North American Journal of Economics and Finance 31 (2015) 94–107

Contents lists available at ScienceDirect

North American Journal of

Economics and Finance

Firm leverage decisions: Does industry matter?

Silvia Z. Islam a,∗, Sarod Khandaker b
a
School of Economics, Finance and Marketing, Royal Melbourne Institute of Technology University (RMIT),
Australia
b
Swin burne Business School, Swin burne University of Technology, Australia

a r t i c l e i n f o a b s t r a c t

Article history: We use a standard capital structure mode l to investigate the

Received 22 March 2014 firm leverage decisions of 1620 companies listed in the Australian
Received in revised form 24 October 2014 Securities Exchange (ASX) across a span of 13 years (2000–2012),
Accepted 27 October 2014
dividing the sample into mining and other industries (non-mining).
Available online 8 November 2014
We also test for significant differences in leverage decisions
between these two groups by applying a dummy variable approach.
Keywords:
Our findings show that fundamental differences exist between min-
Leverage
ing and non-mining companies when making leverage decisions.
Mining firms
Profitability We find evidence that mining firms are more sensitive to profit-
Asset Tangibility ability and asset tangibility where neither profitability nor asset
Dummy variable tangibility has significant association for non-mining firms. Over-
all results suggest that industry-type does matter for firms making
leverage decisions.
© 2014 Elsevier Inc. All rights reserved.

1. Introduction

In the midst of the current credit crunch in financial markets following the global financial crisis,
there has been a dramatic increase in the issuance of shares as many companies around the world
are over levered (Soros, 2008). Highly-leveraged companies are widely expected to under perform
because access to credit has become more difficult (Ceung, Jing, Lu, Rau, & Stouraitis, 2009). Thus,
making a decision on the level of leverage is crucial and the recent global financial crisis provides a

∗ Corresponding author. Tel.: +61 3 9925 1518; fax: +61 3 9925 5624.
E-mail addresses: [email protected] (S.Z. Islam), [email protected] (S. Khandaker).

http://dx.doi.org/10.1016/j.najef.2014.10.005
1062-9408/© 2014 Elsevier Inc. All rights reserved.
 S.Z. Islam, S. Khandaker / North American Journal of Economics and Finance 31 (2015) 94–107 95

graphic example of how highly-leveraged firms are forced to balance their capital structure through
issue of shares (Brusov et al., 2012).
Since the seminal work by Modigliani and Miller (1958), extensive research has been carried out to
examine the leverage decisions of firms. Considerably, less research has been undertaken to identify
the leverage decisions of mining firms in Australia. Da Silva Rosa (1995), Da Silva Rosa Velayuthen and
Walter (2003), How (2000) and Lee, Taylor, and Walter (1996) represent the few studies that investi-
gate Australian mining initial public offerings (IPOs) and the share price performance of mining firms.
Also relevant are major studies published on leverage decisions for Australian firms including studies
by Allen (1991, 1993), Brailsford, Oliver, and Pua (2002); Cassar and Holmes (2003), Chiarella, Pham,
Sim, and Tan (1992), Gatward and Sharpe (1996), and Twite (2001). These studies do not consider any
distinction between mining and firms from other industries (non-mining firms hereafter) in relation
to the capital structure and leverage decision.
Australia is recognized as one of the important exporters of mining resources around the globe
(Fiscore, 2007). During the period 2011–12, the largest contributing industry in industry value added
for Total Selected Industries (TSI) in Australia was mining, accounting for 13.1 per cent ($133.0b) of
the estimate for TSI. In addition, the mining sector also was the largest contributor to operating profits
before tax over the same period, accounting for 23.0 per cent ($83.7b).i The importance of the mining
industry in Australia is further highlighted by the fact that it accounts for almost one-tenth of the
national output (Australian Bureau of Statistics, 1996). It also represents 20 per cent of ASX market
capitalization and one third of listed companies.iii Thus, if capital structure is important to firm value,
then understanding leverage decisions made by mining companies is also important.
In this study, we add to the body of knowledge within empirical studies of capital structure by
examining the dynamic relationships between leverage and the determinants of capital structure
using 715 mining firms and 905 non-mining firms for the period from 2000 to 2012. To examine this
relationship we use the standard capital structure model following previous studies (see for example,
Baker & Wurgler, 2002; Hovakimian, 2006). In the model, leverage is used as a dependent variable
and each of the control variables are used as a common proxy for determinants of leverage or capital
structure used in existing literature. We also examine the fundamental differences between mining
and non-mining firms by applying a dummy variable approach where the Dummy is set to 1 (D = 1) for
non-mining firms and 0 for mining firms (D = 0). The logic behind using dummy variables is to enable
us to include nominal level variables with more than two categories in our multiple regressions. Thus,
applying dummy variables in the original regression model allows us to test for statistically significant
differences in mining verses non-mining firm coefficient estimates.
Our results indicate that the nature of the firm determines the importance of the leverage decision.
Like previous studies, the results shows that profitability, asset tangibility and firm size all play a
vital role when determining the leverage choice of the firm (see for example, Baker & Wurgler, 2002;
Hovakimian, 2006; Myers & Majluf, 1984; Rajan & Zingales, 1995). Profitability seems to be more
important for mining firms where non-mining firms seem more sensitive towards asset tangibility.
Our findings also provide evidence that market timing (proxied by market-to-book ratio) appears to be
important in some cases, but not important for mining firms in particular. Finally, the result identifies
fundamental variation in the results between mining and non-mining firms.
This paper is organized as follows: the next section briefly reviews the existing studies on capital
structure and mining firms in Australia. The third and fourth sections provide the discussion on data
collection, summary statistics, variable definitions and research methodology. The empirical results
are presented in Section 5 and Section 6 summarises the key findings and conclude.

