SCHOOL OF FINANCE AND ECONOMICS

SCHOOL OF FINANCE AND ECONOMICS UTS:BUSINESS WORKING PAPER NO. 16 APRIL, 1992 Australian Tax Changes and Dividend Reinvestment Announcement Effects...
0 downloads 3 Views 344KB Size
SCHOOL OF FINANCE AND ECONOMICS UTS:BUSINESS

WORKING PAPER NO. 16

APRIL, 1992

Australian Tax Changes and Dividend Reinvestment Announcement Effects: A Pre- and Post-Imputation Study

Keith Chan Damien McColough Michael Skully

ISSN: 1036-7373 http://www.business.uts.edu.au/finance/

Australian Tax Changes and Dividend Reinvestment Announcement Effects: A Pre- and Post-Imputation Study

By Keith K.W. Chan † Damien W. McColough ‡ Michael T. Skully ‡

† School of Finance and Economics, University of Technology, Sydney. ‡ School of Banking and Finance, University of New South Wales, Kensington.

Abstract This paper used an event study approach to examine the impact of dividend reinvestment plans on shareholders returns in the pre- and post-imputation environment. The daily share return behaviour indicated that the announcement to introduce a DRP was received indifferently by the market prior to the imputation, but was valued positively afterwards. The results support the suggestion that under imputation the optimal dividend policy is to distribute the maximum franked dividend and implement a DRP to retain cashflows.

1. Introduction

The imputation system eliminates the double taxation of dividends and provides most Australian shareholders with a strong reason to prefer companies that pay franked dividends rather than retain profits. Many companies have responded by substantially increasing their dividends and implemented measures such as dividend reinvestment plans (DRPs) to retain cashflows. Participation in the DRP allows shareholders to receive all or part of their dividends in the form of additional shares in the company, usually at a discount to the market price, rather than in cash.

DRPs have become extremely important for Australian listed companies in raising new equity capital since the implementation of the dividend imputation system on 1 July 1987. In the year to 30 June 1991, there were more than 150 DRPs in operation and their share issues accounted for 32.7 per cent ($2.1 billion) of the total equity capital. In the year to 30 June 1987, new capital raised through some 40 DRPs accounted for only 2.5 per cent ($400 million).

As the number of DRPs increased dramatically following the introduction of dividend imputation, one might expect the market reaction to DRP announcements in the pre- and postimputation environment to differ. This study examines whether any such difference exists. The results indicate that the popularity of DRPs over the past decade was indeed warranted.

The paper is organised as follows. An overview of the DRP and its announcement effects are discussed in Section 2. This is followed by an examination of the implications of the dividend imputation system in Section 3. The methodology and the data are detailed in Section

-1-

4. The results are presented in Section 5. A summary of the paper's major results and conclusions are presented in Section 6.

2. Dividend Reinvestment Plan

Dividend reinvestment plans allow shareholders to have dividends on some or all of their shares in a company automatically reinvested in newly issued shares.1 Shares issued under the DRP are usually at a discount, typically 5 to 10%, from the weighted average market price of the shares traded on the Australian Stock Exchange, typically during the five days after the ex dividend date. Shareholders are not charged brokerage fees, commission or stamp duty for any allotment of shares under the DRP. Where the issue formula results in a fraction of a share, the entitlement is usually rounded up to the next whole share. A transaction statement and a certificate for the new shares are usually forwarded to participants at each dividend payment. Shareholders can vary their participation or withdraw from the DRP at any time.

In the United States, many corporations have long allowed shareholders to automatically reinvest their dividends, but it was not until 26 March 1982 that an Australian company, the Lend Lease Corporation, offered this opportunity to its shareholders with a 10% discount from market price. Similar plans were soon introduced by other companies and by the end of June 1987, there were 38 DRPs in operation.

1

.

One innovation which has not yet been seen in Australia, though common in USA, is an open market plan where shares are purchased from existing shareholders on the open market for reinvestment.

