Does Growth Affect Firms’ Leverage Adjustment Speed? A Study of Indian Firms

Abstract

The significance of firms’ growth opportunities as one of the determinants of leverage is documented in many prior studies. But, there are not enough studies which examine the impact of growth on leverage adjustment speed. In this backdrop, the present study investigates the relationship between growth and leverage adjustment speed. Second, the study also examines the moderating role of two dimensions of target deviation, that is, nature and level of deviation in the relationship between growth and leverage adjustment speed. Using partial adjustment model on a dataset of 28,532 firm-year observations comprising 2,718 listed Indian firms with 4–12 years data for each firm, the study observes faster leverage adjustment speed for high-growth firms (36%) than low-growth firms (24%). The results also confirm the moderating effect of target deviation in the relationship between growth and adjustment speed. Overall, the study concludes that firms’ growth opportunities cause asymmetries in target adjustment speed by altering the costs and benefits of adjustment, and nature and level of target deviation moderates the relationship between growth and adjustment speed. These findings are expected to have substantial practical implications for financial managers in their capital structure decisions.

Keywords

Growth target leverage adjustment speed partial adjustment model India

Introduction

Investigating firms’ adjustment speed towards target leverage and the determinants of such adjustment speed is one of the most important topics in capital structure research in recent past.¹

Target Leverage is that level of leverage where firms establish trade-off between costs and benefits of debt. It is the point where the marginal costs equal the marginal benefits of raising further debt (Jensen & Meckling, 1976; Jensen, 1986; Karus & Litzenberger, 1973; Myers, 1977). Target deviation is the difference between firms’ target leverage and actual leverage (Marsh, 1982). The term adjustment in the context of capital structure dynamics refers to the tendency of firm to rebalance towards its target leverage (Taggart, 1977; Jalilvand & Harris, 1984). Finally, the term adjustment speed refers to the annual rate at which a firm adjusts the gap between its target leverage and actual leverage (Taggart, 1977; Jalilvand & Harris, 1984).

As per trade-off theory, capital structure choice depends on the costs and benefits of using debt and therefore, firms maintain a target level of leverage which establishes trade-off between the two.

The theory also argues that whenever firms deviate from their target leverage due to market frictions, such as information asymmetry, capital market condition, etc., they revert back to target with their subsequent financing choices. This argument has given birth to two important research questions. First, at what speed firms make adjustment towards the target leverage? Second, what are the determinants of such adjustment speed? Prior studies suggest that firms compare the costs against the benefits of making adjustment and adjustment takes place only when the latter exceed the former.

There are a number of studies that investigate firms’ adjustment speed towards target leverage (Basu, 2015; Lemmon, Roberts, & Zender, 2008; Loof, 2004; Miguel & Pindado, 2001; Ozkan, 2001; Shyam-Sunder & Myers, 1999). Similarly, there are some recent studies which examine the determinants of such adjustment speed either by investigating the relationship between different determinants and adjustment speed (Drobetz & Wanzenried, 2006; Mukherjee & Mahakud, 2010; Qian, Tian, & Wirjanto, 2009) or by studying the asymmetries in adjustment speed caused by a particular factor (Byoun, 2008; Dang & Garrett, 2015; Dang et al., 2012; Dang, Kim, & Shin, 2014; Devos, Rahman, & Tsang, 2017; Drobetz, Schilling, & Schroder, 2015; Fama & French, 2002; Ghose, 2017; Ghose & Kabra, 2018; Mai, Meng, & Ye, 2017; Sardo & Serrasqueiro, 2017; Zeitun, Temimi, & Mimouni, 2017).

Furthermore, some studies also investigated the moderating role of nature of target deviation (over-levered or under-levered) on the relationship between different determinants and adjustment speed. Though, many important factors are already covered, a close look into the existing literature suggests that impact of firms’ growth opportunities, an important determinant of adjustment speed (Dang, Kim, & Shin, 2012, 2014), on adjustment speed is not adequately examined. Accordingly, in an attempt to fill this gap, the present study investigates the impact of growth opportunities on firms’ leverage adjustment speed. Furthermore, it also examines the moderating role two dimensions of leverage gap, that is, nature of deviation and level of deviation (i.e., highly deviated or less deviated) on the relationship between growth opportunities and adjustment speed.

The reminder of the article is organized as follows. The second and third sections discuss conceptual framework and empirical models, respectively. The fourth section presents data description and descriptive statistics followed by the fifth section which contains the empirical results. Finally, the sixth section provides conclusion, and the seventh section outlines implications and limitations of the study.

Conceptual Framework

Growth and Adjustment Speed

Firms with higher growth opportunities are generally in need of funds to finance their growing investment opportunities and more often than not their internally generated funds are not enough to meet the requirement (Dang et al., 2012, 2014). This, as per the arguments of pecking order theory, suggests that they are more in need of external financing and as a result, need to visit financial market more frequently than the low-growth firms. This allows the high-growth firms the opportunity to suitably decide the mix of debt and equity in new issues to adjust towards the target capital structure (Dang et al., 2012, 2014; Drobetz & Wanzenried, 2006). This also reduces the costs of adjustment as the same is shared by the costs of financing necessary growing requirements (Dang et al., 2014; Faulkender, Flannery, Hankins, & Smith, 2012). Furthermore, when a part of costs of raising capital are fixed, firms with higher growth opportunities enjoy lower costs of capital due to economies of scale than low-growth firms (Aybar-Arias, Casino-Martinez, & Lopez-Gracia, 2012; Banerjee, Heshmati, & Wihlborg, 2000). Besides, higher growth opportunities also add value to firms which may increase their financial accessibility in the market (Titman & Wessels, 1988). All these arguments indicate that the speed of adjustment should be higher for high-growth firms than the low-growth firms. On the contrary, low-growth firms are generally firms with higher level of free cash flows (Dang, Garrett, & Nguyen, 2011). In order to mitigate the agency costs of free cash flow (Jensen, 1986), these firms may maintain higher level of leverage.²

As per free cash flow agency conflict argument, mangers of firms with free cash flows engage in reckless spending which do not add value to shareholders wealth. Debt in capital structure mitigates this agency problem by making managers discipline since interest payment on debt is mandatory (Jensen. 1986).

Therefore, concern for financial distress and bankruptcy risks may force these firms to adjust quickly towards their target leverage (Dang et al., 2012, 2014). Moreover, with higher level of free cash flows, they may also find it easier to adjust their leverage by suitable alterations in the payout ratio. On the basis of these arguments, one may also predict faster adjustment speed for low-growth firms than the high-growth firms. Since the arguments are contradictory, the study refrains from forming any specific hypothesis and forms a general hypothesis as firms’ growth opportunities cause asymmetry in adjustment speed.

Nature of Deviation, Growth, and Adjustment Speed

High-growth firms are generally in need of more finance and therefore, frequently visit financial market to meet their financing requirements. On the contrary, low-growth firms are generally firms with higher level of retained earnings and free cash flows (Dang et al., 2012, 2014). Based on these arguments one can expect faster adjustment speed for high-growth firms in case of under-leverage than over-leverage. This is because in case of under-leverage, they can meet their financing needs by issuing debt and at the same time can approach towards their target leverage. But, in case of over-leverage, they need to raise funds through equity to approach towards target which is costly due to adverse selection problem (Myers, 1984; Myers & Majluf, 1984).³

Adverse selection problem arises from information asymmetry between insiders and outsiders. As per the arguments of pecking order theory (Myers, 1984; Myers & Majluf, 1984), to avoid adverse selection problem, firms follow a hierarchy in their financing choices i.e. they prefer internal financing over external financing and in case of external financing they prefer debt over equity.

However, opposite results can also be predicted as high-growth firms are expected to be more concerned about agency issues of excessive debt, such as debt overhang and underinvestment problem (DeAngelo & DeAngelo, 2007; Myers, 1977).⁴

Agency theory argues that presence of debt in capital structure makes managers disciplined and as a result acts as an instrument to reduce owner-manager agency conflicts (Jensen & Meckling, 1976). However, on the downside, increasing proportion of debt leads other problems such as debt overhang i.e. a situation where firms are unable to take further debt for investment purpose due to excessive existing debt and underinvestment i.e. a situation where firms forgo future investment opportunities with positive NPVs as the benefits would mostly go to existing debt holders (Myers, 1977; DeAngelo & DeAngelo, 2007).

On the contrary, low-growth firms are expected to have faster adjustment speed when over-levered than under-levered as, with lesser investment opportunities, they need to dispose their free cash flows to avoid agency costs (Jensen, 1986) and retiring debt is less costly than retiring equity due to adverse selection costs (Myers, 1984; Myers & Majluf, 1984). Furthermore, conditional on over-leverage, low-growth firms are expected to adjust faster than high-growth firms. This is because the former, with lesser investment opportunities, may need to retire debt to dispose free cash flows, whereas the latter, with higher investment opportunities, may need to issue equity to raise funds and simultaneously make adjustment towards the target leverage. However, as mentioned earlier, growing firms’ concern for agency costs related to excessive debt and pressure to access the external market for financing growing investment opportunities may lead to opposite results. However, conditional on under-leverage, high-growth firms are expected to adjust faster than low-growth firms as the former can meet their financing requirements and adjust towards target leverage simultaneously by issuing debt. Based on the arguments, the study forms the second hypothesis as interface of nature of deviation with growth opportunities causes asymmetry in adjustment speed.

Level of Deviation, Growth, and Adjustment Speed

Apart from nature of deviation, level of deviation also expected to have moderating effect in the relationship between growth and adjustment speed. As discussed earlier, high-growth firms are frequent visitor in the external market to meet their growing financial requirements (Dang et al., 2012, 2014). Furthermore, high-growth firms are expected to bear higher market risk and therefore, are in need of keeping their financial risk limited. These arguments suggest that high-growth firms should make adjustment as and when they deviate from target leverage. On the contrary, low-growth firms are not frequent visitor in the financial market due to their lesser financing needs. Moreover, since low-growth firms are expected to operate in more certain market, they may be in a position to bear higher financial risk and therefore, avoid making adjustment when the deviation is low. On the basis of these arguments, one can expect opposite pattern in adjustment speed for low-growth firms, that is, faster adjustment speed when high-deviated than less-deviated. However, since low-growth firms are usually cash rich firms, they, in the process of disposing free cash flows, may find it convenient to adjust towards the target leverage when the level of deviation is low. Furthermore, considering the argument that high-growth firms are under compulsion to visit financial market frequently to meet their financing requirements and at the same time keep their financial risk limited, it can be expected that they adjust faster than low-growth firms irrespective of level of deviation. Based on the arguments, the study forms the third hypothesis as interface of level of deviation with growth opportunities causes asymmetry in adjustment speed.

Nature and Level of Deviation, Growth, and Adjustment Speed

Finally, allowing nature and level of target deviation together to moderate the relationship between growth and adjustment speed are expected to cause further asymmetries in adjustment speed. Conditional on over-leverage, low-growth firms, compared to high-growth firms, are expected to adjust faster irrespective of level of deviation. This is because high-growth firms need to issue equity and low-growth firms need to retire debt to deal with their respective financing deficit/surplus and to make simultaneous adjustment towards the target leverage, and as mentioned earlier, dealing in equity is costlier than debt due to adverse selection costs (Myers, 1984; Myers & Majluf, 1984). However, opposite results can be expected due to high-growth firms’ higher pressure to check financial risk, greater concern for agency costs of excessive leverage, such as debt overhang and underinvestment problem (DeAngelo & DeAngelo, 2007; Myers, 1977), and higher pressure to meet financing needs as compared to low-growth firms. On the contrary, conditional on under-leverage, high-growth firms are expected to adjust faster than low-growth firms irrespective of level of deviation. This is because; high-growth firms can simultaneously meet their financing requirements and make adjustment towards the target leverage by increasing debt. However, since low-growth firms are generally cash rich, in order dispose excess cash and to make concurrent adjustment towards target leverage they need to retire equity which is costly due to adverse selection costs (Myers, 1984; Myers & Majluf, 1984).

Again, conditional on over-leverage, high-growth firms are expected to adjust faster in case of less-deviation than high-deviation due to their frequent visit to financial market and their need to keep financial risk contained. However, low-growth firms may prefer to adjust faster when the deviation is high than low to enjoy benefits of debt. This is because they may be less concerned about financial risk due to their operation in relatively stable environment. However, alternative argument suggests that availability of free cash flow will allow them to make faster adjustment when the deviation is low. On the contrary, conditional on under-leverage, high-growth firms are expected to adjust faster when the deviation is low than high due to their frequent financial market transactions, whereas low-growth firms, with higher internally generated funds, are expected to follow opposite pattern to avoid dealing in equity and/or to take advantage of far-off adjustments.

