Martingale EA - 100% per month - long time risk issue

 

Hello All,


I have a EA based on Laguerre and Moving Averages indicators, among others.

In case of wrong signal I use martingale to help the account (with constant steps, incremental steps or Fibonacci steps)

From my statistics the probability of a wrong signal is just about 2% (I used 10 pairs to tests with individual data for 10 years).

I consider a wrong signal every time the EA needs more than 6 martingale steps to correct a wrong signal.


The question now is, I want to put this EA working with 10 pairs, if each one of them as 2% probability of wrong signal it may be the case of two wrong signals for two different pairs at the same time window. For this case my account will be exposed.


I'm wondering if there is any strategy to make my EA better in relation of what it is right now. I'm talking about a EA with 100% profit per pair with 1k initial balance, I think makes sense to work on this.


Thanks for you help

Paulo

paulo.cf.gomes@gmail.com

 
Don't use Martingale. Use Kelly Instead.
 

Could you explain it a little bit, I'm not aware about the Kelly system

Thansk



ubzen:
Don't use Martingale. Use Kelly Instead.
 
Try this link here.
 
Or just simply use 0.1,0.2,0.3,0.4,... lots. Problem with Kelly is when you lose you always reduce lots, how can you then ever cover previous minus?
 
01005379:
Or just simply use 0.1,0.2,0.3,0.4,... lots. Problem with Kelly is when you lose you always reduce lots, how can you then ever cover previous minus?

by winning more than you lose. You cannot always immediately overcome any previous loss. Trying to enforce this will kill you eventually.


You need a profitable trading system and you will win. Position sizing based on previous losses alone without taking the probabilities of the trading setup into account does not make an otherwise unprofitable system profitable.

On the other hand a system that already gives (on average) profitable trading setups does not need any bet progression systems to "overcome" losses. It will overcome all losses by all winners on average by itself.

 
7bit:

by winning more than you lose. You cannot always immediately overcome any previous loss. Trying to enforce this will kill you eventually.


You need a profitable trading system and you will win. Position sizing based on previous losses alone without taking the probabilities of the trading setup into account does not make an otherwise unprofitable system profitable.

On the other hand a system that already gives (on average) profitable trading setups does not need any bet progression systems to "overcome" losses. It will overcome all losses by all winners on average by itself.


I could restring, instead of martingale close the order if the signal inverts but its not so profitable.

My idea is something like, as soon as you enter 6 step of martingale do something in such pair or in a correlation pair to avoid break the account

 
pcfgomes: as soon as you enter 6 step of martingale do something

The only reasonable thing to do is close the position and take the loss. And accept that these things happen on a regular basis and consider this when deciding about the initial lot sizes.


You cannot enforce 100%

 
pcfgomes:

I could restring, instead of martingale close the order if the signal inverts but its not so profitable.

My idea is something like, as soon as you enter 6 step of martingale do something in such pair or in a correlation pair to avoid break the account

I think the point is if you don't take a six step martingale you won't find yourself in a position to have to do something to avoid breaking the account in the first place. Find a system the gives you effective signals for longterm profits and manage your risk effectively to stay alive in the inevitable drawdown periods. Would you really defy a trend by entering the wrong way just because you were wrong the last time? imho stop at step 1 and re evaluate your signal.

V

 
Viffer:

Would you really defy a trend by entering the wrong way just because you were wrong the last time?

Martingale does not necessarily dictate that you enter the same (losing) direction in any subsequent step. You could do the second trade with a new signal and only double the lot size. But this still does not make martingale a desirable MM scheme.
 
pcfgomes:

I could restring, instead of martingale close the order if the signal inverts but its not so profitable.

My idea is something like, as soon as you enter 6 step of martingale do something in such pair or in a correlation pair to avoid break the account


From my video game Madden says it best. "If the guy didn't stop you on third down, you wouldn't be 4th and Long to begin with." I suggest you punt or kick the field goal. Going for 4th and long is usually a bad idea probability wise. The best place to cut your losses is on second and long in Forex.
Reason: