2
$\begingroup$

I want to apply grangers' Test between GSDP and Electricity production in the state. The state is a newly formed in the year 2000 and hence I have only 13 data points, as mentioned below

Year | GSDP

2000-01 | 29539

01 to 02 | 32493

02 to 03 | 38802

03 to 04 | 47862

04 to 05 | 53381

05 to 06 | 66875

06 to 07 | 80255

07 to 08 | 96972

08 to 09 | 99364

09 to 10 | 119420

10 to 11 | 144382

11 to 12 | 163461

12 to 13 | 185060

The test yields that it is not a stationary process. Kpss test yields that it is not a trend stationary process. Hence a the second popular choice comes to be difference stationary process. (Transformation of data yields the same result).

The question is how to convert this process to stationary process ?. Is it advisable to apply Grangers' Test using this limited number of data points?. If not what other methodology I can adopt for the same.

$\endgroup$
  • $\begingroup$ Your series is too short to give any meaningful results. With so few observations the series will most likely not look very mean reverting, hence a non-stationary process. I would get some new data. $\endgroup$ – Plissken Jan 18 '15 at 11:02

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.