Sven Schreiber | 25 Oct 13:55 2014
Picon
Picon

smpl --dummy syntax

Hi,

just a small clarifying question:

smpl some_zero_one_series --dummy

and

smpl some_zero_one_series --restrict

are equivalent, right? For me they seem to give the same result, and
then I asked myself why the '--dummy' option exists at all. Am I missing
something?

thanks,
sven
Artur Bala | 23 Oct 23:23 2014
Picon

replicated matrix in the icon view window

Dear Allin,
I'm testing a bundle function which contains a scalar and a matrix. 
I saved the matrix normally. Then edited it and made some changes.
Eventually, I turn back to the bundle and wanted to save the original matrix with the same name (to replace the old one I modified). 
gretl prompts the warning "A matrix named...already exist. OK to overwrite?" and I choose "Yes". 
The result is that the old matrix is replicated instead of being replaced: same name, same content. If I repeat the operation several time the old matrix is replicated each time.
Besides, when selected "No" to the warning message and cancelled the saving process the old matrix is still replicated.
Best,
Artur 
<div><div dir="ltr">Dear Allin,<div>I'm testing a bundle function which contains a scalar and a matrix.&nbsp;</div>
<div>I saved the matrix normally. Then edited it and made some changes.</div>
<div>Eventually, I turn back to the bundle and wanted to save the original matrix with the same name (to replace the old one I modified).&nbsp;</div>
<div>gretl prompts the warning "A matrix named...already exist. OK to overwrite?" and I choose "Yes".&nbsp;</div>
<div>The result is that the old matrix is replicated instead of being replaced: same name, same content. If I repeat the operation several time the old matrix is replicated each time.</div>
<div>Besides, when selected "No" to the warning message and cancelled the saving process the old matrix is still replicated.</div>
<div>Best,</div>
<div>Artur&nbsp;<br>
</div>
</div></div>
Artur Bala | 22 Oct 20:29 2014
Picon

Re: bivariate probit (in a loop)


There's no doubt your likelihood function is misbehaved here (rho near 1 is worrying) and what you're seeing are numerical problems.


My question: When it comes to rho "near 1" what about the extreme case of rho=1? I tried another biprobit estimation which shows up rho=1...with no error message. At that point, should the likelihood computations have stopped?
(just in case, here's attached the rho-equal-one output)
Artur
gretl version 1.10.0cvs
Current session: 2014-10-22 20:13

# ------------------ estimation modèle 7
? list x = TAILLE INV_ANT CF_ANT VOLAT D_DIV SCOREZ D_V DRD MTB RFR TANGI \
  AGE_BASE REND TAXR DFC_IND2 PROFI SIC2DIGIT
Generated list x
? list x1 = const POT1 x
Generated list x1
? list x2 = const POT2 x
Generated list x2
? smpl (ok(x1) && ok(x2))  --restrict
Full data set: 2590 observations
Current sample: 296 observations
? biprobit external equity x1 ; x2 --save-xbeta --robust --cluster=i_firm
Successive criterion values within tolerance (1e-006)

Model 1: Bivariate probit, using observations 1-296
Standard errors clustered by 82 values of i_firm

              coefficient   std. error      z       p-value 
  ----------------------------------------------------------
 external:
  const         0.110160    1.48425       0.07422   0.9408  
  POT1          7.13424     1.18635       6.014     1.81e-09 ***
  TAILLE        0.0361287   0.0680595     0.5308    0.5955  
  INV_ANT      −0.265339    0.516228     −0.5140    0.6073  
  CF_ANT        2.09666     1.30696       1.604     0.1087  
  VOLAT         0.781673    5.24522       0.1490    0.8815  
  D_DIV        −0.297728    0.254035     −1.172     0.2412  
  SCOREZ       −0.271968    0.331663     −0.8200    0.4122  
  D_V          −0.236385    1.11381      −0.2122    0.8319  
  DRD          −2.41504     3.57942      −0.6747    0.4999  
  MTB           0.566456    0.112937      5.016     5.28e-07 ***
  RFR          −0.0256260   0.196319     −0.1305    0.8961  
  TANGI        −0.0753952   1.00524      −0.07500   0.9402  
  AGE_BASE     −0.0356837   0.0216510    −1.648     0.0993   *
  REND         −0.246472    0.222779     −1.106     0.2686  
  TAXR         −0.120172    0.521025     −0.2306    0.8176  
  DFC_IND2     −1.57189     3.78622      −0.4152    0.6780  
  PROFI        −2.72475     2.78955      −0.9768    0.3287  
  SIC2DIGIT     0.0622110   0.0667275     0.9323    0.3512  

 equity:
  const         6.97190     2.69931       2.583     0.0098   ***
  POT2          4.30132     0.962711      4.468     7.90e-06 ***
  TAILLE       −0.467652    0.179287     −2.608     0.0091   ***
  INV_ANT       2.95192     1.09282       2.701     0.0069   ***
  CF_ANT       −1.17203     2.01875      −0.5806    0.5615  
  VOLAT        22.1561      7.75291       2.858     0.0043   ***
  D_DIV        −1.24048     0.431633     −2.874     0.0041   ***
  SCOREZ       −1.53191     0.651642     −2.351     0.0187   **
  D_V           0.960271    1.29438       0.7419    0.4582  
  DRD          −4.39569     7.20260      −0.6103    0.5417  
  MTB           0.378008    0.147044      2.571     0.0101   **
  RFR          −0.420355    0.323356     −1.300     0.1936  
  TANGI        −7.06169     2.42802      −2.908     0.0036   ***
  AGE_BASE      0.0755413   0.0532725     1.418     0.1562  
  REND          0.0585621   0.257865      0.2271    0.8203  
  TAXR         −0.914275    1.10068      −0.8306    0.4062  
  DFC_IND2     −3.22315     6.17726      −0.5218    0.6018  
  PROFI       −21.6354      6.27013      −3.451     0.0006   ***
  SIC2DIGIT    −0.125967    0.0890904    −1.414     0.1574  

Log-likelihood      −128.5415   Akaike criterion     335.0830
Schwarz criterion    479.0070   Hannan-Quinn         392.7073

rho = 1

Test of independence -
  Null hypothesis: rho = 0
  Test statistic: Chi-square(1) = 16.4347
  with p-value = 5.03547e-005

# genr matrix predict_prob = $yhat
? genr series predict_external = $yhat[,1]
Replaced series predict_external (ID 63)
? genr series predict_equity = $yhat[,2]
Replaced series predict_equity (ID 64)
? genr series prob_external = cdf(N,predict_external)
Replaced series prob_external (ID 65)
? genr series prob_equity = cdf(N, predict_equity)
Replaced series prob_equity (ID 66)
? genr correct_ext1 = sum((prob_external>=0.5) && external=1)/sum(external=1)
Generated scalar correct_ext1 = 0.512195
? genr correct_ext0 = sum((prob_external<0.5) && external=0)/sum(external=0)
Generated scalar correct_ext0 = 0.96729
? genr correct_ext = sum((prob_external>=0.5)==external)/$nobs
Generated scalar correct_ext = 0.841216
? smpl external=1 --restrict
Full data set: 2590 observations
Current sample: 82 observations
? genr correct_eqt1 = sum((prob_equity>=0.5) && equity=1)/sum(equity=1)
Generated scalar correct_eqt1 = 0.454545
? genr correct_eqt0 = sum((prob_equity<0.5) && equity=0)/sum(equity=0)
Generated scalar correct_eqt0 = 0.985915
? genr correct_eqt = sum((prob_equity>=0.5)==equity)/$nobs
Generated scalar correct_eqt = 0.914634
? smpl full
Full data range: 1 - 2590 (n = 2590)

# Impression des résultats de l'estimation
? print correct_ext1 correct_ext0	correct_ext correct_eqt1 correct_eqt0 \
  correct_eqt

   correct_ext1 =  0.51219512

   correct_ext0 =  0.96728972

    correct_ext =  0.84121622

   correct_eqt1 =  0.45454545

   correct_eqt0 =  0.98591549

    correct_eqt =  0.91463415



----------------------------------- RUN 2 - NO CHANGES OCCURRED 

? list x = TAILLE INV_ANT CF_ANT VOLAT D_DIV SCOREZ D_V DRD MTB RFR TANGI \
  AGE_BASE REND TAXR DFC_IND2 PROFI SIC2DIGIT
Replaced list x
? list x1 = const POT1 x
Replaced list x1
? list x2 = const POT2 x
Replaced list x2
? smpl (ok(x1) && ok(x2))  --restrict
Full data set: 2590 observations
Current sample: 296 observations
? biprobit external equity x1 ; x2 --save-xbeta --robust --cluster=i_firm
Successive criterion values within tolerance (1e-006)

Model 2: Bivariate probit, using observations 1-296
Standard errors clustered by 82 values of i_firm

              coefficient   std. error      z       p-value 
  ----------------------------------------------------------
 external:
  const         0.110160    1.48425       0.07422   0.9408  
  POT1          7.13424     1.18635       6.014     1.81e-09 ***
  TAILLE        0.0361287   0.0680595     0.5308    0.5955  
  INV_ANT      −0.265339    0.516228     −0.5140    0.6073  
  CF_ANT        2.09666     1.30696       1.604     0.1087  
  VOLAT         0.781673    5.24522       0.1490    0.8815  
  D_DIV        −0.297728    0.254035     −1.172     0.2412  
  SCOREZ       −0.271968    0.331663     −0.8200    0.4122  
  D_V          −0.236385    1.11381      −0.2122    0.8319  
  DRD          −2.41504     3.57942      −0.6747    0.4999  
  MTB           0.566456    0.112937      5.016     5.28e-07 ***
  RFR          −0.0256260   0.196319     −0.1305    0.8961  
  TANGI        −0.0753952   1.00524      −0.07500   0.9402  
  AGE_BASE     −0.0356837   0.0216510    −1.648     0.0993   *
  REND         −0.246472    0.222779     −1.106     0.2686  
  TAXR         −0.120172    0.521025     −0.2306    0.8176  
  DFC_IND2     −1.57189     3.78622      −0.4152    0.6780  
  PROFI        −2.72475     2.78955      −0.9768    0.3287  
  SIC2DIGIT     0.0622110   0.0667275     0.9323    0.3512  

 equity:
  const         6.97190     2.69931       2.583     0.0098   ***
  POT2          4.30132     0.962711      4.468     7.90e-06 ***
  TAILLE       −0.467652    0.179287     −2.608     0.0091   ***
  INV_ANT       2.95192     1.09282       2.701     0.0069   ***
  CF_ANT       −1.17203     2.01875      −0.5806    0.5615  
  VOLAT        22.1561      7.75291       2.858     0.0043   ***
  D_DIV        −1.24048     0.431633     −2.874     0.0041   ***
  SCOREZ       −1.53191     0.651642     −2.351     0.0187   **
  D_V           0.960271    1.29438       0.7419    0.4582  
  DRD          −4.39569     7.20260      −0.6103    0.5417  
  MTB           0.378008    0.147044      2.571     0.0101   **
  RFR          −0.420355    0.323356     −1.300     0.1936  
  TANGI        −7.06169     2.42802      −2.908     0.0036   ***
  AGE_BASE      0.0755413   0.0532725     1.418     0.1562  
  REND          0.0585621   0.257865      0.2271    0.8203  
  TAXR         −0.914275    1.10068      −0.8306    0.4062  
  DFC_IND2     −3.22315     6.17726      −0.5218    0.6018  
  PROFI       −21.6354      6.27013      −3.451     0.0006   ***
  SIC2DIGIT    −0.125967    0.0890904    −1.414     0.1574  

Log-likelihood      −128.5415   Akaike criterion     335.0830
Schwarz criterion    479.0070   Hannan-Quinn         392.7073

rho = 1

Test of independence -
  Null hypothesis: rho = 0
  Test statistic: Chi-square(1) = 16.4347
  with p-value = 5.03547e-005

