Annualized Sharpe Ratio Calculation - according to timeframe vote results
Diskusija
Jul 24, 2011 at 14:51
2,534 Views
4 Replies
Biedrs kopš
12 ieraksti
Jul 24, 2011 at 14:51
the current sharpe ratio is not of much use since its not very usefull for comparisons.. which is the whole idea of the ratio
it needs to be annualized so that returns can be compared accross different strategies easier
i.e. daily timeframe you need to adjust (sharpe ratio) x sqrt(250)
minute timeframe need to adjust (sharpe ratio) x sqrt(250*11*60)
and so on..
then we should see numbers between 1-2> if the strategies are good i.e.
it needs to be annualized so that returns can be compared accross different strategies easier
i.e. daily timeframe you need to adjust (sharpe ratio) x sqrt(250)
minute timeframe need to adjust (sharpe ratio) x sqrt(250*11*60)
and so on..
then we should see numbers between 1-2> if the strategies are good i.e.
"Every System is Profitable, One only doesn't know the conditions that it is profitable in"
Biedrs kopš
179 ieraksti
Jul 25, 2011 at 21:43
(labots Jul 25, 2011 at 21:51)
I think the Sharp is not OK. With profitable strategies that have a lot of losing position, will never be high, despite the very good or excellent also gain.
Only your proposal is such that when my current Sharp 0.18 for 1 minute timeframe after conversion will come out ~ 73!
Only your proposal is such that when my current Sharp 0.18 for 1 minute timeframe after conversion will come out ~ 73!
I'm lazy.
Biedrs kopš
12 ieraksti
Aug 10, 2011 at 20:03
yes i see, im not sure though if i got the annualisation right for the minute timeframe, need to check it
"Every System is Profitable, One only doesn't know the conditions that it is profitable in"
Biedrs kopš
3 ieraksti
Dec 27, 2012 at 15:47
The main problem with SR on myfxbook is that the formula that used is not clear. To me at least.
Seems like a daily SR, but in that case I do not understand the way the 'standard deviation' is calculated.
Other problem, and this seems to be a zero divide or other formula implementation problem, is, that with many strategies the value displayed is 0.00 even though there is a normal profit curve displayed. And there is no real reason I can think of for that.
I have no problem with having only daily SR displayed as the number of trading days in a year may differ, and anyone can do simple multiplication by sqrt(250) or sqrt(251).
The formula for the calculation needs to be clarified (and displayed along with stats, instead of just the childish 'the higher the better' comment) and the zero problem fixed.
Simplest way to calculate SR would probably be to take profit value divided by number of days (the profit curve charts are in daily timeframe too) and divide it by std dev on those same values. It's pretty straightforward I believe.
Seems like a daily SR, but in that case I do not understand the way the 'standard deviation' is calculated.
Other problem, and this seems to be a zero divide or other formula implementation problem, is, that with many strategies the value displayed is 0.00 even though there is a normal profit curve displayed. And there is no real reason I can think of for that.
I have no problem with having only daily SR displayed as the number of trading days in a year may differ, and anyone can do simple multiplication by sqrt(250) or sqrt(251).
The formula for the calculation needs to be clarified (and displayed along with stats, instead of just the childish 'the higher the better' comment) and the zero problem fixed.
Simplest way to calculate SR would probably be to take profit value divided by number of days (the profit curve charts are in daily timeframe too) and divide it by std dev on those same values. It's pretty straightforward I believe.
Biedrs kopš
3 ieraksti
Jan 22, 2013 at 12:52
Okay, after further looking at it, the SR calculation on myfxbook seems to be a simple mean of net profit of all analyzed positions divided the standard deviation of those net profits.
This is NOT hos SR should be calculated. It needs to have a time component to it. As I said my previous post, the easiest way to do it would be if they took calculated the mean of daily profits and divided them by their standard deviations. Daily profits are already calculated and displayed on the Profit chart. This would calculate the correct DAILY Sharpe ratio.
Would be lovely if this was fixed.
This is NOT hos SR should be calculated. It needs to have a time component to it. As I said my previous post, the easiest way to do it would be if they took calculated the mean of daily profits and divided them by their standard deviations. Daily profits are already calculated and displayed on the Profit chart. This would calculate the correct DAILY Sharpe ratio.
Would be lovely if this was fixed.
*Spams netiks pieļauts, un tā rezultātā var slēgt kontu.
Tip: Posting an image/youtube url will automatically embed it in your post!
Tip: Type the @ sign to auto complete a username participating in this discussion.
