I have time series variables that are integrated of different order hence I want to undertake bounding test cointegration test byPesaran at al (2001). What procedure is used to carry out this year using eviews ?
(1) Testing for unit root and making sure that none of the variables is I(2) by using the conventional ADF, PP and also the structural break embedded unit root test
(2)ensure the dependent variable is I(1) and perhaps at least one of the independent variable is also I(1).
Plot the graph of the series to see if there's structural break in the series and you can as well confirm this by using the chow test. If there's structural break , include it as exogenous variable because this can cause model instability
3. Select appropriate lags by using the various information criterion.
4. Estimate your ARDL , check the ECM If its negative, significant and less than 1.
5. Do other diagnostic test such as autocorrelation, heteroskedasticity,Cu sum and cusum square test for model stability.
1. Get a feel for the nature of the data.
2. testing unit root and make sure that all the variables are i(0) or i(1) or mix of the two, i(0) and i(1). But note that it cant be used when there is i(2) or higher.
3. Determine the appropriate lag structure of the model selected such AIC, SC.
4. Ensure the model is dynamically stable.
5. Perform the Bounds Test to see if there is evidence of a long-run relationship.
6. If a long-run relationship exists, estimates long-run levels model and then ECM.
7. Use the results of the model estimated above to measure a short-run dynamic effect and long-run equilibrium relationship between the variables.
Hi, you can also watch this YouTube channel https://www.youtube.com/watch?v=qovhIG_WzoE.
Dear Rolle,
I recommend/endorse the procedure or steps outlined by Md Sumon Ali , and emphasise that (if correctly estimated) the ECM coefficient should have a negative (-) sign and statistically significant (at 5 percent) as indicated by Hamid Muili .
All the best!
Dear Rolle,
I also recommend @ Md Simon Ali's and Hamid Muili's procedures.
My kind regards.
Dear Rolle,
The decision to adopt ARDL Bound Testing or any other one can be made based on the outcome of Unit Root testing. Where the order of integration of the variables must be between I(0) and I(1) and not I(2).
Hi
first of all check the variables Unit root test.
Then if your some your variable become I(0) and some of them become I(1) then you can use from Pesaran (2001).
first do your PP and ADF test then select your lags and do your long run form and bound test.
Thanks.
- Badges
- Science topic
- Similar topics
- Engineering
- Modeling