when you perform the VR test you need to specify a sampling interval of the returns i.e. 5, 10, 15, 20 periods etc... 'q' the sampling frequency. perhaps then You can calculate the VR in Excel.

The usual formula for the variance
of a ratio is the following first-order Taylor series approximation:

V(x/y) = [Q^2] [R(x) + R(y) – 2R(x,y)]

where
Q = E(x)/E(y) is the ratio of the means,
R(x) = V(x) / {[E(x)]^2} is the "relative variance" of x,
R(y) = V(y) / {[E(y)]^2} is the "relative variance" of y, and
R(x,y) = Cov(x,y) / {E(x)E(y)} is the "relative covariance" of x&y.

And a second-order Taylor series approximation for the mean of a ratio
is the following:

E(x/y) = Q [ 1 + R(y) – R(x,y)]

hope the above reference will come to help, its difficult i believe to do it.

Atelokin_85

October 22, 2011 at 12:00 am

Really thank you for your help. I will let you know if i managed finally

you need some sort of equation to have values to add them in excel

http://www.mathworks.fr/help/toolbox/econ/vratiotest.html

http://www.freepatentsonline.com/article/Academy-Accounting-Financial-Studies-Journal/179817644.html

when you perform the VR test you need to specify a sampling interval of the returns i.e. 5, 10, 15, 20 periods etc... 'q' the sampling frequency. perhaps then You can calculate the VR in Excel.

Lo and MacKinlay Variance Ratio test

http://www.wilmott.com/messageview.cfm?catid=38&threadid=52056

tLo-Mackinlay variance ratio test

tp://stata.com/statalist/archive/2006-08/msg00045.html

Variance ratio example

http://autospreader.wordpress.com/2010/03/20/variance-ratio-example/

The usual formula for the variance

of a ratio is the following first-order Taylor series approximation:

V(x/y) = [Q^2] [R(x) + R(y) – 2R(x,y)]

where

Q = E(x)/E(y) is the ratio of the means,

R(x) = V(x) / {[E(x)]^2} is the "relative variance" of x,

R(y) = V(y) / {[E(y)]^2} is the "relative variance" of y, and

R(x,y) = Cov(x,y) / {E(x)E(y)} is the "relative covariance" of x&y.

And a second-order Taylor series approximation for the mean of a ratio

is the following:

E(x/y) = Q [ 1 + R(y) – R(x,y)]

hope the above reference will come to help, its difficult i believe to do it.

Really thank you for your help. I will let you know if i managed finally

Cheers