Printer Friendly

On the motivation for paying scrip dividends.

This paper investigates why some companies favor paying dividends in the form of shares rather than cash. By comparing firms that pay scrip dividends to firms that do not, I test the hypotheses that scrip dividends are motivated by tax savings, cash shortage, and signaling. The results show that this option is not offered to save in taxes, to signal future prospects, or to reduce the cash-flow problem. I do find, however, that scrip-paying firms have low growth options and high cash flow, suggesting that the scrip-dividend payment may not be motivated by financial considerations.

Recently, an increasing number of companies in the United Kingdom have given their shareholders the option to receive shares, called scrip dividends, in lieu of cash. Other forms of dividend distributions, such as cash dividends and share repurchases, involve cash outflows, but the scrip-dividend allows shareholders to choose between receiving dividends either in cash or in shares by signing an election form. Unlike stock dividends offered in the US, where the recipient shareholder is not taxed and does not generally have an opportunity to opt for cash (e.g., McNichols and Dravid, 1990), both scrip and cash dividends are taxed at the personal income tax rate. Moreover, unlike dividend reinvestment plans, in which the new shares are issued at a discount of 5% to 10% (e.g., Chan, McColough, and Skully, 1993), scrip dividends do not provide shares at a discount. They represent a capitalization of part of the firms' distributable profits, a way to issue new shares without offering any discount.

However, scrip dividends offer much the same benefits as stock dividends and dividend reinvestment plans. By offering this option, companies in the UK are able to retain cash without altering their payout policies. Further, they can save in taxes because the scrip-dividend option is not subject to the advanced corporation tax (ACT) payable on cash dividends. (ACT is a personal income tax at the basic rate that is paid by the company on behalf of its shareholders on each unit of gross dividends.) At the same time, scrip dividends provide shareholders with the opportunity to increase their holdings without incurring any transaction costs. These benefits can outweigh the potential administrative costs, which may include the expense of advertising the option; preparing, printing, and mailing the scrip-dividend prospectus; and printing and allocating the new shares.

However, scrip dividends are controversial because they are not offered to foreign investors, and they are taxed differently from cash dividends for some domestic investors. Under the imputation system in the United Kingdom, cash dividends carry a tax credit that can be claimed by all investors, whatever their personal income tax position. In contrast, the tax credit on scrip dividends can be claimed only by tax-paying individual investors. Individuals who have reliefs and allowances in excess of their income (referred to, below, as non-taxable individual investors), and corporate and tax-exempt investors forgo the tax credit when they opt for a scrip dividend. As a result, their after-tax return on a scrip dividend is lower than that on a cash dividend. This tax discrimination between cash and scrip dividends implies that corporate and non-taxable individual investors prefer cash dividends, and that tax-paying individual shareholders will opt for scrip dividends for which the firm issues additional shares. However, for shareholders who opt for cash dividends, such additional shares can result in a dilution of control, and a loss in future dividend increases if scrip dividends limit the scope for such increases.(1)

This paper investigates why companies favor the payment of dividends in the form of shares rather than cash. I do this by comparing the operational performance and other characteristics of all companies in the UK that distribute scrip dividends with those of a control group of non-scrip-paying, but otherwise similar, firms. I explore the propositions that scrip dividends are motivated by tax savings, lack of cash, and signaling and document the market reaction to this option.

The results show that there are no significant differences in these two groups' tax exposures. Compared to the control sample, firms that issue scrip dividends do not report a higher proportion of irrecoverable ACT in their accounts, nor do they have a higher proportion of their earnings generated abroad on which the ACT cannot be claimed. More than a third of the scrip-paying firms do not carry any ACT recoverable in their accounts, and nearly half do not generate any profit from abroad. Therefore, in contrast to companies' claims, scrip dividends are not issued to save on taxes, nor was this option adopted by all companies in the United Kingdom to overcome the problem of ACT irrecoverability.

The results also show that this option is not driven by cash shortage, nor is it a substitute for external finance and/or a cut in cash dividends. Moreover, the scrip-dividend option is not used to signal future growth prospects; the market does not value this option.

I find strong systematic differences in the size, dividend yield, and growth opportunities of these two groups. Firms that pay scrip dividends are, on average, large, and have high dividend yield, suggesting that the cash saved is substantial. However, these firms already have high cash How balances and low growth opportunities.

Why then do companies issue scrip dividends? The overall results suggest that firms in the UK do not appear to be making the optimal financial choice for the dividend payment vehicle, and that the scrip-dividend option is used to exacerbate the agency costs associated with the free-cash-flow problem. An alternative explanation could be that individual taxpaying investors request this option from their firms. This suggests that these firms are subject to monitoring by small shareholders, while large shareholders, such as corporate investors, are passive owners. I conclude by calling for more research to evaluate the extent to which other nonfinancial factors affect the decision to pay scrip dividends and the reasons why corporate investors fail to prevent companies from offering this option.

The paper is organized as follows. Section I develops the hypotheses. Section II describes the data and the methodology. Section III presents the empirical results and examines the robustness of the findings to the sample period and to alternative measures of the proxy variables. In Section IV, I attempt to rationalize the observed regularities. Conclusions are set out in Section V.

I. Hypotheses Tested

In this section, I classify the testable hypotheses into three main groups: taxation, cash shortage, and signaling. These hypotheses are first discussed independently. In the empirical section, I provide the joint implications.

A. Taxation

Under the classical system of corporation tax, the taxation of dividends at the firm and shareholder levels is not linked, and scrip dividends, like stock dividends in the US, are a cosmetic financial manipulation with no effect on the firm and its shareholders (e.g., Lakonishok and Lev, 1987). However, under the imputation system, a, scrip-dividend option is likely to affect the value of the firm and the post-tax return of its shareholders.

This effect is apparent when the scrip-dividend is compared to the cash alternative. A firm that distributes cash dividends must pay an advanced corporation tax (ACT) equal to the basic rate of income tax on the gross dividend. The ACT is first paid to the tax authorities 14 days after the end of the quarter in which the dividend is paid. It is then deducted from the firm's corporation tax liability, which is usually payable nine months after the end of the accounting period.

