Dividend Imputation Tax Credits

Transcription

1 How Investors and Management Boards Value Dividend Imputation Tax Credits Portfolio Submitted as Partial Fulfillment of the Requirements for the Degree of Doctor of Business Administration Chih-ming Wang Submitted on: International Graduate School of Management University of South Australia

2 Doctor of Business Administration Portfolio Submission Form Name: Chih-ming WANG Student ID No: w Dear Sir/ Madam, To the best of my knowledge, the portfolio contains all of the candidate's own work completed under my supervision, and is worthy of examination. I have approved for submission the portfolio that is being submitted for examination. Signed: Supervisors Dr Arthur Preston and John Rice Date: Date:

3 Supported by: Prof M Chai anagement Committee Date

4 Declaration I declare that this portfolio does not incorporate without acknowledgement any material previously submitted for a degree or diploma in any university; and that to the best of knowledge it does not contain any materials previously published or written by another person except where due reference is made in the text. Signed: Date: Chih-ming WANG II

5 Acknowledgement In writing these papers, I would like to express my sincere appreciation to the tutors of the DBA program, especially my supervisors, Dr Arthur Preston and John Rice. for their guidance, advice and proofreading. Also, I wish thank Chung-sheng Wang for his help in collecting and processing the data. Special gratitude should be given to Ms Marilin Chou and Nancy Chen and their APML colleague's administrative assistance. Though I should take all the responsibility for any error or misleading of the portfolio. Chih-ming WANG V Date: "A I /14)

6 How Investors and Management Boards Value Dividend Imputation Tax Credits Contents Introduction: Portfolio overview Paper one: The imputation tax system: does dividend imputation tax credit matter to share value? Paper two: The ex-dividend days share behaviour of Taiwan's share prices pre and post imputation Paper three: The effect of imputation tax system on payout ratios: a comparison between Australian and Taiwanese companies Appendix 1: Conference acceptance letter (Paper 3) iv

7 Introduction 1. Portfolio Overview This portfolio consists of three linked papers all exploring issues of imputation tax credit. The three papers are: I. The imputation tax system.- does dividend imputation credit matter to share value? The ex-dividend day behaviour of Taiwan's share prices pre and post imputation The effect of imputation tax system on payout ratios: a comparison between Australian and Taiwanese companies A full imputation tax system has been operation in Taiwan since Because imputation tax credits are completely refundable to resident investors in the income tax returns, they should be valued equally to dividends. However, under Taiwanese taxation law, the determination of imputation tax credit is based on the date of record (as depicted in Figure 1) for dividend payment, and can vary and be speculative for investors: therefore they do not know their real total dividend income on ex-dividend day. Paper 1, theoretically explores whether share prices are affected by the amount of imputation tax credit, which is discretionary, and hence variable, and not completely disclosed at ex-dividend day. It is concluded that since dividend imputation credit is as valuable to investors as dividend received, the corporation should announce it at the same time as the declaration of dividend (see Figure 1). Once it is declared, it should not be varied or revoked for any reason.

8 Figure 1: Dividend days Date of declaration Date of record Date of payment Ex-dividend day The aim in Paper 2 was to ascertain whether share prices in Taiwan value imputation tax credits. Theoretically, the drop-off of share prices on ex-dividend day should be the after-tax value of dividend. Using data of listed companies in Taiwan from the three years before and after the introduction of imputation, ex-dividend day share price behaviours were analysed. It was commonly found that share prices generally fall less than the full amount of the dividend when they go ex-dividend, presumably because of tax considerations of investors. Although the average drop-off ratio after imputation of 0.82 was higher than the pre-imputation value of 0.67, the increase was not statistically significant, and hence it was concluded that investors do not value imputation credit the same as dividends. The analysis of the relationship between the imputation credit ratios and drop-off ratios, confirms this finding; which is contrary to that anticipated. It seems that, to some extent, this result is due to a taxation policy that does not require corporations to announce imputation tax credit amount at the same time as the declaration of dividends as described in paper 1. Paper 3 compares the important variations of the imputation tax system between Taiwan and Australia and investigates if these variations affect the dividend policies of listed companies in each country. It has been suggested that under an imputation tax system, the optimal dividend policy for companies is to pay out sufficient dividends to utilize all of their franking credit so as to maximize the returns to shareholders. By empirically studying Taiwanese listed companies' payout ratios three years before and after imputation, it was found that there was no statistically significant increase of payout ratios after the introduction of imputation arrangements and no relationship between imputation credit ratios and payout ratios. It was also vi

9 found that the increase of average payout ratios in Taiwan after imputation was lower relatively to that in Australia. All of these findings are contrary to that anticipated. This may due to the fact that management has not changed their perceptions of tax policy after the introduction of imputation credits and are reluctant to modify the dividend policy in favour of their shareholders. The relationship between the papers is depicted in Figure 2. Figure 2: The relationship between the three papers in the portfolio. A generally accepted view is that dividend (B) does not affect share value(c) although this may not necessarily be the case. The first two papers explore whether imputation tax credit (D) affects share prices (C). Paper 3 is a study of whether imputation tax credit (D) affects dividend policy (B). 2. Contribution of the portfolio The primary contribution made is to the area of imputation tax policy. Under an imputation tax system, the imputation tax credit is additional to dividend due to its refundable nature. Only the amount of imputation credit being attached to a dividend, will be valuable to shareholders. Therefore, an imputation tax policy should require a company to announce the amount of imputation credit per share as the declaration of dividend. Once it is declared, it should not be varied or revoked for any reason. Although it is claimed that income tax is capitalized in the share value as in the vii

