Online Appendix 1
Table O1: Determinants of CMO Compensation: Selection based on both number of other firms in industry that have CMOs and number of other firms in industry with MBA educated executives The table reports two-stage Heckman sample-selection regressions for Total CMO Compensation, CMO Market Compensation, or CMO Delta. The first stage (Selection) is a probit equation where the dependent variable equals one if the firm has an observable CMO, and zero otherwise. The second stage (Outcome) is either an or system regression where the dependent variable is the compensation variable. Industry use of CMO is the proportion of other firms in that industry that have observable CMOs. Proportion of industry with MBA executives is the proportion of other firms in that industry with a top executive that has an MBA. Lambda is the inverse Mills ratio estimated from the Selection equation. The other variables are as defined in Table 1. t-statistics, based on robust standard errors that are clustered by firm, are reported in parentheses. Year fixed effects are included in all specifications. ***, **, and * represent significance at the 1%, 5% and 10% levels respectively. Selection Selection Selection Advertising 0.9909 *** 1.2768 ** 2.7695 *** 0.0322 ** 0.0984 ** 0.0014 1.1538 ** 1.5794 * 2.2661 *** (3.26) (2.47) (7.01) (2.42) (2.00) ( 0.03) (2.54) (1.84) (5.08) R&D 0.4239 *** 0.3503 *** 0.5888 *** 0.2411 *** 0.2360 *** 0.4868 *** 0.4315 *** 0.3077 * 0.5596 *** (5.03) (2.77) (5.35) (8.62) (4.54) (4.37) (3.33) (1.78) (4.42) Industry Concentration 0.2043 ** -0.2326 * 0.2795 ** 0.0935 *** -0.0807 * 0.3312 ** 0.2342-0.1788 0.3167 ** ( 2.06) (-1.89) (2.17) ( 2.84) (-1.74) (2.58) ( 1.64) (-1.60) (2.18) CEO Value of Dependent 0.2281 *** 0.2249 *** 0.0027 0.4395 *** 0.4525 *** 0.1098 *** 0.4675 *** 0.4411 *** 0.0039 Variable (32.91) (5.07) ( 0.29) (49.56) (25.50) (3.17) (67.62) (22.88) ( 0.60) Size 0.3298 *** 0.2996 *** 0.0609 *** 0.0366 *** 0.0324 *** 0.0673 *** 0.1665 *** 0.1487 *** 0.0631 *** (54.61) (10.28) ( 8.06) (19.84) (9.99) ( 9.95) (19.26) (5.77) ( 8.02) Market-to-Book 0.0129 ** 0.0010 0.0299 *** 0.0026 0.0069* 0.0352 *** 0.0350 *** 0.0231 0.0295 *** ( 2.01) (0.10) (3.43) (1.23) (1.90) (4.05) (3.44) (1.31) (2.87) CAPEX 0.2228 0.1866 0.8605 *** 0.2264 *** 0.1966 *** 0.7912 *** 0.2456 0.1956 0.7473 *** (1.54) (0.85) (4.83) (4.72) (2.93) (4.45) (1.11) (0.79) (3.70) Leverage 0.0664-0.0133 0.4943 *** 0.0056 0.0170 0.5162 *** 0.1307-0.2344 * 0.5509 *** (1.00) (-0.13) ( 6.30) (0.26) (0.49) ( 6.58) ( 1.30) (-1.67) ( 6.21) Std. Ret 2.6586 *** 1.9050 *** 0.8056 *** 0.5750 *** 0.4318 *** 0.7162 *** 1.6923 *** 0.9631 * 0.0399 (12.95) (6.18) (2.95) (8.46) (3.90) (2.62) (5.38) (1.72) (0.13) Lagged Compensation Variable 0.0435 0.0237 0.0893 (0.95) (0.81) (1.63) Proportion of industry with 1.4503 *** 1.5323 *** 1.7602 *** MBA executives (4.94) (5.24) (5.41) Industry use of CMO 1.0350 *** 0.9930 *** 0.6186 * (3.57) (3.44) (1.94) Lambda 0.2427 *** - 0.0427 *** -0.0258 * 0.1775 *** -0.1280 ** ( 7.82) (-3.26) ( 4.25) (-1.92) ( 3.76) (-2.03) Constant 2.6105 *** 2.7234 *** 1.0281 *** 0.0842 *** -0.1058 *** 1.0251 *** 0.7867 *** -0.5538 *** 0.9854 *** (36.20) (10.11) ( 11.98) ( 3.93) (-3.85) ( 13.18) ( 7.88) (-3.17) ( 11.11) Observations 23,454 23,454 23,454 23,454 23,454 23,454 19,051 19,051 19,051 F-statistic 413.85 *** 127.60 *** 220.11 *** 94.90 *** 384.82 *** 109.80 *** Adj. R 2 /Pseudo-R 2 0.595 0.638 0.099 0.438 0.483 0.098 0.650 0.709 0.100 2
