Summary Statistics and Correlations

Table 8 presents the descriptive statistics of all variables in our simultaneous equation systems. Our sample period runs from 1 January to 31 December 2002, giving a total of 252 trading days. Panel A of Table 8 displays the descriptive statistics of the liquidity measures and their control variables. The mean of effective spread (ESP) is 1.6267 cents, ranging between 0.6652 cents and 3.6086 cents; the INF of the effective spread has an average value of 0.6610 cents, ranging between 0.2172 cents and 1.3402 cents;²⁷ the CLP for our sample is approximately US$38.30, with a range between US$4.56 and US$121.73; DOLVOL has a mean of US$69.22 millions, with a sample range between US$4.41 millions to US$610.34 millions; and RETSTD has an average value of about 2.6406%, with the sample ranging between 1.3931% and 7.6836%.

26 The values of ESPi and INFi are so small that the estimated coefficients of the control variables of these liquidity measures are also very small; we therefore multiply the measures by 100, and in consequence, ‘cents’

becomes the unit of measurement.

27 Lin et al. (1995) argue that demanders of immediacy services rarely receive prices which are less favorable than the prevailing quotes on the NYSE.

Table 8 Descriptive statistics of all variables

The sample comprises of a total of 244 S&P 500 index constituent stocks listed on the NYSE between 1 January and 31 December 2002. ESP = the effective spread; INF = the dollar value of the information asymmetry component of the effective spread; CLP = the closing price; DOLVOL = the daily dollar volume; RETSTD = the return standard deviation in the previous year; AFR = the annual basis S&P T&D final ranking; CFR = the composite basis S&P T&D final ranking; SIZE = the firm’s total asset at the end of 2002; AIP = the assets-in-place defined as the book value of fix asset divided by total asset; GI = the governance index; EI = the entrenchment index; LEV = the financial leverage defined as the total debt divided by total equity; D1 = the dummy variable for the Financials group; D2 = the dummy variable for Information Technology group; and SG3

= the past three-year sales growth.

Variables Mean Std. Dev. Minimum Maximum

Panel A: The liquidity measures and their control variables

ESP (cents) 1.6267 0.4601 0.6652 3.6086

INF (cent) 0.6610 0.2143 0.2172 1.3402

CLP 38.2961 18.9560 4.5627 121.7333

DOLVOL (millions) 69.2203 85.2788 4.4059 610.3353

RETSTD (%) 2.6406 0.9383 1.3931 7.6836

Panel B: The internal governance measures and their control variables

AFR 4.8320 0.9731 1 8

CFR 7.5697 0.5204 7 9

SIZE (millions) 42,989 116,237 669 887,515

AIP (%) 30.0785 23.2007 0.0000 93.2126

Panel C: The external governance measures and their control variables

GI 9.8443 2.5126 3 16

EI 2.3934 1.3398 0 5

LEV 2.2464 5.3375 0.0000 52.2460

D1 0.0984 0.2984 0 1

D2 0.1844 0.3886 0 1

SG3 (%) 22.0351 39.6143 -102.8659 177.0896

The descriptive statistics of the S&P T&D final rankings, and their control variables, are presented in Panel B of Table 8. The mean of the annual basis T&D final rankings (AFR) is 4.8320, with a range between 1 and 8, whilst the mean of the composite basis T&D final ranking (CFR) is 7.5698, with a range between 1 and 8. Taking note of the difference between these two rankings, AFR has a lower mean but greater range, whilst CFR has a higher mean but smaller range. This characteristic is consistent with the argument of Patel and Dallas (2002), who suggest that the annual basis rankings, which focus only on a firm’s annual reports, could be regarded as the extent of a firm’s voluntary disclosure; conversely, the composite basis ranking, which includes annual reports, 10-Ks and other proxy statements, may be regarded as regulatory disclosure practices. Thus, as a result of the strict laws on investor protection in the US, as well as the stringent disclosure regulations, the firms reveal consistently higher rankings on a composite basis, along with smaller differences between the firms’ composite basis rankings, than in their annual basis rankings.

Panel C of Table 8 displays the descriptive statistics of the external governance measures and their control variables. The mean of the governance index (GI) is 9.8443, ranging between 3 and 16, whilst the mean of the entrenchment index (EI) is 2.3934, ranging between 0 and 5. Bebchuk et al. (2004) argue that some provisions within 24 IRRC provisions might have some problems on the negative relationship with shareholder value. Therefore, they identify six provisions to develop the entrenchment index in order to simplify the negative relationship between the governance index and the firm value.

