Is negative profitability-leverage relation the only support for the pecking order theory in case of Pakistani firms?
Shah, Attaullah ; Ilyas, Jasir
Previous studies on capital structure in Pakistan have reported
evidence in support of the pecking order theory. However, this evidence
is largely based on testing one dimensional relationship between
leverage ratios and firms' profitability. The objective of this
paper is to extensively test the pecking order theory in Pakistan with
well-known pecking order testing models. Specifically, we use a sample
of 321 firms listed on the Karachi Stock Exchange from 2000 to 2009 and
test pecking order theory with models suggested by Shyam-Sunder and
Myers, Frank and Ooyal, Watson and Wilson, and Rajan and Zingales.
Results of these models indicate that there exits only weak evidence in
support of pecking order theory in Pakistan. However, strong support is
found for pecking order theory when leverage ratios are regressed on
profitability ratio, along with a set of control variables. This
discrepancy in the results of the two sets of models needs further
investigation, as well as care in interpreting the results of existing
studies on capital structure in Pakistan. Our results show robustness
even ajter controlling for possible profits understatements or weak
corporate governance practices.
JEL Classification: G10, G21, G32
Keywords: Pecking Order Theory, Profitability-Leverage Relation,
KSE
1. INTRODUCTION
Many theories have been presented and tested to explain corporate
capital structure choices; however, none of these theories has been able
to come up with a comprehensive explanation of the capital structure
choices of firms in different industries and/or countries. Because of
this reason, Brealey, Myers, and Marcus (1999) included corporate
capital structure in seven unanswered subject issues of finance. Debate
over capital structure decision started with the ground-breaking work of
Modigliani and Miller (1958) who argued that corporate capital structure
is inconsequential to the value of a firm and hence there exists no
optimal capital structure. However, Modigliani and Miller reached this
conclusion under the assumption of perfect capital markets. Once the
assumption of perfect capital markets is relaxed and real-world market
imperfections are allowed to play a role in firm-financing decisions,
then optimal capital structure does exist.
Existing capital structure theories build their arguments around
different market frictions such as taxes, information asymmetry, agency
costs, and different types of transaction costs. Among these theories,
the most heavily discussed and empirically tested theories are the
trade-off theory and the pecking order theory. The trade-off hypothesis
was first proposed by Kraus and Litzenberger (1973); however, it was
later modified and refined by a number of studies. The trade-off theory
proposes that there exists a trade-off between the benefits and costs of
debt financing. Debt financing benefits a firm because interest expense
serves as tax-shield. And cost of debt financing arises from the
increase in probability of bankruptcy as debt financing subjects a firm
to fixed periodic interest and principal payments. An optimal capital
structure is reached at a point when benefits and costs of debt
financing from a one additional dollar of debt financing become equal.
The pecking order theory states that financing behaviour of a firm
follows a pecking order because information asymmetry costs are
different for different sources of funds [Myers (1984)]. When funds are
required by a firm, it first uses the internally-generated funds.
Internally available funds can be employed to meet funding requirements
without information costs and time constraints. When the funding
requirements exceed internally available funds, only then the firm opts
for external financing. While choosing between debt and external equity
financing, a firm prefers a less costly source of financing over the
costly one [Myers (1984)]. Equity has information asymmetry problem;
therefore, debt financing is a less costly choice. Information asymmetry
means that managers and potential investors do not have equal
information regarding the firm's future cash flows. Potential
investors know that corporate managers will work in the interest of
existing shareholders and will issue equity only when shares are
overpriced in the market. Therefore, when equity is issued, potential
investors will discount it in view of possible overpricing. This makes
issuance of equity costly for the existing equity holders. Consequently,
a firm will prefer to use debt before issuing equity when external
financing is required. This forms an order in financing behaviour of
firms. Firms first pick internally available funds i.e. retained
earnings as a financing source. If these internal finances are
inadequate to meet the funding requirements then the firm will opt for
external financing in order of preference from least risky debt
(straight debt) to more risky debt (convertible debts), preferred stocks
and lastly equity financing [Myers and Mujluf (1984) Myers, (2001)].
Empirical work has provided evidence, in favour of as well as
against, both the theories. The relationship between profitability and
leverage of a firm is considered as a focal point when it comes to
testing these theories. Under the trade-off theory, it is predicted that
profitable firms will try to use more debt financing. It is because
these firms are less risky and hence they will try to gain maximum tax
advantage provided by leverage [Barclay and Smith (2005)]. The tax
advantage associated with debt financing increases after tax cash flow
of the firm. This way the trade-off theory suggests a positive
relationship between profitability and leverage of the firm. Contrary to
the trade-off theory, the pecking order theory predicts a negative
relationship between profitability and leverage because a profitable
firm will have more retained earnings over a period of time. This
reserve of funds could be used as first choice of financing when the
firm is in need of funding for purchase of new assets or financing a
project. Thus, there will be less need for external financing. In
contrast, a less profitable firm will have less to retain and will be
unable to meet its funding requirement with internally generated funds.
Such a firm has to meet its funding requirements through external
financing, which according to the pecking order theory ought to be debt
financing. This way, the pecking order theory predicts a negative
relationship between leverage and profitability.
Many empirical studies have supported the pecking order theory
primarily based on the negative profitability-leverage relationship.
Booth, et al. (2001) studied firms in 10 developing countries and found
a negative relationship between profitability and leverage. Similarly,
[Tong and Green (2005)] reported significant inverse relationship
between the current as well as past profits and leverage. [Qureshi
(2009)] followed the work of Tong and Green (2005) and found support for
the pecking order theory on the basis of a negative relation between the
two variables. Moreover, studies such as Sinan (2010) and Ozkan (2001)
from UK; Sheikh and Wang (2010), Qureshi (2009), Ilyas (2008) and
llijazi and Shah (2004) from Pakistan; and Gaud, et al. (2005) from
Switzerland; and Serrasqueiro and Nunes (2010) from Portugal provided
evidence in favour of a negative relation between the two variables.
A major twist in testing the pecking order theory came with the
study by Shyam-Sunder and Myers (1999). They shifted the focus from
profitability-leverage relationship to a more refined proxy for testing
the pecking order hypothesis. They argued that external funding
requirements of a firm should be matched dollar to dollar by changes in
debt financing. Therefore, if pecking order theory holds, coefficient of
funding deficit should be one in the regression of net debt issues.
[Frank and Goyal (2003)] further modified the approach of Shyam-Sunder
and Myers (1999) to use individual components of funding deficit in the
regression of net changes in leverage levels, instead of using just one
composite figure for funding deficit. Following the approach of these
two studies, a large number of studies have reported mixed support for
the pecking order theory.
Existing studies on this topic in Pakistan [see, e.g., Sheikh and
Wang (2010) Qureshi (2009), Ilyas (2008), Khan and Shah (2007) Hijazi
and Shah (2004) and Booth, et al. (2001)] use the profitability-leverage
relation to explain the financing pattern of Pakistani firms. This
provokes a natural question whether the negative profitability-leverage
relationship is the only support for the pecking order theory in
Pakistan? Review of literature suggests that there are several models
such as models suggested by [Sunder and Myers (1999), Frank and Goyal
(2003) and Watson and Wilson (2002)] to test the pecking order theory.
These models use different assumptions and techniques to confirm the
existence of the pecking order theory apart from just
profitability-leverage relation. Therefore, a need for a comprehensive
study is felt which can find evidence in support of or against the
presence of the pecking order theory in financing pattern of Pakistani
listed firms using a set of recently developed alternative models in
this area. There is additional motivation for testing pecking order
theory in Pakistan. Pecking order theory considers information asymmetry
costs as the prime determinants of firms' financing choices. Since
information asymmetry problems are expected to be higher in developing
and emerging markets [see, e.g. Balasubramanian, et al. (2010), Jabeen
and Shah (2011), Seifert and Gonenc (2008b), Stiglitz (1989)], Pakistan
is a good candidate to test the pecking order theory.
Besides the above, our unique contribution to the literature lies
in the fact that no previous study in Pakistan on the given topic has
controlled for possible earning understatements or poor corporate
governance practices. There is some evidence that corporate governance
practices are weak in Pakistan where insider-controlling shareholders
try to expropriate outside minority shareholders or try to evade taxes
through earning understatements. [See, e.g., Abdullah, et al. (2012)].
