This appendix explains the method that I used to select 100 listed companies randomly for both China and Germany in section 3.2.

(1)

APPENDIX

This appendix explains the method that I used to select 100 listed companies randomly for both China and Germany in section 3.2.

As mentioned in section 3.1, the population of China listed companies is 1453 and the population of Germany listed companies is 648. Now the task is to use Excel Visual Basic for Application (VBA) programming to select 100 listed companies randomly for both China and Germany respectively.

In order to simplify the computer programming, first use numbers 1 to 1453 instead of 1453 different China listed company codes ¹⁵ , for example, 1 represents company code 600000.SS, 2 represents company code 600001.SS,…, 836 represents company code 000001.SZ and 837 represents company code 000002.SZ, etc.. Then open a new Excel file and input numbers (integers) 1 to 1453 in cells C1 to C1453.

The VBA programming is as follows: Click Excel sheet 1 / Tools / Macro / Visual Basic Editor. After choosing Sheet 1 (Sheet 1) from VBAProject area, type the following program in the Visual Basic Editor area:

Sub Randx()

Dim xx(1 To 1453) As Integer For t = 1 To 100

rerand:

x = Int(Rnd() * 1453 + 1) If xx(x) > 0 Then GoTo rerand r = r + 1

Cells(r, 1) = x

xx(x) = r

(2)

Next End Sub

To run this program, first go back to Excel sheet 1 and click Tools / Macro / Macros.

Then select Sheet1.Randx and click Run. After this, 100 randomly selected China listed companies will be showed on cells A1 to A100 of Excel sheet 1.

Again, for the Germany case I can get 100 randomly selected listed companies by

using the same process as I did for China case, substituting integer 1453 for 648 in

computer programming step.

(3)

TABLES

Table 1

Summary statistics for China sample

FI POLS LVR NS DY RATIO_BTM LOG_TA DUM

Mean 3.578100 4.141030 47.41371 64305.22 1.063420 0.133499 8.584227 0.360000 Median 1.425000 2.420000 48.40021 40321.00 0.800000 0.115513 8.433819 0.000000 Maximum 33.89000 30.00000 92.15475 846127.0 6.400000 0.390625 15.48797 1.000000 Minimum 0.000000 0.040000 6.338499 5797.000 0.005000 0.020072 5.893654 0.000000 Std. Dev. 5.527482 5.369389 17.02262 105307.3 0.945828 0.075234 1.489956 0.482418 Skewness 2.699391 2.960266 -0.178068 5.619927 2.724721 0.846428 1.416449 0.583333 Kurtosis 12.09074 12.09078 2.741495 38.61532 14.17096 3.462911 7.133708 1.340278

Jarque-Bera 465.7853 490.3954 0.806905 5811.606 643.6952 12.83354 104.6369 17.14912 Probability 0.000000 0.000000 0.668010 0.000000 0.000000 0.001634 0.000000 0.000189

Sum 357.8100 414.1030 4741.371 6430522. 106.3420 13.34990 858.4227 36.00000 Sum Sq. Dev. 3024.753 2854.203 28687.21 1.10E+12 88.56450 0.560352 219.7769 23.04000

Observations 100 100 100 100 100 100 100 100

Note: FI = foreign investment

POLS = percentage owned by largest shareholder LVR = leverage

NS = number of shareholders DY = dividend yield

RATIO_BTM = book to market ratio

LOG_TA = logarithm of total assets

DUM = dummy variable

(4)

Table 2

Summary statistics for Germany sample

FI POLS LVR NS DY RATIO_BTM LOG_TA DUM

Mean 16.55530 16.27870 0.554239 25.44000 0.014517 0.704056 6.392514 0.510000 Median 11.12500 12.20000 0.602939 22.50000 0.011000 0.551426 5.964339 1.000000 Maximum 86.94000 50.00000 0.924317 74.00000 0.228000 5.835809 12.29086 1.000000 Minimum 0.000000 0.250000 0.120241 2.000000 0.000000 0.120476 2.941804 0.000000 Std. Dev. 16.23096 11.14199 0.187390 18.19497 0.025359 0.718065 2.531791 0.502418 Skewness 1.525182 1.030771 -0.307466 0.566812 6.116558 4.809470 0.692971 -0.040008 Kurtosis 6.110231 3.479437 2.013737 2.391738 51.44585 31.53879 2.518338 1.001601

Jarque-Bera 79.07606 18.66591 5.628569 6.896196 10402.71 3779.110 8.970145 16.66668 Probability 0.000000 0.000088 0.059948 0.031806 0.000000 0.000000 0.011276 0.000240