Finally, both high- and low-growth firms are expected to adjust faster when they are over-levered than under-levered irrespective of level of deviation. This is because: first, maintaining excess leverage involves financial distress and bankruptcy risks. Second, high-growth firms are concerned about excess leverage as that might lead to agency issues, such as debt overhang and underinvestment problem. Finally, with the availability of free cash flows, low-growth firms are expected to find it easier to make adjustment when over-levered than under-levered. However, since high-growth firms are generally in need of funds and costs of raising equity are higher than costs of debt due to asymmetric information, one can argue higher adjustment speed for such firms when under-levered than over-levered. Based on all the arguments, the study forms the fourth and final hypothesis as interface of nature and level of deviation with growth opportunities causes asymmetry in adjustment speed.

Empirical Models and Estimation Techniques

Partial Adjustment Model

In order to fulfill this objective, the study uses partial adjustment model provided by Marc Nerlove (Gujarati, Porter, & Gunasekar, 2012).

The basic partial adjustment model takes the following form:

Y_{it} - Y_{it}_{- 1} = α (Y *_{it} - Y_{it}_{- 1})

(1)

where Y_it is the observed value of variable for unit i at time t, Y_it−1 is the observed value of variable for unit i at time t−1 and Y^*_it is the targeted value of variable for unit i at time t. It means that Y_it − Y_it−1 is the actual change in variable from previous year to current year and Y^*_it − Y_it−1 is the desired change in variable from previous year to current year. α is the adjustment coefficient which shows the speed of adjustment towards the targeted level of variable.

The partial adjustment model has been extensively used in prior capital structure studies in order to examine the capital structure adjustment speed (Banerjee et al., 2000; Byoun, 2008; Chang, Chou, & Huang, 2014; Dang & Garrett, 2015; Dang et al., 2012, 2014; Drobetz & Wanzenried, 2006; Faulkender et al., 2012; Flannery & Rangan, 2006; Flannery Abdeljawad & Nor, 2017; Getzmann, Lang, & Spremann, 2014; Ghose, 2017; Mahakud & Mishra, 2010). Following prior studies, the present study uses the following standard partial adjustment model to investigate firms’ adjustment speed towards their target capital structure⁵

Byoun (2008) and Dang and Garrett (2015) also used constant in Model-1 to account for changes in observed leverage for reasons other than target adjustment. Following them, present study has also cross-checked results by incorporating constant in Model-1(ii) and observed more or less similar results. Nevertheless, results are not reported here for brevity.

L_{it} - L_{it}_{- 1} = α (L *_{it} - L_{it - 1}) + ε_{it}

(2)

{ΔL}_{it} = {α DEV}_{it} + ε_{it}

(3)

As evident, in model (3) ΔL_it is the observed change in leverage ratio from last year to current year (L_it − L_it−1) and DEV_it is the target deviation, that is, the required change from last year’s observed leverage to current year’s target leverage (L^*_it − L_it−1).⁶

In Model-2 L_{_it} and L_{_it-1} are observed leverages for current and last year respectively, and L^{^*}_{_it} is the estimated target leverage for current year.

α shows the speed of adjustment towards the target leverage and ε_it denotes the error term.

In order to examine the impact of growth opportunities on adjustment speed, the study extends model (3) as follows:

{ΔL}_{it} = α_{1} {DEV}_{it} {_D}^{HG}_{it} + α_{2} {DEV}_{it} {_D}^{LG}_{it} + ε_{it}

(4)

In order to examine the moderating effect of nature and extent of target deviation on the relationship between growth opportunities and adjustment speed, the study extends model (3) as follows:

\begin{array}{l} {Δ L}_{it} = α_{1} {DEV}_{it} {_D}^{HG}_{it} {_D}^{OL}_{it} + α_{2} {DEV}_{it} {_D}^{LG}_{it} {_D}^{OL}_{it} + \\ α_{3} {DEV}_{it} {_D}^{HG}_{it} {_D}^{UL}_{it} + α_{4} {DEV}_{it} {_D}^{LG}_{it} {_D}^{UL}_{it} + ε_{it} \end{array}

(5)

\begin{array}{l} {Δ L}_{it} = α_{1} {DEV}_{it} {_D}^{HG}_{it} {_D}^{HD}_{it} + α_{2} {DEV}_{it} {_D}^{LG}_{it} {_D}^{HD}_{it} + \\ α_{3} {DEV}_{it} {_D}^{HG}_{it} {_D}^{LD}_{it} + α_{4} {DEV}_{it} {_D}^{LG}_{it} {_D}^{LD}_{it} + µ_{it} \end{array}

(6)

\begin{array}{l} {Δ L}_{it} = α_{1} {DEV}_{it} {_D}^{HG}_{it} {_D}^{OL}_{it} {_D}^{HD}_{it} + α_{2} {DEV}_{it} {_D}^{HG}_{it} {_D}^{OL}_{it} {_D}^{LD}_{it} + \\ α_{3} {DEV}_{it} {_D}^{LG}_{it} {_D}^{OL}_{it} {_D}^{HD}_{it} + α_{4} {DEV}_{it} {_D}^{LG}_{it} {_D}^{OL}_{it} {_D}^{LD}_{it} + \\ α_{5} {DEV}_{it} {_D}^{HG}_{it} {_D}^{UL}_{it} {_D}^{HD}_{it} + α_{6} {DEV}_{it} {_D}^{HG}_{it} {_D}^{UL}_{it} {_D}^{LD}_{it} + \\ α_{7} {DEV}_{it} {_D}^{LG}_{it} {_D}^{UL}_{it} {_D}^{HD}_{it} + α_{8} {DEV}_{it} {_D}^{LG}_{it} {_D}^{UL}_{it} {_D}^{LD}_{it} + ε_{it} \end{array}

(7)

where D^HG_it and D^LG_it are the dummy variables which, respectively, represent high-growth and low-growth firms.⁷

Growth opportunities (GROW) is measured as ratio of market value to book value of assets. The study uses quintile of GROW for categorizing high-growth and low-growth firms. Firms in fourth (Q4) and fifth (Q5) quintiles are grouped as high-growth firms and firms in first (Q1) and second (Q2) quintiles are grouped as low-growth firms.

D^OL_it and D^UL_it are the dummy variables for firm i at time t representing, respectively, over-levered and under-levered firms under nature of deviation.⁸

Firms are labeled as over-levered when the sign of their target deviation (TDEV) is negative and under-levered when the sign of their target deviation is positive (Byoun, 2008; Chang, et al., 2014; Dang & Garrett, 2015; Abdeljawad & Nor, 2017).

