1、原文 ： The interaction of corporate dividend policy and capital structure decisions under differential tax regimes 1、 The interaction of capital structure and dividend policy Firm values are normalized with respect to the firm with zero debt and zero dividend payout. The panels in the figure indicate
2、that the combined net impact of corporate dividend and capital structure policies on firm value is directly affected by the pertinent tax rates at the time. We next discuss the implications of the model for dividend and capital structure policies under several historical tax regimes. Three represent
3、ative tax regimes (19791981, 19881990, 19932002) were chosen for analysis out of the ten that were in existence at some time during the three decades since 1979. The three representative tax regimes exhibit distinctly different set of tax rates both in terms of absolute values and relative to each o
4、ther. For this reason, these three contrasting regimes provide a suitable setting to test the value implications of our model. If our model provides a reasonable representation of firms capital structure and dividend policy decisions, the three contrasting tax regimes would be the ideal environment
5、to observe the fit between the models predictions and the empirical observations. 2、 Years 19791981 The application of the model using the tax rates from the period 19791981 reveals a subtle effect. The table and the figure depict normalized firm value, VD,/V0,0, as a function of the leverage D and
6、the dividend payout . The gain from leverage is positive only when the firm is at a relatively high payout ratio (above approximately 40%), with the maximum gain occurring at full (100%) payout. Interestingly, at a dividend payout level lower than 40%, increasing leverage lowers firm value. The reve
7、rsal of the leverage effect at lower payout ratios is driven by the relative levels of tax rates. During the years 19791981, the top marginal tax rate for personal income was very high in comparison to the tax rate for corporate income (70% and 46% respectively). In a tax rate environment such as th
8、is, high taxes paid by the bondholders for their interest income proceeds exceed the benefit from the tax deductibility of interest payments at the firm level. Since debt financing can be assumed to have zero NPV, this additional burden is borne by the shareholders. At high levels of dividend payout
9、 on the other hand, the taxation of the dividend income makes dividend payout even more disadvantageous compared to paying interest. In other words, now it would be more beneficial for the firm to borrow and pay interest rather than dividends. The benefit reaped from the tax deductibility of interes
10、t payments tilts the balance in favor of debt financing, and makes leverage more attractive. Another noteworthy observation about the 19791981 tax rate environment is the steep loss in firm value at very low debt levels in response to increasing dividend payout. According to our model, it was possib
11、le for an all-equity firm to experience losses in value up to 58%. The firm could mitigate this loss by maintaining a higher debt level. The tax regime that made the interesting features discussed above possible is not a short-term anomaly confined to the years 19791981. Indeed, the entire period be
12、tween the Great Depression and the late 1970s was characterized by a similar tax rate environment. Our model indicates that optimal policies to maximize firm value under such tax regimes required zero debt and zero dividend payout. This prescription interestingly comports with the observed leverage
13、policies of the time, when numerous prominent companies such as IBM and Coca Cola had little, if any, debt before the 1980s. However, if a firm would need to maintain high dividend payout levels, it would be better off by carrying a relatively high debt level at the same time. Traditional electric u
14、tility companies are examples that appear to fit this mold. 3、 Years 19881990 (and 19911992) The situation during the years 19881990 is unique because during that time the top marginal tax rates on ordinary income (thus on dividend and interest income) were nominally the same as the tax rate on capi
15、tal gains at 28%. In the following 2 years (19911992), the two tax rates remained very close (at 31.0% and 28.9% respectively). The result of the convergence in tax rates is visible in Fig. 2 for the19881990 and 19911992 panels. There is little if any moderating influence of the dividend payout on t
16、he leverage-firm value relation. The maximum theoretical gain from leverage is close to 50% regardless of the level of dividend payout. As discussed and anticipated on the comparative statics for our model, the influence of the dividend payout ratio vanishes due to the near-zero tax rate differentia
17、l ( pd pg) during the years 19881992. 4、 Years 19932002 In contrast to the reversal effect observed under the tax regime during 19791981, and similar to the situation during 19881992, the gain from leverage is always positive under the 19932002 tax regimes. The details of the gain from leverage rela
18、tion and the effect of the dividend payout for the years 19931994 and the year 2002 are available. As a departure from the previous tax regimes discussed above, throughout this decade-long time interval, the gain from leverage is significantly more pronounced for high payout firms. Although at low o
19、r zero debt levels increased dividend payout reduces the firm value, the negative impact of the dividend payout weakens as the debt level increases. In contrast to the maximum potential gain from leverage during 19881992 that reached up to 50%, the tax rate changes throughout the 1990s significantly
20、 reduced the maximum potential gain. the maximum potential gain was near 30% in 1993, and by 1998, approximately 20%,remaining at that level through 2002. 5、 Summary and empirical implications The nature of the combined impact of financial leverage and dividend policy on firm value over the years 19
21、792002 is found to be wide ranging as a direct result of the tax rate changes. We discussed above three distinct tax regime environments in detail. In the first interval 19791981, low leverage and low dividend payout leads to higher firm value. However, given a high dividend payout, the firm is better off by carrying a high debt level. That suggests a simultaneous increase or decrease in leverage and payout for firms. It is less likely to find firms with low leverage and high payout (which results in the minimum possible firm value). The empirical implication
