How market competition regulates the relationship between overinvestment and profitability?

Nội dung text: How market competition regulates the relationship between overinvestment and profitability?

1. HOW MARKET COMPETITION REGULATES THE RELATIONSHIP BETWEEN OVERINVESTMENT AND PROFITABILITY? Chau Van Thuong* Faculty of Accounting – Finance – Banking, Ho Chi Minh City University of Technology (Hutech) ABSTRACT The industrial impact may play an important role in the relation between financial decisions and profitability at a certain level of competition. Thus, the study examines the role to indicate the effect of industry competition on the debt-profitability relationship. Furthermore, overinvestment is supposed to reduce profitability because it worsens agency problems between managers and shareholders. Consequently, the research is the first to point out the different impacts of competition on the debt- profitability relationship in firms with and without overinvestment. Our data from Thomson Reuters covers a wide range of 21 industries of all Vietnamese listed companies in seven years. The methodology goes through two steps: (1) measuring representative variables for competition and overinvestment; (2) running the main regression using SGMM estimator with tangibility and non-debt tax shield as instrumental variables. Our results show that debt is positively associated with profitability and that higher level of competition makes such a relation stronger. However, the interaction between debt and competition becomes weaker in case companies experience the problem of overinvestment. Keywords: Financial leverage, market competition, overinvestment, profitability, Vietnam. 1. INTRODUCTION The debt-profitability relationship attracts much attention and raises many debates in the global science community. Modigliani and Miller [1] suggested that capital structure does not determine firm performance under some assumptions of a perfect capital market. However, an enormous number of empirical studies have subsequently been conducted to confirm their relationship in reality. Surprisingly, almost all of their findings came to the consensus that capital structure is relevant to firm performance through the mechanism of the trade-off effect, limited liability effect, and discipling effect [2-8]. Moreover, the sign of debt remains debatable among empirical studies. The trade-off between the costs and benefits of debt and equity [3], the limited liability [2], and the discipling effect [4, 5] support the positive sign. On the other hand, underinvestment [9] and stakeholders‟ reactions the [10, 11] explain for the negative sign. Besides, the predation theory suggests that when a market is highly concentrated, it is easier for a company in the market to be swallowed by others in case it uses much debt in its capital structure [12-14]. The predation theory emphasizes on the role of market product competition in regulating the real impact of debt on firm performance. In practice, some studies gave empirical evidence on such a role in the US and some emerging markets [15-20]. However, market product competition has not yet been researched in Vietnam since it was added to the list of emerging countries with its high economic growth, trade openness, as well as investment inflow within the two last decades. Moreover, because capital market in 383
2. the country has not fully developed, its banking system has remained the major financing source for investment to Vietnamese enterprises. Thus, debt ratio in essence is a decisive factor of performance in the case of Vietnam [21-23]. Recently, Vietnam‟s global integration accompanied by its openness in various aspects of the economy and its implementation of privatization process among state-owned enterprises (SOEs) leads to the increased level of competition among companies in the market [23, 24]. In short, such a circumstance facilitates the moderation role of industry competition in capital structure of Vietnamese companies. According to Agency Theory, the conflicts of interests between managers and shareholders force a firm to incur huge costs to settle them down [3, 5, 25]. Exposed to the free cash flow, managers enlarge sources under their control through carrying out a lot of investment projects, even those with negative net present value (NPV) in order to satisfy their personal gains. Hence, overinvestment signals a serious agency problem and makes firm performance worse. Generally, with high debt in the capital structure, enterprises operating in a concentrated market and experiencing overinvestment tend to perform inefficiently. In short, debt ratio, industry competition, and overinvestment should be collectively studied in the research. The study aims at analyzing the role of industry competition on the debt-profitability relationship under overinvestment to answer two questions: (1) Does the impact of debt on profitability become better in a competitive market? and (2) Does overinvestment make the relationship worse? The answers to these questions partly contribute to the academic and practical world. For one thing, they provide empirical evidence on the critical role of competition and overinvestment in Vietnam, an emerging market, after the 2008 Global Financial Crisis. For another thing, they help investors set up a suitable investment portfolio and the government make appropriate policies in order to promote the freedom of the market as well as heighten the level of market competition in Vietnam. The original data includes 699 companies listed on Vietnam‟s