There are, however, two main conditions for the recoverability of ACT. The first is that ACT can only be offset against UK taxes. Thus, firms that pay cash dividends out of earnings generated abroad may not be able to recover their ACT. The second condition for the recoverability of ACT is that taxable profits should not exceed gross cash dividends. Although the firm is able to set this surplus ACT against its corporation-tax liabilities of preceding or immediately following periods, the problem is likely to be permanent when, for example, the proportion of overseas earnings is large.(2)

The scrip-dividend option does not involve the payment of ACT at the firm level. Therefore, issuing firms can conserve the cash that would otherwise have been paid out as ACT and bypass the problem of unrelieved ACT, and do so without reducing their dividend payout.(3) The institutional legislation underlying the imputation system implies that scrip dividends are likely to be favored by firms with potentially irrecoverable ACT, i.e., by those with low taxable profits and high accumulated recoverable ACT, and by those that generate a high proportion of their earnings from overseas.

To maximize these cash flow gains, firms are expected to encourage the majority of their shareholders to opt for scrip, rather than cash, dividends. Under the current UK tax system,(4) the tax credit on scrip dividends cannot be claimed by corporate investors or by individuals who have reliefs and allowances in excess of their income tax. For both these types of investors, the scrip option is treated as capital sum. For capital gains purposes, receiving the scrip option is treated as a reorganization of capital, with the new shares being acquired at no cost. Therefore, these investors' net after-tax return from scrip dividends is lower than that on cash dividends.

In contrast, domestic tax-paying individual investors are able to claim the tax credit on scrip dividends. For income-tax purposes, they are placed in the same position as individuals who have received cash dividends and reinvested the after-tax proceeds in the company. Since this option does not involve any transaction costs, it is likely to be favored by tax-paying individual investors. However, given that these investors are not the largest holders in the UK (e.g., Stock Exchange, 1991), the take-up rate is likely to be very low.(5)

In sum, the UK imputation system encourages firms to pay scrip dividends. However, this option is likely to be preferred only by tax-paying domestic individual investors.

B. Cash Shortage

In addition to tax incentives, firms are likely to be motivated to pay scrip, rather than cash, dividends either because of cash shortage or because of the resolution of the agency cost paradigm. In general, when a company has insufficient funds to finance the payment of its cash dividend, it can either raise external funds or cut its distribution. While a new issue leads to an increase in the monitoring of the managerial behavior (Easterbrook, 1984; Hansen and Torregrosa, 1992; and Rozeff, 1982), a dividend cut is likely to convey unfavorable information about the current and/ or future cash flows (e.g., John and Williams, 1985; and Miller and Rock, 1985). Retaining cash by using scrip dividends avoids increased monitoring and may convey less negative information to the market.

Scrip dividends can also provide issuing firms with several other benefits. For example, as with stock dividends (e.g., Eisemann and Moses, 1978), scrip dividends allow firms to retain cash and reduce a cash shortage problem. As with dividend-reinvestment plans (e.g., Wilson, 1982), scrip dividends can reduce the cost of capital for capital-intensive firms with high leverage and poor growth prospects, and when shareholders demand high dividend payouts. Firms such as these can experience difficulties in raising funds, since a new share issue will further depress the share price and borrowing opportunities are limited by already high leverage. Therefore, if the scrip-dividend option is motivated by cash shortage considerations, then scrip-paying firms should have low cash positions, high debt-to-equity ratios, and high dividend yields, and should be less likely to raise external finance. However, the scrip-dividend option is likely to exacerbate the free-cash-flow problem because it offers companies an opportunity to raise equity financing without using the primary market.

When companies pay cash dividends and at the same time finance externally, they reduce the agency conflict between managers and shareholders (e.g., Easterbrook, 1984; and Hansen, Kumar, and Shome, 1994). The agency cost paradigm was first studied by Jensen and Meckling (1976) and then extended explicitly to dividends by Rozeff (1982), Easterbrook (1984), and Jensen (1986). It suggests that when firms are profitable, managers finance their investments from retained earnings. Such firms are also likely to generate cash flow in excess of their positive-NPV investment opportunities. When managers do not have to submit to capital market monitoring to raise the financing needed, they may spend this income either on perquisites or unwise investments and acquisitions, rather than paying out the money to shareholders as cash dividends. To minimize this free-cash-flow problem, investors force managers to pay cash dividends and to raise new finance in the marketplace where they can be directly monitored and disciplined.

In the agency cost framework, the scrip option can worsen the free-cash-flow problem because it allows firms to retain cash, circumvent the capital market monitoring, and bypass the inflexibility that exists in corporate dividend policy (DeAngelo, DeAngelo, and Skinner, 1992). If this option is motivated by the free-cash-flow paradigm, it is likely to be favored by large firms with low growth opportunities, high level of cash flow, and high yield.(6)

C. Signaling

In the presence of asymmetric information between managers and investors, financial decisions, such as cash dividend payments, are likely to signal firm's high quality and favorable earnings prospects (e.g., John and Williams, 1985; Miller and Rock, 1985; Ravid and Sarig, 1991; and Yoon and Starks, 1995).(7) This signaling theory argues that managers have superior private information, and that dividend announcements change the information set available to shareholders, which results in a reevaluation of firms' share prices.

However, if the firm's share price is temporarily depressed and managers have information about a valuable new investment opportunity, simultaneously paying out cash dividends and financing the investment project externally (or, alternatively, cutting the cash dividend) could be prohibitively expensive if the firm cannot effectively inform its investors about these investment opportunities. In this case, managers are likely to issue scrip dividends and to encourage all shareholders to opt for scrip rather than cash dividends. At the same time, managers convey to investors the information that corporate prospects have improved not worsened. As in the case of a stock split (e.g., Asquith, Healy, and Palepu, 1989; and Peterson, Millar, and Rimbey, 1996), paying scrip dividends implies that managers are confident that they will maintain or increase the dividends and the earnings per share even if all the scrip is taken up, and that the firm's prospects will lead to future share price appreciation. These arguments imply that scrip-dividend firms are likely to have larger increases in their subsequent earnings, to maintain or increase their dividends, and to generate higher capital gains than are firms that pay only cash dividends.

