Equity-incentive compensation and payout policy in Europe

Journal of Corporate Finance 30 (2015) 85–97

Contents lists available at ScienceDirect

Journal of Corporate Finance journal homepage: www.elsevier.com/locate/jcorpfin

Equity-incentive compensation and payout policy in Europe Natasha Burns a,⁎, Brian C. McTier b,1, Kristina Minnick c,2 a b c

The University of Texas at San Antonio, United States Washington State University, United States Bentley University, United States

a r t i c l e

i n f o

Article history: Received 15 November 2013 Received in revised form 28 October 2014 Accepted 29 October 2014 Available online 4 November 2014 JEL classification: G35 G30 Keywords: Executive compensation Payout policy Agency costs Corporate governance

a b s t r a c t We examine the effects of executive compensation and investor protection on payout policy in Europe. We find a negative (positive) relationship between both option and restricted stock compensation and dividends (repurchases). However, when the incentive compensation is dividend protected, dividend payouts increase. Firms in weak investor protection countries pay higher dividends consistent with maintaining a reputation for distributing excess free cash flows. However, growth firms in weak investor protection countries reduce dividends (increase repurchases) in relation to increases in equity-incentive compensation. Our results are consistent with growth firms in weak investor protection countries using equity incentives as a substitute for dividends to reduce agency costs. © 2014 Elsevier B.V. All rights reserved.

1. Introduction Existing literature shows that equity-incentive compensation affects the amount and form of payouts in the United States (see for example Fenn and Liang, 2001; Aboody and Kasznick, 2008; Brown et al., 2007). Using a sample of European firms, we examine whether this relation holds outside of the U.S. In particular, Eije and Megginson (2008) show that trends in payout policy in Europe are similar to U.S. trends. Additionally, while the use of equity-based compensation has traditionally been a U.S. phenomenon, European firms have recently increased their use of incentive compensation.3 In this study, we explore how repurchases and dividends in Europe are related to these changes as well as country level investor protection, the use of dividend protection, and individual firm characteristics. Europe provides a rich environment to explore the effect of compensation and investor protection on payouts as investor protection differs across countries, and investor protection affects the agency costs of free cash flow (La Porta et al. (1997, 1998), henceforth LLSV). For firms in weak protection countries, LLSV (2000) posit that dividends can serve to maintain a reputation for better treatment of minority shareholders. Pinkowitz et al. (2006) conclude that in countries with weak investor protection, higher dividend payouts reduce the agency costs of free cash flow. However, the relationship between investor protection and payouts may be altered by compensation if equity-incentive compensation also reduces agency costs. In countries with weaker investor protection, we postulate

⁎ Corresponding author at: College of Business, ONE UTSA Blvd., The University of Texas at San Antonio, San Antonio, TX 78249, United States. Tel.: +1 210 458 6838. E-mail addresses: [email protected] (N. Burns), [email protected] (B.C. McTier), [email protected] (K. Minnick). 1 Tel.: +1 360 546 9516. 2 Tel.: +1 781 891 2941. 3 See Conyon et al. (2013), Cheffins (2003) and Croci et al. (2012).

http://dx.doi.org/10.1016/j.jcorpfin.2014.10.019 0929-1199/© 2014 Elsevier B.V. All rights reserved.

86

N. Burns et al. / Journal of Corporate Finance 30 (2015) 85–97

that equity-incentive compensation and dividends act as substitutes because both can reduce agency costs. Therefore, we expect firms from weak investor protection countries that have more equity-incentive compensation to pay lower dividends. Growth opportunities may affect the interplay between investor protection, incentive compensation, and payouts. Firms with better growth options may pay out more to maintain a favorable reputation with minority shareholders for distributing excess free cash flows as they are more likely to require access to external capital markets, especially firms in countries with weak investor protection.4 However, LLSV (1997, 1998) find that investors in weak investor protection countries are less willing to provide financing to firms, which is particularly problematic for high growth firms in countries with weak investor protections. These firms may prefer to pay fewer dividends and use retained earnings to invest in growth opportunities; yet paying fewer dividends conflicts with potentially using dividends to reduce the agency costs associated with weak investor protection. Durnev and Kim (2005) model that firms with a greater need for external financing and better investment opportunities institute better governance and that this relationship is stronger if the country-level legal environment is weaker. Similarly, to overcome weak investor protection, we hypothesize that high growth firms use more equity-incentive pay to align manager's interests to shareholders thereby reducing the need to pay higher dividends, and increasing financial flexibility.5 We expect high growth firms that use more equity-incentive compensation to have lower (higher) dividend payouts (repurchases) and for this effect to be stronger in countries with weak investor protection. While incentive compensation may reduce agency costs, the composition of incentive compensation would be expected to induce executives to favor a particular form of payout; for example, option and restricted stock compensation could have competing effects on the form of manager payout choice (Aboody and Kasznick, 2008; Brown et al., 2007; Fenn and Liang, 2001). Executive stock options incentivize management to reduce dividends because the value of non-dividend protected options is negatively related to dividend payments (Lambert et al., 1989).6 For U.S. firms, Fenn and Liang (2001) find that options are associated with lower dividends and more repurchases, and suggest that repurchases can be used to offset option-induced reductions in dividends when a firm has a target payout policy.7 While the primary effect of incentive compensation is to reduce agency costs, we also examine whether options and restricted stock have differing effects on payout policy in Europe compared to the U.S. Further, we examine the effect of dividend protection of options and restricted stock on payout policy. Using a sample spanning 15 European countries over the period 2003–2012, we examine the association between payout policy and compensation as Europe increasingly adopts U.S.-like compensation policies. We find that incentive compensation is associated with a reduction in dividends paid and an increase in repurchases. When we decompose equity-incentive compensation into its components, we find similar effects of option and restricted stock incentive compensation on the form of payout in the general cross sectional analysis. This is in contrast to results in the U.S. which shows a negative relationship between options and dividends in the U.S. (Aboody and Kasznick, 2008), but a positive relationship between restricted stock grants and dividends. The negative relation between stock options and dividends in the U.S. is attributed to the fact that options are rarely dividend-protected while restricted stock is more likely to be dividend protected (Minnick and Rosenthal, 2014). We therefore consider the effect of dividend protection on the relationship between incentive compensation and payout policy. We hand collect dividend protection practices for our sample and find that dividend protection of options (restricted stock) is more (less) common in Europe than in the U.S. We find that dividend protected incentive compensation reduces the negative association between equity-based compensation and dividends and is insignificantly associated with repurchases. Importantly, the overall effect that incentive compensation is negatively (positively) associated with dividend payout holds. Moreover we explore the effect of incentive pay on payouts when country-level investor protections or firm growth opportunities are considered. We find that firms in weak protection countries pay higher dividends. This positive relationship is consistent with firms in weak protection countries using dividends to establish a favorable reputation with minority shareholders for distributing excess free cash flow when they are likely to access capital markets. However, these firms reduce dividends at a greater rate with increases in incentive compensation, consistent with the idea that dividends and incentive-compensation both affect agency costs and may be substitutes. We explore the interplay between investor protection, growth opportunities, and incentive compensation and find that in weak protection countries, high growth firms with more incentive compensation reduce dividends at a faster rate than firms in strong investor protection countries. Our analysis of repurchases similarly finds that repurchases increase at a faster rate when more incentive compensation is used in weak protection countries. Considered together, our findings are consistent with growth firms as well as firms in weak investor protection countries using equity incentive compensation in lieu of dividends as a tool to reduce agency costs related to free cash flow. The equity-incentive compensation enables the firm to reduce dividend payouts and use repurchases to distribute excess free cash flow, thereby increasing financial flexibility. Our research extends the literature that examines how compensation affects payout policy in the U.S. (see for examples Lambert et al., 1989; Fenn and Liang, 2001; Chetty and Saez, 2005; and Brown et al., 2007). While Eije and Megginson (2008) examine trends in payouts in Europe, they do not explore the role of compensation. In contemporaneous work, Cesari and Ozkan (2014) study seven European countries and find that dividends are negatively related to CEO option holdings and are not related to CEO equity ownership. In contrast, our research contributes to the literature by examining the role of equity-based compensation (both restricted stock and option compensation) on payout policy in Europe, as well as the effects of a company's dividend protection, growth prospects, and investor protection. Our research adds to several papers that examine agency issues and investor protection around payout policy.

4

LLSV (2000) find some evidence that growth firms in weak protection countries pay greater dividends than mature firms. See e.g. Jagannathan et al. (2000) for a discussion of the importance of financial flexibility on the choice between dividends and repurchases. 6 Dividend protection for options is the practice of compensating managers for dividend payouts, either by decreasing the exercise price of the option, or by adding accumulated dividends plus interest to the stock price upon exercise of the underlying options. 7 Also see Grullon and Michaely (2002) for research related to the substitution of share repurchases for dividends. 5

N. Burns et al. / Journal of Corporate Finance 30 (2015) 85–97

87

Table 1 Distribution of observations. Table 1 reports the yearly breakout of the sample from 2003–2011 by country.

