Lot calculation by Vince - page 2
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
Shuffle around with the calculation of the geometric mean.
:-))) Only now have read for real (for I got used to GURU and their formulas to trust by default) in these formulas, the following picture turns out, namely contradiction, i.e.
here - see underline...
And here - in the Journal tab, I deliberately increased the maximum loss on a trade to 10000, so that the divisor in the formula would be greater and there would be no overflow
you get the following picture:
I.e., see underline below - f=0.9, we increase it further upwards by 0.01 - we have the value f=0.99 in the upper line.
It turns out everything is correct, that according to the formulas given above from his book, the value of TWR will have the maximal value at f=1, and this
fundamentally does not correspond in his words to value of TWR at increase in f - see underlining from his book (extracts), i.e. at growth of f the variable TWR also will unequivocally grow - on his formulas. And he writes the contrary - see underlining that, supposedly, at further increase of f to 1.0 in steps of 0.01 the value of TWR variable will fall - this is nonsense.
f enters the formula as a multiplier, the greater it is, the greater the product...
That's all. So, he is writing something rubbish (I wish I had checked it at once)...
As he wrote: TWR must always have "the highest value" if f = 1.0, whatever the other values of the variables in the formula are, but it's not true... :-)))
What are your thoughts on this issue?
:-))) Only now have read for real (for I got used to GURU and their formulas to trust by default) in these formulas, turns out the following picture, namely contradiction, i.e.
here - see underline...
And here - in the Journal tab, I deliberately increased the maximum loss on the trade to 10000 to make the divisor in the formula bigger and not to overflow
you get the following picture:
I.e., see underline below - f=0.9, we increase it further upwards by 0.01 - we have the value f=0.99 in the upper line.
It turns out everything is correct, that according to the formulas given above from his book, the value of TWR will have the maximum value at f=1, and this
fundamentally does not correspond in his words to value of TWR at increase in f - see underlining from his book (extracts), i.e. at growth of f the variable TWR also will unequivocally grow - on his formulas. And he writes the contrary - see underlining that, supposedly, at further increase of f to 1.0 in steps of 0.01 the value of TWR variable will fall - this is nonsense.
f enters the formula as a multiplier, the bigger it is, the bigger the product...
That's all. So, he is writing something rubbish (I wish I had checked it at once)...
As he wrote: TWR must always have "the highest value" when f = 1.0, for all other values of the variables in the formula, but it's not true... :-)))
What are your thoughts on this issue?
I would start with a formula after all. Make your own formula for optimum f. Although I have already done it with my own data and with different depth of calculations. If the depth is not big, then there is no problem with calculations. At big depth it is necessary to pass to approximate calculations. Formulas for approximate calculations can be derived
Transaction - profit or loss on transaction i (with opposite sign ...).
Accordingly, it turns out:
- if a profit, TWR goes up;
- if a loss, TWR decreases.
So TWR will not grow unambiguously. Either you have all trades in profit, so choose the biggest lot! :)))
I would still start with the formula. Make your own formula for optimum f. Although I have already done it on my own data and with different depth of calculations. If the depth is not great, then there is no problem with the calculations. At big depth it is necessary to pass to approximate calculations. You can derive formulas for approximate calculations.
I see. I'll have a look. I mean, is the formula even correct? Initially?
Seems to be right, there goes the variable "Deal" with a negative sign... I will see how it works on "average" Expert Advisor... At the moment the value of f in this test Expert Advisor is constant and equal to 0.99 when the counter is not overflowing...
And where do you find the greatest_probability? Ahh... You're deliberately asking that yourself.
So TWR will not grow unequivocally. Either you have all trades in profit, and that means you select the biggest lot! :)))
Wooooo!!! :-))) Thanks for the tip, I'm really over 90% profitable trades... :-))) That's why f = 0.99. :-)))
With an "average" EA it would be, there's a variable "deal" with a negative sign... That's right. Looking further.
Where do you find the biggest_lose?
The biggest_loss is in the reports after testing the owl: "biggest loss".
What is the task?
Optimized Expert Advisor, run it in the test for the optimization period + (15-20)% forward, look at the report, find the optimal f in the de-item, put it on a demo account with volumes of lots by this found value using the method of geometric mean optimal f. Pages 30-32 in the trailer.
The biggest loss - is found in the reports after testing the owl: "largest losing trade".
What is the task?
Optimized the Expert Advisor, run it in the test for the optimization period + (15-20)% forward, watch the report, in the de-initialization we find the optimal f, put it on the demo account with lot volumes by this found value using the method of geometric mean optimal f.
Find this value programmatically among the trades you take from the history! ;)
Although this parameter has no effect on anything other than the sign in the formula.
Find that value programmatically! ;)
So I'm looking for it, look at page 30-32 of the trailer - see my previous post, in de-ink - I'm looking for it programmatically... There's no other way.
The biggest loss - is in reports after testing owl: "biggest losing trade".
What is the task?
Optimized Expert Advisor, run in the test for the optimization period + (15-20)% forward, watch the report, in the de-initialize, find the optimal f, put it on the demo account with lots volumes by this found value using the method of geometric mean optimal f.
We need to move from calculating later to calculating on the fly. And of course enter the minimum and maximum risk. The formula can change the lot size in predefined parameters. If you use lot 0, you have to make calculations based on virtual trading.