10 classical tax system; it is not clear the extent that share value will be capitalized by the amount of imputation credits which are themselves uncertain in quantum. Also unclear, is what amount of imputation credits will be realized because this is dependent on dividend policy. Under an imputation tax system, it is widely suggested that the optimal dividend policy for companies is to pay out sufficient dividends to utilize all of their franking credit. Contrary to this expectation, the findings were that management seems reluctant to modify the dividend policy in favour of their shareholders after imputation. Although it might seem that management in this case is not acting in the best interests of their shareholders, the decision of dividend policy is always a complex issue in practice.

11 The imputation tax system does dividend imputation credit matter to share value? Chih-Ming Wane Abstract A full imputation tax system has been operation in Taiwan since Because imputation tax credits are refundable to resident investors in the income tax returns, they should be valued equally to dividends. However, under Taiwanese taxation laws, the determination of imputation tax credit is based on the date of record for dividend payment, and can vary and be speculative for investors; therefore they do not know their real total dividend income on ex-dividend day. This paper explores whether share prices are affected by the amount of imputation tax credit, which is discretionary, and hence variable, and not completely disclosed at ex-dividend day. It is concluded that since dividend imputation credit is as valuable to investors as dividend received, the corporation should announce it at the same time as the declaration of dividend. Once it is declared, it should not be varied or revoked for any reason. Key words: imputation tax credit; share price; drop-off ratio DBA candidate, International Graduate School of Management. University of South Australia. Lecturer, Department of Finance, National Sun Yat-Sen University, Taiwan. Address: 3F-1, Ming-sheng 2nd road. Kaohsiung city. Taiwan. upai bms34.hinet.net. Phone: (07) The author is indebted to the research assistance of Dr Arthur Preston and John Rice. University of South Australia.

12 The imputation tax system does dividend imputation credit matter to share value? 1. Introduction Following the historic democratic election in Taiwan in 1998, the new government chose to mitigate the double taxation of distributed corporate earnings by operating a full imputation tax system. Under the new tax regime, firms were required to attach an imputation credit, which was derived by the income tax paid by the firm and limited by company tax rate, to dividends it distributed. For resident individual shareholders, the imputation credit is grossed-up as taxable dividend income. When the marginal tax rate of taxpayers is lower than the company tax rate, this imputation credit can be used as a rebate for tax payable, with the excess credit reducing the tax paid in respect of other sources of taxable income. Any excess credit is refundable where it exceeds the tax on other income. Due to the refundable nature of dividend imputation credit, it is never -wasted'''. On this basis, it seems that all residential investors in Taiwan should fully value the imputation credit as a part of dividend. Since imputation credit is virtually part of the dividend distributed, it would proper to disclose it at the same time as dividends are announced. However, current Taiwanese taxation law does not require the firms to declare the amount of the tax credit along with the dividend at the time of the resolution of shareholders or directors. Firms only disclose the estimated tax credit amount in their annual financial statements. At the time of payment, a firm must give the shareholders a statement setting out the details of the dividend and its tax credit. At the beginning of next year, shareholders will receive a divided statement including the final figures of tax credit for use in their income tax returns. According to the taxation law in Taiwan, the firm even must correct the final figures of tax credit in case of any errors. This means that imputation credit is always variable and tentative in Taiwan. In capital markets, except for the bottom line (net income) of a firm affecting share prices, it is generally accepted that dividends are probably one of the most important information requirements of investors. Under the imputation tax system enacted in Taiwan, dividends should conceptually include the amount received and the

13 imputation credit attached. If investors do not know the final imputation credit before ex-dividend day, it actually means that they do not know the real total amount of dividend to be paid. In turn, this will influence the ex-dividend day share price behaviour. Traditionally, the rate of tax on dividends exceeds that of capital gains. Therefore, if shares are valued on a net-of-tax basis, then on ex-dividend day, share prices are generally fallen by less than the amount of the dividends due to the tax differential of dividends and capital gains. Evidence supporting such a tax effect includes Elton and Gruber (1970), and Kalay (1982), among others. As a consequence, under the Taiwanese imputation tax system, share prices at ex-dividend day may reflect only the dividend distributed, not including the imputation tax credit. Given this set of circumstances, the following issues are problematic: 1. How should Taiwanese investors value the dividend imputation tax credit for the lack of the information of final imputation tax credit? What would the imputation credits and dividends together affect share prices at ex-dividend day??. In the process of the formation of dividend policy, what are the perspectives of the corporation on the imputation tax credit? How does imputation tax credit affect dividend policy? 3. If investors not fully value the imputation tax credit equivalent to dividends, what are the implications to tax policy makers about the variability and disclosure of it during the dividend announcement? The issues involved in understanding how imputation tax credit affects dividend policy will be addressed in another paper by the author (Wang, 2002a). The purposes of this paper are to review theoretical and empirical research about the interaction among share prices, dividends, taxation and imputation credits, and to explore if dividend imputation credits were relevant to a firm's value. As well, the adequacy of current taxation law is discussed with respect to whether the declaration of attached imputation credit should be made simultaneous with the dividend announcement, so that all of the relevant information is disclosed to the market. To achieve these purposes, the remainder of this paper is structured as follows. Section 2 describes the imputation tax system in Taiwan and highlights the differences in imputation systems operating in other countries. Section 3 reviews the research about share prices, dividends and taxation in classical tax system. Section 4 examines if dividend imputation credits matter to share prices. Section 5 discusses the implications to share market and tax policy, and describes the opportunity for further research. Conclusions are drawn in section 6. 3

