Hi, I was wondering whether someone could explain to me why when we are solving for rho_1 we dont divide the numerator by 499 and the denominator by 500 in the following question as the variance has 500 points and the covariance uses 499 points.
This is the convention used in the formulae for sample autocovariance given on page 40 of the Tables.
Thanks for the reply. Could i also ask why in CT4_2009_04 Q12 iv) we are not summing the final two groups in our chi squared calculation (as the expected number of claims in group 5 is only 1.95 therefore blowing up our statistic?)
Its okay, in the older papers the examiners notes are at the front (missed them) and it said this was a reasonable approach. Cheers.