Table 7

Matching results for the variables before and after entropy balancing

TreatControl (before entropy weighting)Control (after entropy weighting)
VariableMeanVarianceSkewnessMeanVarianceSkewnessMeanVarianceSkewness
Size22.1401.6100.73122.2401.7070.72922.1401.6100.731
Lev0.4220.0430.1810.4280.0390.2140.4220.0430.181
Roa0.0510.002−0.4160.0460.002−0.7920.0510.002−0.416
Cfo0.0470.005−0.1420.0510.004−0.0260.0470.005−0.142
PB3.5617.9252.6643.6779.5052.6463.5617.9252.664
Exfin1.94273.0407.1712.13385.7506.6801.94273.0407.170
Top10.3440.0210.5120.3520.0220.4600.3440.0210.512
Indep0.3730.0031.4460.3770.0031.2430.3730.0031.446
Board2.1450.039−0.2912.1220.040−0.2882.1450.039−0.291
Exeshare0.0780.0212.1090.0750.0202.1500.0780.0212.109
Exepay14.340.5180.16214.430.4410.23214.340.5180.162
Audit0.9820.018−7.2930.9820.018−7.1780.9820.018−7.293

Note(s): This table reports matching results for the variables before and after entropy balancing. We divide the whole sample into two groups based on Commun, in which the high-interaction companies are the treated group and the low-interaction companies are the control group. The control variables in model (2) are weighted to reduce the differences between the two groups in the first moment (mean), second moment (variance) and third moment (skewness). Ultimately, we obtained 16,665 matching sample observations

or Create an Account

Close subscription notice
Close access options