two stock market exchanges namely HOSE and HNX in the period of 2010 – 2016. However, after the data processing and missing removal, the final dataset covers 208 companies in a wide range of 21 industries from Thompson Reuters source. Overinvestment is measured by taking the estimated value of residual from the sub-equation model. Competition is calculated in two ways through the opposite HHI Index and the absolute value of coefficients of the sub-equation model (BI Index). Using System Generalized Method of Moments (SGMM) to handle the endogenous problem caused by the dynamic function, the study indicates the positive relationship between debt ratio and profitability. Furthermore, in a competitive market, the use of debt is more likely to enhance its positive effect on profitability. The result implies the existence of the predation theory in Vietnam‟s product market. Nevertheless, overinvestment among Vietnamese enterprises causes the interaction effect to be weaker due to high agency costs. The estimated results are robust with alternative proxies of both competition and profitability. The paper is divided into 5 sections. Section 2 presents the empirical reviews together with hypothesis development. Research methodology and estimation are explained in Section 4 and 5, respectively. The study ends with the conclusion in Section 6. 2. LITERATURE REVIEW AND HYPOTHESIS DEVELOPMENT: 2.1 Debt ratio and profitability The relationship between debt and performance has raised much debates among various studies in corporate finance. Based on some assumptions of a perfect capital market without taxes, transaction costs, and asymmetric information, Modigliani and Miller [1] propose Capital Structure Irrelevance Theory, supporting the idea that capital structure is irrelevant to operation effectiveness. In reality, the capital 384
3. market has some imperfections. When relaxing the assumption of taxes, Modigliani and Miller [26] demonstrate the beneficial role of tax shields in helping companies obtain the benefits of tax deduction. When relaxing the assumption of asymmetric information, Jensen and Meckling [3] indicate the role of debt in solving the conflicts of interests based on Agency Theory. Asymmetric information makes managers more informed of business activities than shareholders. More seriously, with the free cash flow, managers attempt to establish a more extensive control over their companies through the pursuit of higher perquisites. It requires enterprises a certain amount of costs to align the interests between two sides. In this situation, an increase in debt ratio implies the reduction in the free cash flow and the participation of different partners in the capital market in the monitoring tasks with various disciplines as well as covenants [4, 5, 27]. As a result, the use of debt is supposed to heighten firm performance by decreasing agency problems in an enterprise. Hypothesis 1: financial leverage is positively related to profitability 2.1. Financial leverage, industry competition, and overinvestment The association of debt with industry competition seems to be complicated. The limited liability effect suggests that companies with high debt ratio can aggressively compete with others in the market [2]. Their aggressive behavior will lessen agency problems. However, the impact of such a behavior depends the level of competition as well as the characteristics of products in an industry [28]. Thus, a company with high debt cannot increase their profits due to the limited liability effect. Particularly, when a market is completely competitive, this effect is more obvious because the limited liability causes production to become more aggressive and market prices to decrease. According to Cournot competition, the more substitutable market products are the lower profitability is. The predation theory suggests that firms with high debt are more likely to be disadvantageous in terms of competitiveness compared to those with low debt in their capital structure. Fudenberg and Tirole [29] show that the theory is even more obvious in markets with a high level of concentration. Preexisting companies are inclined to predate newcomers. The predation process will lower the profitability of entrant companies and put them into a gloomy prospect. Financially constrained enterprises are more vulnerable to the predation from other rivals in the market. Supporting this idea, Scharfstein [30] explains the rivalry predation through debt covenants which are employed to mitigate agency problems. Once firms are exposed to restrictions from debt covenants, they are on the verge of liquidation and are forced to leave the market in case of failing to satisfy their obligations. Nevertheless, in a perfectly competitive market where every company only contributes a small part to the whole market‟s production, the rivalry predation tends to be negligible. The use of debt will lead to an increase in other competitors‟ market value; therefore, highly leveraged incumbent companies facilitate entrant ones to enter the market or expand their business activities Chevalier [17]. Consequently, debt makes the product market more competitive. In his extensive analysis, Chevalier [18] suggests that in a concentrated market, companies with high debt are forced to charge high prices than those with low debt, which makes them making sensitive to their rivals‟ predation. Chavalier and Scharfstein [13] explain the debt-competition in another respect using the switching cost model. They argue that during a certain recession when the market is less competitive, highly leveraged enterprises are much inferior to their counterparts with low debt because they have to charge high prices in their products. Thus, business disadvantage is associated with product market competition. A economic recession will result in market concentration, making it possible for companies with low debt to predate those with high debt [20]. Economic downturns are highly correlated with market concentration. Thus, during a recession, debt ratio negatively effects firm performance [15, 16]. Market concentration raises 385
