Effect of Firm Size on the Association between Capital Structure and Profitability

Abstract

This paper examines the association between capital structure and firm profitability through the moderating effect of firm size. The analysis of financing choice and firm profitability in developing countries is interesting because of the differences between the characteristics of their firms and those of companies in advanced economies. This research utilized secondary data obtained from the published annual financial statements of 156 manufacturing companies listed on the Tehran Stock Exchange (TSE) over the period 2011–2019. The given description was evaluated using a panel econometric approach, namely, the fixed-effects regression method. The results displayed that profitability is negatively affected by capital structure decisions. However, firm size is positively related to profitability. The robust findings also illustrate that the size of a firm plays a significant role in improving the influence of capital structure choice on the firm’s profitability. Similar to other emerging economies, in Iran, when internal finances are inadequate, long-term debt is used as an alternative for financing. These results provide evidence to support the hypothesis of the trade-off theory, which explains the relationship between firm size, capital structure, and profitability. These findings provide significant information from a developing country, Iran, that confirms the argument of the trade-off theory and provides substantial guidance for sector management.

1. Introduction

In general, a firm’s capital structure places emphasis on a mixture of debt and equity financing. Different financial choices are essential to a company’s financial health [1]. According to empirical evidence, a firm’s choice of funding sources, whether to raise debt or equity, significantly influences the value of the company [2,3,4]. Selecting an incorrect capital structure could result in a company experiencing financial difficulties and, ultimately, bankruptcy. For example, firms with prominent levels of long-term debt that are not utilized well are more likely to go bankrupt [5,6].

Moreover, financial planning to sustain public confidence involves reducing risk exposure by maintaining the debt-to-equity ratio, because the choice to increase long-term debt demonstrates how management moves risk away from the company to bondholders or creditors, leading to increased agency costs, which are a shareholder’s concern [7]. However, ref. [5] argued that companies with fewer non-debt tax shelters and less liquidity suffer from higher agency costs, and stock issuance during the debt reduction period is more likely to lead to quick failure. This is evidenced by the trade-off theory, which claims that firm leverage is determined by balancing the advantages of debt tax reduction against the costs of bankruptcy. Hence, the capital structure of a company is determined by managers to optimize the company’s net value. This highlights that the decision on capital structure has a significant impact on a firm’s profitability.

Ref. [8] proposed the Modigliani–Miller (MM) theory regarding capital structure issues and believed that firm value is not influenced by capital structure decisions. However, this theory was developed based on the assumption of perfect market efficiency, where no information asymmetry, transaction costs, taxes, or bankruptcy risks exist [9]. Furthermore, ref. [10] updated their earlier proposal with [11] and suggested the trade-off theory. According to this hypothesis, firms tend to raise their long-term debt to reduce their tax payments. The theory also states that a company with little debt on its statement of financial position performs better than a company that just uses financing through equity [12,13]. However, the pecking order theory put forward by [14] argues that firms should prioritize their funding sources, first by internal financing through net operating income, followed by debt; however, when debt is insufficient, equity must be used as a third source.

Prior studies on the relationship between capital structure and firm profitability have provided mixed results in both developed and developing countries; some found evidence to support the pecking order theory and concluded that profitability is inversely affected by leverage, such as the debt ratio [1,4,12,15,16,17,18]. Refs. [3,19] also reported a negative relationship, but in the case of long-term debt only. However, other studies support the trade-off theory and have revealed a positive association [20,21].

This study had several goals; first, it sought to assess the association between the capital structure and profitability of manufacturing companies from a developing country, specifically Iran. Firm size was utilized to build this connection. Furthermore, it also examined the correlation between firm size and profitability and intended to respond to the question of whether firm size has any impact on the relationship between financing choice and profitability. Iran serves as a country in transition that has emerged as a market with low growth and wants to expand its market in other countries. Therefore, Iran was used as the inspiration for our study.

Numerous studies have examined the concrete effect of capital structure on profitability and financial performance; however, the findings differ. For example, refs. [3,18] found a negative relationship, refs. [6,12,22] observed a non-linear connection, and ref. [23] reported a non-existent association. To the best of our knowledge, however, there are still very few studies that have examined the moderating effect of firm size as a sensitive variable on the connection between capital structure and profitability [6,16]. Hence, the current study serves as one of few investigations addressing the above arguments and has three main contributions: First, the argument of capital structure and initial profitability are discussed in a developed market. Consequently, empirical research on this topic turns to developing countries [3]. Recent trends show that developing market companies frequently employ more debt, particularly utilizing short- and long-term debt as a funding source, and they have comparatively higher current assets than fixed assets. As a result, companies in developing countries who have a higher level of debt tend to achieve greater profitability, to the detriment of their firm’s value. Hence, the present literature provides empirical evidence regarding the associations between firm size, capital structure, and profitability from a developing market, specifically Iran. Second, Iran, as an emerging economy, has a weak-to-medium financial system, and debt instead of equity is the primary source of funding [24]. Thus, it is important to determine how these factors affect the financing choice of Iranian firms and also to assess the capital structure of the industrial sector, as the biggest sector in Iran. Lastly, in order to evaluate the impact of capital structure choice on profitability and the moderating effect of firm size on the above correlation, this investigation used six models in its method. Three models without a moderating effect were employed to examine the relationship between capital structure and profitability, and three models with interactions were also used to assess the sensitivity of firm size on the association between financing choice and profitability, which is the novelty of this study.