Austria Belgium Denmark Finland France Germany Ireland Italy Netherlands Norway Portugal Spain Sweden Switzerland United Kingdom Total

2003

2004

2005

2006

2007

2008

2009

2010

2011

Total

1 4 12 16 133 53 36 51 90 27 0 4 58 55 519 1059

5 8 15 53 149 86 46 75 97 52 2 3 107 71 568 1337

5 13 13 63 152 122 43 73 96 51 2 5 84 63 631 1416

8 24 15 61 151 188 44 78 95 80 3 4 113 67 652 1583

12 22 15 60 171 190 51 78 93 96 2 9 125 125 672 1721

10 22 13 52 169 193 41 85 93 88 3 7 113 136 684 1709

11 25 13 47 156 197 41 92 94 84 9 8 103 135 697 1712

12 27 18 58 160 209 48 124 99 94 12 9 111 145 758 1884

19 31 18 55 152 197 42 75 83 90 14 10 115 141 728 1770

83 176 132 465 1393 1435 392 731 840 662 47 59 929 938 5909 14191

LLSV (2000) and Denis and Osobov (2008) find that dividend payouts are higher in common law countries with strong investor protection as an outcome of shareholder power. Alzahrani and Lasfer (2012) show that it is important to account for taxes—in countries with strong shareholder protection: firms substitute repurchases for dividends if there is a dividend tax penalty. We extend this research on investor protection and payout policy by examining how CEO compensation relates to this dynamic between investor protection, growth opportunities, dividend protection, and payout policy. The paper proceeds as follows. In Section 2 we describe our data; Section 3 presents and discusses our empirical results. We offer our conclusions in Section 4. 2. Data, variable measurement, and descriptive statistics We obtain CEO compensation, payouts, stock price, and accounting information from Capital IQ (CIQ) for European countries from 2003–2011. We study this period as CIQ provides more detail on compensation after 2002. Focusing on this period also avoids confounding our results with the creation of the European Union. Conditioning on data availability yields an initial sample of 14,191 firm-years across 15 European countries. Table 1 lists the country composition and number of firms by year. UK firms are the largest part of the sample (41.6%), followed by Germany (10%) and France (9.8%). Nordic countries (Denmark, Finland, Norway, and Sweden) comprise 15.4% of the sample. The Benelux (Belgium and Netherlands) countries (7.2%), Southern European (Italy, Spain, Portugal) (5.9%), and other countries (10%), such as Switzerland, Ireland, and Austria are also represented. Table 2 presents summary statistics for the primary measures of interest. We winsorize all continuous variables at the 0.5% level on each tail to reduce the influence of outliers. Our primary measures of firm payouts are common dividends divided by assets, and the value of repurchases divided by assets. The average dividend payout and repurchase is 1.7% and 0.7% of assets, respectively.8 The average total CEO compensation of firms in our sample is $1.46 million. Fifteen percent of our sample pays any form of incentive compensation. For the subset of firms that pay options or restricted stock, average option and restricted stock compensation is 16% and 14% of total compensation, respectively.9 Panels B and C of Table 2 report summary statistics for payouts and compensation by country.10 Extant research finds that payouts are related to a variety of country characteristics, as well as firm characteristics including firm size, profitability, cash holdings, leverage, risk, and institutional ownership. Firm size, measured by assets, is expected to be positively related to payout (Fama and French, 2001). Our measure of size is the logarithm of assets (Log Assets). Aboody and Kasnick (2008) show that more profitable companies have more free cash flow available for payouts. We measure profitability as the firm's EBITDA scaled by assets (ROA). We measure growth opportunities with Tobin's Q as the market capitalization of the firm's stock plus debt divided by total assets. Cash holdings are also important to payout policy. For instance, Jensen (1986) argues that firms with larger amounts of free cash flow have greater flexibility in their dividend policies. Empirically, Opler et al. (1999) find a negative relationship between cash holdings and dividends, while Lee and Suh (2011) provide evidence that repurchasing firms carry more cash than dividend paying firms. Our measure of cash is cash holdings divided by total assets (Cash Ratio). We also include the Debt Ratio, measured as total debt divided by assets. Jensen (1986) contends that higher debt reduces the free cash flow problem, suggesting that highly levered firms have less cash to pay out to shareholders. Kahle (2002) suggests that riskier firms will not increase 8 The average dividend payout is slightly lower for European firms than for US firms (2.2%) over the same time frame, however the average repurchase amount is similar (Minnick and Rosenthal, 2014). We also use dividends and repurchases divided by market value for robustness. 9 In a similarly constructed large sample of U.S. firms from CIQ, 50% offer equity-incentive compensation, with option and restricted stock representing 21% and 22% of total compensation. 10 In unreported summary statistics, we find that when we separate our sample into high growth and low growth firms, the average incentive pay ratio is significantly higher for high growth firms, 5.6% compared to 3.1%. Similarly, when we separate our sample by investor protection, the average incentive pay ratio is significantly higher for firms in weak investor protection, 5.8% compared to 3.0%.

88

N. Burns et al. / Journal of Corporate Finance 30 (2015) 85–97

Table 2 Characteristics of payouts and incentive compensation. Table 2 reports characteristics for the payout and compensation variables we use in our estimation. Panel A reports univariate statistics for the payout variables and compensation variables. The number of observations, mean, median, and standard deviation values are displayed. In addition, we show univariate statistics for the subset of observations in our sample that have payout and compensation values above zero. Dividend to Assets is the total dividend payout in a fiscal year divided by the total assets of the firm. Repurchases to Assets is the total dollar amount of repurchases in a fiscal year divided by the total assets. CEO Options (RSG) Ratio is the ratio of the value of CEO option compensation (RSGs) as reported in Capital IQ divided by the total compensation that a CEO receives. CEO Incentives Ratio is the sum of the option compensation and RSGs a CEO receives divided by the total compensation received. Panels B and C report univariate characteristics by country. Panel A: Univariate characteristics of payouts and incentive compensation

Full sample Dividend to Assets Repurchases to Assets CEO Options Ratio CEO RSG Ratio CEO Incentives Ratio Total Compensation (Th) Firms with non-zero payouts Dividend to Assets Repurchases to Assets CEO Options Ratio CEO RSG Ratio CEO Incentives Ratio

N

Mean

Median

Std. Dev.

14,191 14,191 14,191 14,191 14,191 14,191

0.017 0.007 0.022 0.020 0.044 1461

0.008 0.000 0.000 0.000 0.000 668

0.028 0.028 0.091 0.089 0.136 17,870

8130 4095 1295 1064 2058

0.03 0.03 0.25 0.27 0.30

0.02 0.01 0.20 0.23 0.25

0.03 0.05 0.19 0.19 0.22

Panel B: Univariate characteristics of payouts by country % Div. N0

% Rep. N0

$Div.

$Rep.

Div/A (%)

Rep/A (%)

83

66.3%

21.7%

176

64.8%

34.1%

132

67.4%

25.0%

465

80.2%

21.5%

1393

67.1%

41.5%

1435

53.0%

20.3%

392

54.3%

28.6%

731

57.3%

30.2%

840

53.3%

30.0%

662

42.6%

29.2%

47

78.7%

48.9%

59

52.5%

35.6%

929

61.7%

17.8%

938

55.3%

49.7%

5909

55.5%

26.4%

62.52 10.50 43.27 4.48 62.21 2.47 89.01 8.79 157.39 7.80 126.88 0.96 50.14 1.67 104.24 2.01 84.71 1.42 62.94 0.00 110.40 33.20 166.76 0.34 74.59 3.16 140.11 3.15 65.09 0.72

5.85 0.00 15.98 0.00 93.52 0.00 30.32 0.00 43.88 0.00 26.69 0.00 102.67 0.00 13.39 0.00 76.36 0.00 12.68 0.00 4.23 0.00 4.76 0.00 19.45 0.00 74.16 0.00 29.74 0.00

1.35% 0.87% 1.43% 0.76% 2.38% 0.75% 3.49% 2.50% 1.39% 0.86% 1.50% 0.41% 1.23% 0.77% 1.32% 0.44% 1.60% 0.44% 1.73% 0.00% 1.19% 0.85% 1.30% 0.10% 2.57% 1.56% 1.80% 0.75% 1.71% 0.83%

0.24% 0.00% 0.64% 0.00% 1.39% 0.00% 0.69% 0.00% 0.50% 0.00% 0.70% 0.00% 1.49% 0.00% 0.29% 0.00% 0.99% 0.00% 0.80% 0.00% 0.05% 0.00% 0.32% 0.00% 0.82% 0.00% 1.35% 0.00% 0.66% 0.00%

Obs. Austria Belgium Denmark Finland France Germany Ireland Italy Netherlands Norway Portugal Spain Sweden Switzerland UK

Mean Med. Mean Med. Mean Med. Mean Med. Mean Med. Mean Med. Mean Med. Mean Med. Mean Med. Mean Med. Mean Med. Mean Med. Mean Med. Mean Med. Mean Med.

Panel C: Univariate characteristics of compensation by country Obs. Austria Belgium Denmark Finland France

Mean Med. Mean Med. Mean Med. Mean Med. Mean Med.

