timeseries autocorrelation, residuals and residual autocorrelation

CFA Level II
#1

Let’s say our dataset is as follows: { 5, 8, 4, 7, 6, 5 }. n=6 and our mean = 5.8

now I did a AR (1), dependent variable now is, { ,5, 8,4,7,6}

A Can you show me the method on how to compute the residual for this data set (showing one value would be suffice)

B. Can you show me the method on computing the autocorrelation for residuals.

C. Can you show me the method on how to compute the autocorrelation for this series (assuming it is different than the residual correlations you compute in part B). What’s the difference between B & C?

#2

pepp is back!

This is a very good question that should give an excellent understanding of the concepts behind AR models. I’m not on quant these days, so I can’t try it. Could someone try it? A short dataset…show the different parameters, etc.