10points 3.mq4 - page 333

 

10p3v0.04

David is testing a new 0.04 prior to releasing it here, I was able to forward test it also. It needs much time if we are to optimize it for 19 pairs and the following AUDUSD trade was made yesterday using the default settings.

Read the comments at the bottom of the gif and notice what could have been a different result if the account had been larger. If the EA had been able to open the next trade it would have closed off all the progression in profit.

So it becomes the old martingale problem, can we afford to use an account big enough to be able to withstand the largest progression and if we can will the EA provide a satisfactory return on that size account.

Another option is to limit the number of MaxTrades and set the initial stop to that same level then after the EA hits the MaxTrades level it will only trade up to the next level plus the initial stop and use the dynamic close.

That may be a drastic option but would have prevented this trade hitting the margin requirements and being forcible closed. Of course this problem is unlikely to occur if we use the pair EURUSD only but the whole idea of 0.04 is to be able to trade all pairs one at a time.

John

Files:
audusd.gif  48 kb
 

10p3v0.02 set file

yeoeleven:
My preset is as follows:

Magic=772188

TakeProfit=10.00000000

Lots=0.01000000

InitialStop=160.00000000

TrailingStop=0.00000000

MaxTrades=10

Multiplier=3.00000000

Pips=5

OrderstoProtect=4

Money_management=1

AccountType=1

risk=0.50000000

ReverseSignal=0

Fast_EMA=12

Slow_EMA=26

Signal_SMA=9

Shift=1

TradingRange=0

UseTimeFilter=1

StopTrade=13

StartTrade=18

My TF is M15 and the only pair being used is EURUSDm.

Default from the code of the EA follows:

extern int Magic = 772188;

extern double TakeProfit = 10;

extern double Lots = 0.01;

extern double InitialStop = 160;

extern double TrailingStop = 0;

extern int MaxTrades = 10;

extern double Multiplier = 3;

extern int Pips = 5;

extern int OrderstoProtect = 4;

extern bool Money_management = true;

extern int AccountType = 1; //0: Standard account(NorthFinance,MiG,Alpari) 1: Normal account(FXLQ,FXDD) 2:InterbankFX's NANO Account

extern double risk = 0.5;

extern bool ReverseSignal = false;

extern int Fast_EMA = 12;

extern int Slow_EMA = 26;

extern int Signal_SMA = 9;

extern int Shift = 1;

extern int TradingRange = 0;

extern bool UseTimeFilter=true;

extern int StopTrade = 13;

extern int StartTrade = 18;

Early on I had one progression that went too far for the small account but I just left it run and eventually the price returned and profit was taken.

John

Hi Yeoeleven,

For my live Interbanfx Nano account, should I enter 2 for account type and keep the rest exactly the same as your's above?

Thank you for advising.

Best,

forexjim

 

Preset.

forexjim:
Hi Yeoeleven,

For my live Interbanfx Nano account, should I enter 2 for account type and keep the rest exactly the same as your's above?

Thank you for advising.

Best,

forexjim

My preset saved from the actual trading platform using IBFX nano is below.

My recommendation is to open also a demo account with the same perameters and use it first to see how things pan out. When you actually see how the EA opens units and trades then load it onto a live account.

John

Magic=772188

TakeProfit=10.00000000

Lots=0.01000000

InitialStop=160.00000000

TrailingStop=0.00000000

MaxTrades=10

Multiplier=3.00000000

Pips=5

OrderstoProtect=4

Money_management=1

AccountType=1

risk=0.50000000

ReverseSignal=0

Fast_EMA=12

Slow_EMA=26

Signal_SMA=9

Shift=1

TradingRange=0

UseTimeFilter=1

StopTrade=13

StartTrade=18

Files:
10p3.set  1 kb
 

10p3v002 01/01/99-01/21/08 backtest using John's preset except no $ mgnt & acct 2

yeoeleven:
My preset saved from the actual trading platform using IBFX nano is below.

My recommendation is to open also a demo account with the same perameters and use it first to see how things pan out. When you actually see how the EA opens units and trades then load it onto a live account.