2. Studies on capital structure/leverage decision

There is a substantial literature at both theoretical and empirical level dealing with the important
decision of capital structure. Prior empirical studies on capital structure examine the static trade-off

i
Data is collected from the Australian Bureau of Statistics Mining Statistics Newsletter, 2009–2010, under Australian Industry,
2008–09 (Cat. no. 8155.0).
iii
According to “ASX and Australian Mining” Australian Securities Exchange archived from the original on April 2006.
96 S.Z. Islam, S. Khandaker / North American Journal of Economics and Finance 31 (2015) 94–107

and pecking order theories (Fama & French, 2002; Frank & Goyal, 2003; Shyam-Sunder & Myers, 1999),
agency costs (Jensen, 1986), market timing (Baker & Wurgler, 2002; Bie & Haan, 2007; Hovakimian,
2006; Jenter, 2005) and stock returns (Welch, 2004). Myers and Majluf (1984) and Harris and Raviv
(1991) also surveyed the literature on the capital structure decision or, more specifically, the leverage
decision of US firms (Mackay & Phillips, 2005).
Inevitably analysis of optimal capital structure theory begins with Modigliani and Miller (1958)
who assumed perfect and frictionless capital markets, in which financial innovation would quickly
extinguish any deviation from their predicted equilibrium. Since then the research has progressed
from academic models to practical reality and the literature shows that capital structure choice does
matter and is relevant to firm value (Fischer, Heinkel, & Zechner, 1989; Jensen & Meckling, 1976;
Myers & Majluf, 1984). It has been observed that leverage is related to firm size, growth opportunities,
liquidation, and value of assets and this is consistent with the predictions of trade-off theories (Chang
& Dasgupta, 2009). The studies that report the importance of target leverage as a determinant of
debt/equity choice are also supportive of the trade-off hypothesis (Jalilvand & Harris, 1984; Marsh,
1982). On the other hand, the pecking order model generally outperforms the trade-off model while
explaining the time series variation in leverage (Shyam-Sunder & Myers, 1999). Shyam-Sunder and
Myers (1999) introduce a new test of the pecking order Model that shows that a simple pecking order
model can outperform the static trade-off model while explaining the time-series variation in target
leverage. However, Chirinko and Singha (2000) point out a drawback of Shyam-Sunder and Myers’
(1999) model and argue that their model evaluates neither the pecking order nor the static trade-off
model and suggests that alternative tests are needed to identify the determinants of capital structure
that can better distinguish among competing hypotheses.
In their study, Baker and Wurgler (2002) suggest that choice of financing is hard to explain within
the traditional theories. It is argued that equity market timing is an important aspect of corporate
financial decision-making (Baker & Wurgler, 2002). A further issue that has been addressed in several
recent studies is whether or not external shocks have a persistent impact on the capital structure
of a firm. The persistent impact on the capital structure and the determinants of capital structure
have been the focus of investigation in a number of countries, including the USA (Alti, 2006; Baker
& Wurgler, 2002; Hovakimian, 2006; Leary & Roberts, 2005; Welch, 2004); UK (Bevan & Danbolt,
2002, Bevan & Danbolt, 2004; Marsh, 1982); Australia (Akhtar, 2005; Allen, 1991, 1993; Brailsford
et al., 2002; Cassar & Holmes, 2003; Chiarella et al., 1992; Gatward & Sharpe, 1996; Twite, 2001) and
the Netherlands (Bie & Haan, 2007). In their recent study, Li and Zhao (2014) investigate the impact
of leveraged exchange traded fund (ETF) trading on the market quality of component stocks using
comprehensive liquidity measures (determinants that are used to capture leverage decisions of the
firm) and event-study approach. Their result suggests the effect of leveraged ETF trading on component
stocks is insignificant unlike previous studies. In another study, Schoder (2013) examined the effect of
relative supply-demand conditions on the capital markets of US firm-level investment which shows
that the investment is mostly driven by adverse demand rather than supply conditions.

2.1. Studies on mining firms

In an early study by Ball and Brown (1980), it is reported that mining equities appear to be consid-
erably riskier than industrial equities and that risk does not appear to be diversifiable in their study of
the period 1958–1979. There is also evidence that optimal capital structure is very sensitive to changes
in the level of interest rate. For example, Goldstein, Ju, and Leal (2001) found that the interest rate is
a key input in the dynamic models on capital structure.
Lee et al. (1996) examine 266 industrial IPO firms and find evidence that these IPOs perform poorly
in the market. In addition, How (2000) examines the initial and long-run performance of 130 Aus-
tralian mining IPOs and finds that the average under-pricing of mining IPOs is significantly higher
for industrial firms. Study by Naidu (1986) found significant differences in the aggregate level of
financial leverage between industry groups in Australia where as no such differences can be found
in South Africa. This result indicates that, country factor influence on leverage between these two
countries occurred due to adoption of different leverage strategies. It has also been tested that when
issuing companies are separated into sub-samples based on industry classification, abnormal returns
 S.Z. Islam, S. Khandaker / North American Journal of Economics and Finance 31 (2015) 94–107 97

Table 1
Sample selection (2000–2012).

