Sharpe Ratio - is it correct? - page 3 123 New comment Andriy Moraru 2011.01.16 13:31 #21 So, where's the error here? Does MT5 use some other formula for Sharpe Ratio or did I miss something in my calculations? Slava 2011.01.17 10:46 #22 enivid:So, where's the error here? Does MT5 use some other formula for Sharpe Ratio or did I miss something in my calculations? You have missed divider in the StdDev formula StdDev of HPR = Sqrt(((1.0002 - 1.00014999)^2 + (1.00009998 - 1.00014999)^2))/2) Andriy Moraru 2011.01.17 14:32 #23 stringo: You have missed divider in the StdDev formula StdDev of HPR = Sqrt(((1.0002 - 1.00014999)^2 + (1.00009998 - 1.00014999)^2))/2) Thanks! The results match, but then the phrase in your guide looks rather misleading:The obtained value will be divided by 29 (the amount of trades minus one). In the above example there are 2 trades, that's why I divided it by 1 and not 2. Slava 2011.01.18 13:33 #24 enivid: Thanks! The results match, but then the phrase in your guide looks rather misleading: In the above example there are 2 trades, that's why I divided it by 1 and not 2. We don't use Bessel correction (n/n-1) because we assume standard deviation as biased root mean square deviation. BTW 2 items have not statistical meaning. More items amount - less difference between biased and unbiased square deviations Detect volatility a trading strategy based Standard deviation channel width Andriy Moraru 2011.01.18 19:27 #25 stringo: We don't use Bessel correction (n/n-1) because we assume standard deviation as biased root mean square deviation. BTW 2 items have not statistical meaning. More items amount - less difference between biased and unbiased square deviations That's understandable. It just may be misleading that you show one method in the article and use a slightly different method in the platform itself. 123 New comment
So, where's the error here? Does MT5 use some other formula for Sharpe Ratio or did I miss something in my calculations?
So, where's the error here? Does MT5 use some other formula for Sharpe Ratio or did I miss something in my calculations?
You have missed divider in the StdDev formula
StdDev of HPR = Sqrt(((1.0002 - 1.00014999)^2 + (1.00009998 - 1.00014999)^2))/2)
You have missed divider in the StdDev formula
StdDev of HPR = Sqrt(((1.0002 - 1.00014999)^2 + (1.00009998 - 1.00014999)^2))/2)
Thanks! The results match, but then the phrase in your guide looks rather misleading:
The obtained value will be divided by 29 (the amount of trades minus one).
In the above example there are 2 trades, that's why I divided it by 1 and not 2.
Thanks! The results match, but then the phrase in your guide looks rather misleading:
In the above example there are 2 trades, that's why I divided it by 1 and not 2.
We don't use Bessel correction (n/n-1) because we assume standard deviation as biased root mean square deviation.
BTW 2 items have not statistical meaning. More items amount - less difference between biased and unbiased square deviations
We don't use Bessel correction (n/n-1) because we assume standard deviation as biased root mean square deviation.
BTW 2 items have not statistical meaning. More items amount - less difference between biased and unbiased square deviations