Capital structure and firm value of Vietnamese non-financial listed companies A panel threshold regression analysis

Conditional on having financing needs, firms prefer external equity when the relative cost of equity is low and prefer debt otherwise.Therefore, this theory explained that capital stru[r]

Capital structure and firm value of Vietnamese non-financial listed companies:

A panel threshold regression analysis

Trần Thị Phương Thảo (MSc) Nguyễn Thúy Anh (Doctor) Abstract

The study is aimed at exploring the impact of capital structure on firm performance of Vietnamese non-financial listed companies A threshold regression model estimated on a sample of 122 non-financial companies listed in HOSE and HNX for the period of

2011-2015 confirms that leverage is found to have non-linear effects on firm value.On the other hand, the relationship between debt ratio and firm performance varies in accordance with different changes in debt structure In addition, the findings suggest implications for firms

on flexible usage of financial leverage

Keywords: Panel- threshold effect, Debt ratio

1 Introduction

Capital structure refers to the way in which the company’s assets are financed through a combination of equity and debt for the company’s activity.Managing capital structure properly is important as it would affect the company value.Study the effect of capital structure on firm value will help the enterprises making the decisions of capital restructuring more suitable with a view to maximizing firm value In many studies on capital structure and company value, it is common to apply a linear relationship analysis

In addition, there is no single theory that can fully interpret the impact of capital structure

on the profitability and finally the firm value The aim of study is to apply Hansen’s (1999) advanced panel-threshold regression model to test whether thereisan “optimal” debt ratio which causes there to bethreshold effects.A sample of 122 Vietnamese non-financial listed firms which total asset is over 1000 billion VND during 2011-2015 is conducted in this analysis.We select only large companies in this study because have stable growth and get less effect from economic distress

The tasks of the research are as follows:

- To overview literature on capital structure theories;

- To review previous empirical studies which analyzed a non-linear relationship between capital structure and firm value;

- To apply the panel-threshold regression model to find a non-linear relationship on

a sample of 122 Vietnamese listed companies;

- To conclude and make recommendations for improvement of the use of financial leverage

2 Literature review and Hypotheses

Capital structure theories explain the mix of debt and equity used by firms, determinants

of capital structure and the relationship between capital structure and firm value Major theories underpinning this issue are MM, agency cost, trade-off, pecking order and market timing theory

M&M theory

M&M theory, [ CITATION Mod58 \l 1033 ] based on restrictive assumption of a perfect capital market and stated that the capital structure is irrelevant Firm value will be affected by its own assets, not by any mixture of debt and equity and the optimal capital structure is not existed

However,Modigliani, F and Miller, M (1963)explained that the high debt level in their capital structure leads to lower tax debts and more cash flow after tax, which might increase the market value.An optimal capital structure exists when the company balances the risk of bankruptcy with the tax savings of debt

The Agency theory

The agency theory is initially developed by Berle, A.A and Means, G.C (1932), discovered that managers pursue their own interest instead of maximising returns to the shareholders Jensen and Meckling (1976)demonstrated that there are two kinds of agency costs The agency cost of equityarises because ofthe difference of interest between shareholders and managers and the agency cost of debt is caused by different interests of shareholders and debt holders [CITATION MJe86 \l 1033 ] claimed that with high debt, managers are under pressure to invest in profitable projects to create cash flow to pay interest In other words, debt has a positive effect on a firm’s value

The Tradeoff theory

Myer (1977) explained that a firm will trade off the costs and benefits of debt associated with tax savings and financial distress to create an optimal capital structure for maximizing firm value If the leverage is increased and the tax benefits of debt increase as well and the cost of debt also go up This trade-off theory predicts that target debt ratios will vary from company to company High target debt ratio should applied in companies with safe, tangible assets and profitable In contrast, risky, intangible assets and unprofitablecompanies ought to rely primarily on equity financing [ CITATION Ala73 \

l 1033 ]