Similarly, D^HD_it and D^LD_it are the dummy variables for firm i at time t representing, respectively, highly deviated and less-deviated firms under the extent of deviation.⁹

The study uses quintile of absolute DEV for categorizing high-deviated and less-deviated firms. Firms in fourth (Q4) and fifth (Q5) quintiles are grouped as high-deviated firms and firms in first (Q1) and second (Q2) quintiles are grouped as less-deviated firms.

The study uses ordinary least square (OLS) technique in order to estimate models (4)–(7). For crosschecking, the study adopts two techniques to estimate the standard errors (SE) of the model. First, following Dang and Garrett (2015), it estimates robust SE corrected for heteroscedasticity as suggested by White (1980). Second, it estimates bootstrapped SE which accounts for generated regressors and non-normality of error terms (Faulkender et al., 2012; Gujarati et al., 2012; Pagan, 1984). The coefficients estimated from these models indicate the capital structure adjustment speed for firms falling under different groups. In the absence of differences in adjustment costs and benefits for firms under different groups, there will be no differences in adjustment speed, that is, differences between estimated coefficients (α) will be statistically insignificant.

Dynamic Panel Data Model

All the models to estimate capital structure adjustment speed specified earlier rely on firms’ unobserved target leverage which is most difficult to estimate. Some prior studies use exogenous proxies of target, such as the mean leverage ratio overtime (Jalilvand & Harris, 1984; Shyam-Sunder & Myers, 1999), the moving average leverage ratio (Jalilvand & Harris, 1984; Shyam-Sunder & Myers, 1999), the industry median leverage ratio (Hovakimian et al., 2001), etc., This study uses endogenous approach where it estimates target capital structure as a function of firm and industry specific characteristics. This approach, considered to be more appropriate and is used in majority of the existing studies, also allows the target capital structure to vary across firm and overtime (Dang & Garrett, 2015; Dang et al., 2014; Faulkender et al., 2012; Flannery & Rangan, 2006; Jiang, Jiang, Huang, Kim & Nofsingar, 2017). Specifically, the study estimates the target leverage as follows:

L^{*}_{it} = {λX}_{it}

(8)

where L^*_it is the target leverage ratio for firm i at time t, X_it is a vector of some important determinants of target capital structure for firm i at time t, namely, PROF, TANG, SIZE, GROW, NDTS, MLEV, VOL, UNIQ and DIV (Drobetz & Wanzenried, 2006; Fama & French, 2002; Frank & Goyal, 2009; Ghose & Kabra, 2017; Rajan & Zingales, 1995).¹⁰

The definitions of variables are provided in Table 1.

X_it also includes firm- and time-fixed effects to account for unobserved firm-specific and time-specific heterogeneities, respectively. Substituting L^*_it by model (8), model (2) can be stated as follows:

L_{it} - L_{it} - 1 = {α (λX}_{it} - L_{it}_{- 1}) + ε_{it}

or,

L_{it} = (1 - α) L_{it}_{- 1} + {αλX}_{it} + ε_{it}

or,

L_{it} = {γL}_{it - 1} + {δX}_{it} + ε_{it}

(9)

On replacement of X_it by the determinants of target leverage, model (9) finally takes the following form:

\begin{array}{l} L_{it} = {γL}_{it}_{- 1} + δ_{1} {PROF}_{it} + δ_{2} {TANG}_{it} + δ_{3} {SIZE}_{it} + \\ δ_{4} {GROW}_{it} + δ_{5} {NDTS}_{it} + δ_{6} {MLEV}_{it} + δ_{7} {VOL}_{it} + \\ δ_{8} {UNIQ}_{it} + δ_{9} DIV + η_{i} + ν_{t} + ε_{it} \end{array}

(10)

Many recent studies have estimated the target and the adjustment speed concurrently from model (10) (Basu, 2015; Flannery & Rangan, 2006; Ghose, 2017; Huang & Ritter, 2009; Mukherjee & Mahakud, 2010, 2012). However, as the objectives demand, the present study follows a two-step procedure where the target (L^*_it) is estimated first from model (10) and then asymmetries in adjustment speed are estimated from models (4) to (7) (Byoun, 2008; Dang & Garrett, 2015; Dang et al., 2011; Faulkender et al., 2012; Jiang et al., 2017).

With lagged leverage as one of the explanatory variables, model (10) is a dynamic panel model involving some special estimation issues (Faulkender et al., 2012). The estimated coefficient of lagged leverage is biased upward in case of Pooled OLS estimation ignoring firm fixed effects and is biased downward in case of fixed effect estimation (Huang & Ritter, 2009). Therefore, following prior studies (Drobetz & Wanzenried, 2006; Qian et al., 2009), the study uses the two-step GMM estimation technique (Arellano & Bond, 1991; Arellano & Bover, 1995; Blundell & Bond, 1998), which provides efficient parameter estimates after controlling for firm-specific heterogeneity and endogeneity in explanatory variables. The study uses three standard diagnostic tests to check validity of the model, namely, (a) the explanatory variables are jointly significant (Wald test), (b) there is no second-order autocorrelation (AR2 test), and (c) the instruments used for estimation are valid (Sargan test).

Table 1.

Definition of Variables

Variables	Abbreviations	Definitions
Leverage	LEV	Measured as ratio of total debt to book value of total assets
Profitability	PROF	Measured as ratio of EBIT to total assets
Tangibility	TANG	Measured as ratio of fixed assets to total assets
Size	SIZE	Measured as natural logarithm of total assets
Growth	GROW	Measured as ratio of market value to book value of assets
Non-debt tax shield	NDTS	Measured as ratio of depreciation to total assets
Industry median leverage	MLEV	Measured as median value of leverage ratio for each industry on yearly basis
Earning volatility	VOL	Measured as the moving standard deviation of past 3 years profitability (EBIT to total assets)
Uniqueness	UNIQ	Measured as ratio of selling expenses to net sales
Payout dummy	DIV	Dummy variable which takes the value of 1 when firms pay dividend and 0 otherwise

Source: Authors’ own.

Population and Sample of the Study

The population for the study comprises all listed Indian firms for which data are available on corporate database called Capitaline Plus over a period of 2004–2005 to 2015–2016. Then, following the existing literature, the study applies some standard restrictions on the stated firms to arrive at the final sample (Aybar-Arias et al., 2012; Byoun, 2008; Dang et al., 2014; Haron, Ibrahim, Nor, & Ibrahim, 2013; Ozkan, 2001; Rajan & Zingles, 1995; Wojewodzki et al., 2018). First, the study excludes financial and utility firms as their capital structures are different than the non-financial firms and non-utility firms. It also excludes firms for which industry information is not available in the database. Second, as per the requirement, the study considers only those firms for which data are available for at least four consecutive years.¹¹

The study uses Earning Volatility as one of the explanatory variables in estimating the target leverage. Since, Earning Volatility is measured as three years moving standard deviation of firms’ EBIT to Total Assets Ratio, it needs at least three years consecutive data. Further, the study uses Generalised Method of Moments (GMM) estimation technique to estimate the target leverage which takes first difference of the model to eliminate the panel specific effects. This requires data for another one year. Therefore, the study requires data for at least four consecutive years.

