Euro Blitzkrieg
Filed Under (EA System, Journal) by Casey on 06-06-2009
Tags: Automatic trading, Backtest, Cyborg EA, EuroBlaster, Forex Boomerang, Forex Crusher, Forex Espionage, Forex Transformer, Manual trading, Pipzu
I was trying to backtest other EA in my ‘To Be Tested’ folder… and as I was backtesting and backtesting and backtesting…
An idea popped up on my head…. But… before I dived into what this idea is about… I notice that during all these backtesting… a lot of the graph were see-saw shaped.
While the profit factor is interesting within the initial time period of 2008 till end May 2009… the see saw graph proves discouraging…
Maybe because I am brainwashed by the so called ’smooth’ equity curve… like this one below

(Will talk more about this in another post)
Or recently… like this linear graph below…

hehe.. this guy made no promise to me that he will share his secrets…
Anyway… that wasn’t the point… The point was… all these so call ‘Will make you $300,000 in two months’ may sound pleasing to the ears and bring hope and happiness to the mind… It does not or might not be what it proclaim it to be in reality. There is a reason for all these disclaimer at the bottom of all these BS site you know…
But… get this… all these debate about manual and auto trading got my brain once again churning… Yes… EA are not the holy grail in forex trading. There might not the an absolute robot and it will probably not be… remember on one of my post I equate these Forex EA Robots as droids in the Star Wars movie?
Yes… these EA are the droids who don’t know $hit if it was pooped on… We are still very far away from Artificial Intelligences when it comes to robots… And all these site stating that it is artificial intelligences are BS to the max… Have you seen the type of codes these so call AI uses… so simple and the language that it uses are so ‘loose’ that a software can decompile it so easily…
Ok… so sorry I digress a wee bit…
Anyway… allow me to show you some of the backtesting results I did for a few EA which came about to my idea… (All backtest were from 2008 till end May 2009)

Boomerang: Profit Factor=1.04

Blaster: Profit Factor=10.78

Transformer: Profit Factor=1.17

Pipzu: Profit Factor=1.33
Crusher (Can’t do any backtesting but live results on another site looks promising)
Cyborg (Demo lousy but on Live it rocks!! *fingers cross* so far that is…)
AND the most recent one… Espionage (Profit factor unavailable)

One thing that all of these test have in common is this… They trade or can trade EURUSD… Apart from that, it’s profit factors are also pretty interesting.
If you ever gone to a trading seminar… a trading system if once backtested showed more than 50% chances of winning, it’s a good system. Of course that is manual trading. But what if one can apply this into automatic trading?
I mean hey… all these shows a somewhat good profit factor… No? It just needs to be used properly.
Now here comes my idea… what if I can uses these ‘band of brothers’ at the same time? In a very periodic fashion in the market… Just like how manual traders would go about in their daily trading routine. I mean hey.. these manual traders trade at a certain time… they don’t just sit around 24/5 monitoring looking for opportunity…. do they?
No they don’t…
So the general strategy would be to… Move in at a certain time, Attack and Move out with we kill em all.. Just like what the German did with their fancy Panzer tanks in WWII…
This is an extension of my previous idea or analogy where the robots works while I manage. In this case, this group of ‘brothers’ will be deployed into the forex market to see how much pips in can pull under my watchful eyes. It’s just a matter of clicking ‘Expert Enable’, see how many pips it kills and then ‘Expert Disable’
Interesting no? You see, in manual trading, not only is it about the 2% trading risk, but it is also about timing and adapting.
EA can not adapt… as I said Forex EA has not come to a point that it can adapt… It just runs according to it’s intended program which makes it very rigid.
Everybody is searching for that one ‘Rambo’ to kill em all in the forex market… but I don’t think that is the best strategy. ‘Rambo’ usually hides in Thailand and would not be sold so easily….unless it is a pretty lady with see thru blouse that works for a church (That’s from Rambo V if anyone of you are scratching your heads)
Since I can’t find Rambo now… I’ll make use of these mercenaries and see if they yield me any pips… I might pull some pips I might not… have to find the right time and balance to see if these Panzer robots can gel together. Obviously this is a general strategy… Like General George Patton… we or rather I need more detail before I can strike…
"Take NO Pips!!!" Say… that does not sound right!
Ok.. need to work on my war-cry… but you get the idea…. IT’S WAR I tell you…

But first… it’s training in the demo ground







>>EA can not adapt… as I said Forex EA has not come to a point that it can adapt… It just runs according to it’s intended program which makes it very rigid.
well, this is not 100% true. EA can adapt if it was intended by its developer.
and in MT4 EA developer has a lot of TA indicators to alter EA depending on market conditions. the problem is to determine WHEN and WHAT adaptations to make.
Well… it is rigid for people like me who knows only water about MQ4 language. It adapts according to it’s developer… but not to it’s consumer.
Forex market is one huge psychological pit… so when the market moves, we have to adapt to it… and the EA can’t… the only way to adapt to it is to turn it on and off…
This strategy reminds me of the EA called ” Semi-Auto Robot” based on the FireBird v63 where you click on LONG or SHORT only for the EA to attack once General Patton sees a TREND. But then, trends changes so ever quickly nowadays and if you are sleeping when yr enemies counterattack, you’re dead !
Yes Kentbfea… that was how I got that idea… From FireBird… The idea was interesting but the execution part is the one that needs to be refined…
Hi Casey, regards the Forex Transformer u wrote: ………backtest looks good…
Objection, your honor!
backtest looks really bad… don’t know what u tested, but the 5 digit version is an absolute no go!
Hmmm… need to ask you the criteria which you grade a result of a backtest Onur…
I did another backtest since 1999 and it looks good. It’s not a linear curve, that is one of the criteria which you use to judge…
the criteria is simple: the system has to make profit. backtest graph shows linear down bias until it crashes in the right lower corner jejeje
tested different TF and CCYs