# genr matrix predict_prob = $yhat
? genr series predict_external = $yhat[,1]
Replaced series predict_external (ID 63)
? genr series predict_equity = $yhat[,2]
Replaced series predict_equity (ID 64)
? genr series prob_external = cdf(N,predict_external)
Replaced series prob_external (ID 65)
? genr series prob_equity = cdf(N, predict_equity)
Replaced series prob_equity (ID 66)
? genr correct_ext1 = sum((prob_external>=0.5) && external=1)/sum(external=1)
Replaced scalar correct_ext1 = 0.512195
? genr correct_ext0 = sum((prob_external<0.5) && external=0)/sum(external=0)
Replaced scalar correct_ext0 = 0.96729
? genr correct_ext = sum((prob_external>=0.5)==external)/$nobs
Replaced scalar correct_ext = 0.841216
? smpl external=1 --restrict
Full data set: 2590 observations
Current sample: 82 observations
? genr correct_eqt1 = sum((prob_equity>=0.5) && equity=1)/sum(equity=1)
Replaced scalar correct_eqt1 = 0.454545
? genr correct_eqt0 = sum((prob_equity<0.5) && equity=0)/sum(equity=0)
Replaced scalar correct_eqt0 = 0.985915
? genr correct_eqt = sum((prob_equity>=0.5)==equity)/$nobs
Replaced scalar correct_eqt = 0.914634
? smpl full
Full data range: 1 - 2590 (n = 2590)

# Impression des résultats de l'estimation
? print correct_ext1 correct_ext0	correct_ext correct_eqt1 correct_eqt0 \
  correct_eqt

   correct_ext1 =  0.51219512

   correct_ext0 =  0.96728972

    correct_ext =  0.84121622

   correct_eqt1 =  0.45454545

   correct_eqt0 =  0.98591549

    correct_eqt =  0.91463415






----------------------------------- RUN 3 - Verbose
 
? list x = TAILLE INV_ANT CF_ANT VOLAT D_DIV SCOREZ D_V DRD MTB RFR TANGI \
  AGE_BASE REND TAXR DFC_IND2 PROFI SIC2DIGIT
Replaced list x
? list x1 = const POT1 x
Replaced list x1
? list x2 = const POT2 x
Replaced list x2
? smpl (ok(x1) && ok(x2))  --restrict
Full data set: 2590 observations
Current sample: 296 observations
? biprobit external equity x1 ; x2 --verbose --save-xbeta --robust \
  --cluster=i_firm

Iteration 1: loglikelihood = -126.732319570 (steplength = 1)
Parameters:      0.30456      3.7318  -0.0056409    -0.11407     0.73130      1.4001
                -0.19232    -0.22081    -0.41888    -0.60670     0.33540  0.00056145
                -0.15274   -0.013703    -0.11472    -0.21560    -0.68040    -0.60375
                0.029995
Gradients:       -225.38      35.355     -2658.3     -46.967     -48.452     -4.7018
                 -183.99     -336.77     -53.507     -5.9215     -377.49     -353.63
                 -55.879     -4492.0     -26.460     -82.593     -49.146     -18.681
                 -1061.6 (norm 6.04e+000)

Iteration 2: loglikelihood = -116.854957675 (steplength = 1)
Parameters:      0.34163      6.0807    0.010068    -0.23558      1.5343      2.9072
                -0.24238    -0.30186    -0.40942     -1.7480     0.50877   -0.015752
                -0.19070   -0.026719    -0.22146    -0.26727     -1.2620     -1.9328
                0.056247
Gradients:       -23.968      6.8701     -278.75     -3.8765     -4.7662    -0.48700
                 -19.490     -39.674     -5.9908    -0.85857     -31.911     -43.458
                 -5.1350     -471.05     -1.9053     -8.2734     -5.2663     -1.9668
                 -103.01 (norm 2.64e+000)

Iteration 3: loglikelihood = -115.959422738 (steplength = 1)
Parameters:      0.30892      7.1302    0.019579    -0.28869      1.8496      3.4584
                -0.25586    -0.31740    -0.32406     -2.4688     0.58327   -0.019622
               -0.077858   -0.032896    -0.27326    -0.22249     -1.5644     -2.8691
                0.068263
Gradients:       -3.5930      1.5769     -39.998    -0.51541    -0.77993   -0.091275
                 -2.8437     -6.8115     -1.0150    -0.19892     -5.5161     -7.6243
                -0.51352     -66.552    -0.39267     -1.0756    -0.78307    -0.37986
                 -14.884 (norm 1.21e+000)

Iteration 4: loglikelihood = -115.950074936 (steplength = 1)
Parameters:      0.30123      7.2532    0.020483    -0.29460      1.8852      3.4882
                -0.25722    -0.31876    -0.31499     -2.5441     0.59213   -0.019151
               -0.061960   -0.033572    -0.27929    -0.21387     -1.5878     -2.9856
                0.069803
Gradients:      -0.27425     0.14446     -3.0599   -0.032266   -0.059699  -0.0081430
                -0.21854    -0.54795   -0.084016   -0.017125    -0.42295    -0.60574
               -0.026217     -4.9924   -0.037708   -0.067156   -0.059201   -0.032005
                 -1.0639 (norm 3.53e-001)

Iteration 5: loglikelihood = -115.950073895 (steplength = 1)
Parameters:      0.30109      7.2546    0.020492    -0.29466      1.8856      3.4883
                -0.25723    -0.31877    -0.31491     -2.5448     0.59222   -0.019134
               -0.061801   -0.033579    -0.27936    -0.21376     -1.5879     -2.9868
                0.069822
Gradients:    -0.0027497   0.0014823   -0.031387 -0.00029759 -0.00057514-8.4750e-005
              -0.0022154  -0.0054528 -0.00083026 -0.00015405  -0.0038081  -0.0057266
             -0.00027270   -0.051852 -0.00042101 -0.00060152 -0.00058864 -0.00030506
               -0.010043 (norm 3.52e-002)

Iteration 6: loglikelihood = -115.950073895 (steplength = 0.5)
Parameters:      0.30109      7.2546    0.020492    -0.29466      1.8856      3.4883
                -0.25723    -0.31877    -0.31491     -2.5448     0.59222   -0.019134
               -0.061801   -0.033579    -0.27936    -0.21376     -1.5879     -2.9868
                0.069822
Gradients:  -3.0884e-007 1.6613e-007-3.5403e-006-3.2291e-008-6.3017e-008-1.0351e-008
            -2.4655e-007-6.1279e-007-8.9608e-008-1.4960e-008-4.0093e-007-6.0793e-007
            -3.3477e-008-5.8968e-006-4.8847e-008-6.2128e-008-6.5420e-008-3.2586e-008
            -1.0717e-006 (norm 3.71e-004)


--- FINAL VALUES: 
loglikelihood = -115.950073895 (steplength = 0.5)
Parameters:      0.30109      7.2546    0.020492    -0.29466      1.8856      3.4883
                -0.25723    -0.31877    -0.31491     -2.5448     0.59222   -0.019134
               -0.061801   -0.033579    -0.27936    -0.21376     -1.5879     -2.9868
                0.069822
Gradients:  -1.5442e-007 8.3064e-008-1.7701e-006-1.6146e-008-3.1509e-008-5.1755e-009
            -1.2328e-007-3.0639e-007-4.4804e-008-7.4802e-009-2.0046e-007-3.0396e-007
            -1.6738e-008-2.9484e-006-2.4423e-008-3.1064e-008-3.2710e-008-1.6293e-008
            -5.3587e-007 (norm 2.62e-004)

Successive criterion values within tolerance (1e-008)

Probit, using observations 1-296
Dependent variable: external
Standard errors based on Hessian

              coefficient   std. error      z        p-value 
  -----------------------------------------------------------
  const        0.301095     1.52224       0.1978    0.8432   
  POT1         7.25456      1.06883       6.787     1.14e-011 ***
  TAILLE       0.0204915    0.0762379     0.2688    0.7881   
  INV_ANT     −0.294660     0.540840     −0.5448    0.5859   
  CF_ANT       1.88555      1.24252       1.518     0.1291   
  VOLAT        3.48830      5.31192       0.6567    0.5114   
  D_DIV       −0.257232     0.298702     −0.8612    0.3891   
  SCOREZ      −0.318773     0.290522     −1.097     0.2725   
  D_V         −0.314914     0.928867     −0.3390    0.7346   
  DRD         −2.54478      2.92467      −0.8701    0.3842   
  MTB          0.592223     0.116229      5.095     3.48e-07  ***
  RFR         −0.0191340    0.186414     −0.1026    0.9182   
  TANGI       −0.0618008    0.882095     −0.07006   0.9441   
  AGE_BASE    −0.0335791    0.0210625    −1.594     0.1109   
  REND        −0.279357     0.260663     −1.072     0.2838   
  TAXR        −0.213764     0.558307     −0.3829    0.7018   
  DFC_IND2    −1.58790      3.54576      −0.4478    0.6543   
  PROFI       −2.98678      2.82992      −1.055     0.2912   
  SIC2DIGIT    0.0698219    0.0741635     0.9415    0.3465   

Mean dependent var   0.277027   S.D. dependent var   0.448288
McFadden R-squared   0.336201   Adjusted R-squared   0.227429
Log-likelihood      −115.9501   Akaike criterion     269.9001
Schwarz criterion    340.0170   Hannan-Quinn         297.9735

Test for normality of residual -
  Null hypothesis: error is normally distributed
  Test statistic: Chi-square(2) = 13.6994
  with p-value = 0.00105975

Iteration 1: loglikelihood = -52.5842572593 (steplength = 1)
Parameters:     -0.33466     0.67198   -0.037006     0.28296    -0.13922      3.3417
                -0.22620   0.0067796     0.33270      1.1143    0.068529    0.020298
                -0.52366   0.0032300    0.080058   -0.019956     -1.7180     -2.7484
               0.0036713
Gradients:       -260.48     -36.570     -3038.2     -57.849     -57.115     -6.1820
                 -199.97     -380.57     -62.295     -6.8759     -497.01     -413.79
                 -64.421     -5102.1     -35.551     -95.161     -56.419     -21.829
                 -1266.9 (norm 5.64e+000)

Iteration 2: loglikelihood = -30.1033151997 (steplength = 1)
Parameters:     -0.17922      1.1701   -0.072678     0.58465    -0.29499      6.7094
                -0.37795   -0.031348     0.60270      1.8059     0.14008    0.030686
                -0.98308   0.0071969     0.11724   -0.072590     -2.9877     -5.3662
              0.00020404
Gradients:       -55.084     -7.0777     -657.34     -10.432     -12.417    -0.96932
                 -47.445     -80.946     -12.842     -1.2957     -88.789     -83.364
                 -13.896     -1114.2     -6.1851     -20.314     -12.043     -4.7960
                 -259.32 (norm 3.35e+000)

Iteration 3: loglikelihood = -24.2667468363 (steplength = 1)
Parameters:      0.66757      1.9202    -0.12713     0.99906    -0.22534      11.187
                -0.50592    -0.15831     0.84673      2.7027     0.18765   -0.016966
                 -2.0562    0.013852     0.11876    -0.14206     -4.1577     -8.3711
               -0.018341
Gradients:       -15.627     -1.3144     -187.70     -2.9809     -3.6887    -0.26479
                 -14.026     -23.619     -3.5582    -0.38110     -26.835     -24.150
                 -4.1988     -315.78     -1.5874     -5.7265     -3.4313     -1.4402
                 -73.094 (norm 2.36e+000)

Iteration 4: loglikelihood = -22.0276590435 (steplength = 1)
Parameters:       2.1302      2.6462    -0.21004      1.4851    -0.34673      16.065
                -0.58835    -0.39991      1.0499      2.9902     0.23599    -0.11412
                 -3.5780    0.023943    0.075039    -0.23842     -4.7233     -11.663
               -0.049490
Gradients:       -5.2007    -0.27545     -62.433     -1.0920     -1.3456   -0.098099
                 -4.6427     -8.1640     -1.1711    -0.15334     -10.054     -8.3032
                 -1.5164     -103.99    -0.57669     -1.9117     -1.1382    -0.51464
                 -24.954 (norm 1.79e+000)