% CEO Opts. N0

% CEO RSG N0

% CEO Incent. N 0

CEO Opts To Comp

CEO RSG To Comp

5.5% 0.0% 1.9% 0.0% 4.3% 0.0% 1.6% 0.0% 3.7% 0.0%

0.0% 0.0% 1.5% 0.0% 3.2% 0.0% 3.7% 0.0% 3.1% 0.0%

83

19.3%

0.0%

19.3%

176

8.5%

7.4%

10.2%

132

28.0%

14.4%

41.7%

465

6.9%

10.8%

15.9%

1393

12.1%

10.3%

17.9%

CEO Incent. To Comp 5.6% 0.0% 3.4% 0.0% 7.5% 0.0% 5.4% 0.0% 6.9% 0.0%

$CEO Opts.

$CEO RSG

$CEO Incent.

49.36 0.00 70.49 0.00 75.34 0.00 27.25 0.00 143.06 0.00

0.00 0.00 88.67 0.00 92.20 0.00 126.58 0.00 128.44 0.00

49.36 0.00 165.75 0.00 167.54 0.00 162.10 0.00 325.87 0.00

N. Burns et al. / Journal of Corporate Finance 30 (2015) 85–97

89

Table 2 (continued) Panel C: Univariate characteristics of compensation by country

Germany Ireland Italy Netherlands Norway Portugal Spain Sweden Switzerland UK

Mean Med. Mean Med. Mean Med. Mean Med. Mean Med. Mean Med. Mean Med. Mean Med. Mean Med. Mean Med.

Obs.

% CEO Opts. N0

% CEO RSG N0

% CEO Incent. N 0

CEO Opts To Comp

CEO RSG To Comp

CEO Incent. To Comp

$CEO Opts.

$CEO RSG

$CEO Incent.

1435

10.8%

10.0%

19.1%

392

19.1%

16.8%

26.5%

731

4.2%

2.6%

6.7%

840

19.3%

12.4%

26.9%

662

12.7%

2.7%

15.1%

47

0.0%

0.0%

0.0%

59

5.1%

3.4%

6.8%

929

4.3%

5.1%

8.4%

938

23.8%

28.3%

43.4%

5909

4.3%

2.9%

6.8%

2.5% 0.0% 5.2% 0.0% 1.1% 0.0% 3.9% 0.0% 2.9% 0.0% 0.0% 0.0% 1.2% 0.0% 0.6% 0.0% 5.7% 0.0% 1.2% 0.0%

2.3% 0.0% 5.0% 0.0% 0.6% 0.0% 4.0% 0.0% 0.5% 0.0% 0.0% 0.0% 0.6% 0.0% 0.6% 0.0% 7.8% 0.0% 0.8% 0.0%

4.8% 0.0% 10.3% 0.0% 1.8% 0.0% 8.1% 0.0% 3.4% 0.0% 0.0% 0.0% 1.8% 0.0% 1.3% 0.0% 13.7% 0.0% 2.1% 0.0%

65.18 0.00 222.45 0.00 18.91 0.00 93.14 0.00 27.07 0.00 0.00 0.00 25.47 0.00 15.58 0.00 204.09 0.00 18.60 0.00

99.00 0.00 348.04 0.00 16.01 0.00 144.96 0.00 6.01 0.00 0.00 0.00 28.33 0.00 13.04 0.00 383.55 0.00 28.78 0.00

170.40 0.00 617.32 0.00 40.40 0.00 245.68 0.00 33.07 0.00 0.00 0.00 53.80 0.00 35.66 0.00 638.68 0.00 51.25 0.00

% CEO Opts. (RSG, Incent.) N0 is the percent of firms that used CEO Opts. (RSG, Incent.) pay. CEO Opts. (RSG, Incent.) to Comp is CEO Opts. (RSG, Incent.) divided by CEO total compensation. $CEO Opts. (RSG, Incent.) is the dollar value (thousands) of CEO Opts. (RSG, Incent.). % Div. (% Rep.) is the percent of firms with dividends (repurchases) payouts. $Div. ($Rep.) is the dollar value (millions) of dividends (repurchases). Div/A (Rep/A) is the value of dividends (repurchases) divided by assets.

their dividends but instead will use a repurchase as a way of paying out free cash flow to investors. We measure risk, SD Return, as the standard deviation of monthly returns over the fiscal year. Institutional ownership (Inst % Owned) has also been shown to be positively related to payout policy. Short et al. (2002) suggest two reasons for this relationship. First, institutions may demand high levels of dividends in order to force firms to go to the capital markets for external funding and, hence, be subject to monitoring by the external market. Second, institutions may counter management's tendency to retain excess free cash flow. Table 3 provides descriptive statistics for firm characteristics. In our sample, the average firm has over $4 billion in assets. The average ROA of our sample is 5.7%. With respect to other financial characteristics, the average European firm has cash holdings of 14.5% and leverage of 21.8% of assets. The average institutional ownership of firms in our sample is 24%. Dividend protection may be more common in European countries than in the U.S. (Liljeblom and Pasternack, 2006; Winter, 2000).11 We hand collect information on dividend protection from firm's annual reports and classify firms by whether the firm dividend protects options, RSGs, or both.12 We show the breakdown by country of the firms that have dividend protection in Table 4. Of the subsample of firms that offer incentive compensation to their CEOs (either options or stock grants), we find that 19% of these firms offer dividend protection. Dividend protection is most common in Finland, France, Germany and the Netherlands, while companies in Portugal and Spain never offer dividend protection. 3. Empirical results 3.1. Evidence on the relation between equity incentive compensation and payouts We examine the relation between compensation and payouts while controlling for firm specific and country measures in a multivariate framework. Because the payout measures are strictly non-negative and censored at zero, we perform our primary tests using multivariate Tobit regression analysis. We also explore the effect of compensation on the likelihood of payouts using a probit analysis. In all specifications, we report a White heteroskedastic-consistent estimator adjusted for clustering at the firm level (Andrews, 1991; Zeileis, 2004), and include year and industry fixed effects. Clustering standard errors by firm and including year fixed effects minimizes bias in standard errors from firm and time effects (Petersen, 2009). Our primary specification is: Payout tþ1 ¼ α þ β1 Compt þ β3 ADRIt þ β3 DTP t þ β5−14 Firm Characteristicst þ ε;

11 Arnold and Gillenkirch (2005) discuss the opération blanche and strike price adjustment method for dividend protection of incentive compensation and theoretically test their effectiveness. Under the opération blanche, the number of shares per option is increased according to the number of additional shares a shareholder could buy from the dividend received; under the strike price adjustment method, the strike price is reduced by the compounded dividend. 12 Interestingly, we note that the majority of firms that protect options do so by reducing the strike price of the option. The methods to protect stock grants are more varied, although they tend to fall into one of three categories: 1) pay out the dividends on unvested stock grants at the dividend grant date; 2) accumulate the dividends that would be received and then pay them out when the stock vests; 3) simply issue new shares in the amount of the dividends.

90

N. Burns et al. / Journal of Corporate Finance 30 (2015) 85–97

Table 3 Univariate analysis of firm characteristics. Table 3 provides mean, median and standard deviation for firm characteristics. Debt ratio is Total Debt divided by Assets; Q is market value of equity plus debt divided by total assets; ROA is Earnings before Interest, Taxes, Depreciation, and Amortization (EBITDA); Cash Ratio is cash divided by Assets and SD Return is the standard deviation of the firm's monthly stock returns over a year. Inst % owned is the percent of the firm's stock held by institutions.

Assets (MM) Debt Ratio Q ROA SD Return Cash Ratio Inst % Owned

Obs.

Mean

Median

Std. Dev.

