13. Vector Error Correction Model (VECM) using EViews || Dr. Dhaval Maheta

[@AshrafulIslam-bp1iv]

How did you determine the t-distribution (critical value) for the long run coefficients? Is there any references? My model is good, also get conintegrating relationship among variable. But I am confused regarding the statistical significance of the long-run coefficients. Please explain or refer any literature. Thanks.

[@rojadhungana8843]

Sir, i selected lag 1 based on SC criteria. When calculating vecm, what lag should I put? As it is said you should loose one lag while doing VECM. Should i use lag 0(1-1) or lag 1 itself?

[@mukaramazhar3847]

Which cointegreation should be used in paper to vonclude the results. Suppose I am checking long-run relationship among GDP PG and FDI.gdp as DV population growth and fdi are InDV. I have run unit root lag lengh VECM but confuse for results conclusion

[@aminaahmedalibelal5676]

Is it advisable to include the optimal lag we obtain when testing for cointegration?

[@DhavalSaifaleeAaryash]

depends on case to case basis

[@АлексейБольдясов]

Hello! Your videos are very usefull. And it open me a lot of opportunities for my economics researches. But I have a question with VECM. Is it necessary to perform seasonal adjustment before cointegration checking and estimation? I analize domestic and world wheat prices. And I think that price pairs for some countries are not cointegrated because it needs to first, exclude seasonality. Thank you in advance

[@DhavalSaifaleeAaryash]

I agree

[@vishwajeetpundkar7841]

Do we check the stationarity of logged variables or their original forms?

[@DhavalSaifaleeAaryash]

We check of both

[@AnamAkram-bn5ez]

sir, i have 8 explanatory variables and johansen cointegration showing 7 cointegrating relationships. VECM will be very complicated in this way. Whats the solution.

[@DhavalSaifaleeAaryash]

Make subset and then run analysis

[@SiwarFerchichi-h9q]

it it necessary to reverse the signs of the coefficients while interpreting the impact of each variable on an other one ?

[@thulabakada]

I know it is only necessary when interpreting the long-run impact, and when looking at the speed of adjustment, you want to see negative signs as that indicates that the variable is adjusting back to equilibrium.

[@tumeloarthurmohale8739]

Good evening Dr, How can i find ECT+ and ECT- in eViews. Am stuck. Kindly assist. Thanks in advance. Kind regards, Tumelo Arthur Mohale

[@DhavalSaifaleeAaryash]

R u talking about NARDL

[@tumeloarthurmohale8739]

@DhavalSaifaleeAaryash Am using Vector Error Correction model. The error term is split into two, the negative and the positive (ect- and ect+)

[@asrawani7190]

in my VECM independent variables, coefficients are negative but statistically significant, e.g in short-run estimates: D_NL New Real GDP _ce1 L1. = -.1084373; z value = -2.45; p value=0.014 NL New Real GDP LD.=.4752317 ; z value = 1.84 ; p value=0.066 NL Tour Forex LD. = -.0560698; z value = -2.64 p value= 0.008 How to interpret

[@aminaahmedalibelal5676]

what about stability diagnostics?

[@aminaahmedalibelal5676]

my variables are daily for the whole year, to estimate the optimal lag, LR FPE AIC and HQ showed the optimal lag is 8. should i have to choose 8? can i test the optimal lags with original data?

[@DhavalSaifaleeAaryash]

No u hv to run VAR to check lags

[@bellisma77]

@@DhavalSaifaleeAaryash yes dr. I did which showed me the choosing by LR FPE AIC and HQ showed the optimal is 8.

[@Pooja-cu9ek]

Sir you tested that model is not homoscedastic is it right in VECM?

[@DhavalSaifaleeAaryash]

For garch we require heteroskedasticity

[@jenemilton4791]

good day sir.. what if my time series are cointegrating but the coefficient of the error correction is positive

[@DhavalSaifaleeAaryash]

it is not good model.

[@Upbbf]

How 0.60 is Less than 0.05

[@DhavalSaifaleeAaryash]

Can u specify the time period