Carga y exploración de
datos
## # A tibble: 6 × 10
## Empresa Nemotecnico Fecha ROE ROA Tamano_Junta Mujeres_Junta
## <fct> <chr> <date> <dbl> <dbl> <dbl> <dbl>
## 1 RIOPAILA AGRICO… IRP.CN 2023-12-31 6.17 4.13 5 0
## 2 RIOPAILA AGRICO… IRP.CN 2022-12-31 9.74 6.58 5 0
## 3 RIOPAILA AGRICO… IRP.CN 2021-12-31 6.05 4.17 5 0
## 4 RIOPAILA AGRICO… IRP.CN 2020-12-31 3.38 2.29 5 0
## 5 RIOPAILA AGRICO… IRP.CN 2019-12-31 1.96 1.26 5 0
## 6 RIOPAILA AGRICO… IRP.CN 2018-12-31 -2.24 -1.36 5 0
## # ℹ 3 more variables: Porc_Mujeres_Junta <dbl>, Rep_Legal <dbl>, Anio <chr>
Análisis
descriptivo
# Cargar paquetes
library(dplyr)
library(plm)
library(lmtest)
# Asegurarse que Anio sea numérico
datos <- datos %>%
mutate(Anio = as.numeric(Anio))
# Crear panel
panel_datos <- pdata.frame(datos, index = c("Empresa", "Anio"))
Modelo ROE
# Modelo para ROE
modelo_roe_fe <- plm(ROE ~ Porc_Mujeres_Junta + Rep_Legal, data = panel_datos, model = "within")
modelo_roe_re <- plm(ROE ~ Porc_Mujeres_Junta + Rep_Legal, data = panel_datos, model = "random")
# Test de Hausman para ROE
hausman_roe <- phtest(modelo_roe_fe, modelo_roe_re)
# Test de Breusch-Pagan (LM test) para ROE (para ver si RE es mejor que OLS)
modelo_roe_pool <- plm(ROE ~ Porc_Mujeres_Junta, data = panel_datos, model = "pooling")
bp_roe <- plmtest(modelo_roe_pool, type = "bp")
summary(modelo_roe_re)
## Oneway (individual) effect Random Effect Model
## (Swamy-Arora's transformation)
##
## Call:
## plm(formula = ROE ~ Porc_Mujeres_Junta + Rep_Legal, data = panel_datos,
## model = "random")
##
## Balanced Panel: n = 32, T = 11, N = 352
##
## Effects:
## var std.dev share
## idiosyncratic 51.062 7.146 0.46
## individual 59.842 7.736 0.54
## theta: 0.7317
##
## Residuals:
## Min. 1st Qu. Median 3rd Qu. Max.
## -35.73353 -2.67152 -0.11726 3.25438 24.42244
##
## Coefficients:
## Estimate Std. Error z-value Pr(>|z|)
## (Intercept) 8.340996 1.514500 5.5074 3.641e-08 ***
## Porc_Mujeres_Junta 0.042303 0.034644 1.2211 0.2221
## Rep_Legal 2.654427 1.702958 1.5587 0.1191
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Total Sum of Squares: 17915
## Residual Sum of Squares: 17732
## R-Squared: 0.010182
## Adj. R-Squared: 0.0045098
## Chisq: 3.59011 on 2 DF, p-value: 0.16612
hausman_roe
##
## Hausman Test
##
## data: ROE ~ Porc_Mujeres_Junta + Rep_Legal
## chisq = 0.25378, df = 2, p-value = 0.8808
## alternative hypothesis: one model is inconsistent
bp_roe
##
## Lagrange Multiplier Test - (Breusch-Pagan)
##
## data: ROE ~ Porc_Mujeres_Junta
## chisq = 476.27, df = 1, p-value < 2.2e-16
## alternative hypothesis: significant effects
Modelo ROA
# Modelo para ROA
modelo_roa_fe <- plm(ROA ~ Porc_Mujeres_Junta + Rep_Legal, data = panel_datos, model = "within")
modelo_roa_re <- plm(ROA ~ Porc_Mujeres_Junta + Rep_Legal, data = panel_datos, model = "random")
# Test de Hausman para ROA
hausman_roa <- phtest(modelo_roa_fe, modelo_roa_re)
# Test de Breusch-Pagan para ROA
modelo_roa_pool <- plm(ROA ~ Porc_Mujeres_Junta, data = panel_datos, model = "pooling")
bp_roa <- plmtest(modelo_roa_pool, type = "bp")
# Resultados
summary(modelo_roa_re)