D. Market Reaction to Scrip Dividend Payment

The market reaction to scrip dividends depends on investors' perception of the costs and benefits of this option. On the announcement date, issuing firms' share prices should increase to capitalize the main benefits of the scrip option such as savings in tax, cash, underwriting fees, and other issue costs (e.g., Scholes and Wolfson, 1989). As with stock dividends (e.g., Grinblatt, Masulis, and Titman, 1984), the announcement of a scrip-dividend option could also be interpreted as a signal of higher future earnings and will result in a rise in share prices. Other information-asymmetry explanations suggest that this option will result in a reduction in shareholder wealth, and thus in a decrease in the announcement-date security prices. For example, if the scrip option exacerbates the agency costs paradigm or if it signals cash shortage and/or financial distress, announcement-date security returns should be negative and/or lower than those of firms that pay only cash dividends.

The announcement-date security returns can also be driven by the trading patterns of taxable individual investors and other shareholders. As previously noted, tax-paying individual shareholders have an opportunity to acquire additional equity in the firm by reinvesting cash dividends without incurring any transaction costs or losing their tax credit. These investors are expected to trigger buying pressure on the announcement date. Other investors are likely to sell their holdings to avoid losing a form of involuntary wealth transfer to participating shareholders, especially if the additional shares result in a dilution of their control and a loss in future dividend increases. In this case, the market reaction cannot be predetermined.

II. Data and Methodology

Scrip dividends that are offered in the United Kingdom during the six-year period from 1987 to 1992 comprise the sample. Scrip-paying firms are drawn from Extel Takeovers, Offers, and New Issues publications. To avoid survivorship bias, all companies that offered this option, even if they have been subsequently acquired or gone out of business, are included as long as their relevant data are available. The sample includes 250 companies that paid the scrip option at least once over the sample period, resulting in 796 observations.(8) MicroExstat is the source of year-end dates and all the accounting data used in this study. MicroView and Datastream provide data on share prices, market values, and dividend yields.

In addition to the test sample, I construct a control sample by matching every company that paid scrip dividends with a similar company that paid only cash dividends. Each year, scrip-paying firms are ranked by industry and a control firm is selected from that industry.

Table 1 provides the MicroExstat industrial classification of firms in the test and control samples.(9) Column 3, Table 1, reports the number of companies in each industry group that paid scrip dividends. Column 4 shows the proportion of these firms relative to the total number of companies listed in each of these industries. On average, the sample firms represent 14% of the total number of companies listed in the London Stock Exchange. However, the distribution of sample firms across industries is not homogeneous. In 16 out of 80 industries (20%), no single firm issues a scrip option. In the remaining industries, this proportion is less than 50% in most cases. However, in the banking, insurance life, and insurance composite sectors, more than half of the listed companies issue scrip dividends. (For example, out of 15 banks, eight paid scrip dividends.) For these three industries, I was unable to match seven scrip-paying firms. Thus, the final control sample includes 243 companies and 784 observations.

[TABULAR DATA 1 NOT REPRODUCIBLE IN ASCII]

In each industry, I attempt to match the two samples by size, defined either in terms of the market value of equity, total assets, or sales. However, the scrip-dividend-paying firms are, on average, much larger than the control firms, even though the control firms are selected to be as close in market value of equity as possible to the test company.(10) Columns 6 and 7 of Table 1 report the distribution of the test and control samples by industry and market value. On average, the market value of test firms is 1,374 million [pounds sterling], while that of the control firms is 859 million [pounds sterling]. Column 8 of Table 1 shows the proportion of cases by industry in which the market value of each test firm is higher than that of the control firm. On average, the test firms are larger than the control firms in 61% of the total cases and in 63 industries (73% of industries covered in the sample). There are only two cases where the market value of the scrip firms is equal to that of the control firm.

A number of proxy variables are defined to test the hypotheses developed in the previous section. These variables are reported in Table 2.

[TABULAR DATA 2, NOT REPRODUCIBLE IN ASCII]

Three proxy variables capture the taxation impact on the decision to issue scrip dividends. The first is the ratio of the reported recoverable ACT to the firm's market value of equity. This variable should measure the magnitude of the ACT carried forward and the probability of losing the tax credit if a cash dividend is paid. The second is the proportion of overseas sales to measure the probability of having to pay cash dividends from earnings generated abroad.(11) Under the tax hypothesis, both these variables should encourage firms to issue scrip dividends. The third is the effective corporation tax rate. This is computed as the ratio of the reported corporation tax expense over profit before tax, and measures the extent to which ACT is irrecoverable. This variable should be negatively related to the likelihood of scrip dividend payment.

I test the cash-shortage hypothesis by comparing the cash and leverage position of the two samples. If the scrip-dividend decision is driven by cash shortage, then the test sample's net cash from operations to total assets should be lower than that of the control firms. Moreover, if firms in financial distress issue scrip dividends, their capacity to raise external finance should be lower -- and their leverage higher -- than that of control firms. These two variables are measured by the ratio of new issue of equity to the market value of equity and total debt to total assets, respectively.

To test the free-cash-flow hypothesis, a variable that measures investment opportunities is first defined and then combined with the firm's cash, position. The investment opportunity set captured in this study is Myers' (1977) notion of assets-in-place relative to growth opportunities. Tobin's q, defined as the ratio of market value of equity plus book value of debt to total assets (e.g., Opler and Titman, 1993), is used to measure growth opportunities.(12) As in Lang, Stulz, and Walkling (1991), a variable identifying firms that simultaneously have low investment opportunities and high cash flow is constructed to separate the free-cash-flow explanation of scrip dividends from the cash-shortage proposition. This variable is equal to net cash flow from operation over total assets if q is lower than unity, and zero otherwise. If managers issue scrip dividends to overinvest, this free-cash-flow variable is expected to be significantly higher for the scrip-paying firms than for the control firms.

