtested audusd and gbpjpy (the pairs I like most) but it was a manual backtest and only for 6 months with no trades taken before Frankfurt open. It looked promising, but using your EA I get a different picture. Not so rosy. If your EA works exactly according to the rules then we have a loser. Or no better than random, let's say. I will have to test more.

Regarding the lock-unlock suggestion - that is a well known form of futile hedging, it simply doesn't work. It has been discussed many times and this is not the place to start that again.
  • #93
  • Mar 18, 2012 8:55am Mar 18, 2012 8:55am
Quoting LazyPawn
Regarding the lock-unlock suggestion - that is a well known form of futile hedging, it simply doesn't work. It has been discussed many times and this is not the place to start that again.
I missed such discussions, therefore I am still in illusions maybe.

If a coder is interested, I will send him the source code to check if the trading-logic is correct. I offered that long ago but nobody asked.
As long as nobody finds errors in my EA, we have nothing else and
have to assume that it is trading according to the rules .....
and as you said: then we have a long-term losing strategy.

Time to jump to another thread ?
  • #94
  • Edited 1:14pm Mar 18, 2012 12:06pm | Edited 1:14pm
Quoting fxtr51
LazyPawn,
somebody once said that a strategy which looks promising must be backtested at least 1 year (maybe better 2,3,4 or more years).

I have seen so many "good looking" strategies which produced profits
for the last 2 weeks ( and usually that is how far some of us look back in the charts ) , but that is gambling.

Forward testing a strategy for several months on a demo takes way too long, so the only solution is an EA .

If the EA cannot show profits in the past, the strategy most probably will not produce profits in the future.

Well,...
Hi fxtr51.I think we need a higher timeframe for filtering those false signals and make it safer to open a trade.. Sorry I'm super newbie here it's been 3 months since I'm studying forex. from basic Candle listick,pivot,stoch,volume,avoid news & etcs. for example if we wanna trade at 1H at least we a need a confirmation from 4H or 1D

For Example The Crazy Ranging and many false signals on 1H Market from 19 January 2012 to 22 February 2012. If we Check on 1Day Chart It's Super Clearly the market At Indecision

I have a request for you. Could You Please make an EA Its Job only for Closing all open trades, that we opened manually?

Rules Close Trade:
Type:Completed Bar (Only all completed bars count)
Close(Buy) Strategy: When Close Price below(Lower than) EMA 12 Low
Close(Sell) Strategy: When Close Price Above(Higher Than) EMA 12 High

Example: I Manually open 1 trade buy EURUSD at london time followed by another new open 2 trades buy EURUSD (market still in perfect condition for buy) at new york time. so at the time I have 3 trades Open.. and then the close trade signal occurred at the tokyo time the close price (Type Completed Bar) lower than EMA 12 LOW. So
all those 3 buy open trades will be closed at the same time by this EA

So basically we open trade manually but closed Automatically!

I'm willing to test this on my live account that just opened special for this system in hope i can contribute something to the system

thanks in advance!
If you have to make a mistake, make sure it is a new one
  • #95
  • Edited 5:04pm Mar 18, 2012 2:41pm | Edited 5:04pm
Komugi,
please test it on a Demo account, not a live account.
(just a recommendation, you will do whatever you want anyways .)

The EA to close manually opened trades is no problem.
As soon as I find time, I will code it.
Edit: Done, it is attached.
For MagicNumber=0 (default) it will close your manually entered trades
on the pair you attach it.
Or you insert a MagicNumber and it will only close trades with that MagicNumber on the pair. ( there are tools which enable you to place trades with a MagicNumber )

The Cobra-EA on H1 shows in backtests that the strategy needs more filters.
The old thing: only take trades in direction of the trend on higher timeframes.

Next question:
How do you define the trend on higher timeframes ?
With MAs or RSI or what ?

I can include such filters into the Cobra-EA but somebody please suggest
precise rules ( and the reason why he thinks that filter will be usefull ).
Attached File(s)
File Type: ex4 Cobra_MA_Close_EA.ex4   9 KB | 435 downloads
  • #96
  • Mar 18, 2012 2:53pm Mar 18, 2012 2:53pm
Hi Guys,

I have been lurking on this thread. I have created an EA, which will trade the trend per the original PDF - or at least pretty close. I have borrowed some code from an candle pattern indicator to help identify the reversal type candles. I have also added a feature to counter trend trade - where all the PA conditions are met, but we have reversal candle situation. I downloaded the minute data from -http://www.fairtradingtech.com/support/ea-backtesting/?utm_source=mt4&utm_medium=mail&utm_campaign=mt4 and have doing extensive backtesting. I will post the EA when i have gotten the bugs out and feel it can be profitable. Thanks!
  • #97
  • Mar 18, 2012 4:13pm Mar 18, 2012 4:13pm
Quoting fxtr51
...The Cobra-EA on H1 shows in backtests that the strategy needs more filters. The old thing: only take trades in direction of the trend on higher time frames...
IMHO every system must take into account important support and resistance areas, from the charts posted it seems the loss trades are mainly from Support and resistance which has been visited again and again by the price action. When I saw this thread I liked the basic idea but straight away I realised that S&R lines were vital for this approach.

I have attached Squalo`s S&R as used by nanningbob successfully in his systems, it has also been used by Stevehopwwod in his EAs, moreover, back-tests during weekends when we have extremely wild fluctuations of spreads is not very accurate.

For the EA if you ask Stevehopwood he may give you some advise.
Attached File(s)
File Type: ex4 10.1DailySR1.ex4   11 KB | 405 downloads
File Type: mq4 10.1DailySR show lines.mq4   10 KB | 415 downloads
  • #98
  • Mar 18, 2012 4:23pm Mar 18, 2012 4:23pm
Astral, I'm trying to learn how to put Fibs on a chart. Would this indicator help with that?
  • #99
  • Mar 18, 2012 4:29pm Mar 18, 2012 4:29pm
Quoting TallCoolOne
Hi Guys,
I have created an EA, which will trade the trend per the original PDF
Mine does that but it is great to compare.

Quote
I will post the EA when i have gotten the bugs out and feel it can be profitable.
I am looking forward to it.
  • #100
  • Mar 18, 2012 4:46pm Mar 18, 2012 4:46pm
Quoting astral77
IMHO every system must take into account important support and resistance areas, from the charts posted it seems the loss trades are mainly from Support and resistance which has been visited again and again by the price action. When I saw this thread I liked the basic idea but straight away I realised that S&R lines were vital for this approach.

I have attached Squalo`s S&R as used by nanningbob successfully in his systems, it has also been used by Stevehopwwod in his EAs, moreover, back-tests during weekends when we have extremely wild fluctuations...
I agree about the S/R areas.
But I need precise rules which then can be coded into the EA.

In other words: Doing all the work alone is not possible.
I have a life besides Forex and I already invested about 50 hours into this.

Posting indicators and pointing me to Steve to ask for advice is like saying:
Do more work on it buddy. We wait until you have coded the best possible
thing. Thanks for it.
Go test the indicators, go learn the code, go do all the backtests .....
we appreciate it.

No offense intended, but I am tired and I cannot do all the work.

Again: I need precise rules for new filters which can be coded into the EA.
Thanks.

Is there an EA which is profitable ?
( I downloaded the Symphony EA from Steve and backtested it, but it only is draining the account similar to the Cobra-EA)
Maybe I missed a really profitable strategy which has been coded into an EA. If someone would point me to such an EA, I would be gratefull.
0 traders viewing now