The research results also show that the profit of the enterprises is influenced by different factors: business size, age, leverage ratio, profitability, debt to equity ratio and structur
THEORETICAL FRAMEWORK AND LITERATURE REVIEW
Theoretical Framework
1.1.1 Overview of business performance of enterprises
The economy - society of our country in 2019 took place amid the world’s slow growth economic situation Trade tensions between the US-China and geopolitical issues significantly increase the instability of the global trading system It has no small impact on global business confidence, investment decisions and trade Unexpected fluctuations in financial markets - international currencies, complex oil prices impact credit growth, psychology and market expectations International organizations are constantly making optimistic forecasts about world economic growth in 2019 Domestic, besides the advantages of positive growth results in 2019, macroeconomics are stable but also face no less difficulties, challenges with complex weather developments affecting crop yields and yields The livestock industry has difficulty with African pig cholera taking place in all 63 provinces and cities under the Central; slowing growth of some key exports; Disbursement of public investment is not planned Vietnam corporations ‘s business performance in 2019
As of December 31, 2019, Vietnam's business landscape boasted 668,505 operating enterprises, representing a significant 9.5% increase from 2018 Notably, 43.0% of these businesses reported profitability, while 8.2% managed to break even Conversely, 48.8% of businesses unfortunately experienced losses during this period.
The total capital used for SXKD of the entire operating enterprise with the results of business production at 12/31/2019 reached VND 43.3 million, increased by 11.4% compared to the same time in 2018
The total net revenue of the entire operating business sector with business production results in 2019 reached more than VND 26.3 million, an increase of 11.4% compared to 2018 with industrial and construction businesses with the largest contribution
In 2019, the total pre-tax profit of the business sector reached VND 889.9 trillion, down 0.5% compared to 2018 The percentage of profitable businesses in 2019 reached 43.0%, down from 2018 (44.1%) The percentage of business enterprises losing in 2019 reached 48.8%, an increase compared to 2018 (48.4%).TOP 50 in 2019 is still a familiar company in the consumer goods, real estate, construction, industry, etc industries, ,…
In particular, the real estate industry group has many leading units honored such as Vinhomes, Novaland, Khang Dien, Nam Long with revenue and profit indicators, Impressive capitalization value
Return on assets (ROA) is a profitability metric that evaluates a company's ability to generate earnings relative to its total assets, which represent the resources and investments it has at its disposal This ratio provides insights into how efficiently a company utilizes its assets in generating profits.
Return on Assets (ROA) = Net Income/Total Assets
The ROA provides investors with information about the interest generated from the amount of investment ( or the amount of assets ) A company's assets are made up of loans and equity Both of these are used to fund company operations The effect of converting investment capital into profit is shown by ROA The higher the ROA, the better because the company is making more money on less investment
When evaluating business performance, profit rate indicators like ROA (Return on Assets) are commonly used ROA reflects the relationship between profit and production costs, indicating the efficiency of resource utilization Different business types require specific performance metrics, with financial indicators being the primary measure for for-profit organizations Therefore, ROA is chosen as an appropriate evaluation tool for assessing the performance of 243 corporations in Vietnam.
2019 Therefore ROA will be presented as a dependent variable later in our model.
Literature Review
1.2.1 Research into factors affecting business performance of enterprises
For enterprises, business performance evaluation is indispensable as it helps them align their employees, resources, and systems to meet their strategic objectives As such, researchers have conducted various studies into factors affecting business performance of enterprises from both theoretical and empirical perspectives However, as those studies vary from different industries and contexts within different periods of time, the results can not be applied into Vietnam’s outlook and characteristics of Vietnamese enterprises –the scope of this thesis
Following are the explanation and analysis of previous studies into the subject a) Prior research in foreign countries
Firstly, “The Determinants of Small and Medium-sized Enterprises Performance in Nigeria” by Idris Isyaku Abdullahi, Sulaiman Chindo (2015) from which data was collected from National Bureau of Statistics of Nigeria and Central Bank of Nigeria database investigated five main factors that could be determinants of small and medium- sized enterprises, which are interest rate, government spending, net export, level of education and instability on SMEs performance in Nigeria SMEs performance was quantified based on share of SMEs contribution to GDP With ADRL approach, the research has identified that only two factors: interest rate and net export had significant and negative impact on SMEs performance Overall, the study pointed out some policies which are for regulators to improve the business performance in Nigeria Moreover, as the economic context of Nigeria is extremely different from that of Vietnam, the findings might not be applicable only in the Nigerian setting
Secondly, to identify factors that are affecting business success of small and medium enterprises in Thailand, Chittithaworn, Islam and Keawchana et al conducted a research in 2011 with regression analysis Eight factors were examined during the research and these factors are SMEs characteristic, management and know-how, products and services, customer and market, the way of doing business and cooperation, resources and finance, strategy, and external environment Based on correlation analysis from 143 companies in Thailand, the results of the survey indicated that SMEs characteristics, customer and markets, the way of doing business, resources and finance, and external environment have significant positive effect on the business success of SMEs Whilst, management know-how, product and services and strategy have no significant effect on performance of business
