Shareholder-Creditor Conflict and Payout Policy: Evidence from Mergers between Lenders and Shareholders

Shareholder-Creditor Conflict and Payout Policy: Evidence from Mergers between Lenders and... Abstract This paper studies how the conflict of interest between shareholders and creditors affects corporate payout policy. Using mergers between lenders and equity holders of the same firm as shocks to the shareholder-creditor conflict, I find that firms pay out less when there is less conflict between shareholders and creditors, suggesting that the shareholder-creditor conflict induces firms to pay out more at the expense of creditors. The effect is stronger for firms in financial distress. Received March 22, 2017; editorial decision October 17, 2017 by Editor Wei Jiang. Authors have furnished an Internet Appendix, which is available on the Oxford University Press Web Site next to the link to the final published paper online. The conflict of interest between shareholders and creditors can induce agency costs in the form of excessive dividend payments, claim dilution, asset substitution, and underinvestment (Jensen and Meckling 1976; Smith and Warner 1979). Excessive dividend payments, in particular, may lead to significant wealth transfers from creditors to shareholders. Black (1976) points out that “there is no easier way for a company to escape the burden of a debt than to pay out all of its assets in the form of a dividend, and leave the creditors holding an empty shell.” While the theory on shareholder-creditor conflict is well established, the empirical relevance of the theory has been understudied. For example, Gilje (2016) finds that exacerbated shareholder-creditor conflict in financial distress causes firms to take less risk, a finding inconsistent with the theory. The same may apply to excessive dividend payment as well. How the shareholder-creditor conflict affects payout policy is therefore empirically important. On the other hand, the existing literature on payout policy has paid little attention to the shareholder-creditor conflict. In a recent survey on payout policy, Farre-Mensa, Michaely, and Schmalz (2014) reviewed no papers related to the shareholder-creditor conflict. In an earlier review article, Allen and Michaely (2003) reviewed only a handful of papers on the relationship between the shareholder-creditor conflict and payout policy, most of which show, at best, indirect evidence for the relevance of the shareholder-creditor conflict. In this paper, I provide direct evidence that the shareholder-creditor conflict affects payout policy. The lack of direct evidence is partly due to the difficulty of empirically measuring the shareholder-creditor conflict. Most existing literature relies on stock and bond price reactions to specific corporate events to infer the shareholder-creditor conflict (Asquith and Wizman 1990; Warga and Welch 1993; Billett, King, and Mauer 2004). Others explore variation in leverage as a proxy for changes in the shareholder-creditor conflict. However, capital structure decisions are often simultaneously determined as investment and payout decisions, and, hence, relying on variation in leverage alone is unlikely to uncover the causal relationship between shareholder-creditor conflict and corporate policy.1 One notable exception is Jiang, Li, and Shao (2010), who use the existence of dual holders, who simultaneously hold equity and debt claims of the same firm, to directly quantify the effect of the shareholder-creditor conflict. In this paper, I build on their idea and also measure the shareholder-creditor conflict with dual holders. Furthermore, I also try to isolate potentially exogenous variation of the conflict by exploiting mergers between shareholders and creditors of the same firm. Under the agency theory (Jensen and Meckling 1976; Myers 1977; Smith and Warner 1979), shareholders may pay excessive dividends at the expense of creditors to maximize shareholder value when the debt contract is in place. Creditors, anticipating potential wealth appropriation by shareholders, often put dividend constraints in the contract and demand higher yields. However, as pointed out by Easterbrook (1984) and John and Kalay (1982), putting overly restrictive dividend constraints often results in overinvestment, and therefore the optimal dividend constraints often do not restore the first-best payout level.2 In equilibrium, a firm still pays out more than the first best in the presence of the shareholder-creditor conflict. When a shareholder merges with a creditor of the same firm, the conflict between the shareholder and the creditor decreases, and, hence, the merger pushes the payout toward the first-best level. I first follow Jiang, Li, and Shao (2010) to identify firms with dual holders and examine the relationship between payout policy and the existence of dual holders using ordinary least squares (OLS) regressions. Consistent with the argument that dual holders mitigate the shareholder-creditor conflict, I find that firms with dual holders pay out less than firms without dual holders. To mitigate the potential endogeneity of dual holders, I next rely on mergers between shareholders and lenders of the same firms to generate plausibly exogenous variation in shareholder-creditor conflict. To construct the sample of mergers between lenders and shareholders, I first identify all mergers between financial firms in SDC and then match the names of the acquirers and targets with lender names in DealScan and shareholder names on Form 13F. I then identify firms who are borrowers of the merging lender and whose stocks are held by the merging institutional shareholder. I further require that the institutional investors hold more than 1% of all shares outstanding at the time of the merger and the lender is allocated more than 10% of the loan at origination. These firms are designated as treated firms. For each treated firm, I then find control firms by matching on firm size, Tobin’s q, institutional ownership, and leverage, and at the same time require the control firms to also have bank loans outstanding. In a difference-in-differences framework with a 6-year window, I find that treated firms reduce payout, as measured by total payout, share repurchases, or cash dividends, relative to control firms after the mergers. The result is consistent with the argument that the shareholder-creditor conflict results in wealth transfers from creditors to shareholders in the form of excessive dividend payout, and the merger between a lender and a shareholder of the same firm aligns the interests of the two and therefore leads to lower payout. The identification of the difference-in-differences estimation relies on the parallel trend condition, that is, the outcome variables have parallel trends in the absence of treatment. While the parallel trend condition is untestable, I follow the advice of Roberts and Whited (2012) to examine the dynamics of the effect of the mergers on payout policy. If the baseline results are driven by nonparallel trends between treated and control firms, the results are likely to show up before the mergers. In contrast, I find that the effect appears only after the mergers, suggesting that the baseline results are unlikely to be driven by nonparallel trends of treated and control firms. The alignment of the interest of shareholders and creditors should increase with the stakes the lenders have in the treated firm. To this end, I find that the strength of the negative effect of the merger on payout increases with the size of the loan allocated to the merging lender. After merging with the lender, the merging shareholder shifts toward maximizing the combined value of equity and loans held, which may then create a conflict between the merging shareholder and other shareholders, whose goal is to maximize the value of equity only. Hence, the extent to which the merging shareholder is able to affect corporate policy will depend on the merging shareholder’s relative power. I measure the relative power as the ratio between the shares held by the merging shareholder and other institutional shareholders, and find that the negative effect of the mergers on payout concentrates in cases in which the merging shareholder is more powerful. Further analysis shows that the effect comes mostly from firms whose loans do not contain dividend restriction covenants. Dividend restriction covenants mitigate the agency cost of excessive payout, and therefore, the mergers, while mitigating the shareholder-creditor conflict, add little value in reducing corporate payout. To ensure that the result is indeed driven by reduced shareholder-creditor conflict, I further explore whether the effect of the merger is stronger for firms in financial distress. Using leverage as a measure of financial distress, I find that the negative effect of the mergers on payout is stronger for firms in financial distress. The result further suggests that the mergers between shareholders and creditors affect payout policy via its impact on the shareholder-creditor conflict. This paper contributes to the literature on payout policy in the presence of shareholder-creditor conflict. Smith and Warner (1979) find that bonds often contain covenants restricting dividend payments due to shareholder-creditor conflict. John and Kalay (1982) derive optimal payout constraints in the presence of shareholder-creditor conflict. Kalay (1982) provides empirical evidence that payout constraints are set to prevent wealth transfer from debt holders to shareholders. Brockman and Unlu (2009) find that creditors demand a more restrictive payout policy ex ante when creditors have weaker rights ex post. Other papers mostly rely on stock and bond price reactions to unexpected dividend changes to infer the relationship between the shareholder-creditor conflict and payout (Handjinicolaou and Kalay 1984; Jayaraman and Shastri 1988; Dhillon and Johnson 1994). This paper also adds to the recent literature on the effect of dual holders. Jiang, Li, and Shao (2010) find that dual holders lower loan spreads, suggesting that dual holders help mitigate shareholder-creditor conflict. Chava,Wang, and Zou (Forthcoming) find that dual holders reduce the use of covenants restricting capital expenditure, and in the event of covenant violation, firms with dual holders are unlikely to suffer a significant drop in debt issuance or investment expenditure. Bodnaruk and Rossi (2016) find that the existence of dual holders of target firms in M&A deals results in higher merger premiums and larger abnormal bond returns. 1. Sample Construction and the Identification Strategy 1.1 Sample construction The sample construction begins with all mergers between financial firms from 1987 to 2011 in the SDC mergers and acquisitions database. I begin the merger sample from 1987 because this is when the DealScan database starts to have a comprehensive coverage of loans. I stop the sample at 2011 because I need 3 years of data after the merger in the analysis. In the second step, I obtain lenders’ information from the LPC DealScan database, and match the lender names with the names of either the acquirers or the targets of the financial mergers. In matching acquirer names, I not only match the names of the lenders directly involved in the merger but also match the names of the parent companies of the lenders and acquirers. Wherever possible, I use the addresses of the companies in both databases to facilitate the match. After this step, I retain all mergers for which either the acquirer or the target can be matched with a lender in the DealScan database. In the third step, I obtain institutional investors’ information from the Thomson Reuter’s 13F database, and match the investors’ names with the unmatched acquirer or target names from the last step. Again, I not only match the names of companies directly involved but also match the names of their parent companies for acquirers. All matches are manually checked to ensure accuracy. This procedure produces a sample of 369 mergers between a lender in the DealScan database and an institutional investor in the 13F database. The next step is to identify treated firms. I first identify all firms whose loans from the merging lenders are still outstanding at the time of the merger. I then require that the merging institutional investor holds stocks of the firm at the end of the quarter immediately before the merger. I require that the lender participates more than 10% of the loan at origination and the institutional shareholder holds more than 1% of all shares outstanding of the firm.3 I exclude all cases in which either the acquirer or the target is a dual holder of the firm before the merger. For firms treated multiple times, I only retain the first time they are treated. I also exclude firms treated again in less than 3 years after receiving the first treatment. I then exclude firms in financial and utility industries and firms with missing key variables. Finally, I only retain treated firms for which key variables are available 1 year before and 1 year after the mergers. This procedure produces a sample of 238 treated firms involved in 61 mergers. I lose some observations when moving away from the treatment date. The number of treated firms is 236 three years before the mergers and 211 three years after the merger. On average, each merger affects about four firms. The distribution of the mergers across time is presented in Table 1. The mergers are fairly evenly distributed across time, with year 2006 having the greatest number of mergers (ten). Table 1 Distribution of mergers between lenders and institutional shareholders Year Freq. % Cum. 1990 1 1.64 1.64 1991 1 1.64 3.28 1995 1 1.64 4.92 1996 2 3.28 8.2 1997 3 4.92 13.11 1998 4 6.56 19.67 1999 2 3.28 22.95 2000 1 1.64 24.59 2001 3 4.92 29.51 2002 6 9.84 39.34 2003 1 1.64 40.98 2004 2 3.28 44.26 2005 4 6.56 50.82 2006 10 16.39 67.21 2007 1 1.64 68.85 2008 5 8.2 77.05 2009 7 11.48 88.52 2010 3 4.92 93.44 2011 4 6.56 100 Total 61 100 Year Freq. % Cum. 1990 1 1.64 1.64 1991 1 1.64 3.28 1995 1 1.64 4.92 1996 2 3.28 8.2 1997 3 4.92 13.11 1998 4 6.56 19.67 1999 2 3.28 22.95 2000 1 1.64 24.59 2001 3 4.92 29.51 2002 6 9.84 39.34 2003 1 1.64 40.98 2004 2 3.28 44.26 2005 4 6.56 50.82 2006 10 16.39 67.21 2007 1 1.64 68.85 2008 5 8.2 77.05 2009 7 11.48 88.52 2010 3 4.92 93.44 2011 4 6.56 100 Total 61 100 This table presents the yearly distribution of the mergers used in this paper. The mergers are merger and acquisition deals between lenders in the DealScan database and institutional shareholders in the Thomson Reuter’s 13F database. Table 1 Distribution of mergers between lenders and institutional shareholders Year Freq. % Cum. 1990 1 1.64 1.64 1991 1 1.64 3.28 1995 1 1.64 4.92 1996 2 3.28 8.2 1997 3 4.92 13.11 1998 4 6.56 19.67 1999 2 3.28 22.95 2000 1 1.64 24.59 2001 3 4.92 29.51 2002 6 9.84 39.34 2003 1 1.64 40.98 2004 2 3.28 44.26 2005 4 6.56 50.82 2006 10 16.39 67.21 2007 1 1.64 68.85 2008 5 8.2 77.05 2009 7 11.48 88.52 2010 3 4.92 93.44 2011 4 6.56 100 Total 61 100 Year Freq. % Cum. 1990 1 1.64 1.64 1991 1 1.64 3.28 1995 1 1.64 4.92 1996 2 3.28 8.2 1997 3 4.92 13.11 1998 4 6.56 19.67 1999 2 3.28 22.95 2000 1 1.64 24.59 2001 3 4.92 29.51 2002 6 9.84 39.34 2003 1 1.64 40.98 2004 2 3.28 44.26 2005 4 6.56 50.82 2006 10 16.39 67.21 2007 1 1.64 68.85 2008 5 8.2 77.05 2009 7 11.48 88.52 2010 3 4.92 93.44 2011 4 6.56 100 Total 61 100 This table presents the yearly distribution of the mergers used in this paper. The mergers are merger and acquisition deals between lenders in the DealScan database and institutional shareholders in the Thomson Reuter’s 13F database. Next, I follow a similar procedure like in Hong and Kacperczyk (2010) to find control firms. First, I exclude all firms ever treated by the mergers (not only those treated firms identified above). Second, I require all control firms to also have bank loans outstanding at the time of the mergers. Third, I exclude all firms with dual holders during the fiscal years of $$[t-3, t+3]$$, in which year $$t$$ is the year during which the merger occurred. Finally, I require control firms to be in the same quintiles sorted based on total assets, Tobin’s q, leverage, and the percentage of institutional ownership. I match control firms based on total assets and Tobin’s q because they are important determinants of payout policy. I match control firms based on leverage and institutional ownership because treated firms, by construction, have debt in their capital structure and are owned by institutional shareholders. I then rank the control firms based on the differences of total assets, Tobin’s q, leverage, and institutional ownership relative to their corresponding treated firms. I compute the rank of the differences for each of these four variables, and then compute the total rank across all four variables. I retain control firms with the five lowest total ranks. That is, for each treated firm, I retain at most five control firms based on their total ranks. This procedure produces a sample of 894 control firms. The number of control firms is 762 three years before the mergers and 705 three years after the merger. The empirical methodology requires specifying a testing window. In choosing the appropriate time window, the trade-off is between a long window that may incorporate information unrelated to the mergers and a short window that contains too few observations. In the baseline specification, I choose a 6-year window, that is, 3 years before and 3 years after the mergers. To ensure a clean identification, I discard firm fiscal years during which the mergers occurred. To ensure robustness, I also try 2-, 4-, and 10-year windows and find similar results. The final step of sample construction involves matching both treated and control firms in the sample with their financial information from Compustat, institutional ownership information from 13F, and detailed loan information from DealScan. 1.2 The identification strategy I use the mergers between lenders and institutional shareholders of the same firm as shocks to the shareholder-creditor conflict. When a lender and an institutional shareholder of the same firm merge, the conflict of interests between the lender and the shareholder is reduced. On the other hand, lenders often lend to hundreds of firms at each point in time and are therefore unlikely to make merger decisions based on factors related to one particular firm. Similarly, institutional shareholders also often hold stocks of many firms at each point in time and are also unlikely to pursue mergers based on factors related to one particular firm.4 As such, the mergers between lenders and institutional shareholders are likely to satisfy both the relevance and the exclusion conditions. To identify the effect of the shareholder-creditor conflict on payout policy, I adopt the difference-in-differences specification as follows: \begin{equation} Y_{\it it}=\alpha_{i}+\alpha_{t}+\beta {\it Treat}_{i} \times {\it Post}_{\it it} +\gamma X_{it-1} +\epsilon_{\it it}, \label{did} \end{equation} (1) where $$Y_{\it it}$$ is measures of payout of firm $$i$$ in year $$t$$; $${\it Treat}_{i}$$ equals one if firm $$i$$ is a treated firm, and zero otherwise; $${\it Post}_{\it it}$$ equals one if the firm year observation is after the announcement of the mergers; $$\alpha_{i}$$ is the firm fixed effects; $$\alpha_{t}$$ is the year fixed effects; and $$X_{\it it-1}$$ is a vector of control variables. In this specification, $${\it Treat}_{i}$$ and $${\it Post}_{\it it}$$ are subsumed by the firm fixed effects and the year fixed effects, respectively. The difference-in-differences coefficient estimate $$\beta$$ captures the marginal effect of the mergers on payout policy. To account for the potential correlation between firms affected by the same merger, I cluster standard errors by merger. However, the results are robust if I instead cluster standard errors by firm. One potential challenge to the identification strategy is that the mergers may be motivated by efficiency gains associated with reduced shareholder-credit conflict. If that is the case, the econometric exogeneity of these mergers no longer holds. However, economically, even if the mergers are motivated by the efficiency gains, $$\beta$$ still captures the effect of reduced shareholder-credit conflict on payout policy, one of the mechanisms through which the efficiency gains are realized. Furthermore, for the efficiency gains associated with reduced shareholder-creditor conflict to drive the mergers, it is still necessary that the merging parties’ stakes in the treated firms are sufficiently large. In the results reported in the Online Appendix, I only focus on mergers in which the merging parties’ stakes are small, and I still find similar results, mitigating the concern that the results are driven by mergers motivated by efficiency gains. 1.3 Variables and summary statistics I use three measures of payout, Payout, defined as total payout (DVC+PRSTKC) scaled by the market value of common equity (PRCC_F $$\times$$ CSHO), Repurchase, defined as share repurchases (PRSTKC) scaled by the market value of common equity, and Dividend, defined as cash dividend (DVC) scaled by the market value of common equity. The control variables include: log assets – the natural logarithm of total assets (AT), Tobin’s q – the market value of total assets (PRCC_F$$\times$$CSHO-CEQ+AT) divided by the book value of total assets (AT), Cash – cash and short term investment (CHE) scaled by total assets (AT), Age – the number of years the firm appeared in Compustat, Leverage – total debt (DLTT $$+$$ DLC) scaled by total assets (AT). Tangibility – total property, plant, and equipment (PPENT) scaled by total assets (AT), and Sales growth – the growth rate of sales (SALE). Table 2 reports the summary statistics of all variables used in the empirical analysis. All variables are winsorized at the 1% and 99% levels.5 The table shows that the average dividend yield of the sample is 1.17%, and the average total payout yield is 3.14%. The average total assets is about $\$$ 3 billion dollars, suggesting that the firms included in the sample are relatively large firms in the Compustat universe. The average Tobin’s q is around 1.73, which is similar to the average Tobin’s q of the Compustat universe. The average leverage ratio is about 27.0%, which is slightly higher than an average Compustat firm. The average firm age in the sample is about 21 years old. Table 2 Summary statistics Obs. Mean SD p25 Median p75 Dividend 5,969 1.170 1.675 0.000 0.000 1.523 Repurchase 5,969 2.251 4.311 0.000 0.046 2.660 Payout 5,969 3.137 5.020 0.000 1.390 4.565 Dividend_At 5,969 0.903 1.704 0.000 0.000 1.196 Payout_At 5,969 3.355 5.555 0.000 1.087 4.146 Repurchase_At 5,969 2.391 4.864 0.000 0.030 2.499 Capex 5,944 7.377 9.237 2.289 4.167 8.114 R&D 5,969 2.496 4.915 0.000 0.000 2.774 Acquisition 5,692 4.620 12.036 0.000 0.119 3.061 Debt financing 5,965 1.400 10.713 $$-$$2.831 0.000 5.009 Equity financing 5,789 2.013 5.681 0.069 0.410 1.314 Change in cash 5,969 0.825 6.221 $$-$$1.306 0.202 2.574 log assets 5,969 7.146 1.501 6.175 7.068 8.212 Tobin’s q 5,969 1.725 0.930 1.158 1.472 1.963 Cash 5,969 0.109 0.130 0.019 0.060 0.147 Leverage 5,969 0.270 0.188 0.133 0.255 0.380 Tangibility 5,969 0.321 0.248 0.123 0.245 0.470 Sales growth 5,969 0.138 0.381 $$-$$0.006 0.080 0.204 Age 5,969 20.971 11.509 11.000 19.000 31.000 Analyst coverage 5,969 8.786 6.785 4.000 7.000 13.000 Obs. Mean SD p25 Median p75 Dividend 5,969 1.170 1.675 0.000 0.000 1.523 Repurchase 5,969 2.251 4.311 0.000 0.046 2.660 Payout 5,969 3.137 5.020 0.000 1.390 4.565 Dividend_At 5,969 0.903 1.704 0.000 0.000 1.196 Payout_At 5,969 3.355 5.555 0.000 1.087 4.146 Repurchase_At 5,969 2.391 4.864 0.000 0.030 2.499 Capex 5,944 7.377 9.237 2.289 4.167 8.114 R&D 5,969 2.496 4.915 0.000 0.000 2.774 Acquisition 5,692 4.620 12.036 0.000 0.119 3.061 Debt financing 5,965 1.400 10.713 $$-$$2.831 0.000 5.009 Equity financing 5,789 2.013 5.681 0.069 0.410 1.314 Change in cash 5,969 0.825 6.221 $$-$$1.306 0.202 2.574 log assets 5,969 7.146 1.501 6.175 7.068 8.212 Tobin’s q 5,969 1.725 0.930 1.158 1.472 1.963 Cash 5,969 0.109 0.130 0.019 0.060 0.147 Leverage 5,969 0.270 0.188 0.133 0.255 0.380 Tangibility 5,969 0.321 0.248 0.123 0.245 0.470 Sales growth 5,969 0.138 0.381 $$-$$0.006 0.080 0.204 Age 5,969 20.971 11.509 11.000 19.000 31.000 Analyst coverage 5,969 8.786 6.785 4.000 7.000 13.000 This table reports the summary statistics of the variables used in this paper. The variables are Dividend, cash dividend (DVC) scaled by market value of common stocks (PRCC_F $$\times$$ CSHO); Payout, total payout (DVC+PRSTKC) scaled by the market value of common stocks; Repurchase, share repurchases (PRSTKC) scaled by the market value of common stocks; Capex, capital expenditure (CAPX) scaled by total assets (AT); R&D, R&D expense (XRD) scaled by total assets (AT); Acquisition, acquisition expense (AQC) scaled by total assets (AT); Debt financing, changes in total liability (DLCCH-DLTIS-DLTR) scaled by total assets; Change in cash, change in cash holding (CHECH) scaled by total assets, log assets, the natural logarithm of total assets (AT); Tobin’s q, market value of total assets (PRCC_F $$\times$$ CSHO$$+$$AT-CEQ) divided by total assets (AT); Cash, cash holding (CHE) scaled by total assets (AT), Leverage, total liability (DLC+DLTT) scaled by total assets (AT); Tangibility, total property, plant, and