Iteration 5: loglikelihood = -21.0928432770 (steplength = 1)
Parameters:       3.9237      3.3492    -0.32264      1.9777    -0.75273      20.118
                -0.67820    -0.65919      1.2795      2.4560     0.30665    -0.24759
                 -5.1936    0.038016    0.035465    -0.41013     -4.6112     -15.422
               -0.080900
Gradients:       -2.0334    -0.13717     -24.494    -0.48961    -0.58114   -0.048747
                 -1.7864     -3.1746    -0.47428   -0.077046     -4.2563     -3.3198
                -0.61547     -40.653    -0.22207    -0.77623    -0.43943    -0.21192
                 -9.9977 (norm 1.38e+000)

Iteration 6: loglikelihood = -20.8350870121 (steplength = 1)
Parameters:       5.3429      3.9341    -0.42036      2.3958     -1.2236      22.253
                -0.78213    -0.81177      1.4900      2.0915     0.39045    -0.38835
                 -6.4703    0.051814    0.026962    -0.53752     -4.3793     -18.290
                -0.11030
Gradients:      -0.86424   -0.098145     -10.321    -0.18448    -0.22136   -0.026769
                -0.70758     -1.2790    -0.20363   -0.029711     -1.4285     -1.5006
                -0.24246     -16.805   -0.039403    -0.33230    -0.18690   -0.077626
                 -3.9892 (norm 1.00e+000)

Iteration 7: loglikelihood = -20.8091107568 (steplength = 1)
Parameters:       5.9131      4.1716    -0.46212      2.5624     -1.3641      23.086
                -0.83520    -0.86663      1.5884      1.9569     0.42768    -0.45973
                 -7.0069    0.058617    0.031473    -0.57980     -4.2216     -19.433
                -0.12573
Gradients:      -0.30435   -0.050855     -3.4799   -0.056347   -0.057021   -0.010394
                -0.20889    -0.44343   -0.073814  -0.0086059    -0.41127    -0.58192
               -0.073279     -5.4520   0.0010671    -0.10905   -0.066156   -0.022316
                 -1.3819 (norm 6.07e-001)

Iteration 8: loglikelihood = -20.8087618346 (steplength = 1)
Parameters:       5.9855      4.2013    -0.46749      2.5830     -1.3767      23.203
                -0.84261    -0.87368      1.6007      1.9479     0.43228    -0.46963
                 -7.0749    0.059549    0.032517    -0.58474     -4.1973     -19.576
                -0.12801
Gradients:     -0.039819  -0.0074390    -0.44536  -0.0069189  -0.0062802  -0.0012307
               -0.025340   -0.057498  -0.0099439 -0.00093631   -0.053767   -0.078247
              -0.0083026    -0.68788 2.1219e-005   -0.014077  -0.0086341  -0.0027223
                -0.18848 (norm 2.18e-001)

Iteration 9: loglikelihood = -20.8087617622 (steplength = 1)
Parameters:       5.9866      4.2018    -0.46757      2.5832     -1.3769      23.205
                -0.84272    -0.87378      1.6008      1.9479     0.43235    -0.46977
                 -7.0759    0.059563    0.032533    -0.58481     -4.1970     -19.578
                -0.12805
Gradients:   -0.00060491 -0.00011343  -0.0067224 -0.00010350-9.2711e-005-1.6785e-005
             -0.00037826 -0.00087400 -0.00015303-1.2306e-005 -0.00082900  -0.0011871
             -0.00011664   -0.010342-6.8660e-006 -0.00021166 -0.00013087-4.0205e-005
              -0.0029613 (norm 2.67e-002)

Iteration 10: loglikelihood = -20.8087617622 (steplength = 0.5)
Parameters:       5.9866      4.2018    -0.46757      2.5832     -1.3769      23.205
                -0.84272    -0.87378      1.6008      1.9479     0.43235    -0.46977
                 -7.0759    0.059563    0.032533    -0.58481     -4.1970     -19.578
                -0.12805
Gradients:  -1.2852e-007-2.4103e-008-1.4239e-006-2.2257e-008-1.9628e-008-3.3888e-009
            -8.0191e-008-1.8715e-007-3.2699e-008-2.4207e-009-1.7654e-007-2.5280e-007
            -2.4095e-008-2.1837e-006-2.4492e-009-4.4818e-008-2.7784e-008-8.4394e-009
            -6.4220e-007 (norm 3.89e-004)


--- FINAL VALUES: 
loglikelihood = -20.8087617622 (steplength = 0.5)
Parameters:       5.9866      4.2018    -0.46757      2.5832     -1.3769      23.205
                -0.84272    -0.87378      1.6008      1.9479     0.43235    -0.46977
                 -7.0759    0.059563    0.032533    -0.58481     -4.1970     -19.578
                -0.12805
Gradients:  -6.4262e-008-1.2052e-008-7.1197e-007-1.1129e-008-9.8139e-009-1.6944e-009
            -4.0096e-008-9.3574e-008-1.6349e-008-1.2104e-009-8.8268e-008-1.2640e-007
            -1.2048e-008-1.0918e-006-1.2246e-009-2.2409e-008-1.3892e-008-4.2197e-009
            -3.2110e-007 (norm 2.75e-004)

Successive criterion values within tolerance (1e-008)

Probit, using observations 1-296
Dependent variable: equity
Standard errors based on Hessian

              coefficient   std. error      z       p-value
  ---------------------------------------------------------
  const         5.98658      4.67475      1.281     0.2003 
  POT2          4.20176      1.57041      2.676     0.0075  ***
  TAILLE       −0.467572     0.307759    −1.519     0.1287 
  INV_ANT       2.58325      1.30845      1.974     0.0484  **
  CF_ANT       −1.37685      2.43441     −0.5656    0.5717 
  VOLAT        23.2046      10.2761       2.258     0.0239  **
  D_DIV        −0.842718     0.715602    −1.178     0.2389 
  SCOREZ       −0.873782     0.931908    −0.9376    0.3484 
  D_V           1.60082      2.12154      0.7546    0.4505 
  DRD           1.94793      7.88598      0.2470    0.8049 
  MTB           0.432349     0.269034     1.607     0.1080 
  RFR          −0.469774     0.536119    −0.8763    0.3809 
  TANGI        −7.07592      3.54591     −1.996     0.0460  **
  AGE_BASE      0.0595628    0.0695811    0.8560    0.3920 
  REND          0.0325328    0.607229     0.05358   0.9573 
  TAXR         −0.584806     1.40269     −0.4169    0.6767 
  DFC_IND2     −4.19696      8.67950     −0.4835    0.6287 
  PROFI       −19.5779       8.88441     −2.204     0.0276  **
  SIC2DIGIT    −0.128049     0.207834    −0.6161    0.5378 

Mean dependent var   0.037162   S.D. dependent var   0.189479
McFadden R-squared   0.557356   Adjusted R-squared   0.153188
Log-likelihood      −20.80876   Akaike criterion     79.61752
Schwarz criterion    149.7344   Hannan-Quinn         107.6909

Test for normality of residual -
  Null hypothesis: error is normally distributed
  Test statistic: Chi-square(2) = 61.7921
  with p-value = 3.81953e-014

Iteration 1: loglikelihood = -130.787148830 (steplength = 1)
Parameters:     0.031968      7.1160    0.039171    -0.21914      1.9315      1.8278
                -0.31693    -0.31342    -0.27102     -2.8971     0.58328   -0.019506
               -0.051445   -0.035786    -0.24773    -0.12462     -1.2948     -2.6384
                0.068354      4.9968      3.7298    -0.34002      2.7074    -0.55119
                  19.516     -1.3326     -1.3491      1.5925     -5.1513     0.29853
                -0.38265     -5.7361    0.041642     0.14496    -0.72172     -1.2335
                 -20.106   -0.097534      1.1582
Gradients:  -1.5367e-007 8.2990e-008-1.7666e-006-1.6409e-008-3.1435e-008-5.1518e-009
            -1.2367e-007-3.0499e-007-4.4339e-008-7.3258e-009-2.0242e-007-3.0174e-007
            -1.6471e-008-2.9471e-006-2.5526e-008-3.0580e-008-3.2561e-008-1.6313e-008
            -5.3234e-007-6.4079e-008-1.1971e-008-7.0997e-007-1.1097e-008-9.7771e-009
            -1.6897e-009-4.0013e-008-9.3349e-008-1.6302e-008-1.2097e-009-8.7972e-008
            -1.2607e-007-1.2016e-008-1.0885e-006-1.1795e-009-2.2347e-008-1.3852e-008
            -4.2035e-009-3.2030e-007      10.107 (norm 5.48e-001)

Iteration 2: loglikelihood = -129.335414642 (steplength = 1)
Parameters:     0.099837      7.0956    0.034049    -0.23373      2.0237      1.7496
                -0.28973    -0.29411    -0.27097     -2.7166     0.57080   -0.015073
               -0.088824   -0.034477    -0.24414    -0.14115     -1.4400     -2.7609
                0.062613      6.9296      4.3261    -0.46778      2.9963     -1.0352
                  21.390     -1.2688     -1.5269      1.2692     -3.8577     0.38035
                -0.50095     -7.0942    0.068132    0.039101    -0.95691     -1.8891
                 -22.183    -0.12754      1.7761
Gradients:      0.099127    0.053624      1.2915   -0.063547    0.018422    0.010023
                 0.48207    0.077144    -0.13227   -0.048258    0.071754     0.73625
                 0.13614      3.9993    0.049949   -0.011269    0.039489    0.010466
                 -2.2003      2.2812     0.70051      21.875     0.47967     0.47742
                0.077409      1.5194      3.9359     0.81009     0.15891      3.8272
                  3.2998     0.20330      42.777   -0.065361     0.80311     0.42448
                 0.18201      14.565      2.6229 (norm 1.27e+000)

Iteration 3: loglikelihood = -128.903327145 (steplength = 1)
Parameters:      0.13616      7.1136    0.033793    -0.26184      2.0819      1.2271
                -0.28933    -0.27933    -0.23812     -2.4925     0.56700   -0.021535
               -0.091844   -0.035046    -0.24700    -0.13708     -1.5904     -2.8106
                0.063020      6.9003      4.2393    -0.45166      2.9852     -1.2219
                  21.092     -1.3171     -1.5043      1.0100     -4.3588     0.38059
                -0.45024     -6.9371    0.071281    0.066235    -0.85429     -3.0193
                 -21.354    -0.13661      2.5276
Gradients:       0.45370   -0.024633      4.1177   -0.025239    0.079541   0.0077866
                 0.21036     0.80788     0.22311    0.056710     0.22088     0.44796
              -0.0095258      6.8164    -0.18614     0.18612    0.083636    0.029139
                  3.4429    -0.57360    -0.24609     -6.2892  -0.0089408    -0.12528
             -0.00094623    -0.30671    -0.75457    -0.20794   -0.025518    -0.12152
                -0.58079   -0.031878     -10.667    0.051376    -0.22078    -0.11979
               -0.032502     -3.1939     0.91562 (norm 6.62e-001)

Iteration 4: loglikelihood = -128.683022741 (steplength = 1)
Parameters:      0.11831      7.1286    0.034864    -0.26368      2.0998     0.92689
                -0.29226    -0.28181    -0.25394     -2.5081     0.56945   -0.020506
               -0.091600   -0.035423    -0.24784    -0.12725     -1.5255     -2.7766
                0.062635      6.8917      4.2914    -0.45026      2.9721     -1.2123
                  21.661     -1.2954     -1.4678     0.93555     -3.8431     0.38492
                -0.44680     -7.0336    0.073234    0.061223    -0.84182     -3.4461
                 -21.417    -0.14103      3.2694
Gradients:      -0.60124     0.13181     -3.8519    0.036287    -0.17099  -0.0037274
               0.0037505     -1.6802    -0.44299    -0.15391     0.94057    -0.42560
                0.047069     -5.6429     0.43499    -0.29145   -0.096033   -0.022574
                 -6.0473     0.59713     0.12474      3.5224   -0.035815     0.17177
               0.0089871   -0.011325      1.7463     0.46029     0.16328    -0.77309
                 0.44581   -0.046883      4.7813    -0.49176     0.28085    0.095682
                0.026031      6.2253     0.44124 (norm 6.98e-001)

