Simbasystem-gbpusd - page 63

New comment
 

Frantacech

frantacech:
Guys, I still find any good EA for pyramiding to direct trend.

Both any martingale, fibogale etc, pyramiding loss and i want pyramiding profit.

Look this

http://www.forexfactory.com/showthread.php?t=11566

Your opinion?

Maybe small tuning - booster, options if enable, ea add 1-3 penging order direct trend and step any 15pips? If execute second, first SL to BE, second SL to -15, still risk only one order?

Hi Frantacech ,

MY OPINION..?..Do you still have to ask?..400:1 leverage(so 100k in account means 400 lots?..go to the casino,ask for comps ..the chicks will come by themselves..you will end up poor the same,but at least you will have interesting stories to explain,Always in the market(the trend WAS your friend wasn`t it ?..now that you are poor you don`t have any friends left),he didn`t solved the "risk" issue..(What did he solve..;) ? ).......

...Vampire sunbathing

Ok..I am going to make a bet..I bet that you are interested with converting 5k into 100k+ minimum...and,if that is the case..what is the best strategy?..1-To run excessive risks? NO..You have to realize that you don`t need to..

Trade your system live,even with 5k...1 year minimum,then go to the banks ,private investors ,etc,for funding..Show them >50% retun with <20% dd..YES..you will earn much more handling other people money with the right reward/risk than pretending to convert 5k into 1mm $..no way my friend..

...Cinderella story..

Regards..take care

Simba

 

It gets even better..

Hi All ,

The Basis:A profitable edge..buy dips in uptrend ..sell rallies in downtrend..simbasystem or whatever..it is really only a philosophical issue

The Multiplier:..Money Management..677K profit for 10k account..in 2.7 years,instead of 44k profit without MM..improvable

The Survivor: Balls to trade 38% dd..?I don`t have them..and my brain tells me there is a better option..just reduce the risk..;)

Regards

Simba

 
popok_sakti:
Hi,

I just try to participate more in this thread, I have made small modifications which maybe usefull. I attached the new EA with small modifications

- check for spread, is spread is bigger than usual, then exit

- change lot calculation from Integer to double

- check minimum/maximum allowable lot from broker and set to minimum/maximum if less/more than minimum/maximum

- fix the ordersend stop loss,for example, for long stop loss, it will 1 point + spread below the chandelier_dn

- make wpr limit for top (-5) & bottom (-95) as extern variable

also make the 10 pips as external variable, which is for reentry criteria

( maybe usefull for optimisation

- print ordersend error if there is order error

- adding posibility of using atr

Hi popok_sakti,

Thanks for your contribution. However, this EA is not ready and is not supposed to be ready for live or demo trading. We want to keep it (its source code) as simple as possible. The Chandelier and SATL values are already difficult enough for the computer to compute and running the backtest on 2.7 years of M5 data isn't exactly a fast thing. We don't want to increase the running time even more When we go to demo trading, we will employ things like spreads, alerts etc. However:

1) If anyone wants to try to forward test this system (although it would be very preliminary), please use popok_sakti's modification.

2) I'll change the lot number from integer to double the way you did it and include it in next release. It's true that we could use minilots and smoother position sizing.

As far as the Ask and Bid issue in LastBuyPrice and LastSellPrice -- it's done the way it is because we want to see how far has the chart (Bid price) moved since last entry. Spread has nothing to do with this, that's the reason for using Bid price only. There might be some other confusion about adding and substracting the spreads around Chandeliers and other things. I've tried to think it through very well and I think it should be the way it is now. If you have doubts, please don't correct it, just show me and we will discuss to avoid confusion -- example: this LastBuyPrice issue is hitting in the eye, yet it is there on purpose

Thanks again for your work

 
SIMBA:
Hi frantacech,

Nothing to be sorry about..I was not offended,I didn`t understood the meanings of your comments..now I understand and allow me at the end of this post to give you some humble "implicit" advice in the form of a story..but first..

You don`t believe results of strategy tester..neither I ,but they are the best next thing,If it works in backtest it can work or not in real trading..but if it doesn`t work in backtest,you better throw it away..in addition we have learnt a lot about a system I trade live,just by a pair of weeks of testing..

Goblin example...I think that this EA was created by bluto and he is an exceptional programmer,and an even better person..so,the only thing I can say is:if the backtest was not done for at least 2 years I do not consider it a valid backtest..if it was done for +2 years,then 1 month double the account is not relevant..let`s wait for 2 years and compare apples with apples..

Your friend...Most people will use their common sense to say that this is not a relevant example..it is exceptional in both profitability and shortness of time,so nothing you can realistically expect to replicate..I will take another venue..

Let me explain an interesting thing,I started trading in 1992,in that era,in Spain,if you wanted to trade you had 2 options,1-buy complicated ,expensive equipment (most fashionable came with satellite connection) to trade from your office..or go to some broker (fortunately the broker was very serious)to trade in the customer`s room(if you have read "reminiscences of a stock operator" ....you will imagine the ambience ) where we had 3 bloomberg screens for Futures traders,separated from the screens used by stock traders in the spanish stock exchange..in front of "our" screens sat everyday from 5 to 10 "daytraders",obviously we shared the 3 screens..and the 2 phones to execute the trades...