I test for the signaling hypothesis by comparing the subsequent financial performance of the scrip-paying firms relative to that of the control firms. If a scrip dividend is used by managers to signal a favorable outlook to the market, then scrip-paying firms' subsequent earnings and dividends should be higher. Barclay and Smith (1995) note that future, abnormal returns can test for the signaling hypothesis. I compute, for each individual company, the one-year-ahead changes in earnings and dividends per share, and the annual cumulative abnormal returns (CAR) from year-end to year-end t+1. I use the market model (Brown and Warner, 1985) in which the coefficients [Alpha] and [Beta] are obtained from an OLS regression of firms' daily returns on their corresponding sector daily returns over the previous accounting year.

I also use the market model to compute the daily abnormal returns around the announcement dates. The expected returns are based on the coefficients of the market model computed over the - 150 to -20 days prior to the event date. Firms announce their intention to pay scrip dividends when they announce their quarterly, interim, or final earnings. I first use all these announcements as events dates and obtain 1,400 events for the test sample and 1,360 for the control sample. For each event, I compute the three-day daily abnormal returns for the test and control firms. I expect these abnormal returns to reflect the market perception of the scrip option. However, these results may not measure correctly the market perception of this option, because some announcements could be expected to recur if the firm has offered this option in its previous quarterly or interim results. I therefore restrict my analysis to the first announcement by excluding, for each firm, any subsequent scrip payments. I find that the results are not affected by this restriction. The reported results are based on the restricted sample to evaluate the market reaction to unexpected scrip option.

III. Empirical Results

This section describes the various empirical tests performed and presents their results.

A. Univariate Tests

To test for the tax impact on the decision to pay scrip dividends, I first classify all companies in the test and control samples into relative ACT recoverable and overseas sales quintiles. These results are reported in Table 3. Panel A of Table 3 shows that 35% of the test and 38% of the control firms do not disclose any ACT recoverable in their reports. This suggests that these firms generate enough taxable profits to deduct their ACT directly from their tax liability. Although 22% of the test firms carry a large proportion of ACT recoverable (more than 1% of the market value of their equity) compared to only 12% of the control firms, the overall difference between the two samples is not statistically significant ([Chi.sup.2] has a p-value of 0.48). Similarly, Panel B of Table 3 shows that the difference in the proportion of total revenue generated abroad between the test and the control firms is not statistically significant ([Chi.sup.2] has a p-value of 0.44). In particular, 47% of the test firms and 44% of the control firms do not generate any overseas turnover. These results imply that scrip dividends are not offered by companies that suffer from ACT problems.

[TABULAR DATA 3 NOT REPRODUCIBLE IN ASCII]

Table 4 reports the mean values and the mean differences of each of the proxy variables for the scrip and control samples. Panels A and B exclude financial companies because of their special tax treatments and specific accounting characteristics. Panel A shows that t-tests for differences in means between all the tax variables are not significant at any conventional confidence level. Firms that issue scrip dividends do not report larger recoverable ACT, do not have lower taxable profits, and do not generate higher overseas sales than the control firms. These results suggest that the payment of the scrip-dividend option is not motivated by tax considerations.

[TABULAR DATA 4 NOT REPRODUCIBLE IN ASCII]

Panel B of Table 4 reports the cash and leverage differences between the test and control samples. Firms that pay scrip dividends appear to be highly levered and exhibit significantly higher dividend yields than does the control group. However, there is no evidence of cash shortage, as scrip-paying firms generate relatively the same level of cash flow and raise the same amount of equity as firms that pay only cash dividends.

Panel C of Table 4 reports the differences in the variables used as a proxy for the free-cash-flow hypothesis. The results show that there are statistically significant differences in the growth opportunities and size between firms that pay scrip dividends and those that do not. Scrip-paying firms have substantially lower growth opportunities, as measured by Tobin's q, than other firms (t = -4.29). As reported in Table 1, scrip-dividend firms are larger than their matched counterparts (t = 10.37). The combination of the-results reported in Panels B and C suggest that scrip-paying firms have the characteristics of mature firms. They are large, have low growth, exhibit significantly higher dividend yield, and use a large proportion of debt mainly to finance their assets-in-place (e.g., Smith and Watts, 1992). Furthermore, scrip-paying firms have high cash positions and low investment opportunities. However, the difference between the two groups is not significant.

Panel D of Table 4 shows that there is no significant difference in the subsequent period's growth in earnings between firms that pay scrip dividends and those that do not. This implies that the cash retained through the scrip dividend may be reinvested to yield the existing shareholders' required rate of return. However, the similarities in dividend growth rates between the two samples imply that the issue of additional shares does not prevent scrip-paying firms from increasing their dividends. Although the average CARs are higher for scrip-paying firms, the difference in means is not significant at any conventional confidence level. These results are striking. They suggest that scrip-paying firms are not likely to generate superior returns to compensate their corporate investors for giving up the cash dividend and tax credit if these shareholders elect to receive scrip.(13)

Panel E of Table 4 reports the announcement-period abnormal returns. Both the test and the control firms generate positive and significant abnormal returns of 0.298% (t = 5.72) and 0.238% (t = 4.75), respectively. This abnormal performance happens only on the announcement dates. However, the differences across both groups are not significant, which suggests that the market values scrip and cash dividends equally, and that the scrip option does not provide any additional information beyond that conveyed by cash dividends.

However, the lack of reaction to a scrip-dividend announcement could be the result of other confounding effects and uncertainties. First, since scrip dividends are announced at the same time as earnings, the announcement-date abnormal returns are likely to be affected by other factors contained in the reports, not linked to the scrip option at all. Second, although companies announce their intention to offer the scrip option when they announce their dividends, the terms of the option are known only after the shares are quoted ex-dividend, i.e., about 15 working days after the announcement date. At this point, the number of shares each shareholder is entitled to is computed as the ratio of cash dividends over the average share price five days after the ex-day. Third, the market is not likely to anticipate the take-up rate because this rate is only known after the election date. Fourth, companies can cancel the scrip option if the share price, on the last date for receipt of elections, is 15% below the average price over the ex-day period. These institutional considerations imply that the market reaction to the scrip option is not confined solely to the announcement date.

