Ex-dividend day price and volume: the case of 2003 dividend tax cut.

To gain further insight into dividend tax clienteles, we separate our sample into four dividend yield groups based on Graham et al. (2003) as follows (right-endpoint inclusive): (18) annualized dividend yield less than or equal to two percent, between two and four percent, between four and six percent, and greater than six percent. Table 5 presents the mean and median PDR of each dividend yield group. In the pre-Act period, the mean (median) PDR is significantly lower than one for each dividend yield group except the highest group which has a PDR that is insignificantly different from one. The mean (median) PDR is higher in the post-Act period than in the pre-Act period for each dividend yield group. The highest dividend yield group has a PDR higher than one in the post-Act period, consistent with the notion that the marginal investor for the highest dividend yield stocks are corporate investors. We also note that the standard deviation decreases in the post-Act period relative to the pre-Act period. We graph the mean PDR of each dividend yield group in Figure 2. (19) We illustrate that the mean PDR of each dividend yield group after the tax change is larger than before the tax change. The entire PDR curve shifts up and is less varied in the post-Act period compared to the pre-Act period.

Contrary to the prediction of the dividend tax clientele theory that a positive relation exists between the PDR and the dividend yield, we find a U-shaped relation in the pre-Act period. As the dividend clientele theory predicts, investors in the high tax brackets tend to hold low dividend paying stocks, which should lead to a low PDR. But in our sample, the PDR of the lowest dividend yield group (group 1--with an annual dividend yield less than or equal to two percent) is significantly higher than that of the next higher dividend yield group. We are not the first to document such a pattern, as similar results appear in Elton and Gruber (1970), and Michaely (1991) among others. Elton and Gruber (1970), however, attribute the high PDR for the low dividend yield stocks to the bias of the data as small dividends drive the PDR misleadingly high or to the large standard deviation of their results. In our study, we winsorize the data to reduce the impact of outliers and correct for heteroskedasticity of the dividend yield effect by assigning lower weights to low dividend yield stocks. Cloyd et al. (2006) discuss several factors which may distort the positive relation between the dividend yield and the ex-dividend day PDR (or excess return). In the post-Act period, the PDR and the dividend yield generally exhibit a positive relation.

[FIGURE 2 OMITTED]

We find similar results, as outlined in Table 6, when analyzing the ex-dividend day excess return behavior across the same dividend yield groups. We find that ex-dividend day excess returns for all the dividend yield groups are significantly positive before the 2003 Act. After the 2003 Act, excess returns decline significantly relative to pre-Act returns for each dividend yield group, and the two highest dividend yield groups have insignificant excess returns. Figure 3 illustrates the arch shape of the excess returns across dividend yield groups, which corresponds with the U-shape of the PDR depicted in Figure 2. Low (high) dividend yield stocks exhibit a positive (negative) relation between the excess return and the dividend yield. Our high dividend yield result is consistent with the findings of Naranjo, Nimalendran, and Ryngaert (2000) who suggest that the clientele for high dividend yield stocks are corporate investors.

ABNORMAL TRADING VOLUME ANALYSIS

As previously hypothesized, we expect that the trading volume around the ex--dividend day will decrease significantly after the 2003 Act (Hypothesis 3) and that the trading volume will be positively related to dividend yield and negatively related to risk and transaction costs (Hypothesis 4). Thus, to analyze tax-induced trading around the ex-dividend day, we estimate abnormal trading volume for ten days surrounding the ex-dividend day for the pre- and post-Act period. We also analyze abnormal trading volume for high dividend yield stocks that we define as an annualized dividend yield greater than four percent. We use four percent since Graham et al. (2003) find that most AV activity occurs in stocks with quarterly dividend yields above one percent and argue that the result is consistent with greater dividend capture activity where the payoff to such activity is highest.