Iteration 5: loglikelihood = -128.621333519 (steplength = 0.5)
Parameters:      0.11507      7.1277    0.035293    -0.26571      2.1044     0.82657
                -0.29404    -0.27748    -0.24541     -2.4597     0.56785   -0.022326
               -0.087657   -0.035554    -0.24732    -0.12339     -1.5409     -2.7632
                0.062484      6.9352      4.2864    -0.45803      2.9611     -1.2008
                  21.844     -1.2768     -1.4996     0.93727     -4.2793     0.38092
                -0.43315     -7.0144    0.074870    0.062418    -0.88175     -3.4072
                 -21.451    -0.13403      3.8183
Gradients:       0.47589    -0.10597      4.0241    0.057730     0.11892    0.015059
                 0.10523      1.0050     0.24478    0.077913     0.24489     0.57507
                0.026595      6.4902    -0.16782     0.22980    0.086153    0.025351
                  3.7838    -0.44130   -0.098045     -3.8472   -0.047351    -0.10644
               -0.011675    -0.10418    -0.90830    -0.21906   -0.068731   -0.085353
                -0.51022   -0.020581     -6.4111     0.12576    -0.22043   -0.079245
               -0.022635     -3.4496     0.11438 (norm 5.81e-001)

Iteration 6: loglikelihood = -128.573640541 (steplength = 1)
Parameters:      0.11330      7.1327    0.035774    -0.26505      2.0974     0.82303
                -0.29698    -0.27368    -0.23736     -2.4369     0.56690   -0.023839
               -0.080888   -0.035602    -0.24653    -0.12221     -1.5646     -2.7463
                0.062276      6.9551      4.2954    -0.46339      2.9573     -1.1787
                  22.073     -1.2524     -1.5209     0.94481     -4.2174     0.37906
                -0.42841     -7.0479    0.074942    0.059161    -0.90156     -3.2731
                 -21.548    -0.12911      4.6216
Gradients:      -0.38635     0.12178     -2.2717  -0.0079471    -0.12927  -0.0043520
                0.081539     -1.0891    -0.29837    -0.10610     0.22262    -0.31238
                0.012434     -2.5962     0.33045    -0.18935   -0.063744   -0.012831
                 -4.0683     0.40962    0.034778      2.3937    0.013515     0.13686
               0.0065716   -0.079457      1.1545     0.31581     0.11226    -0.12441
                 0.35417  -0.0088784      2.6536    -0.35928     0.19538    0.068416
                0.014674      4.2873    0.095297 (norm 5.53e-001)

Iteration 7: loglikelihood = -128.573168148 (steplength = 1)
Parameters:      0.11106      7.1337    0.036003    -0.26526      2.0972     0.79169
                -0.29737    -0.27222    -0.23607     -2.4169     0.56650   -0.025388
               -0.076285   -0.035664    -0.24659    -0.12057     -1.5708     -2.7313
                0.062257      6.9611      4.2952    -0.46571      2.9490     -1.1705
                  22.101     -1.2450     -1.5274     0.95407     -4.4303     0.37777
                -0.41940     -7.0498    0.075415    0.059131    -0.90942     -3.2773
                 -21.581    -0.12672      5.6809
Gradients:       0.31983    -0.12261      2.5860    0.093956     0.10450    0.017622
                0.077328     0.64927     0.15837    0.049508     0.84309     0.54671
                0.049391      3.8254    -0.15591     0.13576    0.061462    0.017669
                  2.4648    -0.31532   -0.010821     -2.5646   -0.093631    -0.10348
               -0.017098   -0.075960    -0.63511    -0.15441   -0.048091    -0.82255
                -0.53945   -0.048729     -3.8311     0.14786    -0.13519   -0.060498
               -0.017289     -2.4255    0.029588 (norm 5.21e-001)

Iteration 8: loglikelihood = -128.559756680 (steplength = 0.5)
Parameters:      0.10620      7.1392    0.036406    -0.26531      2.0956     0.77730
                -0.29777    -0.27378    -0.24421     -2.4417     0.56698   -0.024546
               -0.077167   -0.035676    -0.24556    -0.11901     -1.5607     -2.7071
                0.062121      6.9724      4.3073    -0.46771      2.9571     -1.1761
                  22.198     -1.2396     -1.5322     0.95658     -4.2918     0.37880
                -0.42292     -7.0763    0.075463    0.058072    -0.91423     -3.1883
                 -21.660    -0.12653      7.2337
Gradients:       -1.5977     0.48173     -10.062   -0.067709    -0.50819   -0.029188
                 0.26829     -4.4020     -1.1836    -0.42192     0.52663     -1.1991
                0.015348     -12.748      1.2331    -0.77011    -0.26512   -0.061259
                 -16.406      1.6062     0.16744      10.116    0.068816     0.51058
                0.029712    -0.26731      4.4242      1.1895     0.42389    -0.50510
                  1.2119   -0.014455      12.796     -1.2417     0.77272     0.26680
                0.061742      16.478   0.0099495 (norm 1.08e+000)

Iteration 9: loglikelihood = -128.549850444 (steplength = 0.00195313)
Parameters:      0.10624      7.1388    0.036401    -0.26529      2.0955     0.77724
                -0.29771    -0.27373    -0.24402     -2.4411     0.56698   -0.024574
               -0.077009   -0.035676    -0.24558    -0.11906     -1.5610     -2.7076
                0.062120      6.9752      4.3032    -0.46815      2.9537     -1.1742
                  22.187     -1.2389     -1.5345     0.95747     -4.3942     0.37812
                -0.42076     -7.0661    0.075624    0.058599    -0.91661     -3.2027
                 -21.645    -0.12588      7.2369
Gradients:        1.7241    -0.66642      16.004     0.64175     0.48658     0.11929
                 0.87549      2.7294     0.59518     0.13176      6.6649      4.0825
                 0.38580      24.531    -0.50644     0.62561     0.36631     0.10865
                  9.8272     -1.6928   -0.059440     -15.787    -0.63722    -0.47703
                -0.11798    -0.87422     -2.6519    -0.57489    -0.12497     -6.6202
                 -4.0417    -0.38344     -24.263     0.48322    -0.61283    -0.36054
                -0.10715     -9.5501   0.0017564 (norm 1.24e+000)

Iteration 10: loglikelihood = -128.545745347 (steplength = 0.000976563)
Parameters:      0.10621      7.1386    0.036402    -0.26529      2.0954     0.77714
                -0.29771    -0.27371    -0.24390     -2.4410     0.56697   -0.024568
               -0.076960   -0.035677    -0.24559    -0.11910     -1.5611     -2.7079
                0.062120      6.9779      4.3014    -0.46848      2.9524     -1.1736
                  22.185     -1.2384     -1.5359     0.95790     -4.4389     0.37783
                -0.41997     -7.0618    0.075710    0.058825    -0.91788     -3.2071
                 -21.640    -0.12557      7.2393
Gradients:        1.7214    -0.66525      15.980     0.64065     0.48572     0.11906
                 0.87378      2.7250     0.59419     0.13150      6.6525      4.0766
                 0.38509      24.492    -0.50519     0.62453     0.36573     0.10848
                  9.8106     -1.6901   -0.059517     -15.764    -0.63613    -0.47618
                -0.11776    -0.87252     -2.6477    -0.57394    -0.12473     -6.6079
                 -4.0359    -0.38274     -24.225     0.48202    -0.61178    -0.35998
                -0.10698     -9.5341   0.0017556 (norm 1.23e+000)

Iteration 11: loglikelihood = -128.543273538 (steplength = 0.5)
Parameters:      0.10798      7.1370    0.036216    -0.26532      2.0963     0.78514
                -0.29740    -0.27300    -0.24042     -2.4280     0.56672   -0.025082
               -0.076168   -0.035674    -0.24610    -0.11964     -1.5648     -2.7172
                0.062217      6.9685      4.2982    -0.46597      2.9504     -1.1749
                  22.116     -1.2446     -1.5272     0.95041     -4.4034     0.37846
                -0.41947     -7.0620    0.075455    0.058748    -0.90673     -3.2823
                 -21.595    -0.12734      8.0129
Gradients:        1.3680    -0.55396      13.424     0.59374     0.37690     0.10871
                 0.87292      1.8730     0.37350    0.057283      6.4066      3.6590
                 0.36754      20.952    -0.29367     0.46123     0.30324    0.093000
                  6.6342     -1.3367   -0.027656     -13.208    -0.58921    -0.36735
                -0.10741    -0.87166     -1.7955    -0.35321   -0.050498     -6.3620
                 -3.6182    -0.36518     -20.684     0.27047    -0.44844    -0.29748
               -0.091495     -6.3572   0.0025729 (norm 1.12e+000)

Iteration 12: loglikelihood = -128.542712919 (steplength = 0.125)
Parameters:      0.10917      7.1355    0.036242    -0.26523      2.0959     0.77986
                -0.29790    -0.27285    -0.24006     -2.4273     0.56666   -0.024994
               -0.076919   -0.035671    -0.24598    -0.11955     -1.5685     -2.7178
                0.062140      6.9794      4.3009    -0.46826      2.9524     -1.1736
                  22.168     -1.2397     -1.5344     0.95855     -4.4135     0.37793
                -0.42097     -7.0605    0.075646    0.058798    -0.91563     -3.2166
                 -21.638    -0.12588      8.1291
Gradients:      -0.45797     0.16298     -3.0161   -0.063352    -0.15066   -0.018018
               -0.011231     -1.1667    -0.30843    -0.10665    -0.49887    -0.49869
               -0.030019     -3.9864     0.36084    -0.20085   -0.082196   -0.021017
                 -4.3467     0.47099    0.027723      3.0888    0.065713     0.15540
                0.018729   0.0099856      1.2036     0.31836     0.11007     0.52265
                 0.51532    0.031246      4.0546    -0.37450     0.20509    0.084671
                0.021714      4.4793   0.0011106 (norm 6.01e-001)

Iteration 13: loglikelihood = -128.542617385 (steplength = 0.125)
Parameters:      0.10968      7.1355    0.036187    -0.26519      2.0957     0.78094
                -0.29761    -0.27280    -0.24006     -2.4245     0.56667   -0.025280
               -0.076479   -0.035668    -0.24615    -0.11940     -1.5687     -2.7181
                0.062147      6.9621      4.2995    -0.46666      2.9504     -1.1707
                  22.135     -1.2418     -1.5295     0.95814     -4.4188     0.37789
                -0.41891     -7.0586    0.075464    0.058854    -0.91268     -3.2412
                 -21.615    -0.12613      8.2563
Gradients:      -0.61847     0.17894     -5.3572   -0.079900    -0.14918   -0.016881
                -0.13361     -1.2194    -0.30089   -0.093044    -0.65501    -0.88574
               -0.039740     -8.0666     0.19894    -0.31017    -0.11063   -0.026409
                 -4.3969     0.62890    0.047577      5.4096    0.081776     0.15320
                0.017492     0.13230      1.2505     0.30936    0.096028     0.67485
                 0.89761    0.040741      8.1101    -0.21131     0.31352     0.11260
                0.026964      4.5095  0.00099238 (norm 6.78e-001)

