Estimating Second Gen. Panel ARDL (CCEMG & AMG) Models in STATA

Following the previous video of Panel ARDL (MG, PMG and DFE) this video explores the second generation panel ARDL models named as CCEMG and AMG. The full names of the model are Common Correlated Effects Mean Group and Augmented Mean Group. These models address the cross sectional dependence using the cross sectional average of each variable. This video also guides how to generate short run effects and convergence coefficients.

[mervetuncay]

Thank you very much for your valuable videos sir🌸
Can you please tell me if I should check for heteroskedasticity and serial correlation after conducting this analysis?

[DonaldUgochukwuChukwumaeze]

Thanks for the video. Please, how do one estimate a non-linear/asymmetric ARDL (NARDL) model using DCCE-MG, DCCE-PMG and AMG techniques in Stata?

I am conducting a doctoral research on the asymmetric effects of stock prices on exchange rates in 10 selected African countries using monthly data that span the period 2000-2022.

Techniques for data analysis employed include:
1. Descriptive Statistics
2. Correlation Matrix
3. Cross-Sectional Dependence Test (using Breusch-Pagan (1980) test, since T>N)
4. Slope Homogeneity Test (using Pesaran and Yamagata (2008) test)
5. Second-Generation Panel Unit Root Test (using CADF PURT by Pesaran (2007)
6. Second-Generation Panel Cointegration Test (using Westerlund & Edgerton (2008) test)
7. Heterogeneous Panel Estimation Techniques: (a) DCCE-MG vs DCCE-PMG; and AMG

[akilamegnak]

Thank you very much for the video sir. Please sir I am looking for the short -term and long -term effects of exchange rate deterrents. To achieve I did the dependence tests (there is dependence), a second generation unitary root test; There is heterogeneity and lack of corner. And I wanted to see what is the appropriate estimate method sir?
Thank you sir

[bhawnamudgal]

Sir, when to apply CS-ARDL or CCE, AMG in case of second generation panel ARDL techniques?

[nikolaizaicev]

And where is the ardl? laged variable, diferences, etc?

[avs]

Hi Sir,
It is compulsory to have all significant relationships in CS-ARDL?

[akilamegnak]

please sir i need help for the problem of too many variable specified in MG estimation model.

xtpmg d.lTC d.lPr d.lgv d.liv d.lov d.lter d.De, lr(l.lTC l.lPr l.lgv l.liv l.lov l.lteR l.De ) replace mg

Here is my MG regression but when I run on stata 17 I see the error too many variable specified