4. agency problems in an enterprise, so the characteristics of a competitive market can strengthen the discipling effect of debt and help reduce agency problems costs [4, 31]. According to Agency Theory, the interests between managers and shareholders are diverse [3]. Due to asymmetric information between two sides, managers often take advantage of management to satisfy their personal gains. In doing so, they expand assets under their control to achieve higher perquisites, secure the managing position, and build up an empire of their own [32-34], all of which leads to overinvestment, that is, investment in unprofitable projects. Hence, overinvestment is expected to worsen agency problems [35- 38]. In summary, overinvestment will exacerbate the debt-competition interaction, making firm performance inefficient. Hypothesis 2: the effect of debt on profitability is less severe in a competitive market than a concentrated one. Hypothesis 3: the moderation impact of debt over industry competition is weaker under overinvestment. 3. DATA AND METHODOLOGY 3.1. Research methodology To test these three hypotheses, the study applies the following empirical model Profitabilityi,t = 0 +  1 Profitability i,t-1 +  2 LEV i,t +  3 COM j,t (1) ' +4 LEV i,t x COM j,t +  5 LEV i,t x COM j,t x OI i,t + w i,t +  i,t where Profitabilityi,t and Profitabilityi,t-1 are respectively profitability and its one-period lag measured by alternative proxies of firm i at time t and t-1; is the constant; LEVi,t is the ratio of total debt over total assets of firm i at time t; COM j,t is the proxy for the level of competition in industry j at time t, namely Herfindahl–Hirschman Index (HHI) and Boone Index (BI), which are described in detail below; OIi,t is estimated by the error terms extracted from Equation (4), those with positive signs are considered as overinvestment ; wi,t is a set of control variables described in the variable definition; i,t is the error term9. The representatives for profitability are return on assets (ROA) and return on equity (ROE), which are the earning before interests and tax (EBIT), earning before tax (EBT), earning after tax (EAT) divided by total assets and total equity respectively. Although these variables are thought to be affected by different accounting standards because their calculations are based on accounting books, they are considered to be better than Tobin‟s Q to be the representatives in this research. Demsetz and Lehn [39] suppose that ROA and ROE reflect the present situation, while Tobin‟s Q shows future development. Demsetz and Villalonga [40] emphasize that Tobin‟s Q is often affected by tangible assets whose depreciation is different from the real economic depreciation. Furthermore, the use of ROA and ROE helps mitigate the differences in firm size among companies in various industries. Debt ratio is measured by the ratio of total debt over total assets. The study adds some control variables as determinants of profitability to the regression including sale growth, firm size, and average return on assets. Sale growth (SGRO), the representative for growth opportunities [41, 42], is measured by the differences between sale of firm i at time t and its sale at time t- 9 See Appendix for the description of all variables. 386
5. 1 divided by sale at time t-1. Firm size (Size) is the logarithm of total assets. According to Ghosh [43], average return on assets (MROA) are the moving average of ROA in two consecutive years. The instrumental variables used to handle the endogenous problem in the regression model are tangibility (TANG) and non-debt tax shield (NDTS). TANG is the ratio of tangible assets over total assets. This variable plays a decisive role in a firm‟s access to financing capital [16, 44], especially in developing countries where the regulations to protect lenders and carry out loan contracts are loosely controlled. NDTS is the sum of research and development funds (R&D) and depreciation divided by total assets. To examine the role of industry competition in the relationship between financial leverage and firm performance, the research has to identify the proxies for industry competition. In fact, there are two ways to measure industry competition: structural and non-structural (Lawton, 1999). Structural approach evaluates market concentration using Herfindahl–Hirschman Index (HHI) [16] or the level of concentration within four or five largest companies in a certain industry (CR4 or CR5) [15, 17-20]. The degree of concentration (high HHI, CR4, or CR5) often accompanies lower competition and vice versa. Meanwhile, non-structural approach measures the level of competition from the market‟s behaviors. This measurement is appreciated more highly than structural approach because a high level of concentration does not imply lower competition in the market [45]. In fact, the hypothesis on the relationship between market structure and the effectiveness shows that high concentration is simply the results of the market‟s effectiveness [46]. Some