equipment (PPENT) scaled by total assets (AT); Sales growth, change in sales (SALE) divided by lagged sales; and Age, the number of years since the firm first appear in Compustat. Table 2 Summary statistics Obs. Mean SD p25 Median p75 Dividend 5,969 1.170 1.675 0.000 0.000 1.523 Repurchase 5,969 2.251 4.311 0.000 0.046 2.660 Payout 5,969 3.137 5.020 0.000 1.390 4.565 Dividend_At 5,969 0.903 1.704 0.000 0.000 1.196 Payout_At 5,969 3.355 5.555 0.000 1.087 4.146 Repurchase_At 5,969 2.391 4.864 0.000 0.030 2.499 Capex 5,944 7.377 9.237 2.289 4.167 8.114 R&D 5,969 2.496 4.915 0.000 0.000 2.774 Acquisition 5,692 4.620 12.036 0.000 0.119 3.061 Debt financing 5,965 1.400 10.713 $$-$$2.831 0.000 5.009 Equity financing 5,789 2.013 5.681 0.069 0.410 1.314 Change in cash 5,969 0.825 6.221 $$-$$1.306 0.202 2.574 log assets 5,969 7.146 1.501 6.175 7.068 8.212 Tobin’s q 5,969 1.725 0.930 1.158 1.472 1.963 Cash 5,969 0.109 0.130 0.019 0.060 0.147 Leverage 5,969 0.270 0.188 0.133 0.255 0.380 Tangibility 5,969 0.321 0.248 0.123 0.245 0.470 Sales growth 5,969 0.138 0.381 $$-$$0.006 0.080 0.204 Age 5,969 20.971 11.509 11.000 19.000 31.000 Analyst coverage 5,969 8.786 6.785 4.000 7.000 13.000 Obs. Mean SD p25 Median p75 Dividend 5,969 1.170 1.675 0.000 0.000 1.523 Repurchase 5,969 2.251 4.311 0.000 0.046 2.660 Payout 5,969 3.137 5.020 0.000 1.390 4.565 Dividend_At 5,969 0.903 1.704 0.000 0.000 1.196 Payout_At 5,969 3.355 5.555 0.000 1.087 4.146 Repurchase_At 5,969 2.391 4.864 0.000 0.030 2.499 Capex 5,944 7.377 9.237 2.289 4.167 8.114 R&D 5,969 2.496 4.915 0.000 0.000 2.774 Acquisition 5,692 4.620 12.036 0.000 0.119 3.061 Debt financing 5,965 1.400 10.713 $$-$$2.831 0.000 5.009 Equity financing 5,789 2.013 5.681 0.069 0.410 1.314 Change in cash 5,969 0.825 6.221 $$-$$1.306 0.202 2.574 log assets 5,969 7.146 1.501 6.175 7.068 8.212 Tobin’s q 5,969 1.725 0.930 1.158 1.472 1.963 Cash 5,969 0.109 0.130 0.019 0.060 0.147 Leverage 5,969 0.270 0.188 0.133 0.255 0.380 Tangibility 5,969 0.321 0.248 0.123 0.245 0.470 Sales growth 5,969 0.138 0.381 $$-$$0.006 0.080 0.204 Age 5,969 20.971 11.509 11.000 19.000 31.000 Analyst coverage 5,969 8.786 6.785 4.000 7.000 13.000 This table reports the summary statistics of the variables used in this paper. The variables are Dividend, cash dividend (DVC) scaled by market value of common stocks (PRCC_F $$\times$$ CSHO); Payout, total payout (DVC+PRSTKC) scaled by the market value of common stocks; Repurchase, share repurchases (PRSTKC) scaled by the market value of common stocks; Capex, capital expenditure (CAPX) scaled by total assets (AT); R&D, R&D expense (XRD) scaled by total assets (AT); Acquisition, acquisition expense (AQC) scaled by total assets (AT); Debt financing, changes in total liability (DLCCH-DLTIS-DLTR) scaled by total assets; Change in cash, change in cash holding (CHECH) scaled by total assets, log assets, the natural logarithm of total assets (AT); Tobin’s q, market value of total assets (PRCC_F $$\times$$ CSHO$$+$$AT-CEQ) divided by total assets (AT); Cash, cash holding (CHE) scaled by total assets (AT), Leverage, total liability (DLC+DLTT) scaled by total assets (AT); Tangibility, total property, plant, and equipment (PPENT) scaled by total assets (AT); Sales growth, change in sales (SALE) divided by lagged sales; and Age, the number of years since the firm first appear in Compustat. To ensure that the treated firms and control firms are comparable, I also compare the means of the variables of treated and control firms measured at the fiscal year ending immediately before the mergers. The results are presented in Table 3. The treated and control firms are similar in all dimensions, as the differences of the key variables between the treated and control firms are small and statistically insignificant. Table 3 Comparing treated and control firms before treatment Treat Control Difference Dividend 1.198 1.193 0.005 (0.119) (0.056) (0.012) Payout 3.385 3.377 0.008 (0.351) (0.172) (0.148) log assets 7.175 7.150 0.025 (0.088) (0.045) (0.098) Tobin’s q 1.694 1.645 0.050 (0.056) (0.027) (0.058) Cash 0.087 0.088 –0.001 (0.006) (0.108) (0.008) Leverage 0.271 0.281 –0.010 (0.011) (0.006) (0.013) Sales growth 0.110 0.137 –0.027 (0.016) (0.016) (0.031) Age 16.485 17.049 –0.564 (0.527) (0.302) (0.638) Inst own 81.762 82.876 –0.114 (1.903) (0.948) (2.061) Treat Control Difference Dividend 1.198 1.193 0.005 (0.119) (0.056) (0.012) Payout 3.385 3.377 0.008 (0.351) (0.172) (0.148) log assets 7.175 7.150 0.025 (0.088) (0.045) (0.098) Tobin’s q 1.694 1.645 0.050 (0.056) (0.027) (0.058) Cash 0.087 0.088 –0.001 (0.006) (0.108) (0.008) Leverage 0.271 0.281 –0.010 (0.011) (0.006) (0.013) Sales growth 0.110 0.137 –0.027 (0.016) (0.016) (0.031) Age 16.485 17.049 –0.564 (0.527) (0.302) (0.638) Inst own 81.762 82.876 –0.114 (1.903) (0.948) (2.061) This table reports the t-test results comparing treated and control firms measured at the fiscal year immediately before the merger announcement date, along the main variables. Variable definitions can be found in the note to Table 2. Standard errors are reported in parentheses. Significances of the difference between treated and control firms at 1%, 5%, and 10% levels are indicated by ***, **, and *, respectively. Table 3 Comparing treated and control firms before treatment Treat Control Difference Dividend 1.198 1.193 0.005 (0.119) (0.056) (0.012) Payout 3.385 3.377 0.008 (0.351) (0.172) (0.148) log assets 7.175 7.150 0.025 (0.088) (0.045) (0.098) Tobin’s q 1.694 1.645 0.050 (0.056) (0.027) (0.058) Cash 0.087 0.088 –0.001 (0.006) (0.108) (0.008) Leverage 0.271 0.281 –0.010 (0.011) (0.006) (0.013) Sales growth 0.110 0.137 –0.027 (0.016) (0.016) (0.031) Age 16.485 17.049 –0.564 (0.527) (0.302) (0.638) Inst own 81.762 82.876 –0.114 (1.903) (0.948) (2.061) Treat Control Difference Dividend 1.198 1.193 0.005 (0.119) (0.056) (0.012) Payout 3.385 3.377 0.008 (0.351) (0.172) (0.148) log assets 7.175 7.150 0.025 (0.088) (0.045) (0.098) Tobin’s q 1.694 1.645 0.050 (0.056) (0.027) (0.058) Cash 0.087 0.088 –0.001 (0.006) (0.108) (0.008) Leverage 0.271 0.281 –0.010 (0.011) (0.006) (0.013) Sales growth 0.110 0.137 –0.027 (0.016) (0.016) (0.031) Age 16.485 17.049 –0.564 (0.527) (0.302) (0.638) Inst own 81.762 82.876 –0.114 (1.903) (0.948) (2.061) This table reports the t-test results comparing treated and control firms measured at the fiscal year immediately before the merger announcement date, along the main variables. Variable definitions can be found in the note to Table 2. Standard errors are reported in parentheses. Significances of the difference between treated and control firms at 1%, 5%, and 10% levels are indicated by ***, **, and *, respectively. 2. Main Results 2.1 OLS results of the effect of dual holders on payout policy Before presenting the results on the effect of the mergers on payout policy, I first present the OLS regression results on the effect of dual holders on payout policy on all firms with DealScan loans. This serves as a check of the external validity of the difference-in-differences tests presented below. Because the number of mergers and the total number of firms affected by the mergers are small, the difference-in-differences tests may be subject to the small sample bias. Furthermore, the difference-in-differences tests rely on the specific source of variation in shareholder-creditor conflict generated by the mergers, which may not be generalizable to other sources of variation. To this end, I start the sample construction with all Compustat firms with loans outstanding in the LPC DealScan database from 1987–2014. I exclude firms in the financial and utilities industries. I then follow the procedure in Jiang, Li, and Shao (2010) to identify firms with dual holders. The summary statistics of this sample is provided in Table A1. Comparing Table A1 with Table 2 shows that the firms in this large sample are smaller, younger, and pay out less than those in Table 2. I then estimate the following specification: \begin{equation} Y_{\it it}=\alpha_{i}+\alpha_{t}+\beta {\it Dual}_{\it it}+\gamma X_{it-1} +\epsilon_{\it it}, \label{ols} \end{equation} (2) The results are presented in Table 4. Consistent with the argument that the existence of dual holders reduces shareholder-creditor conflict, the coefficient estimates on Dual are all negative and statistically significant. While these results are important in showing that the effect of dual holders persists on a large sample of firms, having dual holders can be potentially endogenous. To mitigate this problem, I then rely on the mergers between shareholders and lenders to generate plausibly exogenous variation in dual holders, and examine the effect of the mergers on payout policy in a difference-in-differences setting. Table 4 OLS regression results of the effect of dual holders on payout policy Payout Repurchase Dividend (1) (2) (3) (4) (5) (6) Dual –0.814$$^{***}$$ –0.784$$^{***}$$ –0.746$$^{***}$$ –0.724$$^{***}$$ –0.056$$^{*}$$ –0.052$$^{*}$$ (0.292) (0.287) (0.253) (0.252) (0.029) (0.028) log assets 0.692$$^{***}$$ 0.529$$^{***}$$ 0.159$$^{***}$$ (0.065) (0.052) (0.025) Tobin’s q –0.249$$^{***}$$ –0.138$$^{***}$$ –0.079$$^{***}$$ (0.040) (0.033) (0.012) Cash 3.206$$^{***}$$ 2.781$$^{***}$$ 0.226$$^{*}$$ (0.440) (0.363) (0.124) Leverage –3.258$$^{***}$$ –2.531$$^{***}$$ –0.476$$^{***}$$ (0.241) (0.183) (0.084) Tangibility 0.116 0.279 0.034 (0.469) (0.368) (0.162) Sales growth –0.142$$^{***}$$ –0.082$$^{**}$$ –0.036$$^{***}$$ (0.039) (0.033) (0.008) Age –0.056$$^{***}$$ –0.014 –0.034$$^{***}$$ (0.012) (0.010) (0.005) Constant 3.255$$^{***}$$ 1.126$$^{**}$$ 1.430$$^{***}$$ –0.542 1.595$$^{***}$$ 1.225$$^{***}$$ (0.208) (0.444) (0.153) (0.344) (0.078) (0.153) Year fixed effects Yes Yes Yes Yes Yes Yes Firm fixed effects Yes Yes Yes Yes Yes Yes Observations 42,919 42,919 42,919 42,919 42,919 42,919 Adjusted R-squared 0.222 0.239 0.153 0.170 0.567 0.572 Payout Repurchase Dividend (1) (2) (3) (4) (5) (6) Dual –0.814$$^{***}$$ –0.784$$^{***}$$ –0.746$$^{***}$$ –0.724$$^{***}$$ –0.056$$^{*}$$ –0.052$$^{*}$$ (0.292) (0.287) (0.253) (0.252) (0.029) (0.028) log assets 0.692$$^{***}$$ 0.529$$^{***}$$ 0.159$$^{***}$$ (0.065) (0.052) (0.025) Tobin’s q –0.249$$^{***}$$ –0.138$$^{***}$$ –0.079$$^{***}$$ (0.040) (0.033) (0.012) Cash 3.206$$^{***}$$ 2.781$$^{***}$$ 0.226$$^{*}$$ (0.440) (0.363) (0.124) Leverage –3.258$$^{***}$$ –2.531$$^{***}$$ –0.476$$^{***}$$ (0.241) (0.183) (0.084) Tangibility 0.116 0.279 0.034 (0.469) (0.368) (0.162) Sales growth –0.142$$^{***}$$ –0.082$$^{**}$$ –0.036$$^{***}$$ (0.039) (0.033) (0.008) Age –0.056$$^{***}$$ –0.014 –0.034$$^{***}$$ (0.012) (0.010) (0.005) Constant 3.255$$^{***}$$ 1.126$$^{**}$$ 1.430$$^{***}$$ –0.542 1.595$$^{***}$$ 1.225$$^{***}$$ (0.208) (0.444) (0.153) (0.344) (0.078) (0.153) Year fixed effects Yes Yes Yes Yes Yes Yes Firm fixed effects Yes Yes Yes Yes Yes Yes Observations 42,919 42,919 42,919 42,919 42,919 42,919 Adjusted R-squared 0.222 0.239 0.153 0.170 0.567 0.572 This table reports the OLS estimation results of $$Y_{\it it}=\alpha_{\it ij}+\alpha_{t}+\beta {\it Treat}_{\it ij} \times {\it Post}_{\it ijt} +\gamma X_{\it it-1} +\epsilon_{\it ijt}$$. The dependent variable in Columns 1 and 2 is Payout, the dependent variable in Columns 3 and 4 is Repurchase, and the dependent variable in Columns 5 and 6 is Dividend. Dual equals one if the firm is has a dual holder, and zero otherwise. All regressions include year fixed effects and firm fixed effects. Standard errors are clustered by firm. Significance at 1%, 5%, and 10% levels are indicated by ***, **, and *, respectively. Table 4 OLS regression results of the effect of dual holders on payout policy Payout Repurchase Dividend (1) (2) (3) (4) (5) (6) Dual –0.814$$^{***}$$ –0.784$$^{***}$$ –0.746$$^{***}$$ –0.724$$^{***}$$ –0.056$$^{*}$$ –0.052$$^{*}$$ (0.292) (0.287) (0.253) (0.252) (0.029) (0.028) log assets 0.692$$^{***}$$ 0.529$$^{***}$$ 0.159$$^{***}$$ (0.065) (0.052) (0.025) Tobin’s q –0.249$$^{***}$$ –0.138$$^{***}$$ –0.079$$^{***}$$ (0.040) (0.033) (0.012) Cash 3.206$$^{***}$$ 2.781$$^{***}$$ 0.226$$^{*}$$ (0.440) (0.363) (0.124) Leverage –3.258$$^{***}$$ –2.531$$^{***}$$ –0.476$$^{***}$$ (0.241) (0.183) (0.084) Tangibility 0.116 0.279 0.034 (0.469) (0.368) (0.162) Sales growth –0.142$$^{***}$$ –0.082$$^{**}$$ –0.036$$^{***}$$ (0.039) (0.033) (0.008) Age –0.056$$^{***}$$ –0.014 –0.034$$^{***}$$ (0.012) (0.010) (0.005) Constant 3.255$$^{***}$$ 1.126$$^{**}$$ 1.430$$^{***}$$ –0.542 1.595$$^{***}$$ 1.225$$^{***}$$ (0.208) (0.444) (0.153) (0.344) (0.078) (0.153) Year fixed effects Yes Yes Yes Yes Yes Yes Firm fixed effects Yes Yes Yes Yes Yes Yes Observations 42,919 42,919 42,919 42,919 42,919 42,919 Adjusted R-squared 0.222 0.239 0.153 0.170 0.567 0.572 Payout Repurchase Dividend (1) (2) (3) (4) (5) (6) Dual –0.814$$^{***}$$ –0.784$$^{***}$$ –0.746$$^{***}$$ –0.724$$^{***}$$ –0.056$$^{*}$$ –0.052$$^{*}$$ (0.292) (0.287) (0.253) (0.252) (0.029) (0.028) log assets 0.692$$^{***}$$ 0.529$$^{***}$$ 0.159$$^{***}$$ (0.065) (0.052) (0.025) Tobin’s q –0.249$$^{***}$$ –0.138$$^{***}$$ –0.079$$^{***}$$ (0.040) (0.033) (0.012) Cash 3.206$$^{***}$$ 2.781$$^{***}$$ 0.226$$^{*}$$ (0.440) (0.363) (0.124) Leverage –3.258$$^{***}$$ –2.531$$^{***}$$ –0.476$$^{***}$$ (0.241) (0.183) (0.084) Tangibility 0.116 0.279 0.034 (0.469) (0.368) (0.162) Sales growth –0.142$$^{***}$$ –0.082$$^{**}$$ –0.036$$^{***}$$ (0.039) (0.033) (0.008) Age –0.056$$^{***}$$ –0.014 –0.034$$^{***}$$ (0.012) (0.010) (0.005) Constant 3.255$$^{***}$$ 1.126$$^{**}$$ 1.430$$^{***}$$ –0.542 1.595$$^{***}$$ 1.225$$^{***}$$ (0.208) (0.444) (0.153) (0.344) (0.078) (0.153) Year fixed effects Yes Yes Yes Yes Yes Yes Firm fixed effects Yes Yes Yes Yes Yes Yes Observations 42,919 42,919 42,919 42,919 42,919 42,919 Adjusted R-squared 0.222 0.239 0.153 0.170 0.567 0.572 This table reports the OLS estimation results of $$Y_{\it it}=\alpha_{\it ij}+\alpha_{t}+\beta {\it Treat}_{\it ij} \times {\it Post}_{\it ijt} +\gamma X_{\it it-1} +\epsilon_{\it ijt}$$. The dependent variable in Columns 1 and 2 is Payout, the dependent variable in Columns 3 and 4 is Repurchase, and the dependent variable in Columns 5 and 6 is Dividend. Dual equals one if the firm is has a dual holder, and zero otherwise. All regressions include year fixed effects and firm fixed effects. Standard errors are clustered by firm. Significance at 1%, 5%, and 10% levels are indicated by ***, **, and *, respectively. 2.2 Mergers and loan prices The identification relies on the assumption that the mergers between shareholders and lenders of the same firms reduce the shareholder-creditor conflict. Under this assumption, loan prices should increase following the mergers. To test this conjecture, I obtain the secondary loan market data from Thomson Reuters LPC. In collaboration with the Loan Syndication and Trading Association (LSTA), LPC collects daily loan bid and ask prices from market makers. The secondary loan database uses the LIN to identify each individual loan. To obtain the corresponding primary market data of individual loans, I match it to the LPC DealScan data based on LIN and FacilityID. I then run the difference-in-differences specification like in Equation (1), but with the logarithm of loan prices as the dependent variable. To implement, I calculate annual average loan prices and then use the logarithm of the calculated average prices as the dependent variables in the specification. The results are presented in Table 5. Consistent with the argument that the mergers reduce the shareholder-creditor conflict and hence should increase loan prices, the difference-in-differences coefficient estimates are all positive and statistically significant. The economic magnitudes are also significant, with the mergers increasing loan prices by more than 3%. The results therefore support the argument that the mergers between shareholders and lenders of the same firms do in fact result in reduced shareholder-creditor conflict, which pushes up loan prices. Table 5 The effect of mergers on loan prices (1) (2) Treat $$\times$$ Post 0.032$$^{**}$$ 0.033$$^{**}$$ (0.013) (0.014) log assets 0.013 (0.012) Tobin’s q –0.012 (0.010) Cash 0.030 (0.041) Leverage –0.065$$^{**}$$ (0.029) Tangibility –0.101$$^{**}$$ (0.041) Sales growth 0.019$$^{***}$$ (0.006) Age –0.011$$^{**}$$ (0.005) Constant 4.728$$^{***}$$ 4.836$$^{***}$$ (0.071) (0.115) Year fixed effects Yes Yes Facility fixed effects Yes Yes Observations 3,444 3,444 Adjusted R-squared 0.553 0.557 (1) (2) Treat $$\times$$ Post 0.032$$^{**}$$ 0.033$$^{**}$$ (0.013) (0.014) log assets 0.013 (0.012) Tobin’s q –0.012 (0.010) Cash 0.030 (0.041) Leverage –0.065$$^{**}$$ (0.029) Tangibility –0.101$$^{**}$$ (0.041) Sales growth 0.019$$^{***}$$ (0.006) Age –0.011$$^{**}$$ (0.005) Constant 4.728$$^{***}$$ 4.836$$^{***}$$ (0.071) (0.115) Year fixed effects Yes Yes Facility fixed effects Yes Yes Observations 3,444 3,444 Adjusted R-squared 0.553 0.557 This table reports the difference-in-differences estimation results of $$P_{\it it}=\alpha_{i}+\alpha_{t}+\beta {\it Treat}_{i} \times {\it Post}_{\it it} +\epsilon_{\it it}$$ for loan prices. The dependent variables are the logarithms of annual average prices of loan quotes. Post equals one if it is after the merger. All regressions include time fixed effects and facility fixed effects. Standard errors are clustered by merger. Significance at 1%, 5%, and 10% levels are indicated by ***, **, and *, respectively. Table 5 The effect of mergers on loan prices (1) (2) Treat $$\times$$ Post 0.032$$^{**}$$ 0.033$$^{**}$$ (0.013) (0.014) log assets 0.013 (0.012) Tobin’s q –0.012 (0.010) Cash 0.030 (0.041) Leverage –0.065$$^{**}$$ (0.029) Tangibility –0.101$$^{**}$$ (0.041) Sales growth 0.019$$^{***}$$ (0.006) Age –0.011$$^{**}$$ (0.005) Constant 4.728$$^{***}$$ 4.836$$^{***}$$ (0.071) (0.115) Year fixed effects Yes Yes Facility fixed effects Yes Yes Observations 3,444 3,444 Adjusted R-squared 0.553 0.557 (1) (2) Treat $$\times$$ Post 0.032$$^{**}$$ 0.033$$^{**}$$ (0.013) (0.014) log assets 0.013 (0.012) Tobin’s q –0.012 (0.010) Cash 0.030 (0.041) Leverage –0.065$$^{**}$$ (0.029) Tangibility –0.101$$^{**}$$ (0.041) Sales growth 0.019$$^{***}$$ (0.006) Age –0.011$$^{**}$$ (0.005) Constant 4.728$$^{***}$$ 4.836$$^{***}$$ (0.071) (0.115) Year fixed effects Yes Yes Facility fixed effects Yes Yes Observations 3,444 3,444 Adjusted R-squared 0.553 0.557 This table reports the difference-in-differences estimation results of $$P_{\it it}=\alpha_{i}+\alpha_{t}+\beta {\it Treat}_{i} \times {\it Post}_{\it it} +\epsilon_{\it it}$$ for loan prices. The dependent variables are the logarithms of annual average prices of loan quotes. Post equals one if it is after the merger. All regressions include time fixed effects and facility fixed effects. Standard errors are clustered by merger. Significance at 1%, 5%, and 10% levels are indicated by ***, **, and *, respectively. 2.3 Baseline difference-in-differences results I present the baseline results of estimating Equation (1) in Table 6. In Columns 1 and 2, I first present the results for Payout with and without the controls. The mergers, which reduce the conflict between shareholders and creditors, can also potentially affect the control variables, that is, the control variables can be endogenous. Estimating Equation (1) both with and without the controls ensures that the results are not driven by these potentially endogenous control variables. In both columns, the difference-in-differences estimates, that is, the coefficients on Treat $$\times$$ Post are negative and statistically significant. The effect is also economically significant. Taking the coefficient in Column 2, the effect of the merger between a lender and a shareholder leads to a decrease of payout of treated firms, relative to control firms, by more than 25% of the average dividend yield in the sample. Table 6 Baseline results of the effect of mergers on payout policy Payout Repurchase Dividend (1) (2) (3) (4) (5) (6) Treat $$\times$$ Post –0.813$$^{***}$$ –0.832$$^{***}$$ –0.728$$^{***}$$ –0.753$$^{***}$$ –0.061$$^{**}$$ –0.069$$^{***}$$ (0.280) (0.278) (0.261) (0.261) (0.024) (0.022) log assets 1.114$$^{***}$$ 0.912$$^{***}$$ 0.240$$^{***}$$ (0.295) (0.267) (0.085) Tobin’s q –0.074 –0.001 –0.058$$^{*}$$ (0.088) (0.077) (0.033) Cash 3.207$$^{**}$$ 2.552$$^{**}$$ 0.301 (1.393) (1.191) (0.204) Leverage –4.424$$^{***}$$ –4.404$$^{***}$$ –0.108 (0.747) (0.758) (0.176) Tangibility –4.582$$^{***}$$ –4.343$$^{***}$$ –0.200 (1.133) (0.970) (0.279) Sales growth –0.580$$^{***}$$ –0.604$$^{***}$$ 0.018 (0.131) (0.124) (0.035) Age –0.142$$^{***}$$ –0.111$$^{***}$$ –0.041$$^{**}$$ (0.040) (0.034) (0.018) Constant 1.211$$^{*}$$ –1.166 –0.144 –1.708 1.466$$^{***}$$ 0.537 (0.669) (1.642) (0.526) (1.369) (0.364) (0.573) Year fixed effects Yes Yes Yes Yes Yes Yes Firm fixed effects Yes Yes Yes Yes Yes Yes Observations 5,969 5,969 5,969 5,969 5,969 5,969 Adjusted R-squared 0.295 0.314 0.270 0.292 0.611 0.614 Payout Repurchase Dividend (1) (2) (3) (4) (5) (6) Treat $$\times$$ Post –0.813$$^{***}$$ –0.832$$^{***}$$ –0.728$$^{***}$$ –0.753$$^{***}$$ –0.061$$^{**}$$ –0.069$$^{***}$$ (0.280) (0.278) (0.261) (0.261) (0.024) (0.022) log assets 1.114$$^{***}$$ 0.912$$^{***}$$ 0.240$$^{***}$$ (0.295) (0.267) (0.085) Tobin’s q –0.074 –0.001 –0.058$$^{*}$$ (0.088) (0.077) (0.033) Cash 3.207$$^{**}$$ 2.552$$^{**}$$ 0.301 (1.393) (1.191) (0.204) Leverage –4.424$$^{***}$$ –4.404$$^{***}$$ –0.108 (0.747) (0.758) (0.176) Tangibility –4.582$$^{***}$$ –4.343$$^{***}$$ –0.200 (1.133) (0.970) (0.279) Sales growth –0.580$$^{***}$$ –0.604$$^{***}$$ 0.018 (0.131) (0.124) (0.035) Age –0.142$$^{***}$$ –0.111$$^{***}$$ –0.041$$^{**}$$ (0.040) (0.034) (0.018) Constant 1.211$$^{*}$$ –1.166 –0.144 –1.708 1.466$$^{***}$$ 0.537 (0.669) (1.642) (0.526) (1.369) (0.364) (0.573) Year fixed effects Yes Yes Yes Yes Yes Yes Firm fixed effects Yes Yes Yes Yes Yes Yes Observations 5,969 5,969 5,969 5,969 5,969 5,969 Adjusted R-squared 0.295 0.314 0.270 0.292 0.611 0.614 This table reports the baseline difference-in-differences estimation results of $$Y_{\it it}=\alpha_{ij}+\alpha_{t}+\beta {\it Treat}_{\it ij} \times {\it Post}_{\it ijt} +\gamma X_{\it it-1} +\epsilon_{\it ijt}$$. The dependent variable in Columns 1 and 2 is Payout, the dependent variable in Columns 3 and 4 is Repurchase, and the dependent variable in Columns 5 and 6 is Dividend. Treat equals one if the firm is a treated firm of the merger, and zero otherwise. Post equals one if the firm-year observation is after the merger. All regressions include year fixed effects and firm fixed effects. Standard errors are clustered by merger. Significance at 1%, 5%, and 10% levels are indicated by ***, **, and *, respectively. Table 6 Baseline results of the effect of mergers on payout policy Payout Repurchase Dividend (1) (2) (3) (4) (5) (6) Treat $$\times$$ Post –0.813$$^{***}$$ –0.832$$^{***}$$ –0.728$$^{***}$$ –0.753$$^{***}$$ –0.061$$^{**}$$ –0.069$$^{***}$$ (0.280) (0.278) (0.261) (0.261) (0.024) (0.022) log assets 1.114$$^{***}$$ 0.912$$^{***}$$ 0.240$$^{***}$$ (0.295) (0.267) (0.085) Tobin’s q –0.074 –0.001 –0.058$$^{*}$$ (0.088) (0.077) (0.033) Cash 3.207$$^{**}$$ 2.552$$^{**}$$ 0.301 (1.393) (1.191) (0.204) Leverage –4.424$$^{***}$$ –4.404$$^{***}$$ –0.108 (0.747) (0.758) (0.176) Tangibility –4.582$$^{***}$$ –4.343$$^{***}$$ –0.200 (1.133) (0.970) (0.279) Sales growth –0.580$$^{***}$$ –0.604$$^{***}$$ 0.018 (0.131) (0.124) (0.035) Age –0.142$$^{***}$$ –0.111$$^{***}$$ –0.041$$^{**}$$ (0.040) (0.034) (0.018) Constant 1.211$$^{*}$$ –1.166 –0.144 –1.708 1.466$$^{***}$$ 0.537 (0.669) (1.642) (0.526) (1.369) (0.364) (0.573) Year fixed effects Yes Yes Yes Yes Yes Yes Firm fixed effects Yes Yes Yes Yes Yes Yes Observations 5,969 5,969 5,969 5,969 5,969 5,969 Adjusted R-squared 0.295 0.314 0.270 0.292 0.611 0.614 Payout Repurchase Dividend (1) (2) (3) (4) (5) (6) Treat $$\times$$ Post –0.813$$^{***}$$ –0.832$$^{***}$$ –0.728$$^{***}$$ –0.753$$^{***}$$ –0.061$$^{**}$$ –0.069$$^{***}$$ (0.280) (0.278) (0.261) (0.261) (0.024) (0.022) log assets 1.114$$^{***}$$ 0.912$$^{***}$$ 0.240$$^{***}$$ (0.295) (0.267) (0.085) Tobin’s q –0.074 –0.001 –0.058$$^{*}$$ (0.088) (0.077) (0.033) Cash 3.207$$^{**}$$ 2.552$$^{**}$$ 0.301 (1.393) (1.191) (0.204) Leverage –4.424$$^{***}$$ –4.404$$^{***}$$ –0.108 (0.747) (0.758) (0.176) Tangibility –4.582$$^{***}$$ –4.343$$^{***}$$ –0.200 (1.133) (0.970) (0.279) Sales growth –0.580$$^{***}$$ –0.604$$^{***}$$ 0.018 (0.131) (0.124) (0.035) Age –0.142$$^{***}$$ –0.111$$^{***}$$ –0.041$$^{**}$$ (0.040) (0.034) (0.018) Constant 1.211$$^{*}$$ –1.166 –0.144 –1.708 1.466$$^{***}$$ 0.537 (0.669) (1.642) (0.526) (1.369) (0.364) (0.573) Year fixed effects Yes Yes Yes Yes Yes Yes Firm fixed effects Yes Yes Yes Yes Yes Yes Observations 5,969 5,969 5,969 5,969 5,969 5,969 Adjusted R-squared 0.295 0.314 0.270 0.292 0.611 0.614 This table reports the baseline difference-in-differences estimation results of $$Y_{\it it}=\alpha_{ij}+\alpha_{t}+\beta {\it Treat}_{\it ij} \times {\it Post}_{\it ijt} +\gamma X_{\it it-1} +\epsilon_{\it ijt}$$. The dependent variable in Columns 1 and 2 is Payout, the dependent variable in Columns 3 and 4 is Repurchase, and the dependent variable in Columns 5 and 6 is Dividend. Treat equals one if the firm is a treated firm of the merger, and zero otherwise. Post equals one if the firm-year observation is after the merger. All regressions include year fixed effects and firm fixed effects. Standard errors are clustered by merger. Significance at 1%, 5%, and 10% levels are indicated by ***, **, and *, respectively. I then decompose total payout into share repurchases and cash dividends, and present the results in Columns 3–6.6 While the coefficient estimates are still negative and statistically significant in all columns, the coefficients are much larger for share repurchases (Columns 3 and 4) than those for cash dividends (Columns 5 and 6). These results suggest that the mergers, while reducing the shareholder-creditor conflict, have a more pronounced effect on share repurchases. This is probably due to the fact that firms are often able to adjust share repurchases more quickly than cash dividends. 