[Gretl-users] interpreting the “StrucBreak” package for gretl

Yusuf Abduwahab Hassan Sun, 14 Oct 2018 13:02:06 +0000

Good day all.
I am trying to learn the StrucBreak package. can some one please point me
to how to interpret the results below. I tried to follow the sample script
in the package using the PDF guide to the StrucBreak package.


gretl version 2018c
Current session: 2018-10-14 09:33

? include StrucBreak.gfn
StrucBreak 0.21, 2018-01-09 (Riccardo "Jack" Lucchetti and Sven Schreiber)
? open wp.gdt

Read datafile C:\Users\HP\Documents\gretl\wp.gdt
periodicity: 1, maxobs: 44
observations range: 1971 to 2014

Listing 6 variables:
  0) const      1) Nigeria    2) Ghana      3) Togo       4) Senegal
  5) Benin

? list Z = const Nigeria(-1)
? list X = null
? bundle bII = SB_Setup(Nigeria,Z,X,3)
SB_Setup: OK!
? bII.nh = 8 # dustance between breaks
? strings opts = defarray("HetdatYes", "HetvarYes", "HetomegaNo", "HetqNo")
? SB_SetOptions(&bII, opts)
? SB_Global(&bII)

===================================================================
   Options are now:
===================================================================
Robust S.E.:    OFF
Warning: inappropriate if lagged dependent variables
         are included as regressors
Hetdat:  ON
Hetvar:  ON, Hetomega: OFF, Hetq: OFF
beta     :     Nonfixed
Eps = 0.000    Eps1 = 0.100    Max. breaks = 3

Min. distance between breaks = 8


===================================================================
   OUTPUT FROM THE GLOBAL OPTIMIZATION STAGE
===================================================================
Breaks           SSR    Dates

     1    4117.47615    30 (2001)
     2    3394.26266    10 (1981), 30 (2001)
     3    3108.58184    10 (1981), 22 (1993), 30 (2001)

===================================================================
   OUTPUT FROM THE APPLICATION OF INFORMATION CRITERIA
===================================================================
tmp (4 x 3)

      Breaks          BIC          lwz
      0.0000       4.8725       4.9201
      1.0000       4.8242       5.0223
      2.0000       4.8935       5.2483
      3.0000       5.0679       5.5869

The number of breaks chosen by BIC is : 1
The number of breaks chosen by LWZ is : 0
? SB_Tests(&bII)

===================================================================
   OUTPUT FROM THE TESTING PROCEDURES
===================================================================
a) supF tests against a fixed number of breaks
--------------------------------------------------------
    supF(1|0)   supF(2|0)   supF(3|0)
      11.432      11.189      10.897

Critical values:
           supF(1|0)   supF(2|0)   supF(3|0)
10%           10.37        9.43        8.48
5%           12.25       10.58        9.29
2.5%           13.86       11.63       10.14
1%           16.19       12.90       11.12
--------------------------------------------------------
b) Dmax tests against an unknown number of breaks
--------------------------------------------------------
UDmax test: 11.431741
Crit. values:    10%: 10.86 5%: 12.59 2.5%: 14.15 1%: 16.19

........................................................
    WDmax test    (crit. val.)
10%       13.33       11.71
5%       14.37       13.66
2.5%       14.90       15.33
1%       15.87       17.80
********************************************************
supF(l+1|l) tests using global optimizers under the null
--------------------------------------------------------
supF(2|1)     10.93    1981
supF(3|2)     10.70    1993
--------------------------------------------------------
Critical values:   10%    5%  2.5%    1%
       supF(2|1)  12.19 13.83 15.51 17.58
       supF(3|2)  13.20 14.73 16.55 18.31

# eval bII.BreakDates
? SB_Estimate(&bII,"FIX",2)

===================================================================
   OUTPUT FROM ESTIMATION OF THE MODEL WITH 2 BREAKS (fixed)
===================================================================
? SB_PrintEstimates(&bII)

Dependent variable: Nigeria
----------------------------------------------------------------
Subsample 1: 1972 - 1981

              coefficient   std. error     z     p-value
  -------------------------------------------------------
  const        15.0822       7.86913     1.917   0.0553   *
  Nigeria_1     0.729122     0.159461    4.572   4.82e-06 ***

Subsample 2: 1982 - 2001

              coefficient   std. error     z     p-value
  -------------------------------------------------------
  const        53.0242      12.7994      4.143   3.43e-05 ***
  Nigeria_1     0.379843     0.151844    2.502   0.0124   **