Aybar-Arias et al. (2012), Haron et al. (2013) and Wojewodzki et al. (2018) in their studies, respectively, consider firms with at least 7, 3 and 3 years consecutive data as per their requirements. Finally, the study excludes firms with non-positive total assets and net sales. After applying the above-mentioned conditions, the final dataset is arrived at 28,532 firm-year observations comprising 2,718 listed Indian firms with a minimum of 4 years and maximum of 12 years data for each firm for the purpose of analysis.

Results and Discussion

Target Leverage

As per the requirement, the study first estimates target leverage with the help of model (10). Table 2 reports the target estimation results and in order to save space, these results are summarized briefly. The results reveal that impacts of different factors on target leverage are mostly consistent with the theoretical expectations and findings of prior empirical studies. Tangibility, growth, and industry median leverage have a positive impact, whereas profitability and uniqueness have a negative impact on firms’ target leverage. These findings are consistent with the arguments of trade-off, pecking order and agency theory of capital structure. However, impacts of size, non-debt tax shield, earnings volatility and payout dummy are found to be statistically insignificant and, therefore, are less important in Indian context. With regard to validity of model, the results reveal that all explanatory variables are jointly significant (Wald tests), there is no second-order autocorrelation in error terms (AR2 test) and the instruments used in the model are valid (Sargan test).

Table 2.

Target Estimation Results

Variables	Coefficient	Standard Errors	z-Statistics	p-Value
LEV (−1)	0.72	0.04	19.31	0.00
PROF	−0.49	0.04	−14.02	0.00
TANG	0.20	0.03	7.63	0.00
SIZE	0.01	0.01	0.57	0.57
GROW	0.02	0.00	4.76	0.00
NDTS	−0.33	0.20	−1.64	0.10
MLEV	0.27	0.08	3.46	0.00
VOL	0.001	0.05	0.08	0.93
UNIQ	−0.22	0.08	−2.68	0.01
DIV	0.001	0.00	1.06	0.29
AR (1)	−8.40*
AR (2)	0.95
Wald test 1	856.46*
Wald test 2	370.87*
Wald test 3	51.18*
Sargan value	14.87

Source: Capitaline Plus’ database, statistical tool: Stata 14

Notes: (a) Figures showing against AR (1) and AR (2) are z-statistics, respectively, for first-order and second-order autocorrelation coefficients in residuals, (b) figures showing against Wald tests and Sargan test are chi-square statistics and (c) *denote results are significant at 1% level.

Growth and Adjustment Speed

The impact of firms’ growth opportunities on adjustment speed is examined by model (4) and results are reported in Table 3. The results reveal that high-growth firms have an adjustment speed of 36 percent, whereas low-growth firms have an adjustment speed of only 24 percent towards their target leverage. That means the former takes around 2.8 years (1/α) to accomplish their optimum capital structure, whereas the latter takes around 4.2 years for the same. The probability values of Wald test 2 confirm that the adjustment speeds in two groups of firms are statistically different. This means high-growth firms adjust their capital structure faster than low-growth firms by around 12 percent. The faster adjustment speed for high-growth firms possibly indicates their ability to close target deviation by making appropriate changes in debt–equity mix in new issues (Dang et al., 2012, 2014; Drobetz & Wanzenried, 2006). Other possible reasons could be lesser costs of adjustment due to economies of scale in the presence of fixed costs (Aybar-Arias et al., 2012; Banerjee et al., 2000), sharing of adjustment costs by otherwise necessary costs of financing (Dang et al., 2014; Faulkender et al., 2012), better accessibility in the financial market due to better firm value (Titman & Wessels, 1988) and higher concern for financial risk. The faster adjustment speed for high-growth firms confirms the empirical findings of Drobetz and Wanzenried (2006) and Aybar-Arias et al. (2012). To conclude, Wald test (2) confirms the first hypothesis of the study, that is, firms’ growth opportunities cause asymmetry in adjustment speed.

Table 3.

Growth and Adjustment Speed

Variables	Robust SE		Bootstrap SE
	Coefficients	t-Statistics	Coefficients	z-Statistics
DEV_D^HG	0.36	22.88*	0.36	19.19*
DEV_D^LG	0.24	20.16*	0.24	19.21*
Wald test 1 (model)	0.00		0.00
Wald test 2 (HG = LG)	0.00		0.00
Adjusted R²	0.10		0.10

Source: ‘Capitaline Plus’ database, statistical tool: Stata 14.

Notes: (a) Figures showing against Wald tests are probability values of F-statistics and chi-square (χ² statistics in case of robust SE and bootstrapped SE, respectively, and (b) * denote results are significant at 1% level.

Nature of Deviation, Growth, and Adjustment Speed

Table 4 reports the results of model (5), which examines the moderating effect of nature of deviation on the relationship between growth opportunities and adjustment speed. The results reveal that both high- and low-growth firms adjust their capital structure faster towards target when they are over-levered than under-levered (i.e., 41% vs. 32% in case of high-growth firms and 72% vs. 15% in case of low-growth firms). Though the results are similar, reasons behind the results are expected to be different. In case of high-growth firms, the results support the agency costs argument of excessive debt, that is, excess debt leads to debt overhang and underinvestment problem for high-growth firms (Myers, 1977). However, in case of low-growth firms, the results support the free cash flow argument, that is, retiring debt is less costly than retiring equity to dispose of free cash flows. Besides, these results can also be interpreted as the concern for financial distress and bankruptcy risks for both the groups of firms. Furthermore, conditional on over-leverage, low-growth firms adjust faster than high-growth firms (i.e., 72% vs. 41%), which possibly indicates availability of free cash flows to reduce debt for the former. On the contrary, conditional on under-leverage, high-growth firms adjust faster than low-growth firms (i.e., 32% vs. 15%) which is in line with the prediction that the former can achieve twin objectives by issuing debt in case of under-leverage, that is, meeting their financing requirements and approaching towards the target leverage. Overall, Wald test (6) rejects the null hypothesis that all estimated coefficients are equal and confirms that interface of nature of deviation with growth opportunities causes asymmetry in adjustment speed. These results establish that nature of deviation moderates the relationship between growth opportunities and adjustment speed.