2.4 Addressing identification challenges The consistency of the difference-in-differences estimates depends on the parallel trend condition; that is, the outcome variables should have parallel trends in the absence of the treatment. Although the parallel trend condition is untestable, I follow the advice of Roberts and Whited (2012) to conduct a visual examination of the payout policy around the mergers. Specifically, I examine the evolution of Payout, Repurchase, and Dividend around the mergers for treated and control firms separately and the results are presented in Figure 1, with panel A for Payout, panel B for Repurchase, and panel C for Dividend. The figure shows that the payout variables follow similar trends before the events. After the event, however, although the control firms continue their pre-event trend, the treated firms experience an abrupt change of the trend. The results suggest (although do not prove) that the parallel trend condition is likely to be satisfied. Figure 1 View largeDownload slide Payout around mergers This figure shows the evolution of payout policy of treated and control firms around the merger of lenders and institutional investors. Panel A reports the yearly average of Payout, total payout divided by market value of common equity; panel B reports the yearly average of Repurchase, share repurchases divided by total assets; and panel C reports the yearly average of Dividend, cash dividend divided by market value of common equity. Figure 1 View largeDownload slide Payout around mergers This figure shows the evolution of payout policy of treated and control firms around the merger of lenders and institutional investors. Panel A reports the yearly average of Payout, total payout divided by market value of common equity; panel B reports the yearly average of Repurchase, share repurchases divided by total assets; and panel C reports the yearly average of Dividend, cash dividend divided by market value of common equity. To provide further evidence that the baseline results are not driven by preexisting trend differences between treated and control firms, I conduct a dynamic analysis of the effect of the mergers. To implement, I extend the sample period to 6 years before the mergers and then interact each event year dummy with the treatment dummy, that is, I estimate the following \begin{equation} Y_{\it it}=\alpha_{i}+\Sigma_{k=-6}^{k=3}\beta_k {\it Treat}_{i} \times {\it Year}^{k} +\gamma X_{\it it-1} +\epsilon_{\it it}, \label{placebo} \end{equation} (3) where all variables are defined exactly the same as those in Equation (1), except for $${\it Year}^{k}$$’s, which equals one if the fiscal year is $$k$$ years before the merger, and zero otherwise. If the baseline results are truly driven by the mergers, I expect the $$\beta_k$$’s to be close to zero for all $$k<0$$, and the $$\beta_k$$’s to be negative for some $$k>0$$. On the other hand, if the baseline results are driven by preexisting differences between treated and control firms, the effect may show up before the mergers. The coefficient estimates are provided in Figure 2. Consistent with the argument that the baseline results are truly driven by the mergers, the estimates of $$\beta_k$$’s are all small and statistically insignificant for $$k<0$$, but turn negative and statistically significant for some $$k>0$$. The results therefore provide further confidence that the baseline results are unlikely to be driven by nonparallel trends between treated and control firms. Figure 2 View largeDownload slide Dynamics of the coefficient estimates This figure shows the coefficient estimates from estimating $$Y_{\it it}=\alpha_{i}+\Sigma_{k=-6}^{k=3}\beta_k {\it Treat}_{i} \times {\it Year}^{k} +\gamma X_{it-1} +\epsilon_{\it it}$$ for total payout, repurchases, and dividends. Figure 2 View largeDownload slide Dynamics of the coefficient estimates This figure shows the coefficient estimates from estimating $$Y_{\it it}=\alpha_{i}+\Sigma_{k=-6}^{k=3}\beta_k {\it Treat}_{i} \times {\it Year}^{k} +\gamma X_{it-1} +\epsilon_{\it it}$$ for total payout, repurchases, and dividends. In the Online Appendix, I provide a test using a subsample of mergers in which the merging parties stakes in the treated firms, relative to their total loan and stock portfolio, are below the sample median, and find results similar to those in Table 6. When merging parties’ stakes in the treated firms are small, these mergers are less likely to be motivated by efficiency gains associated with these firms, and hence the results are less likely to be subject to the endogeneity concerns. 2.5 Merging lenders’ stakes and the effect of the mergers I then examine how the stakes of the lenders in the treated firm alter the effects of the mergers on payout policy. I argue that the baseline results are driven by shareholders shifting from maximizing shareholder value toward maximizing combined value of stocks and loans they hold after the merger, and hence the effect should be stronger if the merging lender’s stake in the treated firm is larger. On the other hand, if the results are driven by other unobservable factors correlated with the mergers, it is unlikely to be correlated with the stakes of the merging lenders. To this end, I sort the observations into terciles according to the merging lender’s loan size (the total amount of the loan allocated to the lender) scaled by the firm’s total assets and redo the analysis on the top and bottom terciles separately. The results are presented in panel A of Table 7. The difference-in-differences estimates are negative and statistically significant for observations in the top tercile, and are much smaller and statistically insignificant for observations in the bottom tercile. The differences of the estimates are also statistically significant. Overall, the results are consistent with the argument that the mergers reduce the shareholder-creditor conflict and that the effect is stronger if the stakes of the merging lenders are larger. Table 7 Cross-sectional heterogeneity of the effect of mergers on payout policy A. Lender stake Payout Repurchase Dividend Low High Low High Low High (1) (2) (3) (4) (5) (6) Treat $$\times$$ Post –0.290 –1.095$$^{**}$$ –0.312 –0.914$$^{**}$$ 0.045 –0.142$$^{**}$$ (0.528) (0.487) (0.600) (0.446) (0.110) (0.069) Controls Yes Yes Yes Yes Yes Yes Year fixed effects Yes Yes Yes Yes Yes Yes Firm fixed effects Yes Yes Yes Yes Yes Yes Observations 2,056 1,962 2,056 1,962 2,056 1,962 Adjusted R-squared 0.265 0.351 0.242 0.331 0.553 0.634 B. Shareholder power Payout Repurchase Dividend High Low High Low High Low (1) (2) (3) (4) (5) (6) Treat $$\times$$ Post –1.047$$^{***}$$ –0.457 –0.962$$^{**}$$ –0.382 –0.079$$^{**}$$ 0.023 (0.359) (0.565) (0.422) (0.499) (0.038) (0.065) Controls Yes Yes Yes Yes Yes Yes Year fixed effects Yes Yes Yes Yes Yes Yes Firm fixed effects Yes Yes Yes Yes Yes Yes Observations 1,983 1,995 1,983 1,995 1,983 1,995 Adjusted R-squared 0.329 0.338 0.310 0.297 0.600 0.601 C. Dividend restriction covenants Payout Repurchase Dividend Dividend covenant Dividend covenant Dividend covenant Yes No Yes No Yes No (1) (2) (3) (4) (5) (6) Treat $$\times$$ Post –0.311 –0.933$$^{***}$$ –0.259 –0.799$$^{***}$$ –0.025 –0.077$$^{**}$$ (0.526) (0.335) (0.503) (0.323) (0.059) (0.032) Controls Yes Yes Yes Yes Yes Yes Year fixed effects Yes Yes Yes Yes Yes Yes Firm fixed effects Yes Yes Yes Yes Yes Yes Observations 1,168 4,801 1,168 4,801 1,168 4,801 Adjusted R-squared 0.292 0.317 0.242 0.299 0.599 0.616 A. Lender stake Payout Repurchase Dividend Low High Low High Low High (1) (2) (3) (4) (5) (6) Treat $$\times$$ Post –0.290 –1.095$$^{**}$$ –0.312 –0.914$$^{**}$$ 0.045 –0.142$$^{**}$$ (0.528) (0.487) (0.600) (0.446) (0.110) (0.069) Controls Yes Yes Yes Yes Yes Yes Year fixed effects Yes Yes Yes Yes Yes Yes Firm fixed effects Yes Yes Yes Yes Yes Yes Observations 2,056 1,962 2,056 1,962 2,056 1,962 Adjusted R-squared 0.265 0.351 0.242 0.331 0.553 0.634 B. Shareholder power Payout Repurchase Dividend High Low High Low High Low (1) (2) (3) (4) (5) (6) Treat $$\times$$ Post –1.047$$^{***}$$ –0.457 –0.962$$^{**}$$ –0.382 –0.079$$^{**}$$ 0.023 (0.359) (0.565) (0.422) (0.499) (0.038) (0.065) Controls Yes Yes Yes Yes Yes Yes Year fixed effects Yes Yes Yes Yes Yes Yes Firm fixed effects Yes Yes Yes Yes Yes Yes Observations 1,983 1,995 1,983 1,995 1,983 1,995 Adjusted R-squared 0.329 0.338 0.310 0.297 0.600 0.601 C. Dividend restriction covenants Payout Repurchase Dividend Dividend covenant Dividend covenant Dividend covenant Yes No Yes No Yes No (1) (2) (3) (4) (5) (6) Treat $$\times$$ Post –0.311 –0.933$$^{***}$$ –0.259 –0.799$$^{***}$$ –0.025 –0.077$$^{**}$$ (0.526) (0.335) (0.503) (0.323) (0.059) (0.032) Controls Yes Yes Yes Yes Yes Yes Year fixed effects Yes Yes Yes Yes Yes Yes Firm fixed effects Yes Yes Yes Yes Yes Yes Observations 1,168 4,801 1,168 4,801 1,168 4,801 Adjusted R-squared 0.292 0.317 0.242 0.299 0.599 0.616 D. Financial distress Payout Repurchase Dividend High Low High Low High Low (3) (4) (1) (2) (1) (2) Treat $$\times$$ Post –1.153$$^{**}$$ –0.665 –1.082$$^{**}$$ –0.686 –0.076$$^{**}$$ 0.020 (0.509) (0.538) (0.434) (0.420) (0.032) (0.096) Controls Yes Yes Yes Yes Yes Yes Year fixed effects Yes Yes Yes Yes Yes Yes Firm fixed effects Yes Yes Yes Yes Yes Yes Observations 1,993 1,984 1,993 1,984 1,993 1,984 Adjusted R-squared 0.356 0.282 0.320 0.273 0.659 0.603 E. Analyst coverage Payout Repurchase Dividend Analyst coverage Analyst coverage Analyst coverage High Low High Low High Low Treat $$\times$$ Post –0.332 –0.333 –0.322 –0.282 –0.052 –0.053 (0.622) (0.647) (0.551) (0.562) (0.056) (0.089) Controls Yes Yes Yes Yes Yes Yes Year fixed effects Yes Yes Yes Yes Yes Yes Firm fixed effects Yes Yes Yes Yes Yes Yes Observations 2,025 2,301 2,025 2,301 2,025 2,301 Adjusted R-squared 0.337 0.316 0.294 0.316 0.661 0.571 D. Financial distress Payout Repurchase Dividend High Low High Low High Low (3) (4) (1) (2) (1) (2) Treat $$\times$$ Post –1.153$$^{**}$$ –0.665 –1.082$$^{**}$$ –0.686 –0.076$$^{**}$$ 0.020 (0.509) (0.538) (0.434) (0.420) (0.032) (0.096) Controls Yes Yes Yes Yes Yes Yes Year fixed effects Yes Yes Yes Yes Yes Yes Firm fixed effects Yes Yes Yes Yes Yes Yes Observations 1,993 1,984 1,993 1,984 1,993 1,984 Adjusted R-squared 0.356 0.282 0.320 0.273 0.659 0.603 E. Analyst coverage Payout Repurchase Dividend Analyst coverage Analyst coverage Analyst coverage High Low High Low High Low Treat $$\times$$ Post –0.332 –0.333 –0.322 –0.282 –0.052 –0.053 (0.622) (0.647) (0.551) (0.562) (0.056) (0.089) Controls Yes Yes Yes Yes Yes Yes Year fixed effects Yes Yes Yes Yes Yes Yes Firm fixed effects Yes Yes Yes Yes Yes Yes Observations 2,025 2,301 2,025 2,301 2,025 2,301 Adjusted R-squared 0.337 0.316 0.294 0.316 0.661 0.571 This table reports the difference-in-differences estimation results of $$Y_{\it it}=\alpha_{\it ij}+\alpha_{t}+\beta {\it Treat}_{\it ij} \times {\it Post}_{\it ijt} +\gamma X_{\it it-1} +\epsilon_{\it ijt}$$ on subsamples partitioned on measures of the stakes of the merging lenders (panel A), power of the merging shareholder (panel B), dividend restriction covenants (panel C), firm leverage (panel D), and analyst coverage (panel E). The dependent variables in Columns 1 and 2 are Payout, and in Columns 3 and 4 are Repurchase, and in Columns 5 and 6 are Dividend. Treat equals one if the firm is a treated firm of the merger, and zero otherwise. Post equals one if the firm-year observation is after the merger. All regressions include year fixed effects and merger-firm fixed effects. Standard errors are clustered by merger. Significance at 1%, 5%, and 10% levels are indicated by ***, **, and *, respectively. Table 7 Cross-sectional heterogeneity of the effect of mergers on payout policy A. Lender stake Payout Repurchase Dividend Low High Low High Low High (1) (2) (3) (4) (5) (6) Treat $$\times$$ Post –0.290 –1.095$$^{**}$$ –0.312 –0.914$$^{**}$$ 0.045 –0.142$$^{**}$$ (0.528) (0.487) (0.600) (0.446) (0.110) (0.069) Controls Yes Yes Yes Yes Yes Yes Year fixed effects Yes Yes Yes Yes Yes Yes Firm fixed effects Yes Yes Yes Yes Yes Yes Observations 2,056 1,962 2,056 1,962 2,056 1,962 Adjusted R-squared 0.265 0.351 0.242 0.331 0.553 0.634 B. Shareholder power Payout Repurchase Dividend High Low High Low High Low (1) (2) (3) (4) (5) (6) Treat $$\times$$ Post –1.047$$^{***}$$ –0.457 –0.962$$^{**}$$ –0.382 –0.079$$^{**}$$ 0.023 (0.359) (0.565) (0.422) (0.499) (0.038) (0.065) Controls Yes Yes Yes Yes Yes Yes Year fixed effects Yes Yes Yes Yes Yes Yes Firm fixed effects Yes Yes Yes Yes Yes Yes Observations 1,983 1,995 1,983 1,995 1,983 1,995 Adjusted R-squared 0.329 0.338 0.310 0.297 0.600 0.601 C. Dividend restriction covenants Payout Repurchase Dividend Dividend covenant Dividend covenant Dividend covenant Yes No Yes No Yes No (1) (2) (3) (4) (5) (6) Treat $$\times$$ Post –0.311 –0.933$$^{***}$$ –0.259 –0.799$$^{***}$$ –0.025 –0.077$$^{**}$$ (0.526) (0.335) (0.503) (0.323) (0.059) (0.032) Controls Yes Yes Yes Yes Yes Yes Year fixed effects Yes Yes Yes Yes Yes Yes Firm fixed effects Yes Yes Yes Yes Yes Yes Observations 1,168 4,801 1,168 4,801 1,168 4,801 Adjusted R-squared 0.292 0.317 0.242 0.299 0.599 0.616 A. Lender stake Payout Repurchase Dividend Low High Low High Low High (1) (2) (3) (4) (5) (6) Treat $$\times$$ Post –0.290 –1.095$$^{**}$$ –0.312 –0.914$$^{**}$$ 0.045 –0.142$$^{**}$$ (0.528) (0.487) (0.600) (0.446) (0.110) (0.069) Controls Yes Yes Yes Yes Yes Yes Year fixed effects Yes Yes Yes Yes Yes Yes Firm fixed effects Yes Yes Yes Yes Yes Yes Observations 2,056 1,962 2,056 1,962 2,056 1,962 Adjusted R-squared 0.265 0.351 0.242 0.331 0.553 0.634 B. Shareholder power Payout Repurchase Dividend High Low High Low High Low (1) (2) (3) (4) (5) (6) Treat $$\times$$ Post –1.047$$^{***}$$ –0.457 –0.962$$^{**}$$ –0.382 –0.079$$^{**}$$ 0.023 (0.359) (0.565) (0.422) (0.499) (0.038) (0.065) Controls Yes Yes Yes Yes Yes Yes Year fixed effects Yes Yes Yes Yes Yes Yes Firm fixed effects Yes Yes Yes Yes Yes Yes Observations 1,983 1,995 1,983 1,995 1,983 1,995 Adjusted R-squared 0.329 0.338 0.310 0.297 0.600 0.601 C. Dividend restriction covenants Payout Repurchase Dividend Dividend covenant Dividend covenant Dividend covenant Yes No Yes No Yes No (1) (2) (3) (4) (5) (6) Treat $$\times$$ Post –0.311 –0.933$$^{***}$$ –0.259 –0.799$$^{***}$$ –0.025 –0.077$$^{**}$$ (0.526) (0.335) (0.503) (0.323) (0.059) (0.032) Controls Yes Yes Yes Yes Yes Yes Year fixed effects Yes Yes Yes Yes Yes Yes Firm fixed effects Yes Yes Yes Yes Yes Yes Observations 1,168 4,801 1,168 4,801 1,168 4,801 Adjusted R-squared 0.292 0.317 0.242 0.299 0.599 0.616 D. Financial distress Payout Repurchase Dividend High Low High Low High Low (3) (4) (1) (2) (1) (2) Treat $$\times$$ Post –1.153$$^{**}$$ –0.665 –1.082$$^{**}$$ –0.686 –0.076$$^{**}$$ 0.020 (0.509) (0.538) (0.434) (0.420) (0.032) (0.096) Controls Yes Yes Yes Yes Yes Yes Year fixed effects Yes Yes Yes Yes Yes Yes Firm fixed effects Yes Yes Yes Yes Yes Yes Observations 1,993 1,984 1,993 1,984 1,993 1,984 Adjusted R-squared 0.356 0.282 0.320 0.273 0.659 0.603 E. Analyst coverage Payout Repurchase Dividend Analyst coverage Analyst coverage Analyst coverage High Low High Low High Low Treat $$\times$$ Post –0.332 –0.333 –0.322 –0.282 –0.052 –0.053 (0.622) (0.647) (0.551) (0.562) (0.056) (0.089) Controls Yes Yes Yes Yes Yes Yes Year fixed effects Yes Yes Yes Yes Yes Yes Firm fixed effects Yes Yes Yes Yes Yes Yes Observations 2,025 2,301 2,025 2,301 2,025 2,301 Adjusted R-squared 0.337 0.316 0.294 0.316 0.661 0.571 D. Financial distress Payout Repurchase Dividend High Low High Low High Low (3) (4) (1) (2) (1) (2) Treat $$\times$$ Post –1.153$$^{**}$$ –0.665 –1.082$$^{**}$$ –0.686 –0.076$$^{**}$$ 0.020 (0.509) (0.538) (0.434) (0.420) (0.032) (0.096) Controls Yes Yes Yes Yes Yes Yes Year fixed effects Yes Yes Yes Yes Yes Yes Firm fixed effects Yes Yes Yes Yes Yes Yes Observations 1,993 1,984 1,993 1,984 1,993 1,984 Adjusted R-squared 0.356 0.282 0.320 0.273 0.659 0.603 E. Analyst coverage Payout Repurchase Dividend Analyst coverage Analyst coverage Analyst coverage High Low High Low High Low Treat $$\times$$ Post –0.332 –0.333 –0.322 –0.282 –0.052 –0.053 (0.622) (0.647) (0.551) (0.562) (0.056) (0.089) Controls Yes Yes Yes Yes Yes Yes Year fixed effects Yes Yes Yes Yes Yes Yes Firm fixed effects Yes Yes Yes Yes Yes Yes Observations 2,025 2,301 2,025 2,301 2,025 2,301 Adjusted R-squared 0.337 0.316 0.294 0.316 0.661 0.571 This table reports the difference-in-differences estimation results of $$Y_{\it it}=\alpha_{\it ij}+\alpha_{t}+\beta {\it Treat}_{\it ij} \times {\it Post}_{\it ijt} +\gamma X_{\it it-1} +\epsilon_{\it ijt}$$ on subsamples partitioned on measures of the stakes of the merging lenders (panel A), power of the merging shareholder (panel B), dividend restriction covenants (panel C), firm leverage (panel D), and analyst coverage (panel E). The dependent variables in Columns 1 and 2 are Payout, and in Columns 3 and 4 are Repurchase, and in Columns 5 and 6 are Dividend. Treat equals one if the firm is a treated firm of the merger, and zero otherwise. Post equals one if the firm-year observation is after the merger. All regressions include year fixed effects and merger-firm fixed effects. Standard errors are clustered by merger. Significance at 1%, 5%, and 10% levels are indicated by ***, **, and *, respectively. 2.6 Conflict between the merging shareholder and other institutional shareholders While the merger between a shareholder and a lender aligns the interest of the two, the merging shareholder’s incentive to maximizing the combined value of shares and loans will necessarily conflict with those of other shareholders. Hence, the extent to which the merging shareholder can push the dividend policy to benefit creditors will also depend on the relative power between the merging shareholder and other shareholders. To measure the relative power, I first note that institutional shareholders are more powerful in affecting corporate decisions, and hence measure the relative power by the ratio between shares owned by the merging shareholder and other institutional shareholders, which I call the relative power ratio. If the relative power ratio is high, that is, the merging institutional shareholders are more powerful, I expect the effect of the mergers on payout to be more pronounced. I again sort the observations into terciles according to the relative power ratio and redo the analysis on the top and the bottom terciles separately. The results are presented in panel B of Table 7. Consistent with the argument that the relative power of the merging shareholder enhances the effect of the mergers on dividend policy, the difference-in-differences estimates are negative and statistically significant for observations in the top terciles (Columns 1, 3, and 5), that is, for observations in which the merging shareholder is relatively more powerful. In contrast, the estimates in Columns 2, 4, and 6, in which the merging shareholder is less powerful, are much smaller in magnitudes and are statistically insignificant. The differences are again statistically significant. The results confirm that the alignment of interest between the merging lender and the merging shareholder also creates conflict of interest between the merging shareholder and other shareholders, further suggesting that the baseline result is driven by the alignment of interest between the merging shareholder and the merging lender. 2.7 Dividend restriction covenants Syndicated loans often contain covenants restricting dividend payment to mitigate the agency cost associated with excessive dividend payment. If the loan contract already contains the dividend restriction covenants, the effect of the mergers in reducing excessive dividend payment may be limited. If the baseline results are driven by reduced shareholder-creditor conflict, the effect should concentrate in loans without such covenants. I partition the sample according to whether the loan has the dividend restriction covenants and then reestimate the baseline difference-in-differences specification on these two subsamples. The results are presented in panel C of Table 7. The difference-in-differences estimates are both negative and statistically significant in the subsample without the dividend restriction covenants (in Columns 2, 4, and 6). In sharp contrast, the estimates are much smaller and are statistically insignificant in the subsample with dividend restriction covenants. The differences of the estimates between these two subsamples are also statistically significant. Overall, these results suggest that dividend restriction covenants do in fact mitigate the agency cost of excessive dividend payment due to the shareholder-creditor conflict, and hence the reduction in shareholder-creditor conflict due to the mergers has limited effects on dividend payment for firms with these covenants. 2.8 Financial distress and the effects of the mergers The conflict of interest between shareholders and creditors often becomes exaggerated when the firm is in financial distress (Smith and Warner 1979; Gilson, John, and Lang 1990; Gilson and Vetsuypens 1993; Ayotte, Hotchkiss, and Thorburn 2013). It follows that the alignment of interest between shareholders and creditors via the merger should have a stronger effect in resolving the conflict. I therefore test this conjecture to provide further support to the argument that the results are driven by reduced conflict of interests between shareholders and creditors. To test this conjecture, I first sort the firms into terciles based on their leverage measured immediately before the merger, and then re-estimate the difference-in-differences specification on the top and bottom terciles separately. The results are presented in panel D of Table 7. Consistent with the conjecture, the results show that the effect concentrates in the top tercile of firms sorted on leverage, i.e., more financially distressed firms. In contrast, the effect is small and statistically insignificant in nondistressed firms. Overall, the results suggest that while shareholders have stronger incentives to pay excessive dividends at the expense of creditors when the firm is in financial distress, the alignment of the interest of shareholders and creditors also has a stronger effect in mitigating the shareholder-creditor conflict when the firm is in financial distress. The results therefore provide further support to the argument that the mergers affect payout policy via their effects on the shareholder-creditor conflict. 2.9 Signaling versus shareholder-creditor conflict While the baseline results are consistent with the theory of the shareholder-creditor conflict, the baseline results can also be driven by reduced information asymmetry between shareholders and the firm (managers). Lenders often have access to private information, and merging with a lender allows the shareholder to gain access to private information. The signaling theory of dividend policy (Miller and Rock 1985; John and Williams 1985) suggests that firms pay dividends to signal the quality of the firm in the presence of information asymmetry. The merger, which reduces the information asymmetry between shareholders and the firms, therefore can reduce the need of the firm to signal with dividend payout. To mitigate this concern, I first note that according to the signaling theory of John and Williams (1985), cash dividends are more effective than stock repurchases in signaling. Empirically, Grullon and Michaely (2004) also find that share repurchases do not signal future performance. It follows that if the baseline results are indeed driven by signaling, the mergers should, at least, have a larger effect on cash dividends than on share repurchases. However, the results presented in Table 6 show that the magnitudes of the coefficients on share repurchases are much larger than those for cash dividends. To the extent that share repurchases are not an effective tool for signaling, this result suggests that the baseline results are unlikely to be driven by signaling. The signaling hypothesis suggests that the mergers decrease payout via their effects on information asymmetry. It follows that the effect should be more pronounced for firms more subject to the asymmetric information problem ex ante. To test whether that is the case, I follow Li and Zhao (2008) to use analyst coverage as a proxy for information asymmetry and partition the sample according to the number of analyst coverage into terciles.7 The results on the bottom and top terciles are presented in panel E of Table 7. Inconsistent with the signaling hypothesis, the results show that the difference-in-differences coefficients are very similar between the low coverage and high coverage firms, further suggesting that the baseline results are unlikely to be driven by signaling. 