Relationship between tax avoidance and institutional ownership over business cost of debt

Abstract

Beginning with classical theories on finance, such as the capital structure theory, the trade-off theory of capital structure, and the pecking order theory, the literature shows a negative correlation between tax avoidance and institutional ownership with respect to the business cost of debt. However, the impact of tax avoidance and institutional ownership on corporate debt policy in Vietnam is an under-researched topic. The aim of the study is to identify the effect of those mentioned factors on business borrowing policy, using data on 207 companies listed on the Ho Chi Minh City Stock Exchange (HOSE) in Vietnam from 2008 to 2016. The study employs model proposed by Lim in 2009 to achieve mentioned research object with Feasible Generalized Least Squares (FGLS) method to overcome for any defection. The study results show no conclusive empirical evidence of a relationship between business’s cost of debt and tax avoidance and institutional ownership. This result contrasts with the conclusion in previous studies and can be explained by the characteristics of the funding market in Vietnam where financial organizations often focus on business results and management efficiency in making lending decisions and this characteristic is at no sign of change soon.

Keywords:

• Cost of debt
• institutional ownership
• tax avoidance
• HOSE
• HNX

JEL Classification:

• G23
• G32
• H26

PUBLIC INTEREST STATEMENT

The impact of tax avoidance and institutional ownership on corporate debt policy in Vietnam is an under-researched topic. The paper aim is to identify the effect of tax avoidance and institutional ownership on cost of debt for listed firms in Vietnam borrowing. The study uses model proposed by Lim (2009) for the research goal.

There is no conclusive empirical evidence of a relationship between business’s cost of debt and tax avoidance and institutional ownership. In contrasting previous studies, the findings can be explained by the characteristics of the funding market in Vietnam where financial organizations often focus on business results and management efficiency in making lending decisions.

1. Introduction

Businesses take advantage of the current tax regime to lower their tax payments by reducing their taxable income (Noor et al., 2009) and thereby increase current profits as well as the company’s after-tax value (Chung et al., 2002; Noor et al., 2009; Salehi et al., 2019). However, tax avoidance can reduce company value in cases where the costs are directly related to a firm’s tax planning costs, such as adaptation costs and agency costs (Fuadah & Kalsum, 2021; Wang, 2010). According to Graham and Tucker (2005), savings from tax avoidance can be considered in making financial plans, as it is a form of funding that reduces a business’s dependence on external borrowing. In addition, tax avoidance increases financial flexibility, thereby increasing credit quality, reducing bankruptcy risk (Lim, 2009), and lowering a business’s average cost of capital (Monila, 2005).

Tax avoidance behaviour can also represent the subjective actions of the manager of a business for personal purposes. This means that tax avoidance can increase information asymmetry at businesses. Chung et al. (2002) shows that increasing the ownership ratio of institutional shareholders can improve the quality of corporate governance and limit profit manipulation by means of the accounting method. Some empirical research (Utkir, 2012) confirms the controlling effect of institutional shareholders in the Malaysian market, but other literature (Lim, 2009; Sunarto & Widjaja, 2021) shows the opposite in the Korean and Indonesian market. There is also research showing that at well-managed companies, measured by the degree of institutional ownership, tax avoidance has a favourable impact on corporate value (Desai & Dharmapala, 2009). In sum, no consensus has been reached in the debate over managerial opportunism and tax avoidance.

Overall, tax avoidance can help businesses reduce the cost of debt by temporarily taking advantage of saving on the amount paid to the state. Also, it might include the issue of representative costs because of the separation between management and ownership, as a result, tax avoidance may serve the personal needs of the manager. However, to date, no studies have examined the relationship between tax avoidance and institutional ownership with respect to the debt policy of businesses in Vietnam. Therefore, we examine this relationship at companies listed on the HOSE in Vietnam.

