1
$\begingroup$

I was given 2 forecasts: Scenario 1 and Scenario 2 for EU GDP. How can I calculate 2 scenarios (forecasts) for Germany GDP? a) There is a strong correlation between (historic) EU GDP and Germany GDP b) EU GDP is impacted by Germany GDP.

At first I thought that I can predict Germany GDP using sARIMA and later to apply shocks from scenario 1 or scenario 2 to the predicted germany gdp data. However, I am not sure how to make different shocks between EU and Germany, should I scal them using correlation? Any ideas or hints are welcome.

[edit] Added data CSV files

There are 2 CSVs:

a) Example_historic.csv with historic data for Germany and EU

b) Example_forecasts.csv with 2 scenarios for the EU.

Goal: how to build 2 forecasts for Germany?

Improve this question
$\endgroup$
6
  • $\begingroup$ Please post the forecasts you were given ... and also post the historical data for both series $\endgroup$
    – IrishStat
    Mar 26, 2019 at 22:55
  • $\begingroup$ your two scenarios seem identical i.e. the predictions for the euro gdp are the same for the next 12 quarters. $\endgroup$
    – IrishStat
    Mar 28, 2019 at 12:17
  • $\begingroup$ Just to clarify: data between 1980 and 2018 are historic. 2019 - 2021 are forecasts. There are 2 scenarios for EU GDP and 2 missing for Germany. $\endgroup$
    – Lohengrin
    Mar 29, 2019 at 6:24
  • $\begingroup$ They look the same to me ... can you simply post these 24 values 12 quarterly values for each scenario over 3 years $\endgroup$
    – IrishStat
    Mar 29, 2019 at 7:26
  • $\begingroup$ ![enter image description here](i.stack.imgur.com/U948w.png) The second nd third rows are your future values for the euro gdp. They are the same. $\endgroup$
    – IrishStat
    Mar 29, 2019 at 13:16

1 Answer 1

Reset to default
0
$\begingroup$

I introduced your 156 months of data to AUTOBOX which developed a SARIMAX model How to predict the next number in a series while having additional series of data that might affect it? .

The cross-correlation analysis using the pre-whitened series suggested a contemporaneous relationship between the two series. A few anomalous data points were detected along with two downwards level shifts viz 1993/2 and 2006/2 .

The useful equation is here enter image description here with residual plot here enter image description here

The Actual/Fit and Forecast using scenario 1 is enter image description here with forecasts here enter image description here

while using scenario 2 we get enter image description here and enter image description here

The model is presented here enter image description here

Improve this answer
$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.

Not the answer you're looking for? Browse other questions tagged or ask your own question.