## Oneway (individual) effect Random Effect Model
## (Swamy-Arora's transformation)
##
## Call:
## plm(formula = ROA ~ Porc_Mujeres_Junta + Rep_Legal, data = panel_datos,
## model = "random")
##
## Balanced Panel: n = 32, T = 11, N = 352
##
## Effects:
## var std.dev share
## idiosyncratic 9.074 3.012 0.377
## individual 15.009 3.874 0.623
## theta: 0.7718
##
## Residuals:
## Min. 1st Qu. Median 3rd Qu. Max.
## -13.89580 -1.03237 -0.17534 1.03119 11.43483
##
## Coefficients:
## Estimate Std. Error z-value Pr(>|z|)
## (Intercept) 3.138286 0.738613 4.2489 2.148e-05 ***
## Porc_Mujeres_Junta 0.021036 0.014750 1.4262 0.1538
## Rep_Legal 0.401232 0.725002 0.5534 0.5800
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Total Sum of Squares: 3176.6
## Residual Sum of Squares: 3156.6
## R-Squared: 0.0062905
## Adj. R-Squared: 0.00059587
## Chisq: 2.20928 on 2 DF, p-value: 0.33133
hausman_roa
##
## Hausman Test
##
## data: ROA ~ Porc_Mujeres_Junta + Rep_Legal
## chisq = 0.82605, df = 2, p-value = 0.6616
## alternative hypothesis: one model is inconsistent
bp_roa
##
## Lagrange Multiplier Test - (Breusch-Pagan)
##
## data: ROA ~ Porc_Mujeres_Junta
## chisq = 645.59, df = 1, p-value < 2.2e-16
## alternative hypothesis: significant effects
Modelos con variables
de control
modelo_roa_re_ctrl <- plm(ROA ~ Porc_Mujeres_Junta + Tamano_Junta + Rep_Legal + factor(Anio),
data = panel_datos, model = "random")
summary(modelo_roa_re_ctrl)
## Oneway (individual) effect Random Effect Model
## (Swamy-Arora's transformation)
##
## Call:
## plm(formula = ROA ~ Porc_Mujeres_Junta + Tamano_Junta + Rep_Legal +
## factor(Anio), data = panel_datos, model = "random")
##
## Balanced Panel: n = 32, T = 11, N = 352
##
## Effects:
## var std.dev share
## idiosyncratic 9.150 3.025 0.381
## individual 14.868 3.856 0.619
## theta: 0.7698
##
## Residuals:
## Min. 1st Qu. Median 3rd Qu. Max.
## -13.0566 -1.0725 -0.1595 1.0645 11.4355
##
## Coefficients:
## Estimate Std. Error z-value Pr(>|z|)
## (Intercept) 1.7943258 2.0407943 0.8792 0.3793
## Porc_Mujeres_Junta 0.0187030 0.0173722 1.0766 0.2817
## Tamano_Junta 0.2024458 0.2764550 0.7323 0.4640
## Rep_Legal 0.3693875 0.7391765 0.4997 0.6173
## factor(Anio)2014 -0.0079409 0.7551993 -0.0105 0.9916
## factor(Anio)2015 0.0980519 0.7565344 0.1296 0.8969
## factor(Anio)2016 0.2097068 0.7657754 0.2738 0.7842
## factor(Anio)2017 -0.0434061 0.7678077 -0.0565 0.9549
## factor(Anio)2018 -0.4675181 0.7709117 -0.6064 0.5442
## factor(Anio)2019 0.2708168 0.7677299 0.3528 0.7243
## factor(Anio)2020 -0.7344306 0.7763897 -0.9460 0.3442
## factor(Anio)2021 0.4511446 0.7897960 0.5712 0.5679
## factor(Anio)2022 0.9093640 0.7993759 1.1376 0.2553
## factor(Anio)2023 -0.6348824 0.8185599 -0.7756 0.4380
##
## Total Sum of Squares: 3181.3
## Residual Sum of Squares: 3080.4
## R-Squared: 0.031715
## Adj. R-Squared: -0.0055264
## Chisq: 11.0709 on 13 DF, p-value: 0.60488
Verificación de
supuestos
Heterocedasticidad
(Test de Breusch-Pagan)
##
## Breusch-Pagan test
##
## data: modelo_roa_re
## BP = 7.9674, df = 2, p-value = 0.01862
##
## Breusch-Pagan test
##
## data: modelo_roe_re
## BP = 4.1576, df = 2, p-value = 0.1251
Autocorrelación (Test
de Wooldridge)
##
## Durbin-Watson test for serial correlation in panel models
##
## data: ROA ~ Porc_Mujeres_Junta + Rep_Legal
## DW = 1.3455, p-value = 2.615e-10
## alternative hypothesis: serial correlation in idiosyncratic errors
##
## Durbin-Watson test for serial correlation in panel models
##
## data: ROE ~ Porc_Mujeres_Junta + Rep_Legal
## DW = 1.4284, p-value = 2.75e-08
## alternative hypothesis: serial correlation in idiosyncratic errors
Multicolinealidad
(VIF)
## Porc_Mujeres_Junta Rep_Legal
## 1.006886 1.006886
## Porc_Mujeres_Junta Rep_Legal
## 1.006886 1.006886