Several years later, another research also conducted in Thailand is the reflective study of Nimlaor, Trimetsoontorn and Fongsuwan in 2014 “Factors Affecting Business Performance: An Empirical Study in Thailand” which evaluated “the empirical relationships and influencing factors of firm strategy, firm environment and firm characteristics that influence the performance of the Thai garment industry” (Nimlaor, Trimetsoon and Fongsuwan, 2014) With PLS model and data collected from 178 chief executives in garment companies in Thailand, the research revealed that all firm characteristics including leaders and teamwork and firm strategy factor consisting of R&D, product differentiation and brand has direct influence on firm performance Firm environment also have indirect impact on performance of enterprises through sub- factors such as government policy, alliance and international trading system Overall, the researchers in this study has showed that the internal factors of companies, espcially human-related factors also have statistical impact on business performance However, this research is limited in the Thai garment industry For that reason, the conclusion can be varying in other industries due to the unique characteristics of garment industry and therefore, unable to generalize those findings in other countries
In comparison with the research of Chittithaworn, Islam and Keawchana et al., Nimlaor (2011), that of Trimetsoontorn and Fongsuwan (2014)also have similar variables in term of meaning, just only minor differences in categorizing variables and the scope of research (SMEs in all over Thailand and companies in garment industry) However, even if the study is conducted in the same country, it can be also seen from these two researchers that not every factor can have impact on the firm performance, it depends greatly on the scale or the scope of the research b) Previous studies in Vietnam
In Vietnam, factors affecting business performance of enterprises also triggred interests of scholars and practitioners to find out more, especially when the Vietnamese economy is in the international economic integration era Prior researchers in Vietnam focus mainly into the business performance of SMEs in Vietnam for the reason that the number of small and medium-sized enterprises accounts for 93.5% of the total number of enterprises operating in Vietnam according to the 2021 Vietnamese White paper about enterprises However, according to Muriithi (2017) and Tuan(2020) there is no universally accepted definition for small and medium businesses In Vietnam, according to the Government's Decree 39/2018/ND-CP dated March 11, 2018, the classification of Vietnamese enterprises has been announced
First, the research article “Factors affecting performance of small and medium- sized enterprises in Vietnam” by Nguyen Thi Ngoc Oanh, Nguyen Kim Quoc Trung and Nguyen Thi Kim Chi et al (2021), from which data was collected from the Ho Chi Minh Stock Exchange (HOSE), determined the factors affecting the performance of Vietnam’s SMEs in the Ho Chi Minh City area from 2009 to 2019 These factors are firm age, firm size, leverage ratio, revenue growth, gross domestic product (GDP) growth, inflation rate, and quality of local governance ROE is used in this study as the dependent variable to represent the performance of firms With panel data and four methods: OLS, FEM, REM and FGLS, the researchers reach the conclusion that firm age, size and revenue growth positively affect SMEs' performance Among macroeconomic variables, GDP, growth, inflation rate and local governance have “ a notable impact on the performance of enterprises” Overall, the study identified seven factors in four internal factors (from the companies perspectives) and three external factors that have impact on business performance However, the study has ignored the effect of Covid-19 on SME businesses, therefore, with different time range, different dataset, the results can be varying
With the same topic, Vu Ngoc Xuan, Nguyen Thi Phuong Thu and Ngo Tuan Anh (2020) in the research paper “Factors affecting the business performance of enterprises: Evidence at Vietnam small and medium-sized enterprise” have the same variable with the previous research: revenue growth and the number of years of operation but in this study, conductor put great emphasis on internal factors, which are the number of state support, education level of the owner, social relationship of companies The performance of enterprises is measured by ROA (return on assets) With 456 SMEs collected from the Vietnam stock market, the study revealed that only growth rate, profitability of the previous year and industry cohesion have significant and positive influence on firm performance However, age (the number of years operating) did not affect the performance of the firm, which is a contrast conclusion with the study of Nguyen Thi Ngoc Oanh, Nguyen Kim Quoc Trung and Nguyen Thi Kim Chi et al
Last but not least, the research article “ Factors affecting business performance of firms in food industry in Vietnam” by Lai Cao Mai Phuong & Nguyen Thi Loi (2022) also used six various variables to measure the business performance of firms, which are age, size of the firms, leverage ratio, liquidity, the structure of fixed assets Similar with two previous research, revenue growth is also included in those independent variables The results shows that only size, liquidity has statiscally affected the performance of business while the structure of fixed assets, revenue growth has no statistical effect on business performance Overall, the article added various breakthrough findings on the literature However, the study only addressed companies in food industry, so the generalizability of findings also limit only in this industry itself
In sum, we find that the existing literature has contrast outcomes which triggers us to find our own answers with Vietnamese economic outlook and context Moreover, although many models in these aforementioned research are used such as FGLS or PLS-SEM, we want to use OLS model to re-test prior research Therefore, this hole in the literature motivates us to conduct the research.
