EAs are awesome, and fun, and well, let's see who's the awesome EA of the week.
Both forexpeacearmy and 4xproject show an EA named "Equity Builder" doing remarkably well in forward testing ( +11.24% weekly for 7 weeks on peacearmy, and 13.8% weekly for 5 weeks on 4xproject.)
Of course, the internals of that amazing EA are jealously guarded, but they do put an account on mt4stats, at http://fxequitybuilder.mt4stats.com/
That's enough to get a general idea of the investing pattern of that script:
- only one buy or sell at a time (although each buy/sell can be split in up to 8 individual orders)
- orders start tiny, and are placed in progressively bigger lots (+50%) if the price goes in the opposite direction of the order, for up to 8 orders
- in the pattern above, each order is roughly a pip apart. If the price starts to go in the other direction, the bigger lot orders don't get placed.
- there is no stop loss. ever. why didn't anybody think of this before?
- the exit happens at a predefined (presumably manually) limit. The mt4stats account starts at a 30 limit, then 8, then 10, then back to 8. When any one of the orders hit, all of them get closed at once. Simple stuff.
So I wrote a quick EA that follows the above rules. I'm using buy and sell limits, since it's easier than having to track price levels myself, but otherwise, it's pretty faithful. no stop loss, of course.
The secret sauce is of course the entry algorithm, and I have no idea what Equity Builder is doing there.
So I wrote my very own finely tuned algorithm that relies on the too rarely used mathematical function MathRand to take a peek into the future and consistently beat the market. Oh yeah.
As a ForexFactory exclusive, I am now sharing this extraordinary algorithm with the world:
So, how do you think this modern marvel of statistical analysis would perform?
Are you expecting amazingly consistent performance going through the roof?
You, my friend, are correct.
Look at a typical backtest result, done from jan 09 until now.
http://imgur.com/bWZ7o.gif
The numbers speak for themselves. And look at that smooth equity curve.
That's a sure sign you're in a for smooth ride, buddy.
Alright, I think I've got it out of my system now.
Back to your regularly scheduled programming.
Both forexpeacearmy and 4xproject show an EA named "Equity Builder" doing remarkably well in forward testing ( +11.24% weekly for 7 weeks on peacearmy, and 13.8% weekly for 5 weeks on 4xproject.)
Of course, the internals of that amazing EA are jealously guarded, but they do put an account on mt4stats, at http://fxequitybuilder.mt4stats.com/
That's enough to get a general idea of the investing pattern of that script:
- only one buy or sell at a time (although each buy/sell can be split in up to 8 individual orders)
- orders start tiny, and are placed in progressively bigger lots (+50%) if the price goes in the opposite direction of the order, for up to 8 orders
- in the pattern above, each order is roughly a pip apart. If the price starts to go in the other direction, the bigger lot orders don't get placed.
- there is no stop loss. ever. why didn't anybody think of this before?
- the exit happens at a predefined (presumably manually) limit. The mt4stats account starts at a 30 limit, then 8, then 10, then back to 8. When any one of the orders hit, all of them get closed at once. Simple stuff.
So I wrote a quick EA that follows the above rules. I'm using buy and sell limits, since it's easier than having to track price levels myself, but otherwise, it's pretty faithful. no stop loss, of course.
The secret sauce is of course the entry algorithm, and I have no idea what Equity Builder is doing there.
So I wrote my very own finely tuned algorithm that relies on the too rarely used mathematical function MathRand to take a peek into the future and consistently beat the market. Oh yeah.
As a ForexFactory exclusive, I am now sharing this extraordinary algorithm with the world:
Inserted Code
void checkForEntry() { if (OrdersTotal()>0) { return; } // science! if (MathRand()<100) { // yay, we're placing an order. if (MathRand()<16384) { buy(); } else { sell(); } } }
Are you expecting amazingly consistent performance going through the roof?
You, my friend, are correct.
Look at a typical backtest result, done from jan 09 until now.
http://imgur.com/bWZ7o.gif
The numbers speak for themselves. And look at that smooth equity curve.
That's a sure sign you're in a for smooth ride, buddy.
Alright, I think I've got it out of my system now.
Back to your regularly scheduled programming.
How did this get here oh god I am not good with computer.