Applied Economics

ISSN: 0003-6846 (Print) 1466-4283 (Online) Journal homepage: http://www.tandfonline.com/loi/raec20

Mean reverting financial leverage: theory and

evidence from Pakistan

Tanveer Ahsan, Wang Man & Muhammad Azeem Qureshi

To cite this article: Tanveer Ahsan, Wang Man & Muhammad Azeem Qureshi (2015): Mean
reverting financial leverage: theory and evidence from Pakistan, Applied Economics, DOI:
10.1080/00036846.2015.1080802

To link to this article: http://dx.doi.org/10.1080/00036846.2015.1080802

Published online: 11 Sep 2015.

Submit your article to this journal

Article views: 21

View related articles

View Crossmark data

Full Terms & Conditions of access and use can be found at

http://www.tandfonline.com/action/journalInformation?journalCode=raec20

Download by: [University of Sussex Library] Date: 13 September 2015, At: 01:05
 APPLIED ECONOMICS, 2015
http://dx.doi.org/10.1080/00036846.2015.1080802

Mean reverting financial leverage: theory and evidence from Pakistan

Tanveer Ahsana, Wang Mana and Muhammad Azeem Qureshi b

a
School of Accounting, Dongbei University of Finance and Economics, Dalian, P.R. China; bOslo Business School, Oslo and Akershus
University, College of Applied Sciences, Oslo, Norway

ABSTRACT KEYWORDS
Grounding concepts of the two competing theories of capital structure (trade-off theory, Target capital structure;
pecking order theory) are quite opposite to each other. Trade-off theory claims that there is Pakistan; unit root analysis;
an optimal (target) capital structure and firms try to achieve that optimal (target) point. panel data
Whereas pecking order theory argues that there is no optimal (target) capital structure but JEL CLASSIFICATION
the firms follow a specific pattern of financing. Using the two competing theoretic frame- C23; G30; G32
works, this study applies Fisher-type panel unit root test to an unbalanced panel data of
Downloaded by [University of Sussex Library] at 01:05 13 September 2015

13 115 firm-year observations of nonfinancial firms listed on Karachi Stock Exchange Pakistan
spread over 38 years (1973–2010). Overall panel test results, for short-term, long-term, as well
as total leverage support trade-off financing behaviour while individual firm results do not.
Individual firm results show that only 16% of the firms have short-term target, 25% of the
firms have long-term target and 12% of the firms have total target leverage ratio. Further,
industry results explain that most of the industries do have target leverage ratios and
classification of data into profitable and lossmaking firm-year observations explains that
profitable firms clearly follow trade-off financing behaviour while the results for lossmaking
firms do not support trade-off financing behaviour. Our study indicates that it is important for
the government to ensure policies to develop well-balanced financial markets and to improve
accountability systems.

I. Introduction a specific pattern of financing. This also explains that

capital structure of firms mainly depends upon their
Since the seminal work of MM (Modigliani and
investments needs and past profitability.
Miller 1958), a number of capital structure theories
The empirical studies (Fama and French 2002;
have evolved of which trade-off theory (TOT) and
Flannery and Rangan 2006; Huang and Ritter 2009)
pecking order theory (POT) are the two most influ-
and surveys (Graham and Harvey 2001; Brounen, De
ential and competing theoretical frameworks. These
Jong, and Koedijk 2006) provide robust evidence of
theories also contribute to determine influential fac-
dynamic rebalancing in pursuit of their target capital
tors of capital structure. The TOT argues that opti-
structure. They conclude that corporate leverage
mal corporate capital structure is an outcome of
ratios are mean-reverting providing support to
trade-off between costs and benefits associated with
TOT. In Pakistan, initial empirical work used 5
leverage (Modigliani and Miller 1963). This optimal
years’ data and found relationships in favour of
capital structure works as static target for the firms.
POT (Shah and Hijazi 2004). Another comprehen-
Empirical studies (Jensen and Meckling 1976;
sive study used 34 years’ data of listed manufactur-
Myers 1977; Stulz 1990; Ross 1977) also support
ing firms and found the results generally in favour of
this concept of static TOT. On the other hand,
POT (Qureshi 2009). Another significant work used
POT advocates pecking order of firm’s preference
5 years’ data of listed nonfinancial firms and
to finance its investments; first by internal funds
obtained results in favour of both POT and TOT
being less costly, second by external debt, and equity
(Sheikh and Wang 2011). Other studies also
as last resort (Donaldson 1961; Myers and Majluf
explained mixed results in favour of both the the-
1984). This explains that firms do not have strong
ories (Sheikh and Wang 2010; Qureshi, Imdadullah,
reasoning to follow a target capital structure, instead
and Ahsan 2012; Sheikh and Qureshi 2014; Ahsan,

CONTACT: Muhammad Azeem Qureshi Muhammad-Azeem.Qureshi@hioa.no

© 2015 Taylor & Francis
 2 T. AHSAN ET AL.