Table 4.

Nature of Deviation, Growth and Adjustment Speed

Variables	Robust SE		Bootstrap SE
	Coefficients	t-Statistics	Coefficients	z-Statistics
DEV_D^HG_D^OL	0.41	13.58*	0.41	13.75*
DEV_D^LG_D^OL	0.72	17.45*	0.72	20.07*
DEV_D^HG_D^UL	0.32	19.57*	0.32	20.77*
DEV_D^LG_D^UL	0.15	15.15*	0.15	17.36*
Wald test 1 (model)	0.00		0.00
Wald test 2 (HG_OL = HG_UL)	0.00		0.01
Wald test 3 (LG_OL = LG_UL)	0.00		0.00
Wald test 4 (HG_OL = LG_OL)	0.00		0.00
Wald test 5 (HG_UL = LG_UL)	0.00		0.00
Wald test 6 (all are equal)	0.00		0.00
Adjusted R²	0.11		0.11

Source: ‘Capitaline Plus’ database, statistical tool: Stata 14.

Notes: (a) Figures showing against Wald tests are probability values of F-statistics and chi-square (χ²) statistics in case of robust SE and bootstrapped SE, respectively, and (b) * denote results are significant at 1% level.

Level of Deviation, Growth, and Adjustment Speed

The results of model (6) are presented in Table 5, which examines the moderating effect of level of deviation on the relationship between growth opportunities and adjustment speed. As per the results, conditional on both high and low deviation, high-growth firms adjust faster than low-growth firms. To be specific, the former, in comparison to the latter, has 11 percent higher adjustment speed in case of high-deviation and 16 percent higher adjustment speed in case of low-deviation. These findings are consistent with the argument that high-growth firms are under compulsion to visit financial market frequently to meet their financing needs and in the process, approach towards their target leverage with lesser adjustment costs. Moreover, these results, as mentioned earlier, also indicate the possibility of lesser adjustment costs due to economies of scale, better accessibility in external market, sharing of adjustment costs by otherwise costs of financing, etc., and greater concern for financial risk. Wald test (6) shows that all estimated coefficients are not equal and confirms the hypothesis that interface of level of deviation with growth opportunities causes asymmetry in adjustment speed. Overall, the results establish that extent of deviation has moderating effect on the relationship between growth opportunities and target adjustment speed.

Table 5.

Level of Deviation, Growth, and Adjustment Speed

Variables	Robust SE		Bootstrap SE
	Coefficients	t-Statistics	Coefficients	z-Statistics
DEV_D^HG_D^HD	0.35	21.20*	0.35	22.13*
DEV_D^LG_D^HD	0.24	18.42*	0.24	19.61*
DEV_D^HG_D^LD	0.46	6.88*	0.46	5.97*
DEV_D^LG_D^LD	0.30	5.88*	0.30	5.90*
Wald test 1 (model)	0.00		0.00
Wald test 2 (HG_HD = HG_LD)	0.12		0.18
Wald test 3 (LG_HD = LG_LD)	0.25		0.27
Wald test 4 (HG_HD = LG_HD)	0.00		0.00
Wald test 5 (HG_LD = LG_LD)	0.05		0.05
Wald test 6 (all are equal)	0.00		0.00
Adjusted R²	0.10		0.10

Source: ‘Capitaline Plus’ database, statistical tool: Stata 14.

Nature and Level of Deviation, Growth, and Adjustment Speed

Finally, the results of model (7) which considers moderating effect of both nature and level of deviation together while examining relationship between growth and adjustment speed are reported in Table 6. The results reveal that, conditional on over-leverage and high-deviation, low-growth firms make faster adjustment than high-growth firms (i.e., 71% vs. 40%), whereas conditional on under-leverage and high-deviation, high-growth firms make faster adjustment than low-growth firms (i.e., 33% vs. 16%). These results suggest that with higher level of free cash flows, low-growth firms, compared to high-growth firms, can comfortably adjust towards target leverage by retiring debt in case of over-leverage but find it difficult to make adjustment in case of under-leverage as the same requires retirement of equity which suffers from higher information asymmetry costs than debt. The results further reveal that conditional on under-leverage and low deviation, only high-growth firms make adjustment towards the target (i.e., 23%), whereas low-growth firms avoid making any adjustment. This result suggests that with free cash flows, low-growth firms do not attempt to make adjustment as the same requires retirement of equity which suffers from information asymmetry costs. On the contrary, high-growth firms find it easier to make adjustment as by raising debt they can simultaneously adjust towards the target and meet their financing needs.

Table 6.

Nature and Extent of Deviation, Growth, and Adjustment Speed:

Variables	Robust SE		Bootstrap SE
	Coefficients	t-Statistics	Coefficients	z-Statistics
DEV_D^HG_D^OL_D^HD	0.40	12.62*	0.40	11.96*
DEV_D^HG_D^OL_D^LD	0.90	5.84*	0.90	5.88*
DEV_D^LG_D^OL_D^HD	0.71	15.29*	0.71	17.14*
DEV_D^LG_D^OL_D^LD	0.89	6.53*	0.89	7.09*
DEV_D^HG_D^UL_D^HD	0.33	18.48*	0.33	17.21*
DEV_D^HG_D^UL_D^LD	0.23	3.53*	0.23	4.01*
DEV_D^LG_D^UL_D^HD	0.16	14.67*	0.16	15.01*
DEV_D^LG_D^UL_D^LD	0.05	0.96	0.05	0.93
Wald test 1 (model)	0.00		0.00
Wald test 2 (HG_OL_HD = LG_OL_HD)	0.00		0.00
Wald test 3 (HG_OL_LD = LG_OL_LD)	0.96		0.96
Wald test 4 (HG_UL_HD = LG_UL_HD)	0.00		0.00
Wald test 5 (HG_UL_LD = LG_UL_LD)	0.02		0.00
Wald test 6 (HG_OL_HD = HG_OL_LD)	0.00		0.00
Wald test 7 (LG_OL_HD = LG_OL_LD)	0.22		0.17
Wald test 8 (HG_UL_HD = HG_UL_LD)	0.15		0.08
Wald test 9 (LG_UL_HD = LG_UL_LD)	0.02		0.03
Wald test 10 (HG_OL_HD = HG_UL_HD)	0.04		0.04
Wald test 11 (LG_OL_HD = LG_UL_HD)	0.00		0.00
Wald test 12 (HG_OL_LD = HG_UL_LD)	0.00		0.00
Wald test 13 (LG_OL_LD = LG_UL_LD)	0.00		0.00
Wald test 14 (all are equal)	0.00		0.00
Adjusted R²	0.12		0.12

Source: ‘Capitaline Plus’ database, statistical tool: Stata 14.