B. Regression Results

The above cross-sample differences suggest that taxation, financial distress, and signaling are less important motives behind the decision to issue scrip dividends. The cash flow hypothesis is by far the most important determinant of such a financial decision. In this section, I extend the analysis by linking the likelihood of paying scrip dividends to the hypotheses stated in the previous section.

Table 5 reports the results from several multivariate logit regressions in which the dependent variable is one for scrip dividend firms and zero for the control firms. A set of independent variables measure the tax, financial distress, free cash flow, and signaling effects. Columns 1, 2, and 6 exclude financial companies. Column 1 reports the results for the tax hypothesis and shows that neither the tax coefficients nor the regression fit, as measured by [Lambda.sup.2], is significant.(14) The results provide further evidence that taxation is not a major determinant in the scrip-dividend decision.

[TABULAR DATA 5 NOT REPRODUCIBLE IN ASCII]

Column 2 reports the results of the cash-shortage hypothesis and shows that firms that issue scrip dividends have significantly higher debt-to-equity ratios and higher dividend yields. These results suggest that firms issue scrip dividends to maintain cash needed to service their hard contracts and to avoid cutting dividends. However, the positive (though not significant) coefficients of cash flow from operations and new equity issues indicate that these firms are not short of cash.(15) Thus, the results are not consistent with the cash-shortage hypothesis.

To test the free-cash-flow hypothesis, I run separate regressions on q and the cash flow variable NCFO/TA x low q because of the relatively high correlation between these two variables. (See Table 6.(16) The results, reported in Column 3 of Table 5 are consistent with the free-cash-flow hypothesis, as they indicate that firms with low-growth opportunities and net cash flow from operations are more likely to issue scrip dividends. Column 4 of Table 5 reports that large firms and those with lower investment opportunities, as measured by Tobin's q, are more likely to issue scrip dividends than smaller and growth firms.

[TABULAR DATA 6 NOT REPRODUCIBLE IN ASCII]

Column 5 of Table 5 provides the results for the signaling hypothesis. There does not appear to be any statistically significant relation between the likelihood of issuing scrip dividends and the subsequent growth in earnings, dividends, and/or share prices. The regression fit is not significant, which implies that there is no rejection of the joint hypothesis that the coefficients of the signaling variables are zero. This suggests that scrip dividends do not lead to a firm's financial improvement.

Column 6 of Table 5 allows for possible interaction between the hypotheses. In this equation, I exclude the cash flow variable, NCFO/TA x low q, because of the high correlation of this variable with the growth opportunities variable, D(q [is greater than] 1). (See Table 6.) The results show that scrip-paying firms have higher dividend yields, use more debt in their capital structure, are larger, and have less favorable growth potential than control firms. Scrip-paying firms are less likely to suffer a loss in their tax credit and/or to return to the market to seek additional funds, do not generate less funds from their operations to constrain their cash-dividend payment, and are not more likely to generate higher earnings and capital gains than firms that pay only cash dividends.

C. Alternative Proxies

Since the literature does not offer single measures of taxation, cash flow, leverage, relative size, and market index, I test the sensitivity of my results by using a number of alternative variables to proxy for the hypotheses. To test the tax hypothesis, I compute the ratio of ACT recoverable to the tax paid, and the ratio of total tax liability(17) to pre-tax profit. I expect both these variables to be positively related to the probability of issuing a scrip dividend if this option is motivated by tax considerations.

I compute cash-dividend cover (i.e., the ratio of net funds from operations before accounting for dividends to the amount of dividends) and operating profit over total assets and expect these two variables to be lower for scrip-paying firms if this option is driven by cash shortage. These firms should exhibit higher long-term debt over capital employed and higher dividend payout ratio if the scrip-dividend option is driven by cash shortage.

Following Chung and Charoenwong (1991) and Collins and Kothari (1989), I now define growth opportunities, q, as the ratio of market value of equity to the book value of equity. Then, as in Opler and Titman (1993), I compute a dummy variable high CF x low q equal to one if operating profit over total assets is higher than the mean and q is lower than one, and zero otherwise. I also compute a dummy variable low CF x high q to combine firms that are characterized as needing cash (average operating profit to total assets below the mean) to finance large growth opportunities (Tobin's q above unity). This variable should be negative if the decision to use scrip dividends is motivated by the free-cash-flow paradigm.(18)

Finally, I test the sensitivity of the signaling variables by computing the cumulative abnormal returns based on the raw returns and the market model, using the FTA and the FTSE as alternative market indices.(19)

The results reported in Table 7 are similar to those of Table 5. Column 1 provides further support for the free-cash-flow hypothesis. Column 2 shows that the dummy variable low CF x high q has a significant negative coefficient, suggesting that firms that are unlikely to have free-cash-flow problems do not issue scrip dividends.

These results are neither sensitive to the introduction of other proxy variables (Column 6) nor sample-dependent. This is shown in the last two columns of Table 7, where the logit equations are estimated for each of the two equal subperiods, 1987-1989 and 1990-1992. These two subsamples correspond to growth and depression in the UK economy. In these regressions, I exclude the investment growth opportunities variable, q, because of its high correlation with the high CF x low q variable. The dummy variable, high CF x low q, which is designed to capture the prediction of the free-cash-flow theory, is positive and statistically significant in Column 6 and in both subsample periods. The results imply that firms that are more likely to have free-cash-flow problems, i.e., those with higher cash flow and lower growth opportunities, are more likely to issue the scrip option, while those that need cash to finance their growth potentials (low CF x high q) are less likely to issue scrip dividends.

The results, reported in Column 6 and in the two subperiods, fail to provide support for the tax hypothesis, because none of the alternative tax variables is significant. When the model includes only the two tax variables, the regression fit is not statistically significant, implying that the joint hypothesis that the coefficients of the tax variables are zero cannot be rejected. Column 6 also shows that scrip-dividend firms have high payout ratios and high levels of long-term debt in their capital structure. However, these companies do not appear to suffer financial distress or to be short of cash. Although the coefficient of the cash-dividend cover is negative, it is not statistically significant at any conventional confidence level. This implies that the payment of scrip dividends is not motivated by cash shortage. Also, the results do not provide support for the signaling hypothesis; the coefficients of the cumulative abnormal returns using alternative market indices are not significant (Columns 3, 4, and 5).