The rest of the study is laid out as follows: Section 2 describes the literature review and hypotheses development, while the materials and methods are presented in Section 3. Section 4 illustrates the result and discussion, and finally, the conclusion and recommendations are presented in Section 5. The findings of our study have important consequences for portfolio investing and financial analysis. Moreover, the results are also useful for both equity investors and corporate managers that are obviously interested in creating value for capital structure choice.

2. Literature Review and Hypothesis Development

2.1. Theoretical Framework

Capital structure hypothesis initially started at the end of 1950s; the first theory was the irrelevance theory developed by [8], well known as the Modigliani–Miller (MM) theory. According to this concept, there is no ideal capital structure for companies [9,12]. After some revision, Modigliani and Miller proposed a trade-off model based on no tax assumptions [12]. According to [25], the trade-off theory argues that companies accept debt over equity when the benefit of debt, through a reduction in interest expenses prior to calculating income tax, is greater than the marginal cost of the debt. In other words, corporations possess a special optimum capital structure that compares the tax benefits of financing through debt with the costs of debt [9,26,27,28]. Firms that rely on debt financing have a greater value than companies that do not use debt [6,12]. However, debt should be utilized completely; otherwise, paying interest rates may become a burden for the firm.

On the other hand, the pecking order theory can be seen as the most popular hypothesis of capital structure that clarifies how capital structure, firm performance, and profitability are related to each other. This theory was developed by [14] and claims that firms should use a hierarchy in obtaining their source of financing to cover their capital requirements and improve their profitability [26,29], and this is also an effective strategy to reduce the issues of information asymmetry as well [30]. This theory recommends that firms should use retained earnings first because they hold a lower risk. However, when internal financing is not sufficient, firms turn to less risky debt to prevent missing potential investments. As a last option, when they are unable to acquire debt, firms can issue new stocks or equity instead [4,9,12,13,25,26,31].

Ref. [32] investigated the relationship between debt financing and firm value by adopting both pecking-order and trade-off methodologies for Taiwanese listed firms based on 2548 observations. The findings illustrated that when the ratio of debt is lower than 10%, firm value rises by 0.05%, with a 1% rise in the debt ratio. Additionally, when the ratio of debt is in a range between 10% and 33%, firm value is improved only by 0.005% with a 1% rise in the debt ratio. These results are completely evidenced by the trade-off theory, which argues that firms have a static amount of loan that encourages managers to seek the ideal capital structure. Thus, firm value increases when the benefit of debt financing is comparable to the marginal debt expenses.

Ref. [33] evaluates the role of pecking order theory on explaining the behavior of leverage in Pakistani listed firms. The study used 34 years of uniformed financial data for manufacturing firms. The results show that both recent and previous profitability are negatively impacted by leverage, while past dividends are positively affected by leverage. Similarly, ref. [27] examined the conclusive evidence on the pecking order and trade-off theory of capital structure. The study covered non-financial listed firms on the Indonesia Stock Exchange (IDX) during 2014–2017. The findings illustrated that the firms focus on financing decisions depending on their financial health. Firms with both lower and higher levels of leverage who face financial deficiencies frequently raise their amount of debt. Alternatively, corporations with financial surplus prefer to raise their power, regardless of how much debt they have. These findings are strongly supported by the pecking order theory which claims that firms should use internal sources first, such as “retaining earnings”; then, debt must be adopted by firms, and when the debt is not accessible by the firm, equity can be released as a last report.

2.2. Empirical Evidence and Hypotheses Development

Numerous academic studies have examined the potential connection among firm size, capital structure, and firm value in various economic or business industries, such as non-financial firms [6,34]; manufacturing firms [35]; the mining industry [4]; the pharmaceutical sector [36]; the construction and property industry [37]; the banking sector [21]; and listed firms on the stock exchange [3,16,25].