The Pearson correlation coefficients of all variables are presented in Table 9. One issue which immediately draws our attention is the fact that INF and ESP have a strong positive correlation, implying that the higher asymmetric information costs induce higher equity spread, given that order processing costs are largely fixed.

Table 9 Pearson correlation coefficients of all variables

The sample comprises of a total of 244 S&P 500 index constituent stocks listed on the NYSE between 1 January and 31 December 2002. Figures in parentheses are p-values. ESP = the effective spread; INF = the dollar value of the information asymmetry component of the effective spread; CLP = the closing price; DOLVOL = the daily dollar volume; RETSTD = the return standard deviation in the previous year; AFR = the annual basis S&P T&D final ranking; CFR = the composite basis S&P T&D final ranking; SIZE = the firm’s total asset at the end of 2002; AIP = the assets-in-place defined as the book value of fix asset divided by total asset; GI = the governance index; EI = the entrenchment index; LEV = the financial leverage defined as the total debt divided by total equity; and SG3 = the past three-year sales growth.

Variables ESP INF CLP DOLVOL RETSTD AFR CFR SIZE AIP GI EI LEV SG3 ESP 1

INF 0.9379 1 (<.0001)

CLP 0.8743 0.788 1

(<.0001) (<.0001)

DOLVOL 0.1435 -0.0903 0.3586 1

(0.0250) (0.1599) (<.0001)

RETSTD -0.3345 -0.383 -0.4781 -0.0445 1

(<.0001) (<.0001) (<.0001) (0.4893)

AFR 0.0047 0.0188 0.0225 0.0115 -0.1544 1

(0.9421) (0.7707) (0.7262) (0.8584) (0.0158)

CFR -0.0797 -0.041 -0.0379 -0.0796 -0.0956 0.3117 1 (0.2147) (0.5242) (0.5562) (0.2156) (0.1364) (<.0001) SIZE 0.0117 -0.1101 0.1886 0.5738 -0.0821 0.0579 0.0539 1 (0.8555) (0.0861) (0.0031) (<.0001) (0.2010) (0.3675) (0.4023) AIP -0.2061 -0.196 -0.1867 -0.1263 0.0786 0.1079 0.1634 -0.2705 1 (0.0012) (0.0021) (0.0034) (0.0487) (0.221) (0.0926) (0.0106) (<.0001)

GI -0.0413 -0.0295 -0.0967 -0.1483 0.0357 0.0179 -0.0672 -0.0894 0.0258 1

(0.5212) (0.6466) (0.1321) (0.0205) (0.5793) (0.7813) (0.2958) (0.1639) (0.6886) EI -0.0226 0.015 -0.0811 -0.1478 0.0437 -0.0059 -0.0277 -0.1614 0.0353 0.7004 1

(0.7253) (0.8158) (0.2071) (0.0209) (0.4970) (0.9270) (0.6671) (0.0116) (0.5835) (<.0001) LEV 0.0393 -0.0016 0.1468 0.1914 -0.0416 0.0452 0.1109 0.6251 -0.1641 -0.1058 -0.1717 1

(0.5410) (0.9807) (0.0218) (0.0027) (0.5181) (0.4820) (0.0840) (<.0001) (0.0102) (0.0991) (0.0072)

SG3 0.0329 0.0483 0.0278 -0.0621 -0.0381 -0.1254 -0.0999 -0.0383 0.0285 -0.0648 -0.0416 -0.0049 1 (0.6094) (0.4527) (0.6658) (0.3339) (0.5536) (0.0503) (0.1195) (0.5514) (0.6574) (0.3134) (0.5178) (0.9390)

Table 9 also presents the Pearson correlation coefficients of the S&P T&D final rankings, and their control variables. SIZE and AIP reveal positive correlations to the T&D final rankings, although the positive correlation between SIZE and these two rankings is insignificant. RETSTD has a significantly negative correlation with the annual basis T&D final ranking, but has an insignificantly negative correlation with the composite basis T&D ranking. This finding is consistent with the results of the first-stage regression. For the Pearson correlation coefficients of the governance index and their control variables, LEV and SG3 reveal negative correlations to GI and EI, but only the negative correlation between LEV and these two indexes is significant.