Earning understatements or poor corporate governance might contaminate
our results. We have controlled for this possibility in two ways. First,
we estimate all regression models on a full sample of 321 firms, and
also on a sub-sample of 102 firms for which corporate governance
compliance score was available. This score was obtained from Tariq and
Abbas (2013) who measured compliance with the code of corporate
governance of Securities and Exchange Commission of Pakistan on more
than 50 dimensions. The subsample was further divided into two groups of
firms i.e. firms with higher compliance score and firms with lower
compliance score. Then all the regression models were estimated
separately for each group to compare whether corporate governance
practices drive our results. Our second approach to control for possible
earning understatement is to divide firms into three groups based on
25th, 50th, and 75th percentiles of firms' profitability. If
discretionary understatement of earnings have any effect on our
analysis, that should be visible in the results of these three groups.
Separate regressions were estimated to see whether explanatory variables
of interest behave randomly across these groups.
In next section, we review the relevant literature. After that, we
discuss the data sources, sample, and choice of models in Section 3. In
Section 4, results and findings of the empirical analysis are presented
and discussed. Section 5 concludes the paper.
2. REVIEW OF LITERATURE
This section reviews the relevant literature for developing a set
of testable hypotheses. The review specifically focuses on models used
to test pecking order theory of corporate capital structure.
Donaldson (1961) found that majority of firms used internally
generated funds as a first choice of financing even with fairly high PE
ratio. He formulated the pecking order hypothesis which was later on
modified and refined by Myers (1984) and Myers and Majluf (1984) Myers
(1984) proposed the pecking order in the context of asymmetric
information and highlighted the shortcomings of the trade-off theory in
the presence of correction costs to optimal leverage ratio. According to
Myers, a firm adjusts its capital structure to maximise its value by
changing the level of debt. Myers highlighted that trade-off theory
holds only when costs of these corrections are zero. Myers and Majluf
(1984) proposed that firms should issue equity only when balanced
information exists between managers and potential investors. However,
when the condition of balanced information does not hold, equity
issuance can be harmful to the interest of its existing equity holders.
This happens because potential investors know that managers will work in
the interest of existing shareholders and will issue equity only when
shares are overpriced in the market. Therefore, when equity is issued,
potential investors will try to correct the share price downward. They
do so because they feel they are exposed to adverse selection in the
presence of information asymmetry. Thus, information asymmetry between
managers and potential investor makes equity financing costlier. This
led Myers and Majluf to propose that in the presence of information
asymmetry, a firm should depend on past equity reserves or surplus
profits retained over period of time along with savings through
reduction in dividend pay-outs as a first choice of financing. If
internal funds are insufficient, firms would then choose debt financing
before going for equity issuance because debt issuance has lower
information asymmetry costs.
With the increasing focus on pecking order theory, researchers
developed several different models to test this theory under different
assumptions. These models focused primarily on how firms finance their
funding deficits. Among the pioneering works in this area was the study
by Shyam-Sunder and Myers (1999). Their model implies that for the
pecking order theory to hold, a dollar of financing deficit should be
funded by a dollar of debt financing. Thus, in the regression of net
debt issue, funding deficit should return slope coefficient of one. The
results of their study mostly favoured pecking order theory as compared
to trade-off theory. Chirinko and Singha (2000) criticised Shyam-Sunder
and Myers model (SSM) on the grounds that their model contained only
financing deficit and debt financing while equity financing was missing.
If equity, as a last resort, is accommodated in the model then slope
coefficient won't be equal to one as suggested by SSM model.
Furthermore, Chirinko and Singha marked other weaknesses of SSM model
such as it does not speak of the situation in which equity is issued
prior to debt or when debt and equity financing are used in fixed
proportions. Moreover, Frank and Goyal (2003) challenged the
generalisation of the empirical results in Shyam-Sunder and Myers (1999)
on grounds that the sample of 157 firms used in their study was
relatively small for publically traded US firms.
The SSM model argues that change in debt financing is purely a
result of change in funding deficit. A challenge to this argument comes
from target adjustment models which argue that changes in debt financing
show attempts of a firm to adjust to its target capital structure with
the passage of time. A number of studies used SSM model and target
adjustment models to test the pecking order theory. These studies
include [Dang (2005), Hovakimian and Vulanovic (2008), Seifert and
Gonenc (2008b)]. A brief overview of these studies is presented. Dang
(2005) tested the pecking order theory and the trade-off theory using a
sample of UK firms for the period 1996-2003. He found that most of the
tested firms adjust to their ideal leverage ratio with a substantial
speed. This study also tested both theories together is one model and
found that the trade-off theory did well in contrast to the pecking
order theory. Hovakimian and Vulanovic (2008) tested funding of the long
term retiring debt instead of funding deficit in SSM model. Conventional
SSM model regresses financing deficit on new debt financing. The study
argued that doing so was in line with pecking order theory as maturing
debts were financed by new debt after exhausting inside funds. This fact
was evident from negative intercept term which shows employment of
inside funds before new debt funding. The pecking order theory failed
when retiring debt was regressed on outside funding, i.e. debt and
equity together, where the regression produced a positive intercept
term. The study argued this failure is in line with the finding of Leary
and Roberts (2007). And finally, Seifert and Gonenc (2008b) argued that
emerging economies have more information asymmetry problems; therefore,
they should mostly follow pecking order theory. They tested pecking
order theory in 23 emerging economies. Results of their study revealed
that equity financing is preferred over debt financing in these emerging
economies which was inconsistent with pecking order theory. In a more
recent study, Komera and Lukose (2014) tested the role of pecking order
theory in Indian market and found that pecking order theory cannot
explain capital structure of the firms used in the sample.
Frank and Goyal (2003) argued that SSM model uses an aggregated
value for funding deficit, which is less informative. They suggested
that the funding deficit should be disaggregated into its individual
components and then be tested in conventional leverage regressions.
Using this modified model, Frank and Goyal (2003) studied US public
firms over a period 1971-1998 to know how these firm finance their
funding deficits. They found that the sample firms used equity financing
to meet funding deficit. Frank and Goyal also found that support for
pecking order theory declines over a period of time. This declining
support was found in case of both large and small firms. Large firms
somewhat tend to follow pecking order theory in comparison to small
firms. Theoretically, as highlighted by Berger and Udell (1995), small
firms should follow pecking order theory more than large firms as small
firms are more susceptible to information asymmetry problems. Frank and
Goyal argued that small firms did not follow pecking order theory
because most of them went public during 1980s and 90s. Later on, several
studies including Flannery and Rangan (2006) and Huang and Ritter (2007)
reported findings similar to that of Frank and Goyal. Seifert and Gonenc
(2008a) extended the work of Frank and Goyal (2003) to British, German
and Japanese firms along with American firms using OLS and fixed effect
models. They found results similar to Frank and Goyal study with
exception of Japanese firms. Overall results from US, Britain and
Germany do not support the presence of pecking order theory. However,
large sized US and German firms followed pecking order theory.
Importantly large sized US firms with higher profitability were
following pecking order theory but surprisingly in case of large sized
Japanese firms, even firms with low profitability were following pecking
order theory.
Several other studies have used quite different methodologies to
test pecking order theory. For example, [Bharath, et al. (2009)] tested
information asymmetry as a key driver of pecking order theory. They
found that with an increase in information asymmetry, firms avoid their
financing through equity which is in line with pecking order theory.
However, they argued that it does not completely determine the financing
source selection of the firms. In case of highest information symmetry,
only 30 percent of the funding requirements are fulfilled with debt
instead of 100 percent as implied by pecking order theory. They
concluded that information asymmetry is significant but not the sole
determinant of leverage. Another study that used a different approach to
test the pecking order theory was [Autore and Kovacs (2004)] who
investigated the pecking order theory in relation with changing adverse
selection cost over time. The study took dispersion in analysts'
earnings estimates as a measure of adverse selection cost. The study
used pooled and fixed-effect regression models and found that with the
lower adverse selection cost, firms tend to finance themselves via
outside sources, preferably with equity. However, in case of a firm with
higher adverse selection costs, traces of the pecking order theory were
found. The study further found that firm profitability is negatively
associated to adverse selection costs, outside financing and changes in
debt. And finally, [Ghosh and Cai (2004)] used different tests and found
that typically firms which have debt level greater than industrial
average ultimately move towards the industry mean debt ratio. This fact
shows that firms that have debt ratio above the industry average, they
trace the trade-off theory. Such firms try to readjust their debt level
towards the target industrial debt level by lowering it. Whereas those
firms which have a lower level of debt, do not show the same tendency as
they are not bothered by their existing debt level.
Fama and French (2005) argued that the pecking order theory is not
complete capital structure model as information asymmetry problem is not
a prime driver of financing choices. They argued that information
asymmetry problem can be avoided by changing the ways of issuing equity,
for example mergers can be financed with stock, repurchased plans,
employee's stock options and rights offering. Thus equity issuance
is not a last choice of financing as predicted by the pecking order
theory. In their empirical tests, they found that firms do issue equity
generally and retire equity even when firms have funding deficits which
is against the pecking order theory implications. In times of financing
surplus, firms do retire debts. Similar to [Lemmon and Zander (2004)],
Fama and French pointed out that usually firms with funds deficit, low
profitability and good growth opportunities issue equity.