3. Robustness Checks 3.1 How treated firms change their payout? The baseline results show that the treated firms decrease cash dividends and share repurchases relative to control firms. Treated firms can either cut their dividends or become less likely to increase their dividends. On particular concern is that cutting dividends is often extremely costly to the firms (Denis, Denis, and Sarin 1994; Grullon, Michaely, and Swaminathan 2002), which may undermine the firms’ incentives to reduce dividends even when the shareholder-creditor conflict is reduced. In this subsection, I examine whether treated firms indeed cut dividends. In this regard, I use dividend data from CRSP and code a dividend increase as an increase of cash dividends by more than 10%, and code a dividend decrease as a decrease of cash dividends by more than 10% (Denis, Denis, and Sarin 1994). I code a fiscal year with dividend increase (decrease) if a dividend increase (decrease) occurred during the fiscal year. I then replace the dependent variable in Equation (1) with indicator variables of dividend increase and decrease. The results are presented in panel A of Table 8. For dividend increase, the difference-in-differences coefficients are negative and statistically significant, suggesting that treated firms, relative to control firms, become less likely to increase their dividends after the mergers. In contrast, the coefficients for dividend decrease is much smaller and statistically insignificant, suggesting that treated firms do not become more likely to cut their dividends after the mergers. Table 8 The effect of mergers on dividend changes and share repurchases A. Dividend changes Dividend increase Dividend decrease (1) (2) (3) (4) Treat $$\times$$ Post –0.015$$^{**}$$ –0.016$$^{**}$$ –0.005 –0.004 (0.007) (0.007) (0.031) (0.032) Controls Yes Yes Year fixed effects Yes Yes Yes Yes Firm fixed effects Yes Yes Yes Yes Observations 5,969 5,969 5,969 5,969 Adjusted R-squared 0.275 0.275 0.074 0.075 B. Share repurchases Announcement Actual (1) (2) (3) (4) Treat $$\times$$ Post –0.037$$^{***}$$ –0.041$$^{***}$$ –0.015$$^{***}$$ –0.016$$^{***}$$ (0.012) (0.012) (0.004) (0.005) Controls Yes Yes Year fixed effects Yes Yes Yes Yes Firm fixed effects Yes Yes Yes Yes Observations 5,969 5,969 5,969 5,969 Adjusted R-squared 0.164 0.174 0.024 0.027 A. Dividend changes Dividend increase Dividend decrease (1) (2) (3) (4) Treat $$\times$$ Post –0.015$$^{**}$$ –0.016$$^{**}$$ –0.005 –0.004 (0.007) (0.007) (0.031) (0.032) Controls Yes Yes Year fixed effects Yes Yes Yes Yes Firm fixed effects Yes Yes Yes Yes Observations 5,969 5,969 5,969 5,969 Adjusted R-squared 0.275 0.275 0.074 0.075 B. Share repurchases Announcement Actual (1) (2) (3) (4) Treat $$\times$$ Post –0.037$$^{***}$$ –0.041$$^{***}$$ –0.015$$^{***}$$ –0.016$$^{***}$$ (0.012) (0.012) (0.004) (0.005) Controls Yes Yes Year fixed effects Yes Yes Yes Yes Firm fixed effects Yes Yes Yes Yes Observations 5,969 5,969 5,969 5,969 Adjusted R-squared 0.164 0.174 0.024 0.027 This table reports the difference-in-differences estimation results of $$Y_{\it it}=\alpha_{ij}+\alpha_{t}+\beta {\it Treat}_{\it ij} \times {\it Post}_{\it ijt} +\gamma X_{\it it-1} +\epsilon_{\it ijt}$$, with panel A for dividend changes and panel B for share repurchases. In panel A, the dependent variable in Columns 1 and 2 is Dividend increase, which equals one if the firm increases cash dividends during the fiscal year, and zero otherwise, and the dependent variable in Columns 3 and 4 is Dividend decrease, which equals one if the firm decreases cash dividends during the fiscal year, and zero otherwise. In panel B, the dependent variable in Columns 1 and 2 is Announce, which equals one if the firm announces share repurchases during the fiscal year, and zero otherwise, and the dependent variable in Columns 3 and 4 is Actual, which equals one if the firm actually conducts share repurchases during the fiscal year, and zero otherwise. Treat equals one if the firm is a treated firm of the merger, and zero otherwise. Post equals one if the firm-year observation is after the merger. All regressions include year fixed effects and firm fixed effects. Standard errors are clustered by merger. Significance at 1%, 5%, and 10% levels are indicated by ***, **, and *, respectively. Table 8 The effect of mergers on dividend changes and share repurchases A. Dividend changes Dividend increase Dividend decrease (1) (2) (3) (4) Treat $$\times$$ Post –0.015$$^{**}$$ –0.016$$^{**}$$ –0.005 –0.004 (0.007) (0.007) (0.031) (0.032) Controls Yes Yes Year fixed effects Yes Yes Yes Yes Firm fixed effects Yes Yes Yes Yes Observations 5,969 5,969 5,969 5,969 Adjusted R-squared 0.275 0.275 0.074 0.075 B. Share repurchases Announcement Actual (1) (2) (3) (4) Treat $$\times$$ Post –0.037$$^{***}$$ –0.041$$^{***}$$ –0.015$$^{***}$$ –0.016$$^{***}$$ (0.012) (0.012) (0.004) (0.005) Controls Yes Yes Year fixed effects Yes Yes Yes Yes Firm fixed effects Yes Yes Yes Yes Observations 5,969 5,969 5,969 5,969 Adjusted R-squared 0.164 0.174 0.024 0.027 A. Dividend changes Dividend increase Dividend decrease (1) (2) (3) (4) Treat $$\times$$ Post –0.015$$^{**}$$ –0.016$$^{**}$$ –0.005 –0.004 (0.007) (0.007) (0.031) (0.032) Controls Yes Yes Year fixed effects Yes Yes Yes Yes Firm fixed effects Yes Yes Yes Yes Observations 5,969 5,969 5,969 5,969 Adjusted R-squared 0.275 0.275 0.074 0.075 B. Share repurchases Announcement Actual (1) (2) (3) (4) Treat $$\times$$ Post –0.037$$^{***}$$ –0.041$$^{***}$$ –0.015$$^{***}$$ –0.016$$^{***}$$ (0.012) (0.012) (0.004) (0.005) Controls Yes Yes Year fixed effects Yes Yes Yes Yes Firm fixed effects Yes Yes Yes Yes Observations 5,969 5,969 5,969 5,969 Adjusted R-squared 0.164 0.174 0.024 0.027 This table reports the difference-in-differences estimation results of $$Y_{\it it}=\alpha_{ij}+\alpha_{t}+\beta {\it Treat}_{\it ij} \times {\it Post}_{\it ijt} +\gamma X_{\it it-1} +\epsilon_{\it ijt}$$, with panel A for dividend changes and panel B for share repurchases. In panel A, the dependent variable in Columns 1 and 2 is Dividend increase, which equals one if the firm increases cash dividends during the fiscal year, and zero otherwise, and the dependent variable in Columns 3 and 4 is Dividend decrease, which equals one if the firm decreases cash dividends during the fiscal year, and zero otherwise. In panel B, the dependent variable in Columns 1 and 2 is Announce, which equals one if the firm announces share repurchases during the fiscal year, and zero otherwise, and the dependent variable in Columns 3 and 4 is Actual, which equals one if the firm actually conducts share repurchases during the fiscal year, and zero otherwise. Treat equals one if the firm is a treated firm of the merger, and zero otherwise. Post equals one if the firm-year observation is after the merger. All regressions include year fixed effects and firm fixed effects. Standard errors are clustered by merger. Significance at 1%, 5%, and 10% levels are indicated by ***, **, and *, respectively. Similarly, I also examine how the mergers affect treated firms’ propensity to announce share repurchases and to actually conduct share repurchases. To this end, I follow Grullon and Michaely (2004) to collect share repurchases information from the SDC. I code Announce to be one if the firm announces share repurchases during the fiscal year, and zero otherwise; I code Actual to be one if the firm announces share repurchases during the fiscal year and the transactions are completed (which may or may not be completed during the fiscal year). I then replace the dependent variable in Equation (1) with Announce and Actual. The results are presented in panel B of Table 8. The difference-in-differences coefficients are all negative and statistically significant, suggesting that, relative to control firms, treated firms are less likely to announce and complete share repurchases. 3.2 Changing positions by the lenders and the shareholders The DealScan data only report loan allocations at origination, and the lenders can sell their loans in the secondary market. It is therefore possible that the lenders may have already sold the loans at the time of the merger, and consequently the merger may not affect the shareholder-creditor conflict at all. Although such noise may only bias against any finding, I still address this problem in this subsection to ensure the robustness of the results and to ensure that the results are not driven by something else. I first focus on a subsample in which the merging lender is a lead bank of the loan. In syndicated loans, lead banks screen and monitor the borrowers and performing those tasks requires them to have a stake in the firm, that is, lead banks often do not (completely) sell the loans they lead. The estimation results on this subsample are presented in panel A of Table 9. The difference-in-differences estimates all remain negative and statistically significant. Table 9 The effect of changing positions of the lenders and the shareholders A. Lead lenders only Payout Repurchase Dividend (1) (2) (3) (4) (5) (6) Treat $$\times$$ Post –0.962$$^{**}$$ –0.978$$^{**}$$ –0.768$$^{*}$$ –0.773$$^{*}$$ –0.095$$^{***}$$ –0.107$$^{***}$$ (0.466) (0.445) (0.419) (0.395) (0.033) (0.034) Controls Yes Yes Yes Year fixed effects Yes Yes Yes Yes Yes Yes Firm fixed effects Yes Yes Yes Yes Yes Yes Observations 2,468 2,468 2,468 2,468 2,468 2,468 Adjusted R-squared 0.311 0.324 0.290 0.306 0.591 0.594 B. Mergers within 1 year of loan origination Treat $$\times$$ Post –0.943$$^{*}$$ –0.980$$^{**}$$ –0.879$$^{**}$$ –0.914$$^{**}$$ –0.105$$^{*}$$ –0.104$$^{**}$$ (0.490) (0.451) (0.423) (0.401) (0.054) (0.047) Controls Yes Yes Yes Year fixed effects Yes Yes Yes Yes Yes Yes Firm fixed effects Yes Yes Yes Yes Yes Yes Observations 2,090 2,090 2,090 2,090 2,090 2,090 Adjusted R-squared 0.243 0.284 0.212 0.249 0.554 0.558 A. Lead lenders only Payout Repurchase Dividend (1) (2) (3) (4) (5) (6) Treat $$\times$$ Post –0.962$$^{**}$$ –0.978$$^{**}$$ –0.768$$^{*}$$ –0.773$$^{*}$$ –0.095$$^{***}$$ –0.107$$^{***}$$ (0.466) (0.445) (0.419) (0.395) (0.033) (0.034) Controls Yes Yes Yes Year fixed effects Yes Yes Yes Yes Yes Yes Firm fixed effects Yes Yes Yes Yes Yes Yes Observations 2,468 2,468 2,468 2,468 2,468 2,468 Adjusted R-squared 0.311 0.324 0.290 0.306 0.591 0.594 B. Mergers within 1 year of loan origination Treat $$\times$$ Post –0.943$$^{*}$$ –0.980$$^{**}$$ –0.879$$^{**}$$ –0.914$$^{**}$$ –0.105$$^{*}$$ –0.104$$^{**}$$ (0.490) (0.451) (0.423) (0.401) (0.054) (0.047) Controls Yes Yes Yes Year fixed effects Yes Yes Yes Yes Yes Yes Firm fixed effects Yes Yes Yes Yes Yes Yes Observations 2,090 2,090 2,090 2,090 2,090 2,090 Adjusted R-squared 0.243 0.284 0.212 0.249 0.554 0.558 This table reports the difference-in-differences estimation results to address the possibility that lenders or institutional shareholders may change their positions after loan origination or after the merger. In panel A, the estimation is performed on a subsample in which the merging lender is a lead lender of the loan; in panel B, the estimation is performed on mergers that occur within 1 year of loan origination and over a 2-year window around the merger. The dependent variables in Columns 1 and 2 are Payout, in Columns 3 and 4 are Repurchase, and in Columns 5 and 6 are Dividend. Treat equals one if the firm is a treated firm of the merger, and zero otherwise. Post equals one if the firm-year observation is after the merger. All regressions include year fixed effects and merger-firm fixed effects. Standard errors are clustered by merger. Significance at 1%, 5%, and 10% levels are indicated by ***, **, and *, respectively. Table 9 The effect of changing positions of the lenders and the shareholders A. Lead lenders only Payout Repurchase Dividend (1) (2) (3) (4) (5) (6) Treat $$\times$$ Post –0.962$$^{**}$$ –0.978$$^{**}$$ –0.768$$^{*}$$ –0.773$$^{*}$$ –0.095$$^{***}$$ –0.107$$^{***}$$ (0.466) (0.445) (0.419) (0.395) (0.033) (0.034) Controls Yes Yes Yes Year fixed effects Yes Yes Yes Yes Yes Yes Firm fixed effects Yes Yes Yes Yes Yes Yes Observations 2,468 2,468 2,468 2,468 2,468 2,468 Adjusted R-squared 0.311 0.324 0.290 0.306 0.591 0.594 B. Mergers within 1 year of loan origination Treat $$\times$$ Post –0.943$$^{*}$$ –0.980$$^{**}$$ –0.879$$^{**}$$ –0.914$$^{**}$$ –0.105$$^{*}$$ –0.104$$^{**}$$ (0.490) (0.451) (0.423) (0.401) (0.054) (0.047) Controls Yes Yes Yes Year fixed effects Yes Yes Yes Yes Yes Yes Firm fixed effects Yes Yes Yes Yes Yes Yes Observations 2,090 2,090 2,090 2,090 2,090 2,090 Adjusted R-squared 0.243 0.284 0.212 0.249 0.554 0.558 A. Lead lenders only Payout Repurchase Dividend (1) (2) (3) (4) (5) (6) Treat $$\times$$ Post –0.962$$^{**}$$ –0.978$$^{**}$$ –0.768$$^{*}$$ –0.773$$^{*}$$ –0.095$$^{***}$$ –0.107$$^{***}$$ (0.466) (0.445) (0.419) (0.395) (0.033) (0.034) Controls Yes Yes Yes Year fixed effects Yes Yes Yes Yes Yes Yes Firm fixed effects Yes Yes Yes Yes Yes Yes Observations 2,468 2,468 2,468 2,468 2,468 2,468 Adjusted R-squared 0.311 0.324 0.290 0.306 0.591 0.594 B. Mergers within 1 year of loan origination Treat $$\times$$ Post –0.943$$^{*}$$ –0.980$$^{**}$$ –0.879$$^{**}$$ –0.914$$^{**}$$ –0.105$$^{*}$$ –0.104$$^{**}$$ (0.490) (0.451) (0.423) (0.401) (0.054) (0.047) Controls Yes Yes Yes Year fixed effects Yes Yes Yes Yes Yes Yes Firm fixed effects Yes Yes Yes Yes Yes Yes Observations 2,090 2,090 2,090 2,090 2,090 2,090 Adjusted R-squared 0.243 0.284 0.212 0.249 0.554 0.558 This table reports the difference-in-differences estimation results to address the possibility that lenders or institutional shareholders may change their positions after loan origination or after the merger. In panel A, the estimation is performed on a subsample in which the merging lender is a lead lender of the loan; in panel B, the estimation is performed on mergers that occur within 1 year of loan origination and over a 2-year window around the merger. The dependent variables in Columns 1 and 2 are Payout, in Columns 3 and 4 are Repurchase, and in Columns 5 and 6 are Dividend. Treat equals one if the firm is a treated firm of the merger, and zero otherwise. Post equals one if the firm-year observation is after the merger. All regressions include year fixed effects and merger-firm fixed effects. Standard errors are clustered by merger. Significance at 1%, 5%, and 10% levels are indicated by ***, **, and *, respectively. I then focus on another subsample in which the mergers occur within one year of loan origination. The short time between loan origination and the mergers makes it less likely that the lenders sell the loan in the secondary market before the merger. Furthermore, to ensure that the lenders are also more likely to hold the loans in the post period, I also restrict the analysis to a 2-year window, that is, 1 year before and 1 year after the merger. Focusing on the shorter window also makes it unlikely that the institutional investors sell all the shares in the post period. The estimation results are presented in panel B of Table 9. The difference-in-differences estimates are again all negative and statistically significant. Overall, the results in Table 9 suggest that changing positions of lenders or shareholders are not a major concern. 3.3 Other Robustness Tests To further ensure the robustness of the results, I also conduct the following tests.8 First, I re-scale the payout measures by total assets, and find that the results remain robust. Second, the results can be driven by unobservable characteristics of the merging parties. For example, if the merging lender’s ability to press the shareholders not to pay excessive dividends increases over time, payout of all borrowers of the merging lender will decrease over time. On the other hand, if the merging institutional investor becomes increasingly passive and does not press the firm to pay excessive dividends, all firms whose stocks held by the merging investor can also experience a decline in payout. In these cases, the baseline results can simply be driven by unobservable characteristics of the merging lenders or the merging institutional shareholders but not by the alignment of interests between the lenders and the shareholders. The results can also be driven by behavioral changes of the lenders or shareholders after the merger. To mitigate these concerns, I try two alternative matching methods. In the first method, I require the controls firms to also be held by the merging shareholder at the time of the mergers. In the second method, I require the controls firms also have loans outstanding borrowed from the merging lender at the time of the merger. In both cases, I find similar results as those in Table 6, suggesting that the results are not driven by unobservable lender or shareholder characteristics. Third, to show that the mergers do indeed lead to decreases in shareholder-creditor conflict, I show that treated firms, relative to control firms, issue less debt, after the mergers, suggesting that the mergers reduce claim dilution. I also find that treated firms with lower q reduces capital investment, suggesting that the mergers reduce overinvestment. 4. Conclusion This paper examines the effect of the shareholder-creditor conflict on payout policy using a novel identification strategy. I use mergers between lenders and institutional shareholders of the same firm as natural experiments that generate plausibly exogenous variation in the conflict of interest between shareholders and creditors. I find that following the mergers, treated firms, relative to control firms, reduce their payout, suggesting that shareholders pay excessive dividends to themselves at the expense of creditors when the interests of shareholders and creditors are not aligned. Consistent with the argument that the shareholder-creditor conflict often becomes exaggerated when the firm is in financial distress, I find that the effect is stronger for financially distressed firms. The author thanks Allen Berger, Alan Crane (discussant at the 2017 AFA meeting), Wei Jiang (the Editor), Greg Niehaus, Eric Powers, Sergey Tsyplakov, Donghang Zhang; two anonymous referees; and participants at the 2017 AFA meetings for comments and suggestions. The author also thanks Gerard Pinto for excellent research assistance. Supplementary data can be found on The Review of Financial Studies Web site. Footnotes 1 An exception is Gilje (2016), who measures the shareholder-creditor conflict with changes in financial distress driven by exogenous changes in commodity prices. 2 The evidence in Kalay (1982) that the dividend constraints are often not binding and that many loan contracts do not have direct dividend constraints supports the theoretical prediction of Easterbrook (1984) and John and Kalay (1982). 3 The same threshold used by Jiang, Li, and Shao (2010) to identify dual holders. 4 A similar argument is made by Azar, Schmalz, and Tecu (Forthcoming) and Chu (2017). 5 The results are robust when using unwinsorized variables. 6 The coefficients on repurchase and cash dividends do not add to the coefficients on the total payout due to winsorization. 7 The results are similar if I instead use analyst forecast error or forecast dispersion to measure asymmetric information. 8 To save space, I briefly discuss the results, and the tables can be found in the Online Appendix. References Allen, F., and Michaely. R. 2003 . Payout policy. Handbook of the Economics of Finance 1 : 337 – 429 . Google Scholar CrossRef Search ADS Asquith, P., and Wizman. T. A. 1990 . Event risk, covenants, and bondholder returns in leveraged buyouts. Journal of Financial Economics 27 : 195 – 213 . Google Scholar CrossRef Search ADS Ayotte, K., Hotchkiss, E. S. and Thorburn. K. S. 2013 . Governance in financial distress and bankruptcy. In Oxford handbook of corporate governance , 159 – 288 . Oxford University Press . Google Scholar CrossRef Search ADS Azar, J., Schmalz, M. C. and Tecu. I. Forthcoming . Anti-competitive effects of common ownership. Journal of Finance . Billett, M. T., King, T.-H. D. and Mauer. D. C. 2004 . Bondholder wealth effects in mergers and acquisitions: New evidence from the 1980s and 1990s. Journal of Finance 59 : 107 – 135 . Google Scholar CrossRef Search ADS Black, F. 1976 . The dividend puzzle. Journal of Portfolio Management 2 : 5 – 8 . Google Scholar CrossRef Search ADS Bodnaruk, A., and Rossi. M. 2016 . Dual ownership, returns, and voting in mergers. Journal of Financial Economics 120 : 58 – 80 . Google Scholar CrossRef Search ADS Brockman, P., and Unlu. E. 2009 . Dividend policy, creditor rights, and the agency costs of debt. Journal of Financial Economics 92 : 276 – 99 . Google Scholar CrossRef Search ADS Chava, S., Wang, R. and Zou. H. Forthcoming . Covenants, creditors’ simultaneous equity holdings, and firm investment policies. Journal of Financial and Quantitative Analysis . Chu, Y. 2017 . Debt renegotiation and debt overhang: Evidence from lender mergers. Working Paper , University of South Carolina . Google Scholar CrossRef Search ADS Denis, D. J., Denis, D. K. and Sarin. A. 1994 . The information content of dividend changes: Cash flow signaling, overinvestment, and dividend clienteles. Journal of Financial and Quantitative Analysis 29 : 567 – 87 . Google Scholar CrossRef Search ADS Dhillon, U. S., and Johnson. H. 1994 . The effect of dividend changes on stock and bond prices. Journal of Finance 49 : 281 – 9 . Google Scholar CrossRef Search ADS Easterbrook, F. H. 1984 . Two agency-cost explanations of dividends. American Economic Review 74 : 650 – 9 . Farre-Mensa, J., Michaely, R. and Schmalz. M. 2014 . Payout policy. Annual Review of Financial Economics 6 : 75 – 134 . Google Scholar CrossRef Search ADS Gilje, E. P. 2016 . Do firms engage in risk-shifting? Empirical evidence. Review of Financial Studies , 29 : 2925 – 54 . Google Scholar CrossRef Search ADS Gilson, S. C., John, K. and Lang. L. H. 1990 . Troubled debt restructurings: An empirical study of private reorganization of firms in default. Journal of Financial Economics 27 : 315 – 53 . Google Scholar CrossRef Search ADS Gilson, S. C., and Vetsuypens. M. R. 1993 . CEO compensation in financially distressed firms: An empirical analysis. Journal of Finance 48 : 425 – 58 . Google Scholar CrossRef Search ADS Grullon, G., and Michaely. R. 2004 . The information content of share repurchase programs. Journal of Finance 59 : 651 – 80 . Google Scholar CrossRef Search ADS Grullon, G., Michaely, R. and Swaminathan. B. 2002 . Are dividend changes a sign of firm maturity? Journal of Business 75 : 387 – 424 . Google Scholar CrossRef Search ADS Handjinicolaou, G., and Kalay. A. 1984 . Wealth redistributions or changes in firm value: An analysis of returns to bondholders and stockholders around dividend announcements. Journal of Financial Economics 13 : 35 – 63 . Google Scholar CrossRef Search ADS Hong, H., and Kacperczyk. M. 2010 . Competition and bias. Quarterly Journal of Economics 125 : 1683 – 725 . Google Scholar CrossRef Search ADS Jayaraman, N., and Shastri. K. 1988 . The valuation impacts of specially designated dividends. Journal of Financial and Quantitative Analysis 23 : 301 – 12 . Google Scholar CrossRef Search ADS Jensen, M. C., and Meckling. W. H. 1976 . Theory of the firm: Managerial behavior, agency costs and ownership structure. Journal of Financial Economics 3 : 305 – 60 . Google Scholar CrossRef Search ADS Jiang, W., Li, K. and Shao. P. 2010 . When shareholders are creditors: Effects of the simultaneous holding of equity and debt by non-commercial banking institutions. Review of Financial Studies 23 : 3595 – 637 . Google Scholar CrossRef Search ADS John, K., and Kalay. A. 1982 . Costly contracting and optimal payout constraints. Journal of Finance 37 : 457 – 70 . Google Scholar CrossRef Search ADS John, K., and Williams. J. 1985 . Dividends, dilution, and taxes: A signaling equilibrium. Journal of Finance 40 : 1053 – 70 . Google Scholar CrossRef Search ADS Kalay, A. 1982 . Stockholder-bondholder conflict and dividend constraints. Journal of Financial Economics 10 : 211 – 33 . Google Scholar CrossRef Search ADS Li, K., and Zhao. X. 2008 . Asymmetric information and dividend policy. Financial Management 37 : 673 – 94 . Google Scholar CrossRef Search ADS Miller, M. H., and Rock. K. 1985 . Dividend policy under asymmetric information. Journal of Finance 40 : 1031 – 51 . Google Scholar CrossRef Search ADS Myers, S. C. 1977 . Determinants of corporate borrowing. Journal of Financial Economics 5 : 147 – 75 . Google Scholar CrossRef Search ADS Roberts, M., and Whited. T. 2012 . Endogeneity in empirical corporate finance. Handbook of the economics of finance 2 . Smith, C. W., and Warner. J. B. 1979 . On financial contracting: An analysis of bond covenants. Journal of Financial Economics 7 : 117 – 61 . Google Scholar CrossRef Search ADS Warga, A., and Welch. I. 1993 . Bondholder losses in leveraged buyouts. Review of Financial Studies 6 : 959 – 82 . Google Scholar CrossRef Search ADS © The Author(s) 2017. Published by Oxford University Press on behalf of The Society for Financial Studies. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com. This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices) http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png The Review of Financial Studies Oxford University Press

Shareholder-Creditor Conflict and Payout Policy: Evidence from Mergers between Lenders and Shareholders

The Review of Financial Studies , Volume Advance Article (8) – Dec 11, 2017

Loading next page...