14,191 14,191 14,191 14,191 14,191 14,191 14,191

4189 0.218 1.551 0.057 0.133 0.145 24.190

290 0.162 1.021 0.097 0.111 0.089 18.580

15,098 0.221 1.952 0.235 0.088 0.162 23.900

where our Payout measure is either Repurchases or Dividends calculated as repurchases or common dividends divided by assets, respectively. Our compensation measure (Comp) is equity incentive compensation or its separate components of options and restricted stock grants (RSG), divided by total compensation, each measured in the year prior to the payout. ADRI is the “anti-director rights index” from Spamann (2010). We include two country specific controls for taxes and investor protection. We control for the country dividend tax preference using the method of Alzahrani and Lasfer (2012). We calculate the dividend tax preference (DTP) as (1 − the net individual tax rate) divided by (1 − the capital gains tax rate). The net individual tax rate is the tax rate on dividends net of any relief or tax credit applicable to dividends.13 Firm Characteristics include Log Assets, Debt Ratio, Cash Ratio, ROA, Q, Inst % Owned, SDRet., each measured in the year prior to the payout.14 Table 5 presents regression results. Model (1) estimates the probit model of the likelihood that a firm pays dividends. Models (2) through (4) report Tobit regression results with dividend payout as the dependent variable. Each of these models includes incentive compensation as the independent variable of primary interest. Model (2) includes investor protection as an explanatory variable. Model (3) includes country fixed effects and therefore drops investor protection which is subsumed by country fixed effects. Model (4) excludes the UK which makes up 42% of the sample to alleviate concerns that our results are driven by firms in the UK. Models (5) through (8) repeat these regressions but include options and restricted stock separately. The coefficient on incentive compensation in Model (1) is negative and significant with a p-value of less than one percent suggesting that CEOs with more incentive compensation are less likely to pay a dividend. A one unit increase in incentive compensation decreases the odds of a company using dividends by 71%.15 In Models (2) through (4) the coefficient on incentive compensation is reliably negative and significant across each of these estimations. The negative coefficient on incentive compensation implies that higher incentive compensation is associated with lower dividend payouts. A one standard deviation increase in incentive compensation would decrease dividend payouts by 0.3% (Model 2). As the average payout is 2%, this decrease in dividend payouts is economically significant. As prior literature focusing on the U.S. finds that restricted stock grants and option grants may have different effects on payout policy (Aboody and Kasnik, 2008), we separate these components of incentive compensation. We repeat the models of Columns (1) through (4) in Columns (5) through (8) but with equity incentive compensation separated into its components of options and restricted stock grants. Across all specifications both options and restricted stock are negative and significant (p-value at the 1% level). Considering Model (5), a one unit increase in options (RSG) decreases the odds of a company paying dividends by 70% (73%). Model (6) shows that a one standard deviation increase in options (RSG) decreases the dividend payout by 0.24% (0.25%). With respect to control variables, we find larger and more profitable firms are more likely to pay dividends and pay larger dividends, while those with higher leverage and greater risk are less likely to pay dividends and pay smaller dividends. In the probit Models (1) and (5), the negative relationship between cash and the likelihood of dividend payouts is consistent with the idea that firms with greater need to hold cash balances are less likely to pay out dividends, possibly because of precautionary demands (Almeida et al., 2004; Boyle and Guthrie, 2003). However, the Tobit models (2) and (4) show that companies that have higher cash balances pay higher dividends. We find that higher growth firms are less likely to pay dividends as can be inferred by the negative coefficient on Q in model (1), but when they do, the payout ratio is higher (models (2) through (4)). Institutional ownership is associated with a reduction in the likelihood that a company will pay dividends, but we do not find a significant association between actual payouts and institutional ownership. In countries where dividends are taxed preferentially, dividend payouts are higher. Stronger investor protection (ADRI) is associated with lower dividend payouts when controlling for dividend tax preference. This is consistent with Alzahrani and Lasfer (2012) who find that after controlling for the dividend tax preference, managers in strong investor protection countries pay lower dividends.16 We next explore the effect of incentive compensation on repurchases and report the results in Table 6 fitting the same models as in Table 5 with our repurchase variable as the dependent variable. Model (1) (probit) shows that repurchases are more likely in firms

13 We start with the top statutory individual tax rates from the annual OECD tax database www.oecd.org/ctp/taxdatabase). The capital gains tax rate is from Price Waterhouse: Corporate & Individual taxes: A worldwide Summary. 14 We use lag firm characteristic and compensation measures to reduce concerns about endogeneity. 15 The odds ratio is calculated at 1 − exp(−1.2315). 16 In unreported results we verify that when we remove DTP, ADRI is positive with a p-value of 0.1401, providing further evidence that controlling for DTP is important as reported by Alzahrani and Lasfer (2012).

N. Burns et al. / Journal of Corporate Finance 30 (2015) 85–97

91

Table 4 Dividend protection by country. Table 4 presents summaries of the use of dividend protection by country. Num. w/Opt. Prot. (Num. w/RSG Prot.) are firm/years by country where option compensation (RSG compensation) is dividend protected. Num. w/ Both Prot. are firm/years where both options and RSG compensation are dividend protected.

Austria Belgium Denmark Finland France Germany Ireland Italy Netherlands Norway Portugal Spain Sweden Switzerland United Kingdom

Num. Obs.

Num. w/Opt. Prot.

Num. w/RSG Prot.

Num. w/Both Prot.

83 176 132 465 1393 1435 392 731 840 662 47 59 929 938 5909

3 7 19 89 65 69 6 7 45 14 0 0 37 26 25

0 7 8 82 125 123 29 9 57 17 0 0 39 90 106

0 7 8 62 27 26 6 0 17 8 0 0 37 11 13

that have more equity incentive compensation. Models (2) through (4) (Tobit) show a significant positive relation between incentive pay and the level of repurchases for the broad sample of all firms as well as the non-UK firms. A one unit increase in incentive compensation increases the odds of using repurchases by 66% (Model 1), and a one standard deviation increase in incentive compensation increases the repurchase payout by 0.4% (Model 2). These results are consistent with a preference for repurchases when equityincentive compensation is a greater portion of CEO compensation. We also explore whether restricted stock grants and option grants have a different impact on repurchase activity. Regression (5) shows that repurchases are also more likely when restricted stock is greater. In contrast the coefficient on options is positive but insignificant (p-value of 0.13). The Tobit regression models (6) through (8) show higher restricted stock compensation is reliably associated with larger repurchases, for the full sample and non-UK sample.

Table 5 Multivariate analysis of dividend payouts. Table 5 reports coefficients from Probit and Tobit regression analyses with the dependent variable censored at zero. The dependent variable is dividends to total assets. We use robust standard errors clustered by firm and include industry and year fixed effects. ADRI is the “Spamann Antidirector Rights Index” from Spamann (2010) and measures the level of shareholder protection. All other variables are defined in Tables 2 and 3. P-values are reported below the coefficient. ***, **, and * indicate significance at the 1, 5, and 10% levels, respectively.

Incentive Ratio

Europe (Probit)

Europe

Europe (FE)

Europe, No UK

Europe (Probit)

Europe

Europe (FE)

Europe, No UK

(1) −1.2315*** (0.0000)

(2) −0.0252*** (0.0000)

(3) −0.0257*** (0.0000)

(4) −0.0208*** (0.0000)

(5)

(6)

(7)

(8)

−1.2121*** (0.0000) −1.2950*** (0.0000) 0.0616 (0.2341) −0.0027** (0.0168) −0.1405*** (0.0000) 0.2679*** (0.0000) −1.4635*** (0.0000) −0.5226*** (0.0038) 6.2538*** (0.0000) −5.7921*** (0.0000) 0.9599*** (0.0000) 0.0658 (0.1466) −1.4762*** (0.0000) 0.423 14,191

−0.0225*** (0.0002) −0.0285*** (0.0000) −0.0030** (0.0126) −0.0000 (0.3174) 0.0011 (0.1121) 0.0031*** (0.0000) −0.0387*** (0.0000) 0.0095* (0.0759) 0.2254*** (0.0000) −0.1434*** (0.0000) 0.0263*** (0.0000) −0.0203 (0.3171) −0.0196*** (0.0014) 0.283 14,191

−0.0231*** (0.0001) −0.0290*** (0.0000)

−0.0165** (0.0180) −0.0253*** (0.0000) −0.0011 (0.4111) −0.0001 (0.1425) 0.0008 (0.4446) 0.0036*** (0.0000) −0.0402*** (0.0000) 0.0230*** (0.0038) 0.2516*** (0.0000) −0.1651*** (0.0000) 0.0439*** (0.0000) −0.0897*** (0.0021) −0.0488*** (0.0000) 0.273 8282

Options Ratio RSG Ratio ADRI Inst % Owned Q LogAssets DebtRatio CashRatio ROA SDRet DTP Repurchases Intercept R2 Obs.

0.0622 (0.2298) −0.0027** (0.0163) −0.1399*** (0.0000) 0.2677*** (0.0000) −1.4620*** (0.0000) −0.5221*** (0.0038) 6.2516*** (0.0000) −5.7932*** (0.0000) 0.9596*** (0.0000) 0.0652 (0.1502) −1.4792*** (0.0000) 0.424 14,191

−0.0030** (0.0133) −0.0000 (0.3059) 0.0011 (0.1078) 0.0031*** (0.0000) −0.0386*** (0.0000) 0.0095* (0.0759) 0.2254*** (0.0000) −0.1434*** (0.0000) 0.0262*** (0.0000) −0.0206 (0.3123) −0.0197*** (0.0013) 0.283 14,191

−0.0000 (0.3600) 0.0009 (0.1880) 0.0031*** (0.0000) −0.0389*** (0.0000) 0.0110** (0.0400) 0.2252*** (0.0000) −0.1403*** (0.0000) 0.0135*** (0.0010) −0.0171 (0.3963) −0.0209*** (0.0006) 0.284 14,191

−0.0010 (0.4357) −0.0001 (0.1350) 0.0009 (0.4275) 0.0036*** (0.0000) −0.0401*** (0.0000) 0.0229*** (0.0038) 0.2516*** (0.0000) −0.1650*** (0.0000) 0.0439*** (0.0000) −0.0904*** (0.0021) −0.0491*** (0.0000) 0.273 8282

−0.0000 (0.3707) 0.0009 (0.1945) 0.0031*** (0.0000) −0.0389*** (0.0000) 0.0110** (0.0400) 0.2252*** (0.0000) −0.1404*** (0.0000) 0.0136*** (0.0009) −0.0169 (0.4016) −0.0210*** (0.0005) 0.284 14,191

92

N. Burns et al. / Journal of Corporate Finance 30 (2015) 85–97

Table 6 Multivariate analysis of repurchases. Table 6 reports coefficients from Probit and Tobit regression analyses with the dependent variable censored at zero. The dependent variable is repurchases to total assets. We use robust standard errors clustered by firm and include industry and year fixed effects. ADRI is the “Spamann Antidirector Rights Index” from Spamann (2010) and measures the level of shareholder protection. All other variables are defined in Tables 2 and 3. P-values are reported below the coefficient. ***, **, and * indicate significance at the 1, 5, and 10% levels, respectively.

