RDP 9211: Dividends and Taxation: A Preliminary Investigation 3. Influences on the Payout Ratio in the Australian Context
October 1992
- Download the Paper 402KB
This section looks at factors influencing the dividend payout ratio in Australia. It focuses, in particular, on the tax treatment of dividends.
3.1 Taxes
Several major changes to the Australian tax system have been made since 1985. These have been aimed at reducing the distortions affecting company financing decisions and the allocation of investors' savings that were previously present in the tax system. Each of these changes has implications for dividend policy.
The capital gain on an asset acquired after September 1985 is subject to Capital Gains Tax (CGT). If the asset is held for more than one year, only the real gain is taxable. The capital gains tax was introduced to redress the bias in the tax system towards income received in the form of capital gains, and to reduce the distortion towards investment in those more ‘speculative’ assets yielding income in the form of capital gains.
In July 1987 a full imputation system was introduced which eliminated the double taxation of company income. Prior to this, company income distributed as dividends was taxed twice, thus making returns in the form of dividends less attractive than other returns. This may have distorted investors' portfolio allocations away from equity.
The effect of dividend imputation was magnified in July 1988 when a tax of 15 per cent was imposed on the earnings of superannuation funds. Previously, superannuation funds were exempt from tax and therefore the introduction of imputation did not affect their returns. When the tax was introduced on their earnings, they were able to make use of the imputation credit. This increased the attractiveness of equity relative to alternative assets compared with the situation when funds paid no tax.
In the next section we explain how imputation operates and in the second section we examine the effects of the various tax changes on dividend policy.
3.1 (a) Dividend imputation–how it works
Corporate tax law is complex and a full discussion of dividend imputation is beyond the scope of this paper. In this section we present a simplified review of the main aspects of dividend imputation as it applies to several main classes of investors.
Prior to the introduction of dividend imputation, company tax was levied on profits earned and after-tax profits distributed as dividends were taxed again in the hands of shareholders. Dividend imputation ensures that company profits paid out as dividends are taxed only once. Companies credit their “franking accounts” with the amount of income that can be distributed as franked dividends. Credits arise mainly from tax payments made or from receiving franked dividends from the company's shareholdings. Debits arise from paying franked dividends. Table 1 provides estimates of the degree of franking for a sample of industrial and resource companies listed on the Australian Stock Exchange over the period 1987 to 1991.[2]
| Half Year Ended | Industrials | Resources |
|---|---|---|
| December 1987 | 76.8 | 66.9 |
| June 1988 | 84.9 | 52.1 |
| December 1988 | 86.1 | 81.3 |
| June 1989 | 89.0 | 70.1 |
| December 1989 | 90.2 | 92.0 |
| June 1990 | 82.8 | 88.6 |
| December 1990 | 81.7 | 90.2 |
| June 1991 | 82.5 | 95.5 |
| December 1991 | 78.8 | 79.1 |
Source: J.B. Were & Son, Profit Monitor, May 1992 |
||
Shareholders receive franking credits to the extent that company tax has already been paid on dividends received. They are then able to offset these credits against taxable income that is derived from any source, not just dividends. Any excess credits cannot be refunded to a shareholder, offset against a shareholder's Medicare Levy, or carried forward. Therefore, if a shareholder has insufficient tax liabilities against which to offset the imputation credits, the credits will be lost.
Table 2 shows the effects of dividend imputation for four different groups of shareholders subject to different marginal tax rates, namely; individuals, companies, superannuation funds, and non-residents. We assume the following: a fully franked dividend of 61 cents per share is paid and each shareholder holds 100 shares.