-2-

The primary attraction of the dividend reinvestment plan is to help shareholders reinvest their dividends without incurring transaction costs. Participants also benefit from purchasing new shares at a discount to the market price and the convenience of compound dividends. The ease of the DRP transactions is particularly attractive to small investors. Reinvestment is made on a regular basis and share records are produced periodically.

From a corporate viewpoint, dividend reinvestment provides an inexpensive source of additional equity capital and an effective means of increasing dividend payout without committing cashflows. A company usually knows with some precision how much funds will return to the company. Indeed, Fredman and Nichols (1982) found that DRPs provide a substantial flow of working capital, irrespective of market conditions. Companies also benefit from improved shareholder relations: Anderson (1986) found that this was the main reason for Australian companies introducing DRPs.

Dividend reinvestment plans do have some drawbacks. For example, all shareholders pay for the DRP's implementation and administrative costs but only the participants receive the benefit; thus non-participants are subsidising them. Another problem is the possible dilution of the earnings per share (EPS) caused by an expanding equity base. When a discount is offered, this dilution potential is magnified. In fact, a number of companies have recently reduced their discount (e.g. ANZ Bank) or even suspended their DRPs (e.g. Brambles) to reduce the dilution in EPS. DRP cash flows could also create a surplus of unneeded funds where companies have insufficient profitable investment opportunities or already have considerable funds on hand.

-3-

In a perfect market, the introduction of a DRP should have no impact on firm value. However, market imperfections such as transaction costs, taxation and information asymmetry might cause these plans to have value. Anderson (1986) found that the shareholders' potential savings of transaction costs were an important factor in Australian companies deciding to introduce DRPs. The announcement of a firm's intention to introduce a DRP might be construed to have an information effect as the plan will result in new shares being issued and hence the size of the firm's total dollar amount of dividends increased. As with Miller and Modigliani (1961), the willingness to service additional dividend payments may signal that management perceives good earning prospects for at least the immediate future, or that the company has sufficient positive net present value projects available to employ the additional equity, so there will be no EPS dilution. Thus, Dubofsky and Bierman (1988, p.61) contended that a DRP may lead to positive abnormal returns for shareholders if the firm "is signalling favourable information via the decision to utilise the internal financing characteristics of the DRP and/or is reducing the total brokerage fees and transaction costs normally incurred by the firm and its owners."

Thus far, most empirical work on DRPs has concentrated on the impact of a firm announcing its intention to introduce a DRP. They have been conducted on the U.S. data in the form of event studies and use residual analysis to identify abnormal returns associated with the announcement. Perumpral (1983) found that the market reacted positively when a firm announced its intention to introduce a DRP, but that not all announcements produced positive returns. Peterson, Peterson and Moore (1987) examined the announcement effects on utility company DRPs prior to and during a special $750 U.S. tax exemption on DRP reinvestment in public utility shares between 1983 and 1985. They found that DRP announcements had little

-4-

market impact before the reform, but were valued positively during the tax exemption period. Dubofsky and Bierman (1988) examined the market reaction to firms announcing DRPs with a discount from market price feature. Using daily data they found that DRPs were valued positively. These various results suggest that DRPs offer shareholders real benefits.

3. Implications of the Dividend Imputation

The implementation of the dividend imputation tax system from 1 July 1987 provided a catalyst for the growth of DRPs in Australia. Prior to imputation, a company would pay corporate tax on its earnings and then individual shareholders would pay personal tax again if these earnings were paid out in dividends. Imputation, though, now allows Australian resident shareholders to receive a credit for the full amount of income tax paid by the company. These dividends paid out of a company's after-tax profit are referred to as franked dividends. The imputation credit, known as a franking credit, can then be used to offset the shareholders' own tax liabilities.

The benefit imputation affords an Australian resident shareholder depends largely on the shareholder's marginal tax rate and the company's dividend policy. As shown in the appendix 1, fully taxed company profit distributed as dividends is effectively taxed only once at the shareholder's marginal tax rate. Investors with personal marginal tax rates less than the company tax rate (39% in 1991-92) will prefer a 100 per cent dividend payout as they can apply the excess credits against any tax payable on other income. In contrast, in the absence of a capital gains tax, investors with higher tax rates will prefer the company to retain the profits.