companies that are operating effectively can quickly expand their market shares, while those which are ineffective are smaller and smaller in size [47]. Moreover, high concentration sometimes comes from the fierce competition of various companies in the market, leading to the fact that effective companies force ineffective ones to exit the market [48]. Thus, the level of concentration cannot correctly predict the level of competition in the market. To deal with such problems emerged from structural approach, Boone [49] uses a new index to measure market competition, Boone Index (BI). The index measures the sensitivity of firm profitability to the ineffectiveness of the market. Because in a competitive market companies often have to suffer a big loss when they perform ineffectively, firm profitability will increase with how effective a firm performs, and such an increase will be higher in a competitive market [50]. Hence, BI is the proxy that is preferred in studies on industry competition and firm performance [51]. However, to raise the reliability, the research will in turn employ these two alternatives to find out their impact on the relationship between financial leverage and firm performance. According to Beiner, Schmid [52], HHI is measured as the total market shares of each firm in a certain industry. N j Sales HHI ()ijt 2 (2) jt  N j j 1 Sales  j 1 ijt In the formula, HHI jt is HHI of industry j at time t; Salesijt indicates the sales of firm i in industry j at time t. The higher HHI is the higher market concentration becomes (lower market competition). BI is considered to be the index that helps directly evaluate the level of competition in the market. The index is based on the hypothesis of competition and effectiveness with the assumption that in a competitive market, if a firm does not operate effectively, it will incur losses [50, 53, 54]. Therefore, an industry with high competition is expected to have a sharp decrease in variable profits due to the increase in the marginal costs. Then, BI is estimated through the following regression model: VROAit = +  t lnMC ij +  it (3) 387
6. where VROAit is the variable profits calculated by subjecting costs of goods from sales of firm i in industry j divided by total assets; lnMCij is the logarithm of marginal costs which is costs of goods over sales of firm in in industry j; t is the coefficient of the model that is changing overtime. Its absolute value measures the degree of competition. The coefficient sign is expected to be negative. The higher the absolute value is he higher market competition is. Therefore, BI is the absolute value of t . As pointed out in the hypothesis development, market competition is an important factor in analyzing the effect of financial leverage on firm performance. In order to catch such an impact, the interaction between financial leverage and industry competition is added to the regression model. Besides, the research also takes into account the problem of endogeneity in the model which are originated from three major reasons: simultaneity, measurement errors, and omitted variables. To mitigate the simultaneous effect between LEV and ROA, the study uses the lag of LEV due to the fact that financial leverage in the past often affects profits at present but the reserve relationship is impossible. However, in addition to simultaneity, the estimated results are partly affected by omitted variables and measurement errors. Therefore, GMM two-stage least square is used to deal with such a problem. Having the doubt that LEV is endogenous, the research decides to take TANG and NDTS as its instrumental variables. These two instrumental variables are basically considered to be suitable. First, TANG is what the institutions use to evaluate the possibility of their customers‟ paying loans back so that they can make right decisions on lending capital [16, 44]. Thus, the effect of this variable on firm performance is mainly through the financing capital to companies, showing that TANG is an appropriate instrumental variable for LEV [16]. Second, firms with higher non-debt tax shield are expected to have higher financial leverage [55], and non-debt tax shield is not supposed to have the direct impact on earnings before tax and depreciation. This fact suggests that NDTS is an effective instrumental variable for financial leverage. Actually, Fama and French [56] support the empirical evidence for the reverse relationship between non- debt tax shield and financial leverage. In short, the study uses both factors as instrumental variables. Finally, overinvestment is measured through Equation (4) using the fixed-effect technique. The estimated equation is generalized based on the ideas from previous studies [57-64]. The explicit form of Equation (1) is as follows: NewInvestment CashFlow TobinQ FixCapitalIntensity FirmSize i, t 0 1 i , t 2 i , t 3 i , t 4 i , t (4) NewInvestmenti,0t 5RevenueGrowthi , t 6 BusinessRisk i , t 7 Leverage i,,ti  t In the equation, NewInvestmentit, represents for the investment decision; CashFlowit, reflects the cash available in a company after subtracting capital expenditures; TobinQit, is the representative of growth opportunity and market performance; FixCapitalIntensityit, evaluates the ability to generate fixed assets through sales; RevenueGrowthit, demonstrates the growth of the firm; FirmSizeit, shows a company‟s financial constraints; BusinessRiskit, indicates the volatility of firm profitability; Leverageit, is the capital structure of the company. The estimated error-termˆit, taken from the above model is considered as the abnormalities in the investment decision. If the error term‟s value is positive, or ˆit, 0 , ˆit, of 388