| Individual | Super fund | Company | Non-residents | |
|---|---|---|---|---|
| Marginal Tax Rate | 47% | 15% | 39% | |
| Franked dividend | 61 | 61 | 61 | |
| Imputation credit | 39 | 39 | 0 | |
| $61*(τ/1−τ) | ||||
| Assessable Income | 100 | 100 | 61 | |
| Tax payable | 47 | 15 | 24 | |
| –tax credit | 39 | 39 | 24 | |
| Tax payable | 8 | −24 | 0 | 0 |
| After tax return | 53 | 85 | 61 | 61 |
| Before-tax yield | 6.1% | 6.1% | 6.1% | 6.1% |
| After-tax yield | 5.3% | 8.5% | 6.1% | 6.1% |
When paying tax on dividends received, shareholders are required to gross up the dividend by the amount of tax that has been paid by the company on the profit associated with the dividend. Given a 39 per cent company tax rate, τ, the tax paid is equal to
Dividend * τ/(1−τ)
= $61*(0.39/0.61) = $39
Thus for a 39 per cent company tax rate, the $61 dividend received must be grossed up by $39 and the grossed-up value of $100 must be included in their assessable income. Shareholders are then taxed at their marginal rate on the grossed up amount, but receive tax credits for the tax already paid by the company. For the individual shareholder on the top marginal rate of 47 per cent, the tax liability on the $100 is $47, but this is reduced by the tax credit of $39. The shareholder has to pay tax of $8 on the dividends received. Table 2 shows the different effects for the four groups of investors. Note that non-residents do not participate in the dividend imputation system but are exempt from dividend withholding tax if dividends received are franked.
The table shows the before and after tax yields of dividends for the different shareholders. Superannuation funds receive the highest aftertax return on dividends while individual shareholders paying tax at the top marginal rate receive the lowest. The latter group are nevertheless better off than in the previous regime where dividends would have been taxed twice, reducing the after-tax yield to 3.2 per cent (assuming unadjusted equity prices).
3.1 (b) The effect of tax changes on shareholders' preferences for retentions and dividends
In this section, we identify the effect of the changes in tax arrangements for dividends, capital gains and superannuation funds on the four different classes of shareholders previously identified: individuals, companies (other than superannuation funds), superannuation funds and overseas residents. Table 3 outlines the timetable of tax changes as they affect these groups of investors and shows the current tax position of each.
| Individuals | Companies | Super funds | Non-residents | |
|---|---|---|---|---|
| Dividends | Dividend imputation effective from 1 July 1987. Credits can be offset against income tax on the shareholders' income, but not against the Medicare levy. |
Dividends received by a resident company are relieved from tax by the intercorporate dividend rebate. The recipient company is able to pass the imputation credits to its shareholders. | Introduction of a 15 per cent tax on earnings in July 1988. Dividend imputation effective from that date. | Withholding tax on dividends, applied from 1 July 1960. From 1 July 1987 franked dividends exempt from withholding tax. Unfranked dividends are subject to 30% or 15% withholding tax depending on the double taxation treaty. (Most countries are covered by treaties.) |
| Capital Gains | Assets acquired after 19 September 1985 are subject to CGT. If the asset is held for more than one year, it applies to the real component only. |
Same as for individuals. | All assets disposed of after 1 July 1988 subject to CGT at the rate of 15%. If the asset is held for more than one year, it applies to the real component only. | Assets acquired after 19 September 1985 subject toCGT.(Same marginal tax rate as appliesto residents) |
The above-mentioned tax changes should have had an influence on the desired dividend payout ratio of companies by altering the relative value of retained earnings vis-à-vis dividends. Poterba and Summers (1985) define a tax discrimination variable, θ, as the opportunity cost of retained earnings in terms of dividends foregone, both measured after personal tax. If there is no discrimination, the value of θ is one. If θ is less than one, capital gains are favoured by the tax system, if θ is greater than one, dividends are favoured. θ is defined as,
where σ is the effective marginal personal tax rate on dividends and c is the effective marginal personal tax rate on capital gains. Below we consider θ for each of the four groups of investors we have identified and then construct an aggregate measure of θ.
For individuals, there are a wide range of tax bands. Here, we consider two: the first for someone on average income, the second for someone paying the highest marginal personal income tax rate. The only difference this distinction makes is in the size of the tax bias. There is never a case where one of the groups of personal investors considered prefers dividends and the other prefers capital gains. There was a significant bias in favour of capital gains before the introduction of the capital gains tax. Between the introduction of the CGT and imputation, there was a slight bias in favour of capital gains. Since imputation, there has been a bias in favour of receiving dividends. Some individuals who are low marginal rate taxpayers and who have insufficient other income against which to offset the tax credits are effectively taxed at the company tax rate. However, this is not likely to be very important because these people are unlikely to be major investors in equity.
