<< agf Business cycle tools bkfilter >>

Grocer >> Business cycle tools > banerji

banerji

performs banerji test to evaluate the leading profile of a series against another one

CALLING SEQUENCE

[rba]=banerji(Pr,Tr,Pt,Tt, arg1,...,argn)

PARAMETERS

Input

* Pr = a vector of through dates for the reference series

* Tr = a vector of through dates for the reference series

* Pt = a vector of through dates for the competeting series

* Tt = a vector of through dates for the competeting series

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

  - 'lead = xx': the number of leads to test (default is 4)

  -   the string 'noprint' if the user doesn't want to display the results of the regression

 

Output

* rdfa = a results tlist with

  - rba('signi') = confidence for rejection of the null hypothesis of lead not significant

  - rba('sum_ini') = initial sum of time difference

  - rba('dates_extra')= dates of extra-cycle in the reference serie

  - rba('PT_extra') = type of each dates (P or T) in the extra-cycles

DESCRIPTION

This function provides randomization test to evaluate the leading profile of a serie against another one. It tests the null H0: leads not significant.

EXAMPLE

load(GROCERDIR+'/data/BusinessClimate.dat');
 
// performs Bry-Boschan turning points dating procedure
bounds('1990m1','2005m5')
rbe = brybos('clim_be','proc=''bb''');    
rfr =brybos('clim_fr','proc=''bb''');    
bounds()
 
// French business climate is the reference series
Pr = rfr('P') ;
Tr = rfr('T') ;
Pt = rbe('P') ;
Tt = rbe('T') ;
 
// banerji with 4 month lead
rba =banerji(Pr,Tr,Pt,Tt,'lead=4');
// This Example is taken from banerji_d(): after having performed the bry-boschan turing points dating procedure,
// the function tests the leading profile of belgium business climate over french one since the 90's.

AUTHOR

Emmanuel Michaux 2005

Report an issue
<< agf Business cycle tools bkfilter >>