## ARIMA(1,0,0) output Parameter Estimates - I do not understand the output

New Contributor

Joined:

Sep 20, 2017

I ran a simple AR(1) model on 24 rows of trending data just to see the output.  It provided the AR(1) coefficient and the intercept - however, they made no sense to me.  For example, I thought that if the output for an AR(1) is 0.8 and Intercept = 105, then the forecasting equation would be for y(t) = 105 + 0.8*y(t-1).  But the output of the parameter estimates do not fit this.  What am I missing?  Thanks

1 ACCEPTED SOLUTION

Accepted Solutions

Staff

Joined:

Jul 7, 2014

Solution

According to ARIMA (http://www.jmp.com/support/help/13-2/Statistical_Details_for_ARIMA_Models.shtml#409330),

So, the AR(1) model you fit is:  Y(t)-105=0.8*(Y(t-1)-105) +e(t), or  Y(t)=21+0.8*Y(t-1)+e(t)

2 REPLIES

Staff

Joined:

Jul 7, 2014

Solution

According to ARIMA (http://www.jmp.com/support/help/13-2/Statistical_Details_for_ARIMA_Models.shtml#409330),

So, the AR(1) model you fit is:  Y(t)-105=0.8*(Y(t-1)-105) +e(t), or  Y(t)=21+0.8*Y(t-1)+e(t)

New Contributor

Joined:

Sep 20, 2017

Perfect!  Thanks