Factors affecting the business performance of enterprises and research
Some studies suggest that the longer the companies have been working on this specific field, the better their financial results and operation effieciency given the more intensive and broad experience (Coad et al., 2013; Osunsan et al., 2015)
However, research by Loderer et al (2011), Ouimet and Zarutkskie (2014) expressed the concern that the age of enterprises can negatively affect their operation activities, making financial indicators such as ROA and ROE worse Because long-term enterprises often face difficulty in accessing modern production and business technologies and more reluctant to alter their labour structure, their competitive advantage is also lower than that of young enterprises
Age of business = the year chosen to evaluate (2019) - the year of firm’s establishment+1
Some researchers such as Malik (2011), Amoroso (2015), Serrasqueiro and
Nunes (2008), Asimakopoulos et al (2009), Mesut (2013) believe that the firm with more financial potential can easily build up reputation and also have many competitive advantages in trade negotiations and diversified products and services, thereby holding many opportunities to further enhance business performance
However, according to the research results of Kartikasari and Merianti (2016), Shehata et al (2017), firm size measured by total assets has a negative impact on business performance of the enterprise
Leverage ratio is a group of financial measurement that indicate the level and way in which a business uses borrowed capital for business activities, thereby assessing the financial risk (risk of default) of the company
These indicators are often used for the purpose of assessing the ability of an enterprise to fulfill its financial obligations by considering factors such as loan amount, equity, EBIT,
The results of the study by Iqbal et al (2018) and those of Egbunike and Okerekeoti (2018), Zeitun and Tian (2007), Kartiningsih et al (2020) ) has shown that the financial leverage ratio has a significant impact on business performance However, some other studies show that, the more financial leverage is used, the worsen the company business result is (Simerly and Li, 2000; Thuy et al., 2015)
In this essay, we will apply Debt to Asset ratio - one of the key measurement of leverage ratio
Debt- -toAsset= Total Debt/Total Asset
Liquidity ratios are a crucial indicator of a company's financial health and ability to meet short-term financial obligations without external financing They measure the extent to which a company's assets can be quickly converted into cash to cover liabilities These ratios provide insights into a company's cash flow management, working capital efficiency, and financial resilience, allowing stakeholders to assess the company's ability to meet its current debt obligations and manage its cash flow effectively.
Common liquidity ratios include the quick ratio, current ratio, and days sales outstanding
Research results of Akenga (2017), Khidmat et al (2014) confirm that liquidity has a positive impact on enterprises's performance Accordingly, liquidity has a positive impact on corporate performance measured through ROA and ROE However, research by Khalifa and Zurina (2013), Thuy et al (2015) shows that, if the enterprise maintains too high solvency, it means that the enterprise has invested too much in short-term assets , with the increase in operating costs, the profit of the business will decrease Current ratio= Current Asset /Current Liabilities
According to research by Memon et al (2012), asset structure measured through the ratio of long-term assets( fixed asset) to total assets has a negative impact on ROA This means that it has a negative impact on the performance of the business This result is also in line with the study of Zeitun et al (2007)
Asset structure = Fixed asset/ Total Asset
The debt ratio reflects the ability of a business to use its own capital to cover its debts The smaller this index is, the better it reflects the financial viability of the business, presenting its business performance The standard expected value of this indicator is not to exceed 1 So it is consider to have a reverse impact on business performance
Debt ratio= Total Liabilities/Total Shareholder’s Equity
RESEARCH METHODS AND RESEARCH MODEL
Research Design
In the essay, our team used Ordinary least square (OLS) regression analysis to build the model as learnt in Econometrics 1 course OLS is a statistical method of analysis that estimates the relationship between one or more independent variables and a dependent variable; the method estimates the relationship by minimizing the sum of the squares in the difference between the observed and predicted values of the dependent variable configured as a straight line Accordingly, our team analyzed the relationship between dependent variables (ROA), a financial measure that displays the firm's business performance, with six independent variables: the firm's age, size, liquidity ratio, leverage ratio, assets structure and finallly debt to equity ratio
The selection of firm is completed randomly to ensure the variety, applicability and generalization of the factors chosen By looking for enterprise's audit financial reports through company pages and official stock websites namely Cafef, Vietstockfinance, we have chosen 6 independent variables in 2019: the firm's age, size, liquidity ratio, leverage ratio, assets structure and debt to equity ratio and continuosly rechecking the numerical statistic, the credibility of the data source is ensured