The pecking order theory

According ot the pecking order theory, formalized by Myers and Majluf (1984), firms seeking to finance new investments follows hiearchy: first internal funds, then debts issuance and finally equity issuance Retained earning are better than outside funds and debt is better for firms than equity if the firm needs external funds Issuing equity becomes more expensive as asymmetric information insiders and outsiders increase so that firms should issue debt to avoid selling under-priced securities Moreover, transaction costs in obtaining new external funds are higher than the costs of obtaining internal funds The pecking order theory predicts that the most profitable companies generally borrow less since they don’t need outside money, not because of low target debt ratios.Less profitable companiesdebt issue because they need external funds for their investment projects The theory does not deny factors such as tax savings and financial distress but they explained that these factors are less important than manager’s preference for internal over external funds In summary, this theory states that there is a negative relationship between leverage and profitability

The market timing theory

The study by [CITATION Bak \l 1033 ], stated that managers are able to time the equity issues Conditional on having financing needs, firms prefer external equity when the relative cost of equity is low and prefer debt otherwise.Therefore, this theory explained that capital structure decisions are influenced by the market conditions and there is no an optimal capital structure for maximize firm value

Empirical researches

A number of empirical works on capital structure and firm performance provides mixed and contradictory results, and little research has been conducted on emerging or transition economies While most studies conducted in developed countries found a positive relationship between capital structure and firm performance, some studies investigating this relationship in emerging markets have found a negative relationship between capital structure and firm performance Although empirical researchs demonstrated the linkage, few studies has been done on the evidence of non-linear connection between leverage and firm value

In the study of Cheng, Y et al (2010), an advanced panel threshold regression model of

650 Chinese companies over the period from 2001 to 2006 is applied The results confirm that a triple-threshold effect does exist and show an inverted-U correlation between leverage and firm value

Feng-Li Lin; Tsangyao Chang (2011) used a panel of 196 Taiwanese listed companies during the 13-years (1993-2005) period The results substantiate that there is a double threshold effect between debt ratio and firm value When the debt ratio is less than 9.86%, Tobin’Q (proxy for firm value) increases by 0.0546%, with an increse of 1% in the debt ratio When the debt ratio is between 9.86% and 33.33%, Tobin’Q increases by only 0.0057% but when the debt ratio is greater than 33.33%, there is no relationship between debt ratio and firm value

Abd Halim Ahmad and Nur Adiana Hiau Abdullah (2013) investigate the effect of leverage on Malaysian listed firm’s value and indicate that debt is only pertinent to the firm value up to a threshold level of 64.33 percent Additional debt beyond the threshold level does not add to a firm’s value

The literature involving Vietnamese firms is mostly concerned with factors affecting capital structure and a linear relationship between capital structure and firm value In this paper, we apply Hansen’s (1999) advanced panel-threshold regression model to determine whether there is a threshold debt ratio

3 Methodology

Data set includes 122 non-finance listed enterprises in Hochiminh Stock Exchange-HOSEand Hanoi Stock Exchange-HNXfrom 2011- 2015 Financial institutions and insurance firms are excluded since the accounting presentations are different from those

in the other sectors Following the above sample selection process, a total of 610 observations are collected We obtain all year-end financial accounting indicators from the Stoxplus

Following [ CITATION Abd13 \l 1033 ], [ CITATION Fen11 \l 1033 ]we use return

on equity (ROE) and Tobin’s Q to represent the firm value Two categories of dependent variables in the panel threshold estimation of Hansen (1999) In the first category, we use capital structureas the threshold variable to determine whether there is an asymmetric threshold effect of capital structure on the firm value The second category of variable is used to control for other factors including firm size, tangibility, liquidity, growth opportunities, business risk, and cash holding The variables are defined in Table 1

Table 1: Variables used in this study

ROE Return on equity Profit after tax/Total equity

Q Tobin’s Q Total market value of firm/ Total asset value

LEV Capital structure Total debt/ Total assets

SIZE Firm size Natural logarithm of the total assets

TANG Tangibility Fixed assets/ Total assets

LIQ Liquidity Current Assets/ Current Liabilities

GROW Growth Opportunities Percentage change in net sale

Table 2 shows the descriptive statistics of 122 firms included in the sample from 2011 to 2015

Table 2: Sample description Variable Observation

s

Median Standard

Deviation

Minimum Maximum

A panel threshold regression model proposed by [ CITATION Han99 \l 1033 ]is applied The single threshold model is set up as follows:

