Making Outlier and Distribution Robust Estimates Using Panel Quantile ARDL Regression in STATA

Estimation of long #paneldata models having years per country nearing 19 or more tend to be tedious if the data is not normally distributed. This video introduces the #Panel #Quantile #Autoregressive #Distributed #Lag #Model which estimates coefficients using #medians and other specified positions. This model has two advantages, first it is #robust to #outliers / #extreme values and second it can check for changes in coefficient across #distribution.

In this video you will learn
1) how to #encode cross sectional identifiers
2) how to #generate #differenced and #lagged #variables
3) how to use #capture command
4) #estimate panel #quantile equation
5) use the #two-step #ECM equation
6) generating #longrun coefficients
7) Detailed descriptive statistics and panel #unitroot test


Details
Panel Quantile ARDL Model in STATA

〖∆DV〗_it=α_0i+α_1 〖∆IV1〗_it+α_2 〖∆IV2〗_it+⋯+α_n 〖∆IVn〗_it+δ_1 〖DV〗_(it-1)+β_1 〖IV〗_(it-1)+β_2 〖IV〗_(it-1)+⋯+β_n 〖IV〗_(it-1)+ε_it
One set ECM equation
Convergence coefficient δ_1. It must be between -1 and 0
Short run coefficients are α_1, α_2 and α_n
Long run coefficients
Long run for IV1 = β_1⁄δ_1
Long run for IV2 = β_2⁄δ_1
Long run for IVn = β_n⁄δ_1