 
/lp/ou_press/shareholder-creditor-conflict-and-payout-policy-evidence-from-mergers-vkmto9mPkW
Publisher
Oxford University Press
Copyright
© The Author(s) 2017. Published by Oxford University Press on behalf of The Society for Financial Studies. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com.
ISSN
0893-9454
eISSN
1465-7368
D.O.I.
10.1093/rfs/hhx142
Publisher site
See Article on Publisher Site

Abstract

Abstract This paper studies how the conflict of interest between shareholders and creditors affects corporate payout policy. Using mergers between lenders and equity holders of the same firm as shocks to the shareholder-creditor conflict, I find that firms pay out less when there is less conflict between shareholders and creditors, suggesting that the shareholder-creditor conflict induces firms to pay out more at the expense of creditors. The effect is stronger for firms in financial distress. Received March 22, 2017; editorial decision October 17, 2017 by Editor Wei Jiang. Authors have furnished an Internet Appendix, which is available on the Oxford University Press Web Site next to the link to the final published paper online. The conflict of interest between shareholders and creditors can induce agency costs in the form of excessive dividend payments, claim dilution, asset substitution, and underinvestment (Jensen and Meckling 1976; Smith and Warner 1979). Excessive dividend payments, in particular, may lead to significant wealth transfers from creditors to shareholders. Black (1976) points out that “there is no easier way for a company to escape the burden of a debt than to pay out all of its assets in the form of a dividend, and leave the creditors holding an empty shell.” While the theory on shareholder-creditor conflict is well established, the empirical relevance of the theory has been understudied. For example, Gilje (2016) finds that exacerbated shareholder-creditor conflict in financial distress causes firms to take less risk, a finding inconsistent with the theory. The same may apply to excessive dividend payment as well. How the shareholder-creditor conflict affects payout policy is therefore empirically important. On the other hand, the existing literature on payout policy has paid little attention to the shareholder-creditor conflict. In a recent survey on payout policy, Farre-Mensa, Michaely, and Schmalz (2014) reviewed no papers related to the shareholder-creditor conflict. In an earlier review article, Allen and Michaely (2003) reviewed only a handful of papers on the relationship between the shareholder-creditor conflict and payout policy, most of which show, at best, indirect evidence for the relevance of the shareholder-creditor conflict. In this paper, I provide direct evidence that the shareholder-creditor conflict affects payout policy. The lack of direct evidence is partly due to the difficulty of empirically measuring the shareholder-creditor conflict. Most existing literature relies on stock and bond price reactions to specific corporate events to infer the shareholder-creditor conflict (Asquith and Wizman 1990; Warga and Welch 1993; Billett, King, and Mauer 2004). Others explore variation in leverage as a proxy for changes in the shareholder-creditor conflict. However, capital structure decisions are often simultaneously determined as investment and payout decisions, and, hence, relying on variation in leverage alone is unlikely to uncover the causal relationship between shareholder-creditor conflict and corporate policy.1 One notable exception is Jiang, Li, and Shao (2010), who use the existence of dual holders, who simultaneously hold equity and debt claims of the same firm, to directly quantify the effect of the shareholder-creditor conflict. In this paper, I build on their idea and also measure the shareholder-creditor conflict with dual holders. Furthermore, I also try to isolate potentially exogenous variation of the conflict by exploiting mergers between shareholders and creditors of the same firm. Under the agency theory (Jensen and Meckling 1976; Myers 1977; Smith and Warner 1979), shareholders may pay excessive dividends at the expense of creditors to maximize shareholder value when the debt contract is in place. Creditors, anticipating potential wealth appropriation by shareholders, often put dividend constraints in the contract and demand higher yields. However, as pointed out by Easterbrook (1984) and John and Kalay (1982), putting overly restrictive dividend constraints often results in overinvestment, and therefore the optimal dividend constraints often do not restore the first-best payout level.2 In equilibrium, a firm still pays out more than the first best in the presence of the shareholder-creditor conflict. When a shareholder merges with a creditor of the same firm, the conflict between the shareholder and the creditor decreases, and, hence, the merger pushes the payout toward the first-best level. I first follow Jiang, Li, and Shao (2010) to identify firms with dual holders and examine the relationship between payout policy and the existence of dual holders using ordinary least squares (OLS) regressions. Consistent with the argument that dual holders mitigate the shareholder-creditor conflict, I find that firms with dual holders pay out less than firms without dual holders. To mitigate the potential endogeneity of dual holders, I next rely on mergers between shareholders and lenders of the same firms to generate plausibly exogenous variation in shareholder-creditor conflict. To construct the sample of mergers between lenders and shareholders, I first identify all mergers between financial firms in SDC and then match the names of the acquirers and targets with lender names in DealScan and shareholder names on Form 13F. I then identify firms who are borrowers of the merging lender and whose stocks are held by the merging institutional shareholder. I further require that the institutional investors hold more than 1% of all shares outstanding at the time of the merger and the lender is allocated more than 10% of the loan at origination. These firms are designated as treated firms. For each treated firm, I then find control firms by matching on firm size, Tobin’s q, institutional ownership, and leverage, and at the same time require the control firms to also have bank loans outstanding. In a difference-in-differences framework with a 6-year window, I find that treated firms reduce payout, as measured by total payout, share repurchases, or cash dividends, relative to control firms after the mergers. The result is consistent with the argument that the shareholder-creditor conflict results in wealth transfers from creditors to shareholders in the form of excessive dividend payout, and the merger between a lender and a shareholder of the same firm aligns the interests of the two and therefore leads to lower payout. The identification of the difference-in-differences estimation relies on the parallel trend condition, that is, the outcome variables have parallel trends in the absence of treatment. While the parallel trend condition is untestable, I follow the advice of Roberts and Whited (2012) to examine the dynamics of the effect of the mergers on payout policy. If the baseline results are driven by nonparallel trends between treated and control firms, the results are likely to show up before the mergers. In contrast, I find that the effect appears only after the mergers, suggesting that the baseline results are unlikely to be driven by nonparallel trends of treated and control firms. The alignment of the interest of shareholders and creditors should increase with the stakes the lenders have in the treated firm. To this end, I find that the strength of the negative effect of the merger on payout increases with the size of the loan allocated to the merging lender. After merging with the lender, the merging shareholder shifts toward maximizing the combined value of equity and loans held, which may then create a conflict between the merging shareholder and other shareholders, whose goal is to maximize the value of equity only. Hence, the extent to which the merging shareholder is able to affect corporate policy will depend on the merging shareholder’s relative power. I measure the relative power as the ratio between the shares held by the merging shareholder and other institutional shareholders, and find that the negative effect of the mergers on payout concentrates in cases in which the merging shareholder is more powerful. Further analysis shows that the effect comes mostly from firms whose loans do not contain dividend restriction covenants. Dividend restriction covenants mitigate the agency cost of excessive payout, and therefore, the mergers, while mitigating the shareholder-creditor conflict, add little value in reducing corporate payout. To ensure that the result is indeed driven by reduced shareholder-creditor conflict, I further explore whether the effect of the merger is stronger for firms in financial distress. Using leverage as a measure of financial distress, I find that the negative effect of the mergers on payout is stronger for firms in financial distress. The result further suggests that the mergers between shareholders and creditors affect payout policy via its impact on the shareholder-creditor conflict. This paper contributes to the literature on payout policy in the presence of shareholder-creditor conflict. Smith and Warner (1979) find that bonds often contain covenants restricting dividend payments due to shareholder-creditor conflict. John and Kalay (1982) derive optimal payout constraints in the presence of shareholder-creditor conflict. Kalay (1982) provides empirical evidence that payout constraints are set to prevent wealth transfer from debt holders to shareholders. Brockman and Unlu (2009) find that creditors demand a more restrictive payout policy ex ante when creditors have weaker rights ex post. Other papers mostly rely on stock and bond price reactions to unexpected dividend changes to infer the relationship between the shareholder-creditor conflict and payout (Handjinicolaou and Kalay 1984; Jayaraman and Shastri 1988; Dhillon and Johnson 1994). This paper also adds to the recent literature on the effect of dual holders. Jiang, Li, and Shao (2010) find that dual holders lower loan spreads, suggesting that dual holders help mitigate shareholder-creditor conflict. Chava,Wang, and Zou (Forthcoming) find that dual holders reduce the use of covenants restricting capital expenditure, and in the event of covenant violation, firms with dual holders are unlikely to suffer a significant drop in debt issuance or investment expenditure. Bodnaruk and Rossi (2016) find that the existence of dual holders of target firms in M&A deals results in higher merger premiums and larger abnormal bond returns. 1. Sample Construction and the Identification Strategy 1.1 Sample construction The sample construction begins with all mergers between financial firms from 1987 to 2011 in the SDC mergers and acquisitions database. I begin the merger sample from 1987 because this is when the DealScan database starts to have a comprehensive coverage of loans. I stop the sample at 2011 because I need 3 years of data after the merger in the analysis. In the second step, I obtain lenders’ information from the LPC DealScan database, and match the lender names with the names of either the acquirers or the targets of the financial mergers. In matching acquirer names, I not only match the names of the lenders directly involved in the merger but also match the names of the parent companies of the lenders and acquirers. Wherever possible, I use the addresses of the companies in both databases to facilitate the match. After this step, I retain all mergers for which either the acquirer or the target can be matched with a lender in the DealScan database. In the third step, I obtain institutional investors’ information from the Thomson Reuter’s 13F database, and match the investors’ names with the unmatched acquirer or target names from the last step. Again, I not only match the names of companies directly involved but also match the names of their parent companies for acquirers. All matches are manually checked to ensure accuracy. This procedure produces a sample of 369 mergers between a lender in the DealScan database and an institutional investor in the 13F database. The next step is to identify treated firms. I first identify all firms whose loans from the merging lenders are still outstanding at the time of the merger. I then require that the merging institutional investor holds stocks of the firm at the end of the quarter immediately before the merger. I require that the lender participates more than 10% of the loan at origination and the institutional shareholder holds more than 1% of all shares outstanding of the firm.3 I exclude all cases in which either the acquirer or the target is a dual holder of the firm before the merger. For firms treated multiple times, I only retain the first time they are treated. I also exclude firms treated again in less than 3 years after receiving the first treatment. I then exclude firms in financial and utility industries and firms with missing key variables. Finally, I only retain treated firms for which key variables are available 1 year before and 1 year after the mergers. This procedure produces a sample of 238 treated firms involved in 61 mergers. I lose some observations when moving away from the treatment date. The number of treated firms is 236 three years before the mergers and 211 three years after the merger. On average, each merger affects about four firms. The distribution of the mergers across time is presented in Table 1. The mergers are fairly evenly distributed across time, with year 2006 having the greatest number of mergers (ten). Table 1 Distribution of mergers between lenders and institutional shareholders Year Freq. % Cum. 1990 1 1.64 1.64 1991 1 1.64 3.28 1995 1 1.64 4.92 1996 2 3.28 8.2 1997 3 4.92 13.11 1998 4 6.56 19.67 1999 2 3.28 22.95 2000 1 1.64 24.59 2001 3 4.92 29.51 2002 6 9.84 39.34 2003 1 1.64 40.98 2004 2 3.28 44.26 2005 4 6.56 50.82 2006 10 16.39 67.21 2007 1 1.64 68.85 2008 5 8.2 77.05 2009 7 11.48 88.52 2010 3 4.92 93.44 2011 4 6.56 100 Total 61 100 Year Freq. % Cum. 1990 1 1.64 1.64 1991 1 1.64 3.28 1995 1 1.64 4.92 1996 2 3.28 8.2 1997 3 4.92 13.11 1998 4 6.56 19.67 1999 2 3.28 22.95 2000 1 1.64 24.59 2001 3 4.92 29.51 2002 6 9.84 39.34 2003 1 1.64 40.98 2004 2 3.28 44.26 2005 4 6.56 50.82 2006 10 16.39 67.21 2007 1 1.64 68.85 2008 5 8.2 77.05 2009 7 11.48 88.52 2010 3 4.92 93.44 2011 4 6.56 100 Total 61 100 This table presents the yearly distribution of the mergers used in this paper. The mergers are merger and acquisition deals between lenders in the DealScan database and institutional shareholders in the Thomson Reuter’s 13F database. Table 1 Distribution of mergers between lenders and institutional shareholders Year Freq. % Cum. 1990 1 1.64 1.64 1991 1 1.64 3.28 1995 1 1.64 4.92 1996 2 3.28 8.2 1997 3 4.92 13.11 1998 4 6.56 19.67 1999 2 3.28 22.95 2000 1 1.64 24.59 2001 3 4.92 29.51 2002 6 9.84 39.34 2003 1 1.64 40.98 2004 2 3.28 44.26 2005 4 6.56 50.82 2006 10 16.39 67.21 2007 1 1.64 68.85 2008 5 8.2 77.05 2009 7 11.48 88.52 2010 3 4.92 93.44 2011 4 6.56 100 Total 61 100 Year Freq. % Cum. 1990 1 1.64 1.64 1991 1 1.64 3.28 1995 1 1.64 4.92 1996 2 3.28 8.2 1997 3 4.92 13.11 1998 4 6.56 19.67 1999 2 3.28 22.95 2000 1 1.64 24.59 2001 3 4.92 29.51 2002 6 9.84 39.34 2003 1 1.64 40.98 2004 2 3.28 44.26 2005 4 6.56 50.82 2006 10 16.39 67.21 2007 1 1.64 68.85 2008 5 8.2 77.05 2009 7 11.48 88.52 2010 3 4.92 93.44 2011 4 6.56 100 Total 61 100 This table presents the yearly distribution of the mergers used in this paper. The mergers are merger and acquisition deals between lenders in the DealScan database and institutional shareholders in the Thomson Reuter’s 13F database. Next, I follow a similar procedure like in Hong and Kacperczyk (2010) to find control firms. First, I exclude all firms ever treated by the mergers (not only those treated firms identified above). Second, I require all control firms to also have bank loans outstanding at the time of the mergers. Third, I exclude all firms with dual holders during the fiscal years of $$[t-3, t+3]$$, in which year $$t$$ is the year during which the merger occurred. Finally, I require control firms to be in the same quintiles sorted based on total assets, Tobin’s q, leverage, and the percentage of institutional ownership. I match control firms based on total assets and Tobin’s q because they are important determinants of payout policy. I match control firms based on leverage and institutional ownership because treated firms, by construction, have debt in their capital structure and are owned by institutional shareholders. I then rank the control firms based on the differences of total assets, Tobin’s q, leverage, and institutional ownership relative to their corresponding treated firms. I compute the rank of the differences for each of these four variables, and then compute the total rank across all four variables. I retain control firms with the five lowest total ranks. That is, for each treated firm, I retain at most five control firms based on their total ranks. This procedure produces a sample of 894 control firms. The number of control firms is 762 three years before the mergers and 705 three years after the merger. The empirical methodology requires specifying a testing window. In choosing the appropriate time window, the trade-off is between a long window that may incorporate information unrelated to the mergers and a short window that contains too few observations. In the baseline specification, I choose a 6-year window, that is, 3 years before and 3 years after the mergers. To ensure a clean identification, I discard firm fiscal years during which the mergers occurred. To ensure robustness, I also try 2-, 4-, and 10-year windows and find similar results. The final step of sample construction involves matching both treated and control firms in the sample with their financial information from Compustat, institutional ownership information from 13F, and detailed loan information from DealScan. 1.2 The identification strategy I use the mergers between lenders and institutional shareholders of the same firm as shocks to the shareholder-creditor conflict. When a lender and an institutional shareholder of the same firm merge, the conflict of interests between the lender and the shareholder is reduced. On the other hand, lenders often lend to hundreds of firms at each point in time and are therefore unlikely to make merger decisions based on factors related to one particular firm. Similarly, institutional shareholders also often hold stocks of many firms at each point in time and are also unlikely to pursue mergers based on factors related to one particular firm.4 As such, the mergers between lenders and institutional shareholders are likely to satisfy both the relevance and the exclusion conditions. To identify the effect of the shareholder-creditor conflict on payout policy, I adopt the difference-in-differences specification as follows: \begin{equation} Y_{\it it}=\alpha_{i}+\alpha_{t}+\beta {\it Treat}_{i} \times {\it Post}_{\it it} +\gamma X_{it-1} +\epsilon_{\it it}, \label{did} \end{equation} (1) where $$Y_{\it it}$$ is measures of payout of firm $$i$$ in year $$t$$; $${\it Treat}_{i}$$ equals one if firm $$i$$ is a treated firm, and zero otherwise; $${\it Post}_{\it it}$$ equals one if the firm year observation is after the announcement of the mergers; $$\alpha_{i}$$ is the firm fixed effects; $$\alpha_{t}$$ is the year fixed effects; and $$X_{\it it-1}$$ is a vector of control variables. In this specification, $${\it Treat}_{i}$$ and $${\it Post}_{\it it}$$ are subsumed by the firm fixed effects and the year fixed effects, respectively. The difference-in-differences coefficient estimate $$\beta$$ captures the marginal effect of the mergers on payout policy. To account for the potential correlation between firms affected by the same merger, I cluster standard errors by merger. However, the results are robust if I instead cluster standard errors by firm. One potential challenge to the identification strategy is that the mergers may be motivated by efficiency gains associated with reduced shareholder-credit conflict. If that is the case, the econometric exogeneity of these mergers no longer holds. However, economically, even if the mergers are motivated by the efficiency gains, $$\beta$$ still captures the effect of reduced shareholder-credit conflict on payout policy, one of the mechanisms through which the efficiency gains are realized. Furthermore, for the efficiency gains associated with reduced shareholder-creditor conflict to drive the mergers, it is still necessary that the merging parties’ stakes in the treated firms are sufficiently large. In the results reported in the Online Appendix, I only focus on mergers in which the merging parties’ stakes are small, and I still find similar results, mitigating the concern that the results are driven by mergers motivated by efficiency gains. 1.3 Variables and summary statistics I use three measures of payout, Payout, defined as total payout (DVC+PRSTKC) scaled by the market value of common equity (PRCC_F $$\times$$ CSHO), Repurchase, defined as share repurchases (PRSTKC) scaled by the market value of common equity, and Dividend, defined as cash dividend (DVC) scaled by the market value of common equity. The control variables include: log assets – the natural logarithm of total assets (AT), Tobin’s q – the market value of total assets (PRCC_F$$\times$$CSHO-CEQ+AT) divided by the book value of total assets (AT), Cash – cash and short term investment (CHE) scaled by total assets (AT), Age – the number of years the firm appeared in Compustat, Leverage – total debt (DLTT $$+$$ DLC) scaled by total assets (AT). Tangibility – total property, plant, and equipment (PPENT) scaled by total assets (AT), and Sales growth – the growth rate of sales (SALE). Table 2 reports the summary statistics of all variables used in the empirical analysis. All variables are winsorized at the 1% and 99% levels.5 The table shows that the average dividend yield of the sample is 1.17%, and the average total payout yield is 3.14%. The average total assets is about $\$$ 3 billion dollars, suggesting that the firms included in the sample are relatively large firms in the Compustat universe. The average Tobin’s q is around 1.73, which is similar to the average Tobin’s q of the Compustat universe. The average leverage ratio is about 27.0%, which is slightly higher than an average Compustat firm. The average firm age in the sample is about 21 years old. Table 2 Summary statistics Obs. Mean SD p25 Median p75 Dividend 5,969 1.170 1.675 0.000 0.000 1.523 Repurchase 5,969 2.251 4.311 0.000 0.046 2.660 Payout 5,969 3.137 5.020 0.000 1.390 4.565 Dividend_At 5,969 0.903 1.704 0.000 0.000 1.196 Payout_At 5,969 3.355 5.555 0.000 1.087 4.146 Repurchase_At 5,969 2.391 4.864 0.000 0.030 2.499 Capex 5,944 7.377 9.237 2.289 4.167 8.114 R&D 5,969 2.496 4.915 0.000 0.000 2.774 Acquisition 5,692 4.620 12.036 0.000 0.119 3.061 Debt financing 5,965 1.400 10.713 $$-$$2.831 0.000 5.009 Equity financing 5,789 2.013 5.681 0.069 0.410 1.314 Change in cash 5,969 0.825 6.221 $$-$$1.306 0.202 2.574 log assets 5,969 7.146 1.501 6.175 7.068 8.212 Tobin’s q 5,969 1.725 0.930 1.158 1.472 1.963 Cash 5,969 0.109 0.130 0.019 0.060 0.147 Leverage 5,969 0.270 0.188 0.133 0.255 0.380 Tangibility 5,969 0.321 0.248 0.123 0.245 0.470 Sales growth 5,969 0.138 0.381 $$-$$0.006 0.080 0.204 Age 5,969 20.971 11.509 11.000 19.000 31.000 Analyst coverage 5,969 8.786 6.785 4.000 7.000 13.000 Obs. Mean SD p25 Median p75 Dividend 5,969 1.170 1.675 0.000 0.000 1.523 Repurchase 5,969 2.251 4.311 0.000 0.046 2.660 Payout 5,969 3.137 5.020 0.000 1.390 4.565 Dividend_At 5,969 0.903 1.704 0.000 0.000 1.196 Payout_At 5,969 3.355 5.555 0.000 1.087 4.146 Repurchase_At 5,969 2.391 4.864 0.000 0.030 2.499 Capex 5,944 7.377 9.237 2.289 4.167 8.114 R&D 5,969 2.496 4.915 0.000 0.000 2.774 Acquisition 5,692 4.620 12.036 0.000 0.119 3.061 Debt financing 5,965 1.400 10.713 $$-$$2.831 0.000 5.009 Equity financing 5,789 2.013 5.681 0.069 0.410 1.314 Change in cash 5,969 0.825 6.221 $$-$$1.306 0.202 2.574 log assets 5,969 7.146 1.501 6.175 7.068 8.212 Tobin’s q 5,969 1.725 0.930 1.158 1.472 1.963 Cash 5,969 0.109 0.130 0.019 0.060 0.147 Leverage 5,969 0.270 0.188 0.133 0.255 0.380 Tangibility 5,969 0.321 0.248 0.123 0.245 0.470 Sales growth 5,969 0.138 0.381 $$-$$0.006 0.080 0.204 Age 5,969 20.971 11.509 11.000 19.000 31.000 Analyst coverage 5,969 8.786 6.785 4.000 7.000 13.000 This table reports the summary statistics of the variables used in this paper. The variables are Dividend, cash dividend (DVC) scaled by market value of common stocks (PRCC_F $$\times$$ CSHO); Payout, total payout (DVC+PRSTKC) scaled by the market value of common stocks; Repurchase, share repurchases (PRSTKC) scaled by the market value of common stocks; Capex, capital expenditure (CAPX) scaled by total assets (AT); R&D, R&D expense (XRD) scaled by total assets (AT); Acquisition, acquisition expense (AQC) scaled by total assets (AT); Debt financing, changes in total liability (DLCCH-DLTIS-DLTR) scaled by total assets; Change in cash, change in cash holding (CHECH) scaled by total assets, log assets, the natural logarithm of total assets (AT); Tobin’s q, market value of total assets (PRCC_F $$\times$$ CSHO$$+$$AT-CEQ) divided by total assets (AT); Cash, cash holding (CHE) scaled by total assets (AT), Leverage, total liability (DLC+DLTT) scaled by total assets (AT); Tangibility, total property, plant, and equipment (PPENT) scaled by total assets (AT); Sales growth, change in sales (SALE) divided by lagged sales; and Age, the number of years since the firm first appear in Compustat. Table 2 Summary statistics Obs. Mean SD p25 Median p75 Dividend 5,969 1.170 1.675 0.000 0.000 1.523 Repurchase 5,969 2.251 4.311 0.000 0.046 2.660 Payout 5,969 3.137 5.020 0.000 1.390 4.565 Dividend_At 5,969 0.903 1.704 0.000 0.000 1.196 Payout_At 5,969 3.355 5.555 0.000 1.087 4.146 Repurchase_At 5,969 2.391 4.864 0.000 0.030 2.499 Capex 5,944 7.377 9.237 2.289 4.167 8.114 R&D 5,969 2.496 4.915 0.000 0.000 2.774 Acquisition 5,692 4.620 12.036 0.000 0.119 3.061 Debt financing 5,965 1.400 10.713 $$-$$2.831 0.000 5.009 Equity financing 5,789 2.013 5.681 0.069 0.410 1.314 Change in cash 5,969 0.825 6.221 $$-$$1.306 0.202 2.574 log assets 5,969 7.146 1.501 6.175 7.068 8.212 Tobin’s q 5,969 1.725 0.930 1.158 1.472 1.963 Cash 5,969 0.109 0.130 0.019 0.060 0.147 Leverage 5,969 0.270 0.188 0.133 0.255 0.380 Tangibility 5,969 0.321 0.248 0.123 0.245 0.470 Sales growth 5,969 0.138 0.381 $$-$$0.006 0.080 0.204 Age 5,969 20.971 11.509 11.000 19.000 31.000 Analyst coverage 5,969 8.786 6.785 4.000 7.000 13.000 Obs. Mean SD p25 Median p75 Dividend 5,969 1.170 1.675 0.000 0.000 1.523 Repurchase 5,969 2.251 4.311 0.000 0.046 2.660 Payout 5,969 3.137 5.020 0.000 1.390 4.565 Dividend_At 5,969 0.903 1.704 0.000 0.000 1.196 Payout_At 5,969 3.355 5.555 0.000 1.087 4.146 Repurchase_At 5,969 2.391 4.864 0.000 0.030 2.499 Capex 5,944 7.377 9.237 2.289 4.167 8.114 R&D 5,969 2.496 4.915 0.000 0.000 2.774 Acquisition 5,692 4.620 12.036 0.000 0.119 3.061 Debt financing 5,965 1.400 10.713 $$-$$2.831 0.000 5.009 Equity financing 5,789 2.013 5.681 0.069 0.410 1.314 Change in cash 5,969 0.825 6.221 $$-$$1.306 0.202 2.574 log assets 5,969 7.146 1.501 6.175 7.068 8.212 Tobin’s q 5,969 1.725 0.930 1.158 1.472 1.963 Cash 5,969 0.109 0.130 0.019 0.060 0.147 Leverage 5,969 0.270 0.188 0.133 0.255 0.380 Tangibility 5,969 0.321 0.248 0.123 0.245 0.470 Sales growth 5,969 0.138 0.381 $$-$$0.006 0.080 0.204 Age 5,969 20.971 11.509 11.000 19.000 31.000 Analyst coverage 5,969 8.786 6.785 4.000 7.000 13.000 This table reports the summary statistics of the variables used in this paper. The variables are Dividend, cash dividend (DVC) scaled by market value of common stocks (PRCC_F $$\times$$ CSHO); Payout, total payout (DVC+PRSTKC) scaled by the market value of common stocks; Repurchase, share repurchases (PRSTKC) scaled by the market value of common stocks; Capex, capital expenditure (CAPX) scaled by total assets (AT); R&D, R&D expense (XRD) scaled by total assets (AT); Acquisition, acquisition expense (AQC) scaled by total assets (AT); Debt financing, changes in total liability (DLCCH-DLTIS-DLTR) scaled by total assets; Change in cash, change in cash holding (CHECH) scaled by total assets, log assets, the natural logarithm of total assets (AT); Tobin’s q, market value of total assets (PRCC_F $$\times$$ CSHO$$+$$AT-CEQ) divided by total assets (AT); Cash, cash holding (CHE) scaled by total assets (AT), Leverage, total liability (DLC+DLTT) scaled by total assets (AT); Tangibility, total property, plant, and equipment (PPENT) scaled by total assets (AT); Sales growth, change in sales (SALE) divided by lagged sales; and Age, the number of years since the firm first appear in Compustat. To ensure that the treated firms and control firms are comparable, I also compare the means of the variables of treated and control firms measured at the fiscal year ending immediately before the mergers. The results are presented in Table 3. The treated and control firms are similar in all dimensions, as the differences of the key variables between the treated and control firms are small and statistically insignificant. Table 3 Comparing treated and control firms before treatment Treat Control Difference Dividend 1.198 1.193 0.005 (0.119) (0.056) (0.012) Payout 3.385 3.377 0.008 (0.351) (0.172) (0.148) log assets 7.175 7.150 0.025 (0.088) (0.045) (0.098) Tobin’s q 1.694 1.645 0.050 (0.056) (0.027) (0.058) Cash 0.087 0.088 –0.001 (0.006) (0.108) (0.008) Leverage 0.271 0.281 –0.010 (0.011) (0.006) (0.013) Sales growth 0.110 0.137 –0.027 (0.016) (0.016) (0.031) Age 16.485 17.049 –0.564 (0.527) (0.302) (0.638) Inst own 81.762 82.876 –0.114 (1.903) (0.948) (2.061) Treat Control Difference Dividend 1.198 1.193 0.005 (0.119) (0.056) (0.012) Payout 3.385 3.377 0.008 (0.351) (0.172) (0.148) log assets 7.175 7.150 0.025 (0.088) (0.045) (0.098) Tobin’s q 1.694 1.645 0.050 (0.056) (0.027) (0.058) Cash 0.087 0.088 –0.001 (0.006) (0.108) (0.008) Leverage 0.271 0.281 –0.010 (0.011) (0.006) (0.013) Sales growth 0.110 0.137 –0.027 (0.016) (0.016) (0.031) Age 16.485 17.049 –0.564 (0.527) (0.302) (0.638) Inst own 81.762 82.876 –0.114 (1.903) (0.948) (2.061) This table reports the t-test results comparing treated and control firms measured at the fiscal year immediately before the merger announcement date, along the main variables. Variable definitions can be found in the note to Table 2. Standard errors are reported in parentheses. Significances of the difference between treated and control firms at 1%, 5%, and 10% levels are indicated by ***, **, and *, respectively. Table 3 Comparing treated and control firms before treatment Treat Control Difference Dividend 1.198 1.193 0.005 (0.119) (0.056) (0.012) Payout 3.385 3.377 0.008 (0.351) (0.172) (0.148) log assets 7.175 7.150 0.025 (0.088) (0.045) (0.098) Tobin’s q 1.694 1.645 0.050 (0.056) (0.027) (0.058) Cash 0.087 0.088 –0.001 (0.006) (0.108) (0.008) Leverage 0.271 0.281 –0.010 (0.011) (0.006) (0.013) Sales growth 0.110 0.137 –0.027 (0.016) (0.016) (0.031) Age 16.485 17.049 –0.564 (0.527) (0.302) (0.638) Inst own 81.762 82.876 –0.114 (1.903) (0.948) (2.061) Treat Control Difference Dividend 1.198 1.193 0.005 (0.119) (0.056) (0.012) Payout 3.385 3.377 0.008 (0.351) (0.172) (0.148) log assets 7.175 7.150 0.025 (0.088) (0.045) (0.098) Tobin’s q 1.694 1.645 0.050 (0.056) (0.027) (0.058) Cash 0.087 0.088 –0.001 (0.006) (0.108) (0.008) Leverage 0.271 0.281 –0.010 (0.011) (0.006) (0.013) Sales growth 0.110 0.137 –0.027 (0.016) (0.016) (0.031) Age 16.485 17.049 –0.564 (0.527) (0.302) (0.638) Inst own 81.762 82.876 –0.114 (1.903) (0.948) (2.061) This table reports the t-test results comparing treated and control firms measured at the fiscal year immediately before the merger announcement date, along the main variables. Variable definitions can be found in the note to Table 2. Standard errors are reported in parentheses. Significances of the difference between treated and control firms at 1%, 5%, and 10% levels are indicated by ***, **, and *, respectively. 2. Main Results 2.1 OLS results of the effect of dual holders on payout policy Before presenting the results on the effect of the mergers on payout policy, I first present the OLS regression results on the effect of dual holders on payout policy on all firms with DealScan loans. This serves as a check of the external validity of the difference-in-differences tests presented below. Because the number of mergers and the total number of firms affected by the mergers are small, the difference-in-differences tests may be subject to the small sample bias. Furthermore, the difference-in-differences tests rely on the specific source of variation in shareholder-creditor conflict generated by the mergers, which may not be generalizable to other sources of variation. To this end, I start the sample construction with all Compustat firms with loans outstanding in the LPC DealScan database from 1987–2014. I exclude firms in the financial and utilities industries. I then follow the procedure in Jiang, Li, and Shao (2010) to identify firms with dual holders. The summary statistics of this sample is provided in Table A1. Comparing Table A1 with Table 2 shows that the firms in this large sample are smaller, younger, and pay out less than those in Table 2. I then estimate the following specification: \begin{equation} Y_{\it it}=\alpha_{i}+\alpha_{t}+\beta {\it Dual}_{\it it}+\gamma X_{it-1} +\epsilon_{\it it}, \label{ols} \end{equation} (2) The results are presented in Table 4. Consistent with the argument that the existence of dual holders reduces shareholder-creditor conflict, the coefficient estimates on Dual are all negative and statistically significant. While these results are important in showing that the effect of dual holders persists on a large sample of firms, having dual holders can be potentially endogenous. To mitigate this problem, I then rely on the mergers between shareholders and lenders to generate plausibly exogenous variation in dual holders, and examine the effect of the mergers on payout policy in a difference-in-differences setting. Table 4 OLS regression results of the effect of dual holders on payout policy Payout Repurchase Dividend (1) (2) (3) (4) (5) (6) Dual –0.814$$^{***}$$ –0.784$$^{***}$$ –0.746$$^{***}$$ –0.724$$^{***}$$ –0.056$$^{*}$$ –0.052$$^{*}$$ (0.292) (0.287) (0.253) (0.252) (0.029) (0.028) log assets 0.692$$^{***}$$ 0.529$$^{***}$$ 0.159$$^{***}$$ (0.065) (0.052) (0.025) Tobin’s q –0.249$$^{***}$$ –0.138$$^{***}$$ –0.079$$^{***}$$ (0.040) (0.033) (0.012) Cash 3.206$$^{***}$$ 2.781$$^{***}$$ 0.226$$^{*}$$ (0.440) (0.363) (0.124) Leverage –3.258$$^{***}$$ –2.531$$^{***}$$ –0.476$$^{***}$$ (0.241) (0.183) (0.084) Tangibility 0.116 0.279 0.034 (0.469) (0.368) (0.162) Sales growth –0.142$$^{***}$$ –0.082$$^{**}$$ –0.036$$^{***}$$ (0.039) (0.033) (0.008) Age –0.056$$^{***}$$ –0.014 –0.034$$^{***}$$ (0.012) (0.010) (0.005) Constant 3.255$$^{***}$$ 1.126$$^{**}$$ 1.430$$^{***}$$ –0.542 1.595$$^{***}$$ 1.225$$^{***}$$ (0.208) (0.444) (0.153) (0.344) (0.078) (0.153) Year fixed effects Yes Yes Yes Yes Yes Yes Firm fixed effects Yes Yes Yes Yes Yes Yes Observations 42,919 42,919 42,919 42,919 42,919 42,919 Adjusted R-squared 0.222 0.239 0.153 0.170 0.567 0.572 Payout Repurchase Dividend (1) (2) (3) (4) (5) (6) Dual –0.814$$^{***}$$ –0.784$$^{***}$$ –0.746$$^{***}$$ –0.724$$^{***}$$ –0.056$$^{*}$$ –0.052$$^{*}$$ (0.292) (0.287) (0.253) (0.252) (0.029) (0.028) log assets 0.692$$^{***}$$ 0.529$$^{***}$$ 0.159$$^{***}$$ (0.065) (0.052) (0.025) Tobin’s q –0.249$$^{***}$$ –0.138$$^{***}$$ –0.079$$^{***}$$ (0.040) (0.033) (0.012) Cash 3.206$$^{***}$$ 2.781$$^{***}$$ 0.226$$^{*}$$ (0.440) (0.363) (0.124) Leverage –3.258$$^{***}$$ –2.531$$^{***}$$ –0.476$$^{***}$$ (0.241) (0.183) (0.084) Tangibility 0.116 0.279 0.034 (0.469) (0.368) (0.162) Sales growth –0.142$$^{***}$$ –0.082$$^{**}$$ –0.036$$^{***}$$ (0.039) (0.033) (0.008) Age –0.056$$^{***}$$ –0.014 –0.034$$^{***}$$ (0.012) (0.010) (0.005) Constant 3.255$$^{***}$$ 1.126$$^{**}$$ 1.430$$^{***}$$ –0.542 1.595$$^{***}$$ 1.225$$^{***}$$ (0.208) (0.444) (0.153) (0.344) (0.078) (0.153) Year fixed effects Yes Yes Yes Yes Yes Yes Firm fixed effects Yes Yes Yes Yes Yes Yes Observations 42,919 42,919 42,919 42,919 42,919 42,919 Adjusted R-squared 0.222 0.239 0.153 0.170 0.567 0.572 This table reports the OLS estimation results of $$Y_{\it it}=\alpha_{\it ij}+\alpha_{t}+\beta {\it Treat}_{\it ij} \times {\it Post}_{\it ijt} +\gamma X_{\it it-1} +\epsilon_{\it ijt}$$. The dependent variable in Columns 1 and 2 is Payout, the dependent variable in Columns 3 and 4 is Repurchase, and the dependent variable in Columns 5 and 6 is Dividend. Dual equals one if the firm is has a dual holder, and zero otherwise. All regressions include year fixed effects and firm fixed effects. Standard errors are clustered by firm. Significance at 1%, 5%, and 10% levels are indicated by ***, **, and *, respectively. Table 4 OLS regression results of the effect of dual holders on payout policy Payout Repurchase Dividend (1) (2) (3) (4) (5) (6) Dual –0.814$$^{***}$$ –0.784$$^{***}$$ –0.746$$^{***}$$ –0.724$$^{***}$$ –0.056$$^{*}$$ –0.052$$^{*}$$ (0.292) (0.287) (0.253) (0.252) (0.029) (0.028) log assets 0.692$$^{***}$$ 0.529$$^{***}$$ 0.159$$^{***}$$ (0.065) (0.052) (0.025) Tobin’s q –0.249$$^{***}$$ –0.138$$^{***}$$ –0.079$$^{***}$$ (0.040) (0.033) (0.012) Cash 3.206$$^{***}$$ 2.781$$^{***}$$ 0.226$$^{*}$$ (0.440) (0.363) (0.124) Leverage –3.258$$^{***}$$ –2.531$$^{***}$$ –0.476$$^{***}$$ (0.241) (0.183) (0.084) Tangibility 0.116 0.279 0.034 (0.469) (0.368) (0.162) Sales growth –0.142$$^{***}$$ –0.082$$^{**}$$ –0.036$$^{***}$$ (0.039) (0.033) (0.008) Age –0.056$$^{***}$$ –0.014 –0.034$$^{***}$$ (0.012) (0.010) (0.005) Constant 3.255$$^{***}$$ 1.126$$^{**}$$ 1.430$$^{***}$$ –0.542 1.595$$^{***}$$ 1.225$$^{***}$$ (0.208) (0.444) (0.153) (0.344) (0.078) (0.153) Year fixed effects Yes Yes Yes Yes Yes Yes Firm fixed effects Yes Yes Yes Yes Yes Yes Observations 42,919 42,919 42,919 42,919 42,919 42,919 Adjusted R-squared 0.222 0.239 0.153 0.170 0.567 0.572 Payout Repurchase Dividend (1) (2) (3) (4) (5) (6) Dual –0.814$$^{***}$$ –0.784$$^{***}$$ –0.746$$^{***}$$ –0.724$$^{***}$$ –0.056$$^{*}$$ –0.052$$^{*}$$ (0.292) (0.287) (0.253) (0.252) (0.029) (0.028) log assets 0.692$$^{***}$$ 0.529$$^{***}$$ 0.159$$^{***}$$ (0.065) (0.052) (0.025) Tobin’s q –0.249$$^{***}$$ –0.138$$^{***}$$ –0.079$$^{***}$$ (0.040) (0.033) (0.012) Cash 3.206$$^{***}$$ 2.781$$^{***}$$ 0.226$$^{*}$$ (0.440) (0.363) (0.124) Leverage –3.258$$^{***}$$ –2.531$$^{***}$$ –0.476$$^{***}$$ (0.241) (0.183) (0.084) Tangibility 0.116 0.279 0.034 (0.469) (0.368) (0.162) Sales growth –0.142$$^{***}$$ –0.082$$^{**}$$ –0.036$$^{***}$$ (0.039) (0.033) (0.008) Age –0.056$$^{***}$$ –0.014 –0.034$$^{***}$$ (0.012) (0.010) (0.005) Constant 3.255$$^{***}$$ 1.126$$^{**}$$ 1.430$$^{***}$$ –0.542 1.595$$^{***}$$ 1.225$$^{***}$$ (0.208) (0.444) (0.153) (0.344) (0.078) (0.153) Year fixed effects Yes Yes Yes Yes Yes Yes Firm fixed effects Yes Yes Yes Yes Yes Yes Observations 42,919 42,919 42,919 42,919 42,919 42,919 Adjusted R-squared 0.222 0.239 0.153 0.170 0.567 0.572 This table reports the OLS estimation results of $$Y_{\it it}=\alpha_{\it ij}+\alpha_{t}+\beta {\it Treat}_{\it ij} \times {\it Post}_{\it ijt} +\gamma X_{\it it-1} +\epsilon_{\it ijt}$$. The dependent variable in Columns 1 and 2 is Payout, the dependent variable in Columns 3 and 4 is Repurchase, and the dependent variable in Columns 5 and 6 is Dividend. Dual equals one if the firm is has a dual holder, and zero otherwise. All regressions include year fixed effects and firm fixed effects. Standard errors are clustered by firm. Significance at 1%, 5%, and 10% levels are indicated by ***, **, and *, respectively. 2.2 Mergers and loan prices The identification relies on the assumption that the mergers between shareholders and lenders of the same firms reduce the shareholder-creditor conflict. Under this assumption, loan prices should increase following the mergers. To test this conjecture, I obtain the secondary loan market data from Thomson Reuters LPC. In collaboration with the Loan Syndication and Trading Association (LSTA), LPC collects daily loan bid and ask prices from market makers. The secondary loan database uses the LIN to identify each individual loan. To obtain the corresponding primary market data of individual loans, I match it to the LPC DealScan data based on LIN and FacilityID. I then run the difference-in-differences specification like in Equation (1), but with the logarithm of loan prices as the dependent variable. To implement, I calculate annual average loan prices and then use the logarithm of the calculated average prices as the dependent variables in the specification. The results are presented in Table 5. Consistent with the argument that the mergers reduce the shareholder-creditor conflict and hence should increase loan prices, the difference-in-differences coefficient estimates are all positive and statistically significant. The economic magnitudes are also significant, with the mergers increasing loan prices by more than 3%. The results therefore support the argument that the mergers between shareholders and lenders of the same firms do in fact result in reduced shareholder-creditor conflict, which pushes up loan prices. Table 5 The effect of mergers on loan prices (1) (2) Treat $$\times$$ Post 0.032$$^{**}$$ 0.033$$^{**}$$ (0.013) (0.014) log assets 0.013 (0.012) Tobin’s q –0.012 (0.010) Cash 0.030 (0.041) Leverage –0.065$$^{**}$$ (0.029) Tangibility –0.101$$^{**}$$ (0.041) Sales growth 0.019$$^{***}$$ (0.006) Age –0.011$$^{**}$$ (0.005) Constant 4.728$$^{***}$$ 4.836$$^{***}$$ (0.071) (0.115) Year fixed effects Yes Yes Facility fixed effects Yes Yes Observations 3,444 3,444 Adjusted R-squared 0.553 0.557 (1) (2) Treat $$\times$$ Post 0.032$$^{**}$$ 0.033$$^{**}$$ (0.013) (0.014) log assets 0.013 (0.012) Tobin’s q –0.012 (0.010) Cash 0.030 (0.041) Leverage –0.065$$^{**}$$ (0.029) Tangibility –0.101$$^{**}$$ (0.041) Sales growth 0.019$$^{***}$$ (0.006) Age –0.011$$^{**}$$ (0.005) Constant 4.728$$^{***}$$ 4.836$$^{***}$$ (0.071) (0.115) Year fixed effects Yes Yes Facility fixed effects Yes Yes Observations 3,444 3,444 Adjusted R-squared 0.553 0.557 This table reports the difference-in-differences estimation results of $$P_{\it it}=\alpha_{i}+\alpha_{t}+\beta {\it Treat}_{i} \times {\it Post}_{\it it} +\epsilon_{\it it}$$ for loan prices. The dependent variables are the logarithms of annual average prices of loan quotes. Post equals one if it is after the merger. All regressions include time fixed effects and facility fixed effects. Standard errors are clustered by merger. Significance at 1%, 5%, and 10% levels are indicated by ***, **, and *, respectively. Table 5 The effect of mergers on loan prices (1) (2) Treat $$\times$$ Post 0.032$$^{**}$$ 0.033$$^{**}$$ (0.013) (0.014) log assets 0.013 (0.012) Tobin’s q –0.012 (0.010) Cash 0.030 (0.041) Leverage –0.065$$^{**}$$ (0.029) Tangibility –0.101$$^{**}$$ (0.041) Sales growth 0.019$$^{***}$$ (0.006) Age –0.011$$^{**}$$ (0.005) Constant 4.728$$^{***}$$ 4.836$$^{***}$$ (0.071) (0.115) Year fixed effects Yes Yes Facility fixed effects Yes Yes Observations 3,444 3,444 Adjusted R-squared 0.553 0.557 (1) (2) Treat $$\times$$ Post 0.032$$^{**}$$ 0.033$$^{**}$$ (0.013) (0.014) log assets 0.013 (0.012) Tobin’s q –0.012 (0.010) Cash 0.030 (0.041) Leverage –0.065$$^{**}$$ (0.029) Tangibility –0.101$$^{**}$$ (0.041) Sales growth 0.019$$^{***}$$ (0.006) Age –0.011$$^{**}$$ (0.005) Constant 4.728$$^{***}$$ 4.836$$^{***}$$ (0.071) (0.115) Year fixed effects Yes Yes Facility fixed effects Yes Yes Observations 3,444 3,444 Adjusted R-squared 0.553 0.557 This table reports the difference-in-differences estimation results of $$P_{\it it}=\alpha_{i}+\alpha_{t}+\beta {\it Treat}_{i} \times {\it Post}_{\it it} +\epsilon_{\it it}$$ for loan prices. The dependent variables are the logarithms of annual average prices of loan quotes. Post equals one if it is after the merger. All regressions include time fixed effects and facility fixed effects. Standard errors are clustered by merger. Significance at 1%, 5%, and 10% levels are indicated by ***, **, and *, respectively. 2.3 Baseline difference-in-differences results I present the baseline results of estimating Equation (1) in Table 6. In Columns 1 and 2, I first present the results for Payout with and without the controls. The mergers, which reduce the conflict between shareholders and creditors, can also potentially affect the control variables, that is, the control variables can be endogenous. Estimating Equation (1) both with and without the controls ensures that the results are not driven by these potentially endogenous control variables. In both columns, the difference-in-differences estimates, that is, the coefficients on Treat $$\times$$ Post are negative and statistically significant. The effect is also economically significant. Taking the coefficient in Column 2, the effect of the merger between a lender and a shareholder leads to a decrease of payout of treated firms, relative to control firms, by more than 25% of the average dividend yield in the sample. Table 6 Baseline results of the effect of mergers on payout policy Payout Repurchase Dividend (1) (2) (3) (4) (5) (6) Treat $$\times$$ Post –0.813$$^{***}$$ –0.832$$^{***}$$ –0.728$$^{***}$$ –0.753$$^{***}$$ –0.061$$^{**}$$ –0.069$$^{***}$$ (0.280) (0.278) (0.261) (0.261) (0.024) (0.022) log assets 1.114$$^{***}$$ 0.912$$^{***}$$ 0.240$$^{***}$$ (0.295) (0.267) (0.085) Tobin’s q –0.074 –0.001 –0.058$$^{*}$$ (0.088) (0.077) (0.033) Cash 3.207$$^{**}$$ 2.552$$^{**}$$ 0.301 (1.393) (1.191) (0.204) Leverage –4.424$$^{***}$$ –4.404$$^{***}$$ –0.108 (0.747) (0.758) (0.176) Tangibility –4.582$$^{***}$$ –4.343$$^{***}$$ –0.200 (1.133) (0.970) (0.279) Sales growth –0.580$$^{***}$$ –0.604$$^{***}$$ 0.018 (0.131) (0.124) (0.035) Age –0.142$$^{***}$$ –0.111$$^{***}$$ –0.041$$^{**}$$ (0.040) (0.034) (0.018) Constant 1.211$$^{*}$$ –1.166 –0.144 –1.708 1.466$$^{***}$$ 0.537 (0.669) (1.642) (0.526) (1.369) (0.364) (0.573) Year fixed effects Yes Yes Yes Yes Yes Yes Firm fixed effects Yes Yes Yes Yes Yes Yes Observations 5,969 5,969 5,969 5,969 5,969 5,969 Adjusted R-squared 0.295 0.314 0.270 0.292 0.611 0.614 Payout Repurchase Dividend (1) (2) (3) (4) (5) (6) Treat $$\times$$ Post –0.813$$^{***}$$ –0.832$$^{***}$$ –0.728$$^{***}$$ –0.753$$^{***}$$ –0.061$$^{**}$$ –0.069$$^{***}$$ (0.280) (0.278) (0.261) (0.261) (0.024) (0.022) log assets 1.114$$^{***}$$ 0.912$$^{***}$$ 0.240$$^{***}$$ (0.295) (0.267) (0.085) Tobin’s q –0.074 –0.001 –0.058$$^{*}$$ (0.088) (0.077) (0.033) Cash 3.207$$^{**}$$ 2.552$$^{**}$$ 0.301 (1.393) (1.191) (0.204) Leverage –4.424$$^{***}$$ –4.404$$^{***}$$ –0.108 (0.747) (0.758) (0.176) Tangibility –4.582$$^{***}$$ –4.343$$^{***}$$ –0.200 (1.133) (0.970) (0.279) Sales growth –0.580$$^{***}$$ –0.604$$^{***}$$ 0.018 (0.131) (0.124) (0.035) Age –0.142$$^{***}$$ –0.111$$^{***}$$ –0.041$$^{**}$$ (0.040) (0.034) (0.018) Constant 1.211$$^{*}$$ –1.166 –0.144 –1.708 1.466$$^{***}$$ 0.537 (0.669) (1.642) (0.526) (1.369) (0.364) (0.573) Year fixed effects Yes Yes Yes Yes Yes Yes Firm fixed effects Yes Yes Yes Yes Yes Yes Observations 5,969 5,969 5,969 5,969 5,969 5,969 Adjusted R-squared 0.295 0.314 0.270 0.292 0.611 0.614 This table reports the baseline difference-in-differences estimation results of $$Y_{\it it}=\alpha_{ij}+\alpha_{t}+\beta {\it Treat}_{\it ij} \times {\it Post}_{\it ijt} +\gamma X_{\it it-1} +\epsilon_{\it ijt}$$. The dependent variable in Columns 1 and 2 is Payout, the dependent variable in Columns 3 and 4 is Repurchase, and the dependent variable in Columns 5 and 6 is Dividend. Treat equals one if the firm is a treated firm of the merger, and zero otherwise. Post equals one if the firm-year observation is after the merger. All regressions include year fixed effects and firm fixed effects. Standard errors are clustered by merger. Significance at 1%, 5%, and 10% levels are indicated by ***, **, and *, respectively. Table 6 Baseline results of the effect of mergers on payout policy Payout Repurchase Dividend (1) (2) (3) (4) (5) (6) Treat $$\times$$ Post –0.813$$^{***}$$ –0.832$$^{***}$$ –0.728$$^{***}$$ –0.753$$^{***}$$ –0.061$$^{**}$$ –0.069$$^{***}$$ (0.280) (0.278) (0.261) (0.261) (0.024) (0.022) log assets 1.114$$^{***}$$ 0.912$$^{***}$$ 0.240$$^{***}$$ (0.295) (0.267) (0.085) Tobin’s q –0.074 –0.001 –0.058$$^{*}$$ (0.088) (0.077) (0.033) Cash 3.207$$^{**}$$ 2.552$$^{**}$$ 0.301 (1.393) (1.191) (0.204) Leverage –4.424$$^{***}$$ –4.404$$^{***}$$ –0.108 (0.747) (0.758) (0.176) Tangibility –4.582$$^{***}$$ –4.343$$^{***}$$ –0.200 (1.133) (0.970) (0.279) Sales growth –0.580$$^{***}$$ –0.604$$^{***}$$ 0.018 (0.131) (0.124) (0.035) Age –0.142$$^{***}$$ –0.111$$^{***}$$ –0.041$$^{**}$$ (0.040) (0.034) (0.018) Constant 1.211$$^{*}$$ –1.166 –0.144 –1.708 1.466$$^{***}$$ 0.537 (0.669) (1.642) (0.526) (1.369) (0.364) (0.573) Year fixed effects Yes Yes Yes Yes Yes Yes Firm fixed effects Yes Yes Yes Yes Yes Yes Observations 5,969 5,969 5,969 5,969 5,969 5,969 Adjusted R-squared 0.295 0.314 0.270 0.292 0.611 0.614 Payout Repurchase Dividend (1) (2) (3) (4) (5) (6) Treat $$\times$$ Post –0.813$$^{***}$$ –0.832$$^{***}$$ –0.728$$^{***}$$ –0.753$$^{***}$$ –0.061$$^{**}$$ –0.069$$^{***}$$ (0.280) (0.278) (0.261) (0.261) (0.024) (0.022) log assets 1.114$$^{***}$$ 0.912$$^{***}$$ 0.240$$^{***}$$ (0.295) (0.267) (0.085) Tobin’s q –0.074 –0.001 –0.058$$^{*}$$ (0.088) (0.077) (0.033) Cash 3.207$$^{**}$$ 2.552$$^{**}$$ 0.301 (1.393) (1.191) (0.204) Leverage –4.424$$^{***}$$ –4.404$$^{***}$$ –0.108 (0.747) (0.758) (0.176) Tangibility –4.582$$^{***}$$ –4.343$$^{***}$$ –0.200 (1.133) (0.970) (0.279) Sales growth –0.580$$^{***}$$ –0.604$$^{***}$$ 0.018 (0.131) (0.124) (0.035) Age –0.142$$^{***}$$ –0.111$$^{***}$$ –0.041$$^{**}$$ (0.040) (0.034) (0.018) Constant 1.211$$^{*}$$ –1.166 –0.144 –1.708 1.466$$^{***}$$ 0.537 (0.669) (1.642) (0.526) (1.369) (0.364) (0.573) Year fixed effects Yes Yes Yes Yes Yes Yes Firm fixed effects Yes Yes Yes Yes Yes Yes Observations 5,969 5,969 5,969 5,969 5,969 5,969 Adjusted R-squared 0.295 0.314 0.270 0.292 0.611 0.614 This table reports the baseline difference-in-differences estimation results of $$Y_{\it it}=\alpha_{ij}+\alpha_{t}+\beta {\it Treat}_{\it ij} \times {\it Post}_{\it ijt} +\gamma X_{\it it-1} +\epsilon_{\it ijt}$$. The dependent variable in Columns 1 and 2 is Payout, the dependent variable in Columns 3 and 4 is Repurchase, and the dependent variable in Columns 5 and 6 is Dividend. Treat equals one if the firm is a treated firm of the merger, and zero otherwise. Post equals one if the firm-year observation is after the merger. All regressions include year fixed effects and firm fixed effects. Standard errors are clustered by merger. Significance at 1%, 5%, and 10% levels are indicated by ***, **, and *, respectively. I then decompose total payout into share repurchases and cash dividends, and present the results in Columns 3–6.6 While the coefficient estimates are still negative and statistically significant in all columns, the coefficients are much larger for share repurchases (Columns 3 and 4) than those for cash dividends (Columns 5 and 6). These results suggest that the mergers, while reducing the shareholder-creditor conflict, have a more pronounced effect on share repurchases. This is probably due to the fact that firms are often able to adjust share repurchases more quickly than cash dividends. 