Kindly help in analysing the results of Quantile ARDL model output in Gauss 20. I want to know how to get the p-values for short run and long run estimation.
If someone can provide a detailed blog on Quantile ARDL it will by very helpful.
Thanks & best regards
1 Answer
0
Hello,
Thank you for your suggestion about a detailed Quantile ARDL blog. We always appreciate learning about what type of content GAUSS users find useful.
The current QARDL library does not include a feature for computing p-values. However, I believe that you can calculate the p-values using standard t-statistics. The t-statistics can be computed using the covariance matrices and coefficients These are all provided in the qardlOut output structure. To learn more about the qardlOut output structure see the GitHub README page for the qarld library.
For example, the following will compute the p-values for the bt coefficients (note that you will need to set the degrees of freedom df:
// Get standard errors
stderr_bigbt = sqrt(diag(qaOUt.bigbt_cov));
// Get t-values
t_bigbt = qaOut.bigbt./stderr_bigbt;
// Get p-values
p_bigbt = 2*cdftc(abs(t_bigbt), df);
Your Answer
1 Answer
Hello,
Thank you for your suggestion about a detailed Quantile ARDL blog. We always appreciate learning about what type of content GAUSS users find useful.
The current QARDL library does not include a feature for computing p-values. However, I believe that you can calculate the p-values using standard t-statistics. The t-statistics can be computed using the covariance matrices and coefficients These are all provided in the qardlOut output structure. To learn more about the qardlOut output structure see the GitHub README page for the qarld library.
For example, the following will compute the p-values for the bt coefficients (note that you will need to set the degrees of freedom df:
// Get standard errors
stderr_bigbt = sqrt(diag(qaOUt.bigbt_cov));
// Get t-values
t_bigbt = qaOut.bigbt./stderr_bigbt;
// Get p-values
p_bigbt = 2*cdftc(abs(t_bigbt), df);