7. th th firm i and year t is denoted as Overinvestmentit, . This method of calculating overinvestment has been recently adopted by He and Kyaw [65].10 3.2. Research data The research data is collected from Vietnamese listed companies in HOSE and HNX from 2010 to 2016. Based on the classification standard on Vietnam‟s Stock Exchange, the sample is classified into 21 different industries including durable goods, consumer goods, real estates, printing (except the Internet), transportation support, mining, professional contractors, electricity, basic metals, textiles, plastics an rubber, beverages and tobacco, paper, chemicals, non-metal minerals, food, electronic equipment, cultivation, sea transportation, heavy industry and civil construction, houses and buildings. Our data and correlation coefficients are summarized as table 1 and 2 below: Table 1. Descriptive statistics Variable Obs. Mean Std. Dev. Min Max EAT/TA 1,384 0.060576 0.054632 -0.041441 0.239514 EBT/TA 1,384 0.074129 0.065519 -0.041777 0.291967 EBIT/TA 1,384 0.094697 0.060779 -0.016461 0.303093 EAT/Equity 1,384 0.123358 0.090448 -0.128181 0.375616 EBT/Equity 1,384 0.151656 0.107330 -0.125376 0.447676 EBIT/Equity 1,384 0.219343 0.122889 -0.041824 0.561071 MROA 921.0 0.315095 0.232828 0.045080 1.109090 Size 1,384 27.08540 1.293908 23.95720 30.18850 Growth 1,384 0.108733 0.265465 -0.492889 1.131610 Leverage 1,383 0.512923 0.205921 0.103515 0.849271 Competition1 1,388 -0.259280 0.083718 -0.670789 -0.135970 Competition2 1,394 0.475526 0.496635 0.171184 2.278290 Source: Author’s calculation Table 2. Matrix correlation MROA Size Growth Leverage Competition1 Competition2 MROA -0.0876 -0.0261 -0.1467 0.1452 -0.0595 1.0000 (0.0095) (0.4396) (0.0000) (0.0000) (0.0773) Size -0.0876 0.0993 0.2130 -0.1827 0.1329 1.0000 (0.0095) (0.0003) (0.0000) (0.0000) (0.0000) Growth -0.0261 0.0993 1.0000 0.0649 -0.0021 0.029 10 See Appendix for the description of all variables 389
8. MROA Size Growth Leverage Competition1 Competition2 (0.4396) (0.0003) (0.0184) (0.9399) (0.2918) Leverage -0.1467 0.2130 0.0649 -0.0475 0.1292 1.0000 (0.0000) (0.0000) (0.0184) (0.0850) (0.0000) Competition1 0.1452 -0.1827 -0.0021 -0.0475 0.2479 1.0000 (0.0000) (0.0000) (0.9399) (0.0850) (0.0000) Competition2 -0.0595 0.1329 0.029 0.1292 0.2479 1.0000 (0.0773) (0.0000) (0.2918) (0.0000) (0.0000) P-Values are given in the parentheses Source: Author’s calculation 4. RESULTS AND DISCUSSION Table 3 presents the estimation results of Equation (1). The SGMM method is used with the ratio of tangible assets to total assets and non-debt tax shield to total assets as instruments for debt ratio. The first six columns use the HHI index measuring the level of market concentration. Meanwhile, the last six columns use the BI index measuring market competition. Specifically, the lower the HHI, the lower the competition, while BI is in the opposite direction. Therefore, two competition variables, Competition 1 = (-HHI) and Competition 2 = BI, are generated to interpret the impacts of these two indicators on performance in the same direction. The estimated results from the regression indicate that debt is positively associated with profitability. Whereas, two representative variables for industry competition are negatively. These findings are suitable with the disciplining effect and Agency Theory [3-5, 66, 67]. Additionally, the significant positive impact of the two-variable interaction term between debt and competition is clearly shown in the estimation. This result demonstrates that when the market is highly competitive, the positive impact of debt on profitability is stronger when firms are less likely be driven out of the market. The bad effect of debt in a concentrated market becomes less severe when the market becomes more competitive. The estimation is consistent with findings of a recent empirical study in Vietnam [68]. Interestingly, the influence of the debt-competition interaction becomes weaker overinvestment because overinvestment increases agency problems and weakens the good impact of debt. This evidence supports the hypothesis that the debt-profitability nexus is conditional on market competition and overinvestment. The estimation robustness is tested using alternative representatives for industry competition and firm performance. Industry competition is alternatively measured by the residual estimated from the subequation (BI Index) and Herfindahl-Hirschman Index (HHI Index). Furthermore, earnings before interests and taxes (EBIT), earnings before taxes (EBT), and earnings after taxes (EAT) over total assets and total equity are respectively employed to represent for profitability. Consequently, the estimated coefficients of all different proxies reach the consistency in terms of signs and significance levels. All the relevant tests of System Generalized Method of Moments (SGMM) estimator appear to be comfortable in every single regression model in the research. 390