Description Initial sample (firm–years) No. of observations (firm-years)

Mining
No. of firms 757 9841
Exclude outliers 42 546
Final sample 715 9295

Non-Mining
No. of firms 952 12,376
Exclude outliers 47 611
Final sample 905 11,765

Total
No. of firms 1709 22,217
Exclude outliers 89 1157
Final sample 1620 21,060

vary significantly among industrial non-financial, financial and mining companies (Balachandran &
Tanner, October, 2001). It is also found that share issuance decisions are consistent with the market
timing hypothesis, which has strong impact on leverage decisions for mining firms (Elsas, Flannery, &
Garfinkel, 2014).

3. Data and summary statistics

This study uses all available listed firms on the ASX that are collected from DatAnalysis Premium for
the period of 2000–2012 provided by Morning Star. The initial sample consists of 1709 non-financial
companies. Financial companies are excluded from the study due to a lack of critical data, and to
be consistent with previous research (Baker & Wurgler, 2002; Hovakimian, 2006; Leary & Roberts,
2005). We also exclude the extreme outlier applying the above and below 5 per cent range of all the
control variables. Then, to analyse whether the type of industry matters when making leverage deci-
sions, the data are divided into two groups; mining and non-mining firms using the general industry
classification code (GICS).iiiiii Mining firms are represented under Materials (also Metals & Mining)
industry where as non-mining firms include firms from all other industry groups that are listed on
the ASX (except financials). For example; Consumer discretionary, Consumer services, Consumer sta-
ples, Food, beverage &tobacco, Food & staples retailing, Energy, Health care, Industrials, Information
technology and Utilities. Table 1 outlines the process of sample selection.

3.1. Summary statistics for the control variables

Table 2 reports the means, medians and standard deviations for book leverage, market-to-book
values, firm size and other firm characteristics that are important across our sample of mining and
non-mining firm analysis. Table 2 descriptive statistics results are consistent with previous research
(Baker & Wurgler, 2002; Hovakimian, 2006) and show that there is variation in book leverage and
other variables between mining and non-mining firm results, though both show similar patterns for
the control variables.
Fig. 1 also shows the year-wise average in graphical form of each of the control variables for mining
and non-mining firms.

4. Methodology

Panel data analysis with pooled ordinary least square (OLS) fixed effects estimator is used in the
analysis and we follow the standard capital structure model used in previous studies. The motivation

iiiiii
GICS code collected from DatAnalysis Premium data base (Standard & Poor’s and MSCI Barra – GICS Structure).
98 S.Z. Islam, S. Khandaker / North American Journal of Economics and Finance 31 (2015) 94–107

Fig. 1. Year wise average of leverage, market-to-book, firm size, tangibility and profitability.
 S.Z. Islam, S. Khandaker / North American Journal of Economics and Finance 31 (2015) 94–107 99

Table 2
Summary statistics for sample variables for the period, 2000–2012.

Total firms, N = 1620 Mining firms, N = 715 Non-mining firms, N = 905

 
Firm year observations
D
21,060 9295 11,765
Book leverage, A
t−1
Mean 0.278 0.160 0.351
Standard deviation 0.246 0.201 0.243
Minimum 0.000 0.000 0.000
Maximum 1.00 1.00 1.00
M
Market-to-book ratio, B
t−1
Mean 1.910 2.005 1.854
Standard deviation 1.416 1.469 1.381
Minimum 0.220 0.000 0.220
Maximum 7.796 7.763 7.796

Firm size Log(S)t−1

Mean 15.365 13.216 16.382
Standard deviation 3.818 3.74 3.411
Minimum 1.792 1.792 3.583
Maximum 25.13 25.13 24.73
 PPE 
Fixed asset tangibility, A
t−1
Mean 0.242 0.255 0.268
Standard deviation 0.268 0.288 0.268
Minimum 0.00 0.000 0.000
Maximum 1.00 1.00 1.00
 EBITDA 
Profitability, A
t−1
Mean −0.138 −0.202 −0.099
Standard deviation 0.312 0.288 0.319
Minimum −1.45 1.79 −1.45
Maximum 0.270 25.13 0.270

The sample consists of 715 mining and 905 non-mining firms’ data for the period from 2000 to 2012. This table reports the
mean, standard deviation, minimum and maximum values for book value of leverage (book debt to assets), market-to-book ratio
(assets minus book equity plus market equity all divided by assets), firm size (log of total revenue), fixed asset tangibility (net
property, plant and equipment divided by assets) and profitability (operating income before interest, taxes and depreciation
divided by assets).

behind using the panel data for the analysis is because of possible information and estimation efficiency
gains (Gujarati, 2003). The fixed effect method (FEM) is a more general approach to eliminate omitted
variable bias in the multivariate regression analysis that could result from estimation based on the
simple OLS model. For example, using OLS estimation the regression function can be organized as
follows:

Yit = ˛ + ˇXit + εit (1)

The model (1) is expanded as follows applying fixed effects specification:

Yit = ˛i + ˛t + ˇXit + εit (2)

As this is a balanced panel fixed effect model, here, Yit is the dependent variable matrix that repre-
sents leverage in the study over the period for all observations. X is the independent variable matrix.
This represents a matrix of explanatory variables including market-to-book; firm size etc. ˛ is the
intercept and ˇ is the vector of slope coefficients.
The fixed effects model given in (2), assumes that the estimated slope coefficients (ˇ) do not vary
across companies or over time though dummy variable coefficients (˛i , ˛t ) are estimated for each
company/each year to capture unobserved heterogeneity in the data.
100 S.Z. Islam, S. Khandaker / North American Journal of Economics and Finance 31 (2015) 94–107

