Originally Posted by syedrizwan

Which broker are you using? error 130 occurs when the order is to close to the STOPLEVEL and 131 may be a lot sizing issues did you modify the LotDigits parameter in the EA?

Regards,
Riz

Originally Posted by syedrizwan

Which broker are you using? error 130 occurs when the order is to close to the STOPLEVEL and 131 may be a lot sizing issues did you modify the LotDigits parameter in the EA?

Regards,
Riz

Originally Posted by AirforceMook

I'd love to help develop a successful system. I've been messing around with these for some time and never quite mastered it. I still think the concept is absolutely phenomenal, albeit with one major hurdle to overcome before profitability will be hit.

Every martingale system (that I've made, at least) is wildly successful right up until the price begins whipsawing back n' forth and the EA bets far too much, blowing the account. I've tried eliminating all trades at a set loss, etc, but they occur far too often to end up profitable in the end. That is, after all, the point of the martingale system. You'll have massive drawdown only to recover with even larger profits.

Because of this, I'm convinced the only road to a good EA is one that times its reversals well. The fewer reversals needed to recoup the losses, the better.



I think there's two ways to do this, but both are simply trying to identify the exact "tipping point" (peaks and troughs) in the price. This is where you want your reversals to be, otherwise you'll have too many open trades and your free margin will get blown to hell.

Try to directly identify the turns.
I tried looking back and detecting a peak that's larger than the past <X> number of bars. The problem is that when you encounter a large uptrend, it will think just about every bar is a new peak. This results in you betting against the trend the whole way up.

A slightly better method was to detect bars that were in opposition to the current trend. These usually occur at the end of a large movement and are often peaks. Unfortunately, depending on trend detection & other factors, this can result in a lot of false-positives.

Use a lagging indicator that will only move on large turns
MA's are a good example of this. If you watch the slope of the line it will generally stay positive the whole way up until it has reached the top and the trend has turned. The problem is that you reverse far too late to be effective.

I've attached an example pic of a lagging indicator I've been using. It's just a step MA. If the price is currently above both bars, it will stay long. (Reverse for short) You'll notice it still results in a lot of reversals. (4 times there at the end)

=====

I've also noticed that using larger lots with smaller takeprofits tends to work better. My profitability is above 70% on average. This means that the first trade will usually close profitably. If I decrease the lots or increase the TP points this goes down.

I'm also considering the idea of having a swarm of these on different currencies working together. If one currency begins to overwhelm itself, and the rest are able to help (ie they have profit), close ALL trades to minimize the damage. I have yet to code this at all. It unfortunately cannot be backtested, either.

=====

Any thoughts folks?