IV. Why Do Companies Pay Scrip Dividends?

The study's results suggest that scrip-paying firms may not be following optional dividend policies. First, the take-up rate is likely to be low because the dominant shareholders in the UK (the corporate investors) are unlikely to,opt for the scrip dividend. Therefore the financial benefits that firms derive from this option are likely to be lower than the costs.

Second, my analysis is unable to find any financial rationale for this option. In contrast to what is widely claimed, companies do not issue scrip dividends because of ACT recoverability problems. I find that scrip-paying firms are not likely to pay higher taxes or to carry higher ACT recoverable in their accounts than the control firms. Many scrip-paying firms do not report ACT recoverable in their accounts and do not have foreign earnings. These results imply that taxation does not drive all my test firms to pay scrip dividends.

The results also show that scrip dividends do got signal cash shortage and/or financial distress. I find that scrip-paying firms are not less likely to go to the capital market to raise new equity, nor do they generate lower cash flow from their operations than the control firms. Furthermore, the fact that size, which can be a proxy for the inverse probability of default, is strongly and positively related to the likelihood of paying scrip dividends suggests that scrip-paying firms are not financially distressed. Further research will determine the extent to which firms that originally issue scrip dividends because of cash shortage subsequently cut their dividends if the take-up rate is low.

This study's fundamental result is that scrip-paying firms exhibit all the characteristics of companies that are suffering from internal and external agency costs. I find that the test firms are large, mature, and have high yields. Furthermore, my results show that firms with higher cash flows and lower growth opportunities are more likely to issue the scrip option, while firms that need cash to finance their growth potential are less likely to issue scrip dividends. Thus, it seems possible that companies issue scrip dividends to invest the cash saved in non-positive-NPV projects.

However, despite these characteristics, I find that scrip-paying firms do not necessarily waste retained cash. They do not generate lower returns over the subsequent accounting period, and they are not more likely to reduce their earnings or cut in their dividends than the control firms. These results can be extended to show that the severity of potential agency problems is more pronounced in scrip-paying firms with low managerial-share ownership (e.g., Lehn and Poulsen, 1989; and Eckbo and Verma, 1994).

One possible explanation for why companies pay scrip dividends could be that managers issue the scrip option for their own benefit, for example, to trade on insider information. However, the election date is unlikely to coincide with the timing of any insider trading. Furthermore, my results refute this possibility as the scrip-paying firms do not generate any excessive abnormal returns in the post-announcement period. An alternative explanation is offered by Lasfer (1997), who conducted a questionnaire survey among the finance managers of scrip-paying firms and control firms. Managers revealed that the decision between offering the scrip option or paying only cash dividends is substantially affected by shareholder pressure. The managers of scrip-paying companies report that their decision was the result of shareholders' demands at the annual meetings. Managers whose firms pay only cash dividends say that this option is not requested by shareholders. These results suggest that firms are subject to small-shareholder monitoring, and imply that this activism is not limited to takeovers (e.g., Strickland, Wiles, and Zenner, 1996).

Further research is needed to address the question of why, contrary to expectations (e.g., Shleifer and Vishny, 1986), these firms are not subject to monitoring by large corporate investors. This research might assess whether the expected gains from such monitoring exceed the expected costs, as in the model of Admati, Pfleiderer, and Zechner (1994), by analyzing the take-up rate together with the corporate, insiders, and tax-paying individual investors' holdings in scrip-paying firms.

V. Conclusion

This paper investigates why companies might favor the payment of dividends in the form of shares rather than cash. It does so by comparing the financial performance of firms that issued scrip dividends over the period 1987-1992 to industry-matched control firms that paid only cash dividends.

I find that in contrast to what is widely claimed, companies do not use the scrip option to save in taxes or to resolve their ACT problems. I also show that the scrip option is not offered when a firm is short of cash or in financial distress.

However, I do find that scrip-paying firms exhibit all the characteristics of companies that are suffering internal and external agency costs problems; these firms are large, have high yields, and a combination of high cash flow and low growth opportunities. These results suggest that scrip-paying firms may not be behaving rationally.

I suggest a number of areas for further research that will assess the extent to which the granting of this option is driven by taxable investors' requests rather than by any financial rationale. Such research could include an analysis of the take-up rate, and of the holdings of various investors in scrip-paying firms. A deeper understanding of the influences of these additional factors will enable us to understand why firms pay scrip dividends.

(1) Since scrip dividends increase the issued share capital, the total amount of dividends paid must be increased for the dividend-per-share to be increased or maintained. Because this option favors mainly those shareholders taxed at a higher rate of income tax, tax-exempt institutions in the UK have recently become concerned about the widespread adoption of this practice (e.g., The Financial Times, April 19, 1993). Overseas investors are not offered this option because the new shares are not registered under the securities laws of these foreign countries. These investors may share similar concerns.

(2) See Lasfer (1996) for further details on the UK tax system and for an empirical evidence on the impact of taxes on cash dividends. Papaioannou and Savarese (1994) provide an empirical investigation of the impact of taxation on dividends in the US.

(3) The payment of foreign income dividend, offered by companies in the UK since 1993, is another form of bypassing the problem of unrelieved ACT.

(4) See Sections 230 and 249 to 251 of the Inland and Corporation Taxes Act 1988 (as amended) and Sections 141 and 142 of the Taxation of Chargeable Gains Act 1992 for the tax treatment of scrip dividends.

(5) Unfortunately, data on the take-up rate is not readily available. Recently, to alleviate this tax discrimination and to encourage tax-exempt investors to opt for scrip dividends, several companies have offered enhanced scrip dividends, whereby the basis for computing the equivalent number of shares is higher than the net cash dividend by up to 50%. 1 do not analyze the enhanced scrip dividends in this paper because they were issued as "one off" by a relatively small number of companies in 1993.