After the data collection phase, the team processes data using the usual Ordinary least square (OLS), based on data found to test statistical significance and model suitability based on observations as well as similar prior studies From there, we find the best results will be harnessed to serve the analysis phase During the course of the study, the group implemented the knowledge of econometrics 1, macroeconomics, finance–currency, quantitative methods and software that support STATA, Microsoft Excel Microsoft Word to synthesize, process data and finally complete this essay Our group used the Stata software to scale the Ordinary Least Squared method (OLS) to estimate the parameters of multivariate regression models :
• Use the Ramsey Reset Test check to see if the model has missed any variable
• Use Correlation test in Stata software to find the correlation matrix between variables
• Use variance inflation factor (VIF) which measures the correlation and strength of correlation between the explanatory variables in a regression model to detect multicollinearity
• Use Breusch – Pagan test for verification of variance error change of the model and Robust Standard Errors to obtain unbiased standard errors of OLS coefficients under heteroscedasticity
Research model building
Our group collected data of 243 firm’s sample in accordance to the top 500 Vietnam companies of many different fields in 2019 through website Profit500 The selection of firm is completed randomly to ensure the variety, applicability and generalization of the factors chosen.The data is cross-sectional and secondary data concerning a myriad of Vietnam’s firms so we opt to search mainly on the Internet By looking for enterprise's audit financial reports through company pages and official stock websites namely Cafef, Vietstockfinance, for 6 independent variables in 2019: the firm's age, size, liquidity ratio, leverage ratio, assets structure and debt to equity ratio and continuosly rechecking the numerical statistic, the credibility of the data source is ensured
Information about the variables used in the model is described below:
Variables Meaning Units Expectation of sign
4 AST Structure of fixed assets
TABLE 1: Expected signs of variables and explanation of variables table
(Source: Collected by team with STATA)
We created the model below based on the secondary data from the firm’s financial report to investigate the impact of various factors on the business performance of enterprises in Vietnam:
ROA = f (Lev, Size, AST, Age, DE, Liq)
• AST: Structure of fixed assets
In order to investigate the association between above factors on the business performance of enterprises in Vietnam in 2019, our group used the theoretical basis It proposes a theoretical mathematical model as follows:
ROA = B0^ + B1^Lev + B2^Size + B3^AST + B4^Age + B5^DE + B6Liq + ui^
Interpretation of variables in model
1 ROA Returns on Assets Net profit/ total assets
2 Lev Leverage ratio Total debts/Total assets
3 Size Firm size The log of the ratio of the firm's assets
4 AST Structure of fixed assets
5 Age Firm age Age is calculated as one plus the difference between the investigation year and the firm's birth year
6 DE Debt to equity Total debt/ Total stockholders’ equity
7 Liq Liquidity ratio Measured by the current ratio (current assets/ current liabilities)
8 B0 The intercept term of the population regression function
9 B1 The slope coefficient of Lev in population regression function
Shows that if the leverage ratio increases (decreases) by 1% with other factors being constant, ROA increases (decreases) by B1%
10 B2 The slope coefficient of Size in population regression function
Shows that if the size of firms' size increases (decreases) by 1% with other factors being constant, ROA increases (decreases) by B2%
11 B3 The slope coefficient of AST in population regression function
Shows that if the structure of fixed assets increases (decreases) by 1% with other factors being constant, ROA increases (decreases) by B3%
14 B4 The slope coefficient of Age in population regression function
Shows that if the firm's age increases (decreases) by 1% with other factors being constant, ROA increases (decreases) by B4%
15 B5 The slope coefficient of DE in population regression function
Shows that if the DE increases (decreases) by 1% with other factors being constant, ROA increases (decreases) by B5%
16 B6 The slope coefficient of Liq in population regression function
Shows that if the liquidity ratio increases (decreases) by 1% with other factors being constant, ROA increases (decreases) by B6%
17 B0^ The intercept term of the sample regression function
18 B1^ The slope coefficient of Lev in sample regression function
19 B2^ The slope coefficient of Size in sample regression function
20 B3^ The slope coefficient of AST in sample regression function
21 B4^ The slope coefficient of Age in sample regression function
22 B5^ The slope coefficient of DE in sample regression function
23 B6^ The slope coefficient of Liq in sample regression function
24 ui Random error in the population regression function
Represents factors that are not included in the model affecting ROA
25 ui^ Residuals in sample regression function
Data Description
Data of Return on Equity, Firm age, Firm size, Leverage ratio, Liquidity ratio, Structure of fixed assets, and Debt to equity are taken from the official website of CafeF, Finance Vietstock, and the specific firms’ Financial Reports in 2019, which were published in the firm official websites The sample source of the data is taken from the top 500 Vietnam's best profitable firms Out of the 500 highly-ranked-profit firms in Vietnam, 236 enterprises fully met the sample size criteria The data set includes information from 236 firms in Vietnam, which is presented in the data set on the last page
Using Sum command to have the following table:
TABLE 3:The descriptive statistics table (Collected by the team with STATA)
To help readers have a better overview of the variables, our team used the Sum command in STATA to provide the statistical indicators of the variables In addition, this statistical description also helps to predict some errors that may occur when running the model due to the omission of the data.