4.1. Models and definition of variables

The model used in analysis and the definitions for all the explanatory variables used to proxy
for alternative theories are discussed here, for example; market timing, agency costs, and pecking
order theories. The choice of control variables is based on the existing literature (Baker & Wurgler,
2002; Fama & French, 2002; Frank & Goyal, 2003; Leary & Roberts, 2005; Rajan & Zingales, 1995). To
document the relationship between leverage and alternative theories the study uses the following
model:
D D M   PPE   EBITDA 
− = a+b +c +d
A t A t−1 B t−1 A t−1 A t−1
D
+ e log (S)t−1 + f + ut (3)
A t−1

    
Here, in Eq. (3) we consider the changes in book leverage D/A − D/A as the dependent
t t−1

 examining
variable while  the relationship between capital structure and its determinants. Book value
of leverage D/A , is defined as book debt to assets where book debt (D) is the total assets minus book
 in book leverage we
equity. Book equity (E) is defined as total assets less total liabilities. And for changes     
simply take the difference between current and previous periods book leverage D/A − D/A .
  t t−1
The market-to-book ratio M/B is used as a proxy for market timing. It is defined as assets minus book
equity plus market equity all divided by total assets, where market value of equity is, ordinary share
price × shares outstanding. It is assumed that market-to-book may be related to investment oppor-
tunities, growth opportunities and market mispricing. Firms with high market-to-book ratios tend to
issue equity or to use internal funds that lead to debt reduction. Therefore, an inverse relationship
between leverage and market-to-book is postulated.
Three additional control variables are included in the model that are found to be correlated with
leverage in several developed countries (Baker & Wurgler, 2002; Fama& French,  2002; Fama &
MacBeth, 1973; Rajan & Zingales, 1995). They are fixed asset tangibility PPE/A , which is used as
a proxy for agency theory and is defined as net property, plant and equipment divided by total assets.
Agency theory suggests that firms with high leverage are reluctant to invest and thus, firms want to
transfer wealth away from debt holders to equity holders. As a result, lenders require collateral because
the use of secured debt can help ease this problem. Hence, firms unable to provide collateral have to
pay higher interest or be forced to issue equity instead of debt (Scott, 1977).  Therefore, a positive
relationship between leverage and asset tangibility is anticipated. Profitability EBITDA/A , is defined
as earnings before interest, taxes and depreciation divided by total assets. The pecking order theory
suggests that firms prefer to use internal funds because of the informational asymmetry between
managers and outside investors. In addition, profitable firms tend to reduce external equity in order
to minimise the impact on existing ownership. Thus, an inverse relationship between leverage and
profitability is expected. Firm size [Log(S)] is defined as the natural logarithm of total revenue. As large
firms are less likely to face financial distress, sizeis expected
 to have a positive impact on leverage.
Finally, the last control variable is lagged leverage D/A which is included in the model to capture
t−1
time series effects. The last variable, lagged leverage often enters the analysis with a negative sign
and this is consistent with the tendency for leverage to revert toward a long run equilibrium value
over time. Table 3 summarises the relationships proposed by the theory between each explanatory
variable and leverage.
The change in leverage is then decomposed into three components: net equity issues, newly
retained earnings and growth in assets, following Baker and Wurgler (2002). This is used to focus
on how various components of capital structure theory drive each of these three components (net
equity issues, retained earnings and asset growth). The decomposition takes the following form:
D D e  RE   1 1

− =− − − Et−1 − (4)
A t A t−1 A t A t At At−1
 S.Z. Islam, S. Khandaker / North American Journal of Economics and Finance 31 (2015) 94–107 101

Table 3
Expected relationship between corporate factors and leverage.1

Variables Expected Mostly reported in Theories

theoretical relation the literatures

Market-to-book − − Market timing theory

+ Pecking order theory

Tangibility + + Agency theory and

Trade-off theory

Profitability − − Pecking order theory,

+ Trade-off theory and other theory:
dilution of ownership structure

Firm size + + Trade-off theory,

Agency theory and Other theories:
access to the market, economies of
scale.
− Other theory: information asymmetry

Lagged leverage − − Market timing theory

Optimal leverage

Here in (4), net equity issues are defined as the change in bookequity minus
the change in balance
sheet retained earnings divided by total assets and denoted as −(e/A)t . Newly retained earnings

is defined as the change in retained earnings divided by total assets and denoted by −(RE/A)t .
Finally, the residual change in leverage (also known as growth in assets) is defined as lagged book
  divided by
equity total assets minus lagged book equity divided by lagged total assets and denoted as
Et−1 Et−1
− At
− At−1
.

4.2. Test for broad industry effects using dummy variable

In this section, dummy variables are applied in the original Eq. (3) to test for statistically significant
differences between mining and non-mining firm coefficients and for broad industry effects (mining
vs. non-mining firms) on firm leverage choices. The model used for testing is as follows:
D D  e   RE   1 1
 M 
t t
− =− − − Et−1 − =a+b
A t A t−1 At At At At−1 B t−1
 PPE   EBITDA  D M 
+c +d + e log (S)t−1 + f +g
A t−1 A t−1 A t−1 B t−1
 PPE   EBITDA  (5)
∗Dummy + h ∗ Dummy + i ∗ Dummy + j log (S)t−1
A t−1 A t−1
D
∗Dummy + k ∗ Dummy + ut
A t−1

Here, in Eq. (5), the dummy is set to 1 for non-mining firms and set to 0 for mining firms. The first
line provides estimates of the coefficients that apply to mining firms and the second line of coefficients
refers to the difference in the coefficients (the coefficient for the non-mining firms less the coefficient
for the mining firms).