-5-

Unfortunately, inflation adjusted capital gains are taxable unless the shares were purchased before 20 September 1985. Therefore, while imputation has removed the previous double taxation of dividends, the capital gains tax results in the double taxation of retained profits. This adds to the incentive to pay dividends.

Also from 1 July 1988 superannuation funds, which were previously tax exempt, have been subject to an income tax at the rate of 15 per cent and are able to use imputation credits to offset this tax. This has increased their demand for franked dividends.

The demand for franking credits, particularly from small investors and superannuation funds, have caused many companies to increase their dividend payout ratios. Nicol (1991) found that the dividend payout ratios of Australian listed companies able to pay franked dividends have increased substantially post-imputation.

For a company whose shareholders can use franking credits the incidence of tax on equity is reduced whereas the tax treatment of debt remains unchanged as it was under the classical system. Consequently, there is now an incentive for such companies to use more equity finance. Nicol (1991) suggests that the ideal situation for a listed company post-imputation is to pay dividends to the limit of its franking credits while maintaining its investment activities. If new equity needs to be raised as a consequence of this policy, then a company should introduce a DRP to allow shareholders to reinvest the extra dividends in the company, or even to have a DRP underwritten against a minimum level of participation.

-6-

4. Data and Methodology

Theory suggests that DRPs may affect firm valuations in an imperfect world. Skully (1982) and Fletcher (1985) argued that as DRPs become more popular one might expect firms with DRPs to become more desirable and hence valued higher. This study used an event study approach to examine the announcement effects of Australian DRPs. It represents a significant improvement on the previous empirical work in that the DRP announcements were carefully screened against contamination from other price sensitive information. While using only clean announcement dates imposes some bias, Brickley (1986), when examining a similar sample bias in proxy statements, concluded that the clean filtering had no significant impact upon event study conclusions.

Through published sources, 188 companies were found to have or have had DRPs or had made changes to them. From an examination of the Australian Stock Exchange files, 30 firms were found to have unclear announcement dates, and the files of 4 firms were unavailable. This sample of 154 firms was further reduced by data limitations; the School of Banking and Finance, UNSW's Statex Data Base coverage (3 January 1984 - 17 February 1989) precluded daily price data on 29 of these companies. Finally, an examination of the DRP announcements themselves or other announcements near that of the DRP meant 41 more events were dropped from the study. Thus 84 DRP announcements remained, with 32 DRPs announced prior to imputation and 52 announcements post-imputation.

This study used the market model as the return generating process. The market model infers a linear relationship between returns on a security and the returns on the market:

-7-

Rit = αi + βiRmt + εit

(1)

where Rit is the return of security i in period t, Rmt is the return of an equally weighted accumulation index in period t, αi is the intercept coefficient, βi is the slope coefficient and εit is the disturbance term. Brown and Warner (1980, 1985) found that event studies based on a simple market model are as powerful in detecting abnormal returns as other complex methodologies, and Shevlin (1981) confirmed this within an Australian study.

To correct for the bias arising from non-synchronous trading in the Australian market, Dimson's (1979) aggregated coefficients method was used. This requires a multiple regression of security returns against lagged, matching and leading market returns: Rit = αi +

n

Σ 1ßik Rm t+k + εit

(2)

k =- n

A consistent estimate of beta is obtained by aggregating the slope coefficients from the regression: ßi =

n

Σ 2ßik

(3)

k =- n

Following the examination of this problem with Australian data by Sinclair (1981), one lag was used in this study. In addition, the sample data were also corrected for heteroscedasticity and autocorrelation.