Incentive Ratio

Europe (Probit)

Europe

Europe (FE)

Europe, No UK

Europe (Probit)

Europe

Europe (FE)

Europe, No UK

(1) 0.5076*** (0.0000)

(2) 0.0351*** (0.0000)

(3) 0.0310*** (0.0000)

(4) 0.0395*** (0.0000)

(5)

(6)

(7)

(8)

0.2494 (0.1379) 0.7991*** (0.0000) 0.0200 (0.6463) 0.0008 (0.3389) −0.0086 (0.7313) 0.1757*** (0.0000) −0.8351*** (0.0000) 0.3964** (0.0140) 1.0828*** (0.0043) −1.7894*** (0.0000) −0.1270 (0.2805) 0.1141** (0.0403) −1.3848*** (0.0000) 0.142 14,191

0.0184** (0.0472) 0.0516*** (0.0000) −0.0009 (0.6724) 0.0001** (0.0211) 0.0017 (0.1075) 0.0079*** (0.0000) −0.0488*** (0.0000) 0.0364*** (0.0001) 0.0781*** (0.0000) −0.0841*** (0.0000) −0.0002 (0.9711) 0.0334 (0.4818) −0.0742*** (0.0000) 0.122 14,191

0.0147 (0.1195) 0.0472*** (0.0000)

0.0229** (0.0274) 0.0548*** (0.0000) −0.0037* (0.0939) 0.0001* (0.0799) 0.0030** (0.0434) 0.0063*** (0.0000) −0.0358*** (0.0000) 0.0393*** (0.0018) 0.0756*** (0.0012) −0.1054*** (0.0000) −0.0195** (0.0422) −0.0727 (0.1606) −0.0417*** (0.0032) 0.096 8282

Options Ratio RSG Ratio ADRI InstPerOwned Q LogAssets DebtRatio CashRatio ROA SDRet DTP Dividends Intercept R2 Obs.

0.0166 (0.7029) 0.0009 (0.3129) −0.0096 (0.7046) 0.1766*** (0.0000) −0.8406*** (0.0000) 0.3935** (0.0146) 1.0934*** (0.0038) −1.7969*** (0.0000) −0.1244 (0.2907) 0.1133** (0.0417) −1.3718*** (0.0000) 0.141 14,191

−0.0011 (0.6057) 0.0001** (0.0209) 0.0016 (0.1203) 0.0080*** (0.0000) −0.0493*** (0.0000) 0.0364*** (0.0001) 0.0791*** (0.0000) −0.0848*** (0.0000) 0.0000 (0.9984) 0.0316 (0.5057) −0.0735*** (0.0000) 0.122 14,191

0.0001** (0.0102) 0.0016 (0.1252) 0.0079*** (0.0000) −0.0507*** (0.0000) 0.0348*** (0.0002) 0.0793*** (0.0000) −0.0825*** (0.0000) −0.0088 (0.3217) 0.0420 (0.3825) −0.0878*** (0.0000) 0.132 14,191

−0.0039* (0.0739) 0.0001* (0.0788) 0.0029** (0.0488) 0.0064*** (0.0000) −0.0363*** (0.0000) 0.0392*** (0.0019) 0.0763*** (0.0011) −0.1062*** (0.0000) −0.0194** (0.0431) −0.0747 (0.1503) −0.0406*** (0.0040) 0.095 8282

0.0001*** (0.0098) 0.0016 (0.1116) 0.0079*** (0.0000) −0.0504*** (0.0000) 0.0349*** (0.0002) 0.0784*** (0.0000) −0.0818*** (0.0000) −0.0089 (0.3152) 0.0438 (0.3621) −0.0868*** (0.0000) 0.133 14,191

Model (6) shows that a one standard deviation increase in RSG increases the repurchase payout ratio by 0.5%. The coefficient on options is positive in each regression, however loses its significance in the probit and country fixed effects models.17 In summary, dividends (repurchases) are lower (higher) when CEOs are compensated with more equity-based compensation. Further, we do not see a differing effect between options and RSGs in Europe—both decrease dividends and increase repurchase payouts. Our results for options are consistent with findings in the U.S., where higher option pay is related to lower dividend payouts and higher repurchases (Aboody and Kasznick, 2008). However, our results for restricted stock grants are inconsistent with evidence in the U.S., where typically higher RSG payout is related to higher dividends (Minnick and Rosenthal, 2014). One explanation for the finding in the U.S. considers the role dividend protection plays in managers' decisions to use either dividends or repurchases to pay out excess cash. In practice, U.S. firms rarely offer dividend-protection for options: Murphy (1999) reports seven dividend protected option plans in a sample of 618 U.S. firms. Conversely, dividend protection is more common for RSGs in the U.S. (Minnick and Rosenthal, 2014). Based on small sample evidence, Liljeblom and Pasternack (2006) and Winter (2000) find that dividend protection for options is more popular in Europe, at least in specific countries (however, these papers did not look at dividend protection for RSGs). As the lack of dividend protection for options and the use of dividend protection for RSGs potentially offers an explanation for the relationship between corporate payout policy and incentive compensation in the U.S., then it is possible that dividend protection offers greater insight into our results.18 We consider these issues further in the next section. 3.2. Dividend protection and corporate payout policy We explore the role dividend protection plays in influencing corporate payout policy by hand collecting information on dividend protection for our sample of firms. Of the firms that use equity based compensation, we find that 19% offer dividend protection. Our sample provides an interesting opportunity to examine the impact of dividend protection in Europe, where dividend protection for options is more common than in the U.S., and we find that dividend protection for restricted stock is less common than in the U.S. Table 7 presents regression analysis that extends our base analysis by including indicator variables for the use of dividend protected option compensation, RSGs, or both, as well as interaction terms between these indicator variables and incentive compensation. We estimate Tobit models where Models (1)–(4) extend our base analysis using dividend payout as the dependent variable; Models 17 18

Our results are similar if we only include observations where incentive compensation is greater than zero. We thank an anonymous referee for this suggestion.

N. Burns et al. / Journal of Corporate Finance 30 (2015) 85–97

93

Table 7 Multivariate analysis of dividend protection. Table 7 reports coefficients from Tobit regression analyses with the dependent variable censored at zero. The dependent variable in Columns 1–4 is dividends to total assets and in Columns 5–8 is repurchases to total assets. We use robust standard errors clustered by firm and include industry and year fixed effects. ADRI is the “Spamann Antidirector Rights Index” from Spamann (2010) and measures the level of shareholder protection. Options, RSG, and Either Protected are indicator variables that are equal to one if the company offers dividend protection on its options compensation, RSG compensation, or either options or RSG compensation, respectively, and zero otherwise. Options × Options Protected, RSG × RSG Protected, and Incentive × Either Protected are interactions formed by multiplying Options Ratio, RSG Ratio, and Incentive Ratio by the Options Protected, RSG Protected, or Either Protected indicator variables, respectively. All other variables are defined in Tables 2 and 3. P-values are reported below the coefficient. ***, **, and * indicate significance at the 1, 5, and 10% levels, respectively. Dividends (1) Incentive Ratio

Repurchases (2)

−0.0300*** (0.0000) −0.0342*** (0.0000)

0.0500** (0.0147) 0.0160* (0.0955) −0.0001 (0.9616)

Options Protected RSG Protected

InstPerOwned Q LogAssets DebtRatio CashRatio ROA SDRet DTP

Intercept R2 Obs.

(7)

(8)

0.0487*** (0.0000)

0.0217** (0.0254) 0.0473*** (0.0000)

0.0237** (0.0152)

0.0004 (0.8425)

−0.0426 (0.1557)

0.0470** (0.0231) 0.0123 (0.2034) −0.0003 (0.9319) 0.0008 (0.7092)

−0.0015 (0.4841) −0.0029** (0.0152) −0.0000 (0.3368) 0.0012* (0.0994) 0.0031*** (0.0000) −0.0388*** (0.0000) 0.0092* (0.0842) 0.2252*** (0.0000) −0.1433*** (0.0000) 0.0259*** (0.0000)

−0.0026** (0.0335) −0.0000 (0.1593) 0.0011 (0.1127) 0.0029*** (0.0000) −0.0376*** (0.0000) 0.0094* (0.0795) 0.2257*** (0.0000) −0.1442*** (0.0000) 0.0271*** (0.0000)

−0.0030** (0.0115) −0.0000 (0.2626) 0.0011 (0.1402) 0.0030*** (0.0000) −0.0382*** (0.0000) 0.0093* (0.0808) 0.2259*** (0.0000) −0.1447*** (0.0000) 0.0273*** (0.0000)

−0.0029** (0.0180) −0.0000 (0.3009) 0.0012 (0.1024) 0.0031*** (0.0000) −0.0387*** (0.0000) 0.0094* (0.0780) 0.2251*** (0.0000) −0.1434*** (0.0000) 0.0260*** (0.0000)