Sum 1655.530 1627.870 55.42388 2544.000 1.451700 70.40562 639.2514 51.00000 Sum Sq. Dev. 26080.95 12290.26 3.476390 32774.64 0.063665 51.04606 634.5866 24.99000

Observations 100 100 100 100 100 100 100 100

Note: FI = foreign investment

POLS = percentage owned by largest shareholder LVR = leverage

NS = number of shareholders DY = dividend yield

RATIO_BTM = book to market ratio

LOG_TA = logarithm of total assets

DUM = dummy variable

(5)

Table 3

Test the correlations among the independent variables and between dependent variable and independent variables for China data

FI POLS LVR NS DY RATIO_BTM LOG_TA DUM

FI 1.000000

POLS 0.582173 1.000000

LVR 0.118809 0.122226 1.000000

NS -0.047166 -0.028596 0.198816 1.000000

DY 0.051471 -0.011408 -0.101751 0.009496 1.000000

RATIO_BTM -0.149864 -0.027086 0.211739 0.227996 -0.085284 1.000000

LOG_TA 0.186527 0.238554 0.404694 0.669976 0.136342 0.074818 1.000000

DUM -0.068683 0.056477 -0.064771 -0.064684 -0.056808 -0.083667 -0.275462 1.000000

Note: FI = foreign investment

POLS = percentage owned by largest shareholder LVR = leverage

NS = number of shareholders DY = dividend yield

RATIO_BTM = book to market ratio

LOG_TA = logarithm of total assets

DUM = dummy variable

(6)

Table 4

Test the correlations among the independent variables and between dependent variable and independent variables for Germany data

FI POLS LVR NS DY

RATIO_BT

M LOG_TA DUM

FI 1.000000

POLS 0.092675 1.000000

LVR 0.314424 0.047218 1.000000

NS 0.258490 -0.108218 0.306631 1.000000

DY -0.128415 -0.164775 0.035841 0.083851 1.000000

RATIO_BTM -0.132214 -0.046223 -0.157174 -0.307151 0.013583 1.000000 LOG_TA 0.111397 -0.034979 0.614650 0.674674 0.151972 -0.111279 1.000000

DUM -0.074890 0.091784 -0.097694 0.076861 -0.131421 0.109550 0.006353 1.000000

Note: FI = foreign investment

POLS = percentage owned by largest shareholder LVR = leverage

NS = number of shareholders DY = dividend yield

RATIO_BTM = book to market ratio

LOG_TA = logarithm of total assets

DUM = dummy variable

(7)

Table 5

The histogram of the ordinary least squares residuals of China data

0 4 8 12 16 20

-15 -10 -5 0 5 10 15

Series: RESID Sample 1 100 Observations 100 Mean 4.49e-16 Median -0.477791 Maximum 16.26602 Minimum -16.61916 Std. Dev. 4.360786 Skewness 0.194772 Kurtosis 6.694392 Jarque-Bera 57.50116 Probability 0.000000

Table 6

The histogram of the ordinary least square residuals of Germany data

0 2 4 6 8 10 12

-25.0 -12.5 0.0 12.5 25.0 37.5 50.0

Series: RESID

Sample 1 100

Observations 100

Mean -3.11e-15

Median -2.807707

Maximum 56.51111

Minimum -24.35932

Std. Dev. 14.19099

Skewness 1.089108

Kurtosis 4.689147

Jarque-Bera 31.65769

Probability 0.000000

(8)

Table 7

The censored regression for China data

Dependent Variable: FI

Method: ML - Censored Logistic (Quadratic hill climbing) Date: 10/27/07 Time: 08:41

Sample: 1 100

Included observations: 100 Left censoring (value) at zero

Convergence achieved after 8 iterations

Covariance matrix computed using second derivatives

Coefficient Std. Error z-Statistic Prob.

C -4.292213 4.164553 -1.030654 0.3027 POLS 0.756848 0.150998 5.012290 0.0000 LVR 0.018049 0.030970 0.582788 0.5600 NS -3.30E-06 6.09E-06 -0.541836 0.5879 DY 0.484096 0.478014 1.012722 0.3112 RATIO_BTM -3.920204 6.381099 -0.614346 0.5390 LOG_TA 0.387027 0.545025 0.710109 0.4776 DUM 0.299690 1.016738 0.294757 0.7682

Error Distribution

SCALE:C(9) 2.609771 0.265799 9.818605 0.0000

R-squared 0.317474 Mean dependent var 3.578100 Adjusted R-squared 0.257472 S.D. dependent var 5.527482 S.E. of regression 4.763036 Akaike info criterion 5.131043 Sum squared resid 2064.473 Schwarz criterion 5.365508 Log likelihood -247.5521 Hannan-Quinn criter. 5.225935 Avg. log likelihood -2.475521

Left censored obs 24 Right censored obs 0

Uncensored obs 76 Total obs 100

(9)

Table 8

The censored regression for Germany data

Dependent Variable: FI

Method: ML - Censored Logistic (Quadratic hill climbing) Date: 10/27/07 Time: 09:33

Sample: 1 100

Included observations: 100 Left censoring (value) at zero

Convergence achieved after 5 iterations

Covariance matrix computed using second derivatives

Coefficient Std. Error z-Statistic Prob.

