Payout policy and product market competition

Table 4 shows estimates of regressions relating scaled payout policy to the product market competition and other control variables. In Panel A, we report the result that assumes a linear relationship between the HHI and the payout policy. In Panel B, we report the linear

2 Given that HHI is low, it means that firms have a competitive market. When payout policy and HHI have a positive relationship, the payout policy and product market competition have a negative relationship.

relationship between CONC and payout policy. In Table 3, we know that payout policy has a positive relationship with HHI. Next, we examined the correlation between payout policy and product market competition in Table 4. The result shows that a significantly positive relationship exists between PAYOUT 1 and HHI in Column (1) of Panel A. Similarly, there is a significantly positive relationship between PAYOUT 2 and HHI in Column (2) of Panel A.

To check the robustness of the result, we used CONC as a proxy for product market competition in Panel B. The result shows that a significantly positive relationship exists between PAYOUT1 and CONC in Column (1) of Panel B. Similarly, there is a significantly positive relationship between PAYOUT2 and CONC in Column (2) of Panel B. Thus, from the results of Table 3 and Table 4, we conclude that payout policy has a negative relationship with product market competition and all the results are significant.

In Table 4, we also examine the possibility of the free cash flow and the maturity hypothesis. There exists a significantly positive relationship between payout policy and FCF in Columns (1)–(2) of Panel A and in Columns (1)–(2) of Panel B. The results strongly prove that the free cash flow hypothesis is true. However, there is a little discrepancy between the results of Table 4 and the maturity hypothesis. In Column (1) of Panel A, payout policy and growth opportunity have a significantly negative relationship, thus supporting the maturity hypothesis. In Column (2) of Panel A, the results are ambiguous because there is a significantly positive relationship. In Panel B, the results are also not strong and significant to support the maturity hypothesis. In conclusion, we find strong support for the free cash flow hypothesis, but we can not find strong support for the maturity hypothesis.

Table 4 Linear Regression－Payout Policy and Product Market Competition

The table reports linear regressions results. PAYOUT 1 is the ratio of dividends to net income. PAYOUT 2 is the ratio of dividends plus share repurchases to net income. HHI is the three-digit SIC Herfindahl-Hirschman Index from the census of manufacturers. CONC is the proportion of sales in the industry accounted for by the four largest firms (by sales) in the industry. Tobin’s Q is the book value of asset plus market value of equity minus market value of equity, scaled by the book value of assets. FCF is the operating income before depreciation minus interest expenses, income taxes, and capital expenditures, scaled by the book value of total assets.

Leverage is the long-term debt plus short-term debt, scaled by the book value of assets. Size is the natural log of

Panel A:Regression on payout policy with HHI

*, **, and *** indicate significance at 10%, 5% and 1%, respectively.

To be robust, we set two dummy variables to represent a firm’s product market competition in a given year. One was the DUMMYHHI, and the other was the DUMMYCONC. In Panel

A of Table 5, the relationship between payout policy and product market competition is significantly negative in Columns (1)–(2). In Panel B of Table 5, the relationship between payout policy and product market competition is also significantly negative in Columns (1)–(2). The results are consistent with our prior results.

Next, we examined the relationship among payout policy, FCF, and growth opportunity.

In Panel A of Table 5, the relationship between payout policy and FCF is significantly positive in Columns (1)–(2). In Panel B of Table 5, the relationship between payout policy and FCF is also significantly positive in Columns (1)–(2). However, the relationship between payout policy and growth opportunity is still ambiguous in Columns (1)–(2) of Panel A and Panel B. In summary, we get the same results in Table 4.

Table 5 Linear Regression with Dummy Variable－Payout Policy and Product Market Competition

The table reports the results of linear regressions. PAYOUT 1 is the ratio of dividends to net income. PAYOUT 2 is the ratio of dividends plus share repurchases to net income. DUMHHI is 1. If a firm exceeds the within-year median HHI in each year, then it will be zero. DUMCONC is 1. If a firm exceeds the within-year median CONC in each year, then it will be zero. Tobin’s Q is the book value of assets plus market value of equity minus the market value of equity, scaled by the book value of assets. FCF is the operating income before depreciation minus interest expenses, income taxes, and capital expenditures, scaled by the book value of total assets.

Leverage is the long-term debt plus the short-term debt, scaled by the book value of assets. Size is the natural log of the book value of assets. t-statistics are shown in parentheses.

Panel A:Regression on payout policy with DUMHHI

N 4459 4459

Panel B:Regression on payout policy with DUMCONC PAYOUT1 PAYOUT2

*, **, and *** indicate significance at 10%, 5%, and 1%, respectively.

We also used panel regression to explore the relationship between payout policy and product market competition. Before the examination, the regressions in Column (1) of Panel A and Panel B have no explanatory power³. In Panel A of Table 6, payout policy has a significantly negative relationship with product market competition in Columns (1)–(2). In Panel B of Table 6, payout policy also has a significantly negative relationship with product market competition in Columns (1)–(2).

In Panel A of Table 6, payout policy has a significantly positive relationship with FCF in Columns (1)–(2). In Panel B of Table 6, payout policy also has a significantly positive relationship with product market competition in Columns (1)–(2). Nevertheless, the relationship between payout policy and growth opportunity is ambiguous. From Table 4, Table 5, and Table 6, we have the same result, which states that payout policy and product

3 In Panel A and Panel B of Table 6, the regressions do not explain power in Columns (1)-(2) because the results are “NO” in the row of year fixed effects.

market competition have a strong negative relationship, thus supporting the free cash flow hypothesis strongly, but not the maturity hypothesis. With this finding, we consider that some variables affect the relationship between payout policy and growth opportunity. In Table 3, PAYOUT 1 and Tobin’s Q have a significantly negative relationship. In Table 4, Table 5, and Table 6, the relationship between PAYOUT 1 and Tobin’s Q is negative but not significant compared with Table 3. From the table, we suppose that some variables would weaken growth opportunity’s negative effect on payout policy.

Table 6 Panel Regression－Payout Policy and Product Market Competition

The table reports panel regressions results. PAYOUT 1 is the ratio of dividends to net income. PAYOUT 2 is the ratio of dividends plus share repurchases to net income. HHI is the three-digit SIC Herfindahl-Hirschman Index from the Census of Manufacturers. CONC is the proportion of sales in the industry accounted for by the four largest firms (by sales) in the industry. Tobin’s Q is the book value of asset plus market value of equity minus market value of equity, scaled by the book value of assets. FCF is the operating income before depreciation minus interest expenses, income taxes, and capital expenditures, scaled by the book value of total assets.

Leverage is the long-term debt plus the short-term debt, scaled by the book value of assets. Size is the natural log of the book value of assets. t-statistics are shown in parentheses.

Panel A:Panel Regression on payout policy with HHI

PAYOUT1 PAYOUT2

Panel B:Panel Regression on payout policy with CONC

*, **, and *** indicate significance at 10%, 5%, and 1%, respectively.