In conclusion, the review reveals that testing pecking order theory
goes beyond using just profitability-leverage relationship. Second, only
mixed support exists for pecking order and that too is declining in the
recent times.
2.1. Hypotheses of the Study
In light of the literature cited above, we develop and test the
following hypotheses regarding the relevance of pecking order theory to
Pakistani corporate financing behaviour.
[H.sub.1]: Funding deficit determines the debt level of the firm.
[H.sub.2]: Internally available retained earnings are preferred
over debt financing.
[H.sub.3]: Aggregation of funding deficit components is less
informative.
[H.sub.4]: Funding deficit contributes more as a determinant of
leverage as compared to other conventional determinants of leverage.
[H.sub.5]: Retained earnings are preferred over debt financing
whereas debt financing is preferred over equity financing.
3. METHODOLOGY
3.1. Data of Study
Data for the study are taken from State Bank of Pakistan's
publication "Balance Sheet Analysis of Joint Stock Companies Listed
on the Karachi Stock Exchange". Sample period for the study covers
the years 2000 to 2009. Total number of non-financial firms listed in
2000 were 520; however, the number of listed firms decreased to 414 in
(2009). This study selected all firms which had complete data available
during the sample period. After exclusion of outliers and incomplete
data, we were left with a final sample of 321 firms.
3.2. Models to Test Pecking Order Theory
3.2.1. SSM Model
We start with the model developed by [Shyam-Sunder and Mayer (SSM)
(1999)]. This model has also been used by many empirical studies like
[Dang (2005), Hovakimian and Vulanovic (2008), and Seifert and Gonenc
(2008b), and Romera and Lukose (2014)]. These studies used the model
with slight amendments to test the pecking order theory. This model is
not considered a perfect model in general, which was accepted by
Shyam-Sunder and Myers. This model has been heavily criticised in
empirical studies such as [Chirinko and Singha (2000) etc.]. Still due
to its simplicity and good first order approximation this model has been
used in many studies around the globe in testing the pecking order
theory. The pecking order theory suggests that external equity financing
is used only as a last resort; whereas as a first option, firms will use
debt financing when their funding needs exceed the internally available
funds. So every dollar of a firm's deficit is met by each dollar of
debt financing of the firm, which will result in slope coefficient equal
to 1. Thus, this formulation can be expressed in the following form;
[DELTA][D.sub.it] = [alpha] + [[beta].sub.1], [DEF.sub.it] +
[e.sub.it] ... ... ... (1)
Whereas [DEF.sub.t] = [DIV.sub.t] + [X.sub.t] + [DELTA] [W.sub.t] +
[R.sub.t] - [C.sub.t]
In Equation (1), [DELTA][D.sub.it] shows the change in debt level
of a firm i between time t and t-1. This value is expected to be
positive if a firm faces funding deficit i.e. the firm will obtain
external financing. In case the funding deficit is negative, the firm
will retire its debt in that year. [[beta].sub.l], is the pecking order
coefficient. DEF represents the internal funds deficit. Funding deficit
is a combination of dividend payment (DIV), capital expenditure (X), net
increase in working capital ([DELTA]W) and current portion of long term
debt at the beginning of time t {R) minus operating cash flow after
interest and tax (C). All these components are expected to have a
positive relationship with funding deficit except operating cash flow
which should be negatively related to funding deficit. To control for
scale differences, all the variables are scaled by total assets.
Typical definition of funding deficit in SSM also includes current
portion of long term debt (R) as a component. However, Frank and Goyal
(2003) found that, contrary to pecking order theory, this component
showed negative relation with net debt issued. They also argued that
this component already exists in change in working capital component so
it does not need to be repeated. The fact that the funding from internal
sources is preferred over debt financing is represented by term
"a" in Equation (1) which is expected to have zero value.
3.2.2. Frank and Goyal Disaggregation Model
In contrast to SSM model, Frank and Goyal (2003) argued that
aggregation of funding deficit in one value is not very informative.
These components can reveal more information about debt financing
behaviour when studied independently. Therefore, they suggested that
funding deficit as in Equation (2) is more appropriate. Our second model
is adopted from Frank and Goyal (2003)
[DELTA][D.sub.it] = [alpha] + [[beta].sub.Div][DIV.sub.t] +
[[beta].sub.X][X.sub.t] + [[beta].sub.W] [DLETA][W.sub.t] +
[[beta].sub.c] [C.sub.1] + [e.sub.it] ... ... ... (2)
Theoretically, unit change in each of these components of funding
deficit must lead to unit change in debt financing i.e. [[beta].sub.Div]
= [[beta].sub.X] = [[beta].sub.W] a = [[beta].sub.C] = 1 to confirm the
pecking order theory.
3.2.3. Frank and Goyal Conventional Leverage Model
In order to address the omitted variable bias, Frank and Goyal
estimated another equation that incorporates all previously identified
explanatory variables in the leverage regression. In view of this, we
adopt the following model from Frank and Goyal. This model allows us to
find relevant contribution made by the variable of interest (i.e.
funding deficit), in the presence of other conventional variables. The
model is given below:
[DELTA][D.sub.it] - [alpha] + [[beta].sub.T][DELTA][T.sub.it] +
[[beta].sub.G] + [[beta].sub.LS][DELTA][LS.sub.it] + [[beta].sub.1],
[DELTA][P.sub.it] + [[beta].sub.DEF] [DEF.sub.it] + [e.sub.it] ... (3)
In Equation (3) [DELTA]D, [DELTA]T, [DELTA]LS, [DELTA]P, and DEF
show the changes in debt level, tangibility, size, and profitability
from previous period to the current period, and funding deficit of the
firm, respectively. All variables are scaled by total assets, except
growth and size which is the natural log of total assets.
A firm with higher tangibility ratio (i.e. proportion of fixed
asset to total assets) can borrow at a relatively lower rate of interest
by using fixed assets of the firm as collateral. A firm with a higher
percentage of fixed assets is expected to borrow more as compared to a
firm whose cost of borrowing is higher because of less fixed assets
[Bradley, el al. (1984); Rajan and Zingales (1995); Kremp, et al. (1999)
etc.]. From a different perspective, Harris and Raviv (1990) argued that
a firm with lower tangibility has more information asymmetry problem.
Therefore, under the pecking order theory, such a firm will go for more
debt financing in comparison to equity financing after utilisation of
internal funds. It is due to the fact that information asymmetry makes
equity financing as an expensive option. Thus, we expect a negative
relation between funding deficit and debt level of firms that have lower
tangibility ratios. In this study tangibility is measured as a ratio of
fixed assets to total assets.
According to the pecking order theory, a firm will use first
internally generated funds which may not be sufficient for a growing
firm. So next option for such growing firms is to use debt financing
which implies that a growing firm will have a high debt [Drobetz and Fix
(2003)]. Some studies suggest that firms with higher growth are expected
to have lower leverage. This is based on the fact that debt is supported
by assets-in-place rather than growth opportunities [Titman and Wessels
(1988)]. Previous empirical studies have used various proxies for growth
opportunities of a firm such as market-to-book ratio and yearly
percentage changes in capital expenditure and total assets. Firms with
high market-to-book value will opt more for equity financing. It is so
to take advantage of high market value than book value. Later two
proxies are component of funding deficit under SSM model and are
expected to be positively related to debt. This study measures growth as
a geometric mean of percentage increase of total assets of the firm with
respect to the previous year. In this study, it is expected that firms
with higher growth are expected to have higher leverage.
Size of the firm is closely related to risk and bankruptcy costs of
a firm. Large sized firms tend to be more diversified and as a
consequence they have a lower probability of bankruptcy. Thus creditors
will be more willing to lend their funds to larger firms. Examining the
effect of size in the determination of capital structure, Marsh (1982)
and Bennett and Donnelly (1993) found that larger and more capital
intensive companies are likely to employ more debt. On the other hand,
as highlighted by Berger and Udell (1995), small firms should follow
pecking order theory more than large firms as small firms confront
information asymmetry problem more than large firms. The study measures
the size as the natural logarithm of total assets.
About the profitability of the firm, trade-off theory predicts a
positive relationship between leverage and profitability of the firm by
arguing that highly profitable corporations in order to benefit from
debt tax advantages would employ more debt. Finding of many studies,
such as Titman and Wessels (1988), Baskin (1989), Allen (1993),
Michaelas, et al. (1999), Fama and French (2002) and Tong and Green
(2005) challenged this prediction. However, the pecking order theory
predicts that if a firm is profitable then it is more likely that
financing would be from internal sources rather than external sources to
finance their operations and investments. Debt typically grows when
investment exceeds retained earnings and falls when investment is less
than internal funds. Hence a negative relationship between leverage and
profitability is expected. This study measures profitability as net
income of the firm divided by total assets.
Similarly, when funding requirements are in excess of internal
funds, there is a need for external funding. External funding includes
both debt and equity. The pecking order theory predicts that increase in
funding requirement, i.e. funding deficit, results in more debt
financing along with equity. However, the pecking order theory suggests
a preference of debt over equity in the presence of information
asymmetry. This study expects positive relationship between funding
deficit and debt financing. Funding deficit is a sum of dividend
payment, capital expenditure, net increase in working capital minus
operating cash flow after interest and tax.
3.2.4. Watson and Wilson Model
We also use Watson and Wilson (2002) model. This model allows us to
test whether firm prefers debt over equity in situations when internal
funds have already been utilised. The model is given below;
[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] ... ... ... (4)
Whereas [summation][[alpha].sub.i] shows vector of firm fixed
effects, TA shows total assets of the firm, P shows profits, Div shows
dividends, E! shows changes in equity and finally D shows debts of the
firm. In this model, internally available funds are represented as a
remainder of profit after dividend payment. The equation tries to
capture changes in total assets in relation to changes in equity and
liabilities.
3.2.5. Rajan and Zingales Model
Lastly we use Rajan and Zingales (1995) model to know whether
negative leverage-profitability relationship responds differently to
different levels of profitability of Pakistani firms. Firms are
categorised into three groups based on their profit levels by using
25th, 50th and 75th percentiles. These three groups are named as low
profitable firms, average profitable firms and high profitable firms.
The following model is tested for this purpose;
[D.sub.it] = [alpha] + [[beta].sub.1]([T.sub.it]) +
[[beta].sub.2][G.sub.it] + [[beta].sub.3][LS.sub.it] +
[[beta].sub.4][P.sub.it] + [e.sub.it] ... ... ... (5)
Where [D.sub.it] is the total debt of firm i at time t, scaled by
total assets, [T.sub.it] is the ratio of tangible assets to total
assets, G is the geometric mean of annual percentage increase in total
assets, and P is the ratio of net income to total assets.
Under the pecking order theory, profitability of a firm should be
negatively related to the debt financing of the firms as internal funds
are preferred over external funding. Using Rajan and Zingales model for
the above stated groups our interest lies in knowing whether the
negative relation between profitability and debt financing remains the
same for each level of profit or not. If it so, it would mean that the
negative profitability-leverage relation holds true regardless of the
level of profitability of the firm. If not so, it would mean the
profitability-leverage relation is determined by the level of profits
firms generate.
3.3. Panel Data Analysis
Since we use data on both cross-sectional and time-series
dimensions, we employ different variants of panel data models for
analysis. One might argue that many of the models are in difference form
and hence fixed effects might not be an issue, still for comparison
purpose we report results from random effects and fixed effects models.
For formal selection between these two models, we employed the Hausman
(1978) specification test. We also estimated pooled OLS for all models,
but we do not report results as results from pooled OLS and random
effects models were virtually similar.
3.4. Robustness Checks
There is an anecdotal evidence of weak corporate governance in
Pakistan. For example, there is large scale tax evasion, firms are
closely held and controlled, banks rather than markets dominate
corporate finance, and accounting statements may not reflect the true
state of affairs. In such an environment while it is quite legitimate to
study financing patterns and behaviour of corporations, simple tests of
the theories are not likely to be productive exercises.
In view of the above, we conduct our analysis on full sample of
firms and use several robustness checks to see whether results change
substantially once we account for weak corporate governance or potential
profit understatements in Pakistani firms. For this purpose, we have
borrowed corporate governance compliance index (CGCI) for 102 firms from
Tariq and Abbas (2013). They have developed this index to measure the
extent to which companies follow the Code of Corporate Governance of the
Securities and Exchange Commission of Pakistan. We divided the 102
companies into two groups based on the median value of CGCI. These
groups are named as 'High CG and 'Low CG. Our interest lies in
the comparison of the results from different models estimated on the
data of all firms, Higher CG firms, and Lower CG firms. We want to see
whether our results are driven by weak or good corporate governance
practices.
Our second robustness check to control for the understatement of
profits problem is to estimate Rajan and Zingles (1995) model (see
Section 3.2.5) for three groups of firms which are formed on the basis
of 25th, 50th, and 75th, percentiles of firms' profitability. For
each group of firms, we estimate the model and want to see whether the
key variables change their signs or significances. If profits
understatements drive the results, then the coefficients of the
explanatory variables will behave randomly across these groups of firms.
In unreported work, we also estimated all models for high- and
low-profit groups of firms (low group included firms in 25th percentile
of profitability, and high group included firms in the 75th percentile
of profitability) to compare results across different profitability
groups. The results of these regressions show that profits level do
change the basic results. To save space, we do not report these results,
however, they can be supplied upon request from the authors.
4. ANALYSIS AND FINDINGS
In this section, we present and discuss results of the empirical
models. Since we are dealing with panel data, we have to choose between
fixed and random effects models. Almost in all models, the Hausman
(1978) specification test indicated to use fixed effects model. However,
for comparison purpose, we report results from both fixed and random
effects models. In all of the regression tables from Table 4.1 to 4.4,
first column shows names of the variables, second and third columns
report results from the random and fixed effects models, respectively.
As discussed in Section 3.4, we also report results from a subset of
firms for which corporate governance compliance data were obtained from
Tariq and Zaheer (2013). A total of 102 firms are included in this
sub-sample. Fourth column of the regression tables report results for
all these 102 firms whereas fifth and sixth columns report results for
two groups of firms which are divided according to the median value of
the compliance index. Firms with high corporate governance compliance
index are named as 'High CG' whereas firms with low compliance
index are named as 'Low CG'.
4.1. Results from SSM Model
Results of Shayam-Saunders and Myers (SSM) model are given in Table
1. Standard errors are reported in parentheses, whereas ***, **, and *
show significance at 1 percent, 5 percent, and 10 percent, respectively.
Under the pecking order theory, SSM argue that the slope
coefficient of funding deficit must be equal to one. A firm uses debt
funding when their funding needs exceed the internally available funds.
Every dollar increase in funding deficit after utilisation of internally
available funds will be met by a dollar of debt financing. Whereas
equity funding is used only as a last resort and is relatively rare. The
results in Table 4.1 show that the slope coefficient of funding deficit
is 0.085 and [R.sup.2] has a value 0.0136 in fixed the effects model.
These findings show only a weak support for the pecking order theory.
The positive coefficient of funding deficit is in accordance with
prediction of SSM model, but the coefficient value of funding deficit is
too low against its expected value that should be near 1. In fact the
coefficient of funding deficit was reported 0.76 by Shayam-Saunders and
Myers (1999) and 0.75 by Frank and Goyal (2003) when they used a sample
of firms that had no gaps in the data. However, when they relaxed the
continuous data restriction and estimated the equation on full sample,
the funding deficit coefficient declined to 0.28. Further, they noted
that with the passage of time, SSM model showed declining support for
the pecking order theory. Another important fact is that the pecking
order theory predicts the preference of internal source of funding over
debt financing. But our results show that intercept has significant
positive value. This fact is against the pecking order theory.
Therefore, the study rejects the hypothesis that internally available
retained earnings are preferred over debt financing.
Comparing results in 'High CG' and 'Low CG'
groups, we observe that results in these two groups are not different.
In fact, financial deficit seems to have no influence on debt ratio in
the full sample of 102 firms for which corporate governance data is
available or in the high or low compliant groups. This shows that
corporate governance practices do not alter the results.
4.1.2. Results of Frank and Goyal Disaggregation Model
Second model used in this study is the model of Frank and Goyal
(2003) who proposed to disaggregate funding deficit factor in the SSM
model. Table 4.2 presents results of Frank and Goyal disaggregation
model. Standard errors are reported in parentheses, whereas ***, **, and
* show significance at 1 percent, 5 percent, and 10 percent,
respectively.
Results in Table 4.2 suggest that the coefficients of each
component of the funding deficit are significantly different from one.
Under the pecking order theory, one unit change in any component of the
funding deficit should lead to a unit change in debt level. The results
do not support this prediction of the pecking order theory as proposed
in SSM model. Results show that aggregation of the components of funding
deficit term is less informative. Since aggregation of funding deficit
is not justified, study of the individual components can reveal more
information.
Further, we find that the coefficient of dividends (DIV) is
positive and is statistically significant only in the random effects
model. The positive coefficient implies that dividend paying firms use
more debt financing. Since its coefficient is marginally significant, it
supports the pecking order theory to some extent. The coefficient value
of capital expenditure (X) ranges from 0.51 (in the fixed effect
regression) to 0.65 (in the lower CG regression), and is positively
related to debt financing. The positive relation between capital
expenditure and debt financing is in accordance with both the pecking
order theory and the trade-off theory. Under the pecking order theory,
once internal funds are employed, increase in capital expenditure will
increase funding deficit of the firm which will in turn increase debt
financing. Under the trade-off theory, capital expenditures create
tangible assets which can be used by the firm as collateral against debt
financing.
Internally available operating cash flows (C) show a negative
relation with changes in debt in all models. This finding is in line
with the financing behaviour pattern laid down in the pecking order
theory. However, if one considers profitability as a proxy of future
growth opportunities, the trade-off theory would then also predict a
negative relationship between profitability and debt financing. Working
capital (AW) shows negative relation with the changes in debt. Pecking
order theory predicts that after we control for internally generated
funds, working capital needs should be financed dollar for dollar from
debt financing. Thus, pecking order theory fails here.
In conclusion, we find only weak support for the pecking order
theory using the Frank and Goyal model. This is evident not only from
fairly small coefficients of the individual components of funding
deficit, but also some of the components of the funding deficit yielded
unexpected signs.
Results for firms that have corporate governance data are reported
under column headings (3), (4) and (5) in Table 4.2. It is interesting
to see that the results of the sub-sample are almost similar in
statistical significance and coefficient signs as the full sample.
Further, there is significant difference in the results of firms that
score high on corporate governance compliance index (High CG) and firms
that score low on this index (Low CG).
4.1.3. Results of Frank and Goyal Conventional Leverage Model
In order to avoid omitted variable bias and to know the
contribution made by each funding deficit variable, we follow the work
of Frank and Goyal (2003) to modify Equation 1 into Equation 3 by adding
previously identified explanatory variables. Table 4.3 presents results
of Frank and Goyal model using conventional leverage regression.
Result in Table 4.4 shows that tangibility is negatively related to
changes in debt levels of the Pakistani firms. Negative sign of the
coefficient of tangibility is in accordance with the pecking order
theory as highlighted by Harris and Raviv (1990). Rationale of this
negative relation is that firms with low tangibility have more
information asymmetry problems. This makes equity financing more
expensive for them, which in turn makes debt financing attractive after
utilisation of internal finds. However, it is noted that this negative
relation is in contrast to the findings of Hijazi and Shah (2004) and
Ilyas (2008) who also studied the factors determining the leverage of
Pakistani firms.
As expected under pecking order theory, slope coefficient of growth
variable has a positive sign and is statistically significant. The
observed relation between growth and change in debt level shows that
growing Pakistani firms funding requirements exceed the internally
available funds. Thus, these firms go for debt financing [Drobets and
Fix (2003)]. Hence such firms behave in a pecking order. This finding is
similar to the finding of Hijazi and Shah (2004).
The variable LS (a proxy for firm size) did not show the predicted
sign under the pecking order theory. Its coefficient is positive and
significant. Under the pecking order theory, smaller firms tend to use
more debt financing as they have more asymmetric information problems
[Berger and Udell (1995), Rajan and Zingales (1995), Frank and Goyal
(2003)]. In contrast, the trade-off theory predicts a positive relation
between firm size and debt financing as larger firms have more assets.
Larger size increases the firm's ability to obtain more debt.
Similarly, if size is taken as inverse proxy of probability of
bankruptcy then larger size firms have a lower probability of bankruptcy
that allows them to use more debt financing [Rajan and Zingales (1995)].
Negative slope of profitability (AP) is in accordance with the
pecking order theory but in contrast to the trade-off theory. This shows
that Pakistani firms employ their internal funds generated by profits
before debt financing. Another possible explanation for this negative
relation is that Pakistani firms may use profits to pay their debts.
This negative relation was also found by Hijasiand Shah (2004) and Ilyas
(2008).
The funding deficit variable (DEF) showed predicted relation with
changes in debt i.e. it is positive and significant. However, its
coefficient remains very low. Positive slope of funding deficit shows
that with increasing funding deficit, internally available funds become
inadequate and hence firms choose debt financing.
Overall the coefficients of the explanatory variables show
predicted signs under the pecking order theory except size of the firm.
Importantly, funding deficit explained less of the variation in debt
level of the sample firms in the presence of other variables.
Profitability and growth seem to be important determinants of debt level
of Pakistani firms. Thus, this study rejects the hypothesis that funding
deficit contributes more as a determinant of leverage as compared to
other conventional determinants of leverage.
Comparing results of firms that are grouped on the basis of lower
and higher compliance with code of corporate governance of SECP, one can
see not much of a difference. Majority of the variables have their
statistical significances and coefficient signs similar in both the
groups, with the exception of size, which is marginally significant in
'Low CG' group and insignificant in the 'High CG'
group.
4.1.4. Results of Watson and Wilson Model
We use Watson and Wilson (2002) model to investigate a firm's
choice between debt and equity funding once internal funds are utilised.
Table 4.4 presents the results of the Watson and Wilson model.
Results from the fixed effects model show that the coefficient of
profitability has value of 0.4610 which is greater than the coefficient
of equity financing which has a value of 0.4151 but less than the
coefficient of debt financing having value of 1.0032. These estimates
are significant at 1 percent level of significance. Under the pecking
order theory, the coefficients of debt must be greater than equity
funding but lesser than internally available funds. The pecking order
theory suggests that debt financing is utilised before equity financing
which is used only in extreme circumstances when funding needs exceed
internally available funds. Contrary to the prediction of pecking order
theory, results of Watson and Wilson model suggest that external debt
financing is preferred over other sources of funding. Second preference
is given to internal source of financing i.e. profits and as a last
resort equity financing is picked by Pakistani firms at times of funding
deficit. Results lead to rejection of the hypothesis that retained
earnings are preferred over debt financing but accept that debt
financing is preferred over equity financing. Thus, the results of
Watson and Wilson model show only partial support for the pecking order
theory in case of Pakistani firms.
4.1.5. Results of Rajan and Zingales Model
In view of poor corporate governance practices which might lead to
understatement of profits to avoid taxes or expropriate minority
shareholders in Pakistan [see, e.g., Abdullah, el al. (2012)], we are
concerned that our results might be contaminated by reported earning
understatements. As a robustness check, we categorise firms into three
groups based on 25th, 50th, and 75th percentiles of the firms'
profitability to see whether our results behave randomly across
different reported profitability levels. These groups are named as low
profit, medium profit, and high profit firms. Then we estimate Rajan and
Zingales (1995) conventional leverage regression for each group
separately. Table 4.5 presents results of Rajan and Zingales model.
Under the pecking order theory, firms having low profits and
funding requirements will consider debt financing before they consider
equity financing. Whereas firms with high profits will cover funding
requirements with internal funds i.e. retained earnings. However,
average profitable firms will have moderate external financing mostly
from debt financing. So in each case, profitability of the firm must be
negatively related to debt financing of the firm. Low profitable firms
have highest slope coefficient value of -2.1296 for P. Then, average
profitable firms have slope coefficient value of-1.8278. High profitable
firms have lowest slope coefficient value of-0.5837. Results of Rajan
and Zingales model show that for each level of profitability of firms,
profitability is negatively related to debt financing. This negative
relationship between profitability of the firm and its leverage is
statistically significant in each category of profitability of the
Pakistani firms. Thus, we can conclude that level of profitability of
Pakistani firms does not affect negative profitability-leverage
relationship. This finding is in line with the previous studies on
capital structure in Pakistan [see, e.g., Shah and Hijazi (2004), Shah
and Khan (2007)].
5. CONCLUSION
Previous studies in Pakistan on corporate capital structure used
only a single dimension to test pecking order theory where they
presented negative profitability-leverage relationship as an evidence in
support of the pecking order theory. The objective of this study was to
test whether or not empirical support exists for the pecking order
theory in Pakistani firms when we employ a wide range of models that use
different assumptions and hence employ different econometric techniques.
For this purpose, we used financial data of 321 non-financial Pakistani
firms listed on the KSE over the period 2000-2009. Results of
Shayam-Sunder and Myers model showed that funding deficit is not matched
dollar for dollar by changes in debt financing. However, results showed
that there was a positive relationship between funding deficits and debt
levels of the sampled firms. Moreover, SSM model yielded positive
intercept term which is expected to have a zero value under the pecking
order theory. Positive intercept term means that internal funds were not
preferred over other sources of financing at times of funding deficit.
Our conclusion based on the above findings is that funding deficit has
less explanatory power in determining the debt level of the Pakistani
firms.
We also used the disaggregated model of Frank and Goyal (2003).
Results revealed that the aggregation of funding deficit term is not
justified. When studied individually, all of the components of funding
deficit showed expected signs with change in debt level of firms, except
changes in working capital. Capital expenditure showed statistically
significant and positive relationship with changes in debt level of the
firms. Whereas, operating cash flows and changes in working capital
showed negative relationships with changes in debt ratios. The negative
sign for working capital is in contradiction to the pecking order
theory. Dividend payout showed insignificant negative relation with
leverage, which is also contrary to the pecking order theory. Overall
results from this model were mixed. Thirdly, we tested the impact of
funding deficit in the presence of other determinants such as
tangibility, size, growth and profitability of the firms on their debt
ratios. We found that the contribution of funding deficit was negligible
in explaining changes in debt ratios in the presence of other variables.
Profitability and growth seem to be the most important determinants of
changes in debt levels of Pakistani firms. Profitability was negatively
related to debt ratio and was statistically significant. Firm size
showed positive relation with debt ratio which indicates that larger
firms can take more debt. This finding is also contrary to the
prediction of the pecking order theory. Coefficient signs and
significances of tangibility, growth, and profitability variables
support the pecking order theory. Lastly we found that level of
profitability of Pakistani firms does not affect the negative
profitability-leverage relation. Whether the firms are less profitable
or more profitable, a consistent negative relationship between
profitability and leverage was observed. As a further robustness check,
we used data on compliance with the SECP code of corporate governance to
see whether our results behave differently among firms that show high
and low compliance with the code. We found that our results do not
change with the level of compliance.
Overall, we found very weak evidence in support of the pecking
order theory using funding deficit regressions. However, strong support
is found for pecking order theory when leverage ratios are regressed on
profitability ratio, along with a set of control variables. This
discrepancy in the results of the two sets of models needs further
investigation, as well as care in interpreting the results of existing
studies on capital structure in Pakistan.
REFERENCES
Abdullah, F., A. Shah, and S. U. Khan (2012) Firm Performance and
the Nature of Agency Problems in Insiders-controlled Firms: Evidence
from Pakistan. The Pakistan Development Review 51:4, 161-183.
Allen, D. (1993) The Pecking Order Hypothesis: Australian Evidence.
Applied Financial Economics 3, 101-120.
Autore, D. and T. Kovacs (2004) The Pecking Order Theory and
Time-Varying Adverse Selection Costs. Virginia Tech University. (Working
Paper).
Barclay, M. and C. W. Smith (2005) The Capital Structure Puzzle:
Another Look at the Evidence. Journal of Applied Corporate Finance 12:1,
8-20.
Baskin, J. (1989) An Empirical Investigation of the Pecking Order
Theory. Financial Management 18, 26-35.
Bennet, M. and R. Donnelly (1993) The Determinants of Capital
Structure: Some U.K. Evidence. British Accounting Review 25:1, 43-59.
Berger, A. N. and G. Udells (1995) Relationship Lending and Lines
of Credit in Small Firm Finance. Journal of Business 68, 351-381.
Bharath, S. T., P. Pasquariello, and G. Wu (2009) Does Asymmetric
Information Drive Capital Structure Decisions. Review of Financial
Studies 22:8, 321 1-3243.
Booth, L., V. Aivazian, A. Demirguc-Kunt, and V. Maksimovic (2001)
Capital Structures in Developing Countries. The Journal of Finance
LVI:1, 87-130.
Bradley, M., G. A. Jarrell, and E. H. Kim (1984) On the Existence
of an Optimal Capital Structure: Theory and Evidence. Journal of Finance
39.
Brealey, R. A., S. C. Myers, and A. L. Marcus (1999) Fundamentals
of Corporate Finance (2 ed.). McGraw Hill.
Brennan, M. and A. Kraus (1987) Efficient Financing under
Asymmetric Information. Journal of Finance 42, 1225-1243.
Byoun, S. and J. C. Rhim (2005) Tests of the Pecking Order Theory
and the Tradeoff Theory of Optimal Capital Structure. Global Business
Finance Review 10:2, 1-16.
Chang, H. L., H. Y. Liang, C. W. Su, and M. N. Zhu (2010) The
Choices of Capital Structure. African Journal of Business Management
4:15, 3332-3336.
Cheng, S. and C. Shiu (2007) Investor Protection and Capital
Structure: International Evidence. Journal of Multinational Financial
Management 17, 30-44.
Chirinko, R. S., and A. R. Singha (2000) Testing Static Tradeoff
Against Pecking Order Models of Capital Structure: A Critical Comment.
Journal of Financial Economics 58:3, 417-425.
Dang, V. A. (2005) Testing the Trade-off and Pecking Order Theory:
Some UK Evidence. Paper presented at the Money Macro and Finance (MMF)
Research Group Conference.
Donaldson, G. (1961) Corporate Debt Capacity: A Study of Corporate
Debt Policy and the Determination of Corporate Debt Capacity. Paper
presented at the Division of Research, Harvard Graduate School of
Business Administration.
Drobetz, W. and R. Fix (2003) What are the Determinants of the
Capital Structure? Some Evidence for Switzerland. University of Basel.
WWZ / Department of Finance. (Working Paper No. 4/03).
Fama, E. F. and K. R. French (2005) Financing Decisions: Who Issues
Stock? Journal of Financial Economics 76, 549-582.
Fama, E. F. and R. French Kenneth (2002) Testing Tradeoff and
Pecking Order Predictions about Dividends and Debt. Review of Financial
Studies 15, 1-33.
Flannery, M. J. and K. P. Rangan (2006) Partial Adjustment Target
Capital Structures. Journal of Financial Economics 79, 469-506.
Frank, M. Z. and V. K. Goyal (2003) Testing of Pecking Order Theory
of Capital Structure. Journal of Financial Economics 67, 217-248.
Frank, M. Z. and V. K. Goyal (2007) Capital Structure Decisions:
Which Factors are Reliably Important? University of Minnesota and HKUST.
(Working Paper).
Frank, M. Z. and V. K. Goyal (2009) Profits and Capital Structure.
University of Minnesota and Hong Kong University of Science and
Technology. (Working Paper).
Friend, I. and L. H. P. Lang (1988) An Empirical Test of the Impact
of Managerial Self-interest on Corporate Capital Structure. Journal of
Finance 43, 271-281.
Gaud, P., E. Jani, M. Hoesli, and A. Bender (2005) The Capital
Structure of Swiss Companies: An Empirical Analysis using Dynamic Panel
Data. European Financial Management 11, 51-69.
Ghosh, A. and F. Cai (2004) Optimal Capital Structure Vs. Pecking
Order Theory: A Further Test. Journal of Business and Economics Research
2:8.
Gomes, A. R. and G. M. P. (2007) Private and Public Security
Issuance by Public Firms: The Role of Asymmetric Information. Washington
University in St. Louis and University of Maryland. (Working Paper).
Harris, M. and A. Raviv(1990) Capital Structure and the
Informational Role of Debt. The Journal of Finance 45:2, 321-349.
Hausman, J. A. (1978) Specification Tests in Econometrics.
Econometrica: Journal of the Econometric Society, 1251-1271.
Hijazi, S. T. and A. Shah (2004) The Determinants of Capital
Structure in Stock Exchange Listed Non-Financial Firms in Pakistan. The
Pakistan Development Review 43:4, 605-618.
Hovakimian, A. and M. Vulanovic (2008) Corporate Financing of
Maturing Long-Term Debt. Paper presented at the Asian Finance
Association.
Hovakimian, A., G. Hovakimian, and H. Tehranian (2004) Determinants
of Target Capital Structure: The Case of Dual Debt and Equity Issues.
Journal of Financial Economics 71, 517-540.
Hovakimian, A., T. Opler, and S. Titman (2001) The Debt-equity
Choice. Journal of Financial and Quantitative Analysis 36, 1-24.
Huang, R. and J. R. Ritter (2007) Testing Theories of Capital
Structure and Estimating the Speed of Adjustment. Journal of Financial
and Quantitative Analysis. (Forthcoming).
Ilyas, J. (2008) The Determinants of Capital Structure: Analysis of
Non-Financial Firms Listed in Karachi Stock Exchange in Pakistan.
Journal of Managerial Sciences 2:3, 275-304.
Jabeen, M. and A. Shah (2011) A Review on Family Ownership and
Information Asymmetry. African Journal of Business Management 5:35,
13550-13558.
Jalilvand, A. and R. S. Harris (1984) Corporate Behaviour in
Adjusting to Capital Structure and Dividend Targets: An Econometric
Study. The Journal of Finance 39:1, 127-145.
Jensen, M. C. and W. H. Meckling (1976) Theory of the Firm:
Managerial Behaviour, Agency Costs and Ownership Structure. Journal of
Financial Economics 3:4, 305-360.
