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: times
See also: rvlm lo robwhittle roblm lobrob gph sclmtest neweywest

Macro: kpss
Description: Calculation of the KPSS statistics for I(0) against long-memory alternatives. We consider the two tests, denoted by KPSS_mu and KPSS_t, and respectively based on a regression on a constant mu, and on a constant and a time trend t. As in the KPSS paper, we consider for the autocorrelation consistent variance estimator the truncation lags denoted by L0, L4, and L12. The quantlet returns the value of the truncation lag, the type of the test, i.e., const or trend, the estimated statistic, and the critical value for a 95 percent confidence interval for I(0). If the value of the test exceeds the critical value, a star symbol * is displayed after the test statistic.


Reference(s):

Usage: k = kpss(y)
Input:
y vector
Output:
k vector

Example:
;Non-stationarity test on the series dmus58.dat. Since the evaluated
;statistics are over the critical value for I(0), for all the
;truncation orders, the star symbol is displayed after all the
;computed statistics.
library("times")
x = read("dmus58.dat") ; 25477 obs: log FX rates
x = x[1:1000]
k = kpss(x) ; 
k
Result:

Contents of k
[1,] "   Order   Test   Statistic  Crit. Value "
[2,] "_______________________________________ "
[3,] ""
[4,] " L0 =  0   const   77.7016 *   0.4630"
[5,] " L4 =  7   const    9.7786 *   0.4630"
[6,] " L12= 21   const    3.5996 *   0.4630"
[7,] " L0 =  0   trend   11.1167 *   0.1460"
[8,] " L4 =  7   trend    1.4222 *   0.1460"
[9,] " L12= 21   trend    0.5412 *   0.1460"

Library: times
See also: rvlm lo robwhittle roblm lobrob gph sclmtest neweywest

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

Author: Gilles Teyssiere, 980425. Revised 990120
(C) MD*TECH Method and Data Technologies, 28.6.1999