Results of a subsample of family business, when the independent variable is ROE
| Indicator/Variables | (1) | (2) | (3) | (4) | (5) |
|---|---|---|---|---|---|
| const | 11.58 | 8.408 | 9.041 | 17.73 | 3.582 |
| ESG | 0.127*** | ||||
| Envir | 0.036 | ||||
| Social | 0.105*** | ||||
| Governance | 0.161*** | ||||
| env_bin | 4.364 | ||||
| lev | 0.04 | 0.059 | 0.051 | 0.03 | 0.056 |
| tac | 0.162** | 0.154** | 0.165** | 0.148** | 0.142** |
| cash_ratio | 66.21*** | 70.7*** | 67.04*** | 65.91*** | 71.13*** |
| size | −0.564 | −0.395 | −0.433 | −0.975* | −0.318 |
| bs | −0.209 | −0.012 | −0.219 | −0.004 | 0.045 |
| EPS | 0.466*** | 0.974** | 0.461*** | 0.487*** | 0.906** |
| sector | 1.011*** | 0.473*** | 0.91*** | 1.064*** | 0.471*** |
| rsquared | 0.361 | 0.346 | 0.362 | 0.375 | 0.348 |
| rsquared_overall | 0.361 | 0.346 | 0.362 | 0.375 | 0.348 |
| rsquared_betwee | 0.286 | 0.269 | 0.287 | 0.287 | 0.28 |
| rsquared_within | 0.418 | 0.419 | 0.419 | 0.404 | 0.415 |
| Number of obs | 319 | 319 | 319 | 319 | 319 |
| Years periods | 9 | 9 | 9 | 9 | 9 |
| Indicator/Variables | (1) | (2) | (3) | (4) | (5) |
|---|---|---|---|---|---|
| const | 11.58 | 8.408 | 9.041 | 17.73 | 3.582 |
| Envir | 0.036 | ||||
| Social | |||||
| Governance | |||||
| env_bin | 4.364 | ||||
| lev | 0.04 | 0.059 | 0.051 | 0.03 | 0.056 |
| tac | 0.162 | 0.154 | 0.165 | 0.148 | 0.142 |
| cash_ratio | 66.21 | 70.7 | 67.04 | 65.91 | 71.13 |
| size | −0.564 | −0.395 | −0.433 | −0.975 | −0.318 |
| bs | −0.209 | −0.012 | −0.219 | −0.004 | 0.045 |
| 0.466 | 0.974 | 0.461 | 0.487 | 0.906 | |
| sector | 1.011 | 0.473 | 0.91 | 1.064 | 0.471 |
| rsquared | 0.361 | 0.346 | 0.362 | 0.375 | 0.348 |
| rsquared_overall | 0.361 | 0.346 | 0.362 | 0.375 | 0.348 |
| rsquared_betwee | 0.286 | 0.269 | 0.287 | 0.287 | 0.28 |
| rsquared_within | 0.418 | 0.419 | 0.419 | 0.404 | 0.415 |
| Number of obs | 319 | 319 | 319 | 319 | 319 |
| Years periods | 9 | 9 | 9 | 9 | 9 |
The results are the parameters of the fixed-effects model, which includes both individual and time effects. When applying the Breusch and Pagan (1979) test, we found evidence to reject the null hypothesis of homoscedasticity in columns 1, 3 and 4, using robust standard errors. The variable sector and the binary variable family are not included in the results because the fixed-effect model absorbs time-invariant characteristics. The ***, ** and *indicate parameter significance at 1, 5 and 10%, respectively