ROA The average rate of Returns on Assets in 236 firms in 2019 was 8.974915, the standard deviation was 8.471996, the lowest ROA was -4, and the highest was 67.02.
Lev The average rate of the Leverage ratio in 236 firms in 2019 was 1.916229, the standard deviation was 21.40812, the lowest Lev was 0.02, and the highest was 329.3
Size The average rate of Size in 236 firms in 2019 was 29.04157, the standard deviation was 1.566282, the lowest Size was 21.58, and the highest was 34.47
AST The average rate of Structure of fixed assets in 236 firms in 2019 was 1.961271, the standard deviation was 20.94076, the lowest AST was 0, and the highest AST was 303.98
Age The average rate of Age in 236 firms in 2019 was 18.96186, the standard deviation was 14.31955, the lowest Age was 1, and the highest Age was 54
DE The average rate of Debt to equity in 236 firms in 2019 was 72.5678, the standard deviation was 45.17625, the lowest DE was 2, and the highest DE was 160
Liq The average rate of Liquidity ratio in 236 firms in 2019 was 2.216271, the standard deviation was 2.363462, the lowest Liq was 0.13, and the highest Liq was 22.18
It can be seen that the number of observations of the sample is quite large, and the values of the variables are also covered, so the sample can be representative of the population.
Using Corr demand to obtain the following table:
ROA Lev Size AST Age DE Liq
TABLE 4: Correlation table (Collected by team with STATA)
From this result, we can point out the following correlations among variables:
Analyze the relationship between the independent variables and the dependent variable
• r(ROA, Lev) = -0.0033 => The correlation coefficient is negative => Lev has negative relation with ROA
• r(ROA, Size) = -0.2679 => The correlation coefficient is negative => Size has negative relation with ROA
• r(ROA, AST) = -0.0259 => The correlation coefficient is negative => AST has negative relation with ROA
• r(ROA, Age) = 0.2312 => The correlation coefficient is positive => Age has positive relation with ROA
• r(ROA, DE) = -0.4312 => The correlation coefficient is negative => DE has negative relation with ROA
• r(ROA, Liq) = 0.1891 => The correlation coefficient is positive => Liq has positive relation with ROA
• Analyze the relationship among the independent variables
Positive relationship Negative relationship r(Lev, AST) = 0.9425 r(Lev, Age) = 0.1507 r(Size, DE) = 0.2589 r(AST, Age) = 0.1527 r(Age, Liq) = 0.0558 r(Lev, Size) = -0.3079 r(Lev, DE) = -0.0397 r(Lev, Liq) = -0,0090 r(Size, AST) = -0.3793 r(Size, Age) = -0.1168 r(Size, Liq) = -0.1457 r(AST, DE) = -0.0532 r(AST, Liq) = -0.0146 r(Age, DE = -0.1845 r(DE, Liq) = -0.3928
TABLE 5: Relationships among the independent variables
Conclusion: Because the correlation coefficient between the independent variables and the dependent variable are all non-zero, the dependent variable has a dependency on the independent variable; and between the independent variables there is also an interdependence which the Firm Age has the strongest correlation with inROA.
