Great EA in backtest! - page 47
You are missing trading opportunities:
- Free trading apps
- Over 8,000 signals for copying
- Economic news for exploring financial markets
Registration
Log in
You agree to website policy and terms of use
If you do not have an account, please register
could the broker possibly have access to the html that the back test are stored on. Since there htlm there stored on a server.
Mine used to make its first trade with ,27 lots on 500$. With .7 risk My posts of back tests will reflect this. However. Now it starts off with .1 lot. then goes to .2 then .3, and doesn't produce the 65k results with 10 lot max that it used too. And doesn't even get to the 10 lot max in a 1 year back test.
I can even down load a fresh version and run it. same thing.... .1 lot to start with.
if I use risk of 3 it uses a lot size too big for 500$ start. I can start off with 1.5 risk is the most. and it starts with .3 lots.Well as you can see from what I posted when I used risk 3 it started off with .41 lots. I don't know how to account for all the variances and abnormalities produced in metatrader's backtester. I simply know that variances and abnormalities exist. that makes it a shakey estimatation of anything. I don't know that I would lay any of this on the broker's servers. I am far more inclined to believe that the platform itself is unstable in how it processes data when it executes a test. At one time I approached making some kind of code to print out the data set that is processed in an EA when it processes it and print it to a file. But then I realized that simply printing out the raw data of OHLC doesn't do anything to uncover how the backtester is varying in it's processing from one time to the next.
If theres one thing I really dont believe would be the brokers taking time to monitor the 10000's of backtests ran every few minutes.. would be insane.. wouldnt make sense for them to minipulate that either.. they would prefer you see golden eggs and fire up a real account and start snaking from that..
But could they see what you are doing? sure its possible..
with that in mind I may fire up some tools to monitor what data MT4 is actually passing thru my ports..Do you have any tools to evaluate how the backtester itself is processing data to verify it's consistency?
you see I have saved strategy test reports using identical settings on EA's which vary in the final outcome from 11.8 million to 4.1 million. There is no way that a a few new quotes (as many as would come in even in a whole day) on the end of a date range of several months could possibly account for that much variance in results. the new quotes for one day simply can't pull off that kind of variance in a multi month test.
If theres one thing I really dont believe would be the brokers taking time to monitor the 10000's of backtests ran every few minutes.. would be insane.. wouldnt make sense for them to minipulate that either.. they would prefer you see golden eggs and fire up a real account and start snaking from that..
But could they see what you are doing? sure its possible..
with that in mind I may fire up some tools to monitor what data MT4 is actually passing thru my ports..There risk departments only job is to moniter us and how they can limmit our ability to make money. There servers and vast superior computers can moniter us and make frequent changes. With out a human needing to do anything.
A series of great back tests can produce a red flag. The fact that alot of people are useing the same EA and producing huge results, acompanied with my sudden use of the EA on a real account could have generated a flag.
The results of the tests are one thing, but the function of my auto lot has changed.
The way it places an order is different. even when I download a fresh version and don't touch it, accept to run it.
you have a point, but if thats the case they should just close shop... They earn on the spread, the more orders the more $... but we've seen that arguement all over these forums...lol
On a brighter note, I just had a chat with support at MB Trading and asked if they were looking inot offering MT4 like EFX is...they said MT4 as a platform for thier FX accounts is 1-2 months away!
This is really bad news I think. But even when it not flawed I'd say it's still not ready. Backtesting from Januar 2006 and untill today is just fine. But, when i backtesting fall 2004 it totaly wipes out the account here. This is with the North Finance demo account using 1 min data from alapri.
Hearing this stuff you guys are talking about, kind of reinforces my suspicion abount the brokers using the MT4 platform basically are crocks. Why don't any of the brokers pushing the big cash want to go with it? Why do they implement functionality that enables them to see what kind of indicators you are using? Would they need this kind of functionality if they were inn for some honest buisness with you?
you have a point, but if thats the case they should just close shop... They earn on the spread, the more orders the more $... but we've seen that arguement all over these forums...lol On a brighter note, I just had a chat with support at MB Trading and asked if they were looking inot offering MT4 like EFX is...they said MT4 as a platform for thier FX accounts is 1-2 months away!
yea only reason I am not with MB is they don't offer charting. If they get with meta4 I am there. As they say they welcome scalpers stradigy.
I was happy with fxdd untill now.
fxspeedster said no complaints with interbankfx either.
anyone have them?
you have a point, but if thats the case they should just close shop... They earn on the spread, the more orders the more $... but we've seen that arguement all over these forums...lol On a brighter note, I just had a chat with support at MB Trading and asked if they were looking inot offering MT4 like EFX is...they said MT4 as a platform for thier FX accounts is 1-2 months away!
I wish I had a dollar for every time in the last two years I've been told...it's two months away....
so what are they saying? that they are going to start using metatrader? or that they have another software that is like metatrader?
yea only reason I am not with MB is they don't offer charting. If they get with meta4 I am there. As they say they welcome scalpers stradigy.
I was happy with fxdd untill now.
fxspeedster said no complaints with interbankfx either.
anyone have them?I have interbank and I only have one complaint thus far. That is that when they close for holiday's they don't guarantee stops or any executions of any kind. If it's a national holiday according to the holidays' they close on. if you have a trade in play it will not take profit or stop loss it will be in play regardless of the market's movement or your trade's execution orders. When they say they are closed they really mean that you are trapped in whatever you have in play and nothing will execute until the next day they are open. It may work for or against you but you will have no control over it.
honestly though, if all fxdd does is change the way it places the lot size or the way it accepts the code for lot size.and not the actual data feed then its ok. I'll just increase the risk as account grows. Or use a set lot size.
I really didn't intend to launch this into a discussion of brokers. As I said before, I don't see this as a broker issue at this point. I see it as a platform problem with metatrader. Further it is a problem of lack of interest on the part of metaquotes who made the metatrader software to fix an instability in their platform. The fact that so many brokers use the metatrader platform only says to me that there are none of them really interested in the success of their clients, none of them, nor any of their clients have required this instability to be addressed and resolved.
It's a real shame. It's so close to actually being a valid tool. If it were actually what it appears to be on the surface it would be the best thing going that I know of. They have tried to make something at least, they just fell short of making it valid by leaving it with a flaw that makes it inconsistent.
It still works to execute manual trades but that is about all. Thing is that's not what it was created to be. It was created to do automated API strategy execution and development. It almost does that. alas only almost. If they would only stabalize the backtester so it is consistent. That would go along way to making it actually a real tool instead of just a play toy.