2. Literature review and hypotheses development

From 1986, with the remark reform usually referred as Doi Moi, the Vietnamese market has emerged from the socialist centrally planned model. Over fifty years, substantial changes have help Vietnam to attract vast amount of investment from the around the globe. However, a gap between the development process of Vietnam and the international community understanding about the country exists because there is limited research volume on Vietnam economic. One of aspect is the Vietnam financial market mechanism.

Hanlon and Heitzman (2009) define tax avoidance as the reduction in tax per currency unit of pre-tax accounting profit. Tax evasion, by contrast, is defined as the transfer of value from the government to shareholders (Desai & Dharmapala, 2009). The difference between taxable income and accounting income is affected by many different factors in two main systems: financial accounting standards and tax rules. Financial accounting standards adhere to certain fundamental principles set by the GAAP (Generally Accepted Accounting Principles), which help to describe financial transactions and to provide useful information for relevant stakeholders. Tax rules, however, are determined by political conditions, as legislators enact tax laws to increase the state’s income from taxes, encouraging or discouraging certain activities in the economy.

Many studies have explored why some businesses avoid taxes more than others. Researchers approach this question from different perspectives. For example, some arguments are based on business characteristics, the field of operations, size, or the age of the business. Others explain it based on ownership structure and organizational characteristics (Desai & Dharmapala, 2004, 2009, 2011; Graham & Tucker, 2005). An increase in tax avoidance leads to two perceptions of the consequences of such actions: first, tax avoidance is interpreted as increasing other tax incentives; second, it involves agency costs, which managers can use as a tool to cover up opportunistic behaviour. In the first perspective, according to Graham and Tucker (2005), dodging taxes is the act of taking advantage of tax incentives, such as using debt. This view suggests that tax avoidance can be an alternative for external borrowing, so it should have a negative relationship with the cost of debt, and this relationship can be stronger with a higher level of institutional ownership (Lim, 2009). The second perspective emphasizes the correlation between tax avoidance and agency costs, and tax avoidance can be a cover for actions that divert real profits to managers.

Tax avoidance is the act in which businesses take advantage of legal provisions to minimize the amount of tax paid, whereas tax evasion involves providing false information to limit the amount of tax paid (Sandmo, 2005). Because tax evasion is illegal, taxpayers who try to avoid taxes in this way are concerned about the possibility of their actions being discovered. Tax avoidance, in contrast, is a legally sanctioned activity, using tax provisions to reduce their tax liability by converting labour income into capital income to take advantage of lower tax rates.

The theory of the company’s capital structure introduced by Modigliani and Miller in 1958 can be summarized as two cases with and without the effect of taxes related to firm value and capital cost. In the case of no taxation, the value of the company with debt is equal to the value of the firm without debt. Meanwhile, the average cost of capital is constant, regardless of changes in the capital structure when the firm owes no taxes. If taxes are owed, the value of a company employing debt is equal to the value of the debtless company plus the present value of a tax shield (a reduction in taxable income attained through allowable deductions from charity donations, amortization and depreciation). With respect to the cost of capital, if taxes are owed, this theory holds that the required return on equity also increases with increasing use of financial leverage. The benefit from a tax shield helps to reduce the Weighted Average Cost of Capital (WACC). However, the soaring required rate of return on equity, as an increase in the use of financial leverage, triggers equity risk. The capital structure of Modigliani and Miller (1958) is reviewed in this study to explain why businesses do not use the maximum debt to gain benefits from the tax shield.

Kraus and Litzenberger (1973) propose using the trade-off theory of capital structure to explain why businesses are often financed in part by both debt and equity. In this theory, they suggest that the use of debt financing is costly, most notably the cost of financial exhaustion. Therefore, businesses cannot fully finance with loans. Oddly, for every additional percentage increase in debt, the benefit of the tax shield increases, and so does the cost of financial exhaustion. When the present benefit from the tax shield does not exceed the cost of financial exhaustion, borrowing no longer benefits the business. Because of this, companies always seek to optimize their total business value based on this equilibrium principle to determine how much debt and how much equity are optimal for their capital structure. Some authors rely on capital structure theory in studying the relationship between tax avoidance and debt costs (Bhojraj & Sengupta, 2003; Desai & Dharmapala, 2009; Graham & Tucker, 2005; Lim, 2009; Nguyen Minh et al., 2021).

Jensen and Meckling (1976) introduce the concept of agency costs as the aggregated costs of an organized contract. Originating in the distinction between ownership and management at companies, managers are often at better understanding the true value of assets and current and potential risks, hence, causing information asymmetry. In addition, decentralization in business can have consequences. For instance, managers directly run business activities, so they can take actions to maximize their personal benefits. However, because of asymmetric information, managers can make decisions that harm the interests of investors. In this study, agency costs arising from information asymmetry explain how managers implement tax avoidance for personal gain, leading to the risk that it will reduce business creditworthiness and increase the cost of debt.

The pecking order theory (Myers & Majluf, 1984) argues that firms prefer internal sources of funding. Myers and Majluf (1984) showed the priority, in descending order, of corporate funding as follows: (1) retained earnings, (2) direct borrowing, (3) convertible debt, (4) ordinary shares, (5) non-convertible preference shares, and (6) convertible preference shares. This order helps to explain why businesses consider tax avoidance an internal resource that can be used to minimize external funding.

According to Bhojraj and Sengupta (2003), the business cost of debt can be affected by its characteristics, such as bankruptcy risk, agency costs, and information asymmetry. Therefore, businesses often avoid tax to increase their financial surplus and improve their credit quality, thereby reducing the cost of debt. Graham and Tucker (2005) study 44 enterprises with tax avoidance behaviour in the period 1975–2000, with similar results. They show that businesses often use tax avoidance to replace debt usage and reduce the cost of borrowing. Based on these research results, we propose the following hypothesis:

H1: Tax avoidance is negatively correlated with the cost of debt.

Desai and Dharmapala (2009) construct a model that shows the correlation between tax avoidance and profit-distorting behaviour. To hide tax avoidance behaviour from tax authorities, managers can take actions that limit shareholder control. Ashbaugh-Skaife et al. (2006) explain the cost of debt based on the agency theory. Accordingly, creditors might be disadvantaged by information asymmetry caused by the behaviour of managers or shareholders who take advantage by transferring lenders’ assets to themselves. In addition, institutional ownership is negatively correlated with tax avoidance, which is explained by Desai and Dharmapala (2009). Institutional shareholders can use their dominance to limit managerial tax evasion behaviour, while also limiting information asymmetry. This is also the result in Chung et al. (2002). Based on their research, institutional ownership can influence the cost of indirect debt because it is believed that the higher proportion of institutional ownership of a business lowers the agency cost and the cost of debt. Bhojraj and Sengupta (2003) and Nguyen Minh and Hiep (2019) provide empirical evidence on the direct effect of institutional ownership on the cost of debt at a firm. Companies with high institutional ownership often have lower debt costs due to higher credibility. Hereby, we propose our second hypothesis as follows:

H2: The proportion of institutional ownership is negatively correlated with the business cost of debt.

3. Research design

3.1. Research model

To test H1 and H2, we use the model proposed by Lim (2009) as follows:

$CO{D}_{i,t}={\alpha }_1+{\alpha }_{2}BT{D}_{i,t}+{\alpha }_{3}T{A}_{i,t}+{\alpha }_{4}INS{T}_{i,t}+{\alpha }_{5}AG{E}_{i,t}+{\alpha }_{6}LEVERAG{E}_{i,t}+{\alpha }_{7}CF{O}_{i,t}+{\alpha }_{8}SIZ{E}_{i,t}++{\epsilon }_{it}$

Table 1 shows a brief summary for variables of the research model. Also, the expected correlation with independent variables are showed in Table 2 together with relevant previous studies in the same subject matter.

Table 1. Summary of research variables

Variable code	Type of variable	Definition	Measurement
COD	Dependent	Cost of debt	Interest expenses divided by business total debt
BTD	Independent	Book-tax difference	The difference between reporting revenue and the implied revenue derived from the tax payable and the corresponding tax rate
TA	Independent	Total accrual	As mentioned in Desai and Dharmapala (2009) study to measure tax avoidance on accruals, based on the argument that tax savings may come from other purposes of income management and tax avoidance.
INST	Independent	Institutional ownership	The percentage of share held by large financial organizations
AGE	Control	Business age	Study year—business founded year
LEVERAGE	Control	Leverage ratio	Total liabilities divided by total assets
CFO	Control	Cash from operations	Total cash receipts from operating activities divided by total assets
SIZE	Control	Size of business	Logarithm of total assets

Table 2. Summary of expected correlation of the variables

Variables	Expected correlation with dependent variable	Previous studies
BTD	-	Bhojraj and Sengupta (2003), Graham and Tucker (2005), Desai and Dharmapala (2009)
TA	-	Bhojraj and Sengupta (2003), Graham and Tucker (2005), Desai and Dharmapala (2004)
INST	-	Ashbaugh-Skaife et al. (2006), Bhojraj and Sengupta (2003), Desai and Dharmapala (2004)
AGE	-	Lim (2009)
LEVERAGE	+	Petersen and Rajan (1994)
CFO	+	Lim (2009)
SIZE	-	Carey et al. (1993)

3.2. Data collection

Secondary data, including financial statements of listed companies, is collected from vietstock.com. The study period is from 2008 to 2016. To ensure uniformity in the data, we omitted businesses with special financial characteristics, consisting of finance and insurance businesses, banks, real estate companies, companies whose financial information was not disclosed during the study period, and businesses with negative income tax due. The final panel data is with 207 enterprises and a total of 1,863 observations.

Tax rates are an important part of determining the implied revenue from tax paid. However, from 2008 to 2016, Vietnam corporate income tax experienced many fluctuations as detailed in Table 2.

4. Empirical results and discussion

4.1. Statistical description of the data

In Table 4, the cost of debt (COD) averages 0.054%; the difference between reporting revenue and implied revenue derived from the tax paid and the corresponding tax rate (BTD) averages VND 265.649 billion, with a maximum of VND 6,608.416 billion and a minimum of VND −31.608 billion; the average business total accrual (TA) is VND 28.838 billion; the highest is VND 27,860 billion, and the lowest is VND −36,162 billion; the average proportion of institutional ownership (INST) of businesses listed on the HOSE is 0.24%, and some businesses have no institutional ownership. Business age (AGE), Leverage ratio (LEVERAGE), cash from operations (CFO), and sizes (SIZE) in the sample average 26.299, 48.1%, 0.064, and 27.774 respectively. The tax rates in Vietnam for the research period are showed in Table 3.

Table 3. Summary of tax rates in Vietnam for the research period

Effective date	Business affected	Tax rate	Source
Before 1 July 2013	All businesses unless specifically excluded.	25%	Circular 123/2012/TT-BTC issued 2012,
	Businesses in oil and gas prospecting, exploration, and exploitation in Vietnam.	32%–50%
	Businesses in searching, exploring, and exploiting rare and precious mineral resources, except oil and gas.	50%
	Business in rare and precious resource industry with 70% or more of allocated mines in economically disadvantaged areas.	40%
From 7 January 2013, to 31 December 2013	Businesses established under Vietnamese law with annual turnover not exceeding VND 20 billion.	20%	Circular 121/2013/TT-BTC issued 2013,
	Businesses with annual turnover exceeding VND 20 billion.	25%
	Businesses established after 1 July 2013 (except for those eligible for tax incentives)	25%
From 1 January 2014, to 31 December 2015	Businesses established under Vietnamese law, including with annual turnover not exceeding VND 20 billion.	20%	Circular 78/2014/TT-BTC, issued 2014,
	Businesses with annual turnover exceeding VND 20 billion.	22%
Since 1 January 2016	All businesses except for those eligible for tax incentives.	20%	Circular 96/2015/TT-BTC, issued 2015,

Table 4. Descriptive statistics for variables in the model (N = 1,863)

Variable	Unit of measurement	Mean	Maximum	Minimum	Standard deviation
COD	%	0.054	2.649	0.065	0.084
BTD	Bill. VND	265.649	6,608.416	−31.608	682.707
TA	Bill. VND	28.838	27,860	−36,162	2243.765
INST	%	0.241	30.372	0	0.732
AGE	Year	26.299	89	8	13.757
LEVERAGE	%	48.1	97.0	0.1	0.228
CFO	Times	0.064	1.038	−0.798	0.141
SIZE	Logarithm	27.774	34.545	20.215	1.586

Explanation for these variables is at section 3.1 of this paper.

4.2. Correlation and the variance inflation factor (VIF)

The results verified the correlation coefficients between the variables in the proposed model to test the likelihood of multicollinearity (Table 4).

According to these test results, none of the variables have a VIF greater than 5 (Table 5) and no pairs of variables have excessively high correlation, and the correlation coefficients are less than 0.5. Only the pair SIZE and BTD are highly correlated, with a coefficient of 0.679 (Table 6). It is indicating that the model has a low likelihood of multicollinearity.

Table 5. Variance inflation factor

Variable	VIF	1/VIF
SIZE	2.30	0.434
BTD	1.95	0.513
LEVERAGE	1.39	0.719
CFO	1.12	0.889
TA	1.03	0.968
AGE	1.03	0.969
INST	1.00	0.995
Mean VIF	1.40

Table 6. Correlation matrix

	BTD	TA	INST	AGE	LEVERAGE	CFO	SIZE
BTD	1.000
TA	−0.069	1.000
INST	0.037	−0.027	1.000
AGE	0.126	−0.051	−0.002	1.000
LEVERAGE	0.243	−0.077	0.025	0.072	1.000
CFO	0.030	−0.117	0.031	0.045	−0.271	1.000
SIZE	0.679	−0.087	0.048	0.039	0.467	−0.116	1.000

4.3. Regression results

To determine which model to use, we carried out OLS (Ordinary Least Squares), REM (Random Effects Model), and FEM (Fixed Effects Model) tests. The result the Breusch and Pagan Lagrangian tests was Prob > chi2 = 0.0000, indicating that the REM model is better than OLS (Table 7). The results of the Hausman test, Prob > chi2 = 0.165, show that FEM is a better fit for our proposed research model (Table 8). However, the Wald test (Prob > chi2 = 0.0000) result shows evidence of heteroskedasticity (Table 9). This defect in the model is addressed using the FGLS (Feasible Generalized Least Squares) method. The regression results are in Table 10.

Table 7. Breusch and Pagan Lagrangian multiplier test for random effects

	Var	sd = sqrt(Var)
cod	0.007	0.084
e	0.005	0.073
u	0.001	0.038
Test: Var(u) = 0 chibar2(01) = 327.44 Prob > chibar2 = 0.0000

Table 8. Hausman test

Test: Ho: difference in coefficients not systematic

chi2(4) = (b-B)’ [(V_b-V_B) ^ (−1)] (b-B)

= 6.49

Prob>chi2 = 0.165

Table 9. Modified Wald test for groupwise heteroskedasticity in fixed effect regression model