5. Empirical results

5.1. Determinants of annual changes in leverage

Table 4 presents the regression results of annual changes in leverage on each of the five determi-
nants included in the leverage Eq. (3) and the decomposition results using Eq. (4). We find that asset
 102
Table 4
Determinants of changes in book leverage and components (for the period, 2000–2012).

Analysis of annual change in book leverage and its components of market-to-book ratio, fixed assets, profitability, firm size and lagged leverage for Mining and Non-mining firms are
presented in Table  effects
4. Fixed  panel  analysis  below.
is used for the model   
D D et REt 1  M PPE    

S.Z. Islam, S. Khandaker / North American Journal of Economics and Finance 31 (2015) 94–107
1
A
− A
= − − − Et−1 At
− At−1
= a + b +c + d EBITDA
A
+ e log (S)t−1 + f DA + ut
t t−1 At At B t−1 A t−1
t−1 t−1
D
The intercept, a, is not reported. N is the number of observations used in the analysis. Book value of leverage is defined as book debt to assets, at time t. The market-to-book
M  PPE  A
t
ratio is equal to assets minus book equity plus market equity divided by assets. Fixed assets tangibility, is defined as net property, plant and equipment divided by
B
 EBITDA  A

assets. Profitability is defined as operating income before interest, taxes, depreciation and amortization. Firm size is defined as the log of total revenue, (Log(S)t−1 ). The
A
 et 
explanatory variables are measured at time, t − 1. Panel A reports the annual change in leverage. Effect of net equity issues is reported in panel B where net equity issues, is
 REt At
defined as the change in book equity minus the change in retained earnings divided by assets. The newly retained earnings component is reported in Panel C and it is
1 1
 At

defined as the change in retained earnings divided by assets. Finally, panel D reports the components of residual change in leverage Et−1 At
− At−1
that depends on the total
growth in assets.a Robust t-statistics are reported in parentheses.

 
M/Bt−1 PPE/At−1 EBITDA/At−1 Log(S)t−1 D/A R2
t−1

Estimate b t(b) c t(c) d t(d) e t(e) f t(f)

Panel A: Changes in book leverage (( ⁄A)t ) D

Mining firms, N = 9295 −0.010 (−0.74) 0.103 (2.28)* −0.024 (−2.14)* 0.023 (1.54) −0.417 (−2.77)* 0.54
Non-mining firms, N = 11,765 0.003 (0.71) 0.042 (1.33) 0.061 (1.77) 0.030 (1.86) −0.541 (−8.50)** 0.58
 
Panel B: Changes in book leverage through net equity issues −e/At
Mining firms −0.004 (−0.06) −0.281 (−3.18)* −0.021 (−0.49) 0.056 (1.81) 0.401 (0.96) 0.22
Non−mining firms −0.047 (−3.74)* −0.015 (−0.33) 0.080 (1.35) 0.070 (5.19)* −0.144 (−1.30) 0.21
 
Panel C: Changes in book leverage through newly retained earnings −(RE/At )
Mining firms −0.018 (−0.23) 0.212 (1.03) −0.044 (−1.15) −0.020 (−0.47) −1.227 (−1.70) 0.25
Non-mining firms 0.024 (2.19)* −0.015 (−0.18) −0.155 (−1.80) −0.092 (−4.90)** −0.391 (−2.10)* 0.23
  
1 1
Panel D: Changes in book leverage through growth in assets − Et−1 −
At At−1
Mining firms 0.012 (0.52) 0.172 (1.33) 0.042 (6.05)** −0.013 (−0.54) 0.408 (1.36) 0.19
Non-mining firms 0.027 (4.81)** 0.073 (0.64) 0.136 (3.89)** 0.053 (2.93)* −0.006 (−0.04) 0.21
a
The total growth in assets is the combination of net equity issues, net debt issues and newly retained earnings.
*
Less than 5% level significance.
**
Less than 1% significance.
 S.Z. Islam, S. Khandaker / North American Journal of Economics and Finance 31 (2015) 94–107 103

tangibility shows a significant positive effect on leverage for mining firms and it suggests that mining
firms can hold more debt if more assets are in place. It is evident from the Panel B that the higher
the assets lower the equity issues, which helps to explain that mining firms will tend to issue debt if
they need external finance. The effect of tangibility is insignificant for non-mining firms. The result
also shows that profitability has a significant negative effect for mining firms and a positive effect for
non-mining firms. It appears that mining firms tend to follow the pecking order model more closely
than non-mining firms. This suggests that the higher the leverage, the greater the cash flow for mining
firms. This effect is not driven by equity issues or retained earnings but rather is driven by growth
options (see Panel B–D of Table 3). Further, we document that the market timing has no significant
effect on leverage for non-mining firms or mining firms. However, the decomposition results show
that the effect of market timing on leverage operates through each of the components, especially
for non-mining firms. Another result worth noting is the effect of lagged leverage that is the mean
reversion term. It shows a significant negative effect on leverage as expected.
The overall results reported in Table 4 suggests that mining firms appear to behave as if the pecking
order model is important for them, which suggests that mining firms tend to use internally generated
cash flows. This is consistent with the prior literature that documents the existence of the pecking
order hypothesis in Australian companies (Allen, 1991, 1993; Gatward & Sharpe, 1996) and we also
find evidence of fundamental differences between mining and non-mining firms.