2.2.1. Capital Structure and Profitability

As the current research seeks to objectively explore the connection between firm size, capital structure, and profitability, our review of the literature will concentrate on work in this field. It is commonly believed that in the actual world, where markets are not perfect, capital structure has a significant influence on firm profitability or firm performance, and theoretically, this is proven by many hypotheses in the literature [34]. However, previous investigations provide a variety of findings about this connection. According to [38], the exact relationship between financing choices and firm profitability may vary depending on the conditions. We found that the nature of the association between capital structure decisions and profitability and performance tend to be influenced by certain circumstances, such as the size of the firm, the economic condition, and the country’s level of development.

In this context, numerous studies demonstrate that capital structure influences firm performance favorably in developed countries that have established strong financial or economic systems [38,39]. However, other investigations have reported a negative relationship between capital structure decisions and profitability in developing countries [1,3,17,18,31,40]. These results are supported by the pecking order theory which claims that highly productive firms do not rely heavily on external financing.

Further, using balanced panel data of several manufacturing firms listed on the Iraqi stock exchange in the 2004–2020 period, ref. [12] empirically tested the correlation between capital structure and finance performance that is measured using return on assets (ROA) and market-to-book value (MBV). They found that ROA and MBV are inversely impacted by leverage. Likewise, ref. [41] examined a similar industry based on the number of manufacturing firms listed on the Indonesia Stock Exchange over the years 2008–2015 and showed that there is a strong and negative relationship between capital structure and firm profitability. In other industries, such as banks, refs. [42,43] reported a negative and significant relationship between capital structure and firm performance, while [44,45] found a positive relationship. Other studies indicated a non-linear association, meaning that both negative and positive connections were found [34,46,47,48], and some other investigations found a non-significant relationship [49,50]. Thus, the following hypothesis is proposed:

Hypothesis 1. 

Profitability is significantly affected by capital structure.

2.2.2. Firm Size, Capital Structure, and Profitability

Despite the extensive testing of the association between capital structure and profitability, firm size as a moderator is expected to have a significant contribution to this relationship. Previous investigations indicated that multiple market values are caused by different firm sizes [3,51,52,53,54,55,56,57]. In addition, refs. [3,51,52,56,58,59] found that firm size has a positive and considerable influence on profitability. Further, [53] illustrated that small companies are more capable than big companies to evaluate their value. However, ref. [4] demonstrated an adverse connection. Others, such as [54], reported mixed findings, and no indicative association was found by [55].

In addition, a few studies have examined the moderating effect of firm size on the relationship between capital structure and financial performance. Ref. [60] conducted a study on non-linear associations among the size of the firm, capital structure, and profitability. The study used a huge sample of 1194 companies in the Indian industrial sector, which is publicly traded. The period considered was 2005 to 2014. The findings clearly indicated that the impact of the capital structure on firm profitability can be determined by the size of the firm. Likewise, ref. [6] examined the sensitivity of firm size to the relationship between financing decisions and the value of the firm. The study utilized the market values and financial statements of 1638 firms listed on the Indonesian stock exchange over the period 2012–2018. The results showed that the size of a company influences the effect of capital structure decisions on the company’s value. The extensive theory and empirical evidence from prior investigations that are related to the relationship between the size of the firm and financing choices, and between the firm size and profitability, illustrate that firm size is a reliable factor for determining capital structure decisions and profitability [3,6,51,52,56,58,59,60,61]. Firm size can be seen as a crucial component of capital structure due to its ability to serve as an indicator of a company’s sustainability, and it also justifies additional studies into how company size affects the connection between financing choice, profitability, and market value [6]. Based on the above arguments, we propose the second hypothesis as follows:

Hypothesis 2. 

Firm size is expected to moderate the relationship between capital structure and profitability.

3. Methodology

3.1. Sample and Data Collection

We obtained the financial data from annual reports of manufacturing firms listed on the Teheran Stock Exchange (TSX) during 2011–2019. Initially, our preliminary sample included all manufacturing firms. This is because the industrial sector has a significant role on Iranian economic growth and has an enormous capital structure that consists of debt and equity [24], and this information is published in their annual report adequately. However, the COVID-19 pandemic and other financial crises led to the delisting and closure of some listed manufacturing firms. Hence, the study collected data from 156 manufacturing firms during 2011–2019. In total, there were 1404 observations. The sample firms were spread out throughout thirteen different industries, as identified by the Tehran Stock Exchange (TSX) (Tehran, Iran).

3.2. Variables and Their Measurements

3.2.1. Profitability

Profitability is considered to be a dependent variable in this study, and theoretically, capital structure, as an internal factor, directly affects profitability. In the previous literature, profitability was measured using different accounting indicators, such as return on assets, return on equity, Tobin’s Q, earnings per share, and assets turnover [16,17,18,42,46,62,63,64,65,66]. This study used return on assets (ROA), Tobin’s Q (TOBQ), and earnings per share.

