Why some great coders and trading system developers are ignoring Metatrader 5? - page 11
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
If I want to make my own Signals (and I want it) so I will try optimize the settings in strategy tester, and in this case of MT5 - it will take short time for me.
Besides, I am having idea to convert some EAs to mql5 just to find the settings. Because it is easy to optimize EA in MT5 than for MT4.
Because good settings is 50% of the success (the other 50% is good coding and robust manual trading system).
Why would they be ? what relevance does that code have, can you explain please ?
Those are multi timeframe codes.
First one is BrainExpert_v2 EA, and the second one is Electra absolute (v1.17). I traded them for few years.
It is about this one:
"At every-tick modeling, tester pumps all necessary timeframes for the symbol under test independently"
Because I compared. As I explained - we had whole subsection in past for that in some forum.
Anyway, this question is not for me.
I am not using strategy tester results as a proof that EA is profitable.
I am using strategy tester just to optimize the settings only.
Some people might call that curve fitting . . .
In case of backtesting/optimization only - yes.
But if someone created EA, optimized the settings, traded it on demo (forward tested) with this settings for few weeks or few months - so it may be more good than backtesting only.
The most difficult part of this job for MT4 - optimize the setting.
Of course, in case of robust manual trading systems - we do not need to optimize the settings for the long time as we know the settings.
If I want to make Signal to be a vendor so I will convert few profitable MT4 EAs to MT5 just to find good settings.
Scalp_net EA for example .... I spent few months optimizing with MT4 ... and I found good settings for EURUSD M5 only.
It was many years ago. After that - I forward tested it till today using VPS for example.
And I still can not find the settings for the other pairs in case of MT4.
But it will take from few minutes to few hours in case of MT5.
And I still can not find the settings for the other pairs in case of MT4.
But it will take from few minutes to few hours in case of MT5.
As far as I am concerned that suggests your EA is no good.
Perhaps you can explain why you give more weight to 30 days of forward testing on one currency pair than 4 years of Strategy Tester testing on multiple pairs ?
As far as I am concerned that suggests your EA is no good.
Perhaps you can explain why you give more weight to 30 days of forward testing on one currency pair than 4 years of Strategy Tester testing on multiple pairs ?
Do not understand sorry ... we are talking about "let's compare backtesting results with forward testing", right?
As to forward testing so I did it since 2006 on many pairs. Because I trust forward testing more than backtesting.
Do not understand sorry ... we are talking about "let's compare backtesting results with forward testing", right?
As to forward testing so I did it since 2006 on many pairs. Because I trust forward testing more than backtesting.
You said . . .
"How long I forward tested with comparing with backtesting for same period?
many EAs ... from 1 week upto few months"
Why do you think 1 week or a few months is a representative sample time to test an EA ?
We were talking about comparing forward test with backtesting with same condition.
If we are talking about forward testing only (without this comparing) so I think - from 1 week up to 3 months (depends on timeframe) may be enough to show the people that some EA may be profitable (or the settings of some EA may be good settings). Because, some people say, that lifetime of some system on lower timeframe is тщ more than 3 months. After that - we need to improve this system, or create new one.
So, basicly, it was not necessary to do what I did - forward testing EAs since 2006 to have account history for many years.
But ... it is just an opinion ...
But ... it is just an opinion ...
Yep, and one I don't share.
Because, some people say, that lifetime of some system on lower timeframe is тщ more than 3 months. After that - we need to improve this system, or create new one.
If after three months, you need to create a new EA, then the quality of your strategy and your EA could be put into question. If your EA wants to be profitable, you should only adapt the input parameters to the current market situation and your baseline strategy should still be valid.
Otherwise, you just have an EA that works sometimes and sometimes does not, therefore you'll end up losing your money and your time, creating EA's all the time...
It depends on the strategy. There are some strategy that can work for many years with same settings. As to me so I was optimizing the settimngs one time only. But I am talking about people's expectation. But do you know about opinions I was receiving from the people when I was trying to upload those EAs with forward testing statements? There were saying that they do not need something to be profitable for 6 years, or for 1 year. They need something to be profitable for last week and next week too. Because if the system is profitable for many years with same settings so this system is having good week and bad week as well. And bad week may be last week for example. Besides, ROI - some people do not have illussions about it. But the other people are trying to receive big profit for 1 week or for 1 months, and they do not need EA which is profitable for few years consistently only because annual return for this EA may be lesds than 80%.
Example with Terminator EA - my forward testing; just for example only; ROI for 2,000 dollars initial depositis is 35%; it means - if I deposit 2,000 dollars so I will get 2,714 dollars after one year (initial deposit included):
Most popular topics on many forums are the following:
"I found Holy Glail"