For each DRP announcement, the interval of 100 to 6 days before the announcement day (the estimation period) was used to estimate the coefficients αi and ßi in (1). The period of 5 days before and 5 days after the announcement day was chosen as the event period for

-8-

estimating abnormal returns. The abnormal return, or the residual, for each security was obtained by subtracting the expected return from the ex post actual return.

As a company's risk level may change due to the event being studied, an abnormal return might simply reflect an appropriate compensation for the new risk level. Thus switching regression techniques were used to test whether risk levels changed around DRP announcements. The tests did not reveal any such changes in during either the event or estimation period.

To test the null hypothesis that the average abnormal rates of return are positive, this study followed the work of Dodd and Warner (1983), Dubofsky and Bierman (1988) and employed the z-statistic. This meant that individual abnormal returns were first standardised before they were aggregated to mitigate the effects of outliers.

5. Results

5.1 Overall Results

Table 1 presents the estimates of the average abnormal daily rates of return for the 88 DRP announcements in the sample period of 3 January 1984 to 17 February 1989. The results are consistent with U.S. announcement effect studies [Perumpral (1983), Peterson, Peterson and Moore (1987), Dubofsky and Bierman (1988)]. The daily share return behaviour around the announcement date indicates that dividend reinvestment plans are valued positively by the

-9-

market. There was a statistically significant increase in abnormal returns on the event day. However, the positive reaction was by no means universal across companies with only 53.3 per cent of the residuals being positive on the day of the event.

There was no indication that the market had anticipated the DRP announcements. As shown in Table 2, the z-value for the 5-day interval before the announcement [-5, -1] was statistically insignificant. The market also appeared to adjust quickly to the announcement with average residuals dropped on the days after the event and the z-value for the 5-day interval after the event [+1, +5] proved to be insignificant.

5.2 Pre-Imputation Results

Estimates of the average abnormal returns for the pre-imputation period (January 1984 June 1987) are shown in Table 3: it covers a sample of 32 announcements. The results show no significant abnormal returns on the day of the DRP announcement. This pre-imputation indifference to DRPs was highlighted by the fact that only 38.9% of the residuals were positive on the announcement day. Also, the standard deviation of the cross-sectional residuals was very small on the event day, thus indicating that this indifferent response was rather uniform across firms, though not unanimous. Table 2 shows that there was little anticipation of the DRP announcements nor any significant market reaction after the announcements.

One explanation for this lack of response is that DRPs were then still a new phenomenon to Australian investors. The advantages of dividend reinvestment may not have been fully

- 10 -

appreciated. Also, companies initially introduced these plans to enhance shareholder goodwill rather than to increase firm value.

5.3 Post-Imputation Results

Table 4 presents the results of the residual analysis performed on a sample of 52 DRP announcements post-imputation. These results were markedly different to those of the preimputation period. The average abnormal return on day 0 was significantly positive. Also, on this day more residuals were positive (53.8%) than on any other day in the event period. The standard deviation on day 0 was similar to those across the event period, thus suggesting that the strong positive response was quite uniform across the sample.

There were some anticipation of the announcements of DRPs as evidenced by the significant z-value for the [-5, -1] interval in Table 2. The anticipation may have stemmed from shareholders' demand for greater dividend payout thus putting pressure on companies to introduce the plans. The market however appeared to adjust quickly to the announcements.

6. Conclusion

This paper examined the impact of dividend reinvestment plans on shareholders returns in the pre- and post-imputation environment. The daily share return behaviour indicated that the announcement to introduce a DRP was received indifferently by the market prior to imputation, but was valued positively after its introduction. The marked increase in the number of DRPs

- 11 -

following imputation was justified. DRPs allow companies to satisfy shareholders' demand for greater dividends while retaining sufficient funds for investment purposes. The results support the suggestion that under imputation the optimal dividend policy is to distribute the maximum franked dividend and implement a DRP, possibly underwritten, to retain cashflows.

The strong positive response post-imputation however was not uniform across firms with only half of abnormal returns on the announcement day being positive, implying that the reaction to the DRP may be interlinked with specific firm characteristics as well as the DRP itself.

