[Archive!] WRITING A COUNTRY TOGETHER!!! - page 9
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
In general, I think that analyzing one chart and using it for MTS is utopia, and I realized it a long time ago. Many EAs write excellent MTS for EURUSD, but lose money using another currency, for example USDJPY. Why, for example, in the next year the EURUSD chart will not follow the USDJPY one, and vice versa? That's why it's impossible and even dangerous to analyse a single chart. I just draw a chaotic chart by hand, will your EA make money on it? So I suggest to discuss this topic, who has an opinion?
In general, I think that analyzing one chart and using it for MTS is utopia, and I realized it a long time ago. Many EAs write excellent MTS for EURUSD, but lose money using another currency, for example USDJPY. Why, for example, in the next year the EURUSD chart will not follow the USDJPY one, and vice versa? That's why it's impossible and even dangerous to analyse a single chart. I just draw a chaotic chart by hand, will your EA make money on it? So I suggest to discuss this subject, who has any opinions?
We should discuss it in another branch.
>>This is a topic worth discussing in another thread.
What difference does it make? We can jointly make an idea into an EA and see if it works or not. This thread was created for this very purpose... There may be many ideas... Everyone can propose their own ideas and implement them fully or partially (depending on their level of programming knowledge). Not every successful trader is a programmer.
What's the difference? We can jointly implement the idea into an Expert Advisor and see if it works or not. And this thread was created for that very purpose... There may be many ideas... Everyone can propose their own ideas and implement them fully or partially (depending on their level of programming knowledge). Not every successful trader is a programmer.
Better to concentrate on writing the EA together. In order not to drown in side and distracted topics, of which there are legions, I would try to introduce some joint rules and objectives, so that there is no confusion and tug on the blanket from opening on yesterday's extremum breakout, to mashes on Hrelson, phibs on Pisan, etc.
For example. Here you suggested the idea of opening on the breakout of yesterday's extrema. That should be frozen until the moment when everyone says "Give up, there is nothing more to be gained". And until that moment comes, the Expert Advisors that open upon breaching yesterday's extrema and use additional conditions, filters, closes, etc. that do not worsen the result, can be posted in the thread. For my taste, I would disallow reducing the number of trades so as not to dilute the original idea. I would also allow classification of Expert Advisors that improve readings of other symbols without worsening the results of the first one. What do you think?
I see that without any common rules, everyone will be able to overload the development on himself and the thread turns into a bazaar of the type "I don't like yours, while I have a great one! Look at this!"
For example. You suggested the idea of opening on a breakout of yesterday's extremes. And that's worth freezing until everyone says "Give up, you can't squeeze any more."
So I haven't cancelled it... I already wrote in this thread that I managed to get good results using a multi-currency filter. It's another matter if it will be a daily, H4 or weekly breakdown, it depends on one's taste, one may try everything. I am going to use the idea of price breakdown as well.
This is what I managed to get using my filter (my Expert Advisor is not optimized at all...) The test is from 2000 till now. As you can see the result has improved on the whole history, because the source code was zero. Of course, the result is almost nothing. That's why there is no point in presenting it, so I won't try to fool anybody. Offer me your ideas about breakdown of the price level. I will continue to work on my idea. I will periodically post results. I thought everyone will take this idea as a base and try to realize it themselves with their own methods, but there are not so many people who would like to do that... :(
This is what I managed to get using my filter (my Expert Advisor is not optimized at all...) The test is from 2000 till now. As you can see the result has improved on the whole history, because the source code was zero. Of course, the result is almost nothing. That's why there is no point in presenting it, so I won't try to fool anybody. Offer me your ideas about breakdown of the price level. I will continue to work on my idea. I will periodically post results. I thought everyone will take this idea as a base and try to realize it themselves with their own methods, but as it turns out, there are not so many people who would like to do it... :(
Let's start small - with those who are :)
I suggest we don't stray too far from 2009.
For 2009 - did the results improve? Has the number of deals changed?
Post the source code with the filter.
Let's start small - with those who are :)
I propose not to stray too far from 2009.
Did you have better results in 2009? Have the number of deals changed?
>> Put the source code with the filter.
1. Let's get started.
2. I support.... As you need a system that works today, but successful long-term work is welcome :)
3. Improved, wrote in this thread above.
4. At first I will explain the essence of the filter and add the filter itself, so it would be clear what we are working with and how it works. I will definitely post it today, as soon as I have some free time, I have a lot of work to do (main, not programming) ;)
Just thought I'd drop a picture off
Maybe you'll get some ideas.
Quite agree, except you probably don't have a trend detection algorithm...
And from what I understand on this forum, you can't get it out of anyone here with pliers...
If you have any ideas, let's discuss...
Everything below is IMHO.
I don't want to impose an opinion as I think everyone has to go that route themselves, but the way of adding a trend filter is a bad one. This addition reduces profit, although it can reduce drawdown. But at the same time it increases the time to get out of the drawdown. In general, it worsens the statistical parameters of TS.
In my opinion, it is better to follow the way of drawdown reduction using different methods. For example, try to transfer to LOS (or to a given level) at a given level of profit. Or use a stepped or quadratic trawl that takes into account price movement patterns (wavy ones in mind). Or bound the size of TP and SL to the average daily price movement of the last few days. Or... many things you can think of.
The general purpose is to improve smoothness and fluidity of balance curve.