Iteration 14: loglikelihood = -128.542451799 (steplength = 0.125)
Parameters:      0.10912      7.1351    0.036240    -0.26525      2.0962     0.77973
                -0.29800    -0.27262    -0.23907     -2.4251     0.56658   -0.025016
               -0.076714   -0.035676    -0.24602    -0.11978     -1.5693     -2.7193
                0.062161      6.9812      4.3012    -0.46839      2.9526     -1.1738
                  22.171     -1.2396     -1.5347     0.95864     -4.4065     0.37796
                -0.42120     -7.0613    0.075656    0.058726    -0.91561     -3.2152
                 -21.641    -0.12590      8.3751
Gradients:      -0.42136     0.13112     -3.3312   -0.042671    -0.11423   -0.012949
                -0.10205    -0.89253    -0.23173   -0.080518    -0.43963    -0.41345
               -0.025126     -5.2438     0.21755    -0.23556   -0.071521   -0.015488
                 -3.3717     0.42966    0.011602      3.3677    0.044114     0.11765
                0.013473     0.10089     0.91846     0.23892    0.083129     0.45620
                 0.42113    0.025939      5.2703    -0.22870     0.23830    0.073066
                0.015917      3.4669  0.00088092 (norm 5.61e-001)

Iteration 15: loglikelihood = -128.542352398 (steplength = 0.125)
Parameters:      0.10966      7.1351    0.036185    -0.26521      2.0959     0.78080
                -0.29770    -0.27260    -0.23920     -2.4226     0.56660   -0.025302
               -0.076304   -0.035672    -0.24619    -0.11960     -1.5694     -2.7195
                0.062164      6.9635      4.2998    -0.46679      2.9506     -1.1708
                  22.138     -1.2416     -1.5298     0.95845     -4.4134     0.37791
                -0.41910     -7.0592    0.075472    0.058799    -0.91284     -3.2389
                 -21.618    -0.12611      8.4918
Gradients:      -0.45991     0.13269     -3.9555   -0.059190    -0.11124   -0.012219
               -0.084099    -0.92411    -0.22795   -0.069876    -0.45265    -0.67658
               -0.027885     -5.8366     0.15054    -0.22345   -0.082805   -0.020148
                 -3.3068     0.46627    0.039723      3.9773    0.060234     0.11411
                0.012666    0.083071     0.94542     0.23400    0.072161     0.46629
                 0.68035    0.028528      5.8471    -0.16065     0.22565    0.083960
                0.020461      3.3865  0.00077916 (norm 5.83e-001)

Iteration 16: loglikelihood = -128.542304812 (steplength = 0.125)
Parameters:      0.10910      7.1347    0.036238    -0.26527      2.0964     0.77961
                -0.29808    -0.27245    -0.23832     -2.4233     0.56652   -0.025035
               -0.076562   -0.035679    -0.24606    -0.11996     -1.5699     -2.7205
                0.062176      6.9823      4.3014    -0.46851      2.9528     -1.1739
                  22.173     -1.2394     -1.5349     0.95895     -4.4020     0.37797
                -0.42136     -7.0617    0.075663    0.058679    -0.91578     -3.2130
                 -21.643    -0.12587      8.6071
Gradients:      -0.29548    0.093183     -2.2578   -0.025480   -0.082472  -0.0090357
               -0.069110    -0.64529    -0.17008   -0.060798    -0.28033    -0.24602
               -0.015709     -3.5820     0.17102    -0.16991   -0.049203   -0.010287
                 -2.4613     0.30037   0.0045095      2.2693    0.026228    0.084891
               0.0094180    0.068202     0.66289     0.17518    0.062797     0.29163
                 0.24710    0.016219      3.5816    -0.18013     0.17168    0.050067
                0.010513      2.5284  0.00068911 (norm 4.68e-001)

Iteration 17: loglikelihood = -128.542165497 (steplength = 0.125)
Parameters:      0.10965      7.1348    0.036183    -0.26522      2.0961     0.78069
                -0.29777    -0.27245    -0.23854     -2.4211     0.56655   -0.025319
               -0.076173   -0.035675    -0.24622    -0.11975     -1.5700     -2.7205
                0.062177      6.9644      4.3000    -0.46690      2.9507     -1.1708
                  22.140     -1.2414     -1.5300     0.95875     -4.4096     0.37791
                -0.41923     -7.0595    0.075478    0.058759    -0.91300     -3.2368
                 -21.621    -0.12608      8.7219
Gradients:      -0.35305     0.10039     -3.0481   -0.044579   -0.084133  -0.0088842
               -0.056721    -0.71302    -0.17525   -0.052999    -0.31801    -0.53461
               -0.019918     -4.4357     0.11027    -0.16797   -0.063847   -0.015717
                 -2.5296     0.35632    0.033578      3.0466    0.044986    0.086113
               0.0092053    0.055918     0.72685     0.17944    0.054748     0.32693
                 0.53219    0.020286      4.4211    -0.11864     0.16929    0.064379
                0.015844      2.5842  0.00060959 (norm 5.09e-001)

Iteration 18: loglikelihood = -128.542020137 (steplength = 0.0625)
Parameters:      0.10937      7.1346    0.036210    -0.26525      2.0963     0.78011
                -0.29795    -0.27238    -0.23814     -2.4216     0.56651   -0.025185
               -0.076311   -0.035679    -0.24615    -0.11992     -1.5702     -2.7210
                0.062182      6.9738      4.3008    -0.46775      2.9518     -1.1724
                  22.158     -1.2403     -1.5326     0.95899     -4.4041     0.37794
                -0.42035     -7.0608    0.075573    0.058701    -0.91447     -3.2240
                 -21.633    -0.12596      8.7792
Gradients:      -0.20435    0.065582     -1.4860   -0.012987   -0.059274  -0.0061822
               -0.046089    -0.46462    -0.12494   -0.046337    -0.16567    -0.12489
              -0.0089183     -2.3927     0.13627    -0.12275   -0.033024  -0.0064967
                 -1.7962     0.20664 -0.00075319      1.4783    0.013204    0.060931
               0.0064558    0.045378     0.47584     0.12845    0.047867     0.17293
                 0.12092   0.0091961      2.3719    -0.14382     0.12378    0.033366
               0.0065683      1.8416  0.00053934 (norm 3.87e-001)

Iteration 19: loglikelihood = -128.541970359 (steplength = 0.03125)
Parameters:      0.10950      7.1346    0.036197    -0.26523      2.0962     0.78036
                -0.29788    -0.27239    -0.23820     -2.4210     0.56652   -0.025251
               -0.076221   -0.035678    -0.24619    -0.11987     -1.5702     -2.7210
                0.062182      6.9696      4.3005    -0.46737      2.9513     -1.1717
                  22.150     -1.2408     -1.5314     0.95894     -4.4060     0.37793
                -0.41985     -7.0603    0.075530    0.058721    -0.91382     -3.2296
                 -21.628    -0.12601      8.8079
Gradients:      -0.29070    0.081909     -2.5136   -0.036011   -0.068643  -0.0070368
               -0.042444    -0.58964    -0.14470   -0.043453    -0.24466    -0.44651
               -0.015500     -3.6257    0.088548    -0.13665   -0.052675   -0.013045
                 -2.0811     0.29257    0.029156      2.5033    0.036148    0.070160
               0.0072897    0.041775     0.59973     0.14791    0.044887     0.25121
                 0.44187    0.015740      3.6024   -0.095730     0.13755    0.052934
                0.013092      2.1226  0.00050701 (norm 4.60e-001)

Iteration 20: loglikelihood = -128.541779875 (steplength = 1)
Parameters:      0.11000      7.1342    0.036141    -0.26535      2.0968     0.78149
                -0.29780    -0.27195    -0.23627     -2.4154     0.56644   -0.025561
               -0.075461   -0.035685    -0.24644    -0.12024     -1.5719     -2.7248
                0.062214      6.9756      4.3016    -0.46797      2.9523     -1.1727
                  22.163     -1.2401     -1.5329     0.96023     -4.3932     0.37803
                -0.42080     -7.0622    0.075578    0.058539    -0.91475     -3.2187
                 -21.640    -0.12594      9.7820
Gradients:      -0.18535    0.059345     -1.3468   -0.011498   -0.053687  -0.0055460
               -0.040928    -0.42238    -0.11359   -0.042103    -0.14584    -0.11329
              -0.0078459     -2.1634     0.12359    -0.11118   -0.029946  -0.0058956
                 -1.6312     0.18707 -0.00051947      1.3357    0.011607    0.055143
               0.0057897    0.040280     0.43201     0.11667    0.043492     0.15209
                 0.10848   0.0080705      2.1394    -0.13058     0.11203    0.030177
               0.0059352      1.6711  0.00049404 (norm 3.68e-001)

Iteration 21: loglikelihood = -128.541706611 (steplength = 0.03125)
Parameters:      0.11011      7.1342    0.036131    -0.26534      2.0967     0.78170
                -0.29773    -0.27197    -0.23639     -2.4151     0.56645   -0.025615
               -0.075399   -0.035684    -0.24647    -0.12018     -1.5718     -2.7247
                0.062213      6.9713      4.3013    -0.46759      2.9518     -1.1720
                  22.155     -1.2406     -1.5317     0.96015     -4.3954     0.37801
                -0.42029     -7.0617    0.075534    0.058564    -0.91410     -3.2245
                 -21.634    -0.12599      9.8113
Gradients:       0.74405    -0.29787      8.3385     0.37979     0.18493    0.065213
                 0.68026     0.63103    0.081093   -0.018833      4.2524      2.2215
                 0.24406      14.172    0.037825     0.28453     0.16882    0.050728
                  2.2856    -0.74433    0.011382     -8.3418    -0.37985    -0.18496
               -0.065211    -0.68025    -0.63142   -0.081153    0.018839     -4.2523
                 -2.2224    -0.24407     -14.176   -0.038006    -0.28461    -0.16888
               -0.050746     -2.2867  0.00024158 (norm 8.55e-001)

Iteration 22: loglikelihood = -128.541659105 (steplength = 1)
Parameters:      0.11004      7.1341    0.036142    -0.26535      2.0968     0.78153
                -0.29782    -0.27196    -0.23627     -2.4155     0.56644   -0.025546
               -0.075516   -0.035685    -0.24643    -0.12025     -1.5720     -2.7249
                0.062213      6.9761      4.3016    -0.46805      2.9524     -1.1727
                  22.164     -1.2400     -1.5331     0.96051     -4.3938     0.37801
                -0.42089     -7.0620    0.075584    0.058554    -0.91504     -3.2166
                 -21.641    -0.12589      10.796
Gradients:       0.81869    -0.30872      9.1855     0.39152     0.19204    0.064402
                 0.65837     0.76310     0.10566   -0.018282      4.2082      2.4777
                 0.24352      15.097    0.075598     0.29695     0.18499    0.056010
                  2.6052    -0.81900   -0.018546     -9.1892    -0.39159    -0.19208
               -0.064400    -0.65836    -0.76354    -0.10573    0.018287     -4.2082
                 -2.4787    -0.24353     -15.102   -0.075793    -0.29704    -0.18506
               -0.056030     -2.6064  0.00023440 (norm 8.81e-001)

Iteration 23: loglikelihood = -128.541648112 (steplength = 0.00195313)
Parameters:      0.11000      7.1342    0.036141    -0.26535      2.0968     0.78171
                -0.29781    -0.27196    -0.23626     -2.4155     0.56644   -0.025541
               -0.075506   -0.035685    -0.24643    -0.12025     -1.5719     -2.7249
                0.062215      6.9764      4.3015    -0.46798      2.9524     -1.1730
                  22.162     -1.2402     -1.5329     0.95992     -4.3923     0.37805
                -0.42091     -7.0623    0.075581    0.058541    -0.91454     -3.2200
                 -21.639    -0.12599      10.799
Gradients:      -0.66057     0.18957     -6.3936    -0.17320    -0.17249   -0.024307
                -0.24821     -1.0341    -0.23123   -0.061239     -1.3729     -1.1901
               -0.089213     -10.657    0.055495    -0.34390    -0.12285   -0.029773
                 -4.0017     0.66021    0.041624      6.3892     0.17314     0.17245
                0.024308     0.24821      1.0335     0.23114    0.061242      1.3729
                  1.1889    0.089204      10.651   -0.055700     0.34380     0.12278
                0.029749      4.0003 5.2230e-005 (norm 7.19e-001)

