<< BNG_ur1 Panel unit root tests Chang_IV1 >>

Grocer >> Panel unit root tests > Chang_IV

Chang_IV

Chang Panel Unit Root Test

CALLING SEQUENCE

res = Chang_IV(arg1,...,argn)

PARAMETERS

Input

* arg1,..., argn = arguments that can be:

  - a panel data tlist (in that case there must be an argument 'namevar = xxx' to indicate the name of the variable in the panel, see below)

  - the strings:

    . 'lagorders=x' with x=%nan (if the lags for the ADF test are to be determined automatically) or a (N x 1) or (1 x N) vector of lags for the ADF test

    . 'pmax=x' with x=%nan or a number if for the maximal # of lags for the ADF test

    . 't_order=x' for the trend order with x:

         -1: no constant, no trend

         0: a constant, no trend (default)

         1: a constant and a trend

    . 'IGF=n' with:

         n = 1 for F(x) = x*exp(-ci*|x|) (Default)

         n = 2 for F(x) = I(|x|<K) Trimmed OLS on [-K,K]

          n = 3: F(x) = I(|x|<K)*x

    . 'signif=x' with x the significance level for the individual ADF tests (0.05, 0.01 or 0.1)

    . 'noprint' if the user does not want to print the results of the test

    . 'namevar=xx' where xx is the name of the variable in the panel (only if the data are in a 'panel data' tlist, see help paneldb)

  - a time series

  - a real (n x p) matrix

  - a string equal to the name of a time series or a (nxp) real matrix between quotes (note that there must be several variables of this type tos be able to perform a panel unit root test)

 

Output

* res a results tlist with:

  - res('meth') = 'Chang IV'

  - res('t_order') = the trend order (-1, 0 or 1)

  - res('t_orderlit') = the trend order in plain english

  - res('y') = (T x k) matrix of data

  - res('Sn') = Average IV t-ratio statistic (distributed as N(0,1) under H0)

  - res('Sn_critical') = Critical Values of normal N(0,1) at 1%, 5% and 10%

  - res('Sn_pvalue') = Pvalue of the average IV t-ratio statistic

  - res('Zi') = Individual IV t-ratio statistics

  - res('Zi_critical') = Critical Values of normal N(0,1) at 1%, 5% and 10%

  - res('Zi_pvalue') = Pvalue of the individual IV t-ratio statistics

  - res('pi') = Individual lag orders

  - res('ADF') = Individual ADF statistics for comparison

  - res('AR_IV_ind') = Individual IV estimates of the autoregressive parameter

  - res('var_resIV') = Variance of Residuals of IV estimates for each unit

  - res('var_IV') = Variance of the IV estimator of the autoregressive parameter

  - res('IGF') = Instrument Generating Function used

  - res('K') = Constant K of IGF function (IGF 2 and 3 only)

  - res('ci') = Factor ci for IGF 1 : F(x)=x*exp(-ci*|x|)

  - res('prests') = boolean indicating the presence or absence of a time series in the regression

  - res('namey') = name of the y variable

  - res('dropna') = boolean indicating if NAs have been dropped

  - res('bounds') = if there is a timeseries in the regression, the bounds of the regression

  - res('nonna') = vector indicating position of non-NAs

DESCRIPTION

Chang (2002) Unit Root Tests on Panel Data (ref: Chang (2002): "Nonlinear IV unit root test in panels with cross sectioal dependency", Journal of Econometrics, 110, 261-292), vol. 108, pp. 1–24).

EXAMPLE

// load the database containing the GDP for 25 countries in the OECD over the period 1963-2003
load(GROCERDIR+'/data/gdpan_oecd.dat');
// retrieve the names of all variables in database
listvar=dblist(GROCERDIR+'/macros/grocer/db/gdpan_oecd.dat')
res =  Chang_IV('log('+listvar+')','pmax=4','t_order=0')
// Example taken from function Chang_IV_d(): provides the Chang test for all variables in the database, taken in logarithm,
// with a constant but not trend ('t_order=0'), the Lag order  set to 4 (arg 'pmax=4'),
// the function F(x)=x*exp(-ci*|x|) used as an instrument generating function and a 5% significance level (default values).
 
res = Chang_IV('log('+listvar+')','lagorders=4*ones(size(listvar,1),1)','t_order=0','signif=0.01','IGF=2')
// Provides now the Chang test for all variables in the database, taken in logarithm,
// with a constant but not trend ('t_order=0'), the Lag order to be determined by the program,
// the maximum number of lags set to 4 (arg 'lagorders=4*ones(size(listvar,1),1)'),
// the function F(x)=I(|x|%lt;K) used as an instrument generating function and a 10% significance level.

AUTHOR

Christophe Hurlin 2004 / Eric Dubois 2008

Report an issue
<< BNG_ur1 Panel unit root tests Chang_IV1 >>