3.2.2. Capital Structure

Capital structure can be seen as a combination of debt and equity in the corporate form of funding [67]. In this study, capital structure is an independent variable and can be considered an important factor that explains the findings of this investigation. According to the related theories, capital structure is expected to have a significant impact on firm profitability [12]. Prior investigations measured capital structure using different proxies, such as short-term debt ratio, long-term debt ratio, total debt ratio, equity ratio, equity multiplier, and debt-to-market capitalization ratio [1,12,16,17,18,64,68,69,70,71,72]. Our study measured capital structure using both debt-to-market capitalization ratio and total debt ratio.

3.2.3. Firm Size

Big firms are different from small firms in several ways, and this leads to different impacts on firm profitability [51]. Hence, the correlation between capital structure decisions and profitability can be strengthened or weakened by firm size. The current study used firm size as a dependent variable. Firm size also has a role as a moderator to influence the relationship between capital structure and profitability. Firm size is measured in the literature using total assets, total sales, number of employees, and number of members [1,6,28,69,73,74]. The present study used total sales for measuring firm size.

3.2.4. Control Variables

There are several factors that are thought to influence capital structure decisions, yet some of them are important in certain countries but not in others [75]. In order to assess the relationship between firm size, capital structure, and firm profitability accurately, two control variables were used in our research. We included these variables in the regression models to protect and control the industry characteristics and reduce the selection bias. In line with the previous literature [31,71,76], the present study used sales growth and firm age as control variables. According to [22,31,77], firm performance and profitability are positively impacted by both sales growth and firm age, while [70,71] reported a negative link between performance and firm age.

Table 1. Definition of variables.

Variable		Notation	Proxies	Definition
Dependent Variable	Profitability	ROA	Return On Assets	Net income/Total assets
		TOBQ	Tobin’s Q	(Market value of equity + book value of debt)/book value of assets
		EPS	Earnings Per Share	Net income/Number of outstanding shares
Independent Variable	Capital Structure	TDR	Total Debt Ratio	Total debt/Total assets
		DMCR	Debt to Market Capitalization Ratio	Total debt/(Total debt + Market capitalization)
Exogenous and Moderating Variable	Firm Size	TS	Total Sales	Natural log of total sales
Control Variable		ASG	Annual Sales Growth	The percentage of increasing sales compared to the year before
		FAGE	Firm Age	Natural logarithm of the number of years in service since the company began operations

summarizes the definition of the study variables.

3.3. Empirical Model

We employed an econometric model for testing the study hypotheses as follows:

$$\begin{gathered} Profitabilit{y}_{it}={\beta }_0+{\beta }_{1}Capital structur{e}_{it}+{\beta }_{2}Firm siz{e}_{it} \\ \hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}+{\beta }_3\left(Capital structur{e}_{it}\times Firm siz{e}_{it}\right)+{\beta }_{4}Control{s}_{it}+E_{it} \end{gathered}$$

where $i$ indices the observed companies and $t$ indicates the time, ${\beta }_0$ represents a constant, ${\beta }_1$ to ${\beta }_4$ is an array of independent factors, and capital structure, as an independent variable, is proxied by the total debt ratio (TDR) and the debt-to-market capitalization ratio (DMCR). Firm size is an independent and moderating variable and is proxied by total sales (TS). $Controls$ represents the control variables and $E_{it}$ is an error term. In addition, the impact of capital structure on profitability is investigated using multiple regression methodology. This approach examines the connection between a single dependent variable and multiple independent variables [8]. According to [9,78], for testing two connections, a multiple regression approach can be applied, and it is useful. Hence, the current study used this method to analyze the association between capital structure, profitability, and the moderating impact of firm size. The expanded regression equation for assessing the above relationship can be proposed as follows:

Models without interaction:

$$Profitability {\left(ROA, TQ, EPS\right)}_{it}={\beta }_0+{\beta }_{1}TD{R}_{it}+{\beta }_{2}DMC{R}_{it}+{\beta }_{3}T{S}_{it}+{\beta }_4{C}_{it}+E_{it}$$

Models with interaction:

$$\begin{gathered} Profitability {\left(ROA, TQ, EPS\right)}_{it} \\ \hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}={\beta }_0+{\beta }_{1}TD{R}_{it}+{\beta }_{2}DMC{R}_{it}+{\beta }_{3}T{S}_{it}+{\beta }_4\left(TD{R}_{it}\times T{S}_{it}\right) \\ \hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}+{\beta }_5\left(DMC{R}_{it}\times T{S}_{it}\right)+{\beta }_6{C}_{it}+E_{it} \end{gathered}$$

4. Empirical Results and Discussion

4.1. Descriptive Analysis

Table 2. Statistics summary.