C 2.168475 5.862338 0.369899 0.7115

POLS 0.064525 0.132514 0.486927 0.6263 LVR 38.83551 10.03487 3.870058 0.0001 NS 0.415567 0.132346 3.140003 0.0017 DY -240.3425 118.3142 -2.031392 0.0422 RATIO_BTM -2.626458 3.473354 -0.756174 0.4495 LOG_TA -2.410009 1.032566 -2.334000 0.0196 DUM -1.264543 2.907342 -0.434948 0.6636

Error Distribution

SCALE:C(9) 8.131305 0.731227 11.12009 0.0000

R-squared 0.259125 Mean dependent var 16.55530 Adjusted R-squared 0.193993 S.D. dependent var 16.23096 S.E. of regression 14.57181 Akaike info criterion 7.665650 Sum squared resid 19322.72 Schwarz criterion 7.900116 Log likelihood -374.2825 Hannan-Quinn criter. 7.760543 Avg. log likelihood -3.742825

Left censored obs 11 Right censored obs 0

Uncensored obs 89 Total obs 100

(10)

Table 9

Test the correlations among the independent variables and between dependent variable and independent variables for two-country joint data

FI POLS LVR NS DY

RATIO_BT

M LOG_TA DUM DUM_2

FI 1.000000

POLS 0.386915 1.000000

LVR -0.405115 -0.489363 1.000000

NS -0.200764 -0.237005 0.437364 1.000000

DY -0.283926 -0.359760 0.514666 0.253120 1.000000

RATIO_BTM 0.132507 0.249617 -0.427931 -0.175909 -0.308868 1.000000

LOG_TA -0.127455 -0.249582 0.501981 0.461921 0.340303 -0.299809 1.000000

DUM 0.013796 0.148665 -0.155236 -0.100824 -0.126134 0.136609 -0.151925 1.000000

DUM_2 -0.473696 -0.572040 0.890404 0.397963 0.618850 -0.489670 0.468438 -0.151284 1.000000

Note: FI = foreign investment

POLS = percentage owned by largest shareholder LVR = leverage

NS = number of shareholders DY = dividend yield

RATIO_BTM = book to market ratio

LOG_TA = logarithm of total assets

DUM = dummy variable

(11)

Table 10

The histogram of the ordinary least square residuals of two-country joint data

0 10 20 30 40 50

0 25 50

Series: RESID

Sample 1 200

Observations 200

Mean 1.63e-15

Median -1.601244

Maximum 66.28587

Minimum -21.27598

Std. Dev. 11.73241

Skewness 1.812976

Kurtosis 9.311345

Jarque-Bera 441.5050

Probability 0.000000

(12)

Table 11

The censored regression for two-country joint data

Dependent Variable: FI

Method: ML - Censored Logistic (Quadratic hill climbing) Date: 10/27/07 Time: 09:38

Sample: 1 200

Included observations: 200 Left censoring (value) at zero

Convergence achieved after 6 iterations

Covariance matrix computed using second derivatives

Coefficient Std. Error z-Statistic Prob.

C 9.866733 4.183184 2.358666 0.0183

POLS 0.157523 0.103304 1.524853 0.1273 LVR 0.031823 0.067194 0.473606 0.6358 NS -8.98E-06 1.03E-05 -0.874466 0.3819 DY 0.191834 1.121411 0.171065 0.8642 RATIO_BTM -7.351249 2.958074 -2.485147 0.0129 LOG_TA 0.932811 0.457660 2.038220 0.0415 DUM 0.184041 1.650150 0.111530 0.9112 DUM_2 -17.00242 4.250672 -3.999936 0.0001

Error Distribution

SCALE:C(10) 6.620398 0.446545 14.82581 0.0000

R-squared 0.266151 Mean dependent var 10.06670 Adjusted R-squared 0.231390 S.D. dependent var 13.73220 S.E. of regression 12.03907 Akaike info criterion 6.848012 Sum squared resid 27538.47 Schwarz criterion 7.012928 Log likelihood -674.8012 Hannan-Quinn criter. 6.914751 Avg. log likelihood -3.374006

Left censored obs 35 Right censored obs 0

Uncensored obs 165 Total obs 200