−0.0211 (0.3065) −0.0193*** (0.0017) 0.283 14,191

−0.0273 (0.1986) −0.0213*** (0.0006) 0.282 14,191

−0.0234 (0.2543) −0.0197*** (0.0015) 0.282 14,191

−0.0200 (0.3278) −0.0196*** (0.0014) 0.283 14,191

Dividends Repurchases

(6)

0.0063 (0.7541)

RSG × RSG Protected

ADRI

−0.0285*** (0.0000) −0.0322*** (0.0000)

0.0300*** (0.0005)

Options × Options Protected

(5) 0.0341*** (0.0000)

RSG Ratio

Either Protected

(4)

−0.0332*** (0.0000)

Options Ratio

Incentive × Either Protected

(3)

0.0023 (0.6269)

−0.0348 (0.2162) 0.0120 (0.6939) −0.0064 (0.2632) 0.0039 (0.3921)

0.0096 (0.7565) −0.0026 (0.6619)

−0.0018 (0.6663) −0.0012 (0.6003) 0.0001** (0.0195) 0.0016 (0.1184) 0.0080*** (0.0000) −0.0494*** (0.0000) 0.0363*** (0.0001) 0.0790*** (0.0000) −0.0849*** (0.0000) −0.0001 (0.9854) 0.0312 (0.5113)

−0.0016 (0.4629) 0.0001*** (0.0091) 0.0018* (0.0878) 0.0085*** (0.0000) −0.0520*** (0.0000) 0.0369*** (0.0001) 0.0793*** (0.0000) −0.0840*** (0.0000) −0.0025 (0.6934) 0.0216 (0.6540)

−0.0009 (0.6968) 0.0001** (0.0151) 0.0018* (0.0905) 0.0080*** (0.0000) −0.0493*** (0.0000) 0.0365*** (0.0001) 0.0778*** (0.0000) −0.0831*** (0.0000) −0.0013 (0.8383) 0.0307 (0.5177)

−0.0011 (0.6173) 0.0001** (0.0174) 0.0017 (0.1075) 0.0079*** (0.0000) −0.0487*** (0.0000) 0.0359*** (0.0001) 0.0785*** (0.0000) −0.0835*** (0.0000) −0.0001 (0.9813) 0.0345 (0.4660)

−0.0734*** (0.0000) 0.122 14,191

−0.0711*** (0.0000) 0.121 14,191

−0.0736*** (0.0000) 0.122 14,191

−0.0733*** (0.0000) 0.123 14,191

(5)–(8) use repurchases. Model (1) considers total incentive compensation and includes an indicator variable if either of the components of incentive compensation (options compensation or RSG) is dividend protected, as well as the interaction between incentive compensation and the dividend protection indicator. Models (2) and (3) repeat the analysis using options compensation and RSGs, respectively, and Model (4) includes both option compensation and RSG jointly. The results show that dividend protected option compensation and restricted stock is positively related to dividend payout. Model (1) shows that dividend protected incentive compensation is associated with a higher dividend payout. When we separate the components of equity compensation into options and restricted stock in Models (2) and (3), we find their interaction with protection is positive and significant, although in the joint estimation (shown in Model (4)), the interaction for restricted stock becomes insignificant. We also examine whether dividend protection has any relationship to repurchase payouts, and find no significant relationship. Our findings considering dividend protection suggest that using protection positively affects dividend payouts; however the overall association of both option and restricted stock compensation on dividends (repurchases) is reliably negative (positive). Thus our result that restricted stock is negatively associated with dividend payout in Europe, while prior research shows it is positively associated with dividends in the U.S., may be partially explained by the relatively high level of dividend protection on restricted stock in the U.S.

94

N. Burns et al. / Journal of Corporate Finance 30 (2015) 85–97

3.3. Dividend investor protection and growth opportunities We explore whether incentive effects of compensation on payout policy are altered by investor protection, growth opportunities, and their joint effect. Firms from countries with weak protection may pay higher dividends to maintain a reputation for better treatment of minority shareholders, as posited by LLSV (2000). Firms with higher growth options would be expected to prefer to pay lower dividends and instead use the cash to fund opportunities—this negative relation between growth opportunities and dividends is expected to be stronger in weak protection countries in which raising capital is more difficult. However, paying lower dividends conflicts with dividends being used to bond firms in weak investor protection countries. To overcome weak investor protection, high growth firms might pay their executives with more equity-based pay to align their interests with that of shareholders. Table 8 presents regression analysis that examines the interplay between investor protection and growth opportunities. In Models (1) through (4) the dependent variable is the dividend measure, and in Models (5) through (8) the dependent variable is the repurchases measure. Each of these models uses Tobit estimations. We focus on incentive compensation because, as shown in Tables 5 and 6 and our discussion of dividend protection, restricted stock and options do not have significantly different effects on payout policy. Following Pinkowitz (Pinkowitz et al., 2006), we examine the role of investor protection by separating the sample at the median level of investor protection. In Model (1) we include an indicator variable equal to one (Low ADRI) when investor protection is below the median level and interact it with incentive compensation. In Model (2) we explore whether growth options alters the effect of equity-incentive compensation on dividend payouts. The level of incentive pay is expected to be higher at firms with higher growth options to incentivize management to take advantage of the opportunities. Empirically, we test for this by including an indicator variable equal to one for firms with growth options higher than the median (High Q) and interact this with incentive Table 8 Multivariate analysis of payouts, growth, and investor protection. Table 8 reports coefficients from Tobit regression analyses with the dependent variable censored at zero. The dependent variable in Columns 1–4 is dividends to total assets and in Columns 5–8 is repurchases to total assets. We use robust standard errors clustered by firm and include industry and year fixed effects. ADRI is the “Spamann Antidirector Rights Index” from Spamann (2010) and measures the level of shareholder protection. Low ADRI is a dummy variable that equals 1 if ADRI is below the median ADRI of the sample and 0 otherwise. High Q is a dummy variable that equals 1 if Q is above the median Q of the sample and 0 otherwise. Incentive × Low ADRI and Incentive × High Q are interactions formed by multiplying Incentive Compensation by the Low ADRI or High Q interaction variables, respectively. All other variables are defined in Tables 2 and 3. Columns (3) and (7) are subsamples with firms in countries with Low ADRI (the Low ADRI dummy equals 1). Columns (4) and (8) are subsamples with firms in the remaining countries (the Low ADRI dummy equals 0). P-values are reported below the coefficient. ***, **, and * indicate significance at the 1, 5, and 10% levels, respectively. Dividends

Incentive Pay Ratio Incentive × Low ADRI

(2)

(3)

(4)

(5)

(6)

(7)

(8)

−0.0216*** (0.0000) −0.0046 (0.4846)

−0.0117*** (0.0019)

−0.0053 (0.3987)

−0.0142*** (0.0002)

0.0040 (0.6214) 0.0486*** (0.0001)

0.0037 (0.6605)

0.0222* (0.0790)

−0.0124 (0.1940)

−0.0204*** (0.0003)

−0.0266*** (0.0014)

−0.0163** (0.0181)

0.0450*** (0.0002)

0.0405** (0.0150)

0.0301** (0.0291)

0.0039 (0.2575)

0.0013 (0.5421)

0.0001** (0.0342) 0.0016 (0.1218) 0.0080*** (0.0000) −0.0474*** (0.0000) 0.0374*** (0.0001) 0.0777*** (0.0000) −0.0867*** (0.0000) −0.0084 (0.1745) 0.0407 (0.3891)

0.0028 (0.1711) −0.0009 (0.6647) 0.0001** (0.0188)

0.0002* (0.0604)

0.0001 (0.2634)

0.0080*** (0.0000) −0.0473*** (0.0000) 0.0378*** (0.0000) 0.0761*** (0.0001) −0.0798*** (0.0000) −0.0008 (0.8972) 0.0429 (0.3998)

0.0074*** (0.0000) −0.0447*** (0.0000) 0.0447*** (0.0026) 0.0764** (0.0231) −0.0952*** (0.0001) −0.0256** (0.0291) −0.0725 (0.3013)

0.0082*** (0.0000) −0.0421*** (0.0000) 0.0338*** (0.0002) 0.0648*** (0.0000) −0.0633*** (0.0002) 0.0018 (0.8328) 0.2245*** (0.0011)

−0.0630*** (0.0000) 0.125 14,191

−0.0675*** (0.0000) 0.121 14,227

−0.0594*** (0.0001) 0.097 6830

−0.0917*** (0.0000) 0.150 7361

Incentive × High Q Low ADRI

0.0003 (0.8328)

0.0006 (0.6202)

−0.0000 (0.2675) 0.0011 (0.1114) 0.0031*** (0.0000) −0.0386*** (0.0000) 0.0095* (0.0741) 0.2252*** (0.0000) −0.1435*** (0.0000) 0.0263*** (0.0000)

0.0003 (0.7267) −0.0030** (0.0116) −0.0000 (0.3170)

−0.0000 (0.2660)

−0.0000 (0.6064)

0.0031*** (0.0000) −0.0399*** (0.0000) 0.0112** (0.0364) 0.2315*** (0.0000) −0.1434*** (0.0000) 0.0265*** (0.0000)

0.0038*** (0.0000) −0.0451*** (0.0000) 0.0293*** (0.0005) 0.2633*** (0.0000) −0.1717*** (0.0000) 0.0459*** (0.0000)

0.0024*** (0.0000) −0.0350*** (0.0000) −0.0027 (0.6693) 0.2001*** (0.0000) −0.1205*** (0.0000) 0.0068 (0.1905)

−0.0194 (0.3407) −0.0343*** (0.0000) 0.282 14,191

−0.0141 (0.4862) −0.0185*** (0.0023) 0.284 14,191

−0.1046*** (0.0006) −0.0552*** (0.0000) 0.266 6830

0.0835*** (0.0012) −0.0116 (0.1363) 0.306 7361

ADRI

Q LogAssets DebtRatio CashRatio ROA SDRet DTP

−0.0112*** (0.0001)

0.0040*** (0.0074)

High Q

InstPerOwned

Repurchases

(1)

Dividends Repurchases Intercept R2 Obs.