Jung, K., Y. Kim, and R. Stulz (1996) Timing, Investment
Opportunities, Managerial Discretion, and the Security Issue Decision.
Journal of Financial Economics 42, 159-185.
Kester, C. W. (1986) Capital and Ownership Structure: A Comparison
of United States and Japanese Manufacturing Corporations. Financial
Management 15, 97-113.
Komera, S. and P. J. J. Lukose (2014) Capital Structure Choice,
Information Asymmetry, and Debt Capacity: Evidence from India. Journal
of Economics and Finance, Article in Press.
Kraus, A. and R. H. Litzenberger (1973) A State-Preference Model of
Optimal Financial Leverage. Journal of Finance, 911-922.
Kremp, E. E. S. and D. Gerdesmeier (1999) Estimation of a Debt
Function: Evidence from French and German Firm Panel Data. In A. Sauve
and M. Scheuer (eds.) Corporate Finance in Germany and France. A Joint
Research Project of Deutsche Bundesbank and the Banque de France. (SSRN
Working Paper).
La Porta, R., F. Lopez-de-Salianes, A. Shleifer, and R. W. Vishny
(2002) Investors Protection and Corporate Valuation. Journal of Finance
57, 1147-1170.
Leary, M. T. and M. R. Roberts (2007) The Pecking Order, Debt
Capacity, and Information Asymmetry. University of Pennsylvania.
(Working Paper).
Lemmon, M. L. and J. F. Zender (2004) Debt Capacity and Tests of
Capital Structure Theories. (Working Paper).
Marsh, P. (1982) The Choice between Debt and Equity: An Empirical
Study. Journal of Finance 37, 121-144.
Michaelas, N., F. Chittenden, and P. Poutziouris (1999) Financial
Policy and Capital Structure Choice in UK SMEs: Empirical Evidence from
Company Panel Data. Small Business Economics 12, 113-130.
Mira, F. S. and J. L. Gracia (2003) Pecking Order Versus Trade Off:
An Empirical Approach to the Small and Medium Enterprise Capital
Structure. Instituto Valenciano de Investigaciones Economicas, 1-36.
Modigliani, F. and M. Miller (1958) The Cost of Capital,
Corporation Finance and the Theory of Investment. American Economic
Review 48, 261-297.
Myers, S. C. (1984) Capital Structure Puzzle. 1548-1584. (NBER
Working Paper Series, wl393).
Myers, S. C. (2001) Capital Structure. The Journal of Economic
Perspectives 15, 81102.
Myers, S. C. and N. S. Majluf (1984) Corporate Financing and
Investment Decisions when Firms Have Information that Investors Do Not
Have. Journal of Financial Economics 13, 187-221.
Nadeem, S. and Z. Wang (2010) Financing Behaviour of Textile Firms
in Pakistan. International Journal of Innovation, Management and
Technology 1:2.
Ozkan, A. (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.
Qureshi, M. A. (2009) Does Pecking Order Theory Explain Leverage
Behaviour in Pakistan? Applied Financial Economics 19:17, 1365-1370.
Qureshi, M. A. (2010) Testing Pecking Order Theory and Trade-off
Theory-A System Dynamics Approach. Paper presented at the 28th
International Conference of the System Dynamics Society, Seoul, Korea.
Rajan, R. G. and L. Zingales (1995) What Do We Know about Capital
Structure? Some Evidence from International Data. Journal of Finance 50,
1421-1460.
Saunders, M., P. Lewis, and A. Thornhill (2003) Research Methods
for Business Students. Financial Times Press.
Seifert, B. and H. Gonenc (2008a) The International Evidence on the
Pecking Order Hypothesis. Journal of Multinational Financial Management
18:3, 244-266.
Seifert, B. and H. Gonenc (2008b) Pecking Order Behaviour in
Emerging Markets. Old Dominion University, United States. (Working
Paper).
Serrasqueiro, Z. and P. M. Nunes (2010) Are Trade-off and Pecking
Order Theories Mutually Exclusive in Explaining Capital Structure
Decisions? African Journal of Business Management 11, 2216-2230.
Shah, A. and S. Khan (2007) Determinants of Capital Structure:
Evidence from Pakistani Panel Data. International Review of Business
Research Papers 3:4, 265-282.
Shyam-Sunder, L. and S. C. Myers (1999) Testing Static Tradeoff
Against Pecking Order Models of Capital Structure. Journal of Financial
Economics 51, 219-244.
Sinan, A. (2010) How Do Firm Characteristics Affects Capital
Structure? Some UK Evidence. (MPRA Paper No. 29657).
Strebulaev, 1. A. (2007) Do Tests of Capital Structure Theory Mean
What They Say? Journal of Finance 62, 1747-1788.
Swinnen, S., W. Voordeckers, and S. Vandemaele (2005) Capital
Structure in SMEs: Pecking Order versus Static Trade-Off, Bounded
Rationality and the Behavioural Principle. Paper presented at the
European Financial Management Association.
Tariq, Y. Bin and Z. Abbas (2013) Compliance and Multidimensional
Firm Performance: Evaluating the Efficacy of Rule-based Code of
Corporate Governance. Economic Modelling 35, 565-575.
Titman, S. and R. Wessels (1988) The Determinants of Capital
Structure Choice. Journal of Finance 43, 1-19.
Tong, G. and C. J. Green (2005) Pecking Order or Trade-off
Hypothesis? Evidence on the Capital Structure of Chinese Companies.
Applied Economics 37:19, 2179-2189.
Ullah, F., Q. Abbas, and S. Akbar (2009) The Relevance of Pecking
Order Hypothesis for the Financing of Computer Software and
Biotechnology Small Firms: Some UK Evidence. International
Entrepreneurship and Management Journal 6:3, 301-315.
Vasiliou, D., N. Eriotis, and N. Daskalakis (2009) Testing the
Pecking Order Theory: The Importance of Methodology. Qualitative
Research in Financial Markets 1:2, 85-96.
Vilasuso, J. and A. Minkler (2001) Agency Costs, Asset Specificity,
and the Capital Structure of the Firm. Journal of Economic Behaviour and
Organisation 44, 55-69.
Watson, R. and N. Wilson (2002) Small and Medium Size Enterprise
Financing: A Note on Some of the Empirical Implications of a Pecking
Order. Journal of Business Finance and Accounting 29, 557-578.
Yin, R. K. (2003) Case Study Research, Design and Methods.
California: Sega Publications.
Attaullah Shah <attaullah.shah@imsciences.edu.pk> is a
Faculty Member at the Institute of Management Sciences, Peshawar. Jasir
Ilyas <jasirilyas@yahoo,com> is Faculty Member at City University
of Sciences and Information Technology, Peshawar.
Table 4.1
Results of the SSM Model
[DELTA][D.sub.it] = [alpha] + [beta] [DEF.sub.it] +
[e.sub.it] ... ... ... (1)
VARIABLES (1) (2) (3) (4)
Random Eff. Fixed Eff. CG All High CO
DEF 0.0319 *** 0.0857 *** -0.0045 -0.0005
(0.0085) (0.0155) (0.0163) (0.0262)
Constant 0.0559 *** 0.0582 *** 0.0557 *** 0.0528 ***
(0.0029) (0.0029) (0.0046) (0.0068)
Observations 2.534 2,534 762 383
R-squared 0.0136 0.0001 0.0000
Number of firms 321 321 102 56
VARIABLES (5)
Low CG
DEF -0.0046
(0.0215)
Constant 0.0582 ***
(0.0066)
Observations 379
R-squared 0.0001
Number of firms 56
Table 1: Change in debt level's led by total assets is the
dependent variable. Whereas funding deficit variable (DEF) is
measured as a sum of dividend payment, capital expenditure, net
increase in working capital minus operating cash flow after interest
and tax. Both variables are scaled by total assets. Standard errors
are reported in parentheses, whereas ***, **, and * show
significance at 1 percent, 5 percent, and 10 percent, respectively.
