Page 1 of 1
forecasting
Posted: Thu Jan 02, 2014 2:54 pm
by chisola
hi,
I am Vittorio,
I am studying at university
and I have one problem with eviews:
I have to estimate the revenues of one company until 2022,
(i have the past value of revenues).
how can i estimate with eviews?
i am trying to forecast but I obteined an orizzontal line or this error:
unable to compute due to missing data
thanks
Re: forecasting
Posted: Thu Jan 02, 2014 5:43 pm
by EViews Gareth
You need data for the independent variables over the forecast period.
Re: forecasting
Posted: Thu Jan 02, 2014 8:21 pm
by chisola
i have the balance sheet and the income statement of this company...
i put the values(for example the revenues' growth rates to predict furure revenues) of 5 years to estimate the future values of other 10 years...how can i use forecast option?using other items?how can i write independent variables over period of forecast if it goes from 2013 to 2022?i havent future values...
Re: forecasting
Posted: Fri Jan 03, 2014 7:37 am
by startz
Why don't you show everyone exactly what you've done so far. Maybe your commands, a screen shot, and attach your EViews workfile.
Re: forecasting
Posted: Fri Jan 03, 2014 9:40 am
by chisola
I am sorry.
I have to estimate the future values of income statement and balance sheet's items from 2013 to 2022.
I have this vaulues until 2012...
for example I have the revenues' growth rate from 2008 to 2012.
assuming that this will be fixed at 6% from 2019 to 2022 can i estimate the missing rates(from 2013 to 2018)?
there are other way to estimate?i am trying to click forecast's option of rev's growth rate equation :
Dependent Variable: REV_GROWTH_RATE
Method: Least Squares
Date: 01/03/14 Time: 17:21
Sample (adjusted): 2009 2012
Included observations: 4 after adjustments
Variable Coefficient Std. Error t-Statistic Prob.
C 1.382442 0.302718 4.566771 0.1372
REVENUES 6.49E-05 0.000141 0.461297 0.7248
CORE_BUSINESS -8.71E-05 0.000159 -0.548685 0.6805
R-squared 0.928700 Mean dependent var 0.569875
Adjusted R-squared 0.786099 S.D. dependent var 0.305666
S.E. of regression 0.141369 Akaike info criterion -0.961188
Sum squared resid 0.019985 Schwarz criterion -1.421467
Log likelihood 4.922376 Hannan-Quinn criter. -1.971236
F-statistic 6.512598 Durbin-Watson stat 2.402332
Prob(F-statistic) 0.267021
Re: forecasting
Posted: Fri Jan 03, 2014 9:50 am
by startz
In order to forecast from a regression, you need to provide values of the right hand side variables for the period you wish to forecast. You might want to use an equation for which this is easier for you to do: for example, use lagged revenue growth, REV_GROWTH_RATE(-1), as the independent variable.
Moreover, note that your regression only includes 4 observations.
Re: forecasting
Posted: Fri Jan 03, 2014 1:14 pm
by chisola
can i add other indep var?but I don't think that others are good to explain revenues..what indep var I can add?I put also lagged var now...but it doesn't work...
do u know how people estimate future value of income statement's items?
Re: forecasting
Posted: Fri Jan 03, 2014 3:20 pm
by trubador
What you estimate is the following:
Code: Select all
Y(t) = b0 + b1*X1(t) + b2*X2(t) + e(t)
What you want to forecast is the following:
Code: Select all
Yf(t+h) = b0 + b1*X1(t+h) + b2*X2(t+h)
h is the forecast horizon (1,2,...,6 in your case)
So, as Gareth pointed out, in order to forecast Y(t+h) you need values of X1 and X2 from t+1 to t+6.
Or, as Startz suggested, you can use Y(t-1) as independent variable instead of (not in addition to) X1 and X2.
Then you can use dynamic forecast to generate future values: