Effect of Combin ing Corporate and Personal Income Taxes on Capital Structure Decisions: Evidence from Taiwan Market Effect of Combining Corporate and Personal Income Taxes on Capital

Effect of Combining Corporate and Personal Income Taxes on Capital Structure Decisions: Evidence from Taiwan Market

Yen-Chou Chao Hualien Tax Bureau

Hualien, Taiwan Tel: 866-03-8633134

Email: [email protected]

Huijun Jin

University of Alaska Anchorage

Anchorage, Alaska 99508 Tel: (907) 561-7336

Email: [email protected]

Suresh C. Srivastava

University of Alaska Anchorage

Anchorage, Alaska 99508 Tel: (907) 786 4148

Email: [email protected]

Ken Hung National Dong Hwa University

Hualien, Taiwan Tel: 866-3-8662500, ext. 20112

Email: [email protected]

Correspond to: Suresh C. Srivastava E-mail: [email protected]

Fax: (907) 786 4115

- 1

Effect of Combining Corporate and Personal Income Taxes on Capital Structure Decisions: Evidence from Taiwan Market

ABSTRACT

Taiwan, Republic of China (ROC), introduced a new income tax system, combining

corporate and personal income taxes, on 1 January 1976. This new tax system, resembling those used in many European countries, has had a significant impact on the

debt/equity capital structure of companies domiciled in Taiwan. This study provides historical background concerning taxing policies, an analysis of the new income tax

system (combining the Two Tax Systems into One, or 2 -1-CTS) in Taiwan and examines its impact on the capital structure decisions in Taiwan.

I. INTRODUCTION

There have been many theoretical works published on the correlations between the capital

structure and corporation valuation. First to be introduced is the classic theory on financial

management, capital structure irrelevance theory, by American economists Modigliani and

Miller in 1958. In perfect markets, i. e., in the absence of personal and corporate income

taxes, no transactions costs, and free and fully informed markets, stock price (an indication of

company’s value) has no relation with leverage effects; the stock price will be determined by

the value produced from the company’s assets. Because the mentioned “irrele vance of the

capital structure” is based on many unrealistic assumptions, many later tried in relaxing those

assumptions. In imperfect, real markets, changes in capital structure can affect corporate

values. The theories on the capital structure generally are based on the consideration of rents

and taxes, tax benefits from the debt financing and financial stress from bankruptcy, tradeoffs

among different costs, agency costs due to the conflicting interests among management,

creditors, and sharehold ers, as well as the pecking order theory. Among which, rent and taxes

factors include effective taxes and non-tax-benefits, such as tax shields, tax deference and

deductions (DeAngelo and Masulis, 1980).

In addition, the discussions in previous publications all have mentioned the impacts of

- 2

the effective tax rates on the financing policies of a company; the company’s capital structure

also can affect its effective tax rates. The higher the leverage, the higher the interest payments ,

the lower the effective tax rates. Therefore, leverage and the effective tax rate are inversely

related. On the other hand, other things (such as risk costs, operations, and non-debt tax

shields) held equal, the company tends to take on more debt when it foresees that more tax

shields per unit interest costs and value addition to the company can be brought by higher

effective rates. Similar to the spider network theory in the individual economics, the effective

tax rate and leverage level will be determined by the two forces of opposite directions.

However, when the effective tax rates increase, the company will benefit from larger tax

shields, tempting for more debts and resulting in a positive-slope curve. The equilibrium can

be found at the intersection point where optimal debt ratio and effective tax rate are located.

Although the above discussions are simplified, it implies that capital structure and effective

tax rates are important interdependent factors.

Most of the published literature only emphasize one direction, either only on the impacts

of the effective tax rates (actual rent rates) on a company’s financing behavior, or the effects

of other factors, including capital structure, on the effective tax rates. However, there

generally have been few discussions on the causal or consequential interactions between the

two factors. Only Liyuan Chen (1978) expressed the causal relationship between the effective

tax rates and the company capital structure, using models based on the coordinated regression

equations. However, her model has been deduced from the relationships between the effective

tax rates and the capital structure of a company. It is true that capital structure can

immediately impact the effective tax rate via the influence on the tax shielding effects during

the given period. However, the effective tax rate at the given period cannot immediately and

directly affect the financing policies (capital structure). This effect can only be felt with a

time lag in the future periods. That is to say, present financing policies (capital structure) will

be influenced by the previous effective tax rates and foreseeable future effective tax rates,

- 3

forecasted based on the previous data series. There were three problems before the

combination of the two taxes into one. First, the dividends were double taxed. Second, the

financial decisions were misleading using more debt because the interest payments were tax

deductible costs. In addition, when the profits were distributed to the shareholders, they had

to pay personal income taxes for the dividend income. Therefore, the company would defer

the dividend payment by converting it into retained earnings to avoid the taxation for

shareholders. Third, it led to an unfair taxation system. In the face of the increasing domestic

economic transformations and global competition, the (Republic of China or ROC, Taiwan)

government enacted the combination of the two taxes into one It was aimed at competing

edges of the enterprises by adopting a fair income taxation system and creating low-tax loads

and barrier-free investing environments in the ROC. In addition in the dividend-paid

deduction method, because the dividend the shareholder received contains tax deductibles, it

will greatly reduce the tax loads. This will discourage companies from debt financing too

much, and encourage equity financing, contributing to healthier company finances and

reduced debt risks.

This report compares the significant differences in the capital structure and effective tax

rates before and after the combination of the two-taxes-into-one policy. The report also aims

at analyzing the relationships and their changes between the capital structure and the effective

tax rates (independent variables), and capital structure and the effective tax rates (dependent

variables) before and after the applications of the new taxation system, using ordinary least

square (OLS) regression method.

II. THE 2 TO 1 TAXATION SYSTEM

There generally are two classes of corporate tax systems in the world. The first type is the

independent taxation system, represented by the United States. Under this tax system, the

firms are independent taxable business identities and pay corporate income taxes. The

- 4

shareholders have to pay personal comprehensive income taxes on the dividends received

from the ir investment in the company. The corporate and personal income taxes are

independent. The second type is the combination taxation system, represented in the

European countries. Under this system, the corporations are virtual business entities;

therefore, cannot be taxed independently. The corporations are only the transferring channels

of the profits to the shareholders. Therefore, only one taxation is levied at either the

corporation level or at the dividend distribution time, but not both. This is called the

2-to-1-combination taxation system. It is considered the most signification tax reform since

1964 when income taxes for added values were enacted. The new system enables the

shareholders to deduct their dividend income when paying personal income taxes because the

corporate income had been taxed at the corporate level.

The 2-to-1 combination taxation system can be divided into 8 approaches depending on

the degrees and stages of tax combination, including: (1) partnership approach; (2)

dividend-paid deduction system; (3) dividend-paid credit system; (4) split-rate system; (5)

dividend-exemption system; (6) dividend-credit system; (7) imputation system; and (8)

hybrid systems.

A. Partnership Approach

The partnership approach regards shareholders as partners. Their personal income will

be taxed at appropriate personal income tax brackets regardless of whether the corporate

incomes are distributed. Therefore, there are corporate taxes under this approach; or, even

though it exists, it is more of a tax withholding instrument other than s tax itself. The

partnership approach was advocated by the Canadian Carter Commission in 1966, by the US

Treasury in 1977, and by the Australian Campbell Committee in 1983. This approach

eliminates the double taxation for the corporate shareholders, resulting in more fair tax loads

for various incomes and the progressive personal income tax brackets. However, it creates

financial difficulties for governments because of the disappearance of corporate taxes. In

- 5

addition, it also increases the administration load due to the numerous shareholders and their

frequent trading. If there are more than one type of shares, the dividends are difficult to

determine. At present, only the S-corporations in the US are using this approach.

S-corporations refer to the corporations that are classified under the Subchapter S for

domestic tax laws. This type of corporation generally has less than 35 employees. Unanimous

agreement of all shareholders of these corporations can choose to be taxed as S-corporations

before the 15th day of the third month during the taxable year. In the S-corporations, all the

incomes and costs belong to shareholders. S-corporations are not taxable. Shareholders

receive corporate income, deductions, losses, etc. according to their ratios of shares.

B. Dividend -paid Deduction System

Under the Dividend-paid Deduction System, when the corporations calculate their

taxable incomes, a ll the dividends, partially, or at given percentage, will be subtracted as

business costs. The corporate taxes are levied on the undistributed net income. The

shareholders will then include their dividends in their personal annual taxable income, when

calculating their income taxes. The lower the percentage of the dividends allowed for

corporate cost deductions, the closer this system will be to an independent taxation system.

This tax system aims at eliminating the distortions of financing through stock issuance and

debt financing.

C. Dividend -paid Credit System

The Dividend-paid Credit System allows the corporations to deduct payable corporate

income taxes based on a given percentage of distributable dividends. It has the same

objective as the Dividend-paid Deduction System.

When shareholders receive their dividends, they can deduc t a certain percentage of their

income tax. This system aims at abating the double taxation.

D. Split-rate System

- 6

Under the Split-rate System, there are two income tax rates depending on whether

corporate incomes are to be distributed. When the incomes are distributed, low er tax rates

should be applied; undistributed incomes should be taxed at higher rates. This system aims at

eliminating the distortions in corporate incomes and the undistributed incomes, abating

double taxation; but it will affect the corporation’s capital accumulation.

E. Dividend -exemption System

Under the Dividend-exemption System, the shareholders’ dividends can be completely

or partially exempted in the shareholders’ personal income taxes. The former is called

complete exemption; the latter is called partial exemption. It aims at eliminating or reducing

the double taxation while also rewarding investments. Under this system, the progressive

degrees of personal income tax is lowered and its tax administration is easier. The present

comprehensive personal dividend income is taxed under the partial Dividend-exemption

system when personal income is less than 270,000 Taiwan Yuan.

F. Imputation System

Under the Imputation System, income taxes at the corporate level can be completely or

partially used to reduce shareholders’ personal income taxes. If all the taxes at the corporate

level can be used to reduce shareholders’ income taxes, the system is called complete

imputation system; if partially effective, its is called a partial imputation system. In the partial

imputation system, the higher the percentage that can be reduced from personal income taxes,

the closer the system is to the complete imputation system; the lower the percentage, the

closer it is to the independent taxation systems. Shareholders should include dividends as

actual dividends plus imputable taxes. When the shareholders’ tax rate brackets are higher

than the imputation rates, they need to pay more taxes; otherwise, they can file for tax refunds .

This system can eliminate the distortions resulting from stock issuance and debt financing,

and the problem of undistributed incomes.

G Hybrid System

- 7

The Hybrid System allows different combination methods at corporate and shareholder

levels. Genera lly, the Split-rate system is more commonly used at the corporate level, and the

distributable incomes are taxed at lower rates. At the shareholders’ level, the complete or

partial imputation system and dividend-exemption system, or dividend-credit methods are

used.

III. TWO-TO-ONE TAXATION SYSTEM IN THE REPUBLIC OF CHINA (ROC)

Most of the countries using 2-to-1 combination taxes are in Europe and are also used in

Singapore and Malaysia, Australia and New Zealand. Countries using specific types are listed

in Table 1.

A. Principles for Imputation.

The two-to-one combination tax system in the ROC aims at the illogical double taxation

of corporate income. However, this combination does not mean that one of the taxation

systems is absorbed by the other. Actually, after the combination, corporate income taxes and

personal income taxes still exist; corporate income taxes become withholdings. In the future,

when corporations distribute dividends, they can imputate that portion of corporate income

taxes and distribute it to shareholders as deductions for personal income taxes. When the

shareholders’ marginal tax rates are higher than the corporate tax rates, they need to pay the

difference; otherwise, they can file for a tax refund. As a result, double taxation can be

avoided. In addition, the reinvestments in the companies are exempted from income taxes.

However, this imputation system only applies to citizens of the ROC.

1. Corporate Level

(1) The company needs to establish imputable tax accounts for shareholders, record

their corporate income taxes, and calculate the shareholders’ imputable taxes.

- 8

(2) When the company distributes dividends, it should use the tax withholding rate of

the residual account withholdable taxes over accumulated undistributed income, calculate withholdable taxes according to shareholders’ individual net amount of dividends, and distribute it together with dividends. The upper limits for tax imputation are:

Undistributed income not yet taxed by 10%, 33.33% = 25% / (1 - 25%);

Undistributed income already taxed by 10%, 48.15% = (32.5% / (1 - 32.5%));

When the undistributed income is partially taxed by 10%, use a weighted average of 33.33% and 48.15% based on its proportion in accumulated undistributed income, i. e.,

33.33% × A+ 48.15% × (1-A)

where A is the proportion of corporate income not yet taxed by 10% in the corporation’s accumulated undistributed income.

Shareholders’ withholding tax rate = net dividends* tax withholding rate

Total dividends = net dividends + shareholders’ withholdable taxes

(3) Corporations need to file reports on dividends with authorities before the end of January of the year following when the dividends were distributed; copies also need to be sent to the shareholders no later than February 10 for their filing of personal tax returns.

2 Shareholders’ Level

(1) Domestic shareholders should include the total dividend amount in the ir taxable

income, together with the shareholder’s other taxable income for tax reporting. The corporate tax paid contained on the dividends (from corporations) can be deducted

from total tax; if it is more than one should have paid, the tax return should bring a refund of the excess.

(2) Domestic legal person shareholders ’ net dividend amount should not be included as income for taxation. If one is a corporation organizer, the deductible tax amount

contained in the dividend (distribution) should be included as residual in one’s corporation shareholders’ deductible tax account. If the corporation is an educational,

cultural, public benefit, or charities organization, the deductible tax amount distributed together with the dividends cannot be used to offset one’s taxes payable,

or used in filing the tax return.

- 9

(3) The income tax rates for for-profit firms remain unchanged. For -profit firms still

calcula te and pay corporate income taxes according to the present stipulations, as well as the personal income taxes.

(4) The undistributed corporate income should be taxed at additional 10% for for-profit organization income taxes.

After the application of the combination of the 2-to-1 taxation system, the highest

marginal tax rate is still 25% for for-profit firms. Compared to the 40% marginal tax rates for

comprehensive incomes, there is still a 15% difference. There is the incentive to retain

earnings to avoid shareholders tax loads. On the other hand, Taiwan has stopped taxing some

incomes, such as those from securities trading and land trading. Corporate retained earnings

increase net share values, causing the share price to increase. In addition, the realization of

this gain from this part of capital will not be taxed—and is an additional incentive to retain

earnings.

In most countries using the 2-to-1-combination taxation system, there is not a great

difference in the highest marginal corporate and personal tax rates. In some countries, using

that system, corporate income tax rates are higher than personal income taxes (Table 2). In

addition, tax revenue losses due to the applications of the 2-to-1 combined taxation systems

should not be borne by other taxpayers. That is the reason for the additional 10% for-profit

income taxes levied on retained undistributed earnings. After the 10% additional taxes,

undistributed earnings can be retained without limit. The additional taxes are prepaid taxes;

they can be distributed to shareholders as deductible tax amounts when the dividends are

distributed to shareholders.

IV. REVIEW OF LITERATURE

A. Effective Tax Rates and Determinants

According to Fullerton (1984), effective tax rates can be divided into marginal and

average effective tax rates. Marginal effective tax rates refer to the expected tax loads for

- 10

capital gains from additional unit investment. They are generally applicable to expected

future incomes, reflecting future tax loads; while the average ETR is the proportion of annual

taxes in the total net income, indicating the past tax load. The concept of marginal ETR

originates from the capital cost formula of Hall and Jorgenson (1967). For one investment

project, business owners need to pa y the government taxes and required returns for capital

suppliers. Therefore, the business owners need a minimum pre-tax return rate from the

investment project. The project return rate will be the capital cost when it is in equilibrium.

The difference between the capital utilization cost and after-tax return of the capital suppliers

is the so-called marginal ETR from capital gains.

The effective tax rate discussed in this paper refers only to the effective corporate tax

rates (ETRs), which is the percentage of income taxes on pre-tax net income. This figure

reflects the amount of corporate tax loads and the extent to which corporations enjoy tax

exemptions. The early research on the ETRs primarily emphasizes the relationships between

firm sizes and the ETRs. Taxes are regarded as one of the political costs to firms. Therefore,

the ETR reflects the load to firms of the political costs (Siegfried, 1974; Stickney, 1982;

Zimmerman, 1983; Porcano, 1986; Shevlin and Porter, 1992; Manzon and Smith, 1994; from

Mingjin Chen, 2001). However, the results from research do not agree well with the above

hypothesis. Zimmerman (1983) found that large corporations had higher tax costs, and

advocated that ETRs correlate directly with corporation size. This agrees well with the

political cost hypothesis, which argues that large companies are monitored and watched more

closely by government and the public. For example, governmental tax departments audit

large companies more often. Consumers generally tend to ask for invoices more from large

companies than from smaller ones. As a result, large companies have to entail higher political

cost.

However, Siegfried (1974) and Porcano (1986) found that company size correlates

inversely with ETRs. Their aanalyses indicate that large companies are good at utilizing the

- 11

accelerated depreciation and overseas tax deductions permitted by the tax laws, resulting in

lower ETRs. Therefore, they advocate the political power hypothesis on the relationship

between the ETR and company size. Under this hypothesis, compared with small companies,

large companies have more resources for political lobbying and hiring professionals for tax

planning, in order to reduce their tax loads. In addition, some other scholars find there is

significant correlation between the ETRs and company size (Stickney and McGee, 1982;

Shevlin and Porter, 1992; Manzon and Smith, 1994).

The ETRs also are often used in assessing the fairness of tax distribution resulted from

reforms of taxation systems (Shevlin and Porter, 1992; Gupta and Newberry, 1997; from

Mingjin Chen, 1979). For example, the Tax Reform Act (TRA86) in the United States in

1986 significantly reduced tax benefits, such as accelerated depreciation and investment

deductions; in the meantime, it cut tax rates and enlarged the tax bases, hoping to create a

fairer taxation environment. After studying the changes of the ETRs for the US companies

before and after the TRA86, Shevlin and Porter (1992) classified the effects of the changes

into: (1) changes in rules; (2) changes in tax rates; and (3) changes in reportable income.

Their analysis indicates that, although the TRA86 lowered the stated tax rates, it increased the

corporate ETRs. After controlling income change effects, the TRA86 raised ETRs by the

greater change effects from enlarging the taxation bases, over the effects from the stated tax

rate reductions. Gupta and Newberry (1997) agreed on the analysis with Shevlin and Porter

(1992). Their results also indicate that large r companies had higher ETRs before the TRA86,

compatible to the political cost hypotheis; larger companies should have lower ETRs after the

TRA86, compatible to the political power hypothesis.

Capital structure refers to the combination conditions of the company capital sources, or

broadly refers to the relative ratios among various long-term capital, such as long-term debt

and shareholder’s equity (Jianping Xie, Financial Management, Second Edition).

Theoretically, the ultimate goal of the combination of debt and equity is to maximize

- 12

company valuation. When a company finances by debt, it can enjoy the interest tax shield.

When operations are going well, shareholders do not have to share the income with debtors,

because the debt interest is fixed. However, when debt ratios are too high, bankruptcy risks

will be high, which could endanger shareholders’ equity. Therefore, how to make decisions

on financial planning and how to maximize company valuation have been key issues in

corporate activities.

Company features (such as operational risks, profitability, capital pledge values,

company size, and R & D expenditures) , financing and investing policies have intrinsic

influence on ETRs and capital structure. The changes in the external taxation environments

also affect them. Research indicates that company debt ratios have been effectively lowered

after the application of the 2-to-1-combination (corporate and personal taxes mergers) in

Canada and New Zealand (Schulman et al., 1996). In the meantime, changes in governmental

taxation systems will change corporate tax expenses, affecting the ETRs.

The emphasis of this article is on the changes of the ETRs and capital structure, and their

pertinent influencing factors before (1974, 1975), during (1976,1977) and after (1978,1979)

with the applications of the 2-to-1-combination taxation system.

1. Capital Structure (Leverage, or LEV)

Harberger (1959) and Tmbini (1969) discussed the distortion effects of debt and equity

financing, due to the different treatments of the tax laws on company interest payments and

dividends. For example, the interest payments due to the debt financing can be deducted as

expenses, resulting in a tax shielding effect, compared to the equity financing. Therefore, the

more debts the company takes on, the more interest payments, and the lower effective tax

rates (Stickney and McGee, 1982; Gupta and Newberry, 1997; Tianxiong Zhang, 1999). As a

result, ETRs have a negative correlation with capital structure (leverage).

2. Company Size (SIZE)

Different conclusions can be reached under different hypotheses on the relationships

- 13

between the company sizes and the ETRs. Under the political power hypothesis, there is an

inverse relationship between the company size and ETR, because large companies have more

resources for political activities and for hiring professionals on tax planning, in order to

reduce their tax loads (Siegfried, 1974; Porcano, 1986). Under the political cost hypothesis,

there is a direct relationship between the company size and its ETR (Zimmerman, 1983),

because larger companies are more likely to become the targets for being monitored and

regulated. As a result, larger companies could have higher ETRs.

Siegfried (1974) and Porcano (1986) discovered that company ETRs relates inversely

with its size. Larger companies are good at using overseas tax deductions and accelerated

depreciation, resulting in a lower tax rate. Their research supports the political power

hypothesis. In Taiwain, the studies also showed mixed results on the relationships between

the ETRs and company sizes. Some results showed that the ETRs relate d positively with

company sizes (Suxin Cai, 1971; Chen et al., 2001); some others also showed a negative

correlation, such as the research of Shiming Lin and Zhaoxu Ya ng (1972). This paper will not

further discuss the correlations between the ETRs and company sizes.

3. Income Variance (VAR)

The larger the variability in a company’s income, the lower its ETRs (Gropp, 1997, cited

from Liyuan Chen, 1978). For example, when a maximum variation occurs, during the first

period and the company suffered a loss, but in the second and third periods received net

incomes, the company could carry forward the deferred losses to the next periods when

reporting its taxes, thus reducing the taxes for the profitable periods.

Note: In Taiwan, the Section 39 of the Income Tax Law permits the loss carrying

forward to the next five years.

4. Profitability (PM)

When the company is very profitable, tax loads will increase if tax shields cannot

increase accordingly (Shevlin and Porter, 1992; Gupta and Newberry, 1997). Therefore, the

- 14

company’s ETRs relate inversely with the company’s profitability.

5. Capital Intensity (CAPINT)

The higher the capital intensity of a for -profit organization, the more it will enjoy the

depreciation tax shield and investment deductions (Stickney and McGee, 1982; Gupta and

Newberry, 1997, cited from Mingjin Chen, 2001). In Taiwan, the Stipulation for Encouraging

Industry Upgrading, Section 6, points out that, under the following-purpose expenditure items,

companies can deduct taxable corporate incomes by up to 5-20% if they do not have enough

profits to reach the deduction limit, the residual amount can be carried forward for the

following four years. The expenditure items are: (1) investment on automation equipment and

technology; (2) investment on resource recycling, pollution prevention equipment and

technology; (3) investment in the equipment and technology to utilize new and clean energy,

energy saving and reuse of industrial water; and (4) investment in the technology and

equipment for reducing greenhouse gases emissions, or for improving energy utilization

efficiency. When companies purchase automatic equipment, energy-saving and

pollution-prevention equipment and technology, they can enjoy the investment deductions,

resulting in lower ETRs. Otherwise, the ETRs correlate inversely with capital intensity.

Company investments in R&D can be regarded as expenditures in tax reporting, enjoying the

tax shielding for early expenses and upgrading industry-stipulated deduction benefits. It

should be noted that up to 5-20% of the investment on R&D and personnel training can be

deducted from taxable income for that year; if the R&D expenses are greater than the average

of the R&D expenses of the previous two years, or the personnel training expenses are greater

than the average of the two previous years, 50% of the excess amount can be used to deduct

taxable income in that year; if the income in that year is less than the deductions, the surplus

deductions can be carried forward for four years. Therefore, the higher the R&D ratio, the

more the company enjoys the investment and depreciation tax shielding effects and the lower

the ETRs (Gupta and Newberry, 1997). Otherwise, ETRs inversely relates to RDINT.

- 15

6. Long-term Retained Earning Income

Others items held constant, the higher the ratio of tax-exempt securities and land

transaction income in the total income from corporate investments and capital gains, the

lower the ETRs (Chen et al., 2001; cited from Mingjin Chen, 1979). Before the 2-to1

combination tax law, 80% of dividend income from long-term corporate investments in other

company’s tax-exempts, resulted in lower ETRs. After the 2-to-1-taxation system, net

corporate dividends from investing in other domestic for -profit organizations could be

excluded from taxes, but should be included in distributable income. The tax deductions

distributed to shareholders together with the dividends could not be used to deduct income

taxes for for -profit organizations; instead, they should be included in the shareholders’

deductibles account residuals. When earnings are to be distributed to individual shareholders,

these deductibles should be distributed to shareholders for their personal comprehensive

income tax deductions. In another words, dividend from reinvestment by companies were

completely exempted under the 2-to-1-combination tax system, because the 10% taxes levied

on the undistributed earnings were prepaids for future personal income tax deduction for

shareholders.

B. Research on the Capital Structure Determinants

The factors influencing company capital structure are related to the industry

characteristics. They include the effective tax rates (ETR), Non-debt Tax Shields (DEP),

business risks (VAR), collateral value of assets (CVA), profitability (PM), company size

(SIZE), and growth rate (GR).

Effective Tax Rates (ETR)

Mudigliani and Miller (1963) pointed out that debt leverage has tax shielding effects.