	ROA	TOBQ	EPS	TDR	DMCR	TS	ASG	FAGE
Mean	0.13	2.06	0.96	0.59	0.39	14.35	0.28	3.55
Std. Dev.	0.15	1.67	1.58	0.22	0.22	1.80	0.50	0.42
Minimum	−0.50	0.58	−6.28	0.04	0.02	8.90	−0.83	1.95
Maximum	0.66	20.55	16.90	2.08	0.93	20.58	6.60	4.27
Skewness	0.44	4.42	3.49	0.62	0.31	0.70	4.24	−0.74
Kurtosis	4.188	2.12	24.96	6.13	2.16	3.89	42.07	2.91

shows the descriptive statistics of profitability, capital structure, firm size as a moderator, and corporate characteristics that are significant in our study. The reported analysis illustrates that among the measurements of profitability, TOBQ has the highest mean value of 2.06, with a deviation of 1.6. The minimum and highest values of TOBQ are 0.5 and 20, respectively, while the ROA has the lowest mean value of 0.13, with a standard deviation of 0.1. The lowest and highest values of ROA are −0.5 and 0.6, respectively. EPS, as a last profitability indicator, has an arithmetic mean (M = 0.96, SD = 1.5, Min = −6.2 and Max = 16.9). In addition, TDR and DMCR, as proxies of capital structure, have (M = 0.59, SD = 0.2, Min = 0.04 and Max = 2) and (M = 0.39, SD = 0.2, Min = 0.02 and Max = 0.9), respectively. The indicator of agency cost is TS and has the highest arithmetic mean among all variables (M = 14.3, SD = 1.8, Min = 8.9 and Max = 20.5). The mean value of the control variables is (M = 0.28, SD = 0.5, Min = −0.8 and Max = 6.6) for ASG and (M = 3.5, SD = 0.4, Min = 1.9 and Max = 4.2) for FAGE.

Further, a study by [79,80] demonstrated that standard skewness must be less than 1.9 and standard kurtosis must be less than 3. In our results, which are presented in Table 2, most variables do not achieve the standard range. Hence, we conclude that our data sets are not normally distributed. However, non-normal distribution of data is not expected to be a problem when the study includes a large sample size [81,82]. All variables are positively skewed except the FAGE, which is negatively skewed, as shown in Table 2.

Refs. [83,84] also argued that if the value of kurtosis is less than 2, it is considered “platykurtic” distribution; if it is equal to 3, it is “mesokurtic” distribution; and if it is smaller than 3, it is “leptokurtic” distribution. Therefore, Table 2 displays that ROA, EPS, TDR, TS, and ASG are distributed based on “leptokurtic” distribution; on the other side, DMCR and FAGE are considered “platykurtic” distribution because their value is greater than 3. From the above explanations and by applying the “Jarque-Bera”, normality test, we can confidently accept the alternative and reject the null hypothesis.

4.2. Correlation Analysis

Table 3. Correlation matrix.

Panel A Correlation Results									Panel B VIF
	ROA	TOBQ	EPS	TDR	DMCR	TS	ASG	FAGE	VIF	1/VIF
ROA	1
TOBQ	0.28 ***	1
EPS	0.65 ***	0.16 ***	1
TDR	−0.65 ***	−0.18 ***	−0.32 ***	1					2.08	0.47
DMCR	−0.63 ***	−0.59 ***	−0.38 ***	0.40 **	1				2.16	0.46
TS	0.24 ***	−0.01	0.18 ***	−0.007	−0.02	1			1.01	0.98
ASG	0.27 ***	0.31 ***	0.15 ***	−0.11 ***	−0.23 **	0.11 ***	1		1.07	0.92
FAGE	−0.14 ***	0.05 *	−0.14 ***	0.10 ***	0.05	0.004 *	0.02	1	1.01	0.98
									VIF Mean	1.47

***, **, * Correlation is significant at 1%, 5%, and 10% level, respectively (two-tailed).

(panel A) demonstrates the association between all variables including the dependent, independent, moderating, and control variables. The profitability measurements (ROA, TOBQ, and EPS) are inversely related to the indicators of capital structure (TDR and DMCR). TS, as a moderator, has a significant and positive impact on ROA and EPS, while TOBQ has a negative relationship with TS. ASG, as the first control variable, has a positive association with all profitability proxies. However, the correlation between FAGE with ROA and EPS is negative, but in the case of TOBQ, the relationship is positive.