Wang, and Qureshi 2015). All of these studies use
regression analysis (multiple regression, panel data
regression) and none has used unit root test which is
a useful way not only to typify the available data and
test specific theory but also prevents researchers
from producing spurious relationships (Granger
and Newbold 1974). To overcome this common
limitation observed in earlier studies carried out in
Pakistan, we propose to apply panel data unit root
test to a unique large unbalanced panel data set of
Pakistani listed nonfinancial firms to check if their
capital structure is mean-reverting.
The financial as well as judicial institutions
are weak in Pakistan (www.govindicators.org).
Downloaded by [University of Sussex Library] at 01:05 13 September 2015
used. Leverage of firms reverts towards its mean
(target) value if the data are stationary, and hence
the firms follow TOT. However, nonstationary data
suggest that leverage of firms does not revert towards
its mean (target) value and the firms may follow POT.
We use this concept to gather empirical evidence
about financial behaviour of Pakistani listed nonfi-
nancial firms. We choose only nonfinancial firms
for two reasons. First, financial firms in Pakistan are
regulated and their capital structure decision may be
affected by the regulations. Second, financial firms in
Pakistan are subject to different tax rate as compared
to the tax rates applicable to the nonfinancial firms.
The TOT builds its arguments on the tax rate and

Operating with three stock exchanges and Karachi consequent debt tax shield. As such, excluding finan-
Stock Exchange (KSE) as the biggest stock exchange, cial firms from our data set was necessary considering
Pakistan currently has a listed capital of around US$ the objective of the study. Moreover, we classify the
12 billion, market capitalization of around US$ 74 data into profitable and lossmaking firm-year obser-
billion, whereas the listed debt capital is only US$ vations as according to POT internal funds are most
150 million (www.kse.com.pk). These figures clearly preferable funds and profitability is a main source of
suggest that Pakistan is a bank-based economy and internal funds.
capital market of Pakistan is underdeveloped. That is Instead of sample selection, we use an extended
why raising bank debt is easy for firms here as com- data set of all Pakistani nonfinancial firms listed on
pared to raising equity. Furthermore, higher total KSE taken from the State Bank of Pakistan (SBP)
leverage ratios with the mean of 0.8028 (Table 2) publications for the period 1973 to 2010 (SBP
also confirm it. Moreover, higher average short-term 1973–2010) to examine whether Pakistani firms fol-
leverage ratio (mean = 0.5960) as compared to aver- low target capital structure or not. Our data set
age long-term leverage ratio (mean = 0.2068) suggests comprises of 13 industries and 670 listed firms.
that short-term debt financing is easier to avail as Consequently, it produces 13 115 firm-year observa-
compared to long-term debt financing (Table 2). tions (after excluding potential outliers) over the
Apart from the introduction presented in Section study period. In Table 1, we summarize the data
I, we organize rest of the article in the following used in this study.
manner: Section II describes theoretical framework
and the data as well as introduces the methodology;
Section III presents the empirical results and their Table 1. Number of firms and observations in each industry.
Cumulative
discussion; and finally, Section IV is about the con- Number Number of number of
clusions drawn, policy implications and suggested S. no. Industry of firms observations observations
directions for future research. We provide the refer- 1 Cement 25 482 482
2 Chemical 56 1037 1519
ences at the end. 3 Engineering 67 1290 2809
4 Fuel and energy 34 571 3380
5 Jute 09 210 3590
6 Miscellaneous 88 1666 5256
II. Theoretical framework, data and 7 Paper and board 17 389 5645
methodology 8 Sugar 42 1074 6719
9 Textile 255 4940 11 659
10 Ancillary textile 37 690 12 349
Empirical studies (Frank and Goyal 2009; Jõeveer products
2013) use mean leverage ratio as a proxy of target 11 Tobacco 07 178 12 527
12 Transport and 14 193 12 720
leverage for firms and stationarity property of the communications
data necessitates reversion towards its mean value. 13 Vanaspati 19 395 13 115
Notes: We create 100 percentiles of the data set and drop 1st and 100th
This property can be used to identify if there is target percentile as being extreme. The table presents the firm-year observa-
leverage. For this purpose, unit root analysis can be tions after omitting 1st and 100th percentiles.
 APPLIED ECONOMICS 3

The growing use of panel data in empirical studies
has forced econometrics to extend time series unit root
tests to panel data. Consequently, there are a number
of unit root tests available for panel data such as LLC
(Levin, Lin, and Chu 2002), IPS (Im, Pesaran, and Shin
2003) and Fisher type (Choi 2001; Maddala and Wu
1999) tests. These panel data unit root tests are differ-
ent from each other in their technique and assump-
tions. Considering its advantages (Maddala and Wu
1999; Baltagi 2005), we choose Fisher test. Moreover, it
allows time dimension to be different for each i and as
such it does not require a balanced panel (Choi 2001)
and considered a preferred choice for our study that
uses unbalanced panel of 13 115 firm-year observa-
Downloaded by [University of Sussex Library] at 01:05 13 September 2015
III. Empirical results and their discussion
In this section, first we present descriptive statistics
of leverage ratios used in the study. Then we present
and discuss overall, industry-level and firm-level
empirical results of unit root analysis for three dif-
ferent measurements of leverage ratio.
Descriptive statistics
Table 2 presents three leverage measures along with
their descriptive statistics for overall data set and
separately for profitable and lossmaking firms.
Overall mean value of 0.8028 for total leverage
ratio explains that Pakistani listed nonfinancial