N. Burns et al. / Journal of Corporate Finance 30 (2015) 85–97

95

compensation.19 In Models (3) and (4), we segment the sample into low (Model 3) and high investor protection (Model 4), and examine the impact of high growth opportunities on the relationship between incentive compensation and dividend payouts. Across all models presented in Table 8, equity Incentive compensation is negatively associated with dividend payouts. The coefficient on the Low ADRI indicator in Model (1) shows that countries with lower investor protection have significantly higher dividend payout.20 However, the insignificant coefficient on the interaction between low investor protection and incentive compensation implies that the sensitivity of payouts to incentive compensation's does not change depending on whether a company is in a high or low investor protection country. In Model (2) the main explanatory variables of interest are the High Q indicator and it's interaction with incentive compensation. The results show that when CEOs of growth firms have more equity incentives, dividend payouts are reduced at a faster rate than for lower growth firms. In Models (3) and (4) we explore the interplay among weak investor protection, growth opportunities, and incentive compensation. The results show that high growth firms with more incentive compensation pay lower dividends, in both low and high investor protection countries. When we test the difference between the interaction of incentive compensation and High Q between Models (3) and (4), we find that the coefficient for weak protection countries is greater than for strong protection countries at 5% significance. This suggests that in countries with weak investor protection, high growth firms with more incentive compensation reduce dividends at a faster rate than firms in strong investor protection countries. This is consistent with equity incentives serving to align manager's interest with shareholders, thereby enabling growth firms in weak protection countries to pay lower dividends and maintain financial flexibility. Models (5) through (8) repeat these regressions but with repurchases as the dependent variable. Model (5) shows that companies in weak investor protection countries have lower repurchases; however firms in weak investor protection countries increase repurchases at a faster rate in responses to an increase in incentive compensation. We also find that high growth firms with high incentive compensation increase repurchases at a greater rate in responses to an increase in incentive compensation (Model (6)). Models (7) and (8) show that this relationship is stronger for firms in weak protection countries; the difference in the interaction coefficients between Models (7) and (8) is significant at 10%. We next explore the use of incentive compensation and dividend protection in our separated samples. Using dividend protection might counteract the desired tendency of CEOs to favor lower dividend payouts with greater use of incentive compensation, therefore we examine the use of dividend protection between our subsamples of high and low growth firms. Consistent with growth firms greater incentives to reduce dividends, we find that growth firms use less dividend protection of incentive compensation: 16.7% of growth firms dividend protect incentive compensation compared to 25.8% for the remainder of the sample. Taken together, our analysis of growth and investor protection suggests that high growth companies in weak investor protection countries may use greater equity incentive compensation to align manager interests with shareholders. The equity-incentive compensation enables the firm to reduce dividend payouts and use repurchases to distribute excess free cash flow, thereby increasing financial flexibility in markets in which it is likely harder to raise capital.

3.4. Robustness tests As is the case with much of corporate finance research, the issue of causality may exist. Incentive compensation may affect pay out policy or dividend policy may drive the use of incentive compensation. We deal with these potential endogeneity concerns in two ways. First, in the regressions reported in Tables 5–8, the independent variables, including the compensation variables, are lagged. Second, we consider refitting our primary regression using two-stage least squares with instruments for our compensation variables.21 Table 9 presents this additional robustness check. In Columns (1)–(4), we present the results from our two-stage least squares estimation for dividend payout; in the first stage we use the average incentive compensation in the industry of each country each year as the instrument.22 We follow the method employed by John and Knyazeva (2006), where the industry structure is considered unique to each industry and therefore is expected to be exogenous. The intuition is that the average compensation in the industry is positively associated with an individual firm's compensation, and not likely to have a direct impact on individual payout policy. Consistent with this argument, we find that the average industry incentive compensation (or options compensation or RSG) is highly significant in the first stage regressions (untabulated) predicting incentive compensation. Thus, identification does not reject the null hypothesis that the instruments are valid. As the table shows, we continue to find a negative relation between dividend payout and (predicted) compensation, and a positive relation between repurchase payout and (predicted) compensation, though the coefficients with dividend payouts become insignificant.

4. Conclusion We use a cross-country sample of European firms to explore the role of equity-incentive compensation and investor protection on payouts. We examine whether equity-incentive compensation and dividends act as substitutes for reducing agency costs, and further, 19

In an alternate specification we use an indicator variable equal to one for firms with Q higher than the median Q in its industry, and our results are robust. This result conflicts with Alzahrani and Lasfer (2012) who find that dividends are lower in weak protection countries. However our sample consists only of developed European countries and excludes the U.S. If we include the US in our sample, our results are similar to Alzahrani and Lasfer (2012). 21 Determining exogenous factors to control for the relationships being studied is inexact. 22 We exclude the own firm from the calculation of the average industry return and thus lose a few observations where the own firm does not have any industry peers in the given year. 20

96

N. Burns et al. / Journal of Corporate Finance 30 (2015) 85–97

Table 9 Two-stage instrumental Tobit analysis of payouts. Table 9 reports coefficients from repeating the regressions in Columns (2) and (6) of Tables 5 and 6, respectively, with Tobit two-stage regression analyses. Average options compensation, RSG compensation, and total equity incentive compensation of all firms in the same industry, year, and country are used as instruments for the firm's incentive compensation. The firm of interest is excluded from the calculation of the instruments and observations without industry peers for the given year and country are dropped. The original coefficients from Tables 5 & 6 are reported alongside for convenience. We use robust standard errors clustered by firm and include industry and year fixed effects in the second-stage. All other variables are defined in Tables 5 and 6. P-values are reported below the coefficient. ***, **, and * indicate significance at the 1, 5, and 10% levels, respectively. F-values to test the strength of the instrument from the first stage regressions are reported in the last six rows of the table. F-values labeled as “w/CE” are calculated using robust standard errors, clustered by firm. Dividends

Incentive Ratio

Repurchases

Europe

Europe Col 2, Table 5

Europe

Europe Col 6, Table 5

(1)

(2)

(3)

(4)

−0.0123 (0.6212)

−0.0252*** (0.0000)

Options Ratio

−0.0026** (0.0259) −0.0000 (0.3163) 0.0010 (0.1634) 0.0029*** (0.0000) −0.0379*** (0.0000) 0.0092* (0.0831) 0.2235*** (0.0000) −0.1423*** (0.0000) 0.0269*** (0.0000)

−0.0030** (0.0133) −0.0000 (0.3059) 0.0011 (0.1078) 0.0031*** (0.0000) −0.0386*** (0.0000) 0.0095* (0.0759) 0.2254*** (0.0000) −0.1434*** (0.0000) 0.0262*** (0.0000)

0.1283 (0.2304) −0.1150 (0.1370) −0.0038** (0.0115) −0.0000 (0.3725) 0.0008 (0.2913) 0.0030*** (0.0000) −0.0376*** (0.0000) 0.0075 (0.2022) 0.2248*** (0.0000) −0.1464*** (0.0000) 0.0302*** (0.0000)

−0.0205 (0.3972) −0.0205*** (0.0011) 0.216 14,045

−0.0206 (0.3123) −0.0197*** (0.0013) 0.283 14,191

−0.0072 (0.7944) −0.0180** (0.0106) 0.216 14,045

RSG Ratio ADRI Inst % Owned Q LogAssets DebtRatio CashRatio ROA SDRet DTP

−0.0225*** (0.0002) −0.0285*** (0.0000) −0.0030** (0.0126) −0.0000 (0.3174) 0.0011 (0.1121) 0.0031*** (0.0000) −0.0387*** (0.0000) 0.0095* (0.0759) 0.2254*** (0.0000) −0.1434*** (0.0000) 0.0263*** (0.0000)

−0.0203 (0.3171) −0.0196*** (0.0014) 0.283 14,191

Dividends Repurchases Intercept R2 Obs. 1st. Stg. F-Values: Options Ratio Options Ratio—w/CE RSG Ratio RSG Ratio—w/CE Incentive Ratio Incentive Ratio—w/CE

Europe

Europe Col 2, Table 6

Europe

Europe Col 6, Table 6

(5)

(6)

(7)

(8)

0.1816*** (0.0003)