Firm_valueit = {µi+β ' hit + α 1 LEV it+ ε it if LEV it ≤ γ

Where ROEitand Qit are selected to proxy for Firm_valueit, LEVit is the explanatory

variable and also the threshold variable, γ is the specific estimated threshold value We established four control variables hit which may affect firm performance, including firm size, tangibility, liquidity and growth opportunities The errors εit are assumed to be independent and identically distributed with mean zero and finite variance Depending on whether the actual leverage is smaller, equal to, or larger than the threshold value ( ) to

be estimated, the regimes are distinguished by the different regression slopes, α1 and α2 Furthermore, the model is modified if double thresholds exist The respective formula can

be shown as:

Firm value = {µ i+β µi+β ' hit +α 2 LEV it +ε it if γ 1< LEV it ≤ γ 2 ' hit + α 1 LEV it +ε it if LEV it ≤ γ 1

(2)

The model can be extended to multiply threshold using the same process where the threshold value γ1< γ2, γ2< γ3, γ3< γ4… γn

4 Empirical results

To explore the relationship between leverage and firm value, we set the data series in panel form and apply panel unit root test to determine whether or not the variables in the model are stationary The panel unit root tests of [ CITATION Lev02 \l 1033 ] [ CITATION ImK03 \l 1033 ][ CITATION Dic79 \l 1033 ] Based on the results of stationary test in table 3, we find that almost the variables have stationary characteristics since the nulls of the unit root are mostly rejected This allows further analysis of the panel threshold regression

Table 3: Panel unit-root test results

ROE -67.2160 (0.0000) -22.8396 (0.000) 47.3109 (0.000)

Q -21.0200(0.0000) 3.6113 7.6930(0.000)

LEV -15.4300 (0.0000) -1.9220 (0.0273) 14.1088 (0.000)

SIZE -29.5697 (0.0000) 7.218 8.9181 (0.000)

TANG -8.5634 (0.0000) -6.3426 (0.000) 24.3647 (0.000)

LIQ -7.7724 (0.0000) -8.5125 (0.000) 34.2570 (0.000)

GROW -28.0006 (0.0000) -32.6624 (0.000) 52.4403 (0.000)

To obtain the F-statistics follows with the p-values, bootstrap method is utilized The bootstrap procedure is applied and repeated for 300 times for each of the three panel threshold models The results for single, double and triple thresholds effect is shown in table 4

Table 4: Tests for the threshold effects between debt ratio and proxy variables for firm

value Threshold value F p-value Critical value of F

Single threshold effect test

ROEit0.8682

Tobin’sQit0.2569

62.6 0 26.8 6

0.0000**

* 0.0367**

31.0627 38.4478

19.8542 25.3499

16.1013 19.9157

Double threshold effect test

ROEit0.86820.8088

Tobin’sQit 0.2569 0.1113

16.0 6 19.8 9

0.1967 0.1267

85.7893 39.8669

75.0821 25.2014

56.7617 21.7168

Triple threshold effect test

ROEit0.86820.80880.2339

Tobin’sQit 0.2569 0.1113 0.1814

5.95 14.1 7

0.8567 0.4400

90.8391 45.0231

65.9327 36.4092

55.6887 27.7327

Regarding ROE as a proxy for firm value, the single threshold effects is significant (F-statistic 62.60 and p-value 0.0000) However, the test for double threshold effect and triple threshold are insignificant with the bootstrap p-value of 0.1967 and 0.8567respectively Therefore, we conclude that there is evidence of a single threshold effect of leverage on a firm’s performance for Vietnamese listed firms Table 5 presents the regression slope estimates for regimes

Table 5: Estimation of the coefficients for ROE as proxy for firm value

SIZE -5.64E-06 0.016101 0

LEV

The estimated model from our empirical results is represented as follows:

ROEit = µi – 0.13955 TANG + 0.068582 GROW - 0.15518 LEV (LEV ≤ 86.82%)-0.75275 LEV (LEV >86.82%) + εit

The regimes are distinguished by the different regression slopes In the first regime, where the debt ratio is less than 86.82%, the estimate of coefficient α1 = -0.15518 which is significant at the 5% level and indicates that ROE decreases by 0.15518% with an increase of 1% in debt ratio In the second regime, where the leverage is greater than 86.82%, the estimate of coefficient α2 = -0.75275 which is significant at 1% level and this means that ROE decreases by 0.75275% with an increase of 1% in debt ratio Thus, compare the “low-debt” and “high-debt”, we find that the “high-debt” slope is nearly five times (0.75275/0.15518) more than the “low-debt” slope.Our results are consistent with the pecking order theory and highlight a negative relationship between leverage and profitability In the period from 2011 to 2015, the borrowing interest rate was high, especially in 2011, the interest rate was up to 17-19%/year, which means the higher leverage the companies use, the higher interest expense they have, finally impact to performance In addition, annual changes in firm growth are significantly and positively related to ROE or firm value The interpretation here is that greater growth firms have higher value By contrast, tangibility is negatively related to ROE

