One of the requirements for time-series to be covariance stationary is : 1. Constant and finite expected value over time. What exactly does it mean? Mywife & I are both Level 2 candidates & we have been fighting over it. According to her expected value is same as predicted value but I don’t agree with it as CFA curriculum says E(Yt)=U , where t=1,2,3… T Thanks, AG
It just simply says that that it has a constant mean throughtout time
What happens if one of you passes and the other fails? Because statistically this is going to happen. Will there be some gloating or consoling?
The mean and variance should stay constant over time
- The predicted value is something you get from a model (probably). The covariance stationary condition has nothing to do with a model that you fit to the data. It has only to do with the properties of the underlying data. “Constant” means that there is not some long term trend or other structure to the data. That means that (as CFAI source says) that if I don’t know anything about any of the other observations, E(Y[t]) = k for all t and some k (that notation you use above suggests that the expectation is random as U would generally be a random variable). “Finite” means that k can’t be infinity as nothing in the time series analysis works very well if the expectation is infinite (infinite expectations happen all the time, e.g., if you and your wife start having kids when you get over this rough patch, the number of kids you need to have to have an equal number of boys and girls has an infinite expectation). 2) If this marital problem becomes too burdensome, www.harborbridgect.com
joey - is your wife a psychiatrist there?
Not required. Both of us passed. Reggie Wrote: ------------------------------------------------------- > What happens if one of you passes and the other > fails? Because statistically this is going to > happen. Will there be some gloating or consoling?