tions. In the following equation, we present our panel
data unit root (Fisher test) model:
X
N
P ¼ 2 lnρi
i¼1
where N is cross-sectional (firm-year) dimension
and ρi is the combination of p-values of unit root
tests carried out for each cross section i. Furthermore,
−2 ln ρi follows χ2 distribution with 2 degrees of
freedom.
This model contains the following hypothesis:
HO : j ρi j ¼ 1 for all i’s (all panels are
nonstationary)
HA : j ρi j < 1 for some i’s (some panels are non-
stationary while others are not)
We use Eviews® 7.2 to apply Fisher-type panel unit
root test on different leverage ratios used in this
study. We measure leverage using three proxies:
short-term leverage as the ratio of short-term debt
to total assets; long-term leverage as the ratio of
long-term debt to total assets; and total leverage as
the ratio of total debt to total assets.
firms are highly leveraged. Further, overall mean
value of 0.5960 for short-term leverage ratio as com-
pared to the mean value of 0.2068 for long-term
leverage ratio explains that short-term financing is
the main source for these firms. Furthermore, the
mean leverage ratios of lossmaking firms (0.7999 for
short-term leverage; 0.2969 for long-term leverage
and 1.0969 for total leverage) are higher than those
of profitable firms (mean value of 0.4939 for short-
term leverage; 0.1616 for long-term leverage and
0.6556 for total leverage). This indicates that in a
bank-based economy the only financing source with
lossmaking firms is the banks.
Unit root analysis
In this section, we present the results of unit root
analysis for three leverage ratios separately. We car-
ried out unit root analysis for complete data set
(13 115 firm-year observations) and then for 13
industries separately to overcome any plausible effect
of different number of firm-year observations for
different industries.

Table 2. Leverage measurements and descriptive statistics.

Leverage name Measurement method Obs. Mean SD Min Max
Overall
Short-term leverage (STLit) Short-term debt/total assets 13 115 0.5960 0.5196 0.0016 10.0464
Long-term leverage (LTLit) Long-term debt/total assets 13 115 0.2068 0.3448 0.0000 8.6478
Total leverage (TLit) Total debt/total assets 13 115 0.8028 0.6243 0.0180 10.8182
Profitable firms
Short-term leverage (STLit) Short-term debt/total assets 8741 0.4939 0.3122 0.0020 9.1335
Long-term leverage (LTLit) Long-term debt/total assets 8741 0.1616 0.2613 0.0000 7.6818
Total leverage (TLit) Total debt/total assets 8741 0.6556 0.4058 0.0179 10.8182
Lossmaking firms
Short-term leverage (STLit) Short-term debt/total assets 4374 0.7999 0.7433 0.0016 10.0464
Long-term leverage (LTLit) Long-term debt/total assets 4374 0.2969 0.4559 0.0000 8.6478
Total leverage (TLit) Total debt/total assets 4374 1.0969 0.8424 0.0249 10.0609
 4 T. AHSAN ET AL.

Table 3. Overall, industry- and firm-level unit root results for short-term leverage.
Number of firms Number of observations Probability Number and
Lag length percentage of firms
based on ADF-Fisher ADF- Choi with stationary
Short-term leverage Available Included Dropped Available Included Dropped AIC χ2 z-stat behaviour.
Overall 670 650 20 13 115 11 954 1161 0–9 0.000 0.000 104/650 (16%)
Cement 25 23 02 482 423 59 0–5 0.000 0.000 5/23 (22%)
Chemical 56 54 02 1037 948 89 0–5 0.000 0.003 8/54 (15%)
Engineering 67 63 04 1290 1174 116 0–7 0.010 0.115 7/63 (11%)
Fuel and energy 34 31 03 571 519 52 0–4 0.012 0.006 4/31 (13%)
Jute 09 09 00 210 194 16 0–5 0.275 0.726 1/09 (11%)
Miscellaneous 88 83 05 1666 1523 143 0–7 0.007 0.275 8/83 (9.6%)
Paper and board 17 17 00 389 369 20 0–1 0.007 0.010 2/17 (12%)
Sugar 42 41 01 1074 980 94 0–9 0.000 0.000 6/41 (15%)
Textile 255 252 03 4940 4510 430 0–9 0.000 0.000 52/252 (21%)
Ancillary textile products 37 37 00 690 627 63 0–9 0.000 0.014 7/37 (19%)
Tobacco 07 07 00 178 169 09 0–2 0.193 0.226 1/07 (14%)
Transport and 14 14 00 193 176 17 0–1 0.021 0.064 2/14 (14%)
communications
Vanaspati 19 19 00 395 342 53 0–6 0.484 1.000 1/19 (5.3%)
Downloaded by [University of Sussex Library] at 01:05 13 September 2015