Iteration 24: loglikelihood = -128.541624532 (steplength = 0.015625)
Parameters:      0.11020      7.1341    0.036134    -0.26533      2.0967     0.78146
                -0.29779    -0.27197    -0.23638     -2.4153     0.56645   -0.025584
               -0.075518   -0.035683    -0.24645    -0.12018     -1.5721     -2.7248
                0.062207      6.9736      4.3015    -0.46784      2.9521     -1.1722
                  22.160     -1.2402     -1.5325     0.96053     -4.3953     0.37799
                -0.42059     -7.0617    0.075560    0.058568    -0.91473     -3.2194
                 -21.638    -0.12591      10.819
Gradients:      -0.52319     0.16888     -4.5923    -0.16986    -0.16503   -0.023949
                -0.11163    -0.96734    -0.22299   -0.060240     -1.2090     -1.0429
               -0.080935     -7.0978     0.13501    -0.21844    -0.10267   -0.028113
                 -3.7216     0.52268    0.048188      4.5861     0.16980     0.16498
                0.023951     0.11149     0.96674     0.22289    0.060242      1.2089
                  1.0416    0.080917      7.0884    -0.13529     0.21820     0.10257
                0.028088      3.7200 5.2119e-005 (norm 6.56e-001)

Iteration 25: loglikelihood = -128.541618858 (steplength = 0.00390625)
Parameters:      0.11009      7.1342    0.036133    -0.26534      2.0967     0.78192
                -0.29777    -0.27198    -0.23637     -2.4153     0.56645   -0.025568
               -0.075494   -0.035683    -0.24645    -0.12020     -1.5719     -2.7249
                0.062214      6.9747      4.3012    -0.46768      2.9521     -1.1730
                  22.155     -1.2408     -1.5319     0.95893     -4.3910     0.37811
                -0.42066     -7.0624    0.075555    0.058533    -0.91342     -3.2283
                 -21.634    -0.12616      10.824
Gradients:      -0.65081     0.18642     -6.3063    -0.17001    -0.16946   -0.023855
                -0.24580     -1.0165    -0.22723   -0.060176     -1.3494     -1.1704
               -0.087662     -10.521    0.053229    -0.33974    -0.12094   -0.029252
                 -3.9343     0.65023    0.040802      6.2992     0.16995     0.16942
                0.023856     0.24558      1.0159     0.22712    0.060177      1.3492
                  1.1691    0.087640      10.509   -0.053557     0.33944     0.12083
                0.029226      3.9325 5.1343e-005 (norm 7.13e-001)

Iteration 26: loglikelihood = -128.541578159 (steplength = 0.015625)
Parameters:      0.11029      7.1342    0.036126    -0.26532      2.0966     0.78167
                -0.29774    -0.27199    -0.23649     -2.4151     0.56646   -0.025610
               -0.075507   -0.035682    -0.24646    -0.12012     -1.5721     -2.7247
                0.062206      6.9719      4.3011    -0.46754      2.9518     -1.1723
                  22.153     -1.2408     -1.5315     0.95956     -4.3941     0.37805
                -0.42034     -7.0618    0.075534    0.058560    -0.91363     -3.2277
                 -21.633    -0.12608      10.843
Gradients:      -0.51207     0.16536     -4.4916    -0.16634    -0.16167   -0.023450
                -0.10865    -0.94780    -0.21855   -0.059061     -1.1828     -1.0209
               -0.079205     -6.9387     0.13268    -0.21355    -0.10050   -0.027533
                 -3.6464     0.51107    0.047295      4.4790     0.16626     0.16160
                0.023450     0.10802     0.94697     0.21842    0.059060      1.1821
                  1.0191    0.079159      6.9166    -0.13325     0.21286     0.10033
                0.027502      3.6438 5.1147e-005 (norm 6.49e-001)

Iteration 27: loglikelihood = -128.541555412 (steplength = 0.0078125)
Parameters:      0.11035      7.1341    0.036128    -0.26531      2.0966     0.78137
                -0.29777    -0.27199    -0.23649     -2.4151     0.56645   -0.025601
               -0.075558   -0.035682    -0.24645    -0.12011     -1.5722     -2.7247
                0.062201      6.9726      4.3013    -0.46767      2.9519     -1.1722
                  22.156     -1.2405     -1.5320     0.96000     -4.3948     0.37802
                -0.42043     -7.0617    0.075546    0.058565    -0.91413     -3.2240
                 -21.635    -0.12600      10.853
Gradients:      -0.40391     0.14193     -3.2894    -0.13937    -0.14578   -0.021755
                -0.10676    -0.77810    -0.18745   -0.058137     -1.0691    -0.67279
               -0.070487     -5.4270     0.16988    -0.18815   -0.076996   -0.020089
                 -3.1887     0.40288    0.016670      3.2765     0.13928     0.14571
                0.021754     0.10614     0.77721     0.18732    0.058136      1.0683
                 0.67084    0.070437      5.4045    -0.17046     0.18747    0.076819
                0.020055      3.1859 5.0377e-005 (norm 5.84e-001)

Iteration 28: loglikelihood = -128.541544792 (steplength = 0.00390625)
Parameters:      0.11040      7.1341    0.036126    -0.26531      2.0965     0.78131
                -0.29776    -0.27199    -0.23652     -2.4151     0.56646   -0.025611
               -0.075561   -0.035681    -0.24646    -0.12010     -1.5723     -2.7247
                0.062199      6.9719      4.3012    -0.46763      2.9519     -1.1720
                  22.155     -1.2405     -1.5319     0.96015     -4.3956     0.37801
                -0.42035     -7.0616    0.075540    0.058572    -0.91418     -3.2238
                 -21.635    -0.12598      10.857
Gradients:      -0.50057     0.16153     -4.3921    -0.16246    -0.15790   -0.022886
                -0.10589    -0.92657    -0.21361   -0.057683     -1.1541    -0.99874
               -0.077307     -6.7818     0.12929    -0.20860   -0.098257   -0.026925
                 -3.5631     0.49950    0.046409      4.3787     0.16236     0.15782
                0.022884     0.10527     0.92559     0.21346    0.057681      1.1533
                 0.99663    0.077254      6.7588    -0.12989     0.20791    0.098069
                0.026888      3.5601 4.9990e-005 (norm 6.41e-001)

Iteration 29: loglikelihood = -128.541518218 (steplength = 0.25)
Parameters:      0.11033      7.1341    0.036128    -0.26532      2.0966     0.78131
                -0.29775    -0.27198    -0.23649     -2.4150     0.56646   -0.025618
               -0.075507   -0.035682    -0.24646    -0.12011     -1.5722     -2.7247
                0.062202      6.9718      4.3013    -0.46763      2.9519     -1.1720
                  22.156     -1.2405     -1.5319     0.96015     -4.3956     0.37801
                -0.42034     -7.0617    0.075541    0.058566    -0.91419     -3.2238
                 -21.635    -0.12598      11.196
Gradients:      -0.61455     0.17798     -5.9056    -0.16356    -0.16305   -0.023019
                -0.22508    -0.97410    -0.21860   -0.058230     -1.2877     -1.1099
               -0.083851     -9.8138    0.060240    -0.31667    -0.11460   -0.027968
                 -3.7737     0.61347    0.039292      5.8921     0.16346     0.16297
                0.023018     0.22446     0.97313     0.21844    0.058228      1.2869
                  1.1078    0.083797      9.7906   -0.060842     0.31597     0.11441
                0.027931      3.7707 5.1166e-005 (norm 6.94e-001)

Iteration 30: loglikelihood = -128.541493733 (steplength = 1)
Parameters:      0.11015      7.1342    0.036129    -0.26534      2.0967     0.78171
                -0.29773    -0.27197    -0.23638     -2.4151     0.56646   -0.025625
               -0.075391   -0.035684    -0.24647    -0.12018     -1.5719     -2.7248
                0.062212      6.9719      4.3013    -0.46765      2.9519     -1.1720
                  22.156     -1.2405     -1.5319     0.96029     -4.3957     0.37801
                -0.42036     -7.0617    0.075541    0.058563    -0.91429     -3.2230
                 -21.635    -0.12596      12.058
Gradients:      -0.13899    0.023021    -0.91101   0.0028061   -0.050486   0.0028849
                0.047111    -0.37563    -0.10307   -0.041108     0.34283    0.030185
                0.013309     -1.6335    0.069494    -0.10544   -0.017289  -0.0013872
                 -1.6071     0.13818    0.017239     0.90082  -0.0028802    0.050425
              -0.0028858   -0.047580     0.37490     0.10296    0.041107    -0.34343
               -0.031773   -0.013349      1.6161   -0.069948     0.10492    0.017147
               0.0013592      1.6048 4.9775e-005 (norm 3.29e-001)

Iteration 31: loglikelihood = -128.541484283 (steplength = 1)
Parameters:      0.11015      7.1342    0.036129    -0.26534      2.0967     0.78169
                -0.29773    -0.27197    -0.23638     -2.4150     0.56646   -0.025626
               -0.075391   -0.035684    -0.24647    -0.12018     -1.5719     -2.7248
                0.062211      6.9718      4.3013    -0.46765      2.9519     -1.1720
                  22.156     -1.2405     -1.5319     0.96031     -4.3958     0.37801
                -0.42035     -7.0617    0.075541    0.058563    -0.91430     -3.2230
                 -21.635    -0.12596      12.927
Gradients:       0.23145   -0.065189      2.8173     0.15984    0.084782    0.015704
                 0.18563     0.10926  -0.0044692   -0.013842      1.1791     0.64063
                0.082069      5.7696     0.12375     0.13527    0.047585    0.011922
                 0.87356    -0.23145  -0.0066678     -2.8173    -0.15984   -0.084782
               -0.015704    -0.18563    -0.10926   0.0044705    0.013843     -1.1790
                -0.64063   -0.082068     -5.7697    -0.12376    -0.13527   -0.047585
               -0.011922    -0.87356 2.1246e-005 (norm 4.74e-001)

Iteration 32: loglikelihood = -128.541483126 (steplength = 0.5)
Parameters:      0.11014      7.1342    0.036129    -0.26534      2.0967     0.78171
                -0.29773    -0.27197    -0.23638     -2.4151     0.56646   -0.025624
               -0.075389   -0.035684    -0.24647    -0.12018     -1.5719     -2.7248
                0.062212      6.9720      4.3013    -0.46765      2.9519     -1.1721
                  22.156     -1.2405     -1.5319     0.96021     -4.3955     0.37801
                -0.42036     -7.0617    0.075541    0.058561    -0.91423     -3.2234
                 -21.635    -0.12598      13.470
Gradients:      -0.42166    0.075901     -3.8065    -0.11323    -0.15895  -0.0030781
               -0.047305    -0.72860    -0.17487   -0.063981   -0.097728    -0.33694
               -0.031927     -7.7765  -0.0052671    -0.32008   -0.062471  -0.0085006
                 -3.6105     0.42166    0.036679      3.8065     0.11323     0.15895
               0.0030782    0.047303     0.72860     0.17487    0.063981    0.097732
                 0.33694    0.031927      7.7764   0.0052638     0.32008    0.062471
               0.0085006      3.6105 8.2885e-006 (norm 5.42e-001)

