Name

cadf — ADF statistic for residuals from a cointegrating regression

CALLING SEQUENCE

[rescadf]=cadf(p,l,namey,arg1,…,argn)

PARAMETERS

Input

• p: order of time polynomial in the null-hypothesis

 - p = -1, no deterministic part

 - p =  0, for constant term

 - p =  1, for constant plus time-trend

 - p >  1  returns no critical values

• l: # of lagged changes of the residuals to include in regression

• namey: a time series, a real (nx1) vector or a string equal to the name of a time series or a (nx1) real vector between quotes

• argi: arguments which can be:

 * a time series

 * a real (nx1) vector

 * a string equal to the name of a time series or a (nx1) real vector between quotes

 * the string 'noprint' if the user doesn't want to print the results of the regression

 * the string 'dropna' if the user wants to remove the NA values from the data

Output

• rcadf= a tlist with

  . all of the arguments of the second stage regression

• and

  . rcadf('cointrel') = tlist with all the arguments of the first stage regression (see ols() for a description of all these arguments)

DESCRIPTION

Computes augmented Dickey-Fuller statistic for residuals from a cointegrating regression, allowing for deterministic polynomial trends.

EXAMPLE

1) r = cadf(0,6,'illinos','indiana')
2) r = cadf(1,4,illinos,indiana)
 
Example 1 is taken from function cadf_d. Illinos and indiania are the names of two variables taken from data base jpl.dat 
and these names will be used for printings. First entry is set to 0, which means that no trend is imposed in the
cointegrating regression. Second entry is set to 6, which means that there are 6 lags to the residuals in the second 
stage regression. Example 2 imposes a trend in the cointegrating regression, 4 lags on residuals in the second stage 
regression and the variables are named 'endogenous' and 'exogenous' respectively when the results are printed.

               

AUTHOR

Eric Dubois 2002-2007