The company’s value increases with the degree of leverage (DOL). In the extreme situation,

if only corporate taxes are considered, the company should be financed by 100% debt, in

order to maximize its value. However, this is impossible in reality, even though banks are

- 16

financed by 90% debt. Miller (1977) considered corporate and personal income taxes, and

concluded that various combinations of the Corporate Tax Rates (TC), Tax rates for personal

earnings (TPE), and Tax Rates for dividends (TPD) could result in different tax benefits. They

pointed out that for an economic system, there should be an optimal capital structure; but

some companies might not have optimal structures. In reality, considering that tax laws

permit different treatments for stock dividends and debt interests, companies with high tax

rates tend to use more debt financing because the companies can save more in tax deductions

from debt interest payments. Other things held equal, effective tax rates are directly related to

capital structure.

Non-Debt Tax Shields (DEP)

The amortization of intangible assets, the depletion of natural resources, and the

depreciation of fixed assets can be expensed to decrease income tax payments and cash

outflows through reduced net income, resulting in tax benefits. There is a significant

substitution effect between the non-debt and debt tax shields (Dhaliwal, Trezevant and Wang,

1992; Trezevant, 1994). Therefore, the higher the tax shields, the lower the company debt

ratio. However, purchasing tangible assets through debt financing can enjoy double tax

benefits: depreciation expenses and investment deductions (Bradley, Jarrell, and Kim, 1984).

In this case, the non-debt and debt tax shields are directly related. This will not be covered in

this report.

Business Risks (VAR)

If the company business risks are high, it will reduce the triggering factors leading to

second-best decisions for investments, and consequently decrease debt agency costs from

these behaviors (Myers, 1977). As a result, this company has a better ability to finance by

debt than those companies with lower business risks, i. e., the business risks relate directly to

debt ratio. Generally speaking, a company tries to maintain risks (business and financial risks)

at certain levels. However, business risks are influenced by some external factors that cannot

- 17

be controlled by the company, such as changes in industry technologies, capital intensity, and

market conditions. Therefore, an organization has to adjust its financial risks according to its

business risks. The lower the business risks, the higher a company can bear financial risks

due to debt financing and the better ability to use debt leverage. Actual research results on the

variance of shareholders equity market price as an indicator of business risks suggest that

business risks are directly related with debt ratios (Kim and Sorensen, 1986). However,

research on the variances of the earning before interest, taxes, depreciation and amortization

(EBITDA), as an indicator for business risks , show that business risks are inversely related

(Bradley, Jarrell and Kim, 1984). Some other scholars’ research reveals no significant

relationships between business risks and debt ratios (Wessels, 1988; Ferri and Jones, 1979).

Collateral Values of Assets (CVA)

In general, when companies apply for loans from banks, the banks require the companies

to provide pledges. The higher the values of the pledged assets, the easier the companies can

get their loans. Therefore, tangible assets can enhance company’s ability to debt finance

(Myers, 1974). Scott (1977) points out that debt financing will be increased with pledged

assets, depriving the assets from creditors without pledged assets. Advertising and R&D

accumulate intangible assets in a company, reducing corporate debt (Bradley et al. , 1984).

Grossman and Hart (1982) support that pledged values are inversely related with capital

structure, noting that when the company’s pledged assets are low, its creditors will increase

their monitoring and control of the company, in order to prevent management from abusively

using corporate resources for personal benefits. As a result, a company can use higher debt

ratio to reduce its costs.

Profitability (PM)

Myers and Majluf (1984) proposed the pecking order theory, suggesting the order that a

company tends to finance is internal capital, debt, and equity capital. The companies with

high profit margin can acquire more retained earning from operations. When these companies

- 18

need capital, they can retain earnings to use, then use debt financing if the retained earnings

are not adequate. Therefore, more profitable companies usually have lower debt leverage.

Company Size (SIZE)

Since small- and middle-size companies cannot easily finance through the stock markets,

they have to use more debt financing unless they can raise capital private ly. Their debt ratios

will be limited by the bankruptcy distress costs and probability. The direct bankruptcy costs

of large companies inversely relate to company size (Warner, 1977). However, Myers (1977)

argues that capital structure directly relates to company size because large companies tend to

diversify operations, can finance more easily with financial instruments, have higher credit

classes, and favorable interest rates.

Growth Rate (GR)

A company’s assets can be divided into tangible assets and growth opportunities (Myers,

1977). The latter is regarded as a call option. Part of the company value depends on whether

this call option will be carried out. Since high-growth companies have more flexibility in

selecting future investment plans, management has more slack to choose. They tend to

choose the second best options , benefiting shareholders and harming creditors; in addition, it

makes creditors’ monitoring activities more difficult, increasing agency costs. Even though

growth opportunities can be considered as an asset, they cannot be pledged for debt financing

(Myers, 1977). If most of assets in the company are intangible, or of growth type, its debt

financing ability can be limited. However, high-growth companies have more optional

investment projects; the higher the debt ratio, the more chance that shareholders will choose

to exploit creditors investment projects. As a result, the higher the growth rate, the more the

company will choose to debt finance (Titman, 1984). Kester (1986) found that debt ratio is

directly related to growth rate, which is inconsistent with the result of agency cost theory.

Jensen and Meckling (1976) pointed out that growth rate is inversely related to the short -term

debt leverage and convertible bonds of a company, but is directly related to the long-term

- 19

debt leverage.

C. 2-To-1 C ombination Taxation System Research

Extensive studies on the 2-to-1-combination taxation system (corporate and personal

income taxes) have been undertaken in foreign countries. Schulman et al. (1996) discuss the

effects of the 2-to-1-combination taxation system (2-1-CTS) on corporate capital structure in

New Zealand (enacted in 1988) and Canada (in 1972). In Canada, net income from securities

capital also has been taxed. The sampling periods were before and after the year the 2-1-CTS

in New Zealand (1982-1991) and Canada (1968-1977). The 2-1-CTS effectively reduced

corporate debt ratios in New Zealand. The Effectiveness in Canada was somewhat offset by

the taxation on securities income. These studies indicate that the changes in tax policies can

lower corporate financial leverage, but that is not the ultimate goal of the 2-1-CTS. Only

when the advantages of the 2-1-CTS are introduced into the progressive economic system,

could an entire nation benefit from its application. The advocates of this system propose to

eliminate the distortions of the tax policies on the economy and society, consequently

reducing the economic and societal costs, accelerating capital accumulation, and improving

the general savings rate.

Based on the survey of 268 quoted companies from 18 industries in Taiwan, Chen and

Huang (1980) analyzed several aspects of capital structure two years before and after the

2-1-CTS (1974 vs. 1975 and 1976 vs. 1977). The analyses include d: (1) whether the

corporate debt ratio would be lowered, (2) whether the tax shielding effects on capital

structure would be reduced, (3) whether the non-debt shielding effects on capital structure

would be decreased, and (4) whether the influences of retained earnings on capital structure

would be reduced. In addition, statistics indicate that the average effective tax rates of the

surveyed companies in Taiwan were reduced from 13.7171% before the 2-1CTS to 12.9876%

afterwards.

- 20

IV. RESEARCH METHOD AND DISCUSSIONS

The data in this research are from the Taiwan Economy Newspaper. Samples include quoted

companies on the Taiwan Securities Exchanges. Because Taiwan’s public reporting on

income tax accounting was comparatively late, it did not have contemporary and delayed

income tax cost data until 1973. The period is divided into two periods in this study: before

(1974-1975) and after the 2-1-CTS (1976-1977). T he sampling period is from 1974 to 1979

(6 years). In selecting the sampling period, the companies starting their IPOs in and after

1974 are excluded because of their possibly being influenced by financing in capital markets

during those years. In addition, the financial and insurance businesses are excluded because

of their special nature. Because there was only one company in the “comprehensive” category,

it was included in the category of “other”, which had 18 businesses originally. The companies

that were no longer quoted, were bought out, merged with others, or with incomplete data

were excluded. Due to the consideration of the credibility and effectiveness on this study, it

was required that any of the studied companies needed to have 6 years contemporary

(1974-1979) of data. Only 257 companies were selected using the above criteria.

The statistics on various variables before and after the 2-1-CTS are shown in Tables 2 and 3.

The effective tax rates and debt ratios are listed by industries in Tables 4.

From Tables 2 and 3, the average effective tax rates during 1974-75 and 1976-77 of the

257 companies in this study were 8.824% and 7.496%, respectively; the corresponding

average debt ratios were 0.087 and 0.102. The effective tax rates before and after the

2-1-CTS were far below the nominal tax rate 25% for the for-profit organizations in Taiwan,

indicating that the taxation system benefits and rewards most of the surveyed companies. The

average debt ratios before and after the 2-1-CTS were lower than those for Taiwan’s major

trading partners, the US (0.185) and Japan (0.155) (Wald, 1999).

The Z-factor on the sampling company’s average debt ratio and effective tax rate before

and after the 2-1-CTS were examined to detect significant differences. The average debt ratio

- 21

difference examination in Table 5 indicates there was no expected reduction of the ETR under

the 5% -significance level, because the Z-factor (-2.883) was less than the threshold value

(1.645). This could be caused by the effects of the global economic recession after the

2-1-CTS in Taiwan (especially after 1978-79). With the depressed stock markets and low

interest rates, the costs and risks for financing from the stock markets became increasingly

high. Companies tended to obtain financing from financial institutions. As a result, the

average debt ratio did not fall as expected.

The examination of the average ETR difference in Table 5 indicates that there was an

expected reduction of the ETR under the 5%-significance level. Because the Z-factor (2.184)

was less than the threshold value (1.960), the average ETR was lower afterwards.

This research used the ordinary least squares regression to study the correlations and

their variations among the capital structure and factors affecting the effective tax rates in

Taiwan before and after the 2-1-CTS. The following Equations 4-1 and 4-2 are regressed

between the capital structure and ETR.

ETR it = a 0 + (a1 + a1 x D) x LEVit + (a2+a 2 x D) x SIZEit + (a 3+a3 x D) x VAREBTit + (a 4+a4 x

D) x PMEBTit + (a5+a5 x D) x CAPINTit + (a 6+a6 x D) x RDINTit + (a 7+a7 x D) x

LTREIit + ? it ……………………………………………………………(Equation 4-1)

LEVit = ß0 + (ß1 + ß1 x D) x ETRb5it + (ß2 + ß2 x D) x NDTS it + (ß3 + ß3 x D) x VARBrit + (ß4 +

ß4 x D) x CVA it + (ß5 + ß 5 x D) x PMNiit + (ß6 + ß6 x D) x SIZEit + (ß7 + ß7 x D) x

GRit + uit……………………………………………………………….(Equation

4-2)

i—Number of companies in the study; t—Year; ?it , uit – residual values; and D—virtual

variables of the 2-1-CTS. D = 0 in 1974 and 1975; D = 1, in 1976, 1977, 1978, and 1979.

- 22

In order to understand whether there was a multiple linearity problem produced from

the regressions between the ETRs and capital structure factors, correlation coefficients

analyses were carried out for the capital structure factors. From the debt ratio regression

Equation 4-2, the absolute values of the coefficients between any two variables were less than

0.3. in the ETR regression equation, except that the coefficient between debt leverage (LEV)

and the company size (SIZE) of 0.47206 is high. Other coefficients were all less than 0.3.

According to Anderson et al, multiple linearity could occur when the absolute value of the

coefficients between any two variables are greater than 0.7. Therefore, in this research, the

occurrence of multiple linearity problems was quite slim.

Tables 5-1 and 5-2 show the results on the regression analyses of the factors affecting

the effective tax rates (ETR) and capital structure before and after the 2-1-combination of tax

systems (2-1-CTS).

For the established models, the R2-values were only 0.035233 for the ETR and 0.295862

for capital structure. However, they all reached the 1% significance level in the F-value

examination.

The results in Table 7 indicate that the capital structures (LEV), Variation of EBT (VAR),

R&D expenditure (RDINT) and company size (SIZE) are negatively correlated with the

effective tax rates (ETR). The former three factors agree well with the predictions based on

the hypothesis of political power. However, only the correlations of the LEV and VAR were

significant before the 2-1-CTS. In addition, the profit margins before taxes (PMBT) are

positively correlated with the ETRs, as expected, and also the correlation was significant

before the 2-1-CTS. The long-term retained earning investment (LTREI) was directly related

to the ETRs before the 2-1-ETS, but the correlation was not significant. After the 2-1-CTS, it

was inversely related and not significant, either. Surprisingly enough, the capital intensity

(CAPINT) was directly related to the ETR with significance before the new taxes policies.

The effects and degrees of change of the influencing factors on the ETRs before and after the

- 23

2-1-CTS are given by the absolute values of the coefficients of the regression equations

before and after the new tax system (a n+ a n* ). (Note: n = 1, 2, …, 7). Only the influence of

the capital structure (LEV) was enhanced after the 2-1-CTS (but did not reach the

significance level); the influence of all the other factors decreased (only PM reached the

statist ical significance).

In Table 7, it is shown that the ETR, and profitability (PM) were inversely related to the

capital structure (LEV) before and after the 2-1-CTS. The correlation between the LEV and

ETR was not predicted by the political power theory. Non-debt tax shields (NDTS) and

company size (SIZE) directly related to the LEV before and after the 2-1-CTS, but only

before the 2-1-CTS was it significant statistically. Business risks (BR) and pledged values of

assets (CVA) were inversely related to the LEV before the 2-1-CTS, and were directly related

afterwards. The growth rate (GR) was positively correlated with the LEV before the 2-1-CTS,

which was not expected; but the correlation was negative afterwards.

The effects and degrees of change of the influencing factors on the capital structure

(LEV) before and after the 2-1-CTS are given by the absolute values of the regression

equation coefficient before and after the new tax system (ßn+ ßn* ). (Note: n = 1, 2, …, 7).

Only the influence of the capital structure (LEV) was enhanced after the 2-1-CTS (not

reaching the significance level); the influence of all the other factors decreased (only PM

reached the statistical significance). The results indicate that the influences of the ETRs and

SIZE increased after the 2-1-CTS (not reaching statistical significance); that of all other

factors decreased (only the pledged values of assets, or CVA reached statistical significance).

As mentioned above, the correlations among LEV, VAR, M and RDINT, and the ETR

were expected by the political cost hypothesis. The negative correlation between company

size and the ETR was expected by the political power hypothesis, i. e., large r companies have

more resources for political activities and hiring tax-planning professionals to mitigate their

tax payments.

- 24

Before the 2-1-CTS, the long-term retained earning investments (LTREI) were directly

related to the effective tax rates (ETR); afterwards it was inverse. Before the 2-1-CTS, the

average investments and benefits ratio was 0.26; afterwards it became 3.16. This was a

remarkable increase. The governmental objectives of the 2-1-CTS were to improve corporate

operating efficiency and not having tax rates affect financial preference. Although the system

levies an additional 10% tax on the undistributed operating income, after that there is no limit

on the retained earnings. Companies can use the undistributed retained earnings to reinvest in

other businesses, or to expand the company’s size. This effect can be seen from when before

the new tax system, the total capital Ln was 8117.89; and afterwards it became 16,502.36. In

addition, the income from the reinvestments was not included in the taxable corporate income,

encouraging long-term corporate equity investment, and consequently reducing the effective

tax rates.

Why was the influence of profit margins on the effective tax rates reduced after the

2-1-CTS? After the new tax system, mergers and acquisitions were in their climax period, and

the weight of external investments increased significantly. As a result, the long-term

shareholders’ income from other business investments amounted to a large portion of the

pretax net income, which also was not included in the taxable corporate income. Therefore,

its influence on the ETR was reduced. This phenomenon might weaken the fairness principle

of taxation, based on the corporate income from operations, affecting the operating policies to

some extent, because external investment income was basically exempted from taxation. The

companies focusing on their own businesses were penalized because all their operating

income was taxable. This distortion has rarely been addressed.

In addition, the results indicate that the profit margin and company size are either

positively or negatively correlated to capital structure, as expected by the established theories.

The direct correlation between non-debt tax shields and capital structure was predicted by

Bradley et al. (1984), but the correlation was more statistically significant. The present tax

- 25

system in Taiwan is a parallel system of income taxes and tax rewards, in which deductions at

the corporate levels will increase taxation at the shareholders’ personal level. The effects of

tax deductions due to each individual’s favorable tax-exempt rate for reinvestment outside the

companies disappear because the income from the reinvestment is not taxable after the

2-1-CTS (Qiu, 1977). The shareholder’s comprehensive income taxes increased after the new

tax system because tax deductibles were reduced as a result of the corporate exempted and

deductible income. Therefore, non-debt tax shields are directly related to capital structure. It

was unexpected that the average ETR during the previous five years was negatively

correlated. Why high-ETR companies did not use high debt leverage when the effects of the

debt leverage on the previous 5-year average ETR was increased? Under high effective tax

rates, high leverage can provide substantial tax shields through interest payments, but the

companies have to bear the financial risks of the high debt financing, which result s in

underinvestment (Myers, 1977). Companies have to discard investment projects with positive

net present worth, but which cannot pay debt principals and interest. Once suffering losses,

huge debt interest payments would close down promising companies. It is educational and

psychological to deal with taxes as business expenses, rather than evading taxes. The

negative correlation, without statistical significance, might be the fact that before the

2-1-CTS, the average ETR was far below the nominal tax rate of 25% in Taiwan.

Additionally, there are only three tax brackets in Taiwan (Table 5-3), compared with many

tax brackets in other developed nations, such as the United States. Therefore, the

accumulative degrees are small.

Business risks (BR) were negatively correlated with capital structure before the

2-1-CTS; it was positively correlated afterwards. In both situations, they did not reach the

significance level. After the new tax system, the influence of business risks on the capital

structure decreased, but this correlation is with inadequate statistical significance.

- 26

The correlation of the pledged values of assets (CVA) on capital structure was negative

before the 2-1-CTS and positive afterwards, all with statistical significance. After the new

tax system, the influence of the CVA decreased significantly. This might indicate that in a

knowledge economy, financial organizations no longer evaluate companies entirely on their

tangible assets, such as land, equipment and inventories. They have to rely more on the

assessment of the intangible assets, such as R&D inventions and innovations, marketing

operations, team working abilities and the development potentials.

Before the 2-1-CTS, capital structure and growth rate (GR) were directly and

significantly related, agreeing well with the results of some scholars (Kester, 1986; Jensen

and Meckling, 1976; Titman, 1984). However, it conflicts the agency costs theory. This

might have been caused by the evaluation of future growth opportunities , based on the

average growth rates of operating income during the previous five years. After the new tax

system, the leverage (LEV) and the GR were inversely related, with a statistical significance.

Their correlation deceased after the 2-1-CTS, with a statistical significance because before

(1974 and 1975) and after (1978 and 1979) the 2-1-CTS, the economic situations both in

Taiwan and overseas changed significantly.

VI. CONCLUSIONS AND SUGGESTIONS

The research data are from the database of the Taiwan Economic News, not directly

from sampling companies’ financial reports , and the reliability and integrity of data are based

on tha t database. The effective tax rates were measured by the income taxes over pre-tax net

income. These are the conclusions and recommendations:

(1) The average effective tax rates were significantly lower after the 2-to-1

combination tax system (2-1-CTS).

(2) The company debt leverage ratios did not decrease after the 2-1-CTS as

expected.

- 27

(3) The correlations between the company size and the effective tax rates agreed

well with the political power hypothesis, i. e., that company size related

inversely with its effective tax rates. This was similar to the conclusion of

Huang (1978) that, since more than 90% of the companies in Taiwan are

middle and small in size 1, such tax deductions , as governmental fiscal policies

might not take effect in a timely manner. Subsidies for financing policies

might better benefit small and middle size companies , and might avoid the

unfair taxation system as a result misallocation of societal resources due to

abusive tax deduction methods. In addition, government can directly use fiscal

policies to stimulate businesses, creating jobs.

(4) The functions of income taxes on for-profit organization in Taiwan were

significantly affected by the 2-1-CTS. The objectives of the new tax policies

were an investment environment with low taxes, creating competitive edges

for businesses; and a fair taxation system to sustain the growth the investment

environment.

(5) The education campaigns on the taxes have gained acceptance among the

citizens. Business operators can deal with the tax problems reasonably, using

tax planning rather than evading taxes. They also can include the taxes in the

operating costs.

(6) In a knowledge economy, financial organizations calculate valuation more on

intangible assets, such as RD inventions and innovations, marketing

operations, team-working abilities, and corporate development potentials,

1 According to Ministry of Economic affairs, in 1979, there were 1,098, 185 companies, a increase of 0.64%

from that in 1978. Among them, 1,078,162 were small and middle companies. In 1979, total labor force in

Taiwan were 9,382,540 persons. The large companies employed 12.08% of them; middle and small

companies —77.67%; government —10.24%.

- 28

rather than on tangible assets, such as land, equipment and inventories.

(7) The research results indicate that debt ratios did not decrease as expected after

the 2-1-CTS. This was also at odds with the expectation of government

planner. Why? The author preliminarily concludes that it was caused by the

global economic recession, weak markets and depressed interest rates,

resulting in high costs and risks for companies to finance in the equity. As a

result, companies preferred to obtain debt from financial institutions and the

debt ratios did not decrease. This deserves more investigation.

(8) This research on the growth rate is rough. Maybe a trend analysis could

improve the prediction accuracy.

(9) The direct relationships between capital intensity and the effective tax rates

before and after the 2-1-CTS are at odds with other research results. Is this

because that most of the surveyed companies purchased equipment outside the

favorable tax-deduction scope , or the accelerated deprecation in the tax

accounting treatment was not recorded, or other factors? There is a lot room

for further studies.

- 29

References

Bradley, M., G. A. Jarrell, and E. H. Kim, 1984. “On the existence of optimal capital structure: Theory and evidence”. Journal of Finance 39 (July):857-878.

Chen, Lin, and Chang.2001“The impact of tax-exempt stock and land capital gains on corporate effective tax rates. ”Taiwan Accounting Review, Forthcoming.

Chen, Guanghua, Laixi Li, and Shengjie Zeng, translated; Zhou, Fuxing reviewed, June 1977, Statistics (Vol. 1), 1st Edition, from David R. Anderson, Dennis J. Sweeney, Thomas A. Williams STATISTICS FOR BUSINESS AND ECONOMICS , 6th ed. (In Chinese).

Chen, Guanghua, Laixi Li, and Shengjie Zeng, translated; Zhou, Fuxing reviewed, December 1977, Statistics (Vol. 2), 1st Edition, from David R. Anderson, Dennis J. Sweeney, Thomas A. Williams STATISTICS FOR BUSINESS AND ECONOMICS , 6th ed. (In Chinese).

Chen, Liyuan, 1987, Rent Load and Company Capital Structure, Unpublished Master’s Thesis of Institute of Finance, National University of Politics (In Chinese).

Chen, Longqi, and Yufang Wen, 1980. Research on the factors deciding the capital structure of the quoted companies. Taiwan Banks Quarterly, 53 (1), March, 171-194 (In Chinese).

Chen, Mingjin, 1979. Research on the Factors of Determining the Effective Tax Rates on the Profitable Organizations. Proceedings, Symposium on Theories and Practices of Accounting, Private Dongwu University, 2001 (In Chinese).

Chen, Miaoling and Xiaoling Huang, 1980. Impacts of the combining two taxes into one on enterprise capital structures. Research of Finance and Taxes, 34 (5) (September): 99-113 (In Chinese).

DeAngelo, H., and R.W. Masulis, 1980. “Optimal capital structure under corporate and personal taxation”. Journal of Financial Economics 8 (March), 3-30

Dhaliwal, D., R. Trezevant, and S. Wang, 1992. “Taxes, Investment-related Tax Shields and Capital Structure,” The Journal of the American Taxation Association 14 (1), 1-21. ”

Durand, D. “Cost of Debtand Equity Funds for Business: Trends and Problem of Measurement,” in Conference on Research in Business Finance, National Bureau of Economics Research , New York,1952, pp. 57-60.

Eugene F. Brigham, Louis C. Gapenski, and Phillip R. Daves. Intermediate financial management sixth edition

Ferri, M. G., and W. H. Jones, “Determinants of Financial Structure: A New Methodological Approach”, Journal of Finance, Vol. 34 , No.3, 1979, pp. 631-634

Fullerton, Don, “Which Effective Tax Rate”? National Tax Journal, Washington, Date Mar 1984, Volume 37 , Issue 1, Page 23, 19 Pages

Gropp, R. E., 1997, “The effect of expected effective corporate tax rates on incremental financing decisions.” IMF Staff Papers, 44, pp.485-509.

Grossman, S., and O. Hart (1982), “Corporate financial structur e and managerial incentives.” In J. McCall (ed.)The Economics of Information and Uncertainty. Chicago: University of Chicago Press.

Gupta and Newberry 1997, “Determinants of the Variability in Corporate Effect Tax Rates:

- 30

Evidence from Longitudinal Data.” Journal of Accounting and Public Policy, 16 , No.1, pp. 1-34.

Hall, R.E.,and D.W. Jorgenson 1967, “Tax Policy and Investment Behavior, ”American Economics Review, 57, pp. 391-414.

Huang, Junying, October 1972, Research Methods on Enterprises, 1st Edition, National Translation and Compilation Institute (In Chinese).

Huang, Wanyu, June 1978.Comparison of the major factors influencing the effective tax rates on the high-tech industry in Taiwan: before and after the enacting of the combination of the two taxes. Unpublished Master’s Thesis of the Institute of Financial Management, National Central University (In Chinese).

Huang, Xinyao, June 1980. Research on Capital Structure of the China Telecommunications Company. Unpublished Master’s Thesis of Institute of Operations Management, National University of Transportation (In Chinese).

Jensen, Michael C. and William H.Meckling,1976, Theorem of the firm: managerial behavior, agency costs and ownership structure, Journal of Financial Economics 3 , pp. 305-360.

Jiang, Xiaoming, translated and compiled, Modern Financial Management, edition No. 8, August 1988., from Eugene F. Brigham and Joel F. Houston FUNDAMENTALS OF FINANCIAL MANAGEMENT, 8E, Huatai Culture Corporation Limited (In Chinese).

Kester W. Carl; 1986 “Capital and Ownership Structure: A Comparison of United States and Japanese Manufacturing Corporations” Financial Management; Tampa.

Kim & Sorensen, “Evidence on the Impact of the Agency cost Debt on Corporate Debt Policy”, Journal of Finance and Quantitative Analys is, 1986, pp. 131-144.

Mackie-Mason, J. K. 1990. “Do taxes affect corporate financing decision”? Journal of Finance, 45 (December), 1471-1493.

Manzon,G.and W.R. Smith 1994, “The Effect of the Economic Recovery Tax Act of 1981 and the Tax Reform Act of 1986 on the Distribution of Effective Tax Rates. ” Journal of Accounting and Public Policy, 13, No.4, pp. 349-362.

Marsh, P. R. 1982. “The choice between equity and debt: An empirical study”. Journal of Finance 37, (March):121-144.

Merton H. Miller. 1977. “Debt and Taxes”. Journal of Finance 32 (May), 261-275

Ministry of Financial Administration, 1976. Notes on Combinations on the Two Taxes, from the website of the Bureau of Financial Administration (Taiwan) (In Chinese).

Modigliani, F. F., and M. H. Miller. 1958. “The cost of capital, corporation finance, and the theory of investment”. American Economic Review 48 (June), 261-297

Myers, S. C., “Interaction of Corporate Financing and Investment Decision Implications for Capital Budgeting”, Journal of Finance, March, 1974, pp. 1-25.

Myers, S. C. 1977. “Determinants of corporate borrowings. Journal of Financial Economics 5 (June), 47-175

Myers, S. C. 1984. “The capital structure puzzle”. Journal of Finance 39 (July), 75-592.

National Taxation Bureau, Northern Taiwan Province, Ministry of Financial Administration, April 1976, Introduction to the Combination of Two Taxes (In Chinese).

Porcano, T.,1986, “Corporate Tax Rate: Progressive, Proportional ,or Regressive, ”The

- 31

Journal of the American Taxation Association/ Spring (7) 2,17-31.

Qiu, Yilun, 1978. The Relationships of the Corporate Financing Methods and Growth Opportunities—Case Study on the Self-selection Nominal Modeling. Master’s Thesis of the Institute of Economy, National Taiwan University (In Chinese).

Qiu, Shuwan, 1977. Research on the Impacts of Combination of the two taxes into one on corporate operating capital costs. Master’s Thesis of the Institute of Financial Management, National University of Transportation (In Chinese).

Schulman, C.T., D.W. Thomas, K. F. Sellers& Duane B Kennedy. 1996. “Effects on tax integration and capital gains tax on corporate leverage”. National Tax Journal 49 (March), 31-54.

Shevlin, T. and S. Porter .1992, “The Corporate Tax Comeback in 1987: Some Further Evidence.” The Journal of the American Taxation Association, 14, No.1, pp. 58-79

Siegfried, J., 1974, “Effective Average U.S. Corporation Income Tax Rates” National Tax Journal, 27, No.2, pp. 245-59

Stickney and McGee, 1982, “Effective Corporate Tax Rates: The Effect of Size, Capital Intensity, Leverage, and other Factors.” Journal of Accounting and Public Policy, 1, No.2, pp. 125-52

Titman, Sheridan, “The effect of Capital Structure on a Firm's Liquidation Decision”, Journal of Financial Economics, 13, March 1984, pp. 137-151.

Titman, D., and R. Wessels, 1988. “The determinants of capital structure choice” How? Journal of Finance 43 (March), 1-19.

Trezevant, R. 1994. “How did firms adjust their tax-deductible activities in response to the Economic Recovery Tax Act of 1981? National Tax Journal 47 (June): 253-271. ”

Wald, J. K., 1999, “How Firm Characteristics Affect Capital Structure: An International Comparison.” Journal of Financial Research, 22, 161-87.

Wen, Yuwen, 1980. Research on the additional 10% Income Taxes on Undistributed Profit for Profitable Organizations. Unpublished Master’s Thesis of the Institute of Accounting, National Jiang Cheshek University (In Chinese).

Xue, Mingzheng, and Chanjuan Lin, 1979. Research on the Correlations of the Free Cash Flows and Profit Management. Procceeding, Symposium on Theories and Practices of Accounting, 2001, Private Dongwu University (In Chinese).

Ye, Yinhua, and Xianbi Qiu, 1974. Case study on the correlations of capital structures, shareholders’ structure and company values—agency costs theory. Management Seminars of Taiwan University, 7 (2), (August), 57-90 (In Chinese).

Zang, Shiwai, Jingming Shang, and Yiqing Lin, 1978. Theories on the capital structures of companies and combining two taxes into one. Research of Finance and Taxes, 32 (3) (May): 1-27 (In Chinese).

Zheng, Dingwang, Intermediate Accounting (Vol. 2) , Edition No. 6, January 1987 (In Chinese).

Zimmerman, 1983, “Taxes and Firm Size.” Journal of Accounting and Economics, 5 , No.2, pp. 119-149.

- 32

Table 1: Treatment of Dividends in Different Countries

Complete credit

system

Partial credit

system

Dividend-exemption

system

Dividend-credit

system

Hybrid system

Singapore

Malaysia

Australia

New Zealand

France

Finland

Italy

Norway

Ireland

Portugal

Spain

UK

Canada

Greece

Luxemburg 3

Hong Kong

Denmark 1

Hungary

Japan

Germany 2

Iceland 4

Source: Websites of the ROC Finance Ministry, 1980

1. During 1977-1991, Denmark used the dividend-credit system. Since 1991, it has been adjusted to

30% and 40% tax rates for dividends. The two-to-one combination is no longer used.

2. The hybrid system in Germany uses a double track taxation system. For company tax brackets, it

uses the split-rate system; for shareholders, it uses the partial credit system.

3. In Luxemburg, individual shareholders have a 50% tax exemption, up to the limit of the dividend

from the profits distribution, after the company has paid income taxes. 50% or 100% exemption

might be applied for shareholders dividends.

4. In Iceland, the hybrid system uses a dividend-paid credit system for companies and a partial

dividend exemption system for shareholders.

- 33

Table 2: Corporate and Personal Income Tax Rates

State Corporate Income

Tax rate (%)

Personal Income

Tax rate (%)

Difference between the

two tax rates

Denmark 34 40 a 6

Finland 28 39 11

France 33.3 56.8 23.5

Germany 45 53 8 Greece 40 45 5

Hungary 18 b 48 30

Iceland 33 33.15 0.15 Ireland 38 48 10

Italy 37 51 14

Luxemburg 34.32 c 50 15.68

Norway 28 28 0

Portugal 36 40 4

Spain 35 56 21 UK 33 40 7

Singapore 26 28 2

Malaysia 30 30 0

Japan 37.5 50 12.5

Canada 29.12 d 29 -0.12

Australia 36 47 11

New Zeala nd

33 33 0

Source: Website of the Finance Ministry, 1976, Cited from Wanyu Huang (1978).

a. 30% tax rates when the dividend is less than 33,800 Denmark Kroner (DKK); 40% above that limit.

b. 23% supplement taxes applies in addition to company bracket for dividend distribution.

c. Corporate tax rate is 33%, plus 4% surtax.

d. Corporate tax rate is 28%, plus 4% surtax.

- 34

Table 3: Summary Statistics of the Variables

Panel A: Period1974 to 1975

Variables Average Standard

Deviation

Max Min

ETR 8.8237 10.4616 96.2700 0.0000

Debt Ratio 0.0871 0.0979 0.4977

Size 15.7936 1.0148 19.1188 13.3945

DEP/EBITDA 0.2984 0.9742 11.9486 -8.9932

VAR (EBT) 2.8125 13.9693 184.0615 0.0671

PM (EBT) 10.4179 20.8765 335.2700 -55.7200

CAPINT 0.2944 0.1696 0.8512 0.0034

RDINT 1.0852 2.1838 16.6800

Dividend/pre-tax net income 0.2603 1.2457 18.3987 -10.1486

Previous 5-year ETR Average 10.6914 7.1921 32.9080

VAR (BB) 1.3805 4.5817 74.3620 0.0506

CVA 0.3472 0.1806 0.9252 0.0058

GR 24.1033 121.9191 1841.1660 -19.4420

PM (NI) 9.5096 20.2051 341.2800 -47.4800

Panel B: Period 1976 to 1979.