9. Table 3. Regression Estimation Competition 1 = (-HHI) Competition 2 = BI SGMM Estimations EAT/TA EBT/TA EBIT/TA EAT/Equity EBT/Equity EBIT/Equity EAT/TA EBT/TA EBIT/TA EAT/Equity EBT/Equity EBIT/Equity Lag _ Profitability 0.565 0.583 0.680 0.256 0.439 0.559 0.601 0.685 0.787 0.431 0.532 0.532 (0.0887) (0.0982) (0.0776) (0.121) (0.0721) (0.0615) (0.0701) (0.0777) (0.0832) (0.0502) (0.0611) (0.0686) MROA 0.0143 0.0147 0.0195 0.0762* 0.0822 0.0608* 0.0538 0.0651 0.0506 0.104 0.128 0.110 (0.0194) (0.0239) (0.0154) (0.0448) (0.0290) (0.0330) (0.0177) (0.0210) (0.0219) (0.0307) (0.0407) (0.0513) Size -0.00339 -0.00482 -0.00340 -0.0133 -0.00436 -0.0127 -0.00358* -0.00473 -0.00623 -0.00611 -0.0101 -0.0130 (0.00272) (0.00341) (0.00220) (0.00939) (0.00506) (0.00598) (0.00186) (0.00218) (0.00222) (0.00281) (0.00432) (0.00500) Growth 0.00518 0.00526 0.00841 0.00144 0.0112 0.0237* 0.00777 0.00837 0.00955 0.0182 0.0215 0.0130 (0.00508) (0.00588) (0.00555) (0.0139) (0.0117) (0.0136) (0.00316) (0.00347) (0.00378) (0.00714) (0.00786) (0.00419) Leverage (Lev) 0.243* 0.329* 0.248 0.897* 0.419* 0.851 0.184 0.235 0.314 0.377 0.582 0.738 (0.142) (0.175) (0.112) (0.462) (0.250) (0.298) (0.0884) (0.103) (0.103) (0.134) (0.204) (0.226) Competition1 (Com1) -0.645 -0.819 -0.483 -1.439* -0.649 -1.066* (0.264) (0.310) (0.228) (0.837) (0.473) (0.593) Lev*Com1 1.480 1.848 1.136 3.512 1.842 2.316 (0.538) (0.621) (0.449) (1.557) (0.886) (1.125) Lev*Com1*Overinvestment -0.300* -0.339* -0.209 -1.073* -0.589* -0.208 (0.172) (0.203) (0.145) (0.615) (0.322) (0.421) Competition2 (Com2) -0.173 -0.192 -0.217 -0.308 -0.441 -0.540 (0.0871) (0.0952) (0.0996) (0.138) (0.196) (0.240) Lev*Com2 0.392 0.435 0.470 0.730 1.029 1.096 (0.153) (0.167) (0.177) (0.250) (0.356) (0.420) Lev*Com2*Overinvestment -0.182 -0.194 -0.162* -0.281 -0.346* -0.319 (0.0740) (0.0882) (0.0854) (0.136) (0.183) (0.158) Observations 761 761 761 517 645 645 761 761 761 761 761 517 Number of instruments 22 22 30 23 32 30 26 26 26 32 26 24 Number of groups 164 164 164 160 163 163 164 164 164 164 164 160 F-Statistics 51.24 54.92 219.55 16.32 70.76 151.9 38.85 45.44 70.38 66.82 55.71 60.31 Prob. 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 Arellano-Bond test for AR(1) -1.510 -1.600 -1.760 -2.590 -1.400 -1.520 -1.630 -1.690 -1.820 -1.530 -1.620 -1.780 Prob. 0.130 0.110 0.078 0.010 0.161 0.129 0.102 0.091 0.069 0.126 0.106 0.076 Arellano-Bond test for AR(2) 1.140 0.920 -0.080 0.950 1.080 1.320 0.350 -1.010 -0.770 1.090 1.120 0.480 Prob. 0.256 0.356 0.939 0.340 0.280 0.188 0.729 0.312 0.441 0.275 0.262 0.632 Hansen test of over-identification 14.3 13.3 19.65 13.94 29.02 25.9 15.62 16.22 14.12 31.81 25.98 15.77 Prob. 0.428 0.503 0.605 0.53 0.219 0.256 0.619 0.577 0.721 0.132 0.1 0.469 Standard errors in parentheses *, , corresponding to significance level of 10%, 5% and 1% Source: Author’s calculation 5. CONCLUSIONS The debt-profitability relationship is supposed to be moderated by industry competition and overinvestment because both factors are associated with agency problems within enterprises. Vietnam‟s current situation together with previous empirical studies have indicated the interdependence of debt, competition, and overinvestment and their combined impact on firm performance. With the use of System Generalized Method of Moments (SGMM), the research aims at identifying the role of industry competition in the debt-profitability relationship under overinvestment. The paper clarifies that profitability in Vietnamese listed companies are positively affected by debt ratio in the capital structure. 391
10. Furthermore, the positive influence of debt gets stronger in an industry with high levels of competition, meaning that the bad effect of debt in this situation becomes less severe when entrants are less likely to be kicked out of the market by their competitors. However, the positive sign of the two-variable interaction becomes weaker when overinvestment is taken into consideration because overinvestment increases agency problems in companies. Based on the estimated results, some recommendations are given to both the government and corporate companies. The government should heighten the level of competition through higher economic growth, better market regulations, and more transparent legal practices. Companies should limit the problem of overinvestment or mitigate agency problems by compensating managers with more benefits to increase their commitments toward acting in favor of shareholders‟ interests. REFERENCES [1] Modigliani, F. and M.H. Miller, The cost of capital, corporation finance and the theory of investment. The American economic review, 1958. 48(3): p. 261-297. [2] Brander, J.A. and T.R. Lewis, Oligopoly and financial structure: The limited liability effect. The American Economic Review, 1986: p. 956-970. [3] Jensen, M.C. and W.H. Meckling, Theory of the firm: Managerial behavior, agency costs and ownership structure. Journal of financial economics, 1976. 3(4): p. 305-360. [4] Grossman, S.J. and O.D. Hart, An analysis of the principal-agent problem. Econometrica: Journal of the Econometric Society, 1983: p. 7-45. [5] Jensen, M.C., Agency costs of free cash flow, corporate finance, and takeovers. The American economic review, 1986. 76(2): p. 323-329. [6] Margaritis, D. and M. Psillaki, Capital structure, equity ownership and firm performance. Journal of Banking & Finance, 2010. 34(3): p. 621-632. [7] San, O.T. and T.B. Heng, Capital structure and corporate performance of Malaysian construction sector. International Journal of Humanities and Social Science, 2011. 1(2): p. 28-36. [8] Khan, A.G., The relationship of capital structure decisions with firm performance: A study of the engineering sector of Pakistan. International Journal of Accounting and Financial Reporting, 2012. 2(1): p. 245. [9] Myers, S.C., Determinants of corporate borrowing. Journal of financial economics, 1977. 5(2): p. 147-175. [10] Maksimovic, V. and S. Titman, Financial policy and reputation for product quality. The Review of Financial Studies, 1991. 4(1): p. 175-200. [11] Titman, S., The effect of capital structure on a firm's liquidation decision. Journal of financial economics, 1984. 13(1): p. 137-151. [12] Bolton, P. and D.S. Scharfstein, A theory of predation based on agency problems in financial contracting. The American economic review, 1990: p. 93-106. [13] Chavalier, J. and D. Scharfstein, Capital market imperfections and countercyclical markups. Amer. Econ. Rev, 1996. 86: p. 703-725. [14] Dasgupta, S. and S. Titman, Pricing strategy and financial policy. The Review of Financial Studies, 1998. 11(4): p. 705-737. 392