Iteration 33: loglikelihood = -128.541480599 (steplength = 0.00195313)
Parameters:      0.11016      7.1342    0.036129    -0.26534      2.0967     0.78166
                -0.29773    -0.27197    -0.23639     -2.4150     0.56646   -0.025627
               -0.075395   -0.035684    -0.24647    -0.12017     -1.5719     -2.7247
                0.062211      6.9719      4.3013    -0.46765      2.9519     -1.1720
                  22.156     -1.2405     -1.5319     0.96030     -4.3958     0.37801
                -0.42035     -7.0617    0.075541    0.058563    -0.91430     -3.2230
                 -21.635    -0.12596      13.472
Gradients:      0.044610   -0.010230      1.0296    0.068748    0.012640   0.0062505
                 0.12309    -0.16226   -0.060248   -0.027534     0.61310     0.25814
                0.040030      2.4513     0.12425    0.038337    0.011618   0.0028457
                -0.41105   -0.044640   0.0074586     -1.0299   -0.068749   -0.012641
              -0.0062505    -0.12312     0.16224    0.060246    0.027534    -0.61313
                -0.25817   -0.040032     -2.4521    -0.12427   -0.038365   -0.011622
              -0.0028460     0.41099 4.9637e-006 (norm 3.05e-001)

Iteration 34: loglikelihood = -128.541480419 (steplength = 0.015625)
Parameters:      0.11016      7.1342    0.036129    -0.26534      2.0967     0.78167
                -0.29773    -0.27197    -0.23638     -2.4150     0.56646   -0.025626
               -0.075395   -0.035684    -0.24647    -0.12017     -1.5719     -2.7248
                0.062211      6.9719      4.3013    -0.46765      2.9519     -1.1720
                  22.156     -1.2405     -1.5319     0.96027     -4.3957     0.37801
                -0.42035     -7.0617    0.075541    0.058562    -0.91427     -3.2231
                 -21.635    -0.12597      13.489
Gradients:     -0.091598    0.010110    -0.76163    0.065615   0.0054915   0.0059289
               -0.013273    -0.22664   -0.067912   -0.028360     0.45075     0.11282
                0.031848     -1.0927    0.044367   -0.086561  -0.0083541   0.0012395
                -0.68250    0.091566  0.00080138     0.76121   -0.065616  -0.0054933
              -0.0059289    0.013241     0.22662    0.067911    0.028360    -0.45078
                -0.11285   -0.031850      1.0918   -0.044387    0.086532   0.0083495
              -0.0012398     0.68243 4.9543e-006 (norm 2.91e-001)


--- FINAL VALUES: 
loglikelihood = -128.541480419 (steplength = 0.015625)
Parameters:      0.11016      7.1342    0.036129    -0.26534      2.0967     0.78167
                -0.29773    -0.27197    -0.23638     -2.4150     0.56646   -0.025626
               -0.075395   -0.035684    -0.24647    -0.12017     -1.5719     -2.7248
                0.062211      6.9719      4.3013    -0.46765      2.9519     -1.1720
                  22.156     -1.2405     -1.5319     0.96027     -4.3957     0.37801
                -0.42035     -7.0617    0.075541    0.058562    -0.91427     -3.2231
                 -21.635    -0.12597      13.489
Gradients:      0.044272   -0.010123      1.0175    0.067561    0.012435   0.0061443
                 0.12141    -0.15925   -0.059183   -0.027057     0.60307     0.25417
                0.039363      2.4200     0.12235    0.038063    0.011481   0.0028024
                -0.40322   -0.044312   0.0073540     -1.0180   -0.067562   -0.012437
              -0.0061443    -0.12145     0.15923    0.059181    0.027057    -0.60311
                -0.25421   -0.039366     -2.4211    -0.12237   -0.038100   -0.011487
              -0.0028028     0.40314 4.8610e-006 (norm 3.02e-001)

Successive criterion values within tolerance (1e-006)

Model 3: Bivariate probit, using observations 1-296
Standard errors clustered by 82 values of i_firm

              coefficient   std. error      z       p-value 
  ----------------------------------------------------------
 external:
  const         0.110160    1.48425       0.07422   0.9408  
  POT1          7.13424     1.18635       6.014     1.81e-09 ***
  TAILLE        0.0361287   0.0680595     0.5308    0.5955  
  INV_ANT      −0.265339    0.516228     −0.5140    0.6073  
  CF_ANT        2.09666     1.30696       1.604     0.1087  
  VOLAT         0.781673    5.24522       0.1490    0.8815  
  D_DIV        −0.297728    0.254035     −1.172     0.2412  
  SCOREZ       −0.271968    0.331663     −0.8200    0.4122  
  D_V          −0.236385    1.11381      −0.2122    0.8319  
  DRD          −2.41504     3.57942      −0.6747    0.4999  
  MTB           0.566456    0.112937      5.016     5.28e-07 ***
  RFR          −0.0256260   0.196319     −0.1305    0.8961  
  TANGI        −0.0753952   1.00524      −0.07500   0.9402  
  AGE_BASE     −0.0356837   0.0216510    −1.648     0.0993   *
  REND         −0.246472    0.222779     −1.106     0.2686  
  TAXR         −0.120172    0.521025     −0.2306    0.8176  
  DFC_IND2     −1.57189     3.78622      −0.4152    0.6780  
  PROFI        −2.72475     2.78955      −0.9768    0.3287  
  SIC2DIGIT     0.0622110   0.0667275     0.9323    0.3512  

 equity:
  const         6.97190     2.69931       2.583     0.0098   ***
  POT2          4.30132     0.962711      4.468     7.90e-06 ***
  TAILLE       −0.467652    0.179287     −2.608     0.0091   ***
  INV_ANT       2.95192     1.09282       2.701     0.0069   ***
  CF_ANT       −1.17203     2.01875      −0.5806    0.5615  
  VOLAT        22.1561      7.75291       2.858     0.0043   ***
  D_DIV        −1.24048     0.431633     −2.874     0.0041   ***
  SCOREZ       −1.53191     0.651642     −2.351     0.0187   **
  D_V           0.960271    1.29438       0.7419    0.4582  
  DRD          −4.39569     7.20260      −0.6103    0.5417  
  MTB           0.378008    0.147044      2.571     0.0101   **
  RFR          −0.420355    0.323356     −1.300     0.1936  
  TANGI        −7.06169     2.42802      −2.908     0.0036   ***
  AGE_BASE      0.0755413   0.0532725     1.418     0.1562  
  REND          0.0585621   0.257865      0.2271    0.8203  
  TAXR         −0.914275    1.10068      −0.8306    0.4062  
  DFC_IND2     −3.22315     6.17726      −0.5218    0.6018  
  PROFI       −21.6354      6.27013      −3.451     0.0006   ***
  SIC2DIGIT    −0.125967    0.0890904    −1.414     0.1574  

Log-likelihood      −128.5415   Akaike criterion     335.0830
Schwarz criterion    479.0070   Hannan-Quinn         392.7073

rho = 1

Test of independence -
  Null hypothesis: rho = 0
  Test statistic: Chi-square(1) = 16.4347
  with p-value = 5.03547e-005

# genr matrix predict_prob = $yhat
? genr series predict_external = $yhat[,1]
Replaced series predict_external (ID 63)
? genr series predict_equity = $yhat[,2]
Replaced series predict_equity (ID 64)
? genr series prob_external = cdf(N,predict_external)
Replaced series prob_external (ID 65)
? genr series prob_equity = cdf(N, predict_equity)
Replaced series prob_equity (ID 66)
? genr correct_ext1 = sum((prob_external>=0.5) && external=1)/sum(external=1)
Replaced scalar correct_ext1 = 0.512195
? genr correct_ext0 = sum((prob_external<0.5) && external=0)/sum(external=0)
Replaced scalar correct_ext0 = 0.96729
? genr correct_ext = sum((prob_external>=0.5)==external)/$nobs
Replaced scalar correct_ext = 0.841216
? smpl external=1 --restrict
Full data set: 2590 observations
Current sample: 82 observations
? genr correct_eqt1 = sum((prob_equity>=0.5) && equity=1)/sum(equity=1)
Replaced scalar correct_eqt1 = 0.454545
? genr correct_eqt0 = sum((prob_equity<0.5) && equity=0)/sum(equity=0)
Replaced scalar correct_eqt0 = 0.985915
? genr correct_eqt = sum((prob_equity>=0.5)==equity)/$nobs
Replaced scalar correct_eqt = 0.914634
? smpl full
Full data range: 1 - 2590 (n = 2590)

# Impression des résultats de l'estimation
? print correct_ext1 correct_ext0	correct_ext correct_eqt1 correct_eqt0 \
  correct_eqt

   correct_ext1 =  0.51219512

   correct_ext0 =  0.96728972

    correct_ext =  0.84121622

   correct_eqt1 =  0.45454545

   correct_eqt0 =  0.98591549

    correct_eqt =  0.91463415
<div><div dir="ltr"><div class="gmail_extra"><div class="gmail_quote">
<blockquote class="gmail_quote"><div class="gmail_extra"><div class="gmail_quote"><blockquote class="gmail_quote"><div><div class="h5">
<span><br></span>
There's no doubt your likelihood function is misbehaved here (rho near 1 is worrying) and what you're seeing are numerical problems.<div><div><br></div></div>
</div></div></blockquote></div></div></blockquote>
<div><br></div>
<div>My question: When it comes to rho "near 1" what about the extreme case of rho=1? I tried another biprobit estimation which shows up rho=1...with no error message. At that point, should the likelihood computations have stopped?</div>
<div>(just in case, here's attached the rho-equal-one output)</div>
<div>Artur</div>
</div></div></div></div>
henrique.andrade | 20 Oct 22:29 2014
Picon

Re: Wildcard problem in the latest snapshot

Dear Allin,

I think the problem with wildcards is not completely solved. Please take a look at the following script (you
can find the used dataset attached):

<hansl>
open "Dados.gdt" --quiet

loop for i=1..12 --quiet
    store " <at> dotdir\Projeções ($i passo).gdt" ibcbr proj_*_$i
endloop

loop for i=1..12 --quiet
    store " <at> dotdir\Projeções ($i passo).gdt" *_$i
endloop

</hansl>

This commands work just fine, but that one don't:

<hansl>
loop for i=1..12 --quiet
    store " <at> dotdir\Projeções ($i passo).gdt" ibcbr *_$i
endloop
</hansl>

Best,
Henrique Coêlho de Andrade
Diretoria de Estratégia e Organização
Divisão de Cenários e Estudos Macroeconômicos
Banco do Brasil
henrique.andrade@...
Tel.: (61) 3102-6911

-----F1831737 Carlos Henrique Coelho de Andrade/BancodoBrasil escreveu: -----
Para: Gretl list <gretl-users@...>
De: F1831737 Carlos Henrique Coelho de Andrade/BancodoBrasil
Data: 03/10/2014 12:15
Assunto: Re: [Gretl-users] Wildcard problem in the latest snapshot

Dear Allin,

The error message is gone. But the problem remains :-( I think we found a problem greater than the wildcard
handling. Take a look ate the following script:

<hansl>
open australia.gdt

list L = PAU PUS E
eval varname(L)
delete L
</hansl>

Gretl is not deleting lists.

Atenciosamente,
Henrique Coêlho de Andrade
Diretoria de Estratégia e Organização
Divisão de Cenários e Estudos Macroeconômicos
Banco do Brasil
henrique.andrade@...

Tel.: (61) 3102-6911

-----gretl-users-bounces@... escreveu: -----
Para: Gretl list <gretl-users@...>
De: Allin Cottrell 
Enviado por: gretl-users-bounces@...
Data: 01/10/2014 16:50
Assunto: Re: [Gretl-users] Wildcard problem in the latest snapshot

On Wed, 1 Oct 2014, henrique.andrade@... wrote:

> Dear Allin,
> 
> I think the latest snapshot (Windows and Mac) introduced a problem with the
> use of wildcards. Please take a look at the following script:
> 
> <hansl>
> open australia.gdt
> 
> rename e2 e_m
> rename lpus lp_m
> 
> list L = dataset - *_m
> </hansl>
> 
> When I execute this script I get the error message: "Unexpected symbol '_'".
> (And this message are not marked for translation).

Thanks for the report. That's now fixed in CVS; snapshots will follow.

