Here's the new "Goblin" ea...little brother to Predator - page 41
You are missing trading opportunities:
- Free trading apps
- Over 8,000 signals for copying
- Economic news for exploring financial markets
Registration
Log in
You agree to website policy and terms of use
If you do not have an account, please register
That's exactly what the next version of Goblin ("Goblin TF") is going to do. When a strong trend direction shift is detected, it will close existing positions and begin opening new ones in the new direction...immediately.
bluo,
Great to hear that. And I want to explain my thought further clear.
1. I once mentioned, i hope martingale(or any pyramid method) plays a supporing role in goblin, with trend following playing the main role. I believe that's your intention as well. I think what decides their roles is the timeframe of direction signal. When we use shorter timeframe signal, like 1h JRSX, they definitely changes the direction more frequently than say daily JRSX. That means martingale will have more chance to take over. With daily signal, there is also a chance for martingale to take overl, when the daily trend direction changes. But obviously, this happens much less than 1h signal. Therefore trend following will take a dominant role in the majority of the time, if we follow daily trend, which lasts much longer.
2. As a suggestion for the detailed design of goblin, I think we can throw away 1h and 4h signals.(I am sorry, I know you put a lot of efforts there). On the upper left corner of the chart, we just need one line, 'dailytrend=up' or 'dailytrend=down'. The entry condition will be, once ea is launched, openning long position when the signal shows 'up', and vice versa.
3. I have some observation of daily trend, sometimes when daily trend changes its direction, the price still moves in a relatively tight range, or we can see some retracement. Therefore, I don't suggest cutting existing positions right away when a daily direction change is detected. Let's still use pyramid method(martingale or fibo progression) to deal with it, they are good at ranging market. Let's not give up their strength. I know they might fail to recover the loss, so we can set the maxtrade=7 or 8, to limit the risk. Of course we can try different settings or methods for optimiztion.
Goblin using the strength of daily trend signal(main role) and pyramid method(support role), i believe it will be extremely good, if not a holygrail.
Still, just some suggestion for you to consider, some points might be right, some might be wrong.
Thanks
teamfri
Thank you
Thank you John for bringing the 1.5 progression and Maxtrades 13 idea to this thread, I think that could be a healthy addendum to demo test to potentially reduce drawdown. Golden
There has been talk of using Fibonacci numbers to progress Goblin and I will be testing the latest version with that option next week.
Martingale has been used from the beginning of 10points 3.
There is a 1.5 progression that I have recently become aware of via the 10points 3 thread and this involves making Maxtrades13 or over the progression and I will be testing Goblin1 on a seperate platform with that option.
I have attached a gif showing the drawdowns compared so that we can see what is involved if the levels are breached
JohnAlso, extra to my previous post, I read somewhere of a Trailing TP which I guess if the price continues in the right direction then keep going until maybe a few pips in the wrong direction the sell instead of settling on a set TP price. Does Goblin have this Trailing TP option?
That's exactly what the next version of Goblin ("Goblin TF") is going to do. When a strong trend direction shift is detected, it will close existing positions and begin opening new ones in the new direction...immediately.
bluto
At the moment, I am really not sure what is the best thing to do with the existing positions, when daily trend changes its direction.
If we use daily trend signal, and positions are caught in the wrong direction in the transional period, maybe it is a good idea to close the existing postions.
Or, we let martingale take over at this time, by setting maxtrade at a limited number, as I previously suggested.
I think it needs more testing to find the best answer.
Thanks
teamfri
Looks like a great EA
Gidday
I like the look of this ea but I am have trouble getting it to work with NF data suggestions would be good.
Below is the journal entry that shows up.
I have tried adjusting everything, from the initial deposit to the trailing stop.
I have never seen this before.
2006.12.02 21:09:39 2006.10.16 13:00 Cannot open file 'C:\Program Files\MetaTrader - North Finance\experts\indicators\Turbo_JVEL.ex4' on the EURUSD,H1
cheers
Beno
Indicator
Gidday
I like the look of this ea but I am have trouble getting it to work with NF data suggestions would be good.
Below is the journal entry that shows up.
I have tried adjusting everything, from the initial deposit to the trailing stop.
I have never seen this before.
2006.12.02 21:09:39 2006.10.16 13:00 Cannot open file 'C:\Program Files\MetaTrader - North Finance\experts\indicators\Turbo_JVEL.ex4' on the EURUSD,H1
cheers
BenoI know this sounds too easy but have you downloaded both indicators required and compiled/saved into the indicators file?
It can't open a file that it can't find.