5.2. Dummy variable approach

This study re-estimates the original regression model (Eqs. (3) and (4)) applying dummy variables
(Dummy) to test for significant differences in mining verses non-mining firm coefficient estimates
(Eq. (6)). Consistent with the findings in the existing literature, we observed a fundamental difference
between mining and non-mining firms. The dummy is set to 1 for non-mining firms and to 0 for
mining firms. The results of the mining firm dummy variable are included in an analysis reported in
Table 5. Panel A of Table 5 shows significant differences between mining and non-mining firms for asset
tangibility and profitability at the 10 per cent significant level. Non-mining firms are more sensitive
than mining firms to asset tangibility where mining firms seem more sensitive to profitability. There
is no significant mining firm difference noted for the market-to-book effect with respect to leverage.
Decomposition results reported in Panel B–D of Table 4 generally show differences between mining
and non-mining firms for profitability, firm size and lagged leverage. These significant differences
between mining and non-mining firms are consistent with prior studies in other areas of research
(Balachandran & Tanner, October, 2001; Clements & Johnson, 2000; How, 2000; Lee et al., 1996).

5.3. Wald-test result

We further explore significant differences between mining and non-mining firms by applying a
Wald-Coefficient Restriction test for Eq. (3). The Wald test is a commonly used hypothesis testing
method in multiple regression analysis that can be estimated in Eviews. Wald-test results are pre-
sented in Table 6. There is evidence of significant differences between mining and non-mining firm
results and these results are consistent with those reported in Table 4. In summary, there are statis-
tically important differences documented across the broader populations of mining and non-mining
firms.

5.4. Robustness

We have shown that industry-types have significant association with leverage decisions made by
firms, where mining firms are more sensitive towards profitability, which suggests mining firms will
follow a pecking order model more closely than non-mining firms. We then rerun the regression using
a pooled OLS estimation method. All results from re-testing are consistent with those detailed above.
We find that leverage decisions do vary between mining and non-mining firms and thus these results
are not reported separately.
 104
Table 5
Determinants of change in book leverage with dummy variable (For the period 2000-2012).

Analysis of annual change in book leverage and its components with respect to market-to-book ratio, fixed assets, profitability, firm size and lagged leverage including dummy
 for mining
variable   non-mining
and   firms.  Fixed
 effects
 panel   
analysis is used  below.
for the model     

S.Z. Islam, S. Khandaker / North American Journal of Economics and Finance 31 (2015) 94–107
D D et REt 1 1 M PPE EBITDA D
− =− − − Et−1 − =a+b +c +d + e log (S)t−1 + f
  A t−1
A t
  At
At
 At At−1  B t−1
A t−1  A t−1
A t−1
M PPE EBITDA D
+g ∗ Dummy + h ∗ Dummy + i ∗ Dummy + j log (S)t−1 ∗ Dummy + k ∗ Dummy + ut
B t−1 A t−1
A t−1
A t−1
The intercept, a, is not reported. Total 21,060 observations are used in the analysis. Here, the dummy is set to 1 for non-mining firms and set to 0 for mining firms. Book value of
leverage is defined as book debt to assets. The market-to-book ratio is equal to assets minus book equity plus market equity divided by assets. Fixed assets tangibility is defined as
net property, plant and equipment divided by assets. Profitability is defined as operating income before interest, taxes, depreciation and amortization divided by total assets. Firm
size is defined as the log of total revenue. Panel A reports the annual change in leverage. Effect of net equity issues is reported in panel B where net equity issues, is defined as the
change in book equity minus the change in retained earnings divided by assets. The newly retained earnings component is reported in Panel C and it is defined as the change in
retained earnings divided by assets. Finally, panel D reports the components of residual change in leverage. The coefficients from the regression are reported on two separate lines.
The first line refers to the coefficients estimated for each variable and the second line refers to the coefficients estimated for the product of the dummy and each variable. While the
first line provides estimates of the coefficients that apply to the mining firms, the second line of coefficients refers to the difference in the coefficients (the coefficient for the
non-mining firms less the coefficient for the mining firms). Robust t-statistics are in parenthesis.

 
M/Bt−1 PPE/At−1 EBITDA/At−1 Log(S)t−1 D/A R2
t−1

Estimate b t(b) c t(c) d t(d) e t(e) f t(f)

Panel A: Changes in book leverage (( ⁄A)t )

D

Variables −0.013 (−1.01) 0.104 (2.77)* −0.024 (−0.61) 0.023 (1.49) −0.420 (−2.12)* 0.39
Dummy* variable 0.018 (1.40) −0.074 (−1.87) −0.086 (−1.97) 0.013 (0.55) −0.129 (−0.57)
 
Panel B: Changes in book leverage through net equity issues −e/At
Variables −0.002 (−0.03) −0.156 (−1.35) −0.026 (−0.68) 0.062 (1.60) 0.389 (1.17) 0.21
Dummy* Variable −0.040 (−0.66) 0.140 (1.17) 0.114 (1.88) 0.003 (0.04) −0.525 (−1.39)
 
Panel C: Changes in book leverage through newly retained earnings −(RE/At )
Variables −0.024 (−0.49) 0.054 (0.33) −0.038 (−0.51) 0.002 (0.03) −1.239 (−1.95) 0.21
Dummy* Variable 0.046 (0.89) −0.042 (−0.20) −0.102 (−1.30) −0.086 (−1.13) 0.858 (1.26)
 1 1

Panel D: Changes in book leverage through growth in assets − Et−1 At
− At−1
Variables 0.013 (0.62) 0.207 (2.65) *
0.040 (1.11) −0.040 (−2.22)* 0.431 (2.72)* 0.18
Dummy* Variable 0.012 (0.47) −0.174 (−1.08) 0.074 (1.02) 0.097 (4.03)** −0.462 (−2.53)*
*
Less than 5% level significance.
**
Less than 1% significance.
 S.Z. Islam, S. Khandaker / North American Journal of Economics and Finance 31 (2015) 94–107 105

Table 6
Wald coefficient tests for Eq. (3).