Regarding Tobin’s Q as proxy of firm value, by using bootstrap to make 300 times, F-statistics of 26.86 and P-value of 0.0367 for single threshold effect show significance under 5% significant level The results for double and triple threshold effect reject the null hypothesis of two and three thresholds Table 6 presents the estimated value for Tobin’s

Q as a proxy for firm value

Table 6: Estimation of the coefficients for Tobin’s Q as proxy for firm value

SIZE 0.358537 0.037345 9.6***

LEV

The estimated model from our empirical results is represented as follows:

Qit = µi + 0.358537 SIZE – 0.23137 TANG + 1.117625 LEV (LEV ≤ 25.69%) + εit

In the first regime when the debt ratio is less than 25.69%, the estimated coefficient of Tobin’s Q (α 1= 1.117625) is positive and significant at the 1% level, indicating that Tobin’s Q increases by 1.117625% with an increase of 1% in debt ratio In the second regime, where the debt ratio is greater than 25.69%, the estimated coefficient of Tobin’s

Q (α 1= -0.03453) is negative and insignificant, which means that there is no relationship between debt ratio and firm value.We thus conclude that there exists an optimal debt ratio

is less than 25.69% that increases firm value These findings are consistent with the trade-off theory for which firm may search a “balance” that interest tax shield is equal to the incremental costs through debt financing In the estimations of the coefficients of the control variables, firm size is significantly positively related to Tobin’s Q buttangibility has negative impact on firm value (for both ROE and Tobin’s Q as proxy) By contrast,liquidity and growth opportunity are not significantly related to Tobin’s Q

Conclusions

This paper investigates the non-linear connection between capital structure and company value in 122 large Vietnamese listed firms which total assets is over 1000 billion VND In the first model, ROE is selected to proxy firm value, our results substantiate that debt ratio is negatively related to firm value and there are different estimates of the slope coefficients for each regime However, in the second model, where Tobin’s Q is used as proxy for firm value, this study provides new evidence on the existence of optimal debt ratio to maximize firm profitability If debt ratio is less than 25.69%, we find an increase

of Tobin’s Q but when debt ratio is greater than 25.69%, there is no relationship between debt ratio and Tobin’s Q

In the period from 2011 to 2015, with the high interest rate, financial cost has significantly gone up, resulting in a decrease of profitability of the firms As ROE ratio is only calculated by book value, it is not accurately captured the economic value of the company’s assets Therefore, when Tobin’s Q is used as proxy for firm value,the positive relationship is found in the regime that debt ratio is less than 25.69% and in the last regime, there is insignificant negative relationship between leverage and firm value From the above mentioned findings, financial managers should seek a level of debt for company that balances the benefits of interest tax shield and the incremental cost of debt financing For furtherresearches, it is advisable to use a larger sample and divide the observations into smaller samples based on different time periods or market capitalization.Besides, we need to examine the simultaneous relationship between capital structure and firm value

References

Abd Halim Ahmad; Nur Adiana Hiau Abdullah (2013) Investigation of optimal capital

structure in Malaysia: a panel threshold estimation Studies in Economics and Finance, 30, 108-117.

Alan Kraus; Robert H Litzenberger (1973) A State-Preference Model of Optimal

Financial Leverage Journal of Finance, 911-922.

Allen, F (1993) Strategic management and financial markets Strategic Management

Journal, 11-22.

Baker, M & Wurgler, J (2002) The Determinants of Capital Structure: Capital

Market-Oriented versus Bank-Market-Oriented Institutions Journal of Financial and Quantitative Analysis, 43, 59-92.

Berle, A.A and Means, G.C (1932) The Modern Corporation and Private Property.

New York: The Macmillan Company

Cheng, Y.; Liu, Y.; Chien, C (2010) Capital structure and firm value in China: A panel

threshold regression analysis African Journal of Business Management ,

2500-2507