Variables Average Standard

Deviation

Max Min

ETR 7.4962 12.6901 96.3700

Debt Ratio 0.1024 0.1004 0.5227

Size 16.0529 1.0634 19.6473 13.3823

DEP/EBITDA 0.5389 6.7122 188.8321 -45.6910

VAR (EBT) 3.9424 19.9558 473.1922 0.0594

PM (EBT) 0.0594 272.6144 159.1000

CAPINT 0.2862 0.1782 0.8914 0.0014

RDINT 1.1677 2.8834 54.5900

Dividend/pre-tax net income 3.1589 80.4553 2572.3214 -13.9520

Previous 5-year ETR Average 8.7619 7.1994 40.8640

VAR (BB) 2.9184 11.0509 217.2464 0.0463

CVA 0.3396 0.1894 0.9372 0.0018

GR 17.2177 102.0726 2038.7400 -21.6260

PM (NI) -16.8208 278.6434 135.7400 -8381.0900

- 35

Table 4: Average Debt Ratios and the ETRs for Several Industries Panel A: Period 1974-1975

Cement Foods Plastics Textiles

Electronic

machines Telecommuniations Chemicals Glass

Paper

Manufacturing

No. of

Companies 8 20 16 41 10 11 15 6 7

Debt Ratio 0.11 0.06 0.10 0.06 0.06 0.10 0.07 0.07 0.14

ETR 11.72 6.31 7.15 7.10 14.07 8.76 11.56 8.23 3.55

Steels Rubber Automotive Electronics Construction Transportation Tourism Retailing Others

No. of

Companies 15 8 2 35 19 12 5 8 19

Debt Ratio 0.15 0.08 0.07 0.11 0.07 0.15 0.01 0.07 0.08

ETR 11.93 9.19 11.02 5.38 6.90 12.55 28.29 6.36 11.62

Panel B: Period 1976-1979

Cement Foods Plastics Textiles

Electronic

machines Telecomm Chemicals Glass

Paper

Manufacturing

No. of

Companies 8 20 16 41 10 11 15 6 7

Debt

Ratio 0.15 0.11 0.13 0.09 0.10 0.09 0.10 0.05 0.12

ETR 7.60 5.92 6.49 4.70 11.60 5.26 10.27 4.55 5.72

Steels Rubber Automotive Electronics Construction Transportation Tourism Retailing Others

No. of

Companies 15 8 2 35 19 12 5 8 19

Debt

Ratio 0.13 0.07 0.03 0.14 0.06 0.15 0.08 0.09 0.08

ETR 8.45 7.89 14.43 6.13 6.63 9.83 12.49 12.39 11.42

- 36

Table 5A: Debt Ratio Average Variance

Debt Ratio before the 2-1-CTS Debt Ratio after the 2-1-CT S

Average 0.087054 0.10243

Known Variations 0.009586 0.010074

Observation Value No. 514 1028

Assumed Average variance 0

Z -2.88286682

P (Z ≤ z) single trail 0.001970436

Threshold Value: single tail 1.644853

P (Z ≤ z) double trails 0.003940872

Threshold Value: double tails 1.959961082

Table 5B: ETR Average Variance

Debt Ratio before the 2-1-CTS Debt Ratio after the 2-1-C T S

Average 8.823658 7.496235409

Known Variations 109.4441 161.039

Observation Value No. 514 1028

Assumed Average variance 0

z 2.183511

P (Z ≤ z) single trail 0.014499

Threshold Value: single tail 1.644853

P (Z ≤ z) double trails 0.028998

Threshold Value: double tails 1.959961

- 37

Table 6A: Estimated Coefficient for the ETR Regression Equation

Coefficient ETRb5 SIZE VARBR PMNI NDTS CVA GR

ETRb5 1.00000 -0.16416 -0.15572 0.06858 -0.03256 0.05614 -0.05194 SIZE -0.16416 1.00000 -0.02025 0.00592 -0.00514 -0.08691 0.00864

VARBR -0.15572 -0.02025 1.00000 -0.02309 0.04868 -0.00905 -0.00804

PMNI 0.06858 0.00592 -0.02309 1.00000 0.00275 0.03233 0.01340

NDTS -0.03256 -0.00514 0.04868 0.00275 1.00000 0.06157 -0.00314

CVA 0.05614 -0.08691 -0.00905 0.03233 0.06157 1.00000 -0.13654

GR -0.05194 0.00864 -0.00804 0.01340 -0.00314 -0.13654 1.00000

Table 6B: Estimated Coefficient for the Debt Ratio Regression Equation

Coefficient LEV SIZE VAREBT PMBT CAPINT RDINT LTREI

LEV 1.00000 0.47206 -0.01939 0.02192 0.25449 0.09672 0.03150 SIZE 0.47206 1.00000 -0.01900 0.00390 -0.00316 0.15510 0.02373

VAREBT -0.01939 -0.01900 1.00000 -0.00675 0.05305 -0.04444 0.00096

PMBT 0.02192 0.00390 -0.00675 1.00000 0.03701 0.01146 0.00128

CAPINT 0.25449 -0.00316 0.05305 0.03701 1.00000 -0.03722 0.02567

RDINT 0.09672 0.15510 -0.04444 0.01146 -0.03722 1.00000 -0.01075

LTREI 0.03150 0.02373 0.00096 0.00128 0.02567 -0.01075 1.00000

- 38

Table 7: Regression Analysis of the ETR Determinants

Coefficient Value t-Value P-value

Equation 1 a1 -13.5073 -2.31242** 0.020887

(ETR a1* -1.84835 -0.27002 0.787182

R2:

0.043998

a1+a1* -15.35565

Adj R2: a2 -0.85447 -2.50628** 0.012304

0.035233 a2* 0.086306 0.962107 0.336148

F Value: a2+a2* -0.768164

5.019818 a3 -0.08147 -2.17066** 0.030111

Significance a3* 0.064658 1.545459 0.122442

level: a3+a3* -0.016812

2.96E-09 a4 0.07233 2.861784*** 0.00427

a4* -0.06933 -2.73938*** 0.006227

a4+a4* 0.003

a5 5.889837 1.864339* 0.062466

a5* -5.25025 -1.38493 0.166278

a5+a5* 0.639587

a6 -0.29181 -1.2176 0.223564

a6* 0.161598 0.593163 0.55316

a6+a6* -0.130212

a7 0.066825 0.159252 0.873492

a7* -0.06837 -0.16293 0.870598

a7+a7* -0.001545

*, **, *** significant at 10%, 5? and 1? respectively.

Pooled sample size = 257

- 39

Table 8: Regression Analyses the Capital Structure Determinants

Equation 2 Coefficient Coeff Value t-value P-value

(Capital

Structure)

ß1 -0.00083 -1.57455 0.115567

R2: 0.30226 ß1* -0.00034 -0.53704 0.591319

Adj R 2:

0.295862

ß1+ß1* -0.00117

F value:

47.24949

ß2 0.046515 21.44535*** 2.14E-89

Significance:

1E-108

ß2* -0.00079 -1.08966 0.276037

ß2+ß2* 0.045725

ß3 -0.00108 -1.30925 0.190647

ß3* 0.001296 1.509314 0.131425

ß3+ß3* 0.000216

ß4 -0.0008 -3.20963*** 0.001357

ß4* 0.000808 3.24629*** 0.001195

ß4+ß4* 0.000008

ß5 -0.00456 -1.18558 0.235973

ß5* 0.004446 1.149148 0.250675

ß5+ß5* -0.000114

ß6 0.119026 5.712449*** 1.34E-08

ß6* 0.02948 1.184286 0.236484

ß6+ß6* 0.148506

ß7 0.0001 2.435062** 0.015003

ß7* -0.00011 -2.33163** 0.01985

ß7+ß7* -0.00001

*, **, *** significant at 10%, 5? and 1? respectively.

Pooled sample size = 257

- 40

APPENDIX: Symbols and Definitions of the Variables

Variable Definition

ETR Income taxes for the present period /pre-tax net profits LEV Long-term Debt/Total Assets

SIZE Total Assets

VAR

(EBT) Absolute value of the previous 5-year net profit variable coefficients CAPINT (Fixed Assets-Land)/Total Assets

RDINT R&D Expenses/Net Operating Income

LTREI Investing-Div idend Income / Pre-tax net profit NDTS Depreciation/EBITA

VAR (BB)

Absolute value of the previous 5-year operating income variable coefficients

CVA Total fixed Assets/Total Assets PM

(EBT) Pre-tax net income/ operating net income, or pre-tax net profit margin

PM (NI)

After-tax net income/ operating net income, or after-tax net profit

margin

GR Previous 5-year growth of average operating income