ARDL is a simple but effective approach to estimate long run equilibrium models with short run dynamics.

Nonlnear ARDL (NARDL) extends ARDL to a nonlinear framework.

Both ARDL and NARDL can be applied to panel data.

Is the relationship between inflation and returns of the Iraqi stock market to the general index, which represents the model of the self-regression of distributed distributions, which is used when inflation and the returns of sectors are not static at the level of the original data

ARDL is a simple but effective approach to estimate long run equilibrium models with short run dynamics.

Nonlnear ARDL (NARDL) extends ARDL to a nonlinear framework.

Both ARDL and NARDL can be applied to panel data.

it is better to you use NARDL model or also you can use GMM

Thank you Michael John McAleer and Ayoub Zeraibi .

I shall be exploring the NARDL then

To Manali:

It is my pleasure.

I wish you well in your (N)ARDL modelling.

Yeah, as per as the number of observations are available and the data fulfilled the predestination conditions for the technique.

For long run analysis in panel... it is suitable to use Pool mean group. It like a regression using ardl but specific to panel

For your data , T= 5 years, N=255, that means your time is short and cross section is large. Therefore T

Thanks Ahmed Mohamed Dahir for your valuable insights

Ahmed Mohamed Dahir is right; you should use GMM, because Panel ARDL somehow requires use of MG and/or PMG, which also both require large T and small number of cross-sectional units!

Michael John McAleer ,

I would be thankful if you give me your valuable suggestion on (N)ARDL or GMM panel data model?

Because in my data sample T

To Manali:

ARDL and NARDL are dynamic time series models , and are readily available on many software packages.

Panel data models are also widely available.

GMM is a method of estimation and is also widely available on standard software packages.

If you are familiar to use Stata, You can easily implement GMM or even EViews!. You don't have enough time series observations on each country to estimate ARDL and NARDL! You can only estimate panel ARDL; but, as I said wrote, to estimate panel ARDL, you will use (especially in EViews) MG or PMG, which are not appropriate for your panel.

Since your T>N you can not employ Panel ARDL based on the rule (Pesaran paper on consumption, interest and inflation in 23 OECD countries). I do suggest GMM is the appropriate estimation technique as said by Elder Michael and Mustafa. Best

Since your T>N you can not employ Panel ARDL based on the rule (Pesaran paper on consumption, interest and inflation in 23 OECD countries). I do suggest GMM is the appropriate estimation technique as said by Elder Michael and Mustafa. Best

Mustafa Özer

Thank you so much

I shall be then proceeding with GMM model

Dear Manali,

I agree with my collegues. The model you could use for estimates should correspond to you data structure.

As your data sample consists on : T (5 years)

I have a question you have yearly, monthly or dayly data... because I pas thinking that you have montant data

Clery Nomo , I have yearly data

it is really suitable for you to use NARDL or GMM

I agree with suggestion of using GMM, as suggested by other colleagues and STATA is more user-friendly for doing this.