CHAPTER 3 ESTIMATION RESULTS, MODEL TESTING AND STATISTICAL INFERENCE
Estimates results and Model testing
The estimates results of regression coefficients with the OLS method are stated as below:
TABLE 6: Former estimate of model (Source: Gathered by team by STATA)
Adjusted coefficient of determination 0.2433 p-value 0.0000
Total Sum of Squares (SST) 16867.0584
Explained Sum of Squares (SSE) 4429.72812
Residual Sum of Squares (SSR) 12437.3302
TABLE 7: Results of former regression parameters
(Source: Gathered by team by STATA)
From the above results, regression model with estimates is:
ROA = 48.6400+0.1249lev-1.2604size-0.1840AST+0.0951age-0.0658DE+0.0158LIQ + 𝑢 𝑖
+ Residual Sum of Squares = 12437.3302, this is the sum of squares of residual (error) which can be explained by the model
+ Testing F(6,229) is used to test the suitability of regression function with the degree of freedom of explained ESS is k = 6, the degree of freedom of residual which can not be explained by the model is n k 1= 229 ( k = 6 is the number of independent – – variables of model, n = 236 is the number of observation)
+ The coefficient of determination ( 𝑅 2 )and adjusted 𝑅 2 :
𝑅 2 = 0.2626 indicates that 26.26% is the percentage of the variation of ROA is explained by six independent variables in the model (leverage ratio, size, structure of fixed assets, firm age, debt to equity, liquidity ratio) In other words, 73,74% of the variation of ROA is explained by other variables which are not included in the model However, it is proved that 𝑅 2 increases when the number of independent variables in the model increases whether those variables have statistically meaning or not Therefore, the probability of unecessary variables also increases with high 𝑅 2 , making it difficult to compare the suitability of models which have the same dependent variable and the different number of independent variables To solve this problem, adjusted 𝑅 2 = 0.2433 will be used to compare the suitability of models that have the same dependent variables and sample size but have different independent variables Moreover, adjusted 𝑅 2 also can be applied when researchers decide whether the model should have other new variables or not If adding new variables makes adjusted 𝑅 2 goes up and added ones have stastistical meaning, the researcher will agree to add those variables into model.
Model testing
In the general statistics field as well as in econometrics in particular, the Ramsey Regression Equation Specification Error Test (RESET) is a generic specification error test for linear regression models It examines if non-linear combinations of the fitted values aid in the explanation of the response variable
In this essay, we apply the Ramsey RESET to evaluate the accuracy of the variables used in our model
Using STATA software, test the model with command ovtest, we gain the result:
Ramsey RESET test using powers of the fitted values of roa
H : o model has no omitted variables
We have p-value>5%->Accept Ho
➔ Model has no omitted variables
Multicollinearity, a phenomenon in multivariate regression models, arises when independent variables exhibit substantial intercorrelations This interconnectedness among variables can lead to broader confidence intervals, undermining the reliability of statistical conclusions drawn from the model Consequently, the influence of individual independent variables on the dependent variable may be less accurately determined due to the presence of multicollinearity.
Using STATA software, test the model with VIF command, we gain the result as below:
(Collected by team with STATA)
According to the above table, we can see VIF of all 6 variables, which means that this model has imperfect multicollinearity However, because the VIF of all variables is less than 10, there is acceptable multicollinearity
3.2.3 Variance error change test (heteroskedasticity test)
Breusch Bagan/ Cook-Weisberg test for heteroskedasticity
Using STATA software, test the model with hettest command, we gain the result:
Variables: Fitted values of roa
Prob > chi2 = 0.0000 p-value At 5% significance level, reject 𝐻 𝑜 , accept 𝐻 1 : Model has variance change (heteroskedasticity )
We will later confront and fix this variance change defect at 3.2.5 “Resolve variable variance defects”
The normality test helps to determine how likely it is for a random variable underlying the data set to be normally distributed There are several normality tests such as the Skewness Kurtosis test, the Jarque Bera test,…
Thanks to our Econometrics 1 course, we were able to apply Skewness
Kurtosis test to check on the distribution of our chosen variables
Using the command “predict u, residual” and then “sktest u” we gained the result as below
Skewness/Kurtosis tests for Normality
Pr(Skrewness) Pr(Kurtosis) Adj chi2(2)
TABLE 9: Skewess result (Collected by team with STATA)
• At 5% significant level, reject Ho, accept H1
In STATA software, using reg roa lev size AST Age DE LIQ, robustcommand, we gained the result: roa Coef Std Err t P-value [95% Conf Interval] lev 0.1249475 0.0247321 5.05 0.000 0.0762159 0.1736791 size -1.260403 0.3480958 -3.62 0.000 -1.946283 -0.5745232 AST -0.1840961 0.0349596 -5.27 0.000 -0.2529797 -0.1152124 Age 0.095156 0.0561379 1.70 0.091 -0.0154569 0.2057689
DE -0.0658539 0.0110176 -5.98 0.000 -0.0875627 -0.0441452 LIQ 0.0158386 0.2768568 0.06 0.954 -0.5296738 0.5613511 _cons 48.64007 9.850092 4.94 0.000 29.23168 68.04847 Number of obs 236 F(6, 229) 20.12 Prob > 0.0000 F R-squared 0.2626 Root MSE = 7.3696
TABLE 10: Variable variance defect results
(Collected by team with STATA)
From the above table, we can witness that after implementing Robust Standard Error, the coefficient of 6 independent variables remained the same Furthermore, R- squared, which represents the suitability of the model, was also unchanged However, the standard errors was altered to robust standard errors thereafter changing p-value and t Our group had downsized the effect of variable variance defect
Statistical hypothesis test
3.3.1 Testing the statistical hypothesis test of coefficient terms
According to the estimated result from STATA using the OLS method, we obtained the Sample Regression Function (SRF) as below:
ROA = 48.64007 + 0.01249475*Lev - 1.260403*Size - 0.1840961*AST + 0.095156*Age - 0.0658539*DE + 0.0158386*Liq + ui^
The meaning of estimated coefficients:
• The constant term is estimated to be β0= 48.64007 When every explanatory variable equals 0, the expected value of Return on Assets (ROA) will be 48.64007% (other factors constant).
