How to Estimate and Interpret Panel ARDL using Eviews

[@mosh71]

I am tryign to use ARDL for a panel. Dependent variable is renewable energy capacity. For few countries for first few years there was no capacity (therefore 0). Then for few years may be low value (such as 0.7kw per capita). I think i cannot use logarithm for 0 and value below 0 such as 0.7 will become negative when done logarithm. Will it be ok to use values as they are without making them into logarithm for dependent and independent variables. please advise. Thank you.

[@aroojnaz3885]

What if cointegrating equation (speed of adjustment/ error correction term) value is positive and insignificant (In Short run equation panel of ARDL results)? kindly explain

[@bbouchra1000]

Thank you so much. it is clear and well explained. I have one question please, How can we include regressors being fixed ( for example the initial level of GDP) in the specification. When I do so, I have a message error "Near singular matrix"

[@waliullahchannel]

Don't use time invariant variables

[@mansoorahmed4150]

Thank you for sharing valuable info. I have one question, when i perform panel ARDL model, the results show error of near singular matrix. what does it mean, i cannot run the regression. waiting for ur valuable suggestions.

[@moonsafar5718]

Thank you very much, could you please list steps of estimation PANEL ARDL to follow them. I meant step by step to end of estimation.

[@philippetrape9295]

Just excellent! Thank you very much for keeping it simple and didactic. When I think how Econometrics was taught 30 years ago this is just a great leap forward allowed by software technologies. Kudos! I have two questions if you allow me: (i) You have not commented the Coeq - 0,88 which I suppose is the Error Correction Term (ECT) and the speed of adjustment to long term equilibrium. Correct? (ii) My rough and outdated knowledge of panel econometrics was based on the use of pooled OLS and the fixed Effect vs. Random approach based on the Hausman Test (1978). Is this approach still valid or totally outdated given its lack of consideration for stationarity analysis which became a central issue in time series econometrics? Best regards.

[@lovenepal8756]

Please help me how to get country wise long run coefficient???

[@dr.sureshmago9211]

Thanks sir for this valuable lecture. Please tell while applying ARDL model in panel data, other features of optimal regression equation like high r square and adjusted r square, Durbin-Watson value (near two), significance of Prob(F-statistic ), no serial correlation, No hetroscedasticity and normality of residuals is to be ensured or not need to be checked?? Regards

[@dr.kashikaarora8454]

Same query. Sir, please answer this also

[@econacademy16]

They are not of concern in panel ardl

[@teohcaini5183]

Hi Prof.. Just curious can we run Bound test using panel data in e view?

[@chelseajacob7272]

Hello Prof, if a series is of order 2 what of model will use to do the estimation

[@emregokceli5087]

Hi Sir, thank you for the video. I have just a quick question. Could please let me know whether l should look at cross countries coefficients in the short run? Is it not enough to interpret the results you obtained for whole country? Thanks

[@econacademy16]

It depends on your objectives

[@emregokceli5087]

@@econacademy16 thank your for reply. One more question please. Let's assume that l am looking at the effect of investment on gdp for 10 developing countries in the short and long run. In this case, do l need to interpret the cross countries coefficients in the short run? Actually l could not understand the different between the general regression and cross countries coefficients. Thanks

[@sabashah1160]

I followed steps but got “near singular matrix” error message... please tell me where did I go wrong

[@econacademy16]

There might be multi collenearity in your data or the observations might be less

[@sabashah1160]

What should I do to avoid it

[@econacademy16]

Check multicollinearity

[@sabashah1160]

Thanku

[@oumaimasahtout2082]

thank you for this very explanatory video. if it is possible to make us an ardl panel on stata

[@econacademy16]

I will publish it in a day or two

[@oumaimasahtout2082]

@@econacademy16 Thank you, sir. I have a question for you plz . What's t-statistic in ARDL (stata)

[@econacademy16]

T-stats is used to check significance of a variable

[@pushpkumar8474]

Thank You

[@econacademy16]

You're welcome

[@iqrashahbaz9708]

Its realy helpful. .in short.run conintq value intrprtaion nt mch important?

[@ekhaifa9732]

Can I perform the estimation without constant?

[@econacademy16]

Yes but highly non recommended because it will produce biased results

[@ekhaifa9732]

@@econacademy16 Thanks for your reply. When I use the constant, I receive the message "near sigular matrix", which allows me to eliminate the constant.

[@econacademy16]

@@ekhaifa9732 i think you have less number of observations try using low level of lags instead

[@ekhaifa9732]

@@econacademy16 Thanks for your reply. I have N=12 and T=16. I use 1 lag for all the variables. (I would like to use the Panel ARDL model). But if it doesn't work, which model is preferable in this case?

[@econacademy16]

@@ekhaifa9732 are your variable I(0) and I(1)