2.4 Addressing identification challenges The consistency of the difference-in-differences estimates depends on the parallel trend condition; that is, the outcome variables should have parallel trends in the absence of the treatment. Although the parallel trend condition is untestable, I follow the advice of Roberts and Whited (2012) to conduct a visual examination of the payout policy around the mergers. Specifically, I examine the evolution of Payout, Repurchase, and Dividend around the mergers for treated and control firms separately and the results are presented in Figure 1, with panel A for Payout, panel B for Repurchase, and panel C for Dividend. The figure shows that the payout variables follow similar trends before the events. After the event, however, although the control firms continue their pre-event trend, the treated firms experience an abrupt change of the trend. The results suggest (although do not prove) that the parallel trend condition is likely to be satisfied. Figure 1 View largeDownload slide Payout around mergers This figure shows the evolution of payout policy of treated and control firms around the merger of lenders and institutional investors. Panel A reports the yearly average of Payout, total payout divided by market value of common equity; panel B reports the yearly average of Repurchase, share repurchases divided by total assets; and panel C reports the yearly average of Dividend, cash dividend divided by market value of common equity. Figure 1 View largeDownload slide Payout around mergers This figure shows the evolution of payout policy of treated and control firms around the merger of lenders and institutional investors. Panel A reports the yearly average of Payout, total payout divided by market value of common equity; panel B reports the yearly average of Repurchase, share repurchases divided by total assets; and panel C reports the yearly average of Dividend, cash dividend divided by market value of common equity. To provide further evidence that the baseline results are not driven by preexisting trend differences between treated and control firms, I conduct a dynamic analysis of the effect of the mergers. To implement, I extend the sample period to 6 years before the mergers and then interact each event year dummy with the treatment dummy, that is, I estimate the following \begin{equation} Y_{\it it}=\alpha_{i}+\Sigma_{k=-6}^{k=3}\beta_k {\it Treat}_{i} \times {\it Year}^{k} +\gamma X_{\it it-1} +\epsilon_{\it it}, \label{placebo} \end{equation} (3) where all variables are defined exactly the same as those in Equation (1), except for $${\it Year}^{k}$$’s, which equals one if the fiscal year is $$k$$ years before the merger, and zero otherwise. If the baseline results are truly driven by the mergers, I expect the $$\beta_k$$’s to be close to zero for all $$k<0$$, and the $$\beta_k$$’s to be negative for some $$k>0$$. On the other hand, if the baseline results are driven by preexisting differences between treated and control firms, the effect may show up before the mergers. The coefficient estimates are provided in Figure 2. Consistent with the argument that the baseline results are truly driven by the mergers, the estimates of $$\beta_k$$’s are all small and statistically insignificant for $$k<0$$, but turn negative and statistically significant for some $$k>0$$. The results therefore provide further confidence that the baseline results are unlikely to be driven by nonparallel trends between treated and control firms. Figure 2 View largeDownload slide Dynamics of the coefficient estimates This figure shows the coefficient estimates from estimating $$Y_{\it it}=\alpha_{i}+\Sigma_{k=-6}^{k=3}\beta_k {\it Treat}_{i} \times {\it Year}^{k} +\gamma X_{it-1} +\epsilon_{\it it}$$ for total payout, repurchases, and dividends. Figure 2 View largeDownload slide Dynamics of the coefficient estimates This figure shows the coefficient estimates from estimating $$Y_{\it it}=\alpha_{i}+\Sigma_{k=-6}^{k=3}\beta_k {\it Treat}_{i} \times {\it Year}^{k} +\gamma X_{it-1} +\epsilon_{\it it}$$ for total payout, repurchases, and dividends. In the Online Appendix, I provide a test using a subsample of mergers in which the merging parties stakes in the treated firms, relative to their total loan and stock portfolio, are below the sample median, and find results similar to those in Table 6. When merging parties’ stakes in the treated firms are small, these mergers are less likely to be motivated by efficiency gains associated with these firms, and hence the results are less likely to be subject to the endogeneity concerns. 2.5 Merging lenders’ stakes and the effect of the mergers I then examine how the stakes of the lenders in the treated firm alter the effects of the mergers on payout policy. I argue that the baseline results are driven by shareholders shifting from maximizing shareholder value toward maximizing combined value of stocks and loans they hold after the merger, and hence the effect should be stronger if the merging lender’s stake in the treated firm is larger. On the other hand, if the results are driven by other unobservable factors correlated with the mergers, it is unlikely to be correlated with the stakes of the merging lenders. To this end, I sort the observations into terciles according to the merging lender’s loan size (the total amount of the loan allocated to the lender) scaled by the firm’s total assets and redo the analysis on the top and bottom terciles separately. The results are presented in panel A of Table 7. The difference-in-differences estimates are negative and statistically significant for observations in the top tercile, and are much smaller and statistically insignificant for observations in the bottom tercile. The differences of the estimates are also statistically significant. Overall, the results are consistent with the argument that the mergers reduce the shareholder-creditor conflict and that the effect is stronger if the stakes of the merging lenders are larger. Table 7 Cross-sectional heterogeneity of the effect of mergers on payout policy A. Lender stake Payout Repurchase Dividend Low High Low High Low High (1) (2) (3) (4) (5) (6) Treat $$\times$$ Post –0.290 –1.095$$^{**}$$ –0.312 –0.914$$^{**}$$ 0.045 –0.142$$^{**}$$ (0.528) (0.487) (0.600) (0.446) (0.110) (0.069) Controls Yes Yes Yes Yes Yes Yes Year fixed effects Yes Yes Yes Yes Yes Yes Firm fixed effects Yes Yes Yes Yes Yes Yes Observations 2,056 1,962 2,056 1,962 2,056 1,962 Adjusted R-squared 0.265 0.351 0.242 0.331 0.553 0.634 B. Shareholder power Payout Repurchase Dividend High Low High Low High Low (1) (2) (3) (4) (5) (6) Treat $$\times$$ Post –1.047$$^{***}$$ –0.457 –0.962$$^{**}$$ –0.382 –0.079$$^{**}$$ 0.023 (0.359) (0.565) (0.422) (0.499) (0.038) (0.065) Controls Yes Yes Yes Yes Yes Yes Year fixed effects Yes Yes Yes Yes Yes Yes Firm fixed effects Yes Yes Yes Yes Yes Yes Observations 1,983 1,995 1,983 1,995 1,983 1,995 Adjusted R-squared 0.329 0.338 0.310 0.297 0.600 0.601 C. Dividend restriction covenants Payout Repurchase Dividend Dividend covenant Dividend covenant Dividend covenant Yes No Yes No Yes No (1) (2) (3) (4) (5) (6) Treat $$\times$$ Post –0.311 –0.933$$^{***}$$ –0.259 –0.799$$^{***}$$ –0.025 –0.077$$^{**}$$ (0.526) (0.335) (0.503) (0.323) (0.059) (0.032) Controls Yes Yes Yes Yes Yes Yes Year fixed effects Yes Yes Yes Yes Yes Yes Firm fixed effects Yes Yes Yes Yes Yes Yes Observations 1,168 4,801 1,168 4,801 1,168 4,801 Adjusted R-squared 0.292 0.317 0.242 0.299 0.599 0.616 A. Lender stake Payout Repurchase Dividend Low High Low High Low High (1) (2) (3) (4) (5) (6) Treat $$\times$$ Post –0.290 –1.095$$^{**}$$ –0.312 –0.914$$^{**}$$ 0.045 –0.142$$^{**}$$ (0.528) (0.487) (0.600) (0.446) (0.110) (0.069) Controls Yes Yes Yes Yes Yes Yes Year fixed effects Yes Yes Yes Yes Yes Yes Firm fixed effects Yes Yes Yes Yes Yes Yes Observations 2,056 1,962 2,056 1,962 2,056 1,962 Adjusted R-squared 0.265 0.351 0.242 0.331 0.553 0.634 B. Shareholder power Payout Repurchase Dividend High Low High Low High Low (1) (2) (3) (4) (5) (6) Treat $$\times$$ Post –1.047$$^{***}$$ –0.457 –0.962$$^{**}$$ –0.382 –0.079$$^{**}$$ 0.023 (0.359) (0.565) (0.422) (0.499) (0.038) (0.065) Controls Yes Yes Yes Yes Yes Yes Year fixed effects Yes Yes Yes Yes Yes Yes Firm fixed effects Yes Yes Yes Yes Yes Yes Observations 1,983 1,995 1,983 1,995 1,983 1,995 Adjusted R-squared 0.329 0.338 0.310 0.297 0.600 0.601 C. Dividend restriction covenants Payout Repurchase Dividend Dividend covenant Dividend covenant Dividend covenant Yes No Yes No Yes No (1) (2) (3) (4) (5) (6) Treat $$\times$$ Post –0.311 –0.933$$^{***}$$ –0.259 –0.799$$^{***}$$ –0.025 –0.077$$^{**}$$ (0.526) (0.335) (0.503) (0.323) (0.059) (0.032) Controls Yes Yes Yes Yes Yes Yes Year fixed effects Yes Yes Yes Yes Yes Yes Firm fixed effects Yes Yes Yes Yes Yes Yes Observations 1,168 4,801 1,168 4,801 1,168 4,801 Adjusted R-squared 0.292 0.317 0.242 0.299 0.599 0.616 D. Financial distress Payout Repurchase Dividend High Low High Low High Low (3) (4) (1) (2) (1) (2) Treat $$\times$$ Post –1.153$$^{**}$$ –0.665 –1.082$$^{**}$$ –0.686 –0.076$$^{**}$$ 0.020 (0.509) (0.538) (0.434) (0.420) (0.032) (0.096) Controls Yes Yes Yes Yes Yes Yes Year fixed effects Yes Yes Yes Yes Yes Yes Firm fixed effects Yes Yes Yes Yes Yes Yes Observations 1,993 1,984 1,993 1,984 1,993 1,984 Adjusted R-squared 0.356 0.282 0.320 0.273 0.659 0.603 E. Analyst coverage Payout Repurchase Dividend Analyst coverage Analyst coverage Analyst coverage High Low High Low High Low Treat $$\times$$ Post –0.332 –0.333 –0.322 –0.282 –0.052 –0.053 (0.622) (0.647) (0.551) (0.562) (0.056) (0.089) Controls Yes Yes Yes Yes Yes Yes Year fixed effects Yes Yes Yes Yes Yes Yes Firm fixed effects Yes Yes Yes Yes Yes Yes Observations 2,025 2,301 2,025 2,301 2,025 2,301 Adjusted R-squared 0.337 0.316 0.294 0.316 0.661 0.571 D. Financial distress Payout Repurchase Dividend High Low High Low High Low (3) (4) (1) (2) (1) (2) Treat $$\times$$ Post –1.153$$^{**}$$ –0.665 –1.082$$^{**}$$ –0.686 –0.076$$^{**}$$ 0.020 (0.509) (0.538) (0.434) (0.420) (0.032) (0.096) Controls Yes Yes Yes Yes Yes Yes Year fixed effects Yes Yes Yes Yes Yes Yes Firm fixed effects Yes Yes Yes Yes Yes Yes Observations 1,993 1,984 1,993 1,984 1,993 1,984 Adjusted R-squared 0.356 0.282 0.320 0.273 0.659 0.603 E. Analyst coverage Payout Repurchase Dividend Analyst coverage Analyst coverage Analyst coverage High Low High Low High Low Treat $$\times$$ Post –0.332 –0.333 –0.322 –0.282 –0.052 –0.053 (0.622) (0.647) (0.551) (0.562) (0.056) (0.089) Controls Yes Yes Yes Yes Yes Yes Year fixed effects Yes Yes Yes Yes Yes Yes Firm fixed effects Yes Yes Yes Yes Yes Yes Observations 2,025 2,301 2,025 2,301 2,025 2,301 Adjusted R-squared 0.337 0.316 0.294 0.316 0.661 0.571 This table reports the difference-in-differences estimation results of $$Y_{\it it}=\alpha_{\it ij}+\alpha_{t}+\beta {\it Treat}_{\it ij} \times {\it Post}_{\it ijt} +\gamma X_{\it it-1} +\epsilon_{\it ijt}$$ on subsamples partitioned on measures of the stakes of the merging lenders (panel A), power of the merging shareholder (panel B), dividend restriction covenants (panel C), firm leverage (panel D), and analyst coverage (panel E). The dependent variables in Columns 1 and 2 are Payout, and in Columns 3 and 4 are Repurchase, and in Columns 5 and 6 are Dividend. Treat equals one if the firm is a treated firm of the merger, and zero otherwise. Post equals one if the firm-year observation is after the merger. All regressions include year fixed effects and merger-firm fixed effects. Standard errors are clustered by merger. Significance at 1%, 5%, and 10% levels are indicated by ***, **, and *, respectively. Table 7 Cross-sectional heterogeneity of the effect of mergers on payout policy A. Lender stake Payout Repurchase Dividend Low High Low High Low High (1) (2) (3) (4) (5) (6) Treat $$\times$$ Post –0.290 –1.095$$^{**}$$ –0.312 –0.914$$^{**}$$ 0.045 –0.142$$^{**}$$ (0.528) (0.487) (0.600) (0.446) (0.110) (0.069) Controls Yes Yes Yes Yes Yes Yes Year fixed effects Yes Yes Yes Yes Yes Yes Firm fixed effects Yes Yes Yes Yes Yes Yes Observations 2,056 1,962 2,056 1,962 2,056 1,962 Adjusted R-squared 0.265 0.351 0.242 0.331 0.553 0.634 B. Shareholder power Payout Repurchase Dividend High Low High Low High Low (1) (2) (3) (4) (5) (6) Treat $$\times$$ Post –1.047$$^{***}$$ –0.457 –0.962$$^{**}$$ –0.382 –0.079$$^{**}$$ 0.023 (0.359) (0.565) (0.422) (0.499) (0.038) (0.065) Controls Yes Yes Yes Yes Yes Yes Year fixed effects Yes Yes Yes Yes Yes Yes Firm fixed effects Yes Yes Yes Yes Yes Yes Observations 1,983 1,995 1,983 1,995 1,983 1,995 Adjusted R-squared 0.329 0.338 0.310 0.297 0.600 0.601 C. Dividend restriction covenants Payout Repurchase Dividend Dividend covenant Dividend covenant Dividend covenant Yes No Yes No Yes No (1) (2) (3) (4) (5) (6) Treat $$\times$$ Post –0.311 –0.933$$^{***}$$ –0.259 –0.799$$^{***}$$ –0.025 –0.077$$^{**}$$ (0.526) (0.335) (0.503) (0.323) (0.059) (0.032) Controls Yes Yes Yes Yes Yes Yes Year fixed effects Yes Yes Yes Yes Yes Yes Firm fixed effects Yes Yes Yes Yes Yes Yes Observations 1,168 4,801 1,168 4,801 1,168 4,801 Adjusted R-squared 0.292 0.317 0.242 0.299 0.599 0.616 A. Lender stake Payout Repurchase Dividend Low High Low High Low High (1) (2) (3) (4) (5) (6) Treat $$\times$$ Post –0.290 –1.095$$^{**}$$ –0.312 –0.914$$^{**}$$ 0.045 –0.142$$^{**}$$ (0.528) (0.487) (0.600) (0.446) (0.110) (0.069) Controls Yes Yes Yes Yes Yes Yes Year fixed effects Yes Yes Yes Yes Yes Yes Firm fixed effects Yes Yes Yes Yes Yes Yes Observations 2,056 1,962 2,056 1,962 2,056 1,962 Adjusted R-squared 0.265 0.351 0.242 0.331 0.553 0.634 B. Shareholder power Payout Repurchase Dividend High Low High Low High Low (1) (2) (3) (4) (5) (6) Treat $$\times$$ Post –1.047$$^{***}$$ –0.457 –0.962$$^{**}$$ –0.382 –0.079$$^{**}$$ 0.023 (0.359) (0.565) (0.422) (0.499) (0.038) (0.065) Controls Yes Yes Yes Yes Yes Yes Year fixed effects Yes Yes Yes Yes Yes Yes Firm fixed effects Yes Yes Yes Yes Yes Yes Observations 1,983 1,995 1,983 1,995 1,983 1,995 Adjusted R-squared 0.329 0.338 0.310 0.297 0.600 0.601 C. Dividend restriction covenants Payout Repurchase Dividend Dividend covenant Dividend covenant Dividend covenant Yes No Yes No Yes No (1) (2) (3) (4) (5) (6) Treat $$\times$$ Post –0.311 –0.933$$^{***}$$ –0.259 –0.799$$^{***}$$ –0.025 –0.077$$^{**}$$ (0.526) (0.335) (0.503) (0.323) (0.059) (0.032) Controls Yes Yes Yes Yes Yes Yes Year fixed effects Yes Yes Yes Yes Yes Yes Firm fixed effects Yes Yes Yes Yes Yes Yes Observations 1,168 4,801 1,168 4,801 1,168 4,801 Adjusted R-squared 0.292 0.317 0.242 0.299 0.599 0.616 D. Financial distress Payout Repurchase Dividend High Low High Low High Low (3) (4) (1) (2) (1) (2) Treat $$\times$$ Post –1.153$$^{**}$$ –0.665 –1.082$$^{**}$$ –0.686 –0.076$$^{**}$$ 0.020 (0.509) (0.538) (0.434) (0.420) (0.032) (0.096) Controls Yes Yes Yes Yes Yes Yes Year fixed effects Yes Yes Yes Yes Yes Yes Firm fixed effects Yes Yes Yes Yes Yes Yes Observations 1,993 1,984 1,993 1,984 1,993 1,984 Adjusted R-squared 0.356 0.282 0.320 0.273 0.659 0.603 E. Analyst coverage Payout Repurchase Dividend Analyst coverage Analyst coverage Analyst coverage High Low High Low High Low Treat $$\times$$ Post –0.332 –0.333 –0.322 –0.282 –0.052 –0.053 (0.622) (0.647) (0.551) (0.562) (0.056) (0.089) Controls Yes Yes Yes Yes Yes Yes Year fixed effects Yes Yes Yes Yes Yes Yes Firm fixed effects Yes Yes Yes Yes Yes Yes Observations 2,025 2,301 2,025 2,301 2,025 2,301 Adjusted R-squared 0.337 0.316 0.294 0.316 0.661 0.571 D. Financial distress Payout Repurchase Dividend High Low High Low High Low (3) (4) (1) (2) (1) (2) Treat $$\times$$ Post –1.153$$^{**}$$ –0.665 –1.082$$^{**}$$ –0.686 –0.076$$^{**}$$ 0.020 (0.509) (0.538) (0.434) (0.420) (0.032) (0.096) Controls Yes Yes Yes Yes Yes Yes Year fixed effects Yes Yes Yes Yes Yes Yes Firm fixed effects Yes Yes Yes Yes Yes Yes Observations 1,993 1,984 1,993 1,984 1,993 1,984 Adjusted R-squared 0.356 0.282 0.320 0.273 0.659 0.603 E. Analyst coverage Payout Repurchase Dividend Analyst coverage Analyst coverage Analyst coverage High Low High Low High Low Treat $$\times$$ Post –0.332 –0.333 –0.322 –0.282 –0.052 –0.053 (0.622) (0.647) (0.551) (0.562) (0.056) (0.089) Controls Yes Yes Yes Yes Yes Yes Year fixed effects Yes Yes Yes Yes Yes Yes Firm fixed effects Yes Yes Yes Yes Yes Yes Observations 2,025 2,301 2,025 2,301 2,025 2,301 Adjusted R-squared 0.337 0.316 0.294 0.316 0.661 0.571 This table reports the difference-in-differences estimation results of $$Y_{\it it}=\alpha_{\it ij}+\alpha_{t}+\beta {\it Treat}_{\it ij} \times {\it Post}_{\it ijt} +\gamma X_{\it it-1} +\epsilon_{\it ijt}$$ on subsamples partitioned on measures of the stakes of the merging lenders (panel A), power of the merging shareholder (panel B), dividend restriction covenants (panel C), firm leverage (panel D), and analyst coverage (panel E). The dependent variables in Columns 1 and 2 are Payout, and in Columns 3 and 4 are Repurchase, and in Columns 5 and 6 are Dividend. Treat equals one if the firm is a treated firm of the merger, and zero otherwise. Post equals one if the firm-year observation is after the merger. All regressions include year fixed effects and merger-firm fixed effects. Standard errors are clustered by merger. Significance at 1%, 5%, and 10% levels are indicated by ***, **, and *, respectively. 2.6 Conflict between the merging shareholder and other institutional shareholders While the merger between a shareholder and a lender aligns the interest of the two, the merging shareholder’s incentive to maximizing the combined value of shares and loans will necessarily conflict with those of other shareholders. Hence, the extent to which the merging shareholder can push the dividend policy to benefit creditors will also depend on the relative power between the merging shareholder and other shareholders. To measure the relative power, I first note that institutional shareholders are more powerful in affecting corporate decisions, and hence measure the relative power by the ratio between shares owned by the merging shareholder and other institutional shareholders, which I call the relative power ratio. If the relative power ratio is high, that is, the merging institutional shareholders are more powerful, I expect the effect of the mergers on payout to be more pronounced. I again sort the observations into terciles according to the relative power ratio and redo the analysis on the top and the bottom terciles separately. The results are presented in panel B of Table 7. Consistent with the argument that the relative power of the merging shareholder enhances the effect of the mergers on dividend policy, the difference-in-differences estimates are negative and statistically significant for observations in the top terciles (Columns 1, 3, and 5), that is, for observations in which the merging shareholder is relatively more powerful. In contrast, the estimates in Columns 2, 4, and 6, in which the merging shareholder is less powerful, are much smaller in magnitudes and are statistically insignificant. The differences are again statistically significant. The results confirm that the alignment of interest between the merging lender and the merging shareholder also creates conflict of interest between the merging shareholder and other shareholders, further suggesting that the baseline result is driven by the alignment of interest between the merging shareholder and the merging lender. 2.7 Dividend restriction covenants Syndicated loans often contain covenants restricting dividend payment to mitigate the agency cost associated with excessive dividend payment. If the loan contract already contains the dividend restriction covenants, the effect of the mergers in reducing excessive dividend payment may be limited. If the baseline results are driven by reduced shareholder-creditor conflict, the effect should concentrate in loans without such covenants. I partition the sample according to whether the loan has the dividend restriction covenants and then reestimate the baseline difference-in-differences specification on these two subsamples. The results are presented in panel C of Table 7. The difference-in-differences estimates are both negative and statistically significant in the subsample without the dividend restriction covenants (in Columns 2, 4, and 6). In sharp contrast, the estimates are much smaller and are statistically insignificant in the subsample with dividend restriction covenants. The differences of the estimates between these two subsamples are also statistically significant. Overall, these results suggest that dividend restriction covenants do in fact mitigate the agency cost of excessive dividend payment due to the shareholder-creditor conflict, and hence the reduction in shareholder-creditor conflict due to the mergers has limited effects on dividend payment for firms with these covenants. 2.8 Financial distress and the effects of the mergers The conflict of interest between shareholders and creditors often becomes exaggerated when the firm is in financial distress (Smith and Warner 1979; Gilson, John, and Lang 1990; Gilson and Vetsuypens 1993; Ayotte, Hotchkiss, and Thorburn 2013). It follows that the alignment of interest between shareholders and creditors via the merger should have a stronger effect in resolving the conflict. I therefore test this conjecture to provide further support to the argument that the results are driven by reduced conflict of interests between shareholders and creditors. To test this conjecture, I first sort the firms into terciles based on their leverage measured immediately before the merger, and then re-estimate the difference-in-differences specification on the top and bottom terciles separately. The results are presented in panel D of Table 7. Consistent with the conjecture, the results show that the effect concentrates in the top tercile of firms sorted on leverage, i.e., more financially distressed firms. In contrast, the effect is small and statistically insignificant in nondistressed firms. Overall, the results suggest that while shareholders have stronger incentives to pay excessive dividends at the expense of creditors when the firm is in financial distress, the alignment of the interest of shareholders and creditors also has a stronger effect in mitigating the shareholder-creditor conflict when the firm is in financial distress. The results therefore provide further support to the argument that the mergers affect payout policy via their effects on the shareholder-creditor conflict. 2.9 Signaling versus shareholder-creditor conflict While the baseline results are consistent with the theory of the shareholder-creditor conflict, the baseline results can also be driven by reduced information asymmetry between shareholders and the firm (managers). Lenders often have access to private information, and merging with a lender allows the shareholder to gain access to private information. The signaling theory of dividend policy (Miller and Rock 1985; John and Williams 1985) suggests that firms pay dividends to signal the quality of the firm in the presence of information asymmetry. The merger, which reduces the information asymmetry between shareholders and the firms, therefore can reduce the need of the firm to signal with dividend payout. To mitigate this concern, I first note that according to the signaling theory of John and Williams (1985), cash dividends are more effective than stock repurchases in signaling. Empirically, Grullon and Michaely (2004) also find that share repurchases do not signal future performance. It follows that if the baseline results are indeed driven by signaling, the mergers should, at least, have a larger effect on cash dividends than on share repurchases. However, the results presented in Table 6 show that the magnitudes of the coefficients on share repurchases are much larger than those for cash dividends. To the extent that share repurchases are not an effective tool for signaling, this result suggests that the baseline results are unlikely to be driven by signaling. The signaling hypothesis suggests that the mergers decrease payout via their effects on information asymmetry. It follows that the effect should be more pronounced for firms more subject to the asymmetric information problem ex ante. To test whether that is the case, I follow Li and Zhao (2008) to use analyst coverage as a proxy for information asymmetry and partition the sample according to the number of analyst coverage into terciles.7 The results on the bottom and top terciles are presented in panel E of Table 7. Inconsistent with the signaling hypothesis, the results show that the difference-in-differences coefficients are very similar between the low coverage and high coverage firms, further suggesting that the baseline results are unlikely to be driven by signaling. 