After the first year there were still around 7 traders of which only 2 were the same that when I started(I would say average life span was 4 months and i saw like 20,25 different people there)),your humble Simba and "my friend"..the only reason I was still there being that i was short the pound (and had the balls to ride the move) when George Soros decided it was a good sell too..he probably had insider info on my trade ..

The reason why "my friend" was there is that he converted a 50k$ account into 1.7 MILLION $ in 7 months..I was there I saw him do it,he traded 2 to 3 times a week,opening new positions or adding to already favourable positions that he then kept for 2 weeks to 3months,during that period I would say I never saw him lose a single trade,and,before somebody asks,I didn`t followed him...

Well..one day..he mentioned to me that with the money he had plus bank funding he was thinking about creating a Fund,with about 100mm bucks in it,etc,etc("delirios de grandeza "we say in spanish)...in 4 months he was back to below 100k$-he had the good sense to stop when his account was at around 70k$-he took the money,went to India to meditate,yes it is true..he went to India to "repair" his defectuous tool..his brain,and the astonishing thing is that around 6 months later,he reappeared,opened another account with 50k..and you know what..in 6/7 months he was back to 1.7mm bucks..and you know what..5/6 months later he was back at 70k(I told him he must have some kind of freudian fixation with 7..probably not his lucky number)

My friend was a genius trader with a psychological issue that severely limited his performance..Your friend can be either a genius or a lucky trader..only the passing of time will tell

To be successful in trading you need 3 things:

1-The basis: An edge,a profitable system,be it a 30% winning system with 8 to 1 avg win vs avg loss ratio..or a 80% system with 1 to 1 ratio,etc.with max 20% peak to valley drawdown..

2-The Multiplier:And I will prove it to you now..just look at the attachment,it is the same system that made 44k(13%maxdd) in the tests..now it makes 470k(38%maxdd)..what is the ONLY difference? what is the Multiplier? Money management..my friends

3-The Survivor:The most important issue in tradind is survival,believe me that ,just after my first year of trading,I felt like a 1000 years old vampire having seen how all the "departed ones" had exposed themselves unnecessarily-and happily- to the sun..The Survivor is you,your psychological issues..you have within yourself the capability to be a genius..and to destroy your account much faster than you can grow it.

Please see the attachment.and Hope this has helped somebody.

PS:Today is a typical "vampire killing day"..NFP..put on your solar protection .take care..The Van Helsings of the world are there trying to trick you into the sun..

Regards

Simba

Simba, thank you for another great post, as usual

To all: how are we doing with backtesting, any other conclusions? I'm still testing Chandelier, it's taking much more CPU time than I thought. Perhaps because I have to run it on laptop. Anyway, so far it's going in the direction of the conclusions already made, i.e. so far so good

Great that people are talking again, most of yesterday it seemed most of you were dead

frantacech: Welcome on board, Czech crew now increasing in numbers to two... At zije financnik.cz

 

ea ( adding new small feature )

Hi,

I just try to participate more in this thread, I have made small modifications which maybe usefull. I attached the new EA with small modifications

- check for spread, is spread is bigger than usual, then exit

- change lot calculation from Integer to double

- check minimum/maximum allowable lot from broker and set to minimum/maximum if less/more than minimum/maximum

- fix the ordersend stop loss,for example, for long stop loss, it will 1 point + spread below the chandelier_dn

- make wpr limit for top (-5) & bottom (-95) as extern variable

also make the 10 pips as external variable, which is for reentry criteria

( maybe usefull for optimisation

- print ordersend error if there is order error

- adding posibility of using atr

 

Malcik

Malcik:
Hi popok_sakti,

Thanks for your contribution. However, this EA is not ready and is not supposed to be ready for live or demo trading. We want to keep it (its source code) as simple as possible. The Chandelier and SATL values are already difficult enough for the computer to compute and running the backtest on 2.7 years of M5 data isn't exactly a fast thing. We don't want to increase the running time even more When we go to demo trading, we will employ things like spreads, alerts etc. However:

1) If anyone wants to try to forward test this system (although it would be very preliminary), please use popok_sakti's modification.

2) I'll change the lot number from integer to double the way you did it and include it in next release. It's true that we could use minilots and smoother position sizing.

As far as the Ask and Bid issue in LastBuyPrice and LastSellPrice -- it's done the way it is because we want to see how far has the chart (Bid price) moved since last entry. Spread has nothing to do with this, that's the reason for using Bid price only. There might be some other confusion about adding and substracting the spreads around Chandeliers and other things. I've tried to think it through very well and I think it should be the way it is now. If you have doubts, please don't correct it, just show me and we will discuss to avoid confusion -- example: this LastBuyPrice issue is hitting in the eye, yet it is there on purpose

Thanks again for your work

Hi Malcik

Only one question..if we were trading live,for the period tested,this bid ask issue would make the live results different from the tested ones..isn`t it?

If this were the case I suggest that we should change it in the EA to reflect real market conditions..