Short-term leverage For these industries, χ2 probability rejects null

hypothesis but z-stat probability accepts it. For this
In Table 3, we present the results of unit root ana-
reason, we further classify our data into profitable
lysis for short-term leverage ratio. The results of
and lossmaking firm-year observations and run the
overall analysis find no unit root in short-term lever-
same analysis again.
age ratio and reject the null hypothesis. It shows
Table 4 presents the results of unit root analysis
stationary short-term leverage ratio and trade-off
of lossmaking firms for short-term leverage ratio.
short-term financing behaviour of Pakistani listed
According to the results, there is not even a single
nonfinancial firms. But individual firm results show
industry for which both the probabilities (χ2,
that only 16% of the overall firms follow trade-off
z-stat) simultaneously reject the null hypothesis
financing behaviour.
at 1% significance level. Although for the paper
At industry level, unit root analysis of short-term
and board industry both the probabilities (χ2,
leverage ratio explains mixed results. For cement,
z-stat) simultaneously reject the null hypothesis
chemical, fuel and energy, paper and board, sugar,
at 5% significance level. This explains that the
textile and ancillary textile products, the analysis
firms in the paper and board industry try to follow
rejects the null hypothesis significantly and shows
their short-term leverage targets even in losses.
that short-term leverage ratio of these industries is
For the remaining industries, results explain that
stationary. This finding suggests that the firms in
lossmaking firms do not follow any target leverage
these industries have a target short-term leverage
ratio for short-term financing irrespective of
ratio and follow TOT to make their short-term
industry type they belong to.
leverage decisions. But individual firm results for
Table 5 presents the results of unit root analysis of
these industries explain that only a small percentage
profitable firms for short-term leverage ratio.
of firms follow short-term trade-off financing beha-
According to the results, the probabilities are signif-
viour. Alternatively, the analysis accepts the null
icant enough to reject the null hypothesis for almost
hypothesis significantly for jute, tobacco and vanas-
all the industries except for jute, tobacco and vanas-
pati. This finding suggests that short-term leverage
pati. Results explain that profitable firms do have
ratio of these industries is nonstationary and as such
short-term target leverage ratio except for the firms
the firms belonging to these three industries do not
in the jute, tobacco and vanaspati industry.
have any target for their short-term leverage ratio
and do not follow TOT. Individual firm results for
these industries also support that these industries do
Long-term leverage
not have short-term target leverage ratios. The
results for engineering, transport and communica- We present the results of unit root analysis for long-
tions, and miscellaneous industries are ambiguous. term leverage ratio in Table 6. The results of overall
 APPLIED ECONOMICS 5

Table 4. Overall and industry-level unit root results of lossmaking firms for short-term leverage.
Number of firms Number of observations Lag length Probability
based on ADF-Fisher ADF- Choi
Short-term leverage Available Included Dropped Available Included Dropped AIC χ2 z-stat
Overall 573 415 158 4374 3550 824 0–3 0.000 NA
Cement 22 16 06 159 127 32 0–2 0.153 0.083
Chemical 38 25 13 206 154 52 0–1 0.102 0.529
Engineering 53 36 17 371 292 79 0–3 0.444 0.592
Fuel and energy 19 04 15 62 34 28 0–1 0.599 0.984
Jute 09 07 02 66 55 11 0–1 0.032 0.074
Miscellaneous 73 49 24 502 392 110 0–3 0.000 0.206
Paper and board 15 12 03 121 103 18 0–1 0.000 0.017
Sugar 41 34 07 330 274 56 0–1 0.028 0.222
Textile 238 187 51 2034 1692 342 0–3 0.001 NA
Ancillary textile products 30 20 10 240 196 44 0–2 0.764 0.977
Tobacco 05 03 02 31 23 08 0–0 0.225 0.364
Transport and communications 11 06 05 63 47 16 0–0 0.693 0.751
Vanaspati 19 16 03 189 161 28 0–3 0.615 0.999
Downloaded by [University of Sussex Library] at 01:05 13 September 2015

Table 5. Overall and industry-level unit root results of profitable firms for short-term leverage.
Number of firms Number of observations Lag length Probability
based on ADF-Fisher ADF- Choi
Short-term leverage Available Included Dropped Available Included Dropped AIC χ2 z-stat
Overall 638 533 105 8741 7761 980 0–9 0.000 0.000
Cement 24 21 03 323 279 44 0–5 0.001 0.001
Chemical 51 50 01 831 760 71 0–5 0.000 0.000
Engineering 66 51 15 919 807 112 0–7 0.000 0.000
Fuel and energy 34 31 03 509 458 51 0–4 0.001 0.000
Jute 09 08 01 144 129 15 0–1 0.084 0.419
Miscellaneous 85 68 17 1164 1041 123 0–4 0.004 0.022
Paper and board 14 13 01 268 242 26 0–4 0.000 0.000
Sugar 42 39 03 744 677 67 0–7 0.000 0.000
Textile 239 191 48 2906 2543 363 0–9 0.000 0.000
Ancillary textile products 36 30 06 450 400 50 0–5 0.000 0.000
Tobacco 07 06 01 147 136 11 0–2 0.176 0.163
Transport and communications 12 10 02 130 113 17 0–1 0.019 0.008
Vanaspati 19 15 04 206 176 30 0–4 0.080 0.192

Table 6. Overall industry and firm-level unit root results for long-term leverage.
Number of firms Number of observations Probability Number and
Lag length percentage of firms
based on ADF-Fisher ADF- Choi with stationary
Long-term leverage Available Included Dropped Available Included Dropped AIC χ2 z-stat behaviour.
Overall 670 642 28 13 115 11 828 1287 0–8 0.000 0.000 161/642 (25%)
Cement 25 23 02 482 427 55 0–6 0.000 0.000 7/23 (30%)
Chemical 56 54 02 1037 946 91 0–7 0.000 0.000 16/54 (30%)
Engineering 67 61 06 1290 1166 124 0–4 0.000 0.000 15/61 (25%)
Fuel and energy 34 31 03 571 510 61 0–8 0.000 0.003 6/31 (19%)
Jute 09 09 00 210 199 11 0–1 0.005 0.001 1/09 (11%)
Miscellaneous 88 81 07 1666 1462 204 0–8 0.000 0.000 20/81 (25%)
Paper and board 17 17 00 389 351 38 0–5 0.000 0.000 6/17 (35%)
Sugar 42 41 01 1074 1001 73 0–6 0.000 0.000 12/41 (29%)
Textile 255 250 05 4940 4495 445 0–7 0.000 0.000 61/250 (24%)
Ancillary textile products 37 35 02 690 601 89 0–4 0.000 0.000 10/35 (29%)
Tobacco 07 07 00 178 159 19 0–8 0.005 0.004 1/07 (14%)
Transport and 14 14 00 193 177 16 0–1 0.003 0.033 3/14 (21%)
communications
Vanaspati 19 19 00 395 334 61 0–7 0.002 0.861 3/19 (16%)
Vanaspati (in loss) 19 13 06 189 128 61 0–1 0.318 0.959
Vanaspati (profitable) 19 15 04 206 183 23 0–1 0.000 0.000

analysis find no unit root in long-term leverage ratio behaviour of Pakistani listed nonfinancial firms. But
and reject the null hypothesis. It shows stationary long- individual firm results show that only 25% of the over-
term leverage ratio and trade-off long-term financing all firms follow trade-off financing behaviour.
 6 T. AHSAN ET AL.

