Create a loop for a one step ahead out-of-sample analisys

[Marcos Gonzalez]

Good morning everybody,

I would like to create a kind of loop to perform and out-of-sample analysis for my monthly data adding one more observation each time for the forecats (one step ahead forecats). This is what I am doing:

First, I am running a baseline regression for a variable on itself lagged one period and then I am adding some other variables to check if the forecast improves. So, my baseline model is like this:

reg variable_x l.variable_x

Then I would like to run that regression model (I guess that I can use the command Arima as well) until a specific date, like this,

reg variable_x l.variable_x if date<tm(2012m01)

And after that I want to predict the observation for the next month

predict p_variable_x if date==tm(2012m01)

Then I repeat the previous steps adding one month in each regression and predicting the next month:

reg variable_x l.variable_x if date<tm(2012m02)
predict p_variable_x if date==tm(2012m02)

reg variable_x l.variable_x if date<tm(2012m03)
predict p_variable_x if date==tm(2012m03)

.......

And so on...

I have about 120 months to forecast, so it is quite tedious to do it one by one. So I would like to know if there is a way to create a kind of loop to perform that forecasts.

Thanks for your help

[Phil Bromiley]

You'll increase your chances of a useful answer by following the FAQ on asking questions.

It seems to me that you're trying to manually do stuff Stata has programmed. Look at the time series procedures including forecast. I think this is a problem you can probably find solved in one of the examples in the time series documentation.