14 2. Imputation tax system and imputation credit 2.1 Analysis of the imputation tax system in Taiwan The marginal company income tax rate in Taiwan before and after imputation tax system is 25%. Under imputation tax system, an important policy was launched that any undistributed fiscal net income by the end of the next year attracts an additional 10% company income tax2. From the government's perspective, this has the effect of encouraging companies to reduce retained earnings and to pay higher dividends to shareholders so that shareholders will make little attempt to gain the tax savings by avoiding high personal income tax rates. Under the current tax regime, any income taxes paid by the firm, including the extra 10% undistributed income tax, will be credited to an Imputation Tax Credit Account (ICA), and can be accumulated as imputation credit for any dividend distributed (the ICA will be debited at that time) in the future. On receiving the dividend, resident individual shareholder's assessable dividend income is -grossed up- to include the amount of imputation credit. Taxpayers are then entitled to a rebate of tax equal to the amount of imputation credit in respect of that dividend. When the marginal tax rate of taxpayers is lower than the company tax rate, the excess credit can reduce the tax paid in respect of other sources of taxable income. Any excess credit, which exceeds the tax on the other income, is refundable. Due to its refundable nature, the imputation credit in Taiwan is somewhat like the withheld income tax of dividend income. The amount of imputation credit may be expressed as a ratio of dividend received. This is called an imputation credit ratio (C). The ratio, which is subject to an upper limit of 33.33% and /03 depending on whether the additional 10% undistributed income tax was levied, is equal to income tax paid by the enterprise divided by retained earnings; both are based on the date of record (see Figure 1) for dividend payment. In any case, the imputation credit ratio must be between 0% and 48.15% of divided distributed. All dividends declared on the same dividend day must be attached the same imputation credit ratio. A 48.15% imputation credit ratio attached to the dividend paid is equivalent to 100% franked dividend in Australia. According to Taiwanese taxation laws, the imputation tax credit ratio of a firm should be derived based on the balances of the tax paid and retained earnings on the date of record. Therefore, no investors can know the true amount of imputation credit before or during the ex-dividend days. This may induce the time lag and the possibility of revision of the credit ratio after declaration of dividend. It is in doubt if the share prices on ex-dividend day in Taiwan have adequately reflected the true value of imputation credit, which is a part of dividend income of investors. 4

15 2.2 The Taiwanese CPA's position To remedy the time lag of the information of imputation credit, the Taiwanese CPA Association has declared that the firms should disclose the tentative imputation credit ratios in their financial reports. No matter how carefully the tentative imputation credit ratio is calculated, it's still not a final one. According to the taxation law in Taiwan, the firm must correct the final ratio in cases of any mistake found in the future. Nevertheless, due to the complexity of information effects of dividends and dividend imputations, to disclose the tentative imputation credit ratio in the footnotes of financial reports seems not enough to investors in a share market. 2.3 Comparison of the imputation policy among countries To mitigate the double taxation of distributed corporate earnings, some countries4 have chosen to eliminate income tax at the shareholder level. In such countries, the operating imputation tax system will generally contain a memorandum account (ICA, or franking account in Australia) to control the total credit amount that a firm could attach to dividends distributed. Any income tax paid by the company, and dividends received by the company with imputation credit attached, will be credited to the ICA, while debits are entered when dividends are paid with imputation credit attached. There are a number of significant variations in the treatment of imputation credits in Taiwan when tax law in this area is compared to the law of other countries. These are: 1. The extent that firms can pass imputation credit to shareholders : In Australia, if a dividend is paid under a resolution, the company makes a declaration that each dividend covered by the resolution is franked, and specifies the extent to which it is franked (CCH, I996a). The franking percentage of franking credit may be any percentage up to 100%. The company's declaration must be made before the day the dividend is paid. A company in Australia must frank dividends if it has credits in its imputation account. In contrast, at the same dividend announcement, a firm in Taiwan cannot determine the franking level at its discretion because the imputation credit is determined by a formu1a5. In New Zealand, the amount of credits attached to a dividend is solely left to the discretion of the company. There is no obligation to attach imputation credits, even if a credit balance exists in the ICA6. In the UK, when a resident company pays a dividend, it is required to pay advance corporation tax7 (ACT) equal to one third of the amount of the dividend. ACT, up to a maximum of 25% of taxable income, is offset against corporation tax charged on profits for the period in which the dividend is paid. ACT unused in the year can be carried forward 5