11. [15] Campello, M., Capital structure and product markets interactions: evidence from business cycles. Journal of Financial Economics, 2003. 68(3): p. 353-378. [16] Campello, M., Debt financing: Does it boost or hurt firm performance in product markets? Journal of Financial Economics, 2006. 82(1): p. 135-172. [17] Chevalier, J.A., Capital structure and product-market competition: Empirical evidence from the supermarket industry. The American Economic Review, 1995: p. 415-435. [18] Chevalier, J.A., Do LBO supermarkets charge more? An empirical analysis of the effects of LBOs on supermarket pricing. The Journal of Finance, 1995. 50(4): p. 1095-1112. [19] Kovenock, D. and G.M. Phillips, Capital structure and product market behavior: An examination of plant exit and investment decisions. The Review of Financial Studies, 1997. 10(3): p. 767-803. [20] Opler, T.C. and S. Titman, Financial distress and corporate performance. The Journal of Finance, 1994. 49(3): p. 1015-1040. [21] Fu-Min, C., et al., Capital Structure Decisions and Firm Performance of Vietnamese Soes. Asian Economic and Financial Review, 2014. 4(11): p. 1545. [22] Gueorguiev, D. and E. Malesky, Foreign investment and bribery: a firm-level analysis of corruption in Vietnam. Journal of Asian Economics, 2012. 23(2): p. 111-129. [23] Tran, N.M., W. Nonneman, and A. Jorissen, Privatization of Vietnamese firms and its effects on firm performance. Asian economic and financial review, 2015. 5(2): p. 202. [24] Quy, V.T., N.D. Khuong, and F. WilliamSwierczek, Corporate performance of privatized firms in Vietnam. 2014. [25] Gaver, J.J. and K.M. Gaver, Additional evidence on the association between the investment opportunity set and corporate financing, dividend, and compensation policies. Journal of Accounting and economics, 1993. 16(1-3): p. 125-160. [26] Modigliani, F. and M.H. Miller, Corporate income taxes and the cost of capital: a correction. The American economic review, 1963. 53(3): p. 433-443. [27] Harris, M. and A. Raviv, Capital structure and the informational role of debt. The Journal of Finance, 1990. 45(2): p. 321-349. [28] Wanzenried, G., Capital structure decisions and output market competition under demand uncertainty. International Journal of Industrial Organization, 2003. 21(2): p. 171-200. [29] Fudenberg, D. and J. Tirole, A" signal-jamming" theory of predation. The RAND Journal of Economics, 1986: p. 366-376. [30] Scharfstein, D.O., Analytical performance measures for the miniload automated storage/retrieval system. 1990, Georgia Institute of Technology. [31] Aghion, P., M. Dewatripont, and P. Rey, Corporate governance, competition policy and industrial policy. European Economic Review, 1997. 41(3-5): p. 797-805. [32] Myers, S.C., The capital structure puzzle. The journal of finance, 1984. 39(3): p. 574-592. [33] Brealey, R.A., S.C. Myers, and F. Allen, Brealey, Myers, and Allen on valuation, capital structure, and agency issues. Journal of Applied Corporate Finance, 2008. 20(4): p. 49-57. 393
12. [34] Hail, L., A. Tahoun, and C. Wang, Dividend payouts and information shocks. Journal of Accounting Research, 2014. 52(2): p. 403-456. [35] Liu, N. and D. Bredin, Institutional Investors, Over-investment and Corporate Performance. University College Dublin, 2010. [36] Titman, S., K.J. Wei, and F. Xie, Capital investments and stock returns. Journal of financial and Quantitative Analysis, 2004. 39(4): p. 677-700. [37] Yang, Y.M., Corporate governance, agency conflicts, and equity returns along business cycles. 2005. [38] Fu, F., Overinvestment and the operating performance of SEO firms. Financial Management, 2010. 39(1): p. 249-272. [39] Demsetz, H. and K. Lehn, The structure of corporate ownership: Causes and consequences. Journal of political economy, 1985. 93(6): p. 1155-1177. [40] Demsetz, H. and B. Villalonga, Ownership structure and corporate performance. Journal of corporate finance, 2001. 7(3): p. 209-233. [41] King, M.R. and E. Santor, Family values: Ownership structure, performance and capital structure of Canadian firms. Journal of Banking & Finance, 2008. 32(11): p. 2423-2432. [42] Maury, B., Family ownership and firm performance: Empirical evidence from Western European corporations. Journal of Corporate Finance, 2006. 12(2): p. 321-341. [43] Ghosh, S., Leverage, foreign borrowing and corporate performance: firm-level evidence for India. Applied Economics Letters, 2008. 15(8): p. 607-616. [44] Booth, L., et al., Capital structures in developing countries. The journal of finance, 2001. 56(1): p. 87-130. [45] Guzmán, G.M., et al., Measuring the competitiveness level in furniture SMEs of Spain. International Journal of Economics and Management Sciences, 2012. 1(11): p. 09-19. [46] Demsetz, H., Industry structure, market rivalry, and public policy. The Journal of Law and Economics, 1973. 16(1): p. 1-9. [47] Boone, J., R. Griffith, and R. Harrison. Measuring competition. in Encore Meeting. 2004. [48] Boone, J., Competition: Theoretical parameterizations and empirical measures. Journal of Institutional and Theoretical Economics JITE, 2008. 164(4): p. 587-611. [49] Boone, J., Measuring product market competition. CEPR Discussion Paper, 2000(2636). [50] Boone, J., A new way to measure competition. The Economic Journal, 2008. 118(531): p. 1245- 1261. [51] Boone, J., J.C. van Ours, and H. van der Wiel, When is the price cost margin a safe way to measure changes in competition? De Economist, 2013: p. 1-23. [52] Beiner, S., M.M. Schmid, and G. Wanzenried, Product market competition, managerial incentives and firm valuation. European Financial Management, 2011. 17(2): p. 331-366. [53] Boone, J., R. Griffith, and R. Harrison, Measuring competition (Research Paper No. 022). Advanced Institute of Management, 2005. [54] Boone, J., J.C. Van Ours, and H.v.d. Wiel, How (not) to measure competition. 2007. 394