3. Robustness Checks 3.1 How treated firms change their payout? The baseline results show that the treated firms decrease cash dividends and share repurchases relative to control firms. Treated firms can either cut their dividends or become less likely to increase their dividends. On particular concern is that cutting dividends is often extremely costly to the firms (Denis, Denis, and Sarin 1994; Grullon, Michaely, and Swaminathan 2002), which may undermine the firms’ incentives to reduce dividends even when the shareholder-creditor conflict is reduced. In this subsection, I examine whether treated firms indeed cut dividends. In this regard, I use dividend data from CRSP and code a dividend increase as an increase of cash dividends by more than 10%, and code a dividend decrease as a decrease of cash dividends by more than 10% (Denis, Denis, and Sarin 1994). I code a fiscal year with dividend increase (decrease) if a dividend increase (decrease) occurred during the fiscal year. I then replace the dependent variable in Equation (1) with indicator variables of dividend increase and decrease. The results are presented in panel A of Table 8. For dividend increase, the difference-in-differences coefficients are negative and statistically significant, suggesting that treated firms, relative to control firms, become less likely to increase their dividends after the mergers. In contrast, the coefficients for dividend decrease is much smaller and statistically insignificant, suggesting that treated firms do not become more likely to cut their dividends after the mergers. Table 8 The effect of mergers on dividend changes and share repurchases A. Dividend changes Dividend increase Dividend decrease (1) (2) (3) (4) Treat $$\times$$ Post –0.015$$^{**}$$ –0.016$$^{**}$$ –0.005 –0.004 (0.007) (0.007) (0.031) (0.032) Controls Yes Yes Year fixed effects Yes Yes Yes Yes Firm fixed effects Yes Yes Yes Yes Observations 5,969 5,969 5,969 5,969 Adjusted R-squared 0.275 0.275 0.074 0.075 B. Share repurchases Announcement Actual (1) (2) (3) (4) Treat $$\times$$ Post –0.037$$^{***}$$ –0.041$$^{***}$$ –0.015$$^{***}$$ –0.016$$^{***}$$ (0.012) (0.012) (0.004) (0.005) Controls Yes Yes Year fixed effects Yes Yes Yes Yes Firm fixed effects Yes Yes Yes Yes Observations 5,969 5,969 5,969 5,969 Adjusted R-squared 0.164 0.174 0.024 0.027 A. Dividend changes Dividend increase Dividend decrease (1) (2) (3) (4) Treat $$\times$$ Post –0.015$$^{**}$$ –0.016$$^{**}$$ –0.005 –0.004 (0.007) (0.007) (0.031) (0.032) Controls Yes Yes Year fixed effects Yes Yes Yes Yes Firm fixed effects Yes Yes Yes Yes Observations 5,969 5,969 5,969 5,969 Adjusted R-squared 0.275 0.275 0.074 0.075 B. Share repurchases Announcement Actual (1) (2) (3) (4) Treat $$\times$$ Post –0.037$$^{***}$$ –0.041$$^{***}$$ –0.015$$^{***}$$ –0.016$$^{***}$$ (0.012) (0.012) (0.004) (0.005) Controls Yes Yes Year fixed effects Yes Yes Yes Yes Firm fixed effects Yes Yes Yes Yes Observations 5,969 5,969 5,969 5,969 Adjusted R-squared 0.164 0.174 0.024 0.027 This table reports the difference-in-differences estimation results of $$Y_{\it it}=\alpha_{ij}+\alpha_{t}+\beta {\it Treat}_{\it ij} \times {\it Post}_{\it ijt} +\gamma X_{\it it-1} +\epsilon_{\it ijt}$$, with panel A for dividend changes and panel B for share repurchases. In panel A, the dependent variable in Columns 1 and 2 is Dividend increase, which equals one if the firm increases cash dividends during the fiscal year, and zero otherwise, and the dependent variable in Columns 3 and 4 is Dividend decrease, which equals one if the firm decreases cash dividends during the fiscal year, and zero otherwise. In panel B, the dependent variable in Columns 1 and 2 is Announce, which equals one if the firm announces share repurchases during the fiscal year, and zero otherwise, and the dependent variable in Columns 3 and 4 is Actual, which equals one if the firm actually conducts share repurchases during the fiscal year, and zero otherwise. Treat equals one if the firm is a treated firm of the merger, and zero otherwise. Post equals one if the firm-year observation is after the merger. All regressions include year fixed effects and firm fixed effects. Standard errors are clustered by merger. Significance at 1%, 5%, and 10% levels are indicated by ***, **, and *, respectively. Table 8 The effect of mergers on dividend changes and share repurchases A. Dividend changes Dividend increase Dividend decrease (1) (2) (3) (4) Treat $$\times$$ Post –0.015$$^{**}$$ –0.016$$^{**}$$ –0.005 –0.004 (0.007) (0.007) (0.031) (0.032) Controls Yes Yes Year fixed effects Yes Yes Yes Yes Firm fixed effects Yes Yes Yes Yes Observations 5,969 5,969 5,969 5,969 Adjusted R-squared 0.275 0.275 0.074 0.075 B. Share repurchases Announcement Actual (1) (2) (3) (4) Treat $$\times$$ Post –0.037$$^{***}$$ –0.041$$^{***}$$ –0.015$$^{***}$$ –0.016$$^{***}$$ (0.012) (0.012) (0.004) (0.005) Controls Yes Yes Year fixed effects Yes Yes Yes Yes Firm fixed effects Yes Yes Yes Yes Observations 5,969 5,969 5,969 5,969 Adjusted R-squared 0.164 0.174 0.024 0.027 A. Dividend changes Dividend increase Dividend decrease (1) (2) (3) (4) Treat $$\times$$ Post –0.015$$^{**}$$ –0.016$$^{**}$$ –0.005 –0.004 (0.007) (0.007) (0.031) (0.032) Controls Yes Yes Year fixed effects Yes Yes Yes Yes Firm fixed effects Yes Yes Yes Yes Observations 5,969 5,969 5,969 5,969 Adjusted R-squared 0.275 0.275 0.074 0.075 B. Share repurchases Announcement Actual (1) (2) (3) (4) Treat $$\times$$ Post –0.037$$^{***}$$ –0.041$$^{***}$$ –0.015$$^{***}$$ –0.016$$^{***}$$ (0.012) (0.012) (0.004) (0.005) Controls Yes Yes Year fixed effects Yes Yes Yes Yes Firm fixed effects Yes Yes Yes Yes Observations 5,969 5,969 5,969 5,969 Adjusted R-squared 0.164 0.174 0.024 0.027 This table reports the difference-in-differences estimation results of $$Y_{\it it}=\alpha_{ij}+\alpha_{t}+\beta {\it Treat}_{\it ij} \times {\it Post}_{\it ijt} +\gamma X_{\it it-1} +\epsilon_{\it ijt}$$, with panel A for dividend changes and panel B for share repurchases. In panel A, the dependent variable in Columns 1 and 2 is Dividend increase, which equals one if the firm increases cash dividends during the fiscal year, and zero otherwise, and the dependent variable in Columns 3 and 4 is Dividend decrease, which equals one if the firm decreases cash dividends during the fiscal year, and zero otherwise. In panel B, the dependent variable in Columns 1 and 2 is Announce, which equals one if the firm announces share repurchases during the fiscal year, and zero otherwise, and the dependent variable in Columns 3 and 4 is Actual, which equals one if the firm actually conducts share repurchases during the fiscal year, and zero otherwise. Treat equals one if the firm is a treated firm of the merger, and zero otherwise. Post equals one if the firm-year observation is after the merger. All regressions include year fixed effects and firm fixed effects. Standard errors are clustered by merger. Significance at 1%, 5%, and 10% levels are indicated by ***, **, and *, respectively. Similarly, I also examine how the mergers affect treated firms’ propensity to announce share repurchases and to actually conduct share repurchases. To this end, I follow Grullon and Michaely (2004) to collect share repurchases information from the SDC. I code Announce to be one if the firm announces share repurchases during the fiscal year, and zero otherwise; I code Actual to be one if the firm announces share repurchases during the fiscal year and the transactions are completed (which may or may not be completed during the fiscal year). I then replace the dependent variable in Equation (1) with Announce and Actual. The results are presented in panel B of Table 8. The difference-in-differences coefficients are all negative and statistically significant, suggesting that, relative to control firms, treated firms are less likely to announce and complete share repurchases. 3.2 Changing positions by the lenders and the shareholders The DealScan data only report loan allocations at origination, and the lenders can sell their loans in the secondary market. It is therefore possible that the lenders may have already sold the loans at the time of the merger, and consequently the merger may not affect the shareholder-creditor conflict at all. Although such noise may only bias against any finding, I still address this problem in this subsection to ensure the robustness of the results and to ensure that the results are not driven by something else. I first focus on a subsample in which the merging lender is a lead bank of the loan. In syndicated loans, lead banks screen and monitor the borrowers and performing those tasks requires them to have a stake in the firm, that is, lead banks often do not (completely) sell the loans they lead. The estimation results on this subsample are presented in panel A of Table 9. The difference-in-differences estimates all remain negative and statistically significant. Table 9 The effect of changing positions of the lenders and the shareholders A. Lead lenders only Payout Repurchase Dividend (1) (2) (3) (4) (5) (6) Treat $$\times$$ Post –0.962$$^{**}$$ –0.978$$^{**}$$ –0.768$$^{*}$$ –0.773$$^{*}$$ –0.095$$^{***}$$ –0.107$$^{***}$$ (0.466) (0.445) (0.419) (0.395) (0.033) (0.034) Controls Yes Yes Yes Year fixed effects Yes Yes Yes Yes Yes Yes Firm fixed effects Yes Yes Yes Yes Yes Yes Observations 2,468 2,468 2,468 2,468 2,468 2,468 Adjusted R-squared 0.311 0.324 0.290 0.306 0.591 0.594 B. Mergers within 1 year of loan origination Treat $$\times$$ Post –0.943$$^{*}$$ –0.980$$^{**}$$ –0.879$$^{**}$$ –0.914$$^{**}$$ –0.105$$^{*}$$ –0.104$$^{**}$$ (0.490) (0.451) (0.423) (0.401) (0.054) (0.047) Controls Yes Yes Yes Year fixed effects Yes Yes Yes Yes Yes Yes Firm fixed effects Yes Yes Yes Yes Yes Yes Observations 2,090 2,090 2,090 2,090 2,090 2,090 Adjusted R-squared 0.243 0.284 0.212 0.249 0.554 0.558 A. Lead lenders only Payout Repurchase Dividend (1) (2) (3) (4) (5) (6) Treat $$\times$$ Post –0.962$$^{**}$$ –0.978$$^{**}$$ –0.768$$^{*}$$ –0.773$$^{*}$$ –0.095$$^{***}$$ –0.107$$^{***}$$ (0.466) (0.445) (0.419) (0.395) (0.033) (0.034) Controls Yes Yes Yes Year fixed effects Yes Yes Yes Yes Yes Yes Firm fixed effects Yes Yes Yes Yes Yes Yes Observations 2,468 2,468 2,468 2,468 2,468 2,468 Adjusted R-squared 0.311 0.324 0.290 0.306 0.591 0.594 B. Mergers within 1 year of loan origination Treat $$\times$$ Post –0.943$$^{*}$$ –0.980$$^{**}$$ –0.879$$^{**}$$ –0.914$$^{**}$$ –0.105$$^{*}$$ –0.104$$^{**}$$ (0.490) (0.451) (0.423) (0.401) (0.054) (0.047) Controls Yes Yes Yes Year fixed effects Yes Yes Yes Yes Yes Yes Firm fixed effects Yes Yes Yes Yes Yes Yes Observations 2,090 2,090 2,090 2,090 2,090 2,090 Adjusted R-squared 0.243 0.284 0.212 0.249 0.554 0.558 This table reports the difference-in-differences estimation results to address the possibility that lenders or institutional shareholders may change their positions after loan origination or after the merger. In panel A, the estimation is performed on a subsample in which the merging lender is a lead lender of the loan; in panel B, the estimation is performed on mergers that occur within 1 year of loan origination and over a 2-year window around the merger. The dependent variables in Columns 1 and 2 are Payout, in Columns 3 and 4 are Repurchase, and in Columns 5 and 6 are Dividend. Treat equals one if the firm is a treated firm of the merger, and zero otherwise. Post equals one if the firm-year observation is after the merger. All regressions include year fixed effects and merger-firm fixed effects. Standard errors are clustered by merger. Significance at 1%, 5%, and 10% levels are indicated by ***, **, and *, respectively. Table 9 The effect of changing positions of the lenders and the shareholders A. Lead lenders only Payout Repurchase Dividend (1) (2) (3) (4) (5) (6) Treat $$\times$$ Post –0.962$$^{**}$$ –0.978$$^{**}$$ –0.768$$^{*}$$ –0.773$$^{*}$$ –0.095$$^{***}$$ –0.107$$^{***}$$ (0.466) (0.445) (0.419) (0.395) (0.033) (0.034) Controls Yes Yes Yes Year fixed effects Yes Yes Yes Yes Yes Yes Firm fixed effects Yes Yes Yes Yes Yes Yes Observations 2,468 2,468 2,468 2,468 2,468 2,468 Adjusted R-squared 0.311 0.324 0.290 0.306 0.591 0.594 B. Mergers within 1 year of loan origination Treat $$\times$$ Post –0.943$$^{*}$$ –0.980$$^{**}$$ –0.879$$^{**}$$ –0.914$$^{**}$$ –0.105$$^{*}$$ –0.104$$^{**}$$ (0.490) (0.451) (0.423) (0.401) (0.054) (0.047) Controls Yes Yes Yes Year fixed effects Yes Yes Yes Yes Yes Yes Firm fixed effects Yes Yes Yes Yes Yes Yes Observations 2,090 2,090 2,090 2,090 2,090 2,090 Adjusted R-squared 0.243 0.284 0.212 0.249 0.554 0.558 A. Lead lenders only Payout Repurchase Dividend (1) (2) (3) (4) (5) (6) Treat $$\times$$ Post –0.962$$^{**}$$ –0.978$$^{**}$$ –0.768$$^{*}$$ –0.773$$^{*}$$ –0.095$$^{***}$$ –0.107$$^{***}$$ (0.466) (0.445) (0.419) (0.395) (0.033) (0.034) Controls Yes Yes Yes Year fixed effects Yes Yes Yes Yes Yes Yes Firm fixed effects Yes Yes Yes Yes Yes Yes Observations 2,468 2,468 2,468 2,468 2,468 2,468 Adjusted R-squared 0.311 0.324 0.290 0.306 0.591 0.594 B. Mergers within 1 year of loan origination Treat $$\times$$ Post –0.943$$^{*}$$ –0.980$$^{**}$$ –0.879$$^{**}$$ –0.914$$^{**}$$ –0.105$$^{*}$$ –0.104$$^{**}$$ (0.490) (0.451) (0.423) (0.401) (0.054) (0.047) Controls Yes Yes Yes Year fixed effects Yes Yes Yes Yes Yes Yes Firm fixed effects Yes Yes Yes Yes Yes Yes Observations 2,090 2,090 2,090 2,090 2,090 2,090 Adjusted R-squared 0.243 0.284 0.212 0.249 0.554 0.558 This table reports the difference-in-differences estimation results to address the possibility that lenders or institutional shareholders may change their positions after loan origination or after the merger. In panel A, the estimation is performed on a subsample in which the merging lender is a lead lender of the loan; in panel B, the estimation is performed on mergers that occur within 1 year of loan origination and over a 2-year window around the merger. The dependent variables in Columns 1 and 2 are Payout, in Columns 3 and 4 are Repurchase, and in Columns 5 and 6 are Dividend. Treat equals one if the firm is a treated firm of the merger, and zero otherwise. Post equals one if the firm-year observation is after the merger. All regressions include year fixed effects and merger-firm fixed effects. Standard errors are clustered by merger. Significance at 1%, 5%, and 10% levels are indicated by ***, **, and *, respectively. I then focus on another subsample in which the mergers occur within one year of loan origination. The short time between loan origination and the mergers makes it less likely that the lenders sell the loan in the secondary market before the merger. Furthermore, to ensure that the lenders are also more likely to hold the loans in the post period, I also restrict the analysis to a 2-year window, that is, 1 year before and 1 year after the merger. Focusing on the shorter window also makes it unlikely that the institutional investors sell all the shares in the post period. The estimation results are presented in panel B of Table 9. The difference-in-differences estimates are again all negative and statistically significant. Overall, the results in Table 9 suggest that changing positions of lenders or shareholders are not a major concern. 3.3 Other Robustness Tests To further ensure the robustness of the results, I also conduct the following tests.8 First, I re-scale the payout measures by total assets, and find that the results remain robust. Second, the results can be driven by unobservable characteristics of the merging parties. For example, if the merging lender’s ability to press the shareholders not to pay excessive dividends increases over time, payout of all borrowers of the merging lender will decrease over time. On the other hand, if the merging institutional investor becomes increasingly passive and does not press the firm to pay excessive dividends, all firms whose stocks held by the merging investor can also experience a decline in payout. In these cases, the baseline results can simply be driven by unobservable characteristics of the merging lenders or the merging institutional shareholders but not by the alignment of interests between the lenders and the shareholders. The results can also be driven by behavioral changes of the lenders or shareholders after the merger. To mitigate these concerns, I try two alternative matching methods. In the first method, I require the controls firms to also be held by the merging shareholder at the time of the mergers. In the second method, I require the controls firms also have loans outstanding borrowed from the merging lender at the time of the merger. In both cases, I find similar results as those in Table 6, suggesting that the results are not driven by unobservable lender or shareholder characteristics. Third, to show that the mergers do indeed lead to decreases in shareholder-creditor conflict, I show that treated firms, relative to control firms, issue less debt, after the mergers, suggesting that the mergers reduce claim dilution. I also find that treated firms with lower q reduces capital investment, suggesting that the mergers reduce overinvestment. 4. Conclusion This paper examines the effect of the shareholder-creditor conflict on payout policy using a novel identification strategy. I use mergers between lenders and institutional shareholders of the same firm as natural experiments that generate plausibly exogenous variation in the conflict of interest between shareholders and creditors. I find that following the mergers, treated firms, relative to control firms, reduce their payout, suggesting that shareholders pay excessive dividends to themselves at the expense of creditors when the interests of shareholders and creditors are not aligned. Consistent with the argument that the shareholder-creditor conflict often becomes exaggerated when the firm is in financial distress, I find that the effect is stronger for financially distressed firms. The author thanks Allen Berger, Alan Crane (discussant at the 2017 AFA meeting), Wei Jiang (the Editor), Greg Niehaus, Eric Powers, Sergey Tsyplakov, Donghang Zhang; two anonymous referees; and participants at the 2017 AFA meetings for comments and suggestions. The author also thanks Gerard Pinto for excellent research assistance. Supplementary data can be found on The Review of Financial Studies Web site. Footnotes 1 An exception is Gilje (2016), who measures the shareholder-creditor conflict with changes in financial distress driven by exogenous changes in commodity prices. 2 The evidence in Kalay (1982) that the dividend constraints are often not binding and that many loan contracts do not have direct dividend constraints supports the theoretical prediction of Easterbrook (1984) and John and Kalay (1982). 3 The same threshold used by Jiang, Li, and Shao (2010) to identify dual holders. 4 A similar argument is made by Azar, Schmalz, and Tecu (Forthcoming) and Chu (2017). 5 The results are robust when using unwinsorized variables. 6 The coefficients on repurchase and cash dividends do not add to the coefficients on the total payout due to winsorization. 7 The results are similar if I instead use analyst forecast error or forecast dispersion to measure asymmetric information. 8 To save space, I briefly discuss the results, and the tables can be found in the Online Appendix. References Allen, F., and Michaely. R. 2003 . Payout policy. Handbook of the Economics of Finance 1 : 337 – 429 . Google Scholar CrossRef Search ADS Asquith, P., and Wizman. T. A. 1990 . Event risk, covenants, and bondholder returns in leveraged buyouts. Journal of Financial Economics 27 : 195 – 213 . Google Scholar CrossRef Search ADS Ayotte, K., Hotchkiss, E. S. and Thorburn. K. S. 2013 . Governance in financial distress and bankruptcy. In Oxford handbook of corporate governance , 159 – 288 . Oxford University Press . Google Scholar CrossRef Search ADS Azar, J., Schmalz, M. C. and Tecu. I. Forthcoming . Anti-competitive effects of common ownership. Journal of Finance . Billett, M. T., King, T.-H. D. and Mauer. D. C. 2004 . Bondholder wealth effects in mergers and acquisitions: New evidence from the 1980s and 1990s. Journal of Finance 59 : 107 – 135 . Google Scholar CrossRef Search ADS Black, F. 1976 . The dividend puzzle. Journal of Portfolio Management 2 : 5 – 8 . Google Scholar CrossRef Search ADS Bodnaruk, A., and Rossi. M. 2016 . Dual ownership, returns, and voting in mergers. Journal of Financial Economics 120 : 58 – 80 . Google Scholar CrossRef Search ADS Brockman, P., and Unlu. E. 2009 . Dividend policy, creditor rights, and the agency costs of debt. Journal of Financial Economics 92 : 276 – 99 . Google Scholar CrossRef Search ADS Chava, S., Wang, R. and Zou. H. Forthcoming . Covenants, creditors’ simultaneous equity holdings, and firm investment policies. Journal of Financial and Quantitative Analysis . Chu, Y. 2017 . Debt renegotiation and debt overhang: Evidence from lender mergers. Working Paper , University of South Carolina . Google Scholar CrossRef Search ADS Denis, D. J., Denis, D. K. and Sarin. A. 1994 . The information content of dividend changes: Cash flow signaling, overinvestment, and dividend clienteles. Journal of Financial and Quantitative Analysis 29 : 567 – 87 . Google Scholar CrossRef Search ADS Dhillon, U. S., and Johnson. H. 1994 . The effect of dividend changes on stock and bond prices. Journal of Finance 49 : 281 – 9 . Google Scholar CrossRef Search ADS Easterbrook, F. H. 1984 . Two agency-cost explanations of dividends. American Economic Review 74 : 650 – 9 . Farre-Mensa, J., Michaely, R. and Schmalz. M. 2014 . Payout policy. Annual Review of Financial Economics 6 : 75 – 134 . Google Scholar CrossRef Search ADS Gilje, E. P. 2016 . Do firms engage in risk-shifting? Empirical evidence. Review of Financial Studies , 29 : 2925 – 54 . Google Scholar CrossRef Search ADS Gilson, S. C., John, K. and Lang. L. H. 1990 . Troubled debt restructurings: An empirical study of private reorganization of firms in default. Journal of Financial Economics 27 : 315 – 53 . Google Scholar CrossRef Search ADS Gilson, S. C., and Vetsuypens. M. R. 1993 . CEO compensation in financially distressed firms: An empirical analysis. Journal of Finance 48 : 425 – 58 . Google Scholar CrossRef Search ADS Grullon, G., and Michaely. R. 2004 . The information content of share repurchase programs. Journal of Finance 59 : 651 – 80 . Google Scholar CrossRef Search ADS Grullon, G., Michaely, R. and Swaminathan. B. 2002 . Are dividend changes a sign of firm maturity? Journal of Business 75 : 387 – 424 . Google Scholar CrossRef Search ADS Handjinicolaou, G., and Kalay. A. 1984 . Wealth redistributions or changes in firm value: An analysis of returns to bondholders and stockholders around dividend announcements. Journal of Financial Economics 13 : 35 – 63 . Google Scholar CrossRef Search ADS Hong, H., and Kacperczyk. M. 2010 . Competition and bias. Quarterly Journal of Economics 125 : 1683 – 725 . Google Scholar CrossRef Search ADS Jayaraman, N., and Shastri. K. 1988 . The valuation impacts of specially designated dividends. Journal of Financial and Quantitative Analysis 23 : 301 – 12 . Google Scholar CrossRef Search ADS Jensen, M. C., and Meckling. W. H. 1976 . Theory of the firm: Managerial behavior, agency costs and ownership structure. Journal of Financial Economics 3 : 305 – 60 . Google Scholar CrossRef Search ADS Jiang, W., Li, K. and Shao. P. 2010 . When shareholders are creditors: Effects of the simultaneous holding of equity and debt by non-commercial banking institutions. Review of Financial Studies 23 : 3595 – 637 . Google Scholar CrossRef Search ADS John, K., and Kalay. A. 1982 . Costly contracting and optimal payout constraints. Journal of Finance 37 : 457 – 70 . Google Scholar CrossRef Search ADS John, K., and Williams. J. 1985 . Dividends, dilution, and taxes: A signaling equilibrium. Journal of Finance 40 : 1053 – 70 . Google Scholar CrossRef Search ADS Kalay, A. 1982 . Stockholder-bondholder conflict and dividend constraints. Journal of Financial Economics 10 : 211 – 33 . Google Scholar CrossRef Search ADS Li, K., and Zhao. X. 2008 . Asymmetric information and dividend policy. Financial Management 37 : 673 – 94 . Google Scholar CrossRef Search ADS Miller, M. H., and Rock. K. 1985 . Dividend policy under asymmetric information. Journal of Finance 40 : 1031 – 51 . Google Scholar CrossRef Search ADS Myers, S. C. 1977 . Determinants of corporate borrowing. Journal of Financial Economics 5 : 147 – 75 . Google Scholar CrossRef Search ADS Roberts, M., and Whited. T. 2012 . Endogeneity in empirical corporate finance. Handbook of the economics of finance 2 . Smith, C. W., and Warner. J. B. 1979 . On financial contracting: An analysis of bond covenants. Journal of Financial Economics 7 : 117 – 61 . Google Scholar CrossRef Search ADS Warga, A., and Welch. I. 1993 . Bondholder losses in leveraged buyouts. Review of Financial Studies 6 : 959 – 82 . Google Scholar CrossRef Search ADS © The Author(s) 2017. Published by Oxford University Press on behalf of The Society for Financial Studies. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com. This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices)

Journal

The Review of Financial StudiesOxford University Press

Published: Dec 11, 2017

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off