Table O2 Difference-in-difference analysis of Change in CFO compensation after initiation of advertising In this table, we report the difference in changes in CFO compensation between advertising and nonadvertising firms following the initiation of advertising. Treated firms are firms that switch from not advertising, to advertising, and maintain this status for at least three fiscal years after the inception of advertising. Treated firms are matched to control firms that do not advertise over the same period but are similar with respect to R&D spending, industry concentration, firm size, market-to-book, capital expenditure, leverage, standard deviation returns and total CFO and CEO compensation in the year in which the treated firm initiated advertising. We report the difference in changes in compensation between treated and control firms between the year the treated firm initiated advertising (t) and t + n, where n is 1, 2 or 3. p-values (reported in brackets) testing whether or not the difference-in-difference is greater than zero are reported below each estimate. The variables are as defined in Table 1. None of the values reported are significant at either the 1%, 5%, or 10% level. Panel A: Treated firm matched to one control firm t + 1 t + 2 t + 3 Total CFO Compensation 5.03% 0.84% 10.27% [0.384] [0.892] [0.224] Market CFO Compensation 7.07% -3.07% 3.83% [0.157] [0.593] [0.499] CFO Delta 8.68% -26.40% -8.09% [0.426] [0.200] [0.605] Panel B: Treated firm matched to two control firms t + 1 t + 2 t + 3 Total CFO Compensation 5.78% 1.31% 6.94% [0.254] [0.812] [0.251] Market CFO Compensation 6.91% 0.01% 2.36% [0.113] [0.999] [0.650] CFO Delta 2.24% -27.35% 10.33% [0.813] [0.112] [0.493] 3
Table O3 Deviations from expected CMO total compensation and firm performance The table reports dynamic panel regressions of various measures of firm performance on deviations from expected compensation CMO total compensation, and other firm characteristics. In columns (1) and (2), the dependent variable is return on assets (ROA) which is measured as the operating income before depreciation, scaled by total assets. In columns (3) and (4), the dependent variable is Earnings surprise which is the difference between the firm s actual earnings per share and median consensus analysts forecast, scaled by the price at the end of the fiscal year. In columns (5) and (6), the dependent variable is Annual Stock Return which is the one year holding period stock return. Deviation is the absolute deviation from expected CMO total compensation. Deviation + equals the absolute deviation for positive values of deviation, and is zero otherwise. Deviation equals the absolute deviation for