Regards

Simba

 

popok_sakti

popok_sakti:
Hi,

I just try to participate more in this thread, I have made small modifications which maybe usefull. I attached the new EA with small modifications

- check for spread, is spread is bigger than usual, then exit

- change lot calculation from Integer to double

- check minimum/maximum allowable lot from broker and set to minimum/maximum if less/more than minimum/maximum

- fix the ordersend stop loss,for example, for long stop loss, it will 1 point + spread below the chandelier_dn

- make wpr limit for top (-5) & bottom (-95) as extern variable

also make the 10 pips as external variable, which is for reentry criteria

( maybe usefull for optimisation

- print ordersend error if there is order error

- adding posibility of using atr

Hi popok_sakti ,

Thanks for your contribution..I will test your EA for speed of use and inform

It looks very complete and professional..

Thanks again and Regards

Simba

 

popok_sakti

Hi again ,

3 questions :

1-Is it possible to "block" the spread checking feature,yes i know i can block it by just putting 1000 pips..but there is something like putting 0 or unchecking it that is cleaner ?

2-From your code I think that first when using atr feature the tp 355(or whatever) gets substituted by it..is it so?even if(tp355) left inadvertently checked?

3-out of curiosity..and i don`t know if i understood it,why did you put as default value for the atr tp the 14 days atr divided by 4?

Regards

Simba

 
SIMBA:
Hi Malcik

Only one question..if we were trading live,for the period tested,this bid ask issue would make the live results different from the tested ones..isn`t it?

If this were the case I suggest that we should change it in the EA to reflect real market conditions..

Regards

Simba

Simba, I don't think so... IMHO Strategy Tester artificially makes exact market conditions including spreads -- IMHO when trading live, there shouldn't be any difference due to this issue. What led you to that conlusion?

The reason I do this is that we all do our testing and observations on Bid chart that doesn't know such thing as spread. But when you buy or sell or put a SL then it is hit if and only if the correct one of Ask/Bid hits it.

For example, when we're short and we trail a SL 1 pip above Chandelier, and that Chandelier is at, say, 1.9289, we have our SL placed at 1.9290. And that is computed from Bid prices because all our charts show Bid prices. Now when the SL is hit, it means we BUY our position back for Ask price, therefore it is hit when the Ask price hits it. If we, when entering the order, just blindly set our SL at 1.9290, it would have been hit already as soon as the Bid price got to 1.9286 (1.9290 is Ask; Bid is 4 pips less). So, we would have been out of the position when the price on the chart was still below the Chandelier -- whereas during visual testing we would still see ourselves in position.

The outcome of all this crap is that we enter and exit just the way we see the price on the chart crossing our indicators and traveling certain distances -- therefore the trades end up the same way they would if we tested them visually. On the other hand, we have to be aware that setting TP to 30 means we'll be out at 26. I think this is better than how it is normally done -- exiting at 30 really means 30 but the chart we can see must have traveled 34 pips in order to have 30 pips of profit.

I think it's generally better to think of the market in terms of what it has to offer -- what we see on the chart, and not how large a profit we want. If there is 30 pips of potential, I want 26 pips and the rest goes to the broker. If I wanted 30 pips of profit, I would really be asking for 34 although there's only 30 available.

This is what I don't like about Forex, that there are two prices. In futures for example, there's just one price that is used for measuring the potential and managing the trades... and everyone pays the comission later. He doesn't want the market to give him both profit and comission.

This is just my opinion, I hope I made it clear. If you have different opinion and want the EA to be changed accordingly, that's absolutely fine.

 

Hello Malcik,

Firstly, I want to say Thank You for making this EA.

and I like to say sorry to make things complicated,

For the optimization purposes, all the new feature can be switched off, so it will behave the same as the previous one

For the LastBuyPrice issue, I think it should be the last price we open a buy position. But its just my opinion

Best Regards,

Popok

Malcik:
Hi popok_sakti,

Thanks for your contribution. However, this EA is not ready and is not supposed to be ready for live or demo trading. We want to keep it (its source code) as simple as possible. The Chandelier and SATL values are already difficult enough for the computer to compute and running the backtest on 2.7 years of M5 data isn't exactly a fast thing. We don't want to increase the running time even more When we go to demo trading, we will employ things like spreads, alerts etc. However:

1) If anyone wants to try to forward test this system (although it would be very preliminary), please use popok_sakti's modification.

2) I'll change the lot number from integer to double the way you did it and include it in next release. It's true that we could use minilots and smoother position sizing.

As far as the Ask and Bid issue in LastBuyPrice and LastSellPrice -- it's done the way it is because we want to see how far has the chart (Bid price) moved since last entry. Spread has nothing to do with this, that's the reason for using Bid price only. There might be some other confusion about adding and substracting the spreads around Chandeliers and other things. I've tried to think it through very well and I think it should be the way it is now. If you have doubts, please don't correct it, just show me and we will discuss to avoid confusion -- example: this LastBuyPrice issue is hitting in the eye, yet it is there on purpose

Thanks again for your work
1...565758596061626364656667686970...118
New comment
Reason: