URGENT: AIC and SBIC criterion for forecasts

[stasibab]

Hello,

I would like to now if it is possible to estimate the bic and aic criterion for some forecasts.

For example if I have the 3 forecast results f1,f2,f3 and the original series 'sample'

if i perform a regression for each one, meaning for example dependant variable sample and independent f1, I will get AIC1 and BIC1? Am I right??

Thanx in advance!

[EViews Gareth]

I don't understand the question.

If you run a simple regression, then, yes, the output will produce AIC and BIC.

[stasibab]

The point is to estimate the quality forecast by the bic and sic, I was just not sure if to get the bics and sics have to do a simple regression with each one or with all the forecasts together like: dependant sample and independent f1 f2 f3 ;
it is just when I use one forecast the bics and sics of all three are almost equal and I thought that I did something wrong...

[EViews Gareth]

Well it isn't clear what you mean by "AIC of forecast". The AIC is a statistic based upon a likelihood function. I don't see how that fits in with a forecast.

[stasibab]

Well it isn't clear what you mean by "AIC of forecast". The AIC is a statistic based upon a likelihood function. I don't see how that fits in with a forecast.

To make it more clear I have two series f1, f2 and I want to calculate their AIC1,AIC2 and BIC1,BIC2! This is what I want. I am a bit confused on how to do it with eviews...

[EViews Gareth]

A series doesn't have an AIC, thus your question doesn't make sense.

[stasibab]

Gareth I know this will make you laugh, but your last comment just cleared my mind on a lot of things I was troubled!! Sorry for the stupid question and thank u for the immediate response!!