Skip to main content
Frontiers of Business Research in China

Table 7 Second-stage results of Heckman two-stage estimation

From: Corporate spinoffs and executive compensation

 

Coefficient

z-statistic

P > |z|

Panel A – First Stage PROBIT Regression Results

Intercept

0.246

0.138

0.710

ROA

1.912

0.838

0.360

Leverage

0.818

0.973

0.324

Size

−0.089

1.454

0.228

Investment Level

−0.577

1.716

0.190

Tobin’s Q

−0.032

0.205

0.651

ROA t-2

−3.863

2.954

0.086*

ROA t-3

1.266

0.496

0.481

Std. Dev. of ROA

1.078

0.093

0.761

Number of Segments

0.188

3.627

0.057*

Std. Dev. of Returns

1.387

0.016

0.898

N

  

118

Wald Statistic

  

10.679

Panel B – Second Stage Regression Results

 

Dependent Variable: The log difference in Total Compensation

 

(Year − 1, + 1)

(Year− 1, + 1 to + 3)

 

Intercept (b1)

0.321

0.483**

 

(1.10)

(1.99)

 

Rtn t (b2)

−0.228

−0.19

 

(−0.75)

(−0.65)

 

Rtn t-1 (b3)

−0.152

−0.151

 

(−0.46)

(−0.47)

 

SpD× Rtn t (b4)

0.712*

0.535

 

(1.8)

(1.58)

 

SpD ×Rtn t-1 (b5)

0.921*

0.669*

 

(1.84)

(1.75)

 

SpD (b6)

−0.26

−0.1

 

(−1.36)

(−0.59)

 

Lambda (b7)

0.559

0.306

 

(1.17)

(0.92)

 

N

114

226

 

R2x

0.15

0.095

 
  1. This table reports results on the Heckman two-stage least square regressions. The first stage estimation employs the Probit regression by pooling the spinoff and the matched sample. An indicator variable, with 1 denotes the spinoff sample and 0 denotes the matched sample is used as the dependent variable. For independent variables, ROA is defined as income before extraordinary items available for common shareholders divided by total assets. Size is measured by total assets. Leverage is defined as total debt divided by total shareholders’ equity. Investment level is computed by dividing capital expenditures by total sales. Tobin’s Q is defined as the market value of total assets divided by the book value of total assets, where the market value is computed by book value of total assets plus market value of common equity minus book value of common equity and deferred taxes. Standard deviation of returns is the standard deviation of one year of monthly stock returns. All variables for the first-stage regression are measured at Year − 1 in order to represent the parent firm fundamentals before spinoff and the data are obtained from Compustat or the CRSP database. The second stage panel regression estimation employs the parent sample for the pre- and post-spinoff period. The dependent variable for the second stage is total CEO compensation. The lambda computed from the first stage regression is included in the second stage. Other independent variables are previously defined. T-statistics, which are based on White standard errors robust to within firms’ cluster correlation, are reported in parenthesis. ***, ** and * denote significance at the 1%, 5%, and 10% level respectively. Variation in sample size is due to data availability
Back to article page