Allin
_______________________________________________
Gretl-users mailing list
Gretl-users@...
http://lists.wfu.edu/mailman/listinfo/gretl-users

Attachment (dados.gdt): application/octet-stream, 45 KiB
Attachment (Wildcard problem.inp): application/octet-stream, 427 bytes
Dear Allin,

I think the problem with wildcards is not completely solved. Please take a look at the following script (you
can find the used dataset attached):

<hansl>
open "Dados.gdt" --quiet

loop for i=1..12 --quiet
    store " <at> dotdir\Projeções ($i passo).gdt" ibcbr proj_*_$i
endloop

loop for i=1..12 --quiet
    store " <at> dotdir\Projeções ($i passo).gdt" *_$i
endloop

</hansl>

This commands work just fine, but that one don't:

<hansl>
loop for i=1..12 --quiet
    store " <at> dotdir\Projeções ($i passo).gdt" ibcbr *_$i
endloop
</hansl>

Best,
Henrique Coêlho de Andrade
Diretoria de Estratégia e Organização
Divisão de Cenários e Estudos Macroeconômicos
Banco do Brasil
henrique.andrade@...
Tel.: (61) 3102-6911

-----F1831737 Carlos Henrique Coelho de Andrade/BancodoBrasil escreveu: -----
Para: Gretl list <gretl-users@...>
De: F1831737 Carlos Henrique Coelho de Andrade/BancodoBrasil
Data: 03/10/2014 12:15
Assunto: Re: [Gretl-users] Wildcard problem in the latest snapshot

Dear Allin,

The error message is gone. But the problem remains :-( I think we found a problem greater than the wildcard
handling. Take a look ate the following script:

<hansl>
open australia.gdt

list L = PAU PUS E
eval varname(L)
delete L
</hansl>

Gretl is not deleting lists.

Atenciosamente,
Henrique Coêlho de Andrade
Diretoria de Estratégia e Organização
Divisão de Cenários e Estudos Macroeconômicos
Banco do Brasil
henrique.andrade@...

Tel.: (61) 3102-6911

-----gretl-users-bounces@... escreveu: -----
Para: Gretl list <gretl-users@...>
De: Allin Cottrell 
Enviado por: gretl-users-bounces@...
Data: 01/10/2014 16:50
Assunto: Re: [Gretl-users] Wildcard problem in the latest snapshot

On Wed, 1 Oct 2014, henrique.andrade@... wrote:

> Dear Allin,
> 
> I think the latest snapshot (Windows and Mac) introduced a problem with the
> use of wildcards. Please take a look at the following script:
> 
> <hansl>
> open australia.gdt
> 
> rename e2 e_m
> rename lpus lp_m
> 
> list L = dataset - *_m
> </hansl>
> 
> When I execute this script I get the error message: "Unexpected symbol '_'".
> (And this message are not marked for translation).

Thanks for the report. That's now fixed in CVS; snapshots will follow.

Allin
_______________________________________________
Gretl-users mailing list
Gretl-users@...
http://lists.wfu.edu/mailman/listinfo/gretl-users

Alecos Papadopoulos | 20 Oct 00:57 2014
Picon

Re: The "invcdf" function -Correction needed in Help?

Thanks Sven , your explanations were helpful.

As regards the invcdf function,  according to Help, the invcdf function 
can be used for both the Binomial and the Poisson distributions. But it 
indeed seems not to work for either, I tried it for various parameter 
values.
So perhaps the Help should be corrected?

Alecos Papadopoulos
Athens University of Economics and Business, Greece
Department of Economics
cell:+30-6945-378680
fax: +30-210-8259763
skype:alecos.papadopoulos

On 16/10/2014 19:00, gretl-users-request@... wrote:
>> >Now, the function
>> >invcdf(B, prob, 1, 0.05)
>> >returns "NA"
> A CDF for discrete variables is not strictly increasing and thus not
> invertible. This could be the rationale for NA. Perhaps alternative
> conventions could be used for invcdf(), but you would have to argue that
> it's useful or necessary.
>
> hth,
> sven

David van Herick | 19 Oct 09:18 2014
Picon

Nested Logit

Does anyone know if it is possible to run a nested logit model in gretl?  It seems it must be possible somehow with a hansl script or something, but I can't seem to figure it out.  I am hoping for an FIML version, but would be happy to even find something that will do a two-step LIML version with multinomial logit models that include logsum parameters.

<div><div dir="ltr"><div>Does anyone know if it is possible to run a nested logit model in gretl?&nbsp; It seems it must be possible somehow with a hansl script or something, but I can't seem to figure it out.&nbsp; I am hoping for an FIML version, but would be happy to even find something that will do a two-step LIML version with multinomial logit models that include logsum parameters.<br><br>
</div></div></div>
Matteo Chiorboli | 17 Oct 04:41 2014
Picon

Cointegration tests on many pairs of historic series

Dear Gretl users, 

I need your help for running the Johansen test automatically on (n) historic series, matched in pairs. 
Substantially, I have 30 historic series, and I must test the cointegration between all the possible pairs. Then, existing 435 possible pairs, I need an automatic procedure to make this great number of Johansen tests. 
Overall, I need a procedure that can run all possible cointegration tests on a number of pairs of historic series given by the binomial coefficient (combination), that in the specific case is n=30, k=2, then [30!/(2!*(30-2)!]=435 pairs, and 435 Johansen tests.

Thank you for you help!

Matteo Chiorboli
<div><div dir="ltr">Dear Gretl users,&nbsp;<div><br></div>
<div>I need your help for running the Johansen test automatically on (n) historic series, matched in pairs.&nbsp;</div>
<div>Substantially, I have 30 historic series, and I must test the cointegration between all the possible pairs. Then, existing 435 possible pairs, I need an automatic procedure to make this great number of Johansen tests.&nbsp;</div>
<div>Overall, I need a procedure that can run all possible cointegration tests on a number of pairs of historic series given by the binomial coefficient (combination), that in the specific case is n=30, k=2, then [30!/(2!*(30-2)!]=435 pairs, and 435 Johansen tests.</div>
<div><br></div>
<div>Thank you for you help!</div>
<div><br></div>
<div>Matteo Chiorboli</div>
</div></div>
Alecos Papadopoulos | 16 Oct 05:24 2014
Picon

Re: Bernoulli and the "critical" and "invcdf" functions

Good afternoon.
I run gretl 1.9.92 64-bit for Windows.

For 0.05 < prob < 0.95
The function
critical(B, prob, 1, p=0.05)
returns the value "1".
As the Help says, the function "returns x such that P(X > x) = p". Since 
number of trials is equal to 1, we just have a Bernoulli, with P(X=0) = 
1-prob and P(X=1) =prob. So by returning "1" the function asserts that 
P(X > 1) = 0.05 > 0
If the function returned "0" it would assert that P(X > 0) = 0.05 < prob 
=P(X=1)
I take note of the asymmetry (in the first case, the function "sends 
some probability to the right"), -can this be the rule that leads to the 
function returning "1"?

Now, the function
invcdf(B, prob, 1, 0.05)
returns "NA"
As the Help says the function "returns x such that P(X ≤ x) = p"
So here a return of "1" would assert P(X =<1) = 0.05 < 1, while a return 
of "0" would assert that P(X =<0) =0.05 < 1-Prob.
Here in both cases, returning a number, either 0 or 1, would 
"under-signal" the true probability - so maybe this is the reason why 
here we get "NA"?

It appears as a choice to prefer making the rejection of a null 
hypothesis harder rather than easier... but most probably I am imagining 
things.

Can somebody shed some light?

--

-- 
Alecos Papadopoulos
Athens University of Economics and Business, Greece
Department of Economics
cell:+30-6945-378680
fax: +30-210-8259763
skype:alecos.papadopoulos

_______________________________________________
Gretl-users mailing list
Gretl-users <at> lists.wfu.edu
http://lists.wfu.edu/mailman/listinfo/gretl-users
Summers, Peter | 15 Oct 17:32 2014

crash in icon view

I’ve come across another bug, this one causing a crash on Windows 7. If I open a data set (I’ve tried this with several), open the icon view, then double-click on either the “summary” or “correlations” icon, everything goes gray and I get a message saying “gretl.exe has stopped working. A problem caused the program to stop working correctly. Windows will close the program and notify you if a solution is available.” This is v. 1.10.0cvs, built 10/9. The dataset info and notes icons open fine.

 

TIA,

 

PS

 

 

<div>
<div class="WordSection1">
<p class="MsoNormal">I&rsquo;ve come across another bug, this one causing a crash on Windows 7. If I open a data set (I&rsquo;ve tried this with several), open the icon view, then double-click on either the &ldquo;summary&rdquo; or &ldquo;correlations&rdquo; icon, everything goes gray and I get
 a message saying &ldquo;gretl.exe has stopped working. A problem caused the program to stop working correctly. Windows will close the program and notify you if a solution is available.&rdquo; This is v. 1.10.0cvs, built 10/9. The dataset info and notes icons open fine.<p></p></p>
<p class="MsoNormal"><p>&nbsp;</p></p>
<p class="MsoNormal">TIA,<p></p></p>
<p class="MsoNormal"><p>&nbsp;</p></p>
<p class="MsoNormal">PS<p></p></p>
<p class="MsoNormal"><p>&nbsp;</p></p>
<p class="MsoNormal"><p></p></p>
<p class="MsoNormal"><p>&nbsp;</p></p>
</div>
</div>
Summers, Peter | 15 Oct 16:41 2014

scatterplot bug

Folks,

 

I think I’ve found a bug in the GUI scatterplot command. I generate a scatterplot, then edit it and change the default point marker (eg, from ‘+’ to ‘x’). I click apply to make the change, but when I try to change it back, nothing happens. This is on Windows 7, 64-bit cvs, v. 1.10.0, built 10/9.

 

PS

 

 

<div>
<div class="WordSection1">
<p class="MsoNormal">Folks,<p></p></p>
<p class="MsoNormal"><p>&nbsp;</p></p>
<p class="MsoNormal">I think I&rsquo;ve found a bug in the GUI scatterplot command. I generate a scatterplot, then edit it and change the default point marker (eg, from &lsquo;+&rsquo; to &lsquo;x&rsquo;). I click apply to make the change, but when I try to change it back, nothing happens.
 This is on Windows 7, 64-bit cvs, v. 1.10.0, built 10/9. <p></p></p>
<p class="MsoNormal"><p>&nbsp;</p></p>
<p class="MsoNormal">PS<p></p></p>
<p class="MsoNormal"><p>&nbsp;</p></p>
<p class="MsoNormal"><p></p></p>
<p class="MsoNormal"><p>&nbsp;</p></p>
</div>
</div>
Jan Tille | 15 Oct 08:23 2014
Picon

Is there a possibility to estimate a dynamic panel probit model in gretl?

Dear users,

 

I have a question, is there a fuction to estimate a dynamic probit panel model in gretl? I have only seen panel models and dynamic panel models in the user guide, but neither probit panel nor dynamic probit panel models.

Thanks in advance,

 

Jan

 

 

<div><div class="WordSection1">
<p class="MsoNormal"><span>Dear users,<p></p></span></p>
<p class="MsoNormal"><span><p>&nbsp;</p></span></p>
<p class="MsoNormal"><span lang="EN-US">I have a question, is there a fuction to estimate a dynamic probit panel model in gretl? I have only seen panel models and dynamic panel models in the user guide, but neither probit panel nor dynamic probit panel models.<p></p></span></p>
<p class="MsoNormal"><span lang="EN-US">Thanks in advance,<p></p></span></p>
<p class="MsoNormal"><span lang="EN-US"><p>&nbsp;</p></span></p>
<p class="MsoNormal"><span lang="EN-US">Jan<p></p></span></p>
<p class="MsoNormal"><span lang="EN-US"><p>&nbsp;</p></span></p>
<p class="MsoNormal"><span lang="EN-US"><p>&nbsp;</p></span></p>
</div></div>

Gmane