(6) Jensen (1986) argues that firms suffering from free-cash-flow problems are closely monitored by the market through the imposition of high dividend payments. He attributes the high dividend yields to the high probability of free-cash-flow problems. He also argues that large mature firms are likely to suffer from this problem, but that debt finance reduces the propensity of managerial overspending.

(7) A number of studies show, however, that dividend changes do not convey information about future earnings (see, for example, Jensen and Johnson, 1995; and DeAngelo, DeAngelo, and Skinner, 1992, 1996).

(8) The total sample of firms is divided as follows: 59 companies in 1987, 87 in 1988, 122 in 1989, 149 in 1990. 181 in 1991, and 198 in 1992. Financial companies are included only in the analysis of returns but are excluded in the analysis of the tax and financial ratios because of their special tax treatment and their specific financial characteristics. The sample of industrial and commercial companies includes 202 companies and 640 cases for each of the test and control samples. Excluding financial companies from the analysis of returns and other market variables did not make significant changes in the reported results.

(9) The industrial classifications used in MicroExstat are those used by the London Stock Exchange and the Institute and Faculty of Actuaries.

(10) For example, I find that 66% of the constituents of the FTSE 100 Index, the London Stock Exchange 100 largest companies, pay scrip dividends. The selection of the control sample on the basis of total assets or total sales did not significantly change the composition of the firms included in the control sample.

(11) Overseas profits are not reported in the published accounts.

(12) Lehn and Poulsen (1989); Opler and Titman (1993); and Lang, Stulz, and Walkling (1991) are among the studies that use Tobin's q to test the free-cash-flow hypothesis. If a firm undertakes a value-maximizing level of investment, its average q will exceed unity when its investments are scale-expanding and exhibit decreasing marginal efficiency of capital. Stulz (1990) shows the conditions that must be met for the free-cash-flow hypothesis to hold. Chung and Pruitt (1994) show that this measure of Tobin's q is highly correlated with measures of q based on replacement cost of assets.

(13) The results reported in Panels B and C imply that firms retain cash to invest in negative-NPV projects. However, the findings in Panel D suggest that firms are not likely to waste cash. Although one year subsequent to the payment of scrip dividends may be too short a period, an extension may produce confounding effects and misspecification of long-horizon tests. (See Kothari and Warner, 1997.)

(14) By comparing the restricted and the unrestricted log likelihoods, the chi-squared statistic tests whether all slopes of the logit regression are zero. If the chi-squared is higher than its critical value, then the joint hypothesis that the coefficients of the independent variables are zero is rejected. (See Greene, 1993, Chapter 21, for details.)

(15) I have also used other measures of cash shortage, such as interest cover and net cash flow from operations over interest paid, and find that companies that issue scrip dividends do not appear to be short of cash. See Section III.C for alternative measures of cash flow.

(16) The effect of multicollinearity is to increase the variance, thus making an otherwise significant variable insignificant. Other highly correlated variables in Table 6 include, for example, NI/ME and ACTR/ME. However, none of these variables is significant in the other, separate regressions.

(17) The total tax liability is the sum of ACT recoverable, UK corporation tax, overseas tax, deferred tax, related companies tax, exceptional tax, non-current deferred tax, and tax due within one year.

(18) When I use alternative measures of growth opportunities, such as new investment in fixed assets to total fixed assets, net asset growth, earnings growth and sales growth, as in Lehn and Poulsen (1989), the results do not differ significantly from those reported above. Similarly, I use cash and marketable securities over total assets as a proxy for cash flow (Lang, Stulz, and Walkling, 1991), and corporation tax liability over market value as a proxy for the effective corporation tax rate (Lehn and Poulsen, 1989) without any significant change in the results.

(19) The Financial Times All Shares (FTA) Index is based on share price movements of some 680 (800 after 1993) of the more commonly traded quoted companies in the London Stock Exchange. The Financial Times Stock Exchange (FTSE) index is based on prices of the 100 most valuable British quoted companies, each weighted in proportion of its total market value.

References

Admati, A.R., P. Pfleiderer, and J. Zechner, 1994, "Large Shareholder Activism, Risk Sharing, and Financial Market Equilibrium," Journal of Political Economy (December), 1097-1130.

Asquith P., P. Healy, and K. Palepu, 1989, "Earnings and Stock Splits," The Accounting Review (July), 387-403.

Barclay, M.J. and C.W. Smith, 1995, "The Maturity Structure of Corporate Debt," Journal of Finance (June), 609-631.

Brown, S. and J. Warner, 1985, "Using Daily Stock Returns: The Case of Event Studies," Journal of Financial Economics (March), 3-31.

Chan, K.K.W., D.M. McColough, and M.T. Skully, 1993, "Australian Tax Changes and Dividend Reinvestment Announcement Effects: A Pre- and Post-Imputation Study," Australian Journal of Management (June), 41-62.

Chung, K. and C. Charoenwong, 1991, "Investment Options, Assets in Place and the Risks of Stocks," Financial Management (Autumn), 21-33.

Chung, K.H. and S.W. Pruitt, 1994, "A Simple Approximation of Tobin's q," Financial Management (Autumn), 70-74.

Collins, D.W. and S.P. Kothari, 1989, "An Analysis of Intertemporal and Cross-sectional Determinants of Earnings Response Coefficients," Journal of Accounting and Economics (July), 143-181.

DeAngelo, H., L. DeAngelo, and D. Skinner, 1992, "Dividends and Losses," Journal of Finance (December), 1837-1863.

DeAngelo, H., L. DeAngelo, and D. Skinner, 1996, "Reversal of Fortune: Dividend Signaling and the Disappearance of Sustained Earnings Growth," Journal of Financial Economics (March), 341-371.

Easterbrook, F., 1984, "Two Agency Costs Explanations of Dividends," American Economic Review (September), 650-659.

Eckbo, E. and S. Verma, 1994, "Managerial Shareownership, Voting Power and Cash Dividend Policy," Journal of Corporate Finance (March), 33-62.