[FIGURE 3 OMITTED]

Analyzing the full sample results in Table 7, Panel A, we find significant positive AV for most of the 11 days around the ex-dividend day and AV is much higher on the cum-dividend day (day -1) and ex-dividend day (day 0) than the other days. Panel B of Table 7 shows even higher AV for high dividend yield stocks, which is consistent with the tax-induced trading theory. When the dividend yield is high, the tax benefit of exchanging dividends and capital gains is greater, thus providing greater motivation for investors to trade. In the pre-Act period, the high dividend yield stocks have an average AV of 35.86 percent on the cum-dividend day and 28.62 percent on the ex-dividend day. In the post-Act period, the AV decreases to only 18.02 percent on the cum-dividend day and 19.83 percent on the ex-dividend day. Figure 4 depicts the CAV from days -5 to +5 for the pre- and post-Act period. The CAV for the pre- and post-Act period does not exhibit a pronounced difference for the full sample of dividend paying firms. Figure 5 depicts the CAV only for the high dividend yield stocks. For the high dividend yield sample, the post-Act period CAV is lower than that in the pre-Act period particularly after the cum-dividend day.

We also calculate the CAV as the sum of the AV during the 11-day event window encompassing the ex-dividend day. We further separate the sample into low-, medium- and high-yield dividend paying groups (annual dividend yields less than or equal to two percent, two to four percent, and greater than four percent, respectively). We present the average CAV for the 11-day event window in Table 8. For the full sample, the CAV is 27.29 percent before the 2003 Act and 31.53 percent after the 2003 Act. The difference is not significant at conventional levels. However, when we separate the sample into low-, medium- and high-yield groups, the strong positive relation between dividend yield and trading volume becomes obvious. The magnitude of the CAV for the low-yield group (3.22 percent) and the medium-yield group (0.49 percent) before the 2003 Act is insignificant from zero and much lower than that of the high-yield group (183.14 percent). In the post-act period, the CAV of the low- and medium-yield groups increases to 16.44 percent and 34.58 percent, respectively. But only the change in the medium-yield group is statistically significant. Alternatively, the CAV of the high-yield group declines significantly after the 2003 Act to 101.12 percent. These results are generally consistent with our fourth hypothesis. With the reduction of the tax heterogeneity in the market, ex-dividend day trading volume decreases for high dividend yield stocks as the tax-induced motives to trade decrease. However, the overall trading volume is not significantly different between pre- and post-Act since most of our sample is low- and medium dividend yield stocks. (20) For low- and medium-yield dividend paying stocks, most trading may be for non-tax reasons since a small dividend is not attractive in terms of tax considerations, however, investors may time their trading around the ex-dividend day to capture abnormal returns. Thus, the effect of the reduction in tax heterogeneity on tax-motivated trading does not show up in the full sample results.

[FIGURE 4 OMITTED]

[FIGURE 5 OMITTED]

We also analyze the effect of risk and transaction costs on the ex-dividend day trading volume. Following Michaely and Vila (1996), we regress CAV on dividend yield, idiosyncratic risk, beta, market capitalization, and a post-Act dummy variable for the full sample and each of the three dividend yield groups. Results are shown in Table 9. We find a significant negative relation between idiosyncratic risk and trading volume for high- and low-yield stocks, but an insignificant relation between idiosyncratic risk and trading volume for medium- yield stocks. (21) The coefficients of beta are significantly negative for all groups, consistent with the finding of Michaely and Vila (1996). Also the coefficient on market capitalization is positive and significant in all specifications that is consistent with the prediction that transaction costs decrease trading volume. In addition, we find the dividend yield effect is only significant for the high-yield sample and the full sample. The cumulative trading volume for high dividend yield stocks significantly declines after the 2003 Act even after controlling for risk, transaction costs, and market risk, indicated by the significant negative coefficient on the post-Act dummy variable: -0.820. In general, our results support a positive relation between trading volume and dividend yield, a negative relation between trading volume and risk, and a negative relation between trading volume and transaction costs as predicted by the fourth hypothesis, but our results appear to be driven by high dividend paying stocks.

CONCLUSION