• The regression coefficient of Lev is estimated to be β1 = 0.01249475:
Holding other explanatory variables unchanged, if the Lev increases by 1%, the expected value of ROA will increase by 0.01249475%.
• The regression coefficient of Size is estimated to be β2 = - 1.260403:
Holding other explanatory variables unchanged, if the Size increases by 1%, the expected value of ROA will decrease by 1.260403%.
• The regression coefficient of AST is estimated to be β3 = - 0.1840961:
Holding other explanatory variables unchanged, if the AST increases by 1%, the expected value of ROA will decrease by 0.1840961%.
• The regression coefficient of Age is estimated to be β4 = 0.095156:
Holding other explanatory variables unchanged, if the Age increases by 1%, the expected value of ROA will increase by 0.095156%.
• The regression coefficient of DE is estimated to be β5 = -0.0658539:
Holding other explanatory variables unchanged, if the DE increases by 1%, the expected value of ROA will decrease by 0.0658539 %.
• The regression coefficient of Liq is estimated to be β6 = 0.0158386:
Holding other explanatory variables unchanged, if the Liq increases by 1%, the expected value of ROA will increase by 0.0158386 %.\
3.3.2 Testing the statistical significance of coefficient terms
Error t P- value [95% Conf Interval] lnlev 0.1249475 0.0247321 5.05 0.000 0.0762159 0.1736791 lnsize
TABLE 11: Significance of coefficient terms (Collected by team with STATA)
There are three methods to test the statistical significance of an individual regression coefficient i: The confidence interval method, the critical value method, and the p-value method In this paper, the team will use the p-value method from STATA to test the significance level of the coefficient terms
Reject H0; ln lev is statistically significant at α=5%
Reject H0; ln size is statistically significant at α=5%
Reject H0; ln AST is statistically significant at α=5%
Fail to reject H0; ln Age is statistically significant at α=5%
Reject H0; ln DE is statistically significant at α=5%
Fail to reject H0; ln LIQ is not statistically significant at α=5%
After testing the statistical significance of the coefficient of the independent variables, we can conclude the effect of AST, lev, size, and DE on ROA At a 5% significance level, only the Structure of Fixed assets, Leverage ratio, Size, and Debt to Equity ratio are statistically significant to Es’ ROA or business performance
Meaning of slope coefficients and Interpretation of regression results:
• 𝛽 0 H.64 (0.000): means that if all the other independents variables (AST, lev, size, DE) has a value of 0, the average Return on Assets of enterprises in 2019 is 48.64%
• 𝛽 1 =0.1249475 (0.000): Leverage ratio increasing by 1% will lead to ROA to increase by 0 1249475%, ceteris paribus This shows that the higher the leverage ratio, the more business efficiency
• 𝛽 2 =-1.260403 (0.000): Business size increasing by 1% will lead to ROA to decrease by -1.260403%, ceteris paribus The larger Es may have difficulty to effectively manage the organizational structure, from overcoming bureaucratic issues in the management structure On the other hand, the fact shows that the smaller the Es scale, the higher the profit level
• 𝛽 3 =-0.1840961 (0.000): Structure of fixed assets increasing by 1% will lead to ROA decreasing by -0.1840961%, ceteris paribus If the enterprise reduces the proportion of fixed assets, it will help businesses improve business efficiency It can be explained that: When businesses reduce investment in fixed assets, it will create conditions for improving business efficiency of the business as well as the value of the business In addition, fixed assets in production are essential This research results require that managers choose the right type of fixed assets to invest in to improve business efficiency It is a crucial foundation to help businesses improve their value and long-term performance
• 𝛽 4 =-0.0658539 (0.000): Debt to Equity ratio increasing by 1% will lead to ROA increasing by -0.0658539%, ceteris paribus According to the capital structure theory, the debt ratio increases the profitability of the Es by benefiting from the tax shield However, the debt ratio has a two-sided impact Debt is a lever for firms to increase revenue, thereby increasing profits At the same time, bad debt collection will easily push the firm to the edge of bankruptcy The benefits obtained from borrowing cannot compensate for the costs arising from debts 3.3.3 Testing and Statistical Inference
• 𝐻 0 : 𝑅 2 = 0 (the model is not suitable)
In conclusion, the proposed model is suitable
𝑅 2 = 0.2626 means that four independent variables (leverage ratio, size, structure of fixed assets, debt to equity,) can explain 26.26% the variation of dependent ROA.