Table 7. Overall, industry- and firm-level unit root results for total leverage.
Number of firms Number of observations Probability Number and
Lag length percentage of firms
based on ADF-Fisher ADF- Choi with stationary
Total leverage Available Included Dropped Available Included Dropped AIC χ2 z-stat behaviour.
Overall 670 650 20 13 115 11 909 1206 0–9 0.000 0.000 76/650 (12%)
Cement 25 23 02 482 425 57 0–8 0.005 0.015 4/23 (17%)
Chemical 56 54 02 1037 948 89 0–5 0.000 0.022 8/54 (15%)
Engineering 67 63 04 1290 1175 115 0–7 0.013 0.134 7/63 (11%)
Fuel and energy 34 31 03 571 521 50 0–3 0.003 0.043 5/31 (16%)
Jute 09 09 00 210 199 11 0–1 0.313 0.538 0/09 (00%)
Miscellaneous 88 83 05 1666 1498 168 0–8 0.003 0.377 7/83 (8.4%)
Paper and board 17 17 00 389 361 28 0–3 0.403 0.820 2/17 (12%)
Sugar 42 41 01 1074 968 106 0–9 0.001 0.001 5/41 (12%)
Textile 255 252 03 4940 4512 428 0–8 0.000 0.000 33/252 (13%)
Ancillary textile products 37 37 00 690 620 70 0–6 0.000 0.283 3/37 (8.1%)
Tobacco 07 07 00 178 161 17 0–8 0.466 0.522 1/07 (14%)
Transport and 14 14 00 193 172 21 0–3 0.494 0.842 0/14 (00%)
communications
Vanaspati 19 19 00 395 349 46 0–6 0.981 1.000 1/19 (5.3%)
Downloaded by [University of Sussex Library] at 01:05 13 September 2015

Further, industry-level unit root analysis of Total leverage

long-term leverage ratio explains results in favour
of trade-off model also. The analysis rejects the
null hypothesis for all the industries significantly
and shows that long-term leverage ratios of all
these industries are stationary except for vanaspati.
Accordingly, we can claim that almost all
industries in Pakistan have a target long-term
leverage ratio and follow trade-off behaviour to
make their long-term financing decisions, except
for the firms in the vanaspati industry. Individual
firm results also explain that the percentage of
firms having long-term target leverage ratio is
higher as compared to the percentage of firms
having short-term target leverage ratio for overall
as well as industry-level analysis. For the vanaspati
industry, χ2 probability rejects null hypothesis but
z-stat probability accepts it, suggesting that firms
in the vanaspati industry may have target long-
term leverage. Therefore, we further classify the
vanaspati industry into profitable and lossmaking
firm-year observations and run the analysis again.
Now for lossmaking firms, both the probabilities
(χ2, z-stat) simultaneously accept the null hypoth-
esis and suggest that lossmaking firms of the
vanaspati industry do not have long-term target
leverage ratio. While for profitable firms of the
vanaspati industry both the probabilities (χ2,
z-stat) simultaneously reject the null hypothesis
at 1% significance level explaining that profitable
firms of this industry do have long-term target
leverage ratio.
We present the results of unit root analysis for total
leverage ratio in Table 7. The results of overall ana-
lysis find no unit root in total leverage ratio and
reject the null hypothesis. It shows stationary total
leverage ratio and trade-off total financing behaviour
of Pakistani listed nonfinancial firms. But individual
firm results show that only 12% of the total number
of firms follows trade-off financing behaviour.
Alternatively, industry-level unit root analysis of
total leverage ratio explains mixed results almost
similar to short-term leverage ratio. The reason is
that the short-term financing covers the bigger part
of the total financing of Pakistani listed nonfinancial
firms (mean of 0.5960 for short-term leverage in
Table 2). Our analysis rejects the null hypothesis
significantly for the cement, chemical, fuel and
energy, sugar and textile industries and shows that
the total leverage ratio of these industries is station-
ary. But again individual firm results for these indus-
tries show only a small number of firms having
trade-off financing behaviour for total leverage. For
jute, paper and board, tobacco, transport and com-
munications and vanaspati, the analysis accepts the
null hypothesis significantly and shows that the total
leverage ratio of these industries is nonstationary.
Following this, we observe a trade-off financing
behaviour in the cement, chemical, fuel and energy,
sugar and textile industries. On the other hand, there
is no total target leverage ratio in the jute, paper and
board, tobacco, transport and communications and
vanaspati industries. Accordingly, we can claim that
 APPLIED ECONOMICS 7