In conclusion, dividend reinvestment plans have served a useful function for Australian companies and shareholders and as such are valued positively in the market. This has become even more so following the implementation of the dividend imputation system.

- 12 -

Table 1

Market Model Results of DRP announcements: January 1984 - February 1989

Cumulative

Days

Average

Positive

Residuals

Residuals

Standard

(%)

(%)

Deviation

Average Residuals Z-Value

(%)

-5

-0.445

52.2

0.024064

1.015

-0.445

-4

-0.097

43.3

0.031047

0.316

-0.542

-3

-0.106

44.8

0.029068

1.737

-0.648

-2

0.367

44.8

0.030084

2.009(1)

-0.281

-1

-0.058

43.3

0.028211

0.116

-0.339

0

0.549

53.3

0.029798

2.744(2)

0.210

+1

0.341

50.7

0.013865

0.308

0.551

+2

0.126

46.3

0.027801

0.102

0.677

+3

0.099

43.3

0.082327

1.850

0.776

+4

0.193

43.3

0.035405

0.887

0.969

+5

-0.934

46.3

0.028876

0.454

0.035

Notes: 1. Significant at the 1% level using a 2-tailed test 2. Significant at the 5% level using a 2-tailed test

- 13 -

Table 2 Z-value for Abnormal Returns over Time Interval _________________________________________________ Sample

[-5, -1] [0]

[+1, +5]

_________________________________________________ 1984-1989 Pre-imputation -0.996 Post-imputation

2.7442

0.537 -1.210 1

0.822 2

4.655

2.269

-0.232 -1.420

_________________________________________________ Notes: 1. Significant at the 1% level using a 2-tailed test 2. Significant at the 5% level using a 2-tailed test

- 14 -

Table 3 Market Model Results of DRP announcements: Pre-Imputation

Cumulative

Days

Average

Positive

Residuals

Residuals

Standard

(%)

(%)

Deviation

Average Residuals Z-Value

(%)

-5

0.273

47.2

0.015636

0.255

0.273

-4

-0.117

47.2

0.024194

0.366

0.156

-3

0.129

58.3

0.026687

1.308

0.284

-2

0.184

41.7

0.013224

-0.634

0.468

-1

0.333

41.7

0.046478

-3.523(1)

0.801

0

-0.221

38.9

0.015223

-1.210

0.580

+1

-0.471

47.2

0.014965

0.507

0.109

+2

-0.305

41.7

0.019103

-0.399

-0.196

+3

0.253

47.2

0.025968

0.762

0.057

+4

-0.422

41.7

0.014174

0.434

-0.365

+5

0.074

52.8

0.044811

0.535

-0.291

Notes: 1. Significant at the 5% level using a 2-tailed test

- 15 -

Table 4 Market Model Results of DRP announcements: Post-Imputation

Cumulative

Days

Average

Positive

Residuals

Residuals

Standard

(%)

(%)

Deviation

Average Residuals Z-Value

(%)

-5

-0.916

48.1

0.027003

1.335

-0.916

-4

0.847

36.5

0.032887

0.301

-0.069

-3

-0.120

40.4

0.027911

1.564

-0.190

-2

0.404

38.5

0.037104

-2.315(1)

0.215

-1

-0.071

40.4

0.031825

0.019

0.144

0

1.040

53.8

0.034413

4.655(1)

