UA-cam recently changed the way my content will be monetised. My channel now needs 1,000 subscribers. So it would be amazing if you show your support by both watching my videos and subscribing to my channel if you haven’t done so already. Monetising my videos allows me to invest back into the channel with some new equipment so this small gesture from you will be extremely huge for me. Many thanks for your support….CrunchEconometrix loves to teach, help me stay online.
Thank you so much for the videos. I have a question. Do we have to check the stability of the vecm model? And if the model isn't stable, what should we do? Any methods for it? Thank you.
@@CrunchEconometrix Thank you for the explanation. I want to ask the same thing about stability. I read in a journal that if the model is not stable (there is a modulus > 1) then check the autocorrelation error and non normality distributed to make sure that the model is stable or not. Is it right to conduct these test? Thank you
@@CrunchEconometrix I just want to make sure if I can conduct those test. Because it's hard to find another literature that explain about it except that journal I mention before.
Hi Professor, sorry for asking, is it important to make sure all the independent variables to be significant? Can we still do the forecast if they are not significant? Thank you😊
@@CrunchEconometrix ooo understood Professor😊 my next question is, if the T statistics for my two independent variables are 0.8 and 0.7, are they considered significant too?
Hi Summera, taking the log of a series is just a matter of choice. You can always perform stationary test either on the level or log of the series....it may depend on the functional form of your model.
Dear Professor, Are there any differences between the VAR, VECM and Causal analysis in Time Series and Panel Data? I mean, can we estimate the Time Series and Panel Data using the same procedures, or they differ?
Hi Abdullahi, I mentioned in my video on "Johansen Cointegration Test" that a researcher is disposed to using either of the two...may I know from where (location) you are reaching me?
Hi Serentier, "bad" is not the right word. The outcome simply indicates that the error correction mechanism is not statistically significant in your model. If you change your control variables, you may likely get improved results.
UA-cam recently changed the way my content will be monetised. My channel now needs 1,000 subscribers. So it would be amazing if you show your support by both watching my videos and subscribing to my channel if you haven’t done so already. Monetising my videos allows me to invest back into the channel with some new equipment so this small gesture from you will be extremely huge for me. Many thanks for your support….CrunchEconometrix loves to teach, help me stay online.
Good job Doc
Thanks!!!
Thank you so much for the videos. I have a question. Do we have to check the stability of the vecm model? And if the model isn't stable, what should we do? Any methods for it?
Thank you.
Yes, Azizah testing for VECM stability is very important..Play around with the lags and re-test.
@@CrunchEconometrix I see. Thank you so much for your reply.
@@CrunchEconometrix Thank you for the explanation. I want to ask the same thing about stability. I read in a journal that if the model is not stable (there is a modulus > 1) then check the autocorrelation error and non normality distributed to make sure that the model is stable or not. Is it right to conduct these test?
Thank you
Hi Rifatun, you mentioned that "you read it in a Journal". May I know what confirmation you are asking for?
@@CrunchEconometrix I just want to make sure if I can conduct those test. Because it's hard to find another literature that explain about it except that journal I mention before.
Hi Professor, sorry for asking, is it important to make sure all the independent variables to be significant? Can we still do the forecast if they are not significant? Thank you😊
Hi Airel, your question is ok. The most plausible reasoning is to use significant coefficients for forecasting.
@@CrunchEconometrix it means I can still use the ‘not significant variables’ to do the forecast right?😊
That is not what I said.
@@CrunchEconometrix ooo understood Professor😊 my next question is, if the T statistics for my two independent variables are 0.8 and 0.7, are they considered significant too?
@@CrunchEconometrix Professor, do we need to look at the Adjusted R Square too before forecast based on this VECM results?
Thank you very much for the lecture. My question is, what happens if LM test detects autocorrelation at lag h?
You can re-estimate the model at higher/lower-order lags and re-test.
CrunchEconometrix plz explain why and when we should take log of variables before checking the stationarity of the data?
Thanks
Hi Summera, taking the log of a series is just a matter of choice. You can always perform stationary test either on the level or log of the series....it may depend on the functional form of your model.
Dear Professor,
Are there any differences between the VAR, VECM and Causal analysis in Time Series and Panel Data?
I mean, can we estimate the Time Series and Panel Data using the same procedures, or they differ?
Hi Shaf, the procedures are similar but you will have to seek other online resources in that regard since I don't have a video yet. Thanks.
Thank you so much for the explanation!
@@MShaf-ic9td U're welcome.
Thank you.
We are working with 10 year daily data. So what should be the optimal period selection for variance demposition
With daily data, you can use up to 30 lags. Watch my video on "Optimal lags selection" for more insights.
if the result from trace statistic is different from max - eigen statistic what is the way forward iin this case
Hi Abdullahi, I mentioned in my video on "Johansen Cointegration Test" that a researcher is disposed to using either of the two...may I know from where (location) you are reaching me?
thanks
Hi, if the error correction term is not significant, my P-statistic was [-0.342], is that bad?
Hi Serentier, "bad" is not the right word. The outcome simply indicates that the error correction mechanism is not statistically significant in your model. If you change your control variables, you may likely get improved results.