John
lots of good ideas. Let me point out one thing. If the daily trend is up and the ea takes long positions. Then the price stalls, and starts down. By the time the daily trend direction is turned to down the price can be as much as 100 pips against you. It takes a large move to register as a daily trend change. Good for filtering intraday wipsaw. But slow to indacate a trend change.
What about 4 hr time frame? 1 hr is way to choppy. But if the ea only took trades when all 3 time frames lined up in one direction? And if 4 hr turns against the daily then exit existing positions untill all 3 line up for new entry signal. From the moment all 3 turn the same way it should be good for at least 30 pips.
If all 3 are long, and 1,4 hr turn short, chances are the daily will soon follow.
just some more thoughts.
by the way last week testing I made 1k on aud/usd starting with 5 k account and max trades 10 and 30 pip targets. No s/l
Dave
hello everyone,
to be honest with everyone here i think you all are missing the point of this EA.....no disrespect intended to anyone here, just stating it the way it is, to the chap who almost doubled his 10,000 dollar account in 4 days a couple of posts ago, your wayyyyyyyyy overleveraged, one nice big 500 pip move down or up of unexpected news and your account is done!!!!! and it happens more often than you think....its not about the best indicators to tell you what direction the trend is in, you could have the best indicators in the world telling you to go long and then terrible news hits or a bad worldly event happens and it suddenly drops 500 pips with NO retracement...tell me, what is your strategy if you get an unexpected event that makes the currency your trading drop 500 pips???? if you can honestly answer this question with absolute confidence than your ready to go LIVE..if not, than more work needs to be done on the systems weakness, which is hard , fast trends in one direction with no retracements.....what are the answers?? ive stated them in another post ill state them again:
Phorex Phreak Solution To Big Moves
For this kind of EA to have a chance at any longevity long term and avoiding its prime weakness of the big 400-500 move without a retracement there are 2 steps that MUST be implemented in the EA before even considering using it.......
#1- when a position goes against you and lets say your second order is executed to open, the EA MUST close the first one at a loss, why??? to avoid building up positions and using unecessary margin..so if a 3rd position is opened, that second one is closed at a loss as well, and so on, until you current position covers the losses for the other losing closed positions and profits some..
#2- starting with .01 nano lots, now lets say your max trade is open all the way at 10, this means that your down 5.12 lots, now if the position goes against you another 20 pips, which would be at the 200 pip mark from the first order, open a hedge at 1.05 lots the other way simultaneosly, if it goes against you another 10 pips, add another 1.05 lots, another 10 pips, add another 1.05 lots, all with 30 pip trailing stops on the hedge orders only, to a max of a combined 5.25 lots ( so 5 X 1.05 lots) in order to buffer the other side, that way if you have a nasty 500 pip drop due to an unexpected news anouncement at least you have some built in hedge protection and even might profit from it because your using 5.25 versus the other 5.12 from the original orders....
Until these issues are resolved with the above 2 points this EA remains dangerous at best with no hedging features....let me know your thoughts....
Goblin with Fibonacci Sequence ?
Have anybody tried testing Goblin with these setting ? if so please post the result, I am going to test it with John setting next week.
Regards ,
Fadi
Hey,
frantacech wrote a neat post on hedging pyramid's
The interesting thing of course, is that the progression he uses is a Fibonacci Sequence.
1,1,2,3,5,8,13,21,34 . .
I had to try it instead of a martingale
1,2,4,8,16,32
For lot size increments. You can see that the FibSequence allows you to enter more trades before your risk reaches the same level, obviously your reward scales similiarly.
Heres the code I replaced in the EA.
This code could replace any of the martingale code in the similiar EA's that this one is based on.
I replaced this
else { mylotsi=NormalizeDouble(mylotsi*2,2); }[/CODE]
with this
[CODE]mylotsi=MathRound(MathPow(1.6180339,cnt+1)/MathSqrt(5))*lotsi;It should also adjust for whatever LOTS you choose, ie. .01, .1 or 1.
Heres some results for January to August on 30min EURUSD. *EDIT* (not GBPUSD)I've revised the two lastest editions of Goblin with Fibonacci progression option instead of Martingale.
Goblin1.1 converted to GoblinFibo1.1 is here;
https://www.mql5.com/en/forum/175616/page21
Goblin1.2 converted to GoblinFibo1.2 is here;
https://www.mql5.com/en/forum/175616/page26
I've found 1.1 to give me better results during backtesting, it is what I will continue forward testing this week. I don't have enough forward results from last week to post.
I will attach one more set of backtest results for you.