13. [55] DeAngelo, H. and R.W. Masulis, Optimal capital structure under corporate and personal taxation. Journal of financial economics, 1980. 8(1): p. 3-29. [56] Fama, E.F. and K.R. French, Testing trade-off and pecking order predictions about dividends and debt. Review of financial studies, 2002. 15(1): p. 1-33. [57] Bokpin, G.A. and J.M. Onumah, An empirical analysis of the determinants of corporate investment decisions: Evidence from emerging market firms. International Research Journal of Finance and Economics, 2009. 33: p. 134-141. [58] Carpenter, R.E. and A. Guariglia, Cash flow, investment, and investment opportunities: New tests using UK panel data. Journal of Banking & Finance, 2008. 32(9): p. 1894-1906. [59] Connelly, J.T., Investment policy at family firms: Evidence from Thailand. Journal of Economics and Business, 2016. 83: p. 91-122. [60] Li, D. and L. Zhang, Does q-theory with investment frictions explain anomalies in the cross section of returns? Journal of Financial Economics, 2010. 98(2): p. 297-314. [61] Malm, J., et al., Litigation risk and investment policy. Journal of Economics and Finance, 2016: p. 1-12. [62] Nair, P., Financial Liberalization and Determinants of Investment: A Study of Indian Manufacturing Firms. International Journal of Management of International Business and Economic Systems, 2011. 5(1): p. 121-133. [63] Ruiz-Porras, A. and C. Lopez-Mateo, Corporate governance, market competition and investment decisions in Mexican manufacturing firms. 2011. [64] Richardson, S., Over-investment of free cash flow. Review of accounting studies, 2006. 11(2-3): p. 159-189. [65] He, W. and N.A. Kyaw, Ownership structure and investment decisions of Chinese SOEs. Research in International Business and Finance, 2018. 43: p. 48-57. [66] Weill, L., Leverage and corporate performance: does institutional environment matter? Small Business Economics, 2008. 30(3): p. 251-265. [67] Berger, A.N. and E.B. Di Patti, Capital structure and firm performance: A new approach to testing agency theory and an application to the banking industry. Journal of Banking & Finance, 2006. 30(4): p. 1065-1102. [68] Chau Van Thuong, Tran Le Khang, and Nguyen Cong Thanh, Cau truc von va hieu qua hoat dong cua doanh nghiep: Vai tro cua canh tranh nganh. Tap chi Phat trien Kinh te, 2017. 28-10: p. 56-78. 395
14. APPENDIXES Table A1. Variable Measurement for main econometric model FirmPerformanceit,01   FirmPerformance it ,12,3,4,  LEV it  COM it  LEV it COM it , FirmPerformanceit,0  5LEVit , Overinvestmenti , t 6COM i , t Overinvestment i, t FirmPerforman ceit, 0 7,LEVi t COM i, t Overinvestmenti,,, t ' X i t i t Dependent Variables Variables Denote Definition EBIT (Earnings Before Interest & Tax) divided by Total EBIT/TA Asset EBT/TA EBT (Earnings Before Tax) divided by Total Asset EAT/TA EAT (Earnings After Tax) divided by Total Asset Profitability EBIT (Earnings Before Interest & Tax) divided by Total EBIT/TE Equity EBT/TE EBT (Earnings Before Tax) divided by Total Equity EAT/TE EAT (Earnings After Tax) divided by Total Equity Explanatory Variables Variables Denote Definition Moving average of EBITDA (Earnings Before Interest, Moving average of ROA MROAit, Taxes, Depreciation and Amortization) over Total Asset ratios in three consecutive years Growth Company Growth it, Growth rate of total sale Company Size Sizeit, Natural logarithm of total asset Leverage LEVit, Total liabilities over total asset 1 COMit, HHI index Competition 2 COMit, BI index 396
15. Table A2. Variable Measurement for subequation NewInvestmenti, t 0 1 CashFlow i , t 2 TobinQ i , t 3 FixCapitalIntensity i , t 4 FirmSize i , t NewInvestmenti,0t 5RevenueGrowthi , t 6 BusinessRisk i , t 7 Leverage i,,ti  t Dependent Variables Variables Denote Definition Total investment including long-term and short-term New Investment NewInvestment it, investment divided by total asset Explanatory Variables Variables Denote Definition The cash available in a company after subtracting Cash Flow CashFlow it, capital expenditures The Q ratio is calculated as the market value of a Market Performance TobinQ it, company divided by the firm's assets FixCapitalIntensity The ability to generate fixed assets through sales Fix Capital Intensity it, measured by total fix asset divided by total sales Firm Growth RevenueGrowthit, The growth rate of firm's revenue over year Firm Size FirmSizeit, Natural logarithm of total asset Standard deviation of EBITDA (Earnings Before Business Risk BusinessRiskit, Interest, Taxes, Depreciation and Amortization) over Total Asset ratio in three consecutive years Leverage Leverageit, Total liabilities over total asset All the above variables are calculated by using financial data from Thomson Reuters Eikon Financial Analysis 397