[Normalised restriction (=0)] Fixed effect model

F-stat P value

Change in leverage regression (Table 4, Panel A) 9.42 <0.001

Change through net equity issues (Table 4, Panel B) 3.71 <0.001
Change through newly retained profit (Table 4, Panel C) 6.32 <0.001
Change through growth in assets (Table 4, Panel D) 12.04 <0.001

The change in leverage regression and the dummy variable approach is also re-estimated by chang-
ing the definition of leverage. The dependent variable, leverage is defined as long-term debt plus
short-term debt over total assets for the period t. Thus change in leverage, is defined as leverage
at time (t) minus leverage at time (t − 1). The control variables are the firm characteristics used in
previous research. Using the different definition of leverage does not affect the result, which shows
that mining firms appear to follow the pecking order model more closely than non-mining firms and
there is evidence of fundamental differences across the broader population of mining and non-mining
firms. Thus, as the study yields no new conclusions for these regression analyses, these results are not
reported separately here.

6. Conclusion

Our research makes two contributions to the existing capital structure literature. First, we explore
the relationship between leverage and theories of capital structure using 1620 mining and non-mining
firms from Australia. The pecking order theory suggests that there should be a negative relationship in
cross-section between corporate profitability and debt ratios and Allen (1991, 1993) finds evidence to
support the existence of the pecking order hypothesis in Australian firms. Similarly, Chiarella et al.’s
(1992) results are also supportive of the pecking order theory as well as being consistent with the
findings of Titman and Wessels (1988) which find no support for growth opportunities. Evidence
consistent with this view is provided in this paper, especially for mining firms. We find that mining
firms appear to follow the pecking order model more closely than non-mining firms with respect to
leverage choice compared to other theories of capital structure. Consistent with previous literature
we also find that firm size and a mean reversion term play an important role in leverage decisions for
both types of firms.
Second, this paper examines the fundamental differences between Australian mining and non-
mining firms with respect to firms’ leverage decisions by applying a dummy variable approach. The
results show that except for market-to-book, there are significant differences between mining and
non-mining firms. We also provide evidence of this result by conducting a Wald-test. This test result
suggests that, there appears to be a broad industry differences in the data set. Thus, the important
contribution of this analysis is that it shows leverage decisions are important for both mining and
non-mining firms, though the analysis identifies variation in the results between these two groups.

Acknowledgements

We thank Professor Christine Jubb, (Director) Centre for Enterprise Performance in the Faculty
of Business and Enterprise, Swinburne University for her helpful comments. The first author thanks
Professor Lisa Farrell at the School of Economics, Finance and Marketing, RMIT University for her con-
structive feedback during the “12 weeks writing workshop” in which this paper was revised. And we
also thank the anonymous referees participated in the business research seminar series at Swinburne
University for helpful comments that have greatly improved the paper.

References

Akhtar, S. (2005). The determinants of capital structure for Australian multinational and domestic corporations. The Australian
Graduate School of Management, 30, 321–341.
106 S.Z. Islam, S. Khandaker / North American Journal of Economics and Finance 31 (2015) 94–107