16 and used in later years. The corporate tax is levied at the rate of 33% and at a lower rate of 25% for smaller companies. The refundable features of the imputation credit : From a taxpayer viewpoint, when the marginal tax rate of taxpayers is lower than the company tax rate, the excess credit can reduce the tax paid in respect of other sources of taxable income. However, in Australia, any excess credit is not refundable to shareholders and cannot be carried forward where it exceeds the tax on the other income. Therefore, shareholders on low marginal personal tax rates may be unable to use all the imputation credits to which they are entitled (Parmenter and Seyfort, 1987). If the credit cannot be fully used by the shareholder in the particular year, then the benefit of the credit is lost. So, it seems that in Australia, taxpayers with different tax rates would not value equally for the same amount of dividend imputation credit. By comparison, in New Zealand, if there is still an excess of imputation credits left, these can be converted into a tax loss and carried forward to future years8. In spite of the partial imputed features of the company tax in UK, the personal taxpayers are taxed on the grossed amount at their appropriate rate, with a deduction for the 25% treated as already paid. "The associated tax credit cannot be refunded by the tax authorities but corporate investors can use it to frank their own dividend payments or offset it against their previous tax liability" (Lasfer 1996, p.457). Nevertheless, the taxation laws for imputation credit in Australia, New Zealand and UK to some extent are different to those in Taiwan where any excess of the credit is fully refundable. All of these situations would cause to different valuations of imputation credit and share prices in different countries. The declaration of imputation credit : In Taiwan, the imputation credit ratio is determined by the balances of ICA divided by retained earnings, both of which are based on the balances of date of record (see Figure 1). Therefore a company need not declare the imputation credit ratio at the same time as the dividend announcement. The credit ratio must be corrected in the future if a mistake should be found. It seems the equities market in Taiwan does not completely reflect the value of the imputation credit due to its time lag and tentative features. In Australia, a dividend is franked if the company makes a declaration before the date of payment (see Figure 1) that the dividend is franked to a specified extent (CCH, 1996a). Once the franking declaration is made, it cannot be varied or revoked (CCH 1996b, 160 AQF (2)). A company which pays a frankable dividend to a shareholder must give the shareholder, before or at the time of payment, a statement (CCH 1996b, 160 AQH) in the approved form setting out details of the dividend and 6

17 its franked or non-franked status (CCH, 1996a). It seems that the Australian taxation policy has also not recognized the market effects of imputation credit effectively, although the amount of imputation cannot be varied or revoked. Figure 1: Dividend days Date of Date of Date of declaration record payment Ex-dividend day 3. Contemporary views on dividends and share prices effect 3.1 Irrelevance of dividends Since Miller and Modigliani ( 1961 ) established the hypothesis of irrelevance of dividends in perfect capital markets, some empirical studies have found support for this position while many others have argued against it. According to the dividend irrelevance theory, in perfect markets a company's share value will be affected only by how it makes investment and how much the return on the investment is. It is irrelevant if net income is distributed or how much is paid. Contrary to the dividend irrelevance theory, other studies about the payment of dividends have identified dividend relevance along a number of dimensions: (1) taxation and clientele effect (Pettit, 1977), (2) bird-in-the-hand theory (Gordon. 1963), (3) signaling role of dividend (Ross, 1977), and (4) agency theory (Jensen and Meckling, 1976). All of these theories support the notion that dividends are relevant to share value. The next section examines whether imputation tax credits affects share prices from the perspective of tax on dividends. 3.2 Dividend yield and required returns on investment Due to the differential taxation rates in favor of capital gains in the classical income 7

18 tax system, a number of studies have concentrated on the question of whether a taxation effect is reflected in the share market. Because of the tax advantage of capital gains, it is assumed that investors in a classical income tax system generally prefer to have companies retain most of their net income. It follows that, all other things being equal, investors would be willing to pay more for low-payout companies than high-payout companies. The empirical evidence derived till now is mixed. Brennan (1970) proposes that, for covering tax loss of dividend, the higher the dividend yield is, the higher the required rate of return by investors will be. Black and Scholes (1974) found no difference in expected returns between high-dividend-yield and low-dividend-yield, thus supporting a dividend-neutrality hypothesis. By controlling the information effect, Litzenberger and Ramaswamy (1982) found a positive relationship between returns and dividends. Poterba and Summers (1984), also conclude that taxes change equilibrium relationships between dividend yields and market returns. Based on the UK equity market, in which the relative tax rates of dividend income and capital gains are the reverse of those in the US, Morgan and Thomas (1997) supposed that the yield-return relationship in the UK would be the opposite to that observed in the US. On the contrary, they still found a significant positive relation between dividend yields and returns. This is inconsistent with a tax-based explanation. Fama and French (1998) also unexpectedly found a positive relationship between prices and dividends, concluding that if any tax effects exist, the signaling effects of dividends outweigh the tax penalty on dividends. The unexpected positive relationship of Fama and French's portfolios is interpreted by Harris, Hubbard and Kemsley (2001, p. 573) that if future dividend taxes are capitalized in equity values, then share prices absorb the burden of this expected tax.... dividend yields per se should not capture tax effects Dividend tax and share prices effect Taxes are expected to affect the equilibrium share price behaviour on the day when shares are first quoted ex-dividend (the ex-day) (Lasfer, 1996). Early research, by Campbell and Beranek (1955) and Durand and May (1960) documented drop-off ratios that did not differ greatly from one. Elton and Gruber (1970) argued that, in equilibrium, shares must be priced such that the marginal investor is indifferent to trading on the cum-dividend or ex-dividend day. They found a drop-off ratio average of 0.78, which implies a marginal income tax rate of investors of around 35%. By finding out the positive relationship between dividend yield and the drop-off ratio, they also supported the tax-induced clientele hypothesis. The general model adopted by Elton and Gruber is as follows: 8