Notes: (a) Figures showing against Wald tests are probability values of F-statistics and chi-square (χ²) statistics in case of robust SE and bootstrapped, SE respectively, and (b) * denote results are significant at 1% level.

The results further reveal that conditional on over-leverage, high-growth firms adjust faster in case of low-deviation than high-deviation, that is, 90 percent vs. 40 percent which is consistent with the argument that high-growth firms frequently visit financial market to meet their financing requirements, and in the process, they make adjustment towards target immediately after deviation. Moreover, this result also gives an indication that high-growth firms are highly concerned about financial risk. On the contrary, conditional on under-leverage, low-growth firms make adjustment only when they are highly deviated and avoid making adjustment in case of less-deviation (i.e., 16% vs. 0%). Since, low-growth firms, with excess free cash flow, need to retire equity to dispose of excess cash and to simultaneously adjust towards the target, they possibly wait until they are substantially deviated to avoid adverse selection costs. Moreover, far-off adjustment also allows them to avail costs benefits.

Finally, conditional on high deviation, both high- and low-growth firms adjust their capital structure faster in case of over-leverage than under-leverage (i.e., 40% vs. 33% and 71% vs. 16% for high- and low-growth firms, respectively). Similarly, conditional on low-deviation, the adjustment speed is observed to be faster in case of over-leverage than under-leverage for both the groups of firms (i.e., 90% vs. 23% in case of high-growth firms and 89% vs. 0% in case of low-growth firms). Although, these results in general indicate the concern for financial distress and bankruptcy risks, other possibilities cannot be over-looked. These results can also be taken as an indication of high-growth firms’ concern for agency costs associated with excessive debt, that is, debt overhang and underinvestment problem, and low-growth firms’ ease of making adjustment with available free cash flows. Finally, Wald test (14) confirms the hypothesis that interface of nature and level of deviation with growth opportunities causes asymmetry in adjustment speed, which means both nature and level jointly moderates the relationship between growth opportunities and adjustment speed.

Conclusions and Implications of the Study

This study investigates the relationship between firms’ growth opportunities and their adjustment speed towards target capital structure. Besides, it also examines the moderating role of nature and level of target deviation while examining relationship between growth and adjustment speed. It uses data of 2,718 listed Indian firms (excluding financial and utility firms) over the period of 2004–2005 to 2015–2016 with a minimum of 4 years data for each firm. Using standard partial adjustment model, the study observes that, in general, firms with high-growth opportunities (36%) adjust their capital structure faster towards target than firms with low-growth opportunities (24%) which possibly indicate the ability of former to close target deviation by making appropriate changes in debt–equity mix in new issues. This result can also be taken as an indication of lesser adjustment costs, sharing of adjustment costs by otherwise necessary costs of financing and better accessibility in the financial market.

Furthermore, the results confirm the moderating effect of nature and level of target deviation in the relationship between growth and adjustment speed. Both high- and low-growth firms adjust faster when over-levered than under-levered, irrespective of extent of deviation. Concern for agency costs associated with excessive debt, such as debt overhang and underinvestment problems, are the possible reasons behind the results for high-growth firms. However, in case of low-growth firms, adverse selection cost possibly plays the role, as such firms generally have higher level of internally generated funds and free cash flows. Furthermore, low-growth firms adjust more than high-growth firms in case of over-leverage and high-growth firms adjust more than low-growth firms in case of under-leverage, which again indicate the importance of information asymmetry in capital structure adjustment. Overall, the study concludes that firms’ growth opportunities cause asymmetries in target adjustment speed by altering the costs and benefits of adjustment, and nature and level of target deviation moderates the relationship between growth and adjustment speed. As per authors’ knowledge, this is the sole study which investigates the impact of growth opportunities on adjustment speed in the context of an emerging economy in general and India in particular, and therefore, is an important contribution to the existing literature. Furthermore, the study examines and confirms the moderating effects of nature and level of deviation on the relationship between growth and adjustment speed which are expected to have substantial practical implications for financial managers in their capital structure decisions.

Footnotes

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

References

Abdeljawad

Nor

F. M.

(2017). The capital structure dynamics of Malaysian firms: Timing behavior vs adjustment toward the target. International Journal of Managerial Finance , 13(3), 226–245.

Arellano

Bond

(1991). Some tests of specification for panel data: Monte Carlo evidence and an application to employment equations. Review of Economic Studies , 38(2), 277–297.

Arellano

Bover

(1995). Another look at instrumental variable estimation of error-components models. Journal of Econometrics , 68(1), 29–51.

Aybar-Arias

Casino-Martinez

Lopez-Gracia

(2012). On the adjustment speed of SMEs to their optimal capital structure. Small Business Economics , 39(4), 977–996.

Banerjee

Heshmati

Wihlborg

(2000). The dynamics of capital structure (Working Paper No. 333. SSE/EFI working paper series in Economics and Finance). New York, NY: New York University.

Basu

(2015). Market imperfections and optimal capital structure: Evidence from Indian panel data. Global Business Review , 16(1), 61–83.

Blundell

Bond

(1998). Initial conditions and moment restrictions in dynamic panel data models. Journal of Econometrics , 87, 115–143.

Byoun

(2008). How and when do firms adjust their capital structures towards targets? The Journal of Finance , LXIII(6), 3069–3096.

Chang

Chou

R. K.

Huang

(2014). Corporate governance and the dynamics of capital structure: New evidence. Journal of Banking and Finance , 48, 374–385.

10.

Dang

V. A.

Garrett

(2015). On corporate capital structure adjustments. Finance Research Letters , 14, 56–63.

11.

Dang

V. A.

Garrett

Nguyen

(2011). Asymmetric partial adjustment towards target leverage: International evidence . Unpublished Manuscript, University of Manchester.

12.

Dang

V. A.

Kim

Shin

(2012). Asymmetric capital structure adjustments: New evidence from dynamic panel threshold models. Journal of Empirical Finance , 19(4), 465–482.

13.

Dang

V. A.

Kim

Shin

(2014). Asymmetric adjustment towards optimal capital structure: Evidence from a crisis. International Review of Financial Analysis , 33, 226–242.

14.

DeAngelo

(2007). Capital structure, payout policy and financial flexibility (Working Paper No. FBE 02–06). University of Southern California, Los Angeles.

