Panel A: Non-Dividend Paying Firms Versus Dividend Paying Firms 12/31/1925-
12/31/2011 12/31/1925- 12/31/1972
12/31/1972- 12/31/2011
Sub-Period α Difference
Sub-Period β Difference
Average Return for Non-Dividend Paying Firms 0.0128 0.0136 0.0119 Average Return for Dividend Paying Firms 0.0116 0.0106 0.0128
Return Difference 0.001 0.003 -0.001 (0.81) (1.39) (-0.50) α -0.0045 -0.0032 -0.0049 0.0017 0.283 (-3.61) (-2.13) (-2.36) (0.65) (3.09) β 1.49 1.59 1.31
(H0: β=1) (8.10) (7.18) (7.56) R2 0.80 0.85 0.70
Panel B: NIFD Non-Dividend Paying, NIFD Dividend Paying, IFD Firms (12/31/1972-12/31/2011) ND:NIFD vs. D:NIFD IFD vs. D:NIFD IFD vs. ND:NIFD
Return Difference 0.0024 -0.0042 -0.0066 (1.62) (-1.47) (-3.78) α -0.0018 -0.0092 -0.0091 (-1.21) (-3.20) (-5.49) β 1.33 1.39 1.17
(H0: β=1) (9.41) (5.10) (3.97) R2 0.82 0.55 0.83
In parentheses are t-stats that are Newey and West (1987) adjusted for regressions. Without identifying firms in financial distress, Panel A reports parameter estimates in the regression of monthly returns for an equally weighted portfolio of non-dividend paying firms (ND) versus a portfolio of dividend-paying firms (D) (excluding ETFs and CEFs). In Panel B, firms have data from both CRSP and COMPUSTAT. The acronyms IFD and NIFD stand for “in financial distress” and “not in financial distress.” A firm is IFD if it has negative TTM earnings. There are three portfolios in Panel B (all equally weighted): firms that are NIFD and pay dividends (D:NIFD), firms that are NIFD and do not pay dividends (ND:NIFD), and IFD firms regardless of whether they pay dividends or not. The average number of firms in the D:NIFD, ND:NIFD, and IFD portfolios is 1,598, 1,469, and 1,178. In Panel A of Table 1, over the 12/31/1925–12/31/2011 period, average monthly returns for non-dividend paying firms exceed those of dividend paying firms but the difference is statistically insignificant. This result identifies no risk difference between non-dividend paying and dividend-paying firms. In the regression of portfolio returns for non-dividend paying versus dividend-paying firms, the slope coefficient, β, statistically exceeds unity, �̂�𝛽=1.49, which suggests greater risk for non-dividend paying firms. Since there is no difference in raw-returns but non-dividend paying firms have greater risk, the returns of dividend-paying firms are abnormally high compared with non-dividend paying firms. The alpha estimate is negative and statistically significant, 𝛼𝛼� = −0.0045. Sub period results in Panel A are similar to the entire sample. Raw return differences between dividend-paying and non-dividend paying firms are insignificant, the β-risk of non-dividend paying firms exceeds that of dividend-paying firms, and returns for dividend- paying firms are abnormally greater than non-dividend paying firms. Panel B of Table 1 reports average monthly return differences and parameter estimates for equation (1) in the regression of equally-weighted portfolio returns for one business class versus another. The three
4
The International Journal of Business and Finance Research ♦ VOLUME 9 ♦ NUMBER 2 ♦ 2015
business classes are: NIFD non-dividend paying (ND:NIFD), NIFD dividend-paying (D:NIFD), and IFD firms (regardless of whether they pay dividends or not). We do not distinguish the dividend decisions of IFD firms because they face more serious financial issues than dividend pay-out and Table 2 shows that only a small fraction of IFD firms pay dividends (9%). Removing IFD firms, returns for non-dividend paying firms increase relative to dividend-paying firms in Panel B of Table 1 compared with Panel A. In the first row, the return difference between ND:NIFD and D:NIFD is positive and statistically significant at roughly the 10% level (return difference is 0.0024 and the t-stat is 1.62). In addition, abnormal returns disappear. Higher risk for ND:NIFD firms relative to D:NIFD firms ( �̂�𝛽=1.33) accounts for the raw-return difference. The alpha estimate is insignificant (𝛼𝛼�=−0.0018 and the t-stat is −1.21). In the final two rows of Panel B, high β-risk for IFD firms relative to D:NIFD firms (�̂�𝛽=1.39) and IFD firms relative to ND:NIFD firms (�̂�𝛽=1.17) does not accord with low returns for IFD firms. Abnormal returns are negative and statistically significant in both cases (𝛼𝛼� = −0.0092 and 𝛼𝛼� = −0.0091 , respectively). Beginning in the following section, guided by the Galai and Masulis (1976) view that equity is a call option on the assets of a firm, we investigate the hypothesis that returns decrease with volatility and that this relation accounts for low returns for IFD firms. In addition, we present evidence that high-profitability firms have high returns from high growth-leverage despite high volatility. Table 2: Firms in Financial Distress, NASDAQ, and Dividend-Paying Firms
Fraction of Firms That Are IFD Fraction of Firms That Are
NASDAQ Fraction of Firms That Are
Dividend-Paying Panel a: CRSP (12/31/1972–12/31/2011) Non-Dividend Paying 68% Dividend-Paying 35% All Firms 55% 39% Panel B: CRSP & COMPUSTAT (12/31/1972–12/31/2011) Non-Dividend Paying 42% 70% Dividend-Paying 6% 33% NASDAQ 36% 24% Non-NASDAQ 18% 60% IFD Firms 72% 9% NIFD Firms 49% 52% All Firms 28% 55% 40% Panel C: CRSP, COMPUSTAT & I/B/E/S (1/15/1976–1/19/2012) Non-Dividend Paying 35% 69% Dividend-Paying 6% 29% NASDAQ 30% 27% Non-NASDAQ 13% 66% IFD Firms 69% 12% NIFD Firms 45% 55% All Firms 21% 50% 46%
Acronyms IFD and NIFD stand for “in financial distress” and “not in financial distress.” IFD firms have negative trailing twelve month earnings. Portfolio Analysis In Blazenko and Pavlov’s (2009) dynamic equity-valuation model, expected return decreases with volatility and increases with business growth. Since profitability underlies volatility and growth, we form portfolios with profitability and then explore relations between returns, volatility and growth-leverage. Corporate growth depends on profitability for several reasons. First, since earnings have high persistence (Fama and French, 2006), high earnings occur with good growth prospects that managers exploit with expansion investments. Second, with financing constraints (Froot, Scharfstein and Stein, 1993), managers finance growth largely internally and only when profitability allows. We require firms have data from each of the COMPUSTAT, CRSP, and I/B/E/S databases. CRSP is our source for share price and other stock market