19 Pb Pa + Tc (Pb Pa)- (1) Where: Pb : the cum-dividend price immediately before the share is quoted ex-dividend; Pa : the price immediately after the share is quoted ex-dividend; D : cash dividend to be paid on the share; T0: the investor's tax rate on dividend income; T: the investor's tax rate on realized capital gains. To find out drop-off ratio, Equation (1) can be rearranged as follows: (Pb Pa ) D (1 To) (1 Tc) (2) It is assumed, for simplicity, that there is no discounting of the future capital gains tax to its present value. Therefore, the left hand side of Equation (2). (Pb Pa ) D, is the drop-off ratio (R) as conventionally defined, and the right hand side of Equation (2) will be the simplest version of the Elton and Gruber (1970) model, where the drop-off ratio equal (1 To) (1 Tc). On ex-dividend day, the benchmarks for the expected drop-ratio should be equal to one (1). Kalay (1982) questioned that predictable clientele trading behavior surrounding ex-dividend day would eliminate the observable tax effect. He argued that any departure of the dividend drop-off ratio from one (1) is bounded by the transaction costs, not taxes. Although the drop-off ratio estimated by Kalay is not significantly different from one, it is related to dividend yield. Mixed empirical evidence was provided by Eades. Hess and Kim (1984). They examined the ex-dividend day behaviour across a number of alternatives taxable and non-taxable forms of distribution. Consistent with the tax clientele hypothesis, they found strong evidence of a distribution premium over the ex-dividend day. But the evidence was less convincing when negotiated commission was introduced in the U.S.A. In Canada, Booth and Johnston (1984) similarly provided evidence of equity pricing to give drop-off ratios significantly less than 1. Long (1978) and Sterk and Vandenberg (1990), among others, document that tax policy does affect the value of cash dividend. Wu (1996) also found that there are structural shifts in the aggregate dividend payout and that these shifts coincide with tax law changes. In UK, Lasfer (1996) showed that firms set their dividend policies to minimize their tax liability and to maximize the after-tax return of their shareholders. In addition, and consistent with the hypotheses 9

20 of tax effect on share prices, Lasfer found that the differential taxation of dividend and capital gains resulted in a decrease in ex-day share prices by significantly less than the amount of the dividend. There was no evidence of a tax-induced dividend clientele. 3.4 Survey results of dividend policy and share value Using a more direct approach, some researchers have surveyed corporate managers and institutional investors to determine their views of dividends and share value. Baker, Farrelly and Edelman (1985), surveyed management views on dividend policy, and suggested that corporate managers typically believe that dividend policy affects a firm's value and also that an optimal level of dividend payout exists. Abrutyn and Turner (1990) conducted surveys of chief executives officers (CEO) of 550 of the biggest corporations in the United States. Of the 163 usable replies, only 42% of CEOs claimed to know who their shareholders were. Because managers cannot know whether dividends will result in larger tax liabilities for shareholders, Abrutyn and Turner inferred that a firm couldn't tailor dividend policy to their shareholders' tax status. Therefore, it is hard to see how the tax laws can play a significant role in the determination of the payout ratio. Baker and Powell (1999) obtained data from a mid-1997 mail survey sent to 603 chief financial officers (CFO) of U.S. firms listed on the New York Stock Exchange. Based on 198 usable responses, the empirical results show that most survey respondents believe that dividend policy affects firm value. But respondents are unsure about whether investors prefer that a firm retains funds over paying dividends (Baker and Powell, 1999). This is not surprising given that the investors commonly cannot easily to find out the complexity of differential tax effect on dividends and capital gains. Most of the research described above has been conducted on the taxation effect of dividends in a classical tax system. The generally supported conclusion is that there is a substantial tax effect on ex-dividend share prices. Less clear, is whether the same conclusion can be made for countries that have enacted an imputation tax system. Since the imputation credit is additional to the dividends distributed under imputation system, the taxation effect should theoretically include not only dividends itself but also the imputation credit attached. Therefore, the total effects of share prices will be more complex when dividends, imputation credits and the relative tax effects are considered together at the same time. This is discussed further in the following section. 10