John

Magic=772188

TakeProfit=10.00000000

Lots=0.01000000

InitialStop=160.00000000

TrailingStop=0.00000000

MaxTrades=10

Multiplier=3.00000000

Pips=5

OrderstoProtect=4

Money_management=1

AccountType=1

risk=0.50000000

ReverseSignal=0

Fast_EMA=12

Slow_EMA=26

Signal_SMA=9

Shift=1

TradingRange=0

UseTimeFilter=1

StopTrade=13

StartTrade=18

Hi, John,

Many thanks for your advice again.

Attached please find a backtest from 01/01/99 to 01/21/08 of 10p3v002 at 89.96% Modelling Quality. Unidirectionally, it tells us a balance of $78, which is a far cry from the minimumly needed $18141, is doomed to crash. I was only able to upload the graph and summary. The data is 180MB and cannot be uploaded.

Would appreciate comments.

Best,

forexjim

 

Back test

forexjim:
Hi, John,

Many thanks for your advice again.

Attached please find a backtest from 01/01/99 to 01/21/08 of 10p3v002 at 89.96% Modelling Quality. Unidirectionally, it tells us a balance of $78, which is a far cry from the minimumly needed $18141, is doomed to crash. I was only able to upload the graph and summary. The data is 180MB and cannot be uploaded.

Would appreciate comments.

Best,

forexjim

Jim,

I have no comments about back testing, I am sure that there is value in it but I have never done back testing and am more interested in utilizing account balances that I may be able to duplicate in a live situation then forward testing them.

I have been very lucky with this small demo account which started out at $78 and currently is $166.12 but I understand that the next trade can blow it out of existence. I am encouraged by the EAs results to date and have placed it on a live account with good results, this week it has gained $13.40 using 1c pips.

One thing that does interest me is that whilst my live account gained $13.40 the demo account with the same settings only made $8.42 which is most unlikely.

Keep up the good work I am sure someone may like to comment on the back testing results. I am encouraged that it made more profit than loss.

John

 

10p3v002 progression concern.

yeoeleven:
Jim,

I have no comments about back testing, I am sure that there is value in it but I have never done back testing and am more interested in utilizing account balances that I may be able to duplicate in a live situation then forward testing them.

I have been very lucky with this small demo account which started out at $78 and currently is $166.12 but I understand that the next trade can blow it out of existence. I am encouraged by the EAs results to date and have placed it on a live account with good results, this week it has gained $13.40 using 1c pips.

One thing that does interest me is that whilst my live account gained $13.40 the demo account with the same settings only made $8.42 which is most unlikely.

Keep up the good work I am sure someone may like to comment on the back testing results. I am encouraged that it made more profit than loss.

John

Hi, Yeoeleven,

I went back to look at your posted statement with care. Is it correct that the highest lot size you had to trade was 0.81? In my test the highest lot size was 21.87 and there were 24 of them! Now please consider the difference between your live account and my backtest account. If forex brokers do stick out their hands to grab more pips or greenies over and above their standard spread income, into which account will their hands go? You are at 0.81 because your account is about 3 months old. When it becomes 109 months old like mine, do you think you will only have 24 21.87 lotsize trades? Maybe there are even some 65.61 or higher lotsize trades because of their hands. Would you agree?

Best,

forexjim

 

Lotsize

forexjim:
Hi, Yeoeleven,

I went back to look at your posted statement with care. Is it correct that the highest lot size you had to trade was 0.81? In my test the highest lot size was 21.87 and there were 24 of them! Now please consider the difference between your live account and my backtest account. If forex brokers do stick out their hands to grab more pips or greenies over and above their standard spread income, into which account will their hands go? You are at 0.81 because your account is about 3 months old. When it becomes 109 months old like mine, do you think you will only have 24 21.87 lotsize trades? Maybe there are even some 65.61 or higher lotsize trades because of their hands. Would you agree?

Best,

forexjim

The small demo account only went as far as the money would allow, then it just kept getting further into margin. If the price returns before the margin was reached then all is well. So the lotsize that my account reached is not really relevant perhaps if the funds were there it may have reached a much higher level but providing the price returned before hitting a margin call it is not important. If however the next levels were reached then it would return into profitability much more easily.