Superannuation funds were indifferent between capital gains and receiving dividends priorto the introduction of the 15 per cent tax on their earnings in September 1988. However, the introduction of the tax, at a lower rate than the company tax rate, generated a strong bias towards receiving dividends for this group of shareholders. This is because superannuation funds can offset the extra tax credits against other taxable income.
For overseas investors, the withholding tax is not imposed on fully-franked dividends. This extends the benefits of imputation to overseas investors. However, because the amount of the withholding tax on a dividend will be less than the value of the imputation credit, overseas investors do not receive the full benefits received by domestic investors. Although foreign investors are disadvantaged relative to domestic investors insofar as they are unable to utilise fully tax credits derived from franked dividends, the combination of the introduction of CGT and the abolition of the dividend withholding tax for fully-franked dividends have contributed towards θ being greater than one. This, however, is only half the story, since we have not included the taxes imposed on non-residents in their home country.
To derive an aggregate measure of θ, the θ's for each group of investors need to be given appropriate weights. No data are available on equity holdings by sector. Therefore, the weights for each group are calculated from ABS data on dividends received.[3] Data are available from 1969/70. The calculated weights are plotted in Graph 1. These figures indicate that individuals' direct holdings of equity have fallen significantly over the period, from over 60 per cent in the late 1960s to 14 per cent in 1990/91. Holdings by superannuation funds and life offices rose to 30 per cent in 1990/91 compared with about 7 per cent in the late 1960s. However, this rise has not matched the decline in individuals' holdings over the period. Holdings by overseas investors have also risen over the period.
Graph 2 plots the aggregate measure of θ on the basis of both the average and top rate personal tax rates. Clearly, the net result of the changes in taxes since 1985/86 has been to encourage the payment of dividends by firms.
3.2 Dividend Reinvestment Schemes
A related factor that has influenced the dividend payout ratio in recent years is the increase in dividend reinvestment schemes. These schemes allow shareholders to reinvest their dividend receipts in shares in the company. They resolve the conflict between a company's desire to retain profits and the shareholders' desire for a higher dividend payout by enabling the firm to recapture the dividends it pays out. Shareholders prefer to receive imputation credits sooner rather than later as the value of these credits declines in real value over time.[4]
Dividend reinvestment schemes became an important source of equity finance in the late 1980s, rising from 4 per cent of total equity raised in 1987/88 to 33 per cent in 1990/91 (see Table 4). Approximately one third of dividends were reinvested in 1990/91. Under the schemes, shares are normally priced at a discount to the market price and are attractive to small investors because of the lack of transaction costs. For companies where capital issue costs are high, dividend reinvestment plans are also a favourable option. The success of these reinvestment schemes has encouraged companies to pay a higher level of dividends in the knowledge that they will retain the funds.
| Year to June | 1988 | 1989 | 1990 | 1991 |
|---|---|---|---|---|
| per cent | 3.8 | 14.2 | 28.9 | 32.7 |
Footnotes
The sample is comprised of 126 leading industrials and 14 major resource companies. [2]
The ABS provide data on dividends received by life offices and super funds and are unable to split the data any further. This complicates the analysis of tax changes since life offices are taxed at the company tax rate, while super funds are taxed at 15 per cent. However, as at September 1991 life offices held $61.8 billion of assets in super products, over 70 per cent of total assets, and these assets are taxed at the concessional rate of 15 per cent. Thus, not much is lost by making the simplifying assumption that all dividends are received by super funds. [3]
Also popular following the introduction of dividend imputation was the streaming of dividends so that franking credits were directed to those shareholders who could use them. For instance, investors could elect to receive a fully-franked cash dividend or either a bonus-in-lieu of dividends or an unfranked dividend. If the bonus shares were issued out of the share premium reserve they were not deemed as dividends. The main benefit from the latter accrued to shareholders who had pre-September 1985 shareholdings since the bonus shares issued were exempt from capital gains tax. However, from 1 July 1990 dividend streaming was brought to an end when bonus shares issued out of share premium reserves under streaming arrangments, resulted in a debit to the company's franking account. [4]