21 4. Does dividend imputation credit matter to share value? 4.1 Prior research of imputation tax credit It is generally agreed that the measurement of firm performance is based mostly on net income, not on the dividend paid. However, the amount of dividend may affect the share price in different ways9. Besides considering the performance shown on income statement, investors always look into balance sheet to find out the book value of the company for investment. Because the balance of imputation tax credit account ( ICA) is not part of the assets of a firm, it is not entered into the formal accounting system. Although not part of the assets of a firm, the disclosure of the balance in the footnotes of its financial reports is essential to investors. Especially, in terms of the amount of imputation credit offered by a firm. For example, if the imputation credit is high then there is a greater likelihood that shareholders will take advantage from it. Only when the imputation credit is actually attached to the dividend distributed, is the imputation credit valuable to shareholders. Brown and Walter (1986) examined ex-dividend behaviour of Australian share prices during the period 1974 to 1984 (pre imputation). The dividend drop-off ratio in Australia was significantly less than one, and estimated at around 75% to 80%. Nonetheless, Brown and Walter were not prepared to argue that this discounting is tax induced due to the wide difference in the tax status of Australian shareholders. After the introduction of imputation, Clarke (1992) studied how dividend imputation has affected the market capitalization of listed Australian equities. He found, pre-imputation, that in general the Australian equities market price listed equities in such a manner that reflected a market-wide preference for capital gains over dividends. Post imputation, the indications were that there has been a marked change in the magnitude and direction of the dividend premium. This is contrary to the expectations of the new tax regimes for dividends and capital gains adopted by Australia since September By relating to the three major changes in taxation of capital gains, dividends, superannuation funds in Australia. Brown and Clarke (1993, p ) noted, "while the average drop-off ratio (grossed-up) increased after the introduction of dividend imputation, consistent with the tax preferential hypothesis, the increase was not statistically significant...moreover, we estimate that the market, on average, has taken some time to access the implied value of the tax credit." Brown and Clarke (1993) also documented that by 1990, that is three years after the implementation of an imputation tax system in Australia, shareholders had typically obtained 80% of the 11

22 benefit of the imputed tax credit.. In investigating the development of shareholder clienteles in response to the introduction of dividend imputation, Bellamy (1994, p.284) concluded that "... the marginal investor does consider imputation credits to be valuable, since the ex-dividend price fall was greater for franked dividends than for unfranked dividends." In finding out a model for pricing imputation tax credits under Australian imputation tax system, Wood (1997) claimed that the equilibrium value of imputation tax has important implications for corporate and portfolio investment in the Australian capital market. He suggested that there is a market value of approximately 60% of face value for tax credits associated with securities held by investors. Based on the assumption that dividend taxes are capitalized into share prices. Harris, et al. (2001, p.571) suggested, When dividend taxes are capitalized, share prices absorb the burden of dividend taxation, whether or not a firm pays current dividends. Dividends per se do not produce a tax penalty because paying a dollar of dividend reduces firm value by less than a dollar.... preserving the dividend displacement property of Miller and Modigliani (1961)." They empirically examined whether dividend taxes are capitalized into share prices by focusing on investors' implicit valuation of retained earnings versus paid-in equity. Consistent with dividend tax capitalization, they confirmed accumulated retained earning is valued less per unit than contribution capital. in the United States. By comparing evidence from countries operating partial or full imputation tax system, such as France, Germany and Australia, Harris et al. (2001) also conclude that, with tax relief, i.e. the possibility of a negative tax rate, and the value of imputation credits itself, the tax discount for retained earnings are significantly less than those in classical tax system. 4.2 Drop-off ratio post imputation in Taiwan In the Australian tax system, not all shareholders can fully gain from the imputation tax credits attached because credits can only be credited to income tax payable and are not refundable in anyway. In contrast, in Taiwan imputation credits are equivalent to dividends to all individual shareholders because of their fully refundable nature. Even so, it is still questionable whether imputation credits are valued in the same way as dividends received directly by investors. Theoretically, all else being equal, share prices should reflect the tax effect on both dividend and imputation credit at ex-dividend day. Therefore, to shareholders, the grossed-up dividend income under

23 the imputation system in Taiwan should equal Dx(1 C), where C is the imputation credit ratio. By adding the credit ratio to Equation (1), the Elton and Gruber's (1970) equilibrium model post imputation could be extended as follows: Pb D(1 ±C) (1 TO= Pa + Tc (Pb Pa), (3) Where, Pb share prices before ex-dividend day; Pa: share prices on ex-dividend day; D : dividend received; T0: tax rate of ordinary income; T: tax rate of capital gain; C : ratio of imputation credit to shareholders. Rearranging Equation (3), the traditional drop-off ratio post imputation is as follows: Rp= (Pb Pa ) D= (1 +C) (1 To) (1 TO. (4) By comparing Equations (2) and (4), we can see C> =0, i.e. (0% %), so (1 + C) (I To) + (1 Tc) > = (1 To) ± (I Tv). Because, R, > R, the drop-off ratio after imputation will be greater than those before imputation by the scale of credit ratio. Similarly, from Equation (4) it is possible to predict a positive relationship between the drop-off ratio and the imputation credit ratio. 4.3 Imputation tax credit is additional to share value Based on the empirical conclusions of Brown and Clarke (1993), Bellamy (1994), Wood (1997), Harris et al. (2001), and others, there appears reason to support the notion that imputation credit is an addition to dividends and should be valued to some extend the same as dividend. -A corporation with substantial accumulated franking credits...is thus more attractive to investors, a fact that may be reflected in the corporation's stock price" (Kreiser et al. 1998, p. 59). Therefore, from the viewpoint of investors, the performance and the related share value of a firm should be determined not only by net income, but also the imputation credit to the extent actually attached to the dividend under imputation tax system. This is totally different to the classical tax system. For example, if the share price of company A in classical tax system is $50, then under full imputation, the share price should be more than $50 depending on the 13

