اثرات مالیات و تغییرات نظارتی بر سیاست تقسیم سود بانک های تجاری

The literature concerning the Tax Reform Act of 1986 (TRA) is extensive, but generally does not consider dividend policy changes related to TRA’s passage. One exception is Casey et al., but that work omits banking. An examination of banks is especially apt given TRA’s changes in tax rates and municipal bond categorization. Results show bank dividend policy to be different from other industries, as banks show no relation to past growth rates, beta, or an insider ownership as Rozeff’s model holds. The results support the idea that the lower the taxes, the higher the payout which is contrary to the dividend irrelevancy argument. However, the results are not robust in tests using data from a later period meant to more closely examine changing capitalization requirements’ impact on dividend policy.

In August 1997, Congress passed a balanced budget package that includes cuts in the capital gains rate to 18% by 2001. Along with the changes, comes debate as to whether dividend payout ratios would change depending on if capital gains or ordinary (dividend) income would be more preferable for investors. One way to judge the possible impact of this tax change is to examine the impact of past tax reform on dividend policy. The most recent applicable change in the tax code is the Tax Reform Act of 1986 (TRA). TRA eliminates much of the preferential tax treatment for capital gains versus ordinary (dividend) income. Prior to TRA, the maximum marginal tax rate for dividend income was 50% while realized capital gains were taxed at a 20% rate. As of January 1, 1987, the rates for dividends and realized capital gains changed to 38.5% and 28%, respectively, before being equalized at 33% as of January 1, 1988. In a recent paper, Casey et al. (forthcoming) examine the impact of TRA on dividend policy using the model of Rozeff (1982) model and find that different industries’ reactions to TRA varied. However, their study does not include the commercial banking industry. Given that TRA changes the exemptability of income from certain types of municipal bonds (see Grammatikos and Yourougou, 1990 for discussion) in which banks are traditionally active investors, banking firms’ reactions — including changes in dividend payout policies — could differ from other industries’ reactions. This study addresses that omission by examining banking firms’ dividend policy reactions to TRA.1 The finance literature contains reasons why one could expect increased, decreased, or no changes in dividend payout policies. Two arguments support increased dividend payouts with a reduction in the ordinary income tax rate. First, while capital gains can be deferred indefinitely (see Miller, 1986) or offset by capital losses, the relative relationship between the two income forms changes with TRA’s passage. Dividend income would no longer be taxed at a higher rate — thus, the possibility of greater investor preference for dividend versus capital gains income (positive tax argument). The second argument for increased dividend payouts holds that TRA’s repeal of the investment tax credit (ITC) could decrease investment, as projects would need greater cash flows before being viable. Greater dividend payouts could follow if banks pay dividends out of “left-over” funds (positive ITC argument). Conversely, the findings of Vogt (1994) and Whited (1992) that cash flow has a strong positive influence on investment spending suggests that ITC’s elimination could lead to lower dividend payouts (the negative ITC argument). The implication suggests that managers would offset the loss of the ITC’s cash flow savings by reducing dividend payouts to maintain current cash flow (and investment) levels.2 Finally, Miller and Scholes (1978) argue the irrelevancy of taxes in determining dividend policy given tax avoidance vehicles, thus, leading to an expectation of no change in dividend payouts (the tax irrelevancy argument). Section 2 reviews related literature. Section 3 explains the data and methodology of the paper while Section 4 reports the results. Section 5 concludes the paper.

The results show the importance of examining the banking industry’s dividend payout policy separately. The expected relations between banks’ dividend policy and expected future (loan) growth and the number of shareholders holds for the 1982–1992 sample period as it does for other industries. However, the other three variables of the model of Rozeff (1982) a firm’s past growth rates, beta and an insider ownership measure, have no significant relationship. While past studies find the original five variables’ relationships to hold across different time horizons, the current findings show that these relationships do not hold for the banking industry. Results from this sample period concerning TRA’s influence provide evidence that taxes are an important factor in dividend policy. Specifically, the data support the view that the lower the tax rate on ordinary income (dividends) relative to capital gains, the higher the payout ratio (the positive tax argument). The results can also support the positive ITC argument, as the tax changes cannot be separated in this study’s time frame. Importantly, the findings are contrary to arguments that dividend policy is irrelevant (and are contrary to the negative ITC argument). Also, regulatory capital requirements may play a role in the determination of banks’ dividend payouts. The degree that this relationship holds is uncertain given that the regulatory changes for increased capital occur during the same period as the TRA’s enactment. However, results from the 1992–1996 sample do not yield significant results. The poor performance of the regression models for the 1982–1996 sample period is disturbing, but less so when one considers the likely survivorship bias. The sample consists of only large “healthy” banking firms. Thus, no generalizations should be made regarding small and/or “unhealthy” banks. Clearly, further research concerning bank dividend policies will be needed over time.