Interpretation of VECM result

[shilpa gupta]

Hi All
I have run VECM model in Stata to estimate the long run and short run relationship among variables. 11 variables and quarterly data from 2001 to 2020 have been used for the study. All the variables were non-stationary at l(0) but stationary at I(1). Lag selection criteria such as SBIC and HQIC showed Lag 1 and Vecrank (Trace and maxi-eigenvalue) command suggested 4 cointegration equation. After applying the vec command, the error correction term or speed of adjustment was negative and significant for 3 equations (1^st, 2^nd and 4^th equations) out of 4 equations for target variable.

1) Is the result significant or all the 4 equations have to be significant?

2) What will be the short-run coefficient as there only 1 lag has been taken for the analysis?

I am also attaching the results with this.

Are my results are appropriate ?

Thanks

[George Ford]

The lag is the short run effect, the EC term is the long run effect.

Nothing HAS to be significant.

[shilpa gupta]

Thankyou for your reply.

I have one more doubt. As there is only 1 lag used, result is only showing the EC term for 4 equations but not any lag term for short run effect. So in this case what would be the short run effect ?

[George Ford]

vec command?

[shilpa gupta]

vec command

vec in_ECB in_FDI In_FII in_FReserve In_Trade In_ER In_GL In_DCredit Diff_Inflation_Pre_Year Diff_GDP_Pre_Year monthsInterestDiff, trend(rtrend) rank(4)
> lags(1)

[George Ford]

when you set lags(1), you only get the EC term. Set lags(2).