these industries do not demonstrate trade-off beha-
viour for total financing. Individual firm results for
these industries also support that these industries do
not have total target leverage ratios. We obtain
ambiguous results for the engineering, ancillary tex-
tile products and miscellaneous industries: χ2 prob-
ability rejects null hypothesis but z-stat probability
accepts it. To clarify this ambiguity, we further clas-
sify our data into profitable and lossmaking firm-
year observations and run the same analysis again.
Table 8 presents the results of unit root analysis of
lossmaking firms for total leverage ratio. According
to the results, both the probabilities (χ2, z-stat) sig-
nificantly accept the null hypothesis and explain that
Downloaded by [University of Sussex Library] at 01:05 13 September 2015
the null hypothesis significantly for industries such
as cement, chemical, engineering, miscellaneous,
sugar, textile and ancillary textile products and
show that the total leverage ratio of profitable firms
of these industries is stationary. This means profit-
able firms of these industries follow trade off-theory
and do have total target leverage ratio. For fuel and
energy, jute, paper and board, tobacco, transport and
communications and vanaspati, the analysis accepts
the null hypothesis significantly and shows that the
total leverage ratio of profitable firms of these indus-
tries is nonstationary. Accordingly, we observe a
trade-off financing behaviour in profitable firms of
the cement, chemical, engineering, miscellaneous,

lossmaking firms do not follow any target leverage
ratio for total financing irrespective of the industry
type they belong to.
Table 9 presents the results of unit root analysis of
profitable firms for total leverage ratio. According to
the results, both the probabilities (χ2, z-stat) reject
sugar, textile and ancillary textile products indus-
tries. On the other hand, there is no target total
leverage ratio even in profitable firms of the fuel
and energy, jute, paper and board, tobacco, transport
and communications and vanaspati industries.
Therefore, we can claim that even profitable firms

Table 8. Overall and industry-level unit root results of lossmaking firms for total leverage.
Number of firms Number of observations Lag length Probability
based on ADF-Fisher ADF- Choi
Total Leverage Available Included Dropped Available Included Dropped AIC χ2 z-stat
Overall 573 415 158 4374 3547 827 0–4 0.013 NA
Cement 22 16 06 159 129 30 0–2 0.516 0.607
Chemical 38 25 13 206 155 51 0–1 0.217 0.601
Engineering 53 36 17 371 295 76 0–2 0.059 0.643
Fuel and energy 19 04 15 62 35 27 0–0 0.887 0.930
Jute 09 07 02 66 55 11 0–1 0.561 0.529
Miscellaneous 73 49 24 502 394 108 0–3 0.069 0.666
Paper and board 15 12 03 121 101 20 0–2 0.018 0.490
Sugar 41 34 07 330 275 55 0–1 0.313 0.886
Textile 238 187 51 2034 1685 349 0–4 0.177 NA
Ancillary textile products 30 20 10 240 192 48 0–2 0.109 0.789
Tobacco 05 03 02 31 23 08 0–0 0.331 0.439
Transport and communications 11 06 05 63 45 18 0–1 0.972 0.997
Vanaspati 19 16 03 189 163 26 0–3 0.488 1.000