Moreover, this research employs the sample panel data of 156 manufacturing companies over a 9-year period (2011–2019). Thus, multicollinearity and collinearity issues must be taken into consideration. Multicollinearity and collinearity can be a problem when the variance of two or more variables overlap excessively, and this has significant impact on the results of regression. In order to test the aforementioned issues in our study, correlation analysis was utilized firstly to assess the relationship between the exogenous variables. According to [85,86], when the correlation between two variables is greater than 70%, collinearity is considered an issue. The results in Table 3 (Panel A) display that there is no meaningful relationship between independent variables whose correlations are greater than 70%.

Secondly, we also tested the occurrence of multicollinearity problems using two widespread measurements (variance inflation factor (VIF) and tolerance). The tolerance value should be higher than 0.1, and the acceptable rate for VIF should be less than 6 [87,88,89]. Table 3 (Panel B) illustrates that the maximum value of VIF is 2.16 and the lowest tolerance value is 0.46. Therefore, this study is free from the issues of collinearity and multicollinearity.

4.3. Results of Panel Unit Root Tests

There are many different types of panel unit root tests. To enhance the accuracy of our findings, our study utilized various tests, such as, “Hadri-Z”, developed by [90], PP–Fisher, proposed by [91], and LLC, which was educated by [92]. In LLC, the unit root is assumed to be present in all cross-sections equally. However, the aforementioned assumption was disproved by Hadri-Z- and Fisher-type testing. Thus, the unit root is expected to vary among cross-sections for those types of tests. As shown in

Table 4. Results of unit root test.

Method	Unit Root In	Variables
		ROA	TOBQ	EPS	TDR	DMCR	TS	SG	FAGE
LLC	Level	−24.4 ***	−13.5 *	−26.3 ***	−27.0 ***	−16.6 ***	−15.0 ***	−14.4 ***	−70.7 ***
PP–Fisher	Level	423.5 ***	107.8 *	378.0 ***	364.5 **	682.6 ***	127.2 *	469.7 ***	2855.5 ***
Hadri-Z	Level	29.1 ***	26.1 **	31.7 ***	30.7 ***	30.1 ***	25.2 ***	38.6 ***	25.5 ***
Results		Reject H0	Reject H0	Reject H0	Reject H0	Reject H0	Reject H0	Reject H0	Reject H0

***, **, and * display significance at the 1%, 5%, and 10% level, respectively.

, all variables that were used in our study were determined to be stationary at level I(0). Therefore, the null hypothesis of unit root is rejected by all study variables at the 1%, 5%, and 10% level. ROA, EPS, DMCR, SG, and FAGE are significant at the 1% level, while TDR and TS are significant at 1% in the case of LLC and Hadri-Z only. TOBQ is significant at 10% for LLC and PP-Fisher, and at 5% for Hadri-Z.

4.4. Model Specification

In panel data regression methodology, there are three different approaches, such as the “Common Effects Method (CEM), Random Effects Method (REM)”, and Fixed-Effects Method (FEM). For selecting the ideal model, the first two analytical phases were performed. Initially, the Chow test was used to compare the results between CEM and FEM (if the p-value is less than 5%, then FEM can be selected) [93,94]. In addition, the Hausman test was utilized to assess the differences between REM and FEM (if the p-value is less than 5%, FEM can be selected as a suitable model) [95]. Finally, the Lagrange Multiplier (LM) test was employed in our study to compare the results between FEM and CEM. LM diagnosis is performed when the results of the Chow test and the Hausman test are different.

Table 5. Testing suitable model selection.

Name of the Test	Model 1 ROA		Model 2 TOBQ		Model 3 EPS		Outcomes
Hausman Test	Chi-Sq. Statistic	Prob.	Chi-Sq. Statistic	Prob.	Chi-Sq. Statistic	Prob.	Accept null
	141.44	0.000	74.05	0.000	52.73	0.000
Chow Test	Statistic	Prob.	Statistic	Prob.	Statistic	Prob.	Reject null
	1064.11	0.000	186.36	0.033	1053.54	0.000
Lagrange Multiplier Test	Breusch Pagan	Prob.	Breusch Pagan	Prob.	Breusch Pagan	Prob.	Accept null
	40.77	0.000	50.06	0.000	42.93	0.000

shows the results of the Lagrange Multiplier test, the Chow test, and the Hausman test. The results revealed that the probability value for all models in each test is less than 5%. Therefore, the null hypothesis can be confidently rejected, and we can select the Fixed-Effects Method (FEM) as the appropriate methodology in this study for predicting the relationship between firm size, capital structure, and profitability.

5. Discussion

5.1. The Impact of Capital Structure on Profitability

The major impact of the total debt ratio (TDR) and debt-to-market capitalization ratio (DTMC) on ROA, TOBQ, and EPS is presented in models 1, 2, and 3 without interaction in

Table 6. Results of Fixed-Effects Method (FEM)-ROA and TOBQ.

