Comparing a forecast to the mean

For technical questions regarding estimation of single equations, systems, VARs, Factor analysis and State Space Models in EViews. General econometric questions and advice should go in the Econometric Discussions forum.

Moderators: EViews Gareth, EViews Moderator

miorinnovo
Posts: 199
Joined: Sun Mar 01, 2015 5:01 pm

Comparing a forecast to the mean

Postby miorinnovo » Fri Mar 06, 2015 4:17 pm

Can someone tell me how I can compare a forecasts RMSE to the unconditional mean?

startz
Non-normality and collinearity are NOT problems!
Posts: 3798
Joined: Wed Sep 17, 2008 2:25 pm

Re: Comparing a forecast to the mean

Postby startz » Fri Mar 06, 2015 4:29 pm

Regress on just a constant?

EViews Gareth
Fe ddaethom, fe welon, fe amcangyfrifon
Posts: 13604
Joined: Tue Sep 16, 2008 5:38 pm

Re: Comparing a forecast to the mean

Postby EViews Gareth » Fri Mar 06, 2015 4:36 pm

Code: Select all

series ymean = @mean(y) scalar rmse = @rmse(yf, ymean)

miorinnovo
Posts: 199
Joined: Sun Mar 01, 2015 5:01 pm

Re: Comparing a forecast to the mean

Postby miorinnovo » Fri Mar 06, 2015 5:09 pm

Great idea startz, thanks.

EViews Gareth, I tried your code and it says yf is not defined

startz
Non-normality and collinearity are NOT problems!
Posts: 3798
Joined: Wed Sep 17, 2008 2:25 pm

Re: Comparing a forecast to the mean

Postby startz » Fri Mar 06, 2015 5:12 pm

Gareth meant

Code: Select all

scalar rmse = @rmse(y, ymean)

miorinnovo
Posts: 199
Joined: Sun Mar 01, 2015 5:01 pm

Re: Comparing a forecast to the mean

Postby miorinnovo » Fri Mar 06, 2015 5:18 pm

Amazing. Thank you both! :D

EViews Gareth
Fe ddaethom, fe welon, fe amcangyfrifon
Posts: 13604
Joined: Tue Sep 16, 2008 5:38 pm

Re: Comparing a forecast to the mean

Postby EViews Gareth » Fri Mar 06, 2015 6:14 pm

No, I meant yf, where yf is your forecast of y.

miorinnovo
Posts: 199
Joined: Sun Mar 01, 2015 5:01 pm

Re: Comparing a forecast to the mean

Postby miorinnovo » Fri Mar 06, 2015 6:18 pm

Oh, ok. So is it the RMSE of using the mean to forecast though? If so why do we need the forecast variable in there?

startz
Non-normality and collinearity are NOT problems!
Posts: 3798
Joined: Wed Sep 17, 2008 2:25 pm

Re: Comparing a forecast to the mean

Postby startz » Fri Mar 06, 2015 6:20 pm

Do you want to find the rmse of your forecast compared to the mean, or do you want to compute the rmse of the mean compared to the original series?

miorinnovo
Posts: 199
Joined: Sun Mar 01, 2015 5:01 pm

Re: Comparing a forecast to the mean

Postby miorinnovo » Fri Mar 06, 2015 6:29 pm

So I want to compare my forecasts RMSE compared to the RMSE that I would get if I just used the mean as my forecast.

startz
Non-normality and collinearity are NOT problems!
Posts: 3798
Joined: Wed Sep 17, 2008 2:25 pm

Re: Comparing a forecast to the mean

Postby startz » Fri Mar 06, 2015 6:33 pm

Then y, not yf.

miorinnovo
Posts: 199
Joined: Sun Mar 01, 2015 5:01 pm

Re: Comparing a forecast to the mean

Postby miorinnovo » Fri Mar 06, 2015 6:35 pm

Ok, and that code gives the RMSE of the mean as a forecaster, and

scalar rmse = @rmse(yf, y)
is to find the RMSE of the forecast?

startz
Non-normality and collinearity are NOT problems!
Posts: 3798
Joined: Wed Sep 17, 2008 2:25 pm

Re: Comparing a forecast to the mean

Postby startz » Fri Mar 06, 2015 6:39 pm

yes

miorinnovo
Posts: 199
Joined: Sun Mar 01, 2015 5:01 pm

Re: Comparing a forecast to the mean

Postby miorinnovo » Fri Mar 06, 2015 6:45 pm

Wonderful, thank you very much


Return to “Estimation”

Who is online

Users browsing this forum: No registered users and 2 guests