1.184

+1

0.388

50.0

0.022160

-0.267

1.572

+2

0.573

42.3

0.029700

0.216

2.145

+3

-0.083

38.5

0.091540

-1.674

2.062

+4

0.312

38.5

0.040994

-0.247

2.374

+5

-1.068

34.6

0.033232

-1.204

1.306

Notes: 1. Significant at the 5% level using a 2-tailed test

- 16 -

References Anderson, R.H., 1986, "Dividend reinvestment and Australian companies", Professional Administrator, June/July, 46-47. ANZ McCaughan, 1989, ANZ McCaughan Dividend Guide, June. Australian Stock Exchange, 1991, "Survey of DRP companies", an unpublished Australian Stock Exchange draft paper. Brickley, J.A., 1986, "Interpreting common stock returns around proxy statement disclosures and annual shareholder meetings", Journal of Financial and Quantitative Analysis, 21(3), 343-349. Brown, S.J. and J.B. Warner, 1980, "Measuring security price performance", Journal of Financial Economics, 8, 205-258. Brown, S.J. and J.B. Warner, 1985, "Using daily stock returns: the case of event studies", Journal of Financial Economics, 14, 3-31. Davies, C.S., 1991, "Dividend reinvestment plans in Australia with specific reference to current developments", an unpublished research report, School of Banking and Finance, UNSW, 7 November. Dimson, E., 1979, "Risk measurement when shares are subject to infrequent trading", Journal of Financial Economics, 7, 197-226. Dodd, P. and J.B. Warner, 1983, "On corporate Governance: A study of proxy contests", Journal of Financial Economics, 11, 401-438.

- 18 -

Dubofsky, D.A. and L. Bierman, 1988, "The effect of discount dividend reinvestment plan announcements on equity value", Akron Business and Economic Review, 19 (2) Summer, 58-68. Fletcher, P., 1985, "Clever thinking through the new tax trap", Rydges, August, 52. Fredman, A.J. and J.R. Nichols, 1982, "Sizing up new capital dividend reinvestment plans", California Management Review, 24(4), 77-84. McColough, D., 1991, "Dividend Reinvestment Plans in Australia: an event study," unpublished honours thesis, University of New South Wales. Miller, M. and F. Modigliani, 1961, "Dividend policy, growth, and the valuation of shares", Journal of Business, October, 411-433. Miller, M. and K. Rock, 1985, "Dividend policy under asymmetric information", Journal of Finance, September, 1031-51. Nicol, R.E.G., 1991, "The dividend puzzle: an Australian solution?", Paper presented at the Fourth Finance and Banking Conference, 28-29 November, School of Banking and Finance, University of New South Wales, Kensington. Perumpral, S.E., 1983, "An empirical analysis of the effect of automatic dividend reinvestment plans on securities returns", an unpublished PhD thesis, Virginia Polytechnic Institute. Peterson, P.P., Peterson, D.R. and Moore, N.H., 1987, "The adoption of new issue dividend reinvestment plans and shareholder wealth", The Financial Review, 22 (2), 221-232. Shevlin, T.J., 1981, "Measuring abnormal performance on the Australian securities market", Australian Journal of Management, VI (2), 67-107.

- 19 -

Sinclair, N.A., 1981, "An empirical examination of the required number of leading and lagged variables for ACM Beta estimation", Australian Journal of Management, VI (2), 121-126. Skully, M.T., 1982, Dividend Reinvestment Plans: their development and operation in Australia and the United States, Melbourne; Committee for the Economic Development of Australia.

- 20 -

Appendix The impact of fully franked dividends on differing personal marginal tax rates

SHAREHOLDERS

ASSUMED

TAX PAID

DIVIDEND

GROSSED UP

NOTIONAL TAX

MARGINAL

GROSS

BY THE

PAID

DIVIDEND

PAYABLE BY

TAX RATE

PROFIT

COMPANY

SHAREHOLDER ON GROSSED UP DIVIDEND

21.0

$100

$39.00

$61.00

$100.00

$21.00

25.0

$100

$39.00

$61.00

$100.00

$25.00

30.0

$100

$39.00

$61.00

$100.00

$30.00

38.5

$100

$39.00

$61.00

$100.00

$38.50

42.5

$100

$39.00

$61.00

$100.00

$42.50

46.5

$100

$39.00

$61.00

$100.00

$46.50

47.0

$100

$39.00

$61.00

$100.00

$47.00

Note: The above does not reflect the impact of the 1.25% Medicare levy. The tax credit earned through dividend imputation can be utilised against other income in the year it is earned. There is no opportunity for a refund.

- 21 -