Subsample 3: 2002 - 2014

              coefficient   std. error     z     p-value
  ------------------------------------------------------
  const        61.0134      18.9376      3.222   0.0013  ***
  Nigeria_1     0.552676     0.149888    3.687   0.0002  ***

--------------------------------------------------------------
Confidence intervals for the break dates
--------------------------------------------------------------
The 95% C.I. for the 1th break is: 1976 - 1986
The 90% C.I. for the 1th break is: 1977 - 1985
The 95% C.I. for the 2th break is: 1999 - 2003
The 90% C.I. for the 2th break is: 1999 - 2003
********************************************************


-- 






*Yusuf Abdulwahab Hassan.Department of Economics and Development
Studies.Federal University of Kashere,Gombe.+234
8036830166.yabdulwahab(a)fukashere.edu.ng <yabdulwahab(a)fukashere.edu.ng>*

Good day all.

I am trying to learn the StrucBreak package. can some one please point me to how to interpret the results below. I tried to follow the sample script in the package using the PDF guide to the StrucBreak package.

gretl version 2018c
Current session: 2018-10-14 09:33

? include StrucBreak.gfn
StrucBreak 0.21, 2018-01-09 (Riccardo "Jack" Lucchetti and Sven Schreiber)
? open wp.gdt

Read datafile C:\Users\HP\Documents\gretl\wp.gdt
periodicity: 1, maxobs: 44
observations range: 1971 to 2014

Listing 6 variables:
Â 0) constÂ Â Â Â Â 1) NigeriaÂ Â Â 2) GhanaÂ Â Â Â Â 3) TogoÂ Â Â Â Â Â 4) SenegalÂ
Â 5) BeninÂ Â Â

? list Z = const Nigeria(-1)
? list X = null
? bundle bII = SB_Setup(Nigeria,Z,X,3)
SB_Setup: OK!
? bII.nh = 8 # dustance between breaks
? strings opts = defarray("HetdatYes", "HetvarYes", "HetomegaNo", "HetqNo")
? SB_SetOptions(&bII, opts)
? SB_Global(&bII)

===================================================================
Â Â Options are now:
===================================================================
Robust S.E.:Â Â Â OFF
Warning: inappropriate if lagged dependent variables
Â Â Â Â Â Â Â Â are included as regressors
Hetdat:Â ON
Hetvar:Â ON, Hetomega: OFF, Hetq: OFF
betaÂ Â Â Â :Â Â Â Â Nonfixed
Eps = 0.000Â Â Â Eps1 = 0.100Â Â Â Max. breaks = 3

Min. distance between breaks = 8

===================================================================
Â Â OUTPUT FROM THE GLOBAL OPTIMIZATION STAGE
===================================================================
BreaksÂ Â Â Â Â Â Â Â Â SSRÂ Â Â Dates

Â Â Â Â 1Â Â Â 4117.47615Â Â Â 30 (2001)
Â Â Â Â 2Â Â Â 3394.26266Â Â Â 10 (1981), 30 (2001)
Â Â Â Â 3Â Â Â 3108.58184Â Â Â 10 (1981), 22 (1993), 30 (2001)

===================================================================
Â Â OUTPUT FROM THE APPLICATION OF INFORMATION CRITERIA
===================================================================
tmp (4 x 3)

Â Â Â Â Â BreaksÂ Â Â Â Â Â Â Â Â BICÂ Â Â Â Â Â Â Â Â lwz
Â Â Â Â Â 0.0000Â Â Â Â Â Â 4.8725Â Â Â Â Â Â 4.9201
Â Â Â Â Â 1.0000Â Â Â Â Â Â 4.8242Â Â Â Â Â Â 5.0223
Â Â Â Â Â 2.0000Â Â Â Â Â Â 4.8935Â Â Â Â Â Â 5.2483
Â Â Â Â Â 3.0000Â Â Â Â Â Â 5.0679Â Â Â Â Â Â 5.5869

The number of breaks chosen by BIC is : 1
The number of breaks chosen by LWZ is : 0
? SB_Tests(&bII)

===================================================================
Â Â OUTPUT FROM THE TESTING PROCEDURES
===================================================================
a) supF tests against a fixed number of breaks
--------------------------------------------------------
Â Â Â supF(1|0)Â Â supF(2|0)Â Â supF(3|0)
Â Â Â Â Â 11.432Â Â Â Â Â 11.189Â Â Â Â Â 10.897

