Page 1 of 1

n-step ahead forecasts

Posted: Fri Apr 30, 2010 11:59 am
by lamach
Hi,

I currently have a quarterly model of inflation thus the forecasts i produce from this model are one-step ahead forecasts e.g. QoQ changes

I am unsure how I would produce four-step ahead forecasts e.g. 1yr ahead forecasts for inflation.

The model is as such:

Inflation=Constant+Four quarters of lagged inflation+money growth+error term

Where the 4Q lagged inflation is essentialy a AR(4) model