CONCLUSIONS AND RECOMMENDATIONS FOR VIETNAMESE
Recommendations
First, according to the research result, firm size has a negative impact on the business performance of the business This is also supported by the research results of the study of Kartikasari and Merianti (2016), Shehata et al (2017) Therefore, it is not the higher size of the business which is measured by total assets, the higher the business performance of the business Thus, in the Vietnamese setting, firms should focus on quality over quantity For example, one of the most significant factors which can decide the quality of a company is high-quality human resources, so business owners can open their own human resource training courses, have policies to attract talented people Second, it is also reported that the structure of fixed assers has a negative impact on business performance of enterprises Although this variable was not statistically significant in the study of Nguyễn Th L i and Lị ợ ại Cao Mai Phương (2022) and but in this study, this variable was found to have statistical effect on ROA Therefore, enterprises need to reduce fixed assets by, for example, applying scientific and technological achievements to innovate production lines and improve production capacity 4.0 to diversify and improve product quality of the products
Third, in term of debt to equity, this variable has a negative impact on the business performance of the enterprise This has been demonstrated in the Vietnamese
According to the 2021 Whitepaper, businesses with lower debt-to-equity ratios tend to demonstrate higher corporate performance To achieve this, business owners should prioritize budgeting and debt management strategies to minimize financial risk By avoiding excessive or unsustainable debt, businesses can enhance their overall financial health and contribute to long-term success.
Fourth, as for financial leverage, it is drawn from the conclusion that the financial leverage variable in this study has a positive effect on the business performance of enterprises in Vietnam In other words, the smaller the debt ratio, the higher the business performance Previous empirical studies have studied this connection in several ways, including those by Ruland and Zhou (2005); Chandrakumarmangalam and Govindasamy (2010), and these authors conclude that leverage promote firmperformance Thus, the research team's recommendations for firms are to use effective capital savings in production and business stages combining with speed up working capital turnover in order to reduce the need for capital In addition, reasonable decisions should be made in determining funding sources Moreover, strengthening the exploitation of effective capital mobilization channels also can help businesses to access large and stable capital in the long run
In the model, there are two variables that are not statistically significant, namely the number of years of operation of the enterprise and the liquidity of the enterprise
Therefore, within the scope of the study, the authors recommend that Vietnamese firms should not rely on the above two factors to evaluate their business performance.
Conclusions 31 REFERENCES
Through the research process, our team draw the following conclusions: There are 4 factors affecting the business performance of enterprises, which are firm size, the structure of fixed assets, debt to equity and leverage ratio In which, only financial leverage has a positive effect on business performance of enterprises Among the factors that have a negative effect, firm size has the strongest negative effect and debt to equity has the weakest negative effect on firm performance
Despite extensive research on business performance in Vietnam, the study of its influencing factors remains relevant due to temporal and methodological variations This research aims to re-examine the impact of these factors within a specific time frame and sample size, exploring potential novel insights that may have emerged over time.
In term of recommendations, the research team suggested that Vietnamese enterprises should focus on improving human resources in three aspects: skills, knowledge and attitudes; applying the latest and advanced science and technology, have clear business plans; trying to use capital effectively to improve business efficiency In the current context, Vietnamese firms currently have favorable conditions to carry out the above proposals, so the proposals of the research team are completely reasonable and feasible
With limited qualifications, time and scope of research, this research paper definitely needs contributions and correction for further improvement We would like to thank Dr Dinh Thi Thanh Binh gave us the opportunity to research and learn deeply about this meaningful topic.