negative values of deviation, and is zero otherwise. The other variables are as defined in Table 1. t-statistics, based on robust standard errors that are clustered by firm, are reported in parentheses. Year fixed effects are included in all specifications. ***, **, and * represent significance at the 1%, 5% and 10% levels respectively. ROA Earnings Surprise Stock Return Deviation 0.0081 *** 0.0085 *** 0.0083 *** 0.0073 *** 0.0696 *** 0.0829 ** ( 3.57) ( 2.81) ( 3.78) ( 2.80) ( 2.64) ( 2.16) Size 0.0026 *** 0.0055 *** 0.0084 *** 0.0106 *** 0.0320 *** 0.0302 ** (4.50) (5.61) (11.24) (8.49) (6.11) (2.30) Market-to-Book 0.0126 *** 0.0163 *** 0.0009* 0.0006 0.1180 *** 0.0986 *** (12.36) (8.01) (1.91) (0.96) (14.26) (3.80) CAPEX 0.0865 *** 0.1546 *** 0.0571 *** 0.0892 *** 0.6664 *** 1.4247 *** (3.98) (4.03) (3.93) (3.28) ( 4.55) ( 3.42) Leverage 0.0250 *** 0.0653 *** 0.0743 *** 0.0907 *** 0.2709 *** 0.0438 ( 3.51) ( 5.32) ( 5.78) ( 4.50) (2.87) ( 0.47) Std. Ret 0.0971 *** 0.0394 0.0816 *** 0.0886 *** 2.0196 *** 0.6560 ( 3.68) ( 1.10) (3.42) (2.83) (6.44) (1.22) Lagged Performance 0.7276 *** 0.4759 *** 0.3012 *** 0.0003 0.0804 *** 0.5938 (50.91) (8.12) (8.49) ( 0.00) ( 4.52) (1.45) Constant 0.0157 *** 0.0103 0.0841 *** 0.1031 *** 0.3518 *** 0.3013 * (2.68) (0.94) ( 5.36) ( 5.95) ( 3.53) ( 1.82) Observations 5,739 5,739 5,798 5,798 5,747 5,747 F-statistic 544.96 78.31 55.59 11.49 56.90 53.59 Adj. R 2 0.711 0.682 0.197 0.131 0.202 0.307 4
Table O4 Deviations from expected CMO market compensation and firm performance The table reports dynamic panel regressions of various measures of firm performance on deviations from expected CMO market compensation, and other firm characteristics. In columns (1) and (2), the dependent variable is return on assets (ROA) which is measured as the operating income before depreciation, scaled by total assets. In columns (3) and (4), the dependent variable is Earnings surprise which is the difference between the firm s actual earnings per share and median consensus analysts forecast, scaled by the price at the end of the fiscal year. In columns (5) and (6), the dependent variable is Annual Stock Return which is the one year holding period stock return. Deviation is the absolute deviation from expected CMO market compensation. Deviation + equals the absolute deviation for positive values of deviation, and is zero otherwise. Deviation equals the absolute deviation for negative values of deviation, and is zero otherwise. The other variables are as defined in Table 1. t-statistics, based on robust standard errors that are clustered by firm, are reported in parentheses. Year fixed effects are included in all specifications. ***, **, and * represent significance at the 1%, 5% and 10% levels respectively. ROA Earnings Surprise Stock Return Deviation 0.0198 *** 0.0177 ** 0.0146 ** 0.0171 ** 0.2943 *** 0.2243 ** ( 3.01) ( 2.57) ( 2.21) ( 2.19) ( 4.43) ( 2.21) Size 0.0023 *** 0.0053 *** 0.0081 *** 0.0104 *** 0.0299 *** 0.0272 ** (4.11) (5.43) (11.04) (8.33) (5.57) (1.96) Market-to-Book 0.0125 *** 0.0163 *** 0.0009 * 0.0006 0.1183 *** 0.0980 *** (12.30) (8.01) (1.83) (0.96) (14.35) (3.69) CAPEX 0.0858 *** 0.1564 *** 0.0577 *** 0.0897 *** 0.6726 *** 1.4305 *** (3.94) (4.08) (3.98) (3.33) ( 4.60) ( 3.35) Leverage 0.0242 *** 0.0647 *** 0.0737 *** 0.0900 *** 0.2750 *** 0.0345 ( 3.40) ( 5.26) ( 5.72) ( 4.47) (2.92) ( 0.36) Std. Ret 0.0984 *** 0.0417 0.0784 *** 0.0876 *** 2.0381 *** 0.6471 ( 3.70) ( 1.18) (3.34) (2.83) (6.45) (1.18) Lagged Performance 0.7292 *** 0.4735 *** 0.3012 *** 0.0015 0.0805 *** 0.6047 (51.13) (8.02) (8.46) ( 0.01) ( 4.49) (1.43) Constant 0.0171 *** 0.0120 0.0827 *** 0.1013 *** 0.3198 *** 0.2861 * (2.90) (1.08) ( 5.24) ( 5.84) ( 3.23) ( 1.66) Observations 5,739 5,739 5,798 5,798 5,747 5,747 F-statistic 544.29 *** 74.90 *** 55.12 *** 11.47 *** 57.34 *** 53.89 *** Adj. R 2 0.711 0.681 0.195 0.129 0.203 0.308 5
Table O5 Deviations from CMO Delta and Firm Performance The table reports dynamic panel regressions of various measures of firm performance on deviations from expected CMO Delta, and other firm characteristics. In columns (1) and (2), the dependent variable is return on assets (ROA) which is measured as the operating income before depreciation, scaled by total assets. In columns (3) and (4), the dependent variable is Earnings surprise which is the difference between the firm s actual earnings per share and median consensus analysts forecast, scaled by the price at the end of the fiscal year. In columns (5) and (6), the dependent variable is Annual Stock Return which is the one year holding period stock return. Deviation is the absolute deviation from expected CMO Delta. Deviation + equals the absolute deviation for positive values of deviation, and is zero otherwise. Deviation equals the absolute deviation for negative values of deviation, and is zero otherwise. The other variables are as defined in Table 1. t-statistics, based on robust standard errors that are clustered by firm, are reported in parentheses. Year fixed effects are included in all specifications. ***, **, and * represent significance at the 1%, 5% and 10% levels respectively. ROA Earnings Surprise Stock Return Deviation 0.0039 ** 0.0033 * 0.0046 *** 0.0061 *** 0.0328 ** 0.0133 * ( 2.40) ( 1.79) ( 3.13) ( 3.57) ( 2.53) ( 1.92) Size 0.0023 *** 0.0043 *** 0.0087 *** 0.0109 *** 0.0285 *** 0.0428 *** (3.56) (4.40) (9.76) (7.96) (5.14) (5.29) Market to-book 0.0145 *** 0.0209 *** 0.0020 *** 0.0022 *** 0.1058 *** 0.1317 *** (10.59) (8.76) (3.06) (2.71) (10.71) (8.36) CAPEX 0.0711 *** 0.1450 *** 0.0526 *** 0.1002 *** 0.3386 ** 0.3452 * (2.72) (3.31) (3.13) (3.74) ( 2.39) ( 1.72) Leverage 0.0226 *** 0.0565 *** 0.0687 *** 0.0828 *** 0.0015 0.1090 ( 2.71) ( 4.35) ( 4.65) ( 4.54) ( 0.02) ( 1.48) Std. Ret 0.1259 *** 0.1031 *** 0.0716 ** 0.0756 ** 0.8421 *** 0.5879 ( 4.21) ( 2.75) (2.50) (2.25) (2.79) (1.61) Lagged Performance 0.7438 *** 0.4915 *** 0.3161 *** 0.0573 0.1408 *** 0.2277 (40.66) (7.80) (6.69) ( 0.34) ( 6.58) ( 1.26) Constant 0.0132 ** 0.0021 0.0950 *** 0.1055 *** 0.2547 ** 0.3973 *** (1.97) (0.18) ( 4.53) ( 7.67) ( 2.01) ( 4.76) Observations 4,300 4,300 4,341 4,341 4,318 4,318 F-statistic 426.13 *** 75.09 *** 40.89 *** 9.13 *** 37.01 *** 43.08 *** Adj. R 2 0.720 0.695 0.193 0.110 0.178 0.282 6
