Dear All,

Firstly, sorry I can't attach my result tables here. I am struggling to figure out how to do it, but I hope you can still help me. I will try to be as clear as I can be. Thanks in advance

I have a dataset from 1990 to 2014. I am trying to assess the impact of gdp on emissions using VAR/VECM Model. So far I am still trying to run the VAR model.

To determine optimum lag size I have used the following command:

varsoc lco2 lgdp lpopg lfdi lpopd, maxlag(4)

From some literature I have reviewed, it advised that for annual data set, I use a minimum of 1 lag. But I choose 4 and out of curiosity I also applied the command on 5, 6 and 7 lags. I didn't want to go beyond this cause I still want to have as much observations as possible.


Following this and looking at the results of the AIC, SC, HO and FPE I got the lag size with majority (*). I ran the VAR model using

var lco2 lgdp ltrade lpopg lfdi lpopd, lags(1/4)


My results for lpopg at lag 2 and 3 are omitted. For lpopd I have omitted results at lags 2 and 3 and for lag 4 I have what appears to be a period (.)

When I run the model with 1 and 2 lags using the commands

var lco2 lgdp ltrade lpopg lfdi lpopd, lags(1/1)
var lco2 lgdp ltrade lpopg lfdi lpopd, lags(1/2)

I have no omitted variables but for lpopd at lag 2 I have (.) while values are present at lag 1.

If I go on to run the var model at lags 5, 6, 7, the number of omitted values and (.) for certain variables increases.


Questions
1. Why am I getting the omitted values and (.)
2. should I stick to lag 1 and 2 if so how do I interpret the results for lpopd (.) under lag2