0.0351*** (0.0000) 0.0184** (0.0472) 0.0516*** (0.0000) −0.0009 (0.6724) 0.0001** (0.0211) 0.0017 (0.1075) 0.0079*** (0.0000) −0.0488*** (0.0000) 0.0364*** (0.0001) 0.0781*** (0.0000) −0.0841*** (0.0000) −0.0002 (0.9711) 0.0334 (0.4818) −0.0897*** (0.0021) −0.0488*** (0.0000) 0.273 8282

−0.0005 (0.8333) 0.0000 (0.5528) 0.0011 (0.3288) 0.0062*** (0.0000) −0.0382*** (0.0000) 0.0332*** (0.0009) 0.0802*** (0.0000) −0.0951*** (0.0000) 0.0120 (0.1349) 0.0834 (0.1340)

−0.0011 (0.6057) 0.0001** (0.0209) 0.0016 (0.1203) 0.0080*** (0.0000) −0.0493*** (0.0000) 0.0364*** (0.0001) 0.0791*** (0.0000) −0.0848*** (0.0000) 0.0000 (0.9984) 0.0316 (0.5057)

−0.1473 (0.4251) 0.4128*** (0.0063) 0.0022 (0.4744) 0.0000 (0.6085) 0.0015 (0.2428) 0.0059*** (0.0000) −0.0388*** (0.0000) 0.0371*** (0.0010) 0.0762*** (0.0006) −0.0841*** (0.0000) 0.0035 (0.7204) 0.0854 (0.1723)

−0.0816*** (0.0000) 0.112 14,045

−0.0735*** (0.0000) 0.122 14,191

−0.0875*** (0.0000) 0.112 14,045

42.1*** 12.4*** 89.7*** 8.4*** 217.9*** 35.6***

43.4*** 12.7*** 94.4*** 9.0*** 227.9*** 38.0***

how growth opportunities alters the relationship between investor protection, incentive compensation, and payouts. We also explore the role of dividend protection in corporate payout policy. In general, we find that firms in weak investor protection countries use more incentive compensation and repurchase less often, consistent with these firms using dividends to establish a favorable reputation with minority shareholders for distributing excess free cash flow. However, use of equity-incentive compensation is negatively (positively) associated with dividends (repurchases), and this relationship is stronger in firms from weak protection countries. Considered together, our findings are consistent with growth firms as well as firms in weak investor protection countries using equity incentive compensation in lieu of using dividends as a tool to reduce agency costs. While we propose that the primary effect of incentive compensation is to reduce agency costs, we also examine whether options and restricted stock have differing effects on payout policy in Europe and whether dividend protection alters these effects. In contrast to research findings in the U.S. (Aboody and Kasznik, 2008), we show that options and restricted stock have similar effects on payout policy in European firms. When we consider dividend protection, we find that on average dividend protected incentive compensation reduces the negative association between incentive compensation and dividends. Coupled with recent U.S. studies, our findings for dividend protection across Europe suggest that the positive relationship between restricted stock and dividend payout in the U.S. may be associated with the greater use of dividend protection for restricted stock in the US.

N. Burns et al. / Journal of Corporate Finance 30 (2015) 85–97

97

Although still far less than U.S. CEOS, firms in other countries are increasingly adopting U.S. style incentive compensation. The results presented in this paper provide interested parties with a better understanding of how compensation can have a broader effect on corporate pay-out policy and managerial behavior. Acknowledgments The authors thank John Wald, Len Rosenthal, and seminar participants at the 2013 Financial Management Association meeting, Bentley University, and Washington State University. Brian Anderson provided excellent research assistance. Burns appreciates support from the UTSA College of Business summer grant program. References Aboody, D., Kasznik, R., 2008. Executive stock-based compensation and firms' cash payout: the role of shareholders' tax-related payout preferences. Rev. Acc. Stud. 13, 216–251. Almeida, H., Campello, M., Weisbach, M., 2004. The cash flow sensitivity of cash. J. Financ. 59, 1777–1804. Alzahrani, M., Lasfer, M., 2012. Investor protection, taxation, and dividends. J. Corp. Financ. 18, 745–762. Andrews, D., 1991. Heteroskedasticity and autocorrelation consistent covariance matrix estimation. Econometrica 59, 817–858. Arnold, M.C., Gillenkirch, R.M., 2005. Stock options and dividend protection. J. Inst. Theor. Econ. 161 (3), 453–472. Boyle, G., Guthrie, G., 2003. Investment, uncertainty and liquidity. J. Financ. 58, 2143–2166. Brown, J., Liang, N., Weisbenner, S., 2007. Executive financial incentives and payout policy: firm responses to the 2003 dividend tax cut. J. Financ. 62, 1935–1965. Cesari, A., Ozkan, N., 2014. CEO incentives and payout policy: empirical evidence from Europe. Working Paper. Cheffins, B.R., 2003. Will executive pay globalise along American lines? Corporate Governance: An International Review 11, 8–24. Chetty, R., Saez, E., 2005. Dividend taxes and corporate behavior: evidence from the 2003 dividend tax cut. Q. J. Econ. 3, 791–833. Conyon, M., Fernandes, N., Ferreira, M., Matos, P., Murphy, K., 2013. The executive compensation controversy: a transatlantic analysis (ICS 2011-002). Retrieved September 15, 2013 from Cornell University, ILR School, Institute for Compensation Studies site. http://www.digitalcommons.ilr.cornell.edu/ics/. Croci, E., Gonenc, H., Ozkan, N., 2012. CEO compensation, family control, and institutional investors in continental Europe. J. Bank. Financ. 36 (2), 3318–3335. Denis, D., Osobov, I., 2008. Disappearing dividends, catering incentives and agency costs: international evidence. J. Financ. Econ. 89, 62–82. Durnev, A., Kim, E., 2005. To steal or not to steal: firm attributes, legal environment, and valuation. J. Financ. 60, 1461–1493. Eije, H.E., Megginson, W., 2008. Dividends and share repurchases in the European Union. J. Financ. Econ. 89, 347–374. Fama, E., French, K., 2001. Disappearing dividends: changing firm characteristics or lower propensity to pay? J. Financ. Econ. 60, 3–43. Fenn, G., Liang, N., 2001. Corporate payout policy and managerial stock incentives. J. Financ. Econ. 60, 45–72. Grullon, G., Michaely, R., 2002. Dividends, share repurchases, and the substitution hypothesis. J. Financ. 57, 1649–1684. Jagannathan, M., Stephens, C., Weisbach, M., 2000. Financial flexibility and the choice between dividends and stock repurchases. J. Financ. Econ. 57, 355–384. Jensen, M., 1986. Agency costs of free cash flow, corporate finance and takeovers. Am. Econ. Rev. 76, 323–339. John, K., Knyazeva, A., 2006. Payout Policy, Agency Conflicts and Corporate Governance. SSRN Working Paper. Kahle, K., 2002. When a buyback isn't a buyback: open-market repurchases and employee options. J. Financ. Econ. 63, 235–261. La Porta, R., Lopez-de-Silanes, F., Shleifer, A., Vishny, R., 1997. Legal determinants of external finance. J. Financ. 52, 1131–1150. La Porta, R., Lopez-de-Silanes, F., Shleifer, A., Vishny, R., 1998. Law and finance. J. Polit. Econ. 106, 1113–1155. La Porta, R., Lopez-De-Silanes, F., Shleifer, A., Vishny, R.W., 2000. Agency problems and dividend policies around the world. J. Financ. 55, 1–33. Lambert, R., Lanen, W., Larcker, D., 1989. Executive stock option plans and corporate dividend policy. J. Financ. Quant. Anal. 24, 409–425. Lee, B., Suh, J., 2011. Cash holdings and share repurchase: international evidence. J. Corp. Financ. 17 (5), 1306–1329 (December). Liljeblom, E., Pasternack, D., 2006. Share repurchases, dividends and executive options: the effect of dividend protection. Eur. Financ. Manag. 12, 7–28. Minnick, K., Rosenthal, L., 2014. Stealth compensation: do CEOs increase their pay by influencing dividend policy? J. Corp. Financ. 25 (1), 435–454. Murphy, K.J., 1999. Executive compensation 2485–2563. In: Ashenfelter, Ο., Card, D. (Eds.), Handbook of Labour Economics vol. 3. Elsevier Science, Amsterdam. Opler, T., Pinkowitz, L., Stulz, R., Williamson, R., 1999. The determinants and implications of corporate cash holdings. J. Financ. Econ. 52, 3–46. Petersen, M.A., 2009. Estimating standard errors in finance panel data sets: comparing approaches. Rev. Financ. Stud. 22, 435–480. Pinkowitz, L., Stulz, R., Williamson, R., 2006. Do firms in countries with poor protection of investor rights hold more cash? J. Financ. 6, 2725–2751. Short, H., Keasey, K., Zhang, H., 2002. The link between dividend policy and institutional ownership. J. Corp. Financ. 8, 105–122. Spamann, H., 2010. The “Antidirector Rights Index” revisited. Rev. Financ. Stud. 23, 467–486. Winter, S., 2000. Optionspläne als Instrument wertorientierter Managementvergütung. Peter Lang, Frankfurt/Main. Zeileis, A., 2004. Econometric computing with HC and HAC covariance matrix estimators. J. Stat. Softw. 11, 1–17.