That is why using large accounts allows the trade to go to whatever price the account can afford. I am not saying that my account will go for three years I am saying that it has managed to go for 2 months and the higher it can reach the more price moves it can absorb. The back test will not show you how the account will actually perform when presented with more progression than the balance can absorb.

Better to use a balance that you can fund and forward test and see what happens when the progression can not increase but the price continues to go against. I started forward testing 10p3 when I started this thread in August 2006 and over that time this current series is the best I have tested. It is the first time I have used any 10p3 on my live account. I am fully aware that the next trade can have the potential to ruin my account, I do not need backtesting with a $10000 account to tell me that.

I have tried to explain why your back tests appear to be at odds with my demo account but frankly I don't care. Use the backtesting to optimize specific pairs to improve their performance but use forward testing to find out how the optimizing works. Then when you are happy with the forward testing please post the details here and we all can learn from them and then apply them to a live account.

John

 
yeoeleven:
The small demo account only went as far as the money would allow, then it just kept getting further into margin. If the price returns before the margin was reached then all is well. So the lotsize that my account reached is not really relevant perhaps if the funds were there it may have reached a much higher level but providing the price returned before hitting a margin call it is not important. If however the next levels were reached then it would return into profitability much more easily.

That is why using large accounts allows the trade to go to whatever price the account can afford. I am not saying that my account will go for three years I am saying that it has managed to go for 2 months and the higher it can reach the more price moves it can absorb. The back test will not show you how the account will actually perform when presented with more progression than the balance can absorb.

Better to use a balance that you can fund and forward test and see what happens when the progression can not increase but the price continues to go against. I started forward testing 10p3 when I started this thread in August 2006 and over that time this current series is the best I have tested. It is the first time I have used any 10p3 on my live account. I am fully aware that the next trade can have the potential to ruin my account, I do not need backtesting with a $10000 account to tell me that.

I have tried to explain why your back tests appear to be at odds with my demo account but frankly I don't care. Use the backtesting to optimize specific pairs to improve their performance but use forward testing to find out how the optimizing works. Then when you are happy with the forward testing please post the details here and we all can learn from them and then apply them to a live account.

John

Hi, Yeoeleven,

I hope you are right, because I will do exactly as you advise and hope I am wrong.

Will post statements when available.

Thank you for your advice.

Best,

forexjim

 

Balance levels

forexjim:
Hi, Yeoeleven,

I hope you are right, because I will do exactly as you advise and hope I am wrong.

Will post statements when available.

Thank you for your advice.

Best,

forexjim

Hi Jim,

if your concern is which account would the broker target my small one or your potentially huge one I would have to agree that the larger one with the bigger lotsize would be more likely to be targeted.

I would have to point out that the way to profit from any progressive trading program is to remove profits as they occur. If my demo account was live and I had improved it from $78 up to $166 I would expect to remove at least $50 by now and look to remove another $25 after another $50 had been earned.

Once my original funding had been removed I would not be so worried about a dramatic loss situation and so I would leave more and remove less. Never would my account reach the level that would cause major concern and thus be a target.

If I was fortunate to be able to trade a $10000 account it would not build up as your backtest suggests, profit would be removed on a regular basis and my trading would more likely be done with a direct on line broker.

John

 

Removing Profit

yeoeleven:
Hi Jim,

if your concern is which account would the broker target my small one or your potentially huge one I would have to agree that the larger one with the bigger lotsize would be more likely to be targeted.

I would have to point out that the way to profit from any progressive trading program is to remove profits as they occur. If my demo account was live and I had improved it from $78 up to $166 I would expect to remove at least $50 by now and look to remove another $25 after another $50 had been earned.

Once my original funding had been removed I would not be so worried about a dramatic loss situation and so I would leave more and remove less. Never would my account reach the level that would cause major concern and thus be a target.

If I was fortunate to be able to trade a $10000 account it would not build up as your backtest suggests, profit would be removed on a regular basis and my trading would more likely be done with a direct on line broker.

John

Hi, Yeoeleven,

This advice of David and yours makes sense and I will do the removing from time to time.

Thank you.

Best,

forexjim