H0: sigma(i)^2 = sigma^2 for all i

chi2 (207) = 5.7e+08

Prob>chi2 = 0.000

Table 10. Regression results of the FGLS method

Variables	Coeff.	Std. Error	P>|z|	[95% Conf. Interval]
BTD	0.000	0.000	0.791	−0.000	0.000
INST	0.000	0.002	0.850	−0.004	0.005
TA	0.000	0.000	0.755	−0.000	0.000
AGE	0.000	0.000	0.783	−0.000	0.000
LEVERAGE	−0.002	0.017	0.889	−0.037	0.032
CFO	0.014	0.014	0.321	−0.013	0.041
SIZE	−0.006	0.003	0.092	−0.013	0.000
Cons	0.217	0.095	0.023	0.029	0.405

4.4. Discussion

4.4.1. Tax avoidance and the cost of debt

Business book-tax different (BTD) and total accrual (TA) are applied to measure for tax avoidance behaviour. The regression results in Table 10 indicate the absence of an analytical basis for confirming the correlation between tax avoidance measurement and business cost of debt (COD). Therefore, there is not enough evidence to support H1. This result is inconsistent with the results of Desai and Dharmapala (2009), who stated that tax avoidance is a funding source for business activities and reduces external borrowing. Therefore, it helps businesses reduce COD in two ways. First, it reduces COD by reducing debt financing. Second, the use of less debt helps businesses improve their credit rating in the eyes of creditors, such as commercial banks, thereby reducing the cost of using debt. However, in Vietnam, the use of capital from tax avoidance does not really have an impact on helping to improve the credit rating of these businesses. The result is also in contrast with Bhojraj and Sengupta (2003), Monila (2005), and Lim (2009) suggesting tax avoidance increases business credit quality lowers business’s average cost of debt.

4.4.2. Institutional ownership and the cost of debt

The regression result is insufficient to confirm H2 (P > | z | = 0.850), which demonstrates that institutional ownership (INST) is not a factor in reducing the cost of borrowing in Vietnam. This outcome contrasts with that of Desai and Dharmapala (2004), who found that the greater the organizational ownership, the lower the cost of representation, and the greater the transparency of companies, thereby improving credit ratings and reduce COD. By this, it suggests that banks and credit institutions in the Vietnamese market do not consider the ownership structure of businesses in their evaluation of creditworthiness. Again, this result implies the characteristic of those institutions in fund providing decision making.

In addition, we do not find any correlation between the control variables (TA, AGE, LEVERAGE, CFO, SIZE) and COD for the firms in the sample.

5. Summary and conclusion

The study focuses on identifying the role of specific business factors and the degree of impact of each factor on debt policy using REM with FGLS estimation in our proposed research model. The result is that tax avoidance and institutional ownership have no statistically significant effect on a firm’s COD. This confirms the approach of banks and credit institutions in Vietnam in customer evaluation. More specifically, these lenders often do not view the use of tax avoidance capital and ownership structure as indicators of creditworthiness. Instead, these organizations often focus on business results and management efficiency in making lending decisions. This characteristic of Vietnam lending industry is at no sign of change soon. Also, the study period is only up to 2016. However, as the stability lending industry in Vietnam is certain, it is arguable that extension of research period would not yield any significant different result. Also, the reverse causality, with cost of debt driving tax avoidance, is excluded from the paper as previous research demonstrating other way around (Graham & Tucker, 2005; Lim, 2009). The paper only points out some specific internal characteristics of the enterprise and considers the impact of these characteristics on the cost of debt. In fact, there will be many other factors that can affect the debt policy of an enterprise, including external factors: macro policies on interest rates, inflation, etc. The suggestion is to include external factors in further studies.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

The authors received no direct funding for this research.

Notes on contributors

Pham Minh Vuong

Nguyen Minh Ha is the rector of Ho Chi Minh City Open University, Vietnam. His research interests include issues on finance and economic. He has published in various international journals.

Pham Minh Vuong is a lecturer Ho Chi Minh City Open University, Vietnam. His research interests are accounting and finance issues.

Tran Thi Phuong Trang graduates from Ho Chi Minh City Open University, Vietnam. She is currently working in a commercial in Vietnam.