Allen, D. (1991). The determinants of capital structure of listed Australian companies: The financial manager’s perspective.
Australian Journal of Management, 16, 103–127.
Allen, D. (1993). The pecking order hypothesis: Australian evidence. Applied Financial Economics, 3, 101–112.
Alti, A. (2006). How persistant is the impact of market timing on capital structure. The Journal of Finance, 61, 1681–1710.
Baker, M., & Wurgler, J. (2002). Market timing and capital structure. The Journal of Finance, 57, 1–30.
Balachandran, B., & Tanner, S. (2001). Bonus issues: Australian evidence. Social Science Research Network Electronic Paper Collec-
tion.
Ball, R., & Brown, P. (1980). Risk and return from equity investments in the Australian mining industries: January 1958–February
1979. Australian Journal of Management, 5, 45–66.
Bevan, A., & Danbolt, J. (2002). Capital structure and its determinants in UK—A decompositional analysis. Applied Financial
Economics, 12, 159–170.
Bevan, A., & Danbolt, J. (2004). Testing for inconsistencies in the estimation of UK capital structure determinants. Applied Financial
Economics, 14, 55–66.
Bie, & Haan. (2007). Market timing and capital structure: Evidence for Dutch firms. De Economist, 155, 183–206.
Brailsford, T. J., Oliver, B. R., & Pua, S. L. H. (2002). On the relation between ownership structure and capital structure. Accounting
and Finance, 42, 1–26.
Brusov, P., Tatiana, F., Eskindarov, M., Brusov, P., Orehova, N., & Brusova, A. (2012). Influence of debt financing on the effectiveness
of the finite duration investment project. Applied Financial Economics, 22, 1043–1052.
Cassar, G., & Holmes, S. (2003). Capital structure and financing of SMEs: Australian evidence. Accounting and Finance, 43, 123–147.
Ceung, Y.-L., Jing, L., Lu, T., Rau, P. R., & Stouraitis, A. (2009). Tunneling and propping up: An analysis of related party transactions
by Chinese listed companies. Pacific-Basin Finance Journal, 17, 372–393.
Chang, X., & Dasgupta, S. (2009). Target Behaviour and Financing: How conclusive is the evidence? The Journal of Finance, 64(4),
1767–1796.
Chirinko, R., & Singha, A. (2000). Testing static tradeoff against pecking order models of capital structure: A critical comment.
Journal of Financial Economics, 58, 417–425.
Chiarella, C., Pham, T. M., Sim, A. B., & Tan, M. L. (1992). Determinants of corporate capital structure: Australian evidence.
Pacific-Basin Capital Markets Research, 3, 139–158.
Clements, K., & Johnson, P. (2000). The minerals industry and employment in Western Australia: Assessing its impact in federal
electorates. Resources Policy, 26, 77–89.
Da Silva Rosa, R. (1995). Measuring Long-run performance: The impact of survivorship bias, firm, size and return asymmetry.
In Academy of International Business South-East Asia Regional Conference, Perth (pp. 531–538).
Da Silva Rosa, R., Velayuthen, G., & Walter, T. (2003). The share market performance of Australian venture capital-backed and
non-venture capital backed IPOs. Pacific-Basin Finance Journal, 11, 197–218.
Elsas, R., Flannery, M. J., & Garfinkel, J. A. (2014). Financing major investments: Information about capital structure decisions.
Review of Financial Studies, 18, 1341–1386.
Fama, E., & French, F. K. R. (2002). Testing trade-off and pecking order predictions about dividends and debt. The Review of
Financial Studies, 15, 1–33.
Fama, E., & MacBeth, F. J. D. (1973). Risk, return, and equilibrium: Empirical tests. Journal of Political Economy, 81, 607–636.
Fischer, E., Heinkel, R., & Zechner, J. (1989). Dynamic capital structure choice: Theory and tests. The Journal of Finance, 44,
19–40.
Fiscore, S. (2007). Australian Mining Technology. Engineering and Mining Journal, 208(6), 41–42, 44–49.
Frank, M., & Goyal, V. (2003). Testing the pecking order theory of capital structure. Journal of Financial Economics, 67, 217–248.
Gatward, P., & Sharpe, I. G. (1996). Capital structure dynamics with interrelated adjustment: Australian evidence. Australian
Journal of Management, 21, 89–112.
Goldstein, R., Ju, N., & Lealand, H. (2001). An EBIT test model of dynamic capital structure. Journal of Business, 74, 483–512.
Gujarati, N. D. (2003). Basic econometrics (4th ed., pp. 10020). New York, NY: McGraw-Hill Higher Education, 1221 Avenue of
the Americas.
Harris, M., & Raviv, A. (1991). The theory of capital structure. The Journal of Finance, 46, 297–355.
Hovakimian. (2006). Are observed capital structures determined by equity market timing? Journal of Financial and Quantitative
Analysis, 41, 221–243.
How, J. (2000). Initial and long-run performance of mining IPOs in Australia. Australian Journal of Management, 25, 95–118.
Jalilvand, A., & Harris, R. S. (1984). Corporate behaviour in adjusting to capital structure and dividend targets: An econometric
study. The Journal of Finance, 39, 127–145.
Jensen, M. C. (1986). Agency costs of free cash flow, corporate finance, and takeovers. The American Economic Review, 76, 323–329.
Jensen, M., & Meckling, C. W. H. (1976). Theory of the firm: Managerial behaviour, agency costs and ownership structure. Journal
of Financial Economics, 3, 305–360.
Jenter, D. (2005). Market timing and managerial portfolio decisions. The Journal of Finance, 4, 1903–1949.
Leary, M., & Roberts, T. M. R. (2005). Do firms rebalance their capital structures. The Journal of Finance, 6, 2575–2619.
Lee, J. P., Taylor, L. S., & Walter, S. T. (1996). Australian IPO pricing in the short run and long run. Journal of Banking and Finance,
20, 1189–1210.
Li, M., & Zhao, X. (2014). Impact of leveraged ETF trading on the market quality of component stocks. North American Journal of
Economics and Finance, 28, 90–108.
Mackay, P., & Phillips, G. (2005). How does industry affect firm financial structure? The Review of Financial Studies, 18, 1433–1466.
Marsh, P. (1982). The choice between equity and debt: An empirical study. The Journal of Finance, 37, 121–144.
Modigliani, F., & Miller, M. H. (1958). The cost of capital, corporation finance, and the theory of investment. American Economic
Review, 48, 655–669.
Myers, S., & Majluf, C. N. S. (1984). Corporate financing and investment decision when firms have information that investors do
not have. Journal of Financial Economics, 13, 187–221.
Naidu, G. N. (1986). Capital structure strategies of Australian and South African firms. Management International Review, 26,
52–61.
 S.Z. Islam, S. Khandaker / North American Journal of Economics and Finance 31 (2015) 94–107 107

Rajan, R., & Zingales, G. L. (1995). What do we know about capital structure: Some evidence from international data. The Journal
of Finance, 50, 1421–1460.
Schoder, C. (2013). Credit vs. demand constraints: The determinants of US firm-level investment over the business cycle from
1977–2011. North American Journal of Economics and Finance, 26, 1–27.
Scott, J. (1977). Bankruptcy, secured debt and optimal capital structure. The Journal of Finance, 32, 1–19.
Shyam-Sunder, L., & Myers, S. C. (1999). Testing static trade-off against pecking order models of capital structure. Journal of
Financial Economics, 51, 219–244.
Soros, G. (2008). The new paradigm for financial markets: The credit crisis of 2008 and what it means. Public Affairs Books.
Titman, S., & Wessels, R. (1988). The determinants of capital structure choice. The Journal of Finance, 43, 1–18.
Twite, G. (2001). Capital structure choices and taxes: Evidence from Australian dividend imputation tax system. International
Review of Finance, 2, 217–234.
Welch, I. (2004). Capital structure and stock returns. Journal of Political Economy, 112, 106–131.