Table 9. Overall and industry-level unit root results of profitable firms for total leverage.
Number of firms Number of observations Lag length Probability
based on ADF-Fisher ADF- Choi
Total Leverage Available Included Dropped Available Included Dropped AIC χ2 z-stat
Overall 638 533 105 8741 7769 972 0–6 0.000 0.000
Cement 24 21 03 323 279 44 0–3 0.000 0.000
Chemical 51 50 01 831 750 81 0–5 0.000 0.003
Engineering 66 51 15 919 818 101 0–3 0.000 0.000
Fuel and energy 34 31 03 509 461 48 0–3 0.055 0.077
Jute 09 08 01 144 129 15 0–2 0.327 0.509
Miscellaneous 85 68 17 1164 1020 144 0–6 0.029 0.021
Paper and board 14 13 01 268 247 21 0–4 0.352 0.303
Sugar 42 39 03 744 686 58 0–3 0.007 0.003
Textile 239 191 48 2906 2557 349 0–6 0.000 0.000
Ancillary textile products 36 30 06 450 395 55 0–4 0.000 0.001
Tobacco 07 06 01 147 134 13 0–2 0.291 0.269
Transport and communications 12 10 02 130 112 18 0–1 0.065 0.032
Vanaspati 19 15 04 206 181 25 0–1 0.837 0.902
 8 T. AHSAN ET AL.
of these industries do not demonstrate trade-off
behaviour for total financing.
IV. Conclusions and policy implications
We apply panel unit root test developed by Fisher on
a unique and large data set comprising of 13 115
firm-year observations of Pakistani nonfinancial
firms listed on KSE spread over 38 years (1973–
2010) to provide empirical evidence on their lever-
age behaviour. We use three leverage proxies to
investigate corporate leverage behaviour at three
levels: overall, 13 different industries and at firm
level.
Downloaded by [University of Sussex Library] at 01:05 13 September 2015
Overall analyses for all three leverage ratios depict
stationary behaviour, suggesting that Pakistani listed
nonfinancial firms generally have target leverage
ratios that are mean-reverting in line with the
TOT. Therefore, these firms, in general, carry out
long-run capital structure planning and try to stick
to it. Further, industry-level analysis explains differ-
ent results for three different ratios of leverage. For
short-term leverage ratio, seven industries have
mean-reverting leverage and as such these seven
industries demonstrate trade-off leverage behaviour.
On the other hand, three industries do not have
mean-reverting short-term leverage ratio, suggesting
that these three industries do not depict trade-off
leverage behaviour. For the remaining three indus-
tries, the results are mixed. Individual firm results
show that only 16% of the firms follow trade-off
short-term financing behaviour. Moreover, the clas-
sification into profitable and lossmaking firm-year
observations explains that lossmaking firms do not
have any target for short-term leverage ratio irre-
spective of the industry type. Whereas profitable
firms do have short-term target leverage ratio, except
for the firms in the jute, tobacco and vanaspati
industries. For long-term leverage ratio, all the
industries except for vanaspati have mean-reverting
long-term leverage ratio, suggesting a trade-off
financing pattern. Classification of vanaspati indus-
try into profitable and lossmaking firm-year obser-
vations shows that lossmaking firms of vanaspati
industry do not have any long-term target leverage
ratio while profitable firms of this industry do have
long-term target leverage ratio. Individual firm
results show that 25% of the firms follow trade-off
long-term financing behaviour. For total leverage
ratio, five industries appear to have a mean-reverting
leverage ratio, suggesting trade-off leverage beha-
viour of these five industries. Whereas five other
industries do not appear to have mean-reverting
total leverage ratio, suggesting that these industries
may not observe trade-off leverage behaviour to
make their total financing decisions. The remaining
three industries have ambiguous results.
Classification into profitable and lossmaking firm-
year observations clarifies this ambiguity and
explains that lossmaking firms do not have any
total target leverage ratio irrespective of industry
type while for profitable firms seven industries
(cement, chemical, engineering, miscellaneous,
sugar, textile, ancillary textile products) do have
target total leverage ratio and six industries (fuel
and energy, jute, paper and board, tobacco, transport
and communications, vanaspati) do not have any
target total leverage ratio. Individual firm results
show that only 12% of the firms follow trade-off
total financing behaviour.
The results of the study suggest that firms of most
of the industries do have target short-term and total
leverage ratios and follow TOT. For long-term lever-
age ratio, the results explain that all the firms have
long-term target leverage ratio, except for those
belonging to the vanaspati industry. These findings
suggest that most of the industries demonstrate
trade-off leverage behaviour. They have target lever-
age and try to achieve that target over time.
Moreover, classification of firms into profitable and
lossmaking firm-year observations explains the
results more clearly. The lossmaking firms do not
have any target leverage ratio and they may follow
POT. Further, the leverage ratios of lossmaking
firms are dangerously high such as the mean value
(1.0969) of total leverage ratio is even greater than 1.
The peculiar nature of credit market in Pakistan
provides plausible explanation for this anomaly.
The debt market in Pakistan is underdeveloped. As
such the banking sector is the only option to provide
debt to the corporate sector, especially lossmaking
firms. It is pertinent to note that majority of the
banks are either owned by the government or own-
ers of the corporate sector otherwise a borrower.
Moreover, the majority of the Pakistani firms are
closely held family-owned firms. As such we argue
that the weak financial and judicial institutions and
environment of corporate arena in Pakistan, coupled

