Mean reverting leverage policy in China: theory and evidence from industrial and sectorial level unit root analysis
| Pages | 290-306 |
| DOI | https://doi.org/10.1108/JABS-10-2016-0138 |
| Published date | 06 August 2018 |
| Date | 06 August 2018 |
| Author | Ajid ur Rehman |
| Subject Matter | Strategy,International business |
Mean reverting leverage policy in China:
theory and evidence from industrial and
sectorial level unit root analysis
Ajid ur Rehman
Abstract
Purpose –This study aimsto apply unit root test to investigate the behaviorof Chinese firms toward their
leveragepolicy. The study is based on two influentialand competing theories of capitalstructure.
Design/methodology/approach –This study applies unit root test to investigate the behavior of
Chinese firms towardtheir leverage policy. The study is based on two influentialand competing theories
of capital structure. Trade off theory advocates that firms have a target level of leverage ratio and that
firms try to achieve that optimalleverage ratio, whereas pecking order theory argues thatfirms have no
target level of leverage and that they follow a specific pattern of leverage. For this purpose, this study
appliesa Fisher typeunit root test to 12,808 firm level observations.The data are unbalanced and cover a
period from1991 to 2014.
Findings –The results reveal the presence of a stationary behavior across short-term, long-term and
total leverage policies.For short-term leverage policy, 21 per cent firms show stationarybehavior, while
for long-term,20 per cent show a targeting behavior; for the totalleverage policy 17 per cent of firms are
found to follow a tradeoffmodel. To make the findings more interesting sample was furtherclassified into
profit and lossmaking firms. The study finds that loss makingfirms do not follow a target level of leverage
in China. Furthermore, unit root is applied to all firms before and after crises-2008. It is revealed that
stationarybehavior is more prevalent before crises-2008.
Originality/value –This study is highlyimportant from the point of view that it quantifiesfirms into distinct
categories of following specific model of capital structure. To the best of the author’s knowledge, the
findings of this studyadd to current research knowledge about Chinese firmswith respect to adjustment
behaviortoward a target capital structure.
Keywords Chinese firms, Unit root testing, Panel data, Capital structure
Paper type Research paper
1. Introduction
After Nelson and Plosser’s (1982) findings related to various macroeconomic time series
analysis appeared, many researchers devoted themselves to explore the univariate
characteristics of economic phenomena. The findings of these studies greatly facilitated
researchers and provided useful ways of analyzing available data. This made researchers
to avoid producing spurious regressions (Granger and Newbold, 1974). The advantage of
unit root analysis lies in its ability to explainspecific characteristics of data in light of specific
theories (King et al., 1991). This study provides important evidence about the mean
reverting behavior of leveragepolicy in Chinese firms.
Since the pivotal work of Modigliani and Miller (1958), significant research has been
devoted to study different capitalstructure theories and models. Literature in this regard put
forward two competing and influential theoretical models. These models are trade off and
pecking order model of capital structure. Trade off (denoted as TO from hereon) theory
Ajid ur Rehman is Assistant
Professor at the Faculty of
Management Sciences,
Riphah International
University, Islamabad,
Pakistan.
Received 17 October 2016
Revised 17 October 2016
Accepted 11 January 2017
PAGE 290 jJOURNAL OF ASIA BUSINESS STUDIES jVOL. 12 NO. 3 2018, pp. 290-306, ©Emerald Publishing Limited, ISSN 1558-7894 DOI 10.1108/JABS-10-2016-0138
suggests that a firm follows a specific leverage ratio based on its trade off of the present
value associated with future marginal benefits of debt financing and the associated costs
(financial distress costs). Moreover, TO suggests that firm used to borrow until they
gradually reach to their optimal leverage ratio, and this optimal leverage ratio maximizes
firm value (Bontempi and Golinelli, 2001). In more plausible terms, the consideration of
marginal costs by TO theory suggests a unique optimal capital structure for a firm. Until the
change in perceptions, a capital structure considered optimal currently by a firm remains
optimal for future as well. Bontempi and Golinelli (2001) argue that adjustment costs of an
optimal leverage policy are not the primary concerns of TO theory and TO seldom mention
these costs. Because of these arguments, TO theory is also termed as static TO theory.
Some of the important studies regarding the static nature of TO theory include those of
Jensen and Meckling (1976),Myers(1977),Stulz (1990) and Ross (1977).
Another competing theory put forward by Myers and Majluf (1982) is the pecking order
(denoted as PO from hereon) theory of capital structure. This theory suggests that firms
follow a pecking order in fulfilling their financing needs. Firms prefer to finance their
investment needs first through internal funds because they are less costly and have less
information asymmetry, followed by debt financing; in the last resort firms issue equity to
raise external funds. Myers and Majluf (1982) show that existing shareholders are better
informed about firm market value as compared to new investors. The new investors may
underprice firm’s stock and this may resultin underinvestment. This implies that firms would
like to follow a hierarchal approach in financing their investment needs. In this regard,
internal funds that are less information asymmetric and easily accessible are selected first,
followed by safest debt; in the last resort firms opt for the equity financing. Thus, contrary to
the TO theory, PO theory does notenvisage a target leverage ratio. Instead, in PO, leverage
policy is the result of hierarchal choices and firms internal funds lies at the top of hierarchy,
while financing through equity lies at the bottom of these hierarchal choices. Moreover, in
times when riskless debt is available, firms will never issue equity. In general, debt is less
risky because it is not influenced by inside informationand it is not under the effects of stock
prices. This suggests that PO theoryrequires an exogenous debt constraint.
As far as empirical point of view is considered, TO theoryimplies that firms follow a specific
leverage policy, and thus the actual leverage values fluctuate around that specific value.
This implies that leverage is not determined and estimated by its determinants; instead,it is
measured by using the mean value of historical debt ratios (Taggart, 1977;Marsh, 1982;
Shyam-Sunder and Myers, 1999). The usage of historical means in TO model reduces the
chances of transient variation influence of seasonal business cycles, costs of floatation and
firms’ adjustment (lagged)toward target level of leverage. This implies that in the long run,it
is a firm’s average leverage that matters and not the periodic determinants that influence
firms’ adjustment toward a target debt ratio. This argument also becomes significant in
deciding whether the deviation of firms from a target debt ratio and various cost and
benefits associated with leverage has economic significance to have an effect on the
borrowing decisions of a firm. On the other hand, PO theory manifests in the form of the
relationship between debt valuevariations and available retained earnings.
All of the above arguments about TO and PO imply that unit root analysismay provide solid
empirical estimation of these two theories. In statistical terms, it can be used to find out
whether leverage ratios of firms are stationary or they exhibit unit root. If a firm shows
stationary behavior, it implies that the firm is showing a mean reverting behavior, and it
adjusts to a target and constant leverage ratio. This shows that random shocks in case of
stationary behavior only have a transitory effect on the leverage ratio. Thus, throughunit root
analysis, TO can be estimated, and using historical means for analysis does not bias
empirical results. On the other hand, the presence of unit root supports PO, implying that
leverage policy of the firm does not follow a specifictarget and is instead influenced by the
VOL. 12 NO. 3 2018 jJOURNAL OF ASIA BUSINESS STUDIES jPAGE 291
To continue reading
Request your trial