The perfect mechanical trading system. - page 6
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
By the way... the market is dead :)
That's right, such silence (even in the pound), a bit back and forth (yesterday the pound was 55 p, today about 40)
Z.I. The calm before the storm.
I'm going to try it now, the optimization idea!
Another argument in favour of this approach is the assumption that EAs adjusted to the story for the first time (even if not for a long time) trade quite profitably, an example I think can be the Championship - in the beginning there were many more profitable EAs (I think this was just due to the fact that they were adjusted to the story)
Before writing an EA with a tester inside, try to check this hypothesis manually. Suppose for each of the last 10 months prooptimize some EA on the previous 6 months and report the result.
If only it were that simple...
more details in the attached file (from May to the current time ) optimized the included mt4 MACD Sample.
If you check it online, it will take a long time and requires further development of the Expert Advisor (connection of the filter, activation of the error block, etc.)
How about taking part in writing the code as well? :-), with your programming experience it would go much faster!
Your approach has a point too, I wonder what will happen here.
I'm not quite sure what you mean? If it's just a simple story fit, you're welcome to it:
- we optimize expert for the period (for example) from 01.01.2006 to 31.03.2006
- check optimized parameters for the period from 01.04.2006 to 30.04.2006
- optimize expert for the period (e.g.) from 01.02.2006 to 30.04.2006
- check the optimised parameters for the period from 01.05.2006 to 31.05.2006
etc., i.e. we optimize in previous months and check one next month.
....................
Can I give you another hint, i.e. describe point 3 in more detail (how to calculate this price logarithm?)
In MT this is the MathLog(X) function.
The logarithm of price has many advantages over the price itself.
But for pipsing it won't make any difference.
Not quite sure what you mean? If it's just a simple story fit, you're welcome to it:
- we optimize expert for a period (for example) from 01.01.2006 to 31.03.2006
- check optimized parameters for the period from 01.04.2006 to 30.04.2006
- optimize expert for the period (e.g.) from 01.02.2006 to 30.04.2006
- check the optimised parameters for the period from 01.05.2006 to 31.05.2006
etc., i.e. we optimise in previous months and check one next month.
....................
Could you give me another hint, i.e. describe point 3 in more detail (how to calculate this price logarithm?)
In MT this is the MathLog(X) function.
The logarithm of price has many advantages over the price itself.
But for pipsing it won't make any difference.
If you can be more specific, what are the advantages?
I will also add my own proposal.
I have to tell you right away I have not been able to realize the principle described so far, I cannot yet beat MQL :).
First we need a well-filtered trend line, but not lagging like MA. For example we can use Kravchuk's FATL. Then we calculate its derivative and the second derivative, i.e., acceleration and jerk.
Provided both acceleration and jerk are greater than zero (possibly also introducing some delta for false signals), then we buy, when both acceleration and jerk are less than zero
then sell.
But there will still be a lot of false signals. To filter them, I propose the condition that the selected trend line (eg, FATL) is greater than the fractal in the desired direction, but in a smaller time frame.
Exiting a position using a trailing stop, the stop loss at entry into the position is equal to the average size of a candle with the shadows for the last 24 hours in the desired time frame.
The first derivative is the speed, the second one is acceleration.
Practically, acceleration, even in a larger time frame, changes faster than the speed in the current one.
In short, acceleration can simply not be analyzed, there are many false signals.
The speed is sufficient. But not on one TF, but the sum of several ones.
If so, it will turn out approximately the following: 'Pivot Indicator Priliv.mq4'.
And here is the version for a separate window: http://autograf.dp.ua/.
Sources.
Despite the lag of initial MAs, there is practically no actual lag in extremums.
(I have been using it for about 2 times a month, people either visit ICQ or write to e-mail what they like)
You should also try it.
I'll write like a Chukcha, probably not very well put together, just what comes to mind now ...
1а. The price is a strictly positive value, for this reason complexity may arise.
For example, Price +Delta may turn out to be a negative value.
These complications need to be dealt with ...
1б. The logarithm of Price occupies the +- infinity space.
The difficulties described above do not arise.
2. External factors act on the price not aditively, but multiplicatively.
That is, the change in price of an asset with the same impact force is proportional to the price itself.
I.e. for a $10 stock, a +$5 change is not the same as a +$5 change for a $100 stock.
Probably everyone will consider that 10 +-5 is about the same as 100 +-50
In both cases the profit/loss will be the same (with the same amount invested)
3. From p.2 (and some other considerations) it follows that the distribution function of price increments should obey the log-normal law rather than the normal law. I.e. the increments of the logarithm of the price obey the normal law rather than the increments of the price itself.
..........
etc.
lazy to think and remember now ...
It is already considered an axiom among serious MTS-people ...
But once again - if the price increments are small compared to the price itself, there won't be much difference.
It may be significant when considering:
1. more volatile instruments - stocks for example.
2. when considering large periods of history.
3. when considering several instruments at the same time with significantly different prices
and correspondingly significantly different volatility.
I do not want to remember, to formulate, and to write a long treatise...
I'll write like a Chukcha, probably not very well put together, just what comes to mind now ...
1а. The price is a strictly positive value, for this reason complexity may arise.
For example, Price +Delta may turn out to be a negative value.
These complications need to be dealt with ...
1б. The logarithm of Price occupies the +- infinity space.
The difficulties described above do not arise.
2. External factors act on the price not aditively, but multiplicatively.
That is, the change in price of an asset with the same impact force is proportional to the price itself.
I.e. for a $10 stock, a +$5 change is not the same as a +$5 change for a $100 stock.
Probably everyone would consider that 10 +-5 is about the same as 100 +-50
In both cases, the profit/loss will be the same (for the same amount invested)
3. From p.2 (and some other considerations) it follows that the distribution function of price increments should rather obey a log-normal law than a normal law. i.e. increments in the logarithm of the price follow the normal law rather than the price itself.
..........
etc.
lazy to think and remember now ...
Among serious MTS-people it is already considered an axiom ...
But once again - if the price increments are small compared to the price itself, there won't be much difference.
It may be significant when considering:
1. more volatile instruments - stocks for example.
2. when considering large periods of history.
3. when several instruments with significantly different prices are examined simultaneously
and respectively significantly different volatility.
Thank you!, I think we should get better acquainted with it...... .