Table O6 Variance Inflation Factors The table reports the variance inflation factors for the explanatory variables in the regressions reported in Tables 3, 4 and 5. For each compensation variable, we report the variance inflation factor for each variable estimated following the regression of the determinants of each compensation variable Total Compensation Market Compensation Compensation Delta CMO Tenure 4.84 4.85 4.63 New CMO 4.78 4.79 4.57 Std. Ret 2.23 2.25 2.17 Size 1.74 1.46 1.43 Market-to-Book 1.52 1.51 1.61 Leverage 1.36 1.36 1.41 R&D 1.33 1.36 1.35 CEO Value of Compensation Variable 1.47 1.27 1.29 CAPEX 1.11 1.12 1.12 Advertising 1.03 1.03 1.04 Industry Concentration 1.03 1.03 1.02 7
Table O7 Determinants of compensation: alternative definition of industry concentration The table reports regressions where the dependent variable is total compensation, market compensation, or compensation Delta, of the CMO. Industry concentration is measured at the three-digit SIC code level. The other explanatory variables are as defined in Table 1. t-statistics, based on robust standard errors that are clustered by firm, are reported in parentheses. Year fixed effects are included in all specifications. ***, **, and * represent significance at the 1%, 5% and 10% levels respectively. Total Compensation Market Compensation Delta Advertising 1.6698 *** 1.6396 *** 0.1177 * 0.2160 ** 1.8054 *** 2.3809 ** (6.41) (3.57) (1.83) (2.02) (4.95) (2.30) R&D 0.5234 *** 0.5630 *** 0.2115 *** 0.2648 *** 0.5707 *** 0.3930 ** (6.14) (4.55) (6.92) (5.68) (5.03) (2.00) Industry Concentration 0.1280 * 0.1222 * 0.0733 *** 0.0776 * 0.1925 * 0.1759 ** ( 1.84) ( 1.92) ( 2.91) ( 1.81) ( 1.91) ( 2.05) CEO Value of Dependent 0.2357 *** 0.2262 *** 0.4579 *** 0.4531 *** 0.4407 *** 0.4223 *** Variable (8.44) (4.01) (44.61) (26.54) (68.79) (25.25) Size 0.3162 *** 0.2773 *** 0.0389 *** 0.0388 *** 0.1786 *** 0.1719 *** (20.89) (8.96) (20.28) (11.78) (23.58) (6.96) Market-to-Book 0.0209 *** 0.0128 0.0015 0.0006 0.0225 *** 0.0042 ( 2.98) ( 1.38) ( 0.66) ( 0.20) (2.59) (0.20) CAPEX 0.1800 0.1675 0.1752 *** 0.1487 *** 0.3711 ** 0.3400 (1.51) (1.00) (4.39) (2.61) (2.04) (1.38) Leverage 0.1928 *** 0.2751 *** 0.0301 0.0563 * 0.2636 *** 0.2801 ** ( 3.43) ( 3.00) ( 1.47) ( 1.74) ( 3.01) ( 2.09) Std. Ret 3.0945 *** 2.0140 *** 0.7480 *** 0.4768 *** 2.2229 *** 1.9938 *** (13.42) (6.78) (10.57) (4.76) (8.00) (4.01) CMO Tenure 0.0026 0.0279 0.0156 ** 0.0265 *** 0.0310 0.0036 (0.12) ( 1.11) ( 2.23) ( 2.78) (0.97) ( 0.09) New CMO 0.0004 0.0005 0.0093 0.0025 0.0340 0.0107 ( 0.01) ( 0.02) (0.99) ( 0.22) (0.80) (0.21) Lagged Compensation 0.0990 *** 0.0272 0.0651 (2.65) (1.02) (1.24) Constant 2.0911 *** 2.1181 *** 0.2075 *** 0.1256 ** 1.2481 *** 0.9479 *** (17.40) (8.91) ( 5.05) ( 2.55) ( 4.92) ( 5.93) Observations 8,806 6,191 8,797 6,184 6,330 3,781 F-statistics 105.61 *** 135.03 *** 235.14 *** 111.02 *** 311.13 *** 96.96 *** Adj. R 2 /Pseudo- R 2 0.602 0.648 0.468 0.520 0.618 0.668 8