Table 10 Two-stage least square (2SLS)...

Table 10

Two-stage least square (2SLS) regression

Parameter	First stage	Second STAGE	Exclusion criteria	Second stage	Exclusion criteria
Parameter	(1)	SRV (2)	(3)	ROAV (4)	(5)
Intercept	0.006	0.381		0.344
Lag_EngCEO (IV)	0.498*** (8.450)		0.013 (1.299)		0.010 (0.984)
Predicted_EngCEO		0.027* (1.690)		0.022** (2.031)
ROA	0.000 (0.190)	0.204*** (4.099)		0.181*** (3.790)
Leverage	−0.060* (1.850)	0.001 (0.996)		0.002 (1.201)
FirmSize	0.017*** (3.230)	−0.012*** (−6.579)		−0.009*** (−5.454)
SalesGrowth	0.001 (0.670)	−0.008 (−1.142)		−0.006 (−1.015)
Bsize	−0.002 (0.480)	−0.004** (−2.296)		−0.003** (−2.149)
BInd	−0.015 (0.520)	0.000 (0.964)		0.001 (0.877)
CAPEX	−0.001 (0.250)	0.0002 (1.454)		0.000 (1.279)
CEOGEN	−0.003 (0.100)	0.0005 (0.045)		0.001 (0.056)
CEODUA	0.017 (0.890)	−0.015 (−1.560)		−0.014 (−1.431)
BEng	0.181*** (3.600)	0.0024 (0.123)		0.001 (0.105)
Year and industry FE	Yes	Yes		Yes
R-squared	0.441	0.200	0.00	0.210	0.00
F-statistic	40.97***

Parameter	First stage	Second STAGE	Exclusion criteria	Second stage	Exclusion criteria
Parameter	(1)	SRV (2)	(3)	ROAV (4)	(5)
Intercept	0.006	0.381		0.344
Lag_EngCEO (IV)	0.498*** (8.450)		0.013 (1.299)		0.010 (0.984)
Predicted_EngCEO		0.027* (1.690)		0.022** (2.031)
ROA	0.000 (0.190)	0.204*** (4.099)		0.181*** (3.790)
Leverage	−0.060* (1.850)	0.001 (0.996)		0.002 (1.201)
FirmSize	0.017*** (3.230)	−0.012*** (−6.579)		−0.009*** (−5.454)
SalesGrowth	0.001 (0.670)	−0.008 (−1.142)		−0.006 (−1.015)
Bsize	−0.002 (0.480)	−0.004** (−2.296)		−0.003** (−2.149)
BInd	−0.015 (0.520)	0.000 (0.964)		0.001 (0.877)
CAPEX	−0.001 (0.250)	0.0002 (1.454)		0.000 (1.279)
CEOGEN	−0.003 (0.100)	0.0005 (0.045)		0.001 (0.056)
CEODUA	0.017 (0.890)	−0.015 (−1.560)		−0.014 (−1.431)
BEng	0.181*** (3.600)	0.0024 (0.123)		0.001 (0.105)
Year and industry FE	Yes	Yes		Yes
R-squared	0.441	0.200	0.00	0.210	0.00
F-statistic	40.97***

Note(s): This table reports the results of the IV approach. Column 1 presents the first-stage regression results, estimating EngCEO using Lag_EngCEO as the instrumental variable (IV). Column 2 and 4 reports the second-stage regression estimates, testing the H1 that engineering-trained CEOs influence corporate risk-taking, using the predicted values of the endogenous variable (EngCEO). Column 3 and 5 shows the test of the exclusion criterion for the IV, evaluating whether Lag_EngCEO directly affects corporate risk-taking (SRV) without passing through EngCEO. *, ** and *** indicate statistical significance at 10, 5, and 1% levels, respectively. Variables are as defined in Table 1. t-statistics are reported in parentheses

Source(s): Authors’ own work

[ViewLarge]

Sharing Unavailable