Fixed effects estimator
allows for differing intercepts by cross-sectional unit
slope standard errors in parentheses, p-values in brackets

           const:         6,6168       (0,39494)       [0,00000]
           LNASS:        0,43447      (0,022607)       [0,00000]
          RECINV:        0,34046       (0,12227)       [0,00540]
             CUR:      -0,028227     (0,0061409)       [0,00000]
             LEV:       0,071352      (0,055074)       [0,19524]
             ROA:       -0,20229      (0,065877)       [0,00216]
           DLOSS:       0,034694      (0,016256)       [0,03292]
         DFORSAL:       0,050056      (0,022651)       [0,02720]
            LSEG:       0,079557      (0,034226)       [0,02018]
          GROWTH:       -0,01861      (0,010161)       [0,06714]
          TENURE:      -0,011962      (0,011755)       [0,30894]
         DBIG4_5:         0,1815      (0,039969)       [0,00001]
          AUDCHG:       -0,12032      (0,022849)       [0,00000]
          REPLAG:     0,00027518    (0,00013913)       [0,04805]
             AS5:     -0,0059255      (0,017407)       [0,73358]
              MW:        0,11478      (0,016911)       [0,00000]
           DMERG:      -0,021615      (0,011497)       [0,06023]
           DYEND:        0,34298      (0,089382)       [0,00013]
             D04:     -0,0026861      (0,069166)       [0,96902]
             D05:       0,002703      (0,061857)       [0,96515]
             D06:       0,022061      (0,057413)       [0,70082]
             D07:        0,01961      (0,051107)       [0,70123]

1576 group means were subtracted from the data

Residual variance: 110,621/(4220 - 1597) = 0,0421734
Joint significance of differing group means:
 F(1575, 2623) = 15,4547 with p-value 0
(A low p-value counts against the null hypothesis that the pooled OLS model
is adequate, in favor of the fixed effects alternative.)


Breusch-Pagan test statistic:
 LM = 2278,99 with p-value = prob(chi-square(1) > 2278,99) = 0
(A low p-value counts against the null hypothesis that the pooled OLS model
is adequate, in favor of the random effects alternative.)

Variance estimators:
 between = 0,283045
 within = 0,0421734
Panel is unbalanced: theta varies across units

                         Random effects estimator
           allows for a unit-specific component to the error term
           (standard errors in parentheses, p-values in brackets)

           const:         5,6998        (0,3173)       [0,00000]
           LNASS:        0,50545     (0,0087628)       [0,00000]
          RECINV:        0,50441      (0,076668)       [0,00000]
             CUR:      -0,031813     (0,0051885)       [0,00000]
             LEV:       0,096529      (0,045652)       [0,03454]
             ROA:       -0,24389      (0,060915)       [0,00006]
           DLOSS:        0,05316      (0,015811)       [0,00078]
         DFORSAL:          0,131      (0,019103)       [0,00000]
            LSEG:        0,19641      (0,022799)       [0,00000]
          GROWTH:      -0,023833     (0,0075204)       [0,00154]
          TENURE:      -0,013222      (0,010124)       [0,19165]
         DBIG4_5:        0,18762      (0,031187)       [0,00000]
          AUDCHG:       -0,11928      (0,022126)       [0,00000]
          REPLAG:     0,00016691    (0,00012903)       [0,19590]
             AS5:      -0,010003       (0,01739)       [0,56518]
              MW:        0,13794      (0,016433)       [0,00000]
        Dagicult:       -0,50015       (0,31831)       [0,11619]
         Dmining:        -0,3604       (0,27913)       [0,19672]
           Dfood:       -0,32364       (0,27789)       [0,24423]
       Dtextiles:       -0,25208        (0,2733)       [0,35640]
          Ddrugs:       -0,11158       (0,27274)       [0,68248]
           Dchem:       -0,12496       (0,27501)       [0,64958]
          Drefin:       -0,47633       (0,27582)       [0,08425]
          Drubbr:       -0,28344       (0,27377)       [0,30057]
         Dindust:      -0,022252       (0,27253)       [0,93493]
         Delectr:      -0,034394       (0,27733)       [0,90131]
         Dmisceq:      -0,027584        (0,2722)       [0,91929]
          Dcompu:      -0,087462       (0,27139)       [0,74726]
         Dtransp:       -0,55602       (0,27362)       [0,04221]
           Dutil:        -0,6496       (0,27447)       [0,01799]
         Dretail:       -0,42315       (0,27134)       [0,11896]
           Dbank:       -0,36939       (0,29912)       [0,21694]
         Dservic:      -0,078757       (0,27526)       [0,77480]
           DMERG:      -0,021772      (0,011042)       [0,04871]
           DYEND:       0,010759      (0,031185)       [0,73011]
             D04:      -0,024989      (0,064815)       [0,69985]
             D05:      -0,026265      (0,059034)       [0,65641]
             D06:      -0,015157      (0,055811)       [0,78596]
             D07:      -0,022875      (0,050169)       [0,64844]

Hausman test matrix is not positive definite (this result may be treated as
"fail to reject" the random effects specification).