24 amount of imputation tax credit is. That is, since the imputation tax credits are additional to dividends and net income, the credits should theoretically be capitalized in the share price under the Efficient Markets Hypothesis (EMH)I. On the other hand, assume cum-dividend price of share A is $50 and cash dividend is $ 5 under classical tax system. If there is no dividend tax or transaction cost to shareholders, on the first day of ex-dividend, the share price will be equivalent to $ 45. In contrast, under full imputation tax system, if cum-dividend price were also assumed $50, cash dividend is $5 and a $ 2 dividend imputation tax credit was attached to the cash dividend, then the ex-dividend share price will be $43 since shareholders receive dividend income of $5 and get the refund of $2 from the government. In this case, no matter how the imputation credit was capitalized in the share price, the expected drop-off on ex-divided day under imputation tax system should be greater than that of classical tax system. 4.4 The declaration of imputation tax credit Since imputation credit is as valuable as dividend to shareholders, it becomes very important for corporations to announce it at the same time as the dividend resolution (i.e. the date of declaration as depicted in Figure 1) to give a full picture of dividend yields to shareholders. It is likely that there will be a time lag before imputation credit declared on either the date of record (such as in Taiwan), or on the date of payment (such as in Australia), is reflected in stock markets. Once the imputation credit is declared, it should be deemed a part of dividend. Because the amount of dividend is one of the most important signals to share market, neither dividends nor the imputation credit attached should be varied or revoked for any reason under an imputation tax system. 5. Implications and further research Because investors are always focused on the bottom line (net income) and dividends of a firm, any unexpected changes of net income and dividend will probably affect the share prices. Since there is evidence that dividend taxes and imputation credit are capitalized in the share prices, it seems little reason for not disclosing the final amount of imputation credit at the same time of dividend announcement. This raises the question of whether share prices would be affected and disturbed improperly solely by the imputation tax policy as that in Taiwan. For example: I. If investors do not know the final imputation credit before, or at, ex-dividend day, it means they do not know their real dividend income. It is unknown how well the share prices reflect only the tentative dividend amount. 2. More importantly, according to Taiwanese taxation laws, firms must revise the 14

25 imputation credit ratio should any errors be found. It is unknown whether share prices change in response to this form of correction. 3. Since real dividend income of investors should conceptually include the final imputation credit attached, should there be a mechanism to declare the imputation credit so as to give the market all of the relevant information? All of these situations and questions involved in the imputation tax system, not only have the implications to share market, dividend policy and tax policy makers, but also raise the possibility for further research in these relative fields. In Taiwan, firms only disclose the final imputation credit in the dividend statement in the following February for income tax return purposes. Due to investors not knowing the final imputation credit before ex-dividend day, they don't know the real total amount of dividend and, accordingly, they will have no idea about their actual income tax burden in the future. If dividend imputation credit is relevant to a firm's value, and is important information in capital market, share prices must presumably be affected or distorted by the lack of adequate disclosure of the final imputation credit on the declaration of dividend. Therefore, it is in doubt as to whether imputation credit is fully capitalized in share value in Taiwanese equity market. If it is capitalized in share value, the ex-dividend day expected traditional drop-off value (Rp) after imputation should be greater than that before imputation. As well, there should be a positive relationship between expected traditional drop-off value (Rp) and imputation credit ratio (C) by inferring from Equation (4). To understand how investors treat these situations, the author proposes to empirically explore the ex-dividend day behaviours of Taiwan's share prices before and after tax imputation in another research (Wang, 2002b). 6. Conclusions If imputation credit could be completely refunded by shareholders, the total dividend would be the sum of dividend received and the attached imputation credit. According to the conclusion of Harris et al. (2001), retained earnings are generally valued less per unit than contribution capital, due to tax differential of capital gains and dividends. Therefore, imputation credit, which is addition to dividends and retained earnings also, should be relevant to a firm value by both the imputation credit itself and the relative tax effects. Under the Taiwanese tax policy, the determination of imputation credit is based on the date of record for dividend payment, and is variable and hence never fully disclosed at 15

26 ex-dividend day. This policy causes a time lag for dividend imputation tax credit information. By contrast, in Australia, although the extent of dividend franked (franking level) is certain, the declaration of it being made before the dividend is paid is also too late in this sense. Since dividend imputation credit is as valuable to investors as dividend received, it is probably that share prices to some extent will be distorted or affected by tax policies such as those currently in place in Taiwan and Australia. It is concluded, therefore, that corporations should be required to announce the amount of imputation credit per share at the same time as dividend resolution (i.e. date of declaration). Once it is declared, it should not be varied or revoked for any reason. 16