Critical values:
Â Â Â Â Â Â Â Â Â Â supF(1|0)Â Â supF(2|0)Â Â supF(3|0)
10%Â Â Â Â Â Â Â Â Â 10.37Â Â Â Â Â Â Â 9.43Â Â Â Â Â Â Â 8.48
5%Â Â Â Â Â Â Â Â Â 12.25Â Â Â Â Â Â 10.58Â Â Â Â Â Â Â 9.29
2.5%Â Â Â Â Â Â Â Â Â 13.86Â Â Â Â Â Â 11.63Â Â Â Â Â Â 10.14
1%Â Â Â Â Â Â Â Â Â 16.19Â Â Â Â Â Â 12.90Â Â Â Â Â Â 11.12
--------------------------------------------------------
b) Dmax tests against an unknown number of breaks
--------------------------------------------------------
UDmax test: 11.431741
Crit. values:Â Â Â 10%: 10.86 5%: 12.59 2.5%: 14.15 1%: 16.19

........................................................
Â Â Â WDmax testÂ Â Â (crit. val.)
10%Â Â Â Â Â 13.33Â Â Â Â Â 11.71
5%Â Â Â Â Â 14.37Â Â Â Â Â 13.66
2.5%Â Â Â Â Â 14.90Â Â Â Â Â 15.33
1%Â Â Â Â Â 15.87Â Â Â Â Â 17.80
********************************************************
supF(l+1|l) tests using global optimizers under the null
--------------------------------------------------------
supF(2|1) Â Â Â 10.93Â Â Â 1981
supF(3|2) Â Â Â 10.70Â Â Â 1993
--------------------------------------------------------
Critical values:Â Â 10%Â Â Â 5%Â 2.5%Â Â Â 1%
Â Â Â Â Â Â supF(2|1)Â 12.19 13.83 15.51 17.58
Â Â Â Â Â Â supF(3|2)Â 13.20 14.73 16.55 18.31

# eval bII.BreakDates
? SB_Estimate(&bII,"FIX",2)

===================================================================
Â Â OUTPUT FROM ESTIMATION OF THE MODEL WITH 2 BREAKS (fixed)
===================================================================
? SB_PrintEstimates(&bII)

Dependent variable: Nigeria
----------------------------------------------------------------
Subsample 1: 1972 - 1981

Â Â Â Â Â Â Â Â Â Â Â Â Â coefficientÂ Â std. errorÂ Â Â Â zÂ Â Â Â p-value
Â -------------------------------------------------------
Â constÂ Â Â Â Â Â Â 15.0822Â Â Â Â Â Â 7.86913Â Â Â Â 1.917Â Â 0.0553Â Â *
Â Nigeria_1Â Â Â Â 0.729122Â Â Â Â 0.159461Â Â Â 4.572Â Â 4.82e-06 ***

Subsample 2: 1982 - 2001

Â Â Â Â Â Â Â Â Â Â Â Â Â coefficientÂ Â std. errorÂ Â Â Â zÂ Â Â Â p-value
Â -------------------------------------------------------
Â constÂ Â Â Â Â Â Â 53.0242Â Â Â Â Â 12.7994Â Â Â Â Â 4.143Â Â 3.43e-05 ***
Â Nigeria_1Â Â Â Â 0.379843Â Â Â Â 0.151844Â Â Â 2.502Â Â 0.0124Â Â **

Subsample 3: 2002 - 2014

Â Â Â Â Â Â Â Â Â Â Â Â Â coefficientÂ Â std. errorÂ Â Â Â zÂ Â Â Â p-value
Â ------------------------------------------------------
Â constÂ Â Â Â Â Â Â 61.0134Â Â Â Â Â 18.9376Â Â Â Â Â 3.222Â Â 0.0013Â ***
Â Nigeria_1Â Â Â Â 0.552676Â Â Â Â 0.149888Â Â Â 3.687Â Â 0.0002Â ***

--------------------------------------------------------------
Confidence intervals for the break dates
--------------------------------------------------------------
The 95% C.I. for the 1th break is: 1976 - 1986
The 90% C.I. for the 1th break is: 1977 - 1985
The 95% C.I. for the 2th break is: 1999 - 2003
The 90% C.I. for the 2th break is: 1999 - 2003
********************************************************

Yusuf Abdulwahab Hassan.
Department of Economics and Development Studies.
Federal University of Kashere,Gombe.
+234 8036830166.
yabdulwa...@fukashere.edu.ng

[Gretl-users] interpreting the “StrucBreak” package for gretl

Reply via email to