Variable	Model 1 ROA				Model 2 TOBQ
	Without Interaction		With Interaction		Without Interaction		With Interaction
	Coef.	t-Stat.	Coef.	t-Stat.	Coef.	t-Stat.	Coef.	t-Stat.
C	0.780 ***	8.876	0.428 ***	4.351	−3.545 ***	−2.450	−3.097 *	−1.869
TDR	−0.289 ***	−16.075	0.635 ***	4.835	3.468 ***	11.712	4.530 **	2.047
DMCR	−0.135 ***	−7.948	−0.471 ***	−3.564	−6.813 ***	−24.369	−9.695 ***	−4.348
TS	0.044 ***	7.519	0.077 ***	10.814	0.387 ***	4.010	0.376 ***	3.108
TS×TDR			−0.067 ***	−7.130			−0.075	−0.471
TS×DMCR			0.024 ***	2.608			0.202 *	1.299
ASG	0.040 ***	9.904	0.041 ***	9.234	0.342 ***	4.584	0.340 ***	4.528
FAGE	−0.304 ***	−9.816	−0.333 ***	−10.860	0.155	0.305	0.075	0.146
R-Square	0.788		0.798		0.552		0.557
Adj. R-Square	0.761		0.771		0.494		0.499
F-statistic	28.963		30.318		9.580		9.478
Prob.	0.000		0.000		0.000		0.000

Note: ***, **, and * indicate the level of significance at 1%, 5%, and 10% respectively.

and

Table 7. Results of Fixed-Effects Method (FEM)- EPS.

Variables	Model 3 EPS
	Without Interaction		With Interaction
	Coef.	t-Stat.	Coef.	t-Stat.
C	2.465 **	1.969	−1.244	−0.877
TDR	−1.661 ***	−6.483	5.518 ***	2.913
DMCR	−0.672 ***	−2.780	−0.430 *	−0.225
TS	0.455 ***	5.445	0.774 ***	7.458
TS×TDR			−0.529 ***	−3.860
TS×DMCR			−0.012 *	−0.096
ASG	0.295 ***	4.567	0.263 ***	4.097
FAGE	−1.944 ***	−4.404	−2.122 ***	−4.787
R-Square	0.623		0.631
Adj. R-Square	0.574		0.583
F-statistic	12.840		13.132
Prob.	0.000		0.000

Note: ***, **, and * indicate the level of significance at 1%, 5%, and 10%, respectively.

. These findings display that ROA and EPS are inversely affected by TDR, with a coefficient value of −0.289 and −1.661, respectively, and the results are significant at the 1% level. However, the association between TOBQ and TDR is positive, with a value of 3.468 at the 1% significance level. This implies that a 1% rise in TDR has a marginal effect on ROA and EPS by 0.28 and 1.66 percent, respectively, and in ascending order on TOBQ by 3.46 percent. Additionally, as illustrated in Table 6 and Table 7, DMCR has a negative and significant relationship with all profitability indicators without a moderating effect with a coefficient value of −0.135, −6.813, and −0.672, and these findings are significant at the 1% level. This shows that a 1% increase in DTMC brings about a decrease in ROA, TOBQ, and EPS by 0.13, 6.81, and 0.67 percent, respectively.

These results are consistent with the explanation of the pecking order theory, which argues that companies need to collect their source of funding in a structured way to meet their capital demands and increase their profitability [26]. This theory also suggests that firms must employ retaining earnings as a first option then debt and stock issuance in the case of insufficient debt [25,26]. This suggests that a rise in total debt is related to a decline in profitability. This is illustrated by the reality that using a high percentage of total debts might result in low profitability because they have interest expenses and other costs as well. These findings are also similar with the previous results, such as those of [1,3,12,17,18,31,34,42,43,46,47,48], who reported a negative association between debt ratios, firm performance, and profitability. Thus, the first hypothesis—that profitability is significantly affected by capital structure—is accepted.

Further, Figure 1, Figure 2 and Figure 3 below, Figure 4, Figure 5 and Figure 6 show the graphical residual, fitted, and actual values after regressing the exogenous variables. All figures with and without a moderating effect display a clearly visible trend line, and the close association of the actual and fitted lines increases the homoscedasticity by decreasing the residuals.

5.2. The Sensitivity of Firm Size on the Relation between Capital Structure and Profitability

The panel regression results of the Fixed-Effects Method (FEM) in Table 6 and Table 7 demonstrate that the F-statistics for all the moderating models (models 1, 2, and 3 with interaction) has an appropriate and acceptable rate for the independent variables, and the value of the adjusted R square increased after using the firm size as a moderator. This normally means that the correlation between capital structure and profitability is significantly influenced by the moderating effect of firm size.