with deep-rooted family-based politics, help develop
and sustain the collusive networks of debtor and
creditor. Such a nexus, along with poor mechanisms
of corporate governance and the regulatory bodies
mired by political interferences and nexuses of self-
serving myopic gains, helps provide them easy access
to debt which may become nonperforming loan
before being written off using their nexuses.1
According to a previous study, politically connected
firms borrow 45% more and have 50% higher default
rates (Khwaja and Mian 2005). Under these circum-
stances, it may not only be difficult for these firms
but also for the banks to survive, having serious
implications for the development of the capital mar-
Downloaded by [University of Sussex Library] at 01:05 13 September 2015
ket in Pakistan. Government and multilateral agen-
cies should make some policies to help strengthen
judicial, financial and regulatory institutions in
Pakistan. Consequently, improved governance
mechanism shall help develop banks, bond as well
as stock markets in Pakistan not only to provide
more financing options to the firms operating in
Pakistan but also to create a healthy, competitive
and resilient business environment. On the other
hand, profitable firms of most of the industries do
have target leverage ratios and follow trade-off
financing behaviour. These firms are expected to
strive to achieve their target leverage over time.
Further research can be carried out to find out the
speed of adjustment of these firms towards their
target leverage ratios.
Disclosure statement
No potential conflict of interest was reported by the authors.
ORCID
Muhammad Azeem Qureshi http://orcid.org/0000-0001-
5107-2192
References
Ahsan, T., M. Wang, and M. A. Qureshi. 2015. “Firm,
Industry and Country Level Determinants of Capital
Structure: Evidence from Pakistan.” Working Paper.
1
http://www.sbp.org.pk/ecodata/NPL.pdf of State Bank of Pakistan reports PKR 594 billion (approximately US$ 7 billion) as nonperforming loans of banking
sector in Pakistan. Apex court of Pakistan is currently hearing a case in this regard.
APPLIED ECONOMICS 9
Baltagi, B. H. 2005. Econometric Analysis of Panel Data. 3rd
ed. New York: John Wiley & Sons.
Brounen, D., A. de Jong, and K. Koedijk. 2006. “Capital
Structure Policies in Europe: Survey Evidence.” Journal
of Banking & Finance 30: 1409–1442. doi:10.1016/j.
jbankfin.2005.02.010.
Choi, I. 2001. “Unit Root Tests for Panel Data.” Journal of
International Money and Finance 20: 249–272.
doi:10.1016/S0261-5606(00)00048-6.
Donaldson, G. 1961. Corporate Debt Capacity: A Study of
Corporate Debt Policy and the Determination of Corporate
Debt Capacity. Boston, MA: Harvard, Graduate School of
Business Administration.
Fama, E. F., and K. R. French. 2002. “Testing Trade-Off and
Pecking Order Predictions about Dividends and Debt.”
Review of Financial Studies 15: 1–33. doi:10.1093/rfs/
15.1.1.
Flannery, M. J., and K. P. Rangan. 2006. “Partial Adjustment
toward Target Capital Structures.” Journal of Financial
Economics 79: 469–506. doi:10.1016/j.jfineco.2005.03.004.
Frank, M. Z., and V. K. Goyal. 2009. “Capital Structure
Decisions: Which Factors are Reliably Important?”
Financial Management 38: 1–37. doi:10.1111/
fima.2009.38.issue-1.
Graham, J. R., and C. R. Harvey. 2001. “The Theory and
Practice of Corporate Finance: Evidence from the Field.”
Journal of Financial Economics 60: 187–243. doi:10.1016/
S0304-405X(01)00044-7.
Granger, C. W. J., and P. Newbold. 1974. “Spurious
Regressions in Econometrics.” Journal of Econometrics 2:
111–120. doi:10.1016/0304-4076(74)90034-7.
Huang, R., and J. Ritter. 2009. “Testing Theories of Capital
Structure and Estimating the Speed of Adjustment.”
Journal of Financial and Quantitative Analysis 44: 237–
271. doi:10.1017/S0022109009090152.
Im, K., M. H. Pesaran, and Y. Shin. 2003. “Testing for Unit
Roots in Heterogeneous Panels.” Journal of Econometrics
115: 53–74. doi:10.1016/S0304-4076(03)00092-7.
Jensen, M. C., and W. H. Meckling. 1976. “Theory of the
Firm: Managerial Behavior, Agency Costs and Ownership
Structure.” Journal of Financial Economics 3: 305–360.
doi:10.1016/0304-405X(76)90026-X.
Jõeveer, K. 2013. “Firm, Country and Macroeconomic
Determinants of Capital Structure: Evidence from
Transition Economies.” Journal of Comparative
Economics 41: 294–308. doi:10.1016/j.jce.2012.05.001.
Khwaja, A. I., and A. Mian. 2005. “Do Lenders Favor
Politically Connected Firms? Rent Provision in
an Emerging Financial Market.” The Quarterly Journal
of Economics 120: 1371–1411. doi:10.1162/
003355305775097524.
Levin, A., C.-F. Lin, and C.-S. J. Chu. 2002. “Unit Root Tests
in Panel Data: Asymptotic and Finite-Sample Properties.”
 10 T. AHSAN ET AL.
Journal of Econometrics 108: 1–24. doi:10.1016/S0304-4076
(01)00098-7.
Maddala, G. S., and S. Wu. 1999. “A Comparative Study of
Unit Root Tests with Panel Data and a New Simple Test.”
Oxford Bulletin of Economics and Statistics 61: 631–652.
doi:10.1111/obes.1999.61.issue-S1.
Modigliani, F., and M. H. Miller. 1958. “The Cost of Capital,
Corporation Finance and the Theory of Investment.”
American Economic Review XLVIII: 261–280.
Modigliani, F., and M. H. Miller. 1963. “Corporate Income
Taxes and the Cost of Capital: A Correction.” American
Economic Review 53: 433–443.
Myers, S. C. 1977. “Determinants of Corporate Borrowing.”
Journal of Financial Economics 5 (2): 147–175.
doi:10.1016/0304-405X(77)90015-0.
Myers, S. C., and N. S. Majluf. 1984. “Corporate Financing
and Investment Decisions When Firms Have Information
Downloaded by [University of Sussex Library] at 01:05 13 September 2015
that Investors Do Not Have.” Journal of Financial
Economics 13 (2): 187–221. doi:10.1016/0304-405X(84)
90023-0.
Qureshi, M. A. 2009. “Does Pecking Order Theory Explain
Leverage Behaviour in Pakistan?” Applied Financial
Economics 19: 1365–1370. doi:10.1080/0960310090281
7592.
Qureshi, M. A., M. Imdadullah, and T. Ahsan. 2012. “What
Determines Leverage in Pakistan? A Panel Data Analysis.”
African Journal of Business Management 6: 978–985.
Ross, S. A. 1977. “The Determination of Financial Structure:
The Incentive-Signalling Approach.” The Bell Journal of
Economics 8: 23–40. doi:10.2307/3003485.
Shah, A., and T. Hijazi. 2004. “The Determinants of Capital
Structure of Stock Exchange-listed Non-financial Firms in
Pakistan.” The Pakistan Development Review 43: 605–618.
Sheikh, N. A., and M. A. Qureshi. 2014. “Crowding-Out or
Shying-Away: Impact of Corporate Income Tax on
Capital Structure Choice of Firms in Pakistan.” Applied
Financial Economics 24: 1249–1260. doi:10.1080/
09603107.2014.925053.
Sheikh, N. A., and Z. Wang. 2010. “Financing Behaviour of
Textile Firms in Pakistan.” International Journal of
Innovation, Management and Technology 1: 130–135.
Sheikh, N. A., and Z. Wang. 2011. “Determinants of Capital
Structure: An Empirical Study of Firms in Manufacturing
Industry of Pakistan.” Managerial Finance 37: 117–133.
doi:10.1108/03074351111103668.
Stulz, R. M. 1990. “Managerial Discretion and Optimal
Financing Policies.” Journal of Financial Economics 26:
3–27. doi:10.1016/0304-405X(90)90011-N.