Eisemann, P.C. and E.A. Moses, 1978, "Stock Dividends: Management's View," Financial Analysts Journal (July/ August), 77-80.

Greene, W.H., 1993, Econometric, Analysis, New York, NY, Macmillan.

Grinblatt, M., R. Masulis, and S. Titman, 1984, "The Valuation Effects of Stock Splits and Stock Dividends," Journal of Financial Economics (December), 461-490.

Hansen, R.S. and P. Torregrosa, 1992, "Underwriter Compensation and Corporate Monitoring," Journal of Finance (September), 1537-1556.

Hansen, R.S., R. Kumar, and D.K. Shome, 1994, "Dividend Policy and Corporate Monitoring: Evidence from the Regulated Electric Industry," Financial Management (Spring), 16-22.

Jensen, G.R. and J.M. Johnson, 1995, "The Dynamics of Corporate Dividend Reductions," Financial Management (Winter), 31-51.

Jensen, M.C., 1986, "Agency Costs of Free Cash Flow, Corporate Finance and Takeovers," American Economic Review (May), 323-329.

Jensen, M.C. and W.H. Meckling, 1976, "Theory of the Firm: Managerial Behaviour, Agency Costs and Ownership Structure," Journal of Financial Economics (October), 305 -360.

John, K. and J. Williams, 1985, "Dividend, Dilution and Taxes: A Signaling Equilibrium," Journal of Finance (September), 1053-1070.

Kothari, S.P. and J.B. Warner, 1997, "Measuring Long-Horizon Security Price Performance," Journal of Financial Economics (March), 301-340.

Lakonishok, J. and B. Lev, 1987, "Stock Splits and Stock Dividends: Why, Who and When," Journal of Finance (September), 913-932.

Lang, L.H.P., R.M. Stulz, and R.A. Walkling, 1991, "A Test of the Free Cash Flow Hypothesis: The Case of Bidder Returns," Journal of Financial Economics (October), 315-335.

Lasfer, M.A., 1996, "Taxes and Dividends : The UK Evidence," Journal of Banking and Finance (April), 455-472.

Lasfer, M.A., 1997, "Scrip Dividends: The Management's View," European Financial Management (forthcoming).

Lehn, K. and A. Poulsen, 1989, "Free Cash Flow and Stockholder Gain in Going Private Transactions," Journal of Finance (July) 771-787.

McNichols, M. and A. Dravid, 1990, "Stock Dividends, Stock Splits, and Signaling," Journal of Finance (July), 857-879.

Miller, M.H. and K. Rock, 1985, "Dividend Policy Under Asymmetric Information," Journal of Finance (September), 1031-51.

Myers, S., 1977, "Determinants of Corporate Borrowing," Journal of Financial Economics (November), 147-175.

Opler, T. and S. Titman, 1993, "The Determinants of Leveraged Buyout Activity: Free Cash Flow vs. Financial Distress," Journal of Finance (December), 1985-1999.

Papaioannou, G.J. and C.M. Savarese, 1994, "Corporate Dividend Policy Response to the Tax Reform Act of 1986," Financial Management (Spring), 56-64.

Peterson, C.A, J.A. Millar, and J.N. Rimbey, 1996, "The Economic Consequences of Accounting for Stock Splits and Large Stock Dividends," The Accounting Review (April), 241-253.

Ravid, S.A. and O.H. Sarig, 1991, "Financial Signaling by Committing to Cash Outflows," Journal of Finance and Quantitative Analysis (June), 165-180.

Rozeff, M.S., 1982, "Growth, Beta, and Agency Costs as Determinants of Dividend Payout Ratios," Journal of Financial Research (Fall), 249-260.

Shleifer, A. and R.W. Vishny, 1986, "Large Shareholders and Corporate Control," Journal of Political Economy (June), 461-488.

Scholes, M.S. and M.A. Wolfson, 1989, "Decentralized Investment Banking: The Case of Discount Dividend-Reinvestment and Stock-Purchase Plans," Journal of Financial Economics (September), 7-35.

Smith, C.W. and R.L. Watts, 1992, "The Investment Opportunity Set and Corporate Financing, Dividend and Compensation Policies," Journal of Financial Economics (December), 263-292.

Stock Exchange, 1991, "The Current State of Share Ownership," Stock Exchange Quarterly (Summer).

Strickland, D., K.W. Wiles, and M. Zenner, 1996, "A Requiem for the USA: Is Small Shareholder Monitoring Effective?" Journal of Financial Economics (February), 319-338.

Stulz, R.M., 1990, "Managerial Discretion and Optimal Financing Policies," Journal of Financial Economics (July), 3-27.

Wilson, B., 1982, "Reinvesting Dividends: A Way to Revitalise your Company," Financial Executive (December), 38-44.

Yoon, P.S. and L.T. Starks, 1995, "Signaling, Investment Opportunities and Dividend Announcement," The Review of Financial Studies (Winter), 995-1018.

I am grateful for useful comments to Gordon Gemmill, Mario Levis, S. P. Kothari, Clifford Smith, Meir Statman, Sudi Sudarsanam, Dylan Thomas, two anonymous referees, and to the participants of the 1995 European Financial Management Association Meeting at London, the 1996 European Finance Association Meeting at Oslo, the 1996 Financial Management Association Meeting at New Orleans, and those in Finance Seminars at City University Business School, London, University College, Dublin, Rutgers University, Newark, Neuchatel University, Switzerland, and London Business School. All errors are my responsibility.

M. Ameziane Lasfer is a Lecturer in Finance at the City University Business School, London, England.
COPYRIGHT 1997 Financial Management Association
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1997 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Lasfer, M. Ameziane
Publication:Financial Management
Date:Mar 22, 1997
Words:9176
Previous Article:Control versus firm value: the impact of restrictions on foreign share ownership.
Next Article:Investment analysis of offshore concessions in the Netherlands.
Topics:


Related Articles
United is sticking to the scrip.
United Arab Bank approves scrip dividend of 25%.

Terms of use | Privacy policy | Copyright © 2022 Farlex, Inc. | Feedback | For webmasters |