Moreover, the robust findings also display that “TS*TDR”, as a moderator, has a negative and significant effect on ROA and EPS, with values of −0.067 and −0.529 percent, respectively; however, this result is statistically insignificant in the case of TOBQ, as presented in Table 6 and Table 7, model 2 without interaction. Additionally, “TS* DMCR” is positively related to ROA and TOBQ, with values of 0.024 and 0.202 percent, respectively, but in case of EPS, the association is negative, with a coefficient value of −0.012 percent. These results are significant at the 1% level for ROA and 10% for TOBQ and EPS, as shown in Table 6 and Table 7. These results support the second hypothesis—that firm size as a moderator is sensitive to the correlation between capital structure and profitability. The above findings also support the trade-off theory, which argues that small companies with larger ratios of debt, but less earnings, have a higher firm value. Nevertheless, large companies that have lower levels of debt, but greater profitability, have a lower firm value [6,10,11,51]. This is because big companies take advantage of economies of scope, which are cost savings resulting from the combined manufacture of two separate goods rather than manufacturing them individually. When companies achieve a particular scale, they might realize that integrating the production process is useful for them in terms of cost reduction. For instance, it is more economical to produce an essential product rather than paying another company to produce it. This is because a rise in costs resulting from the joint manufacturing of products is lower than the cost that is initially determined by purchasing a similar product from another company. The study results are also similar to the findings of [3,52,53,54,55,56], who argue that multiple market values are caused by different firm sizes.

6. Conclusions

Our primary objective is to provide empirical evidence about the role of firm size sensitivity on the relationship between capital structure and the profitability of manufacturing companies that were listed on the Tehran Stock Exchange (TSX) during 2011–2019. To achieve the aim of the study, several econometric models were used to analyze the above correlation. The Fixed-Effects Method (FEM), as an acceptable and appropriate methodology, was employed in this study for data analysis. Capital structure is an independent variable and is measured using the total debt ratio (TDR) and debt-to-market capitalization ratio (DMCR). Firm size plays is an independent and moderating variable and is proxied by total sales (TS). The dependent variable, however, is profitability, and it is indicated by three indicators: return on assets (ROA), Tobin’s Q (TOBQ), and earnings per share (EPS).

The findings revealed that without the moderating effect, TDR is related to ROA and EPS negatively, but positively with TOBQ, and the results are statically significant, as illustrated in Table 6 and Table 7. DMCR, as a second proxy of capital structure, has a negative and significant relationship with ROA, TOBQ, and EPS. The relationship between TS and all profitability indicators is positive and statistically significant. In addition, the results of panel regression after the moderating effect demonstrate that ROA, TOBQ, and EPS are negatively correlated with TS*TDR, but the result is statistically insignificant in case of TOBQ only. TS*DMCR is associated with ROA and TOBQ positively, but negatively with EPS. The study results correspond to all research hypotheses and are confirmed by the trade-off theory. The hypothesis of the trade-off theory argues that small- and medium-sized companies with a greater level of debt, but lower profitability, possess a greater firm value. On the other hand, large companies with fewer levels of debt, but greater profitability, have a lower value for the firm. However, increasing the level of long-term debt may bring about a decrease in profitability.

The primary contributions of these findings illustrate how firm size in an emerging economy, such as Iran, has a significant influence on controlling the relationship between capital structure choice and improving profitability. The results indicate that, by improving the effect of capital structure decisions on optimizing firm profitability, big firms, such as industrial firms, have an immense value through improving total assets. When firms boost their value, they can increase the possibility of utilizing debt or other financing strategies that are more ideal. Thus, industrial companies deciding on a capital structure scheme is the most effective approach for improving profitability and increasing their value as well.

This new insight may be used by investors to make decisions regarding investment through determining how important the effect of firm size is on firm profitability. This investigation might also help extend the literature on capital structure and profitability in developing economies by examining the moderating influence of firm size. Moreover, these robust results can also be valuable for executives and stockholders; capital structure decisions are useful for increasing profitability, and the size of the firms has a significant influence on building this relationship.

Some limitations of our study can be highlighted. First, according to the theories of capital structure, this analysis demonstrates that firms achieve the best financing decisions within specific circumstances. Nevertheless, this research does not provide in-depth analysis about the time and conditions needed for the optimum system of capital structure to appear. Therefore, future research is recommended to analyze the firm value in the context of time and circumstances surrounding the ideal capital structure in detail. Secondly, the issue of information asymmetry is also linked to the decision on financial choices, and this investigation is not included in the study model. Hence, future studies are recommended to use information asymmetry as an important variable to examine the above associations. Finally, considering a large number of emerging economies might be significant to determine whether the outcomes can be expanded to all countries or it is particular to countries that have a specific economic condition.