Table 4.2
Results of the Frank and Goyal Disaggregation Model
[DELTA][D.sub.it] = [alpha] + [[beta].sub.Div][DIV.sub.it] +
[[beta].sub.X][X.sub.it] + [[beta].sub.W][DELTA][W.sub.it] +
- [[beta].sub.C][C.sub.it] + [e.sub.it] ... ... ... (2)
Variables (1) (2) (3) (4)
Random Eff. Fixed Eff. CG All High CG
DIV 0.0001 ** 0.0000 0.0000 0.0000
(0.0000) (0.0000) (0.0000) (0.0000)
[DELTA]W -0.0181 ** -0.0900 *** -0.0338 ** -0.0333
-0.0089 (0.0185) (0.0158) (0.0259)
X 0.5661 *** 0.5141 *** 0.6292 *** 0.5884 ***
(0.0271) (0.0300) (0.0497) (0.0735)
c -0.1811 *** -0.2905 *** -0.1801 *** -0.1490 **
(0.0282) (0.0387) (0.0505) (0.0714)
Constant 0.0437 *** 0.0523 *** 0.0492 *** 0.0479 ***
(0.0033) (0.0041) (0.0061) (0.0096)
Observations 2,534 2,534 762 383
R-squared 0.1349 0.1493 0.1966 0.1637
No. of Firms 321 321 102 56
Variables (5)
Low CG
DIV -0.0000
(0.0001)
[DELTA]W -0.0297
(0.0205)
X 0.6581 ***
(0.0679)
c -0.2129 ***
(0.0743)
Constant 0.0519 ***
(0.0081)
Observations 379
R-squared 0.2314
No. of Firms 56
Table 4.2: Change in debt level is dependent variable of the model
and is measured as the difference of total liabilities at time t
with respect to time [t.sub.-1], scaled by total assets Independent
variables are dividend payment (DIV), capital expenditure (X), net
increase in working capital ([DELTA]W) and operating cash flow (C).
All these variables are scaled by total assets. Dividend payment is
the amount of dividend paid by the firm for the year, capital
expenditure is change in net fixed assets with respect to the
previous year, working capital is the difference between current
asset and current liabilities of the firm; and the operating cash
flow after interest and tax is equal to net income plus
depreciation for the year. Standard errors are reported in
parentheses, whereas ***, **, and * show significance at 1 percent,
5 percent, and 10 percent, respectively.
Table 4.3
Results of the Frank and Goyal Model Using Conventional Leverage
Regression [DELTA][D.sub.it] = [alpha] +
[[beta].sub.T][DELTA][T.sub.it] + [[beta].sub.G][G.sub.I] +
[[beta].sub.LS][DELTA][LS.sub.it] + [[beta].sub.I],
[DELTA][P.sub.it] + [[beta].sub.DEF][DEF.sub.it] + [e.sub.it]
Variables (1) (2) (3) (4)
Random Eff. Fixed Eff. CG All High CG
AT -0.0533 *** -0.2828 *** -0.0452 * -0.0577
(0.0154) (0.0355) (0.0244) (0.0362)
G 0.3416 *** .2581 *** 0.3520 *** 0.2843 ***
(0.0307) (.03145) (0.0500) (0.0732)
ALS 0.0081 *** 0.0347 *** 0.0053 0.0018
(0.0025) (0.0066) (0.0038) (0.0055)
AP -0.3476 *** -0.3568 *** -0.2509 *** -0.1797 ***
(0.0335) (0.0453) (0.0515) (0.0682)
DEF 0.0192 * 0.0747 *** -0.0071 -0.0137
(0.0098) (0.0151) (0.0193) (0.0295)
Constant -0.0056 -0.0323 0.0007 0.0340
(0.0183) (0.0498) (0.0298) (0.0465)
Observations 2,534 2,534 762 383
R-squared .0832 0.0740 0.0785 0.0533
Number of Firms 321 321 102 56
Variables (5)
Low CG
AT -0.0373
(0.0338)
G 0.4291 ***
(0.0709)
ALS 0.0098*
(0.0055)
AP -0.3584 ***
(0.0856)
DEF 0.0054
(0.0264)
Constant -0.0390
(0.0410)
Observations 379
R-squared 0.1156
Number of Firms 56
Table 4.3: Change in debt level is dependent variable of the model
and is measured as the difference of total liabilities at time l
with respect to time [t.sub.-1] scaled by total assets. Independent
variables are the tangibility, growth, size, profitability and
funding deficit of the firm. These variables are denoted as
[DELTA]T, G, [DELTA]LS, [DELTA]P, and DEF respectively. This study
measures tangibility as a ratio of change in fixed assets to total
assets with respect to the previous year, growth as a geometric
mean of percentage change in total asset of the firm with respect
to the previous year, size as the change in natural logarithm of
total assets with respect to the previous year, profitability as
change in ratio of net income to total assets of the firm with
respect to the previous year and funding deficit variable as a sum
of dividend payment, capital expenditure, net increase in working
capital minus operating cash flow after interest and tax. Standard
errors are reported in parentheses, whereas ***, **, and * show
significance at 1 percent, 5 percent, and 10 percent, respectively.
Table 4.4
Results of the Watson and Wilson Model
([TA.sub.it]-[TA.sub.it-1])/[TA.sub.it-1] =
[summation][[alpha].sub.i] = [[beta].sub.1] ([P.sub.it] -
[Div.sub.it])/[TA.sub.it-1] + [[beta].sub.2]
([EI.sub.it])/[TA.sub.it-1] + [[beta].sub.3] ([D.sub.it] -
[D.sub.it-1])/[TA.sub.it-1] + [v.sub.it].... (4)
Variables (1) (2) (3) (4)
Random Eff. Fixed Eff. CG All High CG
([P.sub.it] - 0.4014 *** 0.4610 *** 0.2968 *** 0.1904 ***
[Div.sub.it])/
[TA.sub.it-1]
(0.0364) (0.0424) (0.0559) (0.0670)
([EI.sub.it])/ 0.1492 *** 0.4151 *** 0.1226 *** 0.1543 ***
[TA.sub.it-1]
(0.0101) (0.0155) (0.0172) (0.0233)
([D.sub.it]- 1.0286 *** 1.0032 *** 1.1265 *** 1.0299 ***
[D.sub.it-1])/
[TA.sub.it-1]
(0.0190) (0.0183) (0.0334) (0.0433)
Constant -0.0011 -0.1009 *** -0.0040 -0.0104
(0.0047) (0.0061) (0.0081) (0.0116)
Observations 2,534 2,534 762 383
R-squared .5999 0.6476 0.6266 0.6331
Number of firms 321 321 102 56
Variables (5)
Low CG
([P.sub.it] - 0.4958 ***
[Div.sub.it])/
[TA.sub.it-1]
(0.1011)
([EI.sub.it])/ 0.0893 ***
[TA.sub.it-1]
(0.0253)
([D.sub.it]- 1.2154 ***
[D.sub.it-1])/
[TA.sub.it-1]
(0.0501)
Constant 0.0034
(0.0115)
Observations 379
R-squared 0.6332
Number of firms 56
Table 4.4: Dependent variable of the model is changes in total
asset at time l with respect to time [t.sub.-1] measured as a
proportion of total assets and denoted as
([TA.sub.it]-[TA.sub.it-1])/[TA.sub.it-1]. Independent variables
include internally available funds (([P.sub.it] -
[Div.sub.it])/[TA.sub.it-1]), equity funding
(([EI.sub.it])/[TA.sub.it,1]) and debt funding (([D.sub.it]-
[D.sub.it-1])/[TA.sub.it-1]). Internally available fund is measured
as a reminder of profits after paying dividends. Equity funding is
measured as shareholders equity and debt funding as change in total
liabilities. All of the variables are calculated as a proportion of
total assets. Standard errors are reported in parentheses, whereas
***, **, and * show significance at I percent, 5 percent, and 10
percent, respectively.
Table 4.5
Results of Rajan and Zingales Model
[D.sub.it] = [[alpha].sub.1] + [[beta].sub.1]([T.sub.it]) +
[[beta].sub.2][G.sub.it] + [[beta].sub.3][LS.sub.it] +
[[beta].sub.4] [P.sub.it] + [e.sub.it]
Variables (1) (2) (1)
Low Profits Average Profits High Profits
T 0.2685 *** 0.0408 0.0808 *
(0.0694) (0.0560) (0.0440)
LS -0.0388 ** -0.0124 0.0282 ***
(0.0168) (0.0126) (0.0100)
G -0.7248 *** -0.6341 *** 0.0005
(0.1587) (0.1227) (0.0929)
P -2.1296 *** -1.8278 *** -0.5837 ***
(0.2214) (0.3991) (0.1269)
Constant 0.8777 *** 0.8380 *** 0.2834 ***
(0.1565) (0.1064) (0.1062)
Observations 633 633 633
R-squared 0.3131 0.1700 0.1737
Table 4.5: Debt ([D.sub.it]) is the dependent variable of the model
and is measured as the ratio of total liabilities at time t of i,
scaled by total assets. Independent variables include tangibility
(7), growth (G), size (LS) and profitability (P) of the firm. We
measure tangibility as a ratio of fixed assets to total assets,
growth as a geometric mean of percentage changes in total assets,
size as the natural logarithm of total assets, profitability as net
income divided by total assets. Standard errors are reported in
parentheses, whereas ***, **, and * show significance at 1 percent,
5 percent, and 10 percent, respectively
