Machine learning in trading: theory, models, practice and algo-trading - page 3033
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
Line A -maximum will always be above the regression line. The stronger the fluctuations of the trend line, the stronger the movement to 180 degrees (if the balance sheet has not gone to loss) and further to 270, if the balance sheet has gone to loss.
The comparison of the two angles allows us to assess the stability of the balance dynamics. It tells how much the price fluctuated relative to the conditionally linear growth of the balance. It is also possible to plan the expectation matrix, and build an ideal linear balance against it (what is not the target), and calculate deviations in the same way.
The blue line is a new regression trend line.
I think that if you take your line and increase its spread (instability), except for the maximal point, the regression line will remain the same. But the stability will be less.
It is better not to draw lines by eye, but by formula.
I think that if you take your line and increase its spread (instability), except for the maximal point, the regression line will remain the same. But the stability will be less.
That's always in this topic - you tell people an idea, and without testing it, they immediately start to praise it.
That's always in this thread - you tell people an idea, and without testing it, they immediately start praising it.
I did not praise it, I expressed my doubts, because the result is not obvious.
On the graph you have not increased the spread of the original line, but have drawn a completely different one.
Surely the first regression line will correspond to more than one variant of the graph, if you increase exactly the spread. For example, add to each point 10% of its distance to the regression line if it is higher or subtract if it is lower. Other variants of 5, 20, 30 %, etc. will give many variants with different stability, but with almost or exactly the same regression line.
That's how it is:
Except for the maximum point and to it you need to compensate from below to make the line match exactly. I'm sure you can find it. Well, I'm not going to do that. I will deal with the obvious variant, which I expressed earlier.
I was not praising, I was expressing doubts, as the result is not obvious.
On the graph you have not increased the span of the original line, but have drawn a completely different one.
I drew according to your criterion - the coincidence of the maximum and a larger swing. I don't understand what is wrong.
Surely the first regression line will correspond to more than one variant of the graph, if you increase exactly the spread.
Naturally - the method only gives a relative estimation and does not imply restoration of the balance chart movement.
For example, add to each point 10% of its distance to the regression line if it is higher or subtract if it is lower. Other options of 5,20,30 % etc. will give many variants with different stability but with almost or exactly the same regression line.
I added +30% and -30% for clarity - you can see that there are parallel lines to the maximum. And aqua regression crosses black, so it has a smaller degree - the balance is better.
Except for the maximum point and to it it is necessary to compensate from below to make the line coincide exactly. I'm sure you can find it. Well, I will not do it. I will deal with the obvious option, which I expressed earlier.
Do whatever you want - your own version is always better.
I came up with an effective simple method - not the best perhaps, but quite working.
If you make the balance from zero, you can also see that the aqua is better - the angle of the maximum lines shows it, while the regression lines are almost parallel.
I drew according to your criterion - coincidence of the maximum and greater fluctuation. I don't see what's wrong.
Naturally - the method only gives a relative estimation and does not imply restoration of the balance chart movement.
I added +30% and -30% for clarity - you can see that there are parallel lines to the maximum. And aqua regression crosses black, so it has a smaller degree - the balance is better.
Do whatever you want - your own variant is always better.
I have come up with an effective simple method - not the best perhaps, but quite working.
If you make the balance from zero, you can also see that aqua is better - the angle of the maximum lines shows it, while the regression lines are almost parallel.
You have shifted the 2nd line upwards. And I suggested a stronger spread (see my figure). What's above the regression line should be higher (you did), what's below should be made lower (you made it higher too).
But I don't need it. Only if you yourself are interested...
And can you again in 2 words, what do you do? )
And can you again in 2 words, what do you do? )
Inventing their own Sharpe and Sortino) With blackjack and other necessary)
You shifted the 2nd line upwards. I suggested a stronger spread (see my figure). What is above the regression line should be higher (you did it), what is below should be lower (you did it also higher).
But I don't need it. Only if you yourself are interested...
I did exactly as you wrote.
And can we have in 2 words again what you are doing? )
An estimated measure of balance smoothness, taking into account the dynamics of balance growth. You can essentially use it as a fitness function.