Keywords - Function groups - @ A B C D E F G H I J K L M N O P Q R S T U V W X Y Z

Library: panel
See also: panfix panrand pandyn

Quantlet: panunit
Description: computes the unit root statistics for panel data

Usage: output = panunit(z,m,p,d {,T})
Input:
z NT x {2+}k matrix of panel data
m scalar, specifying the position of the variable which should be tested (do not count the time and cross section index)
p scalar, number of lagged differences
d scalar, this parameter selects the deterministic part of the regression, where d = 0: when no deterministics is present, d = 1: a constant is considered, d = 2: represents a trend
T optional scalar, number of time periods for a BALANCED data set, which is a data set with the same T for all individuals (= BALANCED PANEL)
Output:
output string with output table

Example:
library("panel")
z=read("dynpanel")
panunit(z,2,1,1)

Result:
Contents of output
[ 1,] "====================================================="
[ 2,] "Pooled Dickey-Fuller Regression:  2'th variable      "
[ 3,] "====================================================="
[ 4,] "PARAMETERS        Estimate     robust SE      t-value"
[ 5,] "====================================================="
[ 6,] "Lag[1]=            -0.0062       0.0289        -0.215"
[ 7,] "Delta[ 1]=         -0.0057       0.0538        -0.107"
[ 8,] "const=              0.1196       0.0607         1.971"
[ 9,] "====================================================="
[10,] "N*T=    375      N=   25           With constant     "
[11,] "Unit root statistics:                                "
[12,] "STATISTIC      Value  crit. Value (5%)  mean variance"
[13,] "====================================================="
[14,] "B/M (1994)     -0.369     -1.65       0.000     1.000"
[15,] "L/L (1993)      0.845     -1.65      -0.601     0.878"
[16,] "mod. L/L        2.107     -1.65       0.000     1.000"
[17,] "I/P/S (1997)    0.367     -1.65      -1.472     0.954"
[18,] "====================================================="



Author: J. Breitung, 2001010 license MD*Tech
(C) MD*TECH Method and Data Technologies, 05.02.2006