15.

Devos

Rahman

Tsang

(2017). Debt covenants and the speed of capital structure adjustment. Journal of Corporate Finance , 45, 1–18.

16.

Drobetz

Wanzenried

(2006). What determines the speed of adjustment to the target capital structure. Applied Financial Economics , 16, 941–958.

17.

Drobetz

Schilling

D. C.

Schroder

(2015). Heterogeneity in the speed of capital structure adjustment across countries and over the business cycle. European Financial Management , 21(5), 936–973.

18.

Fama

E. F.

French

K. R.

(2002). Testing trade-off and pecking order predictions about dividends and debt. The Review of Financial Studies , 15(2), 1–33.

19.

Faulkender

Flannery

M. J.

Hankins

K. W.

Smith

J. M.

(2012). Cash flows and leverage adjustments. Journal of Financial Economics , 103, 632–646.

20.

Flannery

M. J.

Rangan

K. P.

(2006). Partial adjustment towards target capital structures. Journal of Financial Economics , 79, 469–506.

21.

Frank

M. Z.

Goyal

V. K.

(2009). Capital structure decisions: Which factors are reliably important. Financial Management , 38(1), 1–37.

22.

Getzmann

Lang

Spremann

(2014). Target capital structure and adjustment speed in Asia. Asia-Pacific Journal of Financial Studies , 43, 1–30.

23.

Ghose

(2017). Impact of business group affiliation on capital structure adjustment speed: Evidence from Indian manufacturing sector. Emerging Economy Studies , 3(1), 54–67.

24.

Ghose

Kabra

K. C.

(2017). Capital structure, group affiliation and financial constraints: Indian evidence. Prajnan: Journal of Social and Management Sciences , 46(1), 9–36.

25.

Ghose

Kabra

K. C.

(2018). Dynamic capital structure adjustment and business group affiliations: Indian evidence. Business Perspectives and Research , 6(1), 27–41.

26.

Gujarati

D. N.

Porter

D. C.

Gunasekar

(2012). Basic econometrics (5th ed.). New Delhi, India: Tata McGraw-Hill.

27.

Haron

Ibrahim

Nor

F. M.

Ibrahim

(2013). Factors affecting speed of adjustment to target leverage: Malaysia evidence. Global Business Review , 14(2), 243–262.

28.

Hovakimian

Opler

Titman

(2001). The debt equity choice. Journal of Financial and Quantitative Analysis , 36(1), 1–24.

29.

Huang

Ritter

J. R.

(2009). Testing theories of capital structure and estimating the speed of adjustment. Journal of Financial and Quantitative Analysis , 44(2), 237–271.

30.

Jalilvand

Harris

R. S.

(1984). Corporate behavior in adjusting to capital structure and dividend targets: An econometric study. The Journal of Finance , XXXIX(1), 127–145.

31.

Jensen

M. C.

(1986). Agency costs of free cash flow, corporate finance and takeovers. The American Economic Review , 76(2), 323–329.

32.

Jiang

Huang

Kim

K. A.

Nofsingar

J. R.

(2017). Bank competition and leverage adjustments. Financial Management , 46(4), 995–1022.

33.

Lemmon

M. L.

Roberts

M. R.

Zender

J. F.

(2008). Back to the beginning: Persistence and the cross-section of corporate capital structure. The Journal of Finance , LXIII(4), 1575–1608.

34.

Loof

(2004). Dynamic optimal capital structure and technical change. Structural Change and Economic Dynamics , 15(4), 449–468.

35.

Mahakud

Mishra

A. K.

(2010). Determination of capital structure in India: A partial adjustment approach. International Journal Accounting and Finance , 2(2), 220–235.

36.

Mai

Meng

(2017). Regional variation in the capital structure adjustment speed of listed firms: Evidence from China. Economic Modelling , 64, 288–294.

37.

Miguel

A. D.

Pindado

(2001). Determinants of capital structure: New evidence from Spanish panel data. Journal of Corporate Finance , 7(1), 77–99.

38.

Mukherjee

Mahakud

(2010). Dynamic adjustment towards target capital structure: Evidence from Indian companies. Journal of Advances in Management Research , 7(2), 250–266.

39.

Mukherjee

Mahakud

(2012). Historical market-to-book ratio and corporate capital structure. Global Business Review , 13(2), 339–350.

40.

Myers

S. C.

(1977). Determinants of corporate borrowings. Journal of Financial Economics , 5, 147–175.

41.

Myers

S. C.

(1984). The capital structure puzzle. The Journal of Finance , XXXIX(3), 575–592.

42.

Myers

S. C.

Majluf

(1984). Corporate financing and investment decisions when firms have information that investors do not have. Journal of Financial Economics , 13(2), 187–221.

43.

Ozkan

(2001). Determinants of capital structure and adjustment to long run target: Evidence from UK company panel data. Journal of Business Finance and Accounting , 28(1&2), 175–198.

44.

Pagan

(1984). Econometric issues in the analysis of regressions with generated regressors. International Economic Review, 25 , 154–193.

45.

Qian

Tian

Wirjanto

T. S.

(2009). Do Chinese publicly listed companies adjust their capital structure toward a target level. China Economic Review , 20(4), 662–676.

46.

Rajan

R. G.

Zingales

(1995). What do we know about capital structure? Some evidence from international data. The Journal of Finance , L(5), 1421–1460.

47.

Sardo

Serrasqueiro

(2017). Does dynamic trade-off theory explain Portuguese SME capital structure decisions. Journal of Small Business and Enterprise Development , 24(4), 485–502.

48.

Shyam-Sunder

Myers

S. C.

(1999). Testing of static trade off against pecking order models of capital structure. Journal of Financial Economics , 51(2), 219–244.

49.

Taggart

R. A.

(1977). A model of corporate financing decisions. The Journal of Finance , 32(5), 1467–1484.

50.

Titman

Wessels

(1988). The determinants of capital structure choice. The Journal of Finance , 43, 1–19.

51.

White

(1980). A heteroskedasticity-consistent covariance matrix estimator and a direct test for heteroskedasticity. Econometrica , 48(4), 817–838.

52.

Wojewodzki

Poon

W. P. H.

Shen

(2018). The role of credit rating on capital structure and its speed of adjustment: An international study. The European Journal of Finance , 24(9), 735–760.

53.

Zeitun

Temimi

Mimouni

(2017). Do financial crisis alter the dynamics of corporate capital structure? Evidence from GCC countries. The Quarterly Review of Economics and Finance , 63, 21–33.