27 References 1. Abrutyn, S. & Turner, R. W. 1990, -Taxes and firms' dividend policies: survey results', National Tax Journal, vol. 43, no. 4, pp ?. Baker, H. K., Family, G. E. & Edelman, R. B. 1985, 'A survey of management views on dividend policy', Financial Management. vol. 14, no. 3, pp Baker, H. K. & Powell, G. E. 1999, 'How corporate managers view dividend policy', Quarterly Journal of Business and Economics, vol. 38, no. 2, pp Bellamy, D. E. 1994, 'Evidence of imputation clienteles in Australian equity market', Asia Pacific Journal of Management. vol. 11, no. 2, pp Black, F. & Scholes, M. 1974, 'The effects of dividend yield and dividend policy on common stock prices and returns', Journal of Financial Economics, vol. 1, no. 1, pp Booth, L. D. & Johnston, D. J. 1984, 'The ex-dividend day behavior of Canadian stock prices: Tax changes and clientele effects, Journal of Finance, vol. 39, no. 2, pp Brennan, M. J. 1970, 'Taxes, market valuation and corporate financial policy', National Tax Journal, no. 23 ( December), pp Brigham, E. F., Gapenski, L. C. & Ehrhardt, M. C. 1999, Financial Management: Theory and practice, 9th edn, The Dryden Press, Pp Brown, P. & Clarke, A. 1993, 'The ex-dividend day behaviour of Australian share prices before and after dividend imputation', Australian Journal of Management. vol. 18, no. 1, pp Brown, P. & Walter, T. 1986, 'Ex-dividend day behaviour of Australian share prices', Australian Journal of Management, vol. 11, no. 2, Dec., pp Campbell, J. A. & Beranek, W. 1955, 'Stock price behaviour on ex-dividend dates', Journal of Finance, vol. 10, no. 4, pp CCH 1996a, 1996 Australian master tax guide, CCH Australia limited. Australia. pp CCH 1996b, Income Tax Assessment Act 1936, CCH Australia limited, Australia. The international handbook of corporate and personal taxes 1992, 1St edn, Chapman and Hall, London, UK, Pp Clarke, A. W. 1992, 'The ex-dividend day behaviour of Australian share prices pre and post imputation', Managerial Finance, vol. 18, no. 1, pp Durand, D. & May, A. M. 1960, 'The ex-dividend behavior of American Telephone and Telegraph stock', Journal of Finance, vol. 15, no.1 (March), pp

28 Eades, K. M., Hess, P. J. & Kim, E. H. 1984, 'On interpreting security returns during the ex-dividend period', Journal of Financial Economics, vol. 13, no. 1, pp Elton, E. & Gruber, M. 1970, 'Marginal stockholder tax rates and the clientele effect', Review of Economics and Statistics, vol. 52, no. 1, pp Fama, E. F. & French, K. R. 1998, 'Taxes, financing decisions, and firm value Journal of Finance, no. 53, no. 3, pp Gordon, M. J. 1963, 'Optimal investment and financing policy', Journal of Finance, vol. 18 (May), pp Harris, T. S., Hubbard, R. G. & Kemsley, D. 2001, 'The share price effects of dividend taxes and tax imputation credits', Journal of Public Economics, vol. 79, n. 3, pp Jensen, M. C. & Meckling, W. H. 1976, 'Theory of the firm: managerial behavior, agency costs and capital structure', Journal of Financial Economics. vol. 3, no. 4 (October), pp Kalay, A. 1982, 'The ex-dividend day behavior of stock prices: A re-examination of the clientele effect', Journal of Finance, vol. 37, no. 4, pp Kreiser, L., Butcher, B. & Jowitt, E. 1998, 'Corporate dividend strategy in an integrated income tax environment', International Tax Journal, winter, vol. 24, no. 1, pp Lasfer, M. A. 1996, 'Taxes and dividends: The UK evidence', Journal of Banking & Finance, vol. 20, no. 3, pp Litzenberger, R. H. & Ramaswamy, K. 1982, 'The effects of dividends on common stock prices: Tax effects or information effects?', Journal of Finance, vol. 37, no.2, pp Long, J. B. 1978, 'The market valuation of cash dividends: A case to consider', Journal of Financial Economics, vol. 6, no. 2/3, pp Miller, M. H. & Modigliani, F. 1961, 'Dividend policy, growth and the valuation of shares', Journal of Business, vol. 34, no.4, pp Morgan, G. & Thomas, S. 1998, 'Taxes, dividend yields and returns in the UK equity market', Journal of Banking and Finance, no. 22, pp OECD 1991, Taxing profits in a global economy: Domestic and international issues, Paris, France. Parmenter, B. R. & Seyfort, A. 1987, 'The imputation system of company tax: context, characteristics